.. Jun". 1
2:
1.. , A.
n. ‘ . s . J
twink...“ \ x.
a. a?“
Luau"!
..
1...? n
a .inut x
5.8..

.. n .i v;
yKunu} $.28."
to. . I; :3}:

. vii .1 .

.nnfi...
. .

I...

>n..,...l.§

3..

 

. 1.0.. iii? a. gin...» _

>3. .3.

9:557

 

 

93.
Cd
Bit/3%
<lz;t-
030;
0:27-—
-5:
-.s.>3
E

 

 

This is to certify that the
thesis entitled

FACTORS INFLUENCING HERPETOFAUNAL DIVERSITY ON
DIFFERING LAND OWNERSHIP TYPES IN A HUMAN-
DOMINATED LANDSCAPE

presented by
Tracy E. Grazia

has been accepted towards fulfillment
of the requirements for the

Master of degree in Fisheries and Wildlife

 

Science

 

/.
Majdl‘ Professdr’s Signature
8 / 7 /07

Date

MSU is an afiinnaflveaction, equal-opportunity employer

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:lClRC/Date0ue,indd-p.1

 

 

FACTORS INFLUENCING HERPETOFAUNAL DIVERSITY ON DIFFERING
LAND OWNERSHIP TYPES IN A HUMAN-DOMINATED LANDSCAPE

By

Tracy E. Grazia

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Fisheries and Wildlife

2007

FACTO
LN

no! bcm tic.
and fray?”
mphiliiin til
cycle. the} hi
herpclufiuni
in southcn .\
warships a!

Ll'lldilil'lls ml

iifiStTlilc hf."
l

detected 3
. l I‘

and Lint-“fl"

‘mli‘nanl in

 

rages, .
will; I
'lrp
I
it?" ..
Jand and
j I
.m(.' 1

M in eff.

ABSTRACT
FACTORS INFLUENCING HERPETOFAUNAL DIVERSITY ON DIFFERING
LAND OWNERSHIP TYPES IN A HUMAN-DOMINATED LANDSCAPE
By

Tracy E. Grazia

In recent decades, much attention has focused on the global decline of
herpetofauna] populations throughout the world. Although all causes of declines have
not been clearly determined, anthropogenic habitat modification including habitat loss
and fragmentation is the best-documented cause of herpetofauna] declines, particularly
amphibian declines. Because herpetofauna use multiple distinct habitats during their life
cycle, they have the potential to be affected by a variety of factors. I investigated
herpetofauna] community dynamics on differing land ownership types (public vs. private)
in southern Michigan from 2005-2006 to determine 1) if differences between land
ownerships and between sample years existed, 2) if local habitat and environmental
conditions influence herpetofauna, and 3) if landscape pattern metrics could be used to
describe herpetofauna] communities. Differences between land ownership types were not
detected. Soil associations and warmer temperatures were found to influence richness
and diversity, and percent canopy cover and distance to the nearest water body were
important in structuring herpetofauna] communities. Herpetofauna exhibited significant
relationships with surrounding landscape attributes including a positive association with
wetland and forest cover, and a negative association with agriculture. My results
demonstrate that conservation strategies will need to consider factors at multiple spatial

scales to effectively understand existing and future herpetofauna] populations.

Fur.
Rmurgcsl
intmznm
Rit‘ff. ~ rAl
mm; :hmi
lll‘l'iiimitlt‘ >l

Rtilti-tl. lhcii

md LCilL‘li} i."

““11 time. T

EthU-m. (I;

.14 ',l I A
mu \V‘diiicf

 

ACKNOWLEDGEMENTS

Funding for this project was provided by the Michigan Department of Natural
Resources (MDNR), Wildlife Division; Michigan State University; and the Safari Club
International. 1 am indebted to my graduate committee members: Kelly Millenbah, Gary
Roloff, and Jim Harding for their input throughout this project and for elevating my
critical thinking skills. To my major advisor, Dr. Kelly Millenbah, thank you for your
invaluable support, encouragement, and guidance throughout this project. To Dr. Gary
Roloff, thank you for sharing your statistical knowledge throughout this project. To Jim
Harding, thank you for sharing your invaluable herpetological knowledge throughout this
project. I would like to give a special thanks to Dr. Dana Infante for all of her help with
reviewing this lengthy document.

The data collection could not have been completed without the valuable
assistance of field technicians: David Dimitrie, Carolyn Gillen, Kile Kucher, Todd Thorn,
and Leticia Villarreal who suffered through long days, poison ivy, and wood frog after
wood frog. Thanks to everyone who helped install the drift fences and pitfall traps: Matt
Einheuser, Geoffrey Grisdale, Susan Jones, and Fred Simmons and to the MSU Fisheries
and Wildlife Club for construction of over one hundred fimnel traps.

Thanks to Lila Boyko, Mark and Debra Carlson, Tom Cooley, Earl Flegler,
Wayne Flood, Edward Gildner, Scott and Wendi Kribs, Gloria Miller, Tom and Brenda

Reich, Bumell H. Selleck, PhD. (deceased) and Alice A. Selleck, Vernon and Sue

iii

 

Stephens, Charles and Mary VanLoan, Clifford and Margaret Welsch, Lisa Williams,
Livingston Land Conservancy, Michigan Wildlife Conservancy, Southwest Michigan
Land Conservancy, Woldumar Nature Center, and all of the other participating private
land owners that allowed me access to their property. This project would not have been
possible without their support and interest in the project. Special thanks to everyone who
helped me locate private land owners willing to participate in this project. Much
appreciation is owed to those that helped me see this project come to fruition. Other
contributors to this project include: Bryan and Jordan Burroughs, Rebecca Christoffel,
Jon Deroba, Ali Felix, Dr. Dan Hayes, Dan Linden, Krissy Wildman, MSU, Department
of Fisheries and Wildlife; Peter Kurtz, MSU, Department of Geography; YuMan Lee,
MNFI; Dr. Mike Lannoo, Laura Guderyhan, Ball State University; Dr. Kevin McGari gal,
UMass; Amanda Hathaway, MUCC; MDNR, Wildlife Division; Sarah Panken, and
Jeremy Smith. Also, special thanks to Jill Cruth, Julie Traver, and Mary Witchell for all
of their help and patience.

I would like to acknowledge all of the amazing women in the Millenbah lab:
Lauren Bailey, Katie Kahl, Nikki Lamp, Nichole Rubeck-Schurtz, Trixi Smith, and Adria
VanLoan. Without out your friendship, guidance, support, and words of encouragement,
I would not have completed this project. You made the transition to graduate school
enjoyable and I could not have done it without you. Thank you to Lauren and Adam for
taking me in when I first arrived and making me feel welcome. Thank you to Katie and
Jody for putting out the welcome mat and helping me through the writing process. I
would also like to acknowledge Pamela Roy for her support and friendship. Thank you

for keeping me sane during our year of statistics. To Heather Plant, Amy Jacobs, Marion

iv

Coopd. E:
me MUS"
pgfit’tiilli Lil

l 3:
career in n;

1

metal to r
Haters pr
Gix'erxmn. R
.er Bcrcnh;
encouragcm
limit you ii-

JOLITI‘I’CV.

 

 

Cooper, Elena Sachs, Kathy Sullivan, Abi King, Trisha Crabill, thank you for supporting
me through this journey. Thank you to Scott Hereford for encouraging me to reach my
potential and to gain confidence in myself.

I am indebted to my parents, Jack and Carolyn, for encouraging me to pursue a
career in natural resources and for instilling in me an appreciation of nature. I am
grateful to my husband, Brian Schaffler for his love, patience, and support throughout my
Master’s program. To my brother, Todd, grandmother, Irene Methe, Janice and Rick
Giverson, Rob and Karen Methe, and extended family, Bud and Edna Schafiler, Pam and
Art Berenbaum, Rhonda Schafiler and John Wydra, I am forever grateful for your
encouragement and support. To my grandmother, Grace Grazia and uncle, Gene Grazia,
thank you for your love and support; I wish you could have been there for the whole

journey.

llSl OF TAB
llSl OF FlGl

CHAPTEI

INTRODL'C
Nt’t‘d for
The Impu
Nllt'lhgll‘l
The Valu.
Species C
The Imp‘i

(DALAXI
STITDY A.
UTERATI

CHAPTER
ST.

INTRODI

IETHOD~
Dnfi R.
Clflcrb I
Area TL."
Amgmp
“Mk-R.
Data Ar.

 

RESLIT: I
Smile, '
Spck'l'cg
Si—‘cht
3pm...
Sl‘fllc.
HCJ'Dc!

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix

LIST OF FIGURES ......................................................................................................... xxi

CHAPTER 1: EVALUATION OF THE HERPETOFAUNAL DIVERSITY OF
MICHIGAN’S STATE GAME AND WILDLIFE AREAS

INTRODUCTION ........................................................................................................... 1
Need for Research ....................................................................................................... l
The Importance of Herpetofauna as Biological Indicators ......................................... 3
Michigan’s State Game and Wildlife Areas ................................................................ 5
The Value of Private Land to Biodiversity Conservation ........................................... 8
Species Comparisons between Private and Public Land ........................................... 10
The Importance of Landscape Pattern Metrics on Herpetofauna .............................. 1]

GOAL AND OBJECTIVES ......................................................................................... l4

STUDY AREA .............................................................................................................. 15

LITERATURE CITED ................................................................................................. 22

CHAPTER 2: DESCRIPTION OF THE HERPETOFAUNAL COMMUNITIES IN
STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

INTRODUCTION ......................................................................................................... 29
METHODS ................................................................................................................... 32
Drift Fence Arrays ..................................................................................................... 32
Coverboards .............................................................................................................. 34
Area Time-Constrained Surveys ............................................................................... 35
Anuran Call Surveys ................................................................................................. 35
Mark-Recapture Techniques ..................................................................................... 37
Data Analysis ............................................................................................................ 40
RESULTS ..................................................................................................................... 42
Species Inventory ...................................................................................................... 42
Species Richness ....................................................................................................... 45
Species Frequency of Occurrence ............................................................................. 51
Species Abundance ................................................................................................... 56
Species Diversity ....................................................................................................... 70
Herpetofauna] Community Similarity ....................................................................... 72
DISCUSSION ............................................................................................................... 72

vi

C 0111me
(rampart:

LITERATI
APPENDI.‘
.IPPENDD
APPENDIX
APPENDIX

CHAPTER
MICHIGA.‘

IVTRODU

\IETHODS
Silt Strut}
St

Vt
“mun

 

 

 

Comparisons between Land Ownerships .................................................................. 78

Comparisons between Years ..................................................................................... 79
LITERATURE CITED ................................................................................................. 83
APPENDIX 2.1. ............................................................................................................ 90
APPENDIX 2.2. .......................................................................................................... 123
APPENDIX 2.3. .......................................................................................................... 126
APPENDIX 2.4. .......................................................................................................... 127

CHAPTER 3: COMPARISON OF THE HERPETOFAUNAL COMMUNITIES ON
MICHIGAN ’8 STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

INTRODUCTION ....................................................................................................... 128
METHODS ................................................................................................................. 1 3 1
Site Stratification Variables .................................................................................... 131
Soil Associations ......................................................................................... l3 1
Vegetation Types ........................................................................................ 132
Herpetofauna ........................................................................................................... 1 32
Drifi Fence Arrays ....................................................................................... 132
Coverboards ................................................................................................ 136
Mark-Recapture Techniques ....................................................................... 138
Vegetation Variables ............................................................................................... 140
Ground Vegetation Percent Cover .............................................................. 140

Canopy Percent Cover and Composition .................................................... 141

Coarse Woody Debris ................................................................................. 141

Weather Variables ................................................................................................... 141
Precipitation ................................................................................................ 141
Temperature ................................................................................................ 142

Spatial Feature ......................................................................................................... 142
Proximity to Water Body ............................................................................ 142
Statistical Analyses ................................................................................................. 143
Mixed Model with Repeated Measures — Considering Weather Effects 143

Mixed model — Focus on All Variables (Across Sample Periods) .............. 144
Constrained Ordination ............................................................................... 145
RESULTS ................................................................................................................... 153
Mixed Model with Repeated Measures — Considering Weather Effects ................ 157
Amphibian Species Richness ...................................................................... 157
Herpetofaunal Species Richness ................................................................. 157
Herpetofaunal Species Diversity ................................................................. 166

Mixed model — Focus on All Variables (Across Sample Periods) .......................... 166

vii

Ar
Hc
Hr

Constraim
DlSCL’SSlG
LITERATE]

CHAPTER
INTRODLT

METHODS
DIIIT Fen;
Emerita:
AIC‘CI Timi
Mark-Ru“
Landscap
Statistical

.\l

Ci

RESELTS _
Mixed M

 

 

 

 

Amphibian Species Richness ...................................................................... 166

Herpetofaunal Species Richness ................................................................. 175
Herpetofaunal Species Diversity ................................................................. 175
Constrained Ordination ........................................................................................... 175
DISCUSSION ............................................................................................................. 197
LITERATURE CITED ............................................................................................... 205

CHAPTER 4: EVALUATION OF THE EFFECTS OF LANDSCAPE PATTERN
METRICS ON HERPETOFAUNA

INTRODUCTION ....................................................................................................... 21 1
METHODS ................................................................................................................. 214
Drift Fence Arrays ................................................................................................... 214
Coverboards ............................................................................................................ 216
Area Time-Constrained Surveys ............................................................................. 217
Mark-Recapture Techniques ................................................................................... 217
Landscape Analysis ................................................................................................. 219
Statistical Analysis .................................................................................................. 223
Mixed Model ............................................................................................... 223
Constrained Ordination ............................................................................... 224
RESULTS ................................................................................................................... 230
Mixed Model — 100 m Scale ................................................................................... 232
Amphibian Species Richness ...................................................................... 232
Herpetofaunal Species Richness ................................................................. 235
Herpetofaunal Species Diversity ................................................................. 235

Mixed Model - 200 m Scale ................................................................................... 242
Amphibian Species Richness ...................................................................... 242
Herpetofaunal Species Richness ................................................................. 242
Herpetofaunal Species Diversity ................................................................. 242

Mixed Model — 1000 m Scale ................................................................................. 248
Amphibian Species Richness ...................................................................... 248
Herpetofaunal Species Richness ................................................................. 253
Herpetofaunal Species Diversity ................................................................. 253
Constrained Ordination ........................................................................................... 253
DISCUSSION ............................................................................................................. 267
LITERATURE CITED ............................................................................................... 276

viii

CHAPTER
\ITC ll 10A)

Table 1.]. \
Assc
Nita
Pcnir

lat-ls 1.2. S
(STA
CORN
Stud}
2005

STAT

i151: 1.3.. it

Praia

Tabic 1.4. S”
dii‘i‘t'i'r;

53mph

CHAPTER 3:
STATE ()3 \1

Tabltll D,
[WWDH
id'ld) I

Tahiti? I7

in an-
“\iiki-i.

 

T, r
Iibifi‘ 23' K,

   

Pm
Call
Tibial, \i

ii ind

 

LIST OF TABLES

CHAPTER 1: EVALUATION OF THE HERPETOFAUNAL DIVERSITY OF
MICHIGAN’S STATE GAME AND WILDLIFE AREAS

Table 1.1. Vegetation categories according to the Integrated Forest Monitoring,
Assessment, and Prescription (IF MAP) system (Michigan Department of
Natural Resources 2001b) used in study site selection in the southern Lower
Peninsula of Michigan during the 2005 and 2006 field seasons .......................... 17

Table 1.2. Soil associations (and codes) according to US. General Soil Map
(STATSGO) for Michigan data (Soil Survey Staff, Natural Resources
Conservation Service, United States Department of Agriculture) used in
study sites selection in the southern Lower Peninsula of Michigan during the
2005 and 2006 field seasons. Descriptions of the first three major
STATSGO components are listed ........................................................................ 18

Table 1.3. Total hectares (acres) of study sites sampled in the southern Lower
Peninsula of Michigan during the 2005 and 2006 field seasons .......................... 19

Table 1.4. Site information for 82 survey sites located on 8 different SGWA and 24
different private lands in the southern Lower Peninsula of Michigan
sampled during the 2005 and 2006 field seasons ................................................. 2]

CHAPTER 2: DESCRIPTION OF THE HERPETOFAUNAL COMMUNITIES IN
STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

Table 2.1 . Dates drift fence arrays, pitfall traps, funnel traps and coverboards were
opened in southern Lower Peninsula of Michigan on SGWA and private
lands in summer 2005 and 2006 ........................................................................... 33

Table 2.2. Five decay classes of the downed logs (US. Forest Service 2004) used
in area time-constrained surveys in the southern Lower Peninsula of
Michigan on SGWA and private lands in summer 2005 and 2006 ...................... 36

Table 2.3. Survey periods for conducting frog call surveys in the southern Lower
Peninsula of Michigan, using Michigan Department of Natural Resources
call survey protocol (Sargent 2000). .................................................................... 38

Table 2.4. Windspeed classifications used to determine if frog call surveys should

be conducted based on the Beaufort scale. Surveys were not conducted if
windspeeds > 3 on the Beaufort scale (Berry et al. 1934) ................................... 39

ix

In

BIC

Tibia“ l I I")

Table 2.5. Species of amphibians and reptiles documented to occur in Lower
Peninsula of Michigan, as well as in the study area. Species ranges were

determined from Conant and Collins (1991) and Harding (1997). *Species

were observed in the study areas in summer 2005 and 2006 based on

captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys including incidental

observations .......................................................................................................... 43

Table 2.6. Capture techniques and survey methods used to detect individual species
in the southern Lower Peninsula of Michigan during the 2005 and 2006 field
seasons .................................................................................................................. 46

Table 2.7. Overall herpetofaunal Species richness on SGWA and private lands in
the southern Lower Peninsula of Michigan in summer 2005 and 2006 based
on captures by drift fence arrays coupled with pitfall traps, fiinnel traps, and
coverboards and area-time constrained survey including incidental
observations .......................................................................................................... 48

Table 2.8. Mean (SE) herpetofaunal species richness of SGWA and private lands
by month in the southern Lower Peninsula of Michigan in summer 2005 and
2006 based on captures by drift fence arrays coupled with pitfall traps,
fimnel traps, and coverboards and area-time constrained surveys ....................... 49

Table 2.9. Overall herpetofaunal species richness by year in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift
fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained survey including incidental observations ......................... 50

Table 2.10. List of herpetofaunal species captured and their combined capture
abundance in the southern Lower Peninsula of Michigan in summer 2005
and 2006 from drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 59

Table 2.11. Total herpetofaunal species captured and their monthly capture
abundance in the southern Lower Peninsula of Michigan in summer 2005
and 2006 from drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 60

Table 2.12. List of herpetofaunal species captured by year and their combined
capture abundance in the southern Lower Peninsula of Michigan in summer
2005 and 2006 from drift fence arrays coupled with pitfall traps, funnel
traps, and coverboards and area-time constrained surveys .................................. 64

Table 2.13. Species diversity indices in the southern Lower Peninsula of Michigan
based on capture data from drift fence arrays coupled with pitfall traps,
funnel traps, and coverboards and area-time constrained surveys ....................... 71

‘ 1
Appendix -
cap!

Appendix 3
$334:
and

Appendix 3
sex)
Zillif

Appendix I
and ;
Slit!

CHAPTER
MICHIGAN

Table 3.]. 5
(STA
C003
Sm.
300i

STA'

Table 3.3. L
Mun:
Dt'pg

Tibltij E

OP'CT;
lend:

”Die 3.4, I

in {if
iii,

 

Spy,
SQ?»
Sur“

pifr.
SUI.

 

Appendix 2.1. Measurements (snout to vent length (SVL) and weight) of Anurans
captured in the southern Lower Peninsula of Michigan in 2005 and 2006 .......... 90

Appendix 2.2. Measurements (snout to vent length (SVL) and weight of
salamanders captured in the southern Lower Peninsula of Michigan in 2005
and 2006 ............................................................................................................... 123

Appendix 2.3. Measurements (snout to vent length (SVL), tail length, weight, and
sex) of snakes captured in the southern Lower Peninsula of Michigan in
2005 and 2006 ...................................................................................................... 126

Appendix 2.4. Measurements (sex, carapace length, carapace width, plastron length
and plastron width) of turtles captured in the southern Lower Peninsula of
Michigan in 2005 and 2006 .................................................................................. 127

CHAPTER 3: COMPARISON OF THE HERPETOFAUNA COMMUNITIES IN
MICHIGAN’S STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

Table 3.1. Soil associations (and codes) according to US. General Soil Map
(STATSGO) for Michigan data (Soil Survey Staff, Natural Resources
Conservation Service, United States Department of Agriculture) used in
study sites selection in the southern Lower Peninsula of Michigan during the
2005 and 2006 field seasons. Descriptions of the first three major
STATSGO components are listed ........................................................................ 133

Table 3.2. Landscape categories (and codes) according to the Integrated Forest
Monitoring, Assessment, and Prescription (IF MAP) system (Michigan
Department of Natural Resources 2001b) ............................................................ 134

Table 3.3. Dates drift fence arrays, pitfall traps, funnel traps and coverboards were
opened in southern Lower Peninsula of Michigan on SGWA and private
lands in summer 2005 and 2006 ........................................................................... 137

Table 3.4. Five decay classes of the downed logs (US. Forest Service 2004) used
in area time-constrained surveys in the southern Lower Peninsula of
Michigan on SGWA and private lands in summer 2005 and 2006 ...................... 139

Table 3.5. Names and mean standardized abundance (SE) of the 12 herpetofaunal
species included in the constrained ordination analysis. Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 146

xi

Table 3.6. .\
soutl'
sumr
Inixc

T351537. C
(Toni
on 8(
sum
pilfui

SUDC

lablc3.8. 51
south
capiu
C0\CT
0b>tr

Table 3.9. Sr
RHHEI
capiur
COICd
obscn

Table 3.10. 5
south
taplu'
COTCT
ObScr

liblm l. .
mpg;
Pmr
REL:
arm. '

 

Ta5163.12.
“Ill

 

Ian-i
300!
fun:
I
'r .
“Ale 3.]:
”Nil
P11» .

 

Table 3.6. Mean (SE) and range of habitat and weather variables recorded in the
southern Lower Peninsula of Michigan on SGWA and private lands in
summer 2005 and 2006 and used in model selection for repeated measures,
mixed models, and constrained ordination ........................................................... 148

Table 3.7. Gradient lengths of the ordination axes from the Detrended
Correspondence Analysis (DCA). Herpetofaunal community data collected
on SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 151

Table 3.8. Species observed by land ownership on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained survey including incidental
observations .......................................................................................................... 1 54

Table 3.9. Species observed by soil type on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained survey including incidental
observations .......................................................................................................... 1 55

Table 3.10. Species observed by vegetation type on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained survey including incidental
observations .......................................................................................................... 1 56

Table 3.11. AIC scores for 5 best amphibian species richness mixed modelsa with
repeated measures on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift
fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys. ............................................................................. 158

Table 3.12. Type 3 fixed effects for the amphibian species richness mixed model
with repeated measures (a = 0.05). Data collected on SGWA and private
lands in the southern Lower Peninsula of Michigan in summer 2005 and
2006 based on captures by drift fence arrays coupled with pitfall traps,
funnel traps, and coverboards and area-time constrained surveys ....................... 159

Table 3.13. Solution for random effects for the amphibian species richness mixed

model with repeated measures (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005

xii

and 31
funnel

Table 3.14. .\
nudd
in IIIC
deCII
UUp&

Table 3.15. .-‘
Mill 1
Penin
Tenee
meat

Table 3.16. '
mfldt‘
P11) i'.
and I
Turn

Table 3,171
”100'.
P”);

Table 3.18
mix.
land

31 ll}.

 

 

and 2006 based on captures by drift fence arrays coupled with pitfall traps,
funnel traps, and coverboards and area-time constrained surveys ....................... 160

Table 3.14. Mean comparisons by soil for the amphibian species richness mixed

model with repeated measures. Data collected on SGWA and private lands

in the southern Lower Peninsula of Michigan in summer 2005 and 2006

based on captures by drift fence arrays coupled with pitfall traps, funnel

traps, and coverboards and area-time constrained surveys. ................................. 161

Table 3.15. AIC scores for 5 best herpetofaunal species richness mixed modelsa

with repeated measures on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift

fence arrays coupled with pitfall traps, funnel traps, and coverboards and

area-time constrained surveys .............................................................................. 162

Table 3.16. Type 3 tests of fixed effects for herpetofaunal species richness mixed

model with repeated measures (a = 0.05). Data collected on SGWA and

private lands in the southern Lower Peninsula of Michigan in summer 2005

and 2006 based on captures by drift fence arrays coupled with pitfall traps,

fiJnnel traps, and coverboards and area-time constrained surveys ....................... 163

Table 3.17. Solution for random effects for herpetofaunal species richness mixed

model with repeated measures (a = 0.05). Data collected on SGWA and

private lands in the southern Lower Peninsula of Michigan in summer 2005

and 2006 based on captures by drift fence arrays coupled with pitfall traps,

fimnel traps, and coverboards and area-time constrained surveys ....................... 164

Table 3.18. Mean comparisons for soil types for herpetofaunal species richness

mixed model with repeated measures. Data collected on SGWA and private

lands in the southern Lower Peninsula of Michigan in summer 2005 and

2006 based on captures by drift fence arrays coupled with pitfall traps,

funnel traps, and coverboards and area-time constrained surveys ....................... 165

Table 3.19. AIC scores for 5 best herpetofaunal species diversity mixed models8

with repeated measures on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift

fence arrays coupled with pitfall traps, funnel traps, and coverboards and

area-time constrained surveys. ............................................................................. 167

Table 3.20. Type 3 tests of fixed effects for herpetofaunal species diversity mixed

model with repeated measures (a = 0.05). Data collected on SGWA and

private lands in the southern Lower Peninsula of Michigan in summer 2005

and 2006 based on captures by drift fence arrays coupled with pitfall traps,

funnel traps, and coverboards and area-time constrained surveys ....................... 168

xiii

Table 3.21.
mud
pm.
and

‘

tuna

Table 3.33.
50“
5mm
pitii
sunc

Table 3.33.
0.051
Penn:
fence
area:

CCTPIU
Ctll'r’r

Table 3.35. 1
mode
Penn
It'llt‘c
area-f

Tfik326
min“.

UL

Petr
Tim.

 

area-

 

 

Table 3.21 . Solution for random effects for herpetofaunal species diversity mixed
model with repeated measures (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005
and 2006 based on captures by drift fence arrays coupled with pitfall traps,
funnel traps, and coverboards and area—time constrained surveys ....................... 169

Table 3.22. AIC scores for 5 best amphibian species richness mixed models8 on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 170

Table 3.23. Type 3 fixed effects for amphibian species richness mixed model (a =
0.05). Data collected on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift
fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys. ............................................................................. 171

Table 3.24. Solution for random effects for amphibian species richness mixed
mode] (or = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 172

Table 3.25. Mean comparisons by soil for amphibian species richness mixed
model. Data collected on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift
fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys. ............................................................................. 173

Table 3.26. Mean comparisons by vegetation for amphibian species richness mixed
mode]. Data collected on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift
fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys. ............................................................................. 174

Table 3.27. AIC scores for 5 best herpetofaunal species richness mixed modelsa on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 176

Table 3.28. Type 3 Analysis of Variance for herpetofaunal species richness mixed

model (a = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on

xiv

Table 3.311.
mm
Pen
Tent
am

Table331.
mix:
Lou

bid

 

 

 

captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys .................... - 177

 

Table 3.29. Solution for random effects for herpetofaunal species richness mixed

model (a = 0.05). Data collected on SGWA and private lands in the

southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 178

Table 3.30. Mean comparisons by soil for herpetofaunal species richness mixed

mode]. Data collected on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift

fence arrays coupled with pitfall traps, funnel traps, and coverboards and

area-time constrained surveys. ............................................................................. 179

Table 3.31. Mean comparisons by vegetation for herpetofaunal species richness

mixed model. Data collected on SGWA and private lands in the southern

Lower Peninsula of Michigan in summer 2005 and 2006 based on captures

by drift fence arrays coupled with pitfall traps, funnel traps, and coverboards

and area-time constrained surveys ....................................................................... 180

Table 3.32. AIC scores for 5 best herpetofaunal species diversity mixed modelsa on

SGWA and private lands in the southern Lower Peninsula of Michigan in

summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, fiinnel traps, and coverboards and area-time constrained

surveys .................................................................................................................. 181

Table 3.33. Type 3 analysis of variance for herpetofaunal species diversity mixed

model (or = 0.05). Data collected on SGWA and private lands in the

southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, firnnel traps, and
coverboards and area-time constrained surveys ................................................... 182

Table 3.34. Solution for random effects for herpetofaunal species diversity mixed

model (a = 0.05). Data collected on SGWA and private lands in the

southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 183

Table 3.35. Mean comparisons by soil for herpetofaunal species diversity mixed

 

model. Data collected on SGWA and private lands in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift

fence arrays coupled with pitfall traps, funnel traps, and coverboards and

area-time constrained surveys. ............................................................................. 184

XV

Table 3.3-6.

T331633:
eacl
and
:1 III

Tab1e3.3i\.
\ m
pm
and

CHAPTER
METRICS

Table-1.1. ]

Tablet}. [
in ii
bllcl

Table-I3. I.

DCpJ

Table 44. .\l
SPCCI
SCI-‘3‘!
511m
Pitt.

SUP. ,

 

 

Table 3.36. Summary of the results of the constrained ordination explained by the
environmental variables. Herpetofaunal community data and environmental
data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006. .............................................................. 185

Table 3.37. Summary of Monte Carlo permutation tests showing significance of
each axis in the constrained ordination (a = 0.05). Data collected on SGWA
and private lands in the southern Lower Peninsula of Michigan in summer
2005 and 2006. ..................................................................................................... 187

Table 3.38. Summary of Monte Carlo permutation tests showing significance of
variables in explanatory data set (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005
and 2006. .............................................................................................................. 188

CHAPTER 4: EVALUATION OF THE EFFECTS OF LANDSCAPE PATTERN
METRICS ON HERPETOFAUNA

Table 4.1. Dates drift fence arrays, pitfall traps, funnel traps and coverboards were
opened in southern Lower Peninsula of Michigan on SGWA and private
lands in summer 2005 and 2006 ........................................................................... 215

Table 4.2. Five decay classes of the downed logs (US. Forest Service 2004) used
in area time-constrained surveys in the southern Lower Peninsula of
Michigan on SGWA and private lands in summer 2005 and 2006 ...................... 218

Table 4.3. Landscape classes and IF MAP categories developed by Michigan
Department of Natural Resources 1999-2001. ..................................................... 220

Table 4.4. Names and mean standardized abundance (SE) of the 12 herpetofaunal
species included in the constrained ordination analysis. Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 225

Table 4.5. Gradient lengths of the ordination axes fiom the Detrended
Correspondence Analysis (DCA). Herpetofaunal community data collected
on SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 228

Table 4.6. Summary of land cover types measured at the three spatial scales in the
southern Lower Peninsula of Michigan in 2005 and 2006. All land cover

xvi

bees
their
100 r
PM“

Table 4.7. S
the St
It‘prL‘
V. 11161
dl>ldl
read.
each
1103K
(SE1
TUTTI}

Table 4.8. .1
in St;
twp

Cons!

Table 4.9, I
SPCei
SCH
SUT‘JT
Pill}
SUT‘N

Tahlfi 4.11)
11.1.,»
1011
So.
Slim
. I
Pitt.
Sur.

Ill"
Ali.

COL

 

Cl 1"”

Spy
31.1

types represented the total area (m2) within each AD. Values are means,

their associated standard errors (SE) from all study sites in each data set:

100 m (N = 75); 200 m (N =55); 1000 m (N =25), and Prop. represents the
proportion of the total AD area comprised by each land cover. .......................... 231

Table 4.7. Summary of landscape variables measured at the three spatial scales in
the southern Lower Peninsula of Michigan in 2005 and 2006. Water depth
represented the average of ground water depth (m) within each AD; surface
water represented the distance from survey site (111) to nearest surface water;
distance- road represented the distance from survey site (m) to the nearest
road; road length represented the total length of roads (m) contained within
each AD; distance-urban represented the distance from survey site (m) to the
nearest urban area. Values are means and their associated standard errors
(SE) from all study sites in each data set: 100 m (N = 75); 200 m (N =55);
1000 m (N =25) .................................................................................................... 233

Table 4.8. AIC scores for 5 best amphibian species richness mixed models8 at 100
m scale on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys. ............................................................................................. 23

Table 4.9. Type 3 fixed effects and solution for random effects for amphibian
species richness mixed model at 100 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, firnnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 236

Table 4.10. Correlations between landscape variables and amphibian richness,
herpetofaunal species richness, and herpetofaunal species diversity at the
100 m scale using Pearson’s Correlation Coefficients. Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, firnnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 237

Table 4.11. AIC scores for 5 best herpetofaunal species richness mixed modelsa at
100 m scale on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys. ............................................................................................. 238

Table 4.12. Type 3 fixed effects and solution for random effects for herpetofaunal

species richness mixed mode] at 100 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in

xvii

51111111
pitfall
sun e

Table 4.13. 1‘
1011 n
\Iiehi
coupl

£01151:

Table 4.14. l
speeii
8011'
51111111".
pitfall
sun ej

Table 415. .
111 SL1:
311.1".
C(iupi
C0115?

Table 4.16.
S .L...
8011
sum"
piih
sun

Table 4.11
he»

m M,
and '-
31!.“

Ira;

 

C01.

Table 4. ] r,

Spa;
8(1

 

summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, flannel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 239

Table 4.13. AIC scores for 5 best herpetofaunal species diversity mixed modelsa at
100 m scale on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys .............................................................................................. 240

Table 4.14. Type 3 fixed effects and solution for random effects for herpetofaunal
species diversity mixed model at 100 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 241

Table 4.15. AIC scores for 5 best amphibian species richness mixed models":1 at 200
m scale on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys. ............................................................................................. 243

Table 4.16. Type 3 fixed effects and solution for random effects for amphibian
species richness mixed model at 200 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 244

Table 4.17. Correlations between landscape variables and amphibian richness,
herpetofauna species richness, and herpetofaunal species diversity at the 200
m scale using Pearson’s Correlation Coefficients. Data collected on SGWA
and private lands in the southern Lower Peninsula of Michigan in summer
2005 and 2006 based on captures by drift fence arrays coupled with pitfall
traps, funnel traps, and coverboards and area-time constrained surveys. ............ 245

Table 4.18. AIC scores for 5 best herpetofaunal species richness mixed modelsa at
200 m scale on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys .............................................................................................. 246

Table 4.19. Type 3 fixed effects and solution for random effects for herpetofaunal

species richness mixed model at 200 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in

xviii

sumir.
pitfall

sunej.

Tab1e4.30.
300 1‘.‘
Mich; I

coup‘

Ctinslr

Table 4.21. '
speci-
SG\\
sumi‘.
pill}.

SUI) 1‘

Table 4.22,
lIlI ll
0T M
aria}
Time
Table 4.23,
3pm

So "

SUIT‘?

pitt- '
.. I
5U“

Tibic 41;
her?
111
So.
3UP.
1111:.
SUP,

Table 41:
111
Of “

C117” ’

1117

1

 

Tel"burl

5p,

 

summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 247

Table 4.20. AIC scores for 5 best herpetofaunal species diversity mixed models3 at
200 m scale on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys. ............................................................................................. 249

Table 4.21. Type 3 fixed effects and solution for random effects for herpetofaunal
species diversity mixed model at 200 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 250

Table 4.22. AIC scores for 5 best amphibian Species richness mixed modelsal at
1000 in scale on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area-
time constrained surveys ...................................................................................... 251

Table 4.23. Type 3 fixed effects and solution for random effects for amphibian
species richness mixed model at 1000 m scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, firnnel traps, and coverboards and area-time constrained
surveys .................................................................................................................. 252

Table 4.24. Correlations between landscape variables and amphibian richness,
herpetofauna Species richness, and herpetofaunal species diversity at the
1000 m scale using Pearson’s Correlation Coefficients. Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area—time constrained
surveys .................................................................................................................. 254

Table 4.25. AIC scores for 5 best herpetofaunal species richness mixed models8 at
1000 m scale on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area-
time constrained surveys. ..................................................................................... 255

Table 4.26. Type 3 fixed effects and solution for random effects for herpetofaunal
species richness mixed model at 1000 m scale (or = 0.05). Data collected on

xix

SC
SUI

pit:

SUI

Table 4.3"

Table 4.2%.

Spt‘l
501

Table 4.39.
cm
SGT
sum:

Table 4.31).
Sign}
CUIIL

blitl

Table .13] .
Slit
Ebb.
urb_

“ch

 

0113'
(St.
in i
so;

 

SGWA and private lands in the southern Lower Peninsula of Michigan in

summer 2005 and 2006 based on captures by drift fence arrays coupled with

pitfall traps, funnel traps, and coverboards and area-time constrained

surveys .................................................................................................................. 256

Table 4.27. AIC scores for 5 best herpetofaunal species diversity mixed modelsa at

1000 m scale on SGWA and private lands in the southern Lower Peninsula

of Michigan in summer 2005 and 2006 based on captures by drift fence

arrays coupled with pitfall traps, funnel traps, and coverboards and area-

time constrained surveys. ..................................................................................... 257

Table 4.28. Type 3 fixed effects and solution for random effects for herpetofaunal

species richness mixed model at 1000 in scale (a = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in

summer 2005 and 2006 based on captures by drift fence arrays coupled with

pitfall traps, funnel traps, and coverboards and area-time constrained

surveys .................................................................................................................. 258

Table 4.29. Summary of the results of the constrained ordination for herpetofaunal

community data explained by the landscape variables. Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006 ........................................................................................ 259

Table 4.30. Summary of Monte Carlo permutation tests for landscape data showing

significance of each axis in the constrained ordination (a = 0.05). Data
collected on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 .................................................................... 261

Table 4.31. Summary of Monte Carlo permutation tests for landscape scale data

showing significance of variables in explanatory data set (a = 0.05).
Forest200 represented by the total area of forest (m2) in the 200 m AD;
urban100 represented the total area of urban (m2) in the 100 m AD;

wetland] 000 represented the total area of wetland (m2) in the 1000 m AD;
ownership represented the land ownership where the survey site was located

(SGWA vs. private); wetland200 represented the total area of wetland (m2)
in the 200 m AD. Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 ..................... 262

XX

lnaees. in th

CHAPTER
111C HIGAA

Pig. 1.1. 10
of Region 1
represent pr.
“a: located

CHAPTER 1
STATE GA.‘

fig. 3.]. 01
south
capta
CUVL'I
OTlSCI

Fig 3.3. Fie
Penir
fence
area-
nunti
mint:
T6111:

be 3.3. 11.

 

 

LIST OF FIGURES

Images in this thesis are presented in color.

CHAPTER 1: EVALUATION OF THE HERPETOFAUNAL DIVERSITY OF
MICHIGAN’S STATE GAME AND WILDLIFE AREAS

Fig. 1.1. Location of study Sites in the Lansing sub-subsection of Ionia subsection

of Region 1 ecoregion within Michigan’s Peninsula, 2005 and 2006. Nirmbers

represent private land owners and SGWA surveyed in each county. One SGWA

was located in 2 different counties ................................................................................... 20

CHAPTER 2: DESCRIPTION OF THE HERPETOFAUNAL COMMUNITIES IN
STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

Fig. 2.1. Overall herpetofaunal Species occurrence in 82 sampling sites in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys including incidental
observations .......................................................................................................... 52

Fig. 2.2. Frequency of herpetofaunal species occurrence in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drift
fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys. The top dashed line represents the maximum
number of sites that species occurred; the bottom dashed line represents the
minimum number of site that species occurred; and the straight line
represents the mean number of sites that species occurred .................................. 53

Fig. 2.3. Herpetofaunal species occurrence on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
capture data from drift fence arrays, pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 54

Fig. 2.4. Herpetofaunal species occurrence by year in the southern Lower
Peninsula of Michigan in summer 2005 and 2006-based on captures by drift
fence arrays coupled with pitfall traps, firnnel traps, and coverboards and
area-time constrained surveys including incidental observations ........................ 55

Fig. 2.5. Frequency of Anuran occurrence in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on frog call surveys ......................... 57

Fig. 2.6. Frequency of Anuran occurrence by year in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on frog call surveys ..................... 58

xxi

Fig 3.7. T0
soutl
capti

Ct“ L’

Fig. 2.3. I.»
south
captu
cm CT

  
 

fig. 3.10. A‘
lamb

31111:"

 

Fig. 2.7. Total captures of herpetofaunal species on SGWA by month in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
capture data from drift fence arrays, pitfall traps, funnel traps, and
coverboards and area—time constrained surveys ................................................... 62

Fig. 2.8. Total captures of herpetofaunal species on private lands by month in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
capture data from drift fence arrays, pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 63

Fig. 2.9. Mean relative abundance of herpetofaunal species in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on capture data
from drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys ................................................... 66

Fig. 2.10. Mean relative abundance of herpetofaunal species on SGWA and private
lands in the southern Lower Peninsula of Michigan in summer 2005 and
2006 based on capture data from drift fence arrays coupled with pitfall traps,
funnel traps, and coverboards and area-time constrained surveys ....................... 67

Fig. 2.1 1. Mean relative abundance of herpetofaunal species by year in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
capture data from drift fence arrays coupled with pitfall traps, funnel traps,
and coverboards and area-time constrained surveys ............................................ 68

Fig. 2.12. Frequency of species occurrence versus species mean abundance
(standardized for trap effort) in the southern Lower Peninsula of Michigan
in summer 2005 and 2006 based on capture data from drift fence arrays with
pitfall traps, funnel traps, and coverboards and area-time constrained
surveys. Species labeled are those with high frequencies of occurrence
(>35) and high mean abundances (>0.5) with the exception of eastern tiger
salamander and red-backed salamander, which occurred infrequently, but
had high mean abundances ................................................................................... 69

CHAPTER 3: COMPARISON OF THE HERPETOFAUNAL COMMUNITIES IN
MICHIGAN’S STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

Fig. 3.1. Joint plot of species and sample sites. Species shown in red letters and
represented by codes (see Table 3.5) and Sites shown in black numbers.
Species fall outside of sites indicating that a linear response model is
appropriate. Herpetofaunal community data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005
and 2006 based on captures by drift fence arrays coupled with pitfall traps,
funnel traps, and coverboards and area-time constrained surveys ....................... 152

xxii

Fig. 3.2. Con
(500 T
occur
are re
extent
10 the
specie
the er
\anal
ordin.

CC'llxlI

 

 

 

Fig. 3.2. Constrained ordination diagram (biplot). Species are represented by codes
(see Table 3.5), the proximity of species in ordination space indicates
occurrence in similar environmental conditions. Environmental variables
are represented by vectors, which point toward rate of maximum change and
extend in both directions. The length of the vector indicates its importance
to the constrained ordination (ter Braak 1986). Perpendiculars drawn from
species to vectors give the approximate ranking of that species response to
the environmental variable and indicate the species optimum on that
variable (ter Braak 1986) A smaller angle between the vector and the
ordination axis indicates a greater relationship of the variable to the derived
constrained ordination gradient (Grand and Mello 2004) .................................... 189

Fig. 3.3. Plot of eastern American toad abundance in a Generalized Additive Model
(GAM) surface for distance to water (m) on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006.

Increased size of circles indicates greater abundance .......................................... 190

Fig. 3.4. Plot of eastern American toad abundance in a Generalized Additive Model
(GAM) surface for overstory canopy cover (%) on SGWA and private lands
in the southern Lower Peninsula of Michigan in summer 2005 and 2006.
Increased size of circles indicates increase in abundance .................................... 191

Fig. 3.5. Plot of wood frog abundance in a Generalized Additive Model (GAM)
surface for overstory canopy cover (%) on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006.
Increased size of circles indicates increase in abundance .................................... 192

Fig. 3.6. Plot of wood frog abundance in a Generalized Additive Model (GAM)
surface for litter depth (cm) on SGWA and private lands in the southern
Lower Peninsula of Michigan in summer 2005 and 2006. Increased size of
circles indicates increase in abundance ................................................................ 193

Fig. 3.7. Plot of spring peeper abundance in a Generalized Additive Model (GAM)
surface for leaf litter depth (cm) on SGWA and private lands in the southern
Lower Peninsula of Michigan in summer 2005 and 2006. Increased size of
circles indicates increase in abundance ................................................................ 194

Fig. 3.8. Plot of red-backed salamander abundance in a Generalized Additive
Model (GAM) surface for overstory canopy cover (%) on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005
and 2006. Increased size of circles indicates increase in abundance .................. 195

Fig. 3.9. Plot of green frog abundance in a Generalized Additive Model (GAM)
surface for overstory canopy cover (%) on SGWA and private lands in the

xxiii

T“.
([0

‘1)

H

C HAPT
.‘leTRll
Fig. 4.1.
r
S
a
P
31
ft

Fig. 42.
re
51‘
Er

 

southern Lower Peninsula of Michigan in summer 2005 and 2006.

Increased size of circles indicates increase in abundance .................................... 196

Fig. 3.10. Percent of total variance in the herpetofaunal community data set in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys explained by habitat
variables, site stratification variables, and weather variables, as well as
variance explained by each combination of factors. Of the variation in the

herpetofaunal community, 71.8% of the variance was left unexplained ............. 198

CHAPTER 4: EVALUATION OF THE EFFECTS OF LANDSCAPE PATTERN
METRICS ON HERPETOFAUNA

Fig. 4.]. Joint plot of species and sample sites. Species shown in red letters and
represented by codes (see Table 4.4) and sites shown in black numbers.
Species fall outside of Sites indicating that a linear response model is
appropriate. Herpetofaunal community data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005
and 2006 based on captures by drift fence arrays coupled with pitfall traps,

firnnel traps, and coverboards and area-time constrained surveys ....................... 229

Fig. 4.2. Redundancy Analysis ordination diagram (biplot). Species are
represented by codes (see Table 4.4), the proximity of species in ordination
space indicates occurrence in similar environmental conditions.
Environmental variables are represented by vectors, which point toward rate
of maximum change and extend in both directions. The length of the vector
indicates its importance to the constrained ordination (ter Braak 1986).
Perpendiculars drawn from species to vectors give the approximate ranking
of that species response to the environmental variable and indicate the
species optimum on that variable (ter Braak 1986). A smaller angle
between the vector and the ordination axis indicates a greater relationship of
the variable to the derived constrained ordination gradient (Grand and Mello

2004). Forest200 represented by the total area of forest (m2) at a 200 m
scale; urbanIOO represented the total area of urban (m2) at a 100 m scale;

wetlandIOOO represented the total area of wetland (m2) at a 1000 m scale;
ownership represented the land ownership where the survey site was located

(SGWA vs. private); wetland200 represented the total area of wetland (m2)

at a 200 In scale .................................................................................................... 263

Fig. 4.3. Plot of eastern American toad abundance in a Generalized Additive Model
(GAM) surface for urban land class (m2) at a scale of 100 III. Data collected
on SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006. Increased size of circles indicates increase in

abundance ............................................................................................................. 264

xxiv

fig. 4.4. PIOI
sudae
SGW.
summ
abund.

   

fig-1.5. Plt‘l
surfue
SG\V1
sunini
abund

1g.ie Per
expla
“uhn
“e0;
exnh
Lout
by d:
ands

 

 

Fig. 4.4. Plot of spring peeper abundance in a Generalized Additive Model (GAM)

surface for wetland land class (m2) at a scale of 1000 m. Data collected on

SGWA and private lands in the southern Lower Peninsula of Michigan in

summer 2005 and 2006. Increased size of circles indicates increase in

abundance ............................................................................................................. 265

Fig. 4.5. Plot of wood frog abundance in a Generalized Additive Model (GAM)

surface for forest land class (m2) at a scale of 200 m. Data collected on

SGWA and private lands in the southern Lower Peninsula of Michigan in

summer 2005 and 2006. Increased Size of circles indicates increase in

abundance ............................................................................................................. 266

Fig. 4.6. Percent of total variance in the herpetofaunal community data set
explained by urban land class within 100m of survey site, wetland land class
within 1000 m of the survey site, and scale200 (comprised of forest and
wetland land classes within 200 m of survey sites), as well as variance
explained by the combination of factors. Data collected in the southern
Lower Peninsula of Michigan in summer 2005 and 2006 based on captures
by drift fence arrays coupled with pitfall traps, funnel traps, and coverboards
and area-time constrained surveys ....................................................................... 268

XXV

(HAW

Need for I

Thr
Stale dim 113
occurrfil at
extensiVe “
Ill 1.116 SOUTT
ditched and
31101).

Alie
ebnsen 3110
to set t. e fit
lirgtim‘unur
iDernpse} I
the \Iiehig;

Depanmer.

 

manages ti
Illusion is

Wildlife r.

 

CHAPTER 1: EVALUATION OF THE HERPETOFAUNAL DIVERSITY OF
MICHIGAN ’8 STATE GAME AND WILDLIFE AREAS

INTRODUCTION
Need for Research

The Michigan landscape has been transformed since the pioneers settled in the
state during the early nineteenth century. The settlement of Michigan by Europeans
occurred at a slower pace than other states to the south, possibly due in part to the
extensive wetlands in southern Michigan. The pattern of human emigration and rich soils
in the southern portion of the state resulted in forests being cleared and wetlands being
ditched and tiled to make way for highly productive agricultural practices (Dempsey
2001).

Michigan has a long history of resource conservation and enacted several
conservation laws as early as 1859 to ban the netting of fish on inland waters, as well as
to set the first closed hunting seasons for species such as white-tailed deer (Odocoileus
virginianus), wild turkey (Meleagris gallopavo), and mallards (Anas platyrhynchos)
(Dempsey 2001). In 1921, the Department of Conservation was established, and in 1968,
the Michigan Department of Natural Resources (MDNR) was created (Michigan
Department of Natural Resources 2001 a). Currently, the Wildlife Division of MDNR
manages the state’s wildlife resources and supporting habitat, and the mission of this
division is to promote the enhancement, conservation, and restoration of the state’s
wildlife resources and ecosystems (Michigan Department of Natural Resources 2004).

In Michigan, state game and wildlife areas (SGWA) have been established and
maintained to improve and restore wildlife populations and habitat. In accordance with

MDNR’S mission, these areas are presumed to be essential components in the

ainSm‘ai
tall-13536 I
acqulk‘d i
with JIM
applicant
grateu‘idc

Th.
Midlife ha
and ri’plllt“
apes Am
histories tit
and C then
amphibian I
lll'elx‘h ant-

bnefsini tp

  
  
   
  

\IIL ligan . \

Compan' $4 1 :'

13mm in, ‘

\.

blniihm

conservation of biological diversity throughout the state. Ecologically, SGWA are
valuable because they are lands set aside specifically for wildlife. They were originally
acquired from lands not suitable for farming in the southern portion of the state and
policies provided for purchasing these lands required that they be larger tracts of
minimally-developed lands with the potential for wildlife restoration (Gordinier 1960,
Application for Federal Assistance for the Michigan Department of Natural Resources-
Statewide Land Acquisition Grant October 16, 1998).

The purpose of this study was to determine whether SGWA are providing unique
wildlife habitat to the southern Michigan landscape. To answer this question, amphibians
and reptiles were used as one indicator of biological diversity on differing land ownership
types. Amphibians are potential environmental indicators because their complex life
histories depend on both aquatic and terrestrial habitats (Vitt and Caldwell 1990, Stebbins
and Cohen 1995, Alford and Richards 1999, Collins and Storfer 2003) and reductions in
amphibian diversity and abundance have been shown to indicate ecosystem stress
(Welsch and Olliver 1998). To help provide context for the research project, I provide a
brief synopsis of 1) the role of herpetofauna as biological indicators, 2) background on
Michigan’s SGWA 3) the value of private lands to biodiversity conservation, 4)
comparisons of research done on private and public lands and 5) the effect of landscape
pattern metrics on herpetofauna. These main topics influenced study design and

hypothesis in my work.

The Imports
:Anipl
AnpbduanS'
Stebbins and
0411c“ and
AMMde
ehangesirirc
lllt?)t\'l11bll :
thhards 1111
lmmt‘brute
Amp
Illens 1 0“ 5
formal Eco
dcIC‘C'llll‘iber

~.

eIIC‘LT] 2‘ CITE \

w -

mmmlut:

 

for 5111111; L»

abundance
PH COM“
1

“Wand-

Stein .

1

I THE“,
C .

,
‘TIVSCnt

The Importance of Herpetofauna as Biological Indicators

Amphibians can be key indicators of environmental change for several reasons.
Amphibians use aquatic and terrestrial habitats during their life cycles (McDiarmid 1994,
Stebbins and Cohen 1995), their permeable skin allows for absorption of toxicants
(Maxell and Hokit 1999) and water (Vitt and Caldwell 1990, Stebbins and Cohen 1995,
Alford and Richards 1999, Collins and Storfer 2003), making them more susceptible to
changes in temperature, rainfall, and environmental toxins (Alford and Richards 1999),
they exhibit sensitivity to water chemistry during egg and larval development (Alford and
Richards 1999), and they are an important trophic linkage (providing efficient transfer of
invertebrate resources to higher-level organisms) (Stebbins and Cohen 1995).

Amphibians are thought to play a critical role in ecosystem dynamics (Burton and
Likens 1975, Perkins and Hunter 2006). They are the most abundant vertebrate in many
forested ecosystems, playing a crucial role in structuring communities of forest floor
decomposers (Burton and Likens 1975). Welsh and Ollivier (1998) demonstrated the
effectiveness of using amphibians as indicators of ecosystem stress. They found that
stream amphibian densities were considerably lower in streams impacted by
sedimentation. Welsh and Droege (2001) promoted the use of Plethodontid salamanders
for biodiversity monitoring of North American forests, claiming that changes in
abundance correspond to several microsite conditions including leaf litter, soil moisture,
pH, coarse woody debris, available burrows, and canopy cover. Plethodontid
salamanders demonstrate site fidelity, low fecundity, hold relatively small territories and
are long-lived; characteristics that make variations in salamander counts more likely to

represent substantial environmental changes (Welsh and Droege 2001).

Recent
populations ha
311.111. Sex era.
including the I
Duellman 19"
non-natixe Sp;
illlfi'CllOuS d1.
ultraxiolet B .
climate than
Cbeinical p .j
CUIIITIS and \

the £31136 UT

Pillltttin E (1'

.-,

i ,.
ht USE Sc

 

amp}, ll) 1 dr

‘r-L -.,
“Hi-III]

S111;

 

Recent reports of malformed amphibians and the global decline in amphibian
populations have received considerable attention in the last decade (Kiesecker et al.
2001). Several factors have been attributed to the decline of amphibian populations
including the loss, deterioration, and fragmentation of habitat (Blaustein et a1 1994,
Duelhnan 1999, Blaustein and Kiesecker 2002), the introduction of and competition with
non-native species (Blaustein and Kiesecker 2002, Kats and F errer 2003), emerging
infectious diseases (Alford and Richards 1999, Reaser and Blaustein 2005), increased
ultraviolet B (UV-B) radiation (Alford and Richards 1999, Kiesecker et al. 2001), global
climate change (Alford and Richards 1999, Kiesecker et al. 2001), and the effects of
chemical pollutants in the environment (Alford and Richards 1999, Duellman 1999,
Collins and Storfer 2003). Complex interactions among multiple factors are most likely
the cause of the global amphibian decline, which further complicates the issue of
protecting amphibians (Blaustein and Kiesecker 2002).

Although the causes of amphibian declines have not been clearly determined,
habitat modification including habitat loss and fragmentation is the best-documented
cause of amphibian declines (Alford and Richards 1999). Both habitat loss and habitat
fragmentation are considered to be main causes of decline in the Midwest United States
(Lannoo 1998). The loss and degradation of wetlands is a major threat to most
amphibian species and directly impacts breeding (Houlahan and Findlay 2003). Species
that use several, different habitats during varying parts of their life cycle such as
amphibians may be particularly vulnerable to changes in the landscape such as habitat

modification, fragmentation, and actual habitat loss (Regosin et al 2005).

011161 13

the climate or I
maspheric pt
Cornell 19.83.
nt‘alit} ot'bi
lanal defornii
Alford and R:
premature 111:?
lAl'ford and b
Amphibian e

the atmospbL

1999), Dep

€335 £11111 la."

SEnergixtie;

Pill 111:4 m

blt’fah (1110
long effect

3'- ‘ .

 

 

Other causes of decline have been more difficult to document such as changes in
the climate or chemical pollutants in the environment. The effects of low pH fiom
atmospheric pollution can impact amphibian populations (Pough 1979, Dunson and
Connell 1982, Cummins 1989). For example, low pH levels have been linked to the
mortality of both embryos and larvae causing incomplete absorption of the yolk plug,
larval deformities, and development to cease prematurely (Pough 1979, Cummins 1989,
Alford and Richards 1999). Sublethal effects of acidification include delayed or
premature hatching of eggs, a reduction in larval body size, slower larval growth rates
(Alford and Richards 1999), and reduction in larval swimming ability (Kutka 1994).
Amphibian eggs and larvae are also susceptible to greater UV-B radiation resulting from
the atmospheric pollution of chlorofluorocarbons (Alford and Richards 1999, Duellman
1999). Depletion of the stratospheric ozone has lead to greater exposure of amphibian
eggs and larvae to UV-B radiation (Duellman 1999). UV-B radiation alone or
synergistically with other agents such as pathogens can have detrimental effects, such as
high egg mortality, in certain amphibian Species (e.g., see Kiesecker and Blaustein 1995).
Metals and chemicals used in insecticides and herbicides have also been shown to have
toxic effects on amphibians (Alford and Richards 1999, Metts et al. 2005).

Michigan’s State Game and Wildlife Areas

The Wildlife Division of MDNR was developed to conserve the State’s wildlife
and its habitats. The current mission of this program is to: “enhance, restore, and
conserve the State’s wildlife resources, natural communities, and ecosystems for the
benefit of Michigan’s citizens, visitors, and future generations” (Michigan Department of

Natural Resources 2004). Historically, the Division was called the Game Division

lblichigan D
native u‘ildll
game specie:
Ankney 1W
disappearing
and com em
rapidly and r
reduce over;
H011e1'er. .1.
reducing per;
21,104). Mitre
filabllsl‘icd‘

ecolot'ica] P'

 

(Michigan Department of Natural Resources 2004) and it focused on the reintroduction of
native wildlife species such as wild turkey and the introduction of non-native wildlife
game species such as ring-necked pheasant (Phasianus colchicus) (Burroughs 1946,
Ankney 1988). Non-natives were being introduced to occupy available niches left by
disappearing native wildlife that resulted from human encroachment, intensive logging,
and conversion to agriculture. Some native and non-native wildlife populations expanded
rapidly and exceeded the capacity of the existing habitat to support them. In an effort to
reduce overabundance, population numbers were reduced to match the available habitat.
However, as time passed, research demonstrated that management focus Shift from
reducing populations to managing habitats (Michigan Department of Natural Resources
2004). More recently, efforts to conserve plants and non-game species have been
established, with an emphasis on the importance of protecting critical landscapes and
ecological processes as demonstrated by completion and implementation of the Michigan
Wildlife Action Plan (Eagle et al. 2005) and the Biodiversity Conservation Planning
Processes (Michigan Department of Natural Resources 2005).

With passage of enabling legislation Michigan’s state game areas were
established on September 29, 1939 (Application for Federal Assistance for the Michigan
Department of Natural Resources-Statewide Land Acquisition Grant October 16, 1998).
The Wildlife Aid in Restoration Act, commonly called the Pittman-Robertson Wildlife
Restoration Act, provided state funds for management and restoration of wildlife. Funds
fiom this act are made available through an excise tax on sporting arms and ammunition
(University of New Mexico Center for Wildlife Law and R. S. Musgrave 1998). For the

first time in history, funds were made available for wildlife projects, including acquisition

and in
FLINT-’3
pUICha:
the Mid

Ought"!

(A

3CC ‘1Id \\
A3315?”
.ACqUISl U
e-intlict II
and timbc

federal A

ACQUISIIIU

 

and improvement of wildlife habitat, and many SGWA were created from lands
purchased with funds generated by this act. For the most part, SGWA have been
purchased and maintained with hunters’ money (Application for Federal Assistance for
the Michigan Department of Natural Resources-Statewide Land Acquisition Grant
October 16, 1998).

State game and wildlife areas are managed “...to improve wildlife habitat in
accord with established principles of wildlife management” (Application for Federal
Assistance for the Michigan Department of Natural Resources-Statewide Land
Acquisition Grant October 16, 1998). Outdoor recreation is allowed when uses do not
conflict with the main objectives of wildlife restoration and hunting. Sharecrop farming
and timber harvesting is also allowed if the result is a benefit to wildlife (Application for
Federal Assistance for the Michigan Department of Natural Resources-Statewide Land
Acquisition Grant October 16, 1998).

The majority of SGWA are located in the southern half of the Lower Peninsula.
Private and public lands are interspersed in this landscape making it difficult to manage
SGWA because privately owned parcels contained within a SGWA (known as private in-
holdings) are often densely developed, and this development places intensive pressure
and use on the adjacent public lands (Application for Federal Assistance for the Michigan
Department of Natural Resources-Statewide Land Acquisition Grant October 16, 1998).
Private in-holdings generally resulted from the unwillingness of private owners to sell
their land at the time the SGWA was established.

MDNR still uses Pittman-Robertson funds to acquire additional lands, particularly

private in-holdings that are of high priority as they become available (Michigan

Department.
determined
configuratit
often ITICSC
purchased I
Rest'iurecs-'
contain resi
ownership .
retlueing rh
AllCl‘ilgdn [
l0. 199M.
Wired b} l1.
dC‘CI'EflSeg L
Regime“.
The Valur
Th.
and the ex

dl" @1511} (

 

 

de'SCTTliqj

4‘. _, _

 

leafs. II". t‘

IE ;
d “J 3T1

6“ (2(1th

Department of Natural Resources 2004). Priority rating for a desired in-holding is
determined by the priority level of the associated SGWA based on size, ownership
configuration, public access, and the potential for the habitat to support wildlife. Most
often these priority in-holdings are not developed, although developed areas can be
purchased (Application for Federal Assistance for the Michigan Department of Natural
Resources-Statewide Land Acquisition Grant October 16, 1998). Private in-holdings can
contain residences and the lands can be subdivided which directly conflicts with SGWA
ownership objectives because a safety zone is required around public buildings thereby
reducing the area available for hunting (Application for Federal Assistance for the
Michigan Department of Natural Resources-Statewide Land Acquisition Grant October
16, 1998). Blocking ownership in SGWA by acquiring these in-holdings reduces threats
posed by future development, increases wildlife habitat management flexibility, and
decreases the costs associated with land administration (Michigan Department of Natural
Resources 2004).
The Value of Private Land to Biodiversity Conservation

The rapid alteration of many natural habitats, the increase of human encroachment
and the ever changing anthropogenic impacts on wildlife has led to efforts to biological
diversity conservation. Biological diversity has many different definitions, but can be
described as the variety of life and processes that occur at all levels including species
diversity, genetic diversity, and community diversity (Primack 2004). In the past 150
years, the rate of anthropogenic alteration of ecosystems has increased dramatically and
led to an unprecedented loss of species (Wilson 1988). Historically, efforts have focused

on conserving public lands (Morrison and Humphrey 2001); however, biologists and

rnanili-‘CI'S "C
\l'alCOV‘e ( W
pmfllt’ 133d:

Pm:
compllSlEg i
comm 311m
research h»
and Humphl
lmds has p."
bier. dcx clu
state emim~
suggtstcd ll"
laid on net

Wm‘d and

 

SOtht’m h

aCCf’Ull-lll‘.”
Cl -:

”K'lwgzcal

 

 

mules in

managers now recognize the need to also conserve private lands. According to Bean and
Wilcove (1997), “. . .most endangered species depend significantly on habitats found on
private lands.”

Private lands are the dominant form of land ownership in the United States,
comprising approximately 60% of the total (Hilty and Merenlender 2003); however,
conservation efforts have primarily focused on publicly held lands (Knight 1999). More
research has been conducted on public lands, most likely due to easier access (Morrison
and Humphrey 2001). Scientific knowledge gained from biodiversity research on public
lands has provided input to federal and state land management agencies. Policies have
been developed from this scientific input and these policies trickle down from federal to
state entities, as well as to county level governments (Knight 1999). It has been
suggested that private lands conservation has been avoided due to potential conflicts with
land owner property rights (Knight 1999). In Michigan, 79% of the land base is privately
owned and 29% publicly held (Michigan Department of Natural Resources 2000). In the
southern half of Michigan’s Lower Peninsula, even less land is publicly owned,
accounting for only 4% of the total.

Conservation biologists managing for biological diversity have stated that
biological diversity transcends land owner boundaries (Primack 2004). Biodiversity
resides in a diverse landscape of private and public ownerships (Saterson et al. 2004).
Landscapes are complex and dynamic and can be affected by activities on adjoining
properties. Failure to include private land in studies may result in unrepresentative
sampling, particularly at larger spatial scales (Hilty and Merenlender 2003). For

example, in the state of Florida, only 9% of the land is publicly owned (Myers 2001).

\ {Of
( CZ."
on p1.
Dam t

Speci:

done It
resents
M6762}:
fixated
35min:

inflate};-

l'aluahlc

O‘t'ntd la;-

 

Morrison and Humphrey (2001) found the majority of breeding crested caracaras
(Caracara cheriway) nested on privately owned land. Caracara nests were rarely found
on publicly owned land, most of which was being managed as natural areas to support
native plant and wildlife communities.

Species Comparisons between Private and Public Land

The importance of private lands has been established, but little research has been
done to quantify their biodiversity contribution (Hilty and Merenlender 2003). Even less
research exists that quantifies the differences between land ownership types. Hilty and
Merenlender (2003) found that only 27% of studies they reviewed had at least one site
located on private land. Apprehension by private land owners and the complexities
associated with access to private lands no doubt contribute to the limited amount of
research currently conducted on private lands (Hilty and Merenlender 2003).

Research comparing results between private and public land can provide some
valuable insight for managers. For example, research done by Spies et al. (1994) looked
at the fragmentation process in managed forest landscapes on both privately and publicly
owned lands. Variability in timber harvest rates within ownerships was higher in
privately owned land and the decline in closed-canopy forest was also greater. Morrison
and Humphrey (2001) investigated patterns of distribution and reproductive activity of
breeding pairs of crested caracara relative to land ownership and use. Eighty-two percent
of active nests were found on privately-owned ranches. Small et al. (1991) found that
ruffed grouse (Bonasa umbellus) mortality related to hunting on public lands versus
private lands was higher (P _<_ 0.01) for both adults (73 % vs. 13 %) and juveniles (56 %

vs. 9 %), respectively.

10

V
mics. it :
land on n
obtainui.
and Wild
reference
owners tl
proving t'
Valuable ;
The Imp.

L;
1117 :31 gem
is lmPOfi;
leflljgc‘
Wironm,

 

arr-Didi);

While research has been done to compare differences between the land ownership
types, it is often limited by the ability to gain access to the privately-owned land fiom the
land owner. Beneficial research can be conducted when access to private property is
obtained. For example, restoration of privately owned wetlands through the US. Fish
and Wildlife Service Partners for Fish and Wildlife Program were compared with
reference wetlands on SGWA (Thompson 2004). The ability to access cooperative land
owners through an established Partner’s program no doubt helped facilitate this research,
proving that the removal of the existing barriers to private lands could indeed provide
valuable insight to complex nature of landscape management needs and concerns.

The Importance of Landscape Pattern Metrics on Herpetofauna

Landscapes are characterized by the size and type of patches, as well as the spatial
arrangement of patches within the landscape (Dunning et al. 1992). The size of the patch
is important because smaller patches are more likely influenced by external factors.
Likewise, organisms living in larger patches are less likely to be affected by
environmental and biotic changes associated with edge effects (Saunders et al. 1991).
Additionally, larger patches have greater species richness (MacArthur and Wilson 1967).
The shape of the patch is also important, particularly for smaller patches because of
potential increases in edge to interior ratio as shape complexity increases (Saunders et al.
1991)

Landscape configuration refers to the spatial arrangement of patches within the
landscape (McGari gal 2002). Landscape configuration is important to consider for
amphibians and reptiles, especially because they tend to require distinct habitats for

different life stages (Guerry and Hunter 2002). Landscape complementation refers to the

11

pr‘.‘

u’ndlllti]

l0 jp‘l
”jut?

 

proximity of critical habitat patches for species that require multiple patches during its
life cycle (Dunning et al. 1992). Critical resources such as breeding ponds or foraging
patches can be complemented by resources in close proximity to another patch (Dunning
et al. 1992). Species that have limited dispersal abilities and small home ranges, such as
amphibians, can be greatly affected by landscape complementation (Guerry and Hunter
2002).

Anthropogenic land use such as agriculture or urban tends to reduce habitat
availability, habitat suitability, and landscape connectivity and therefore can negatively
impact amphibian abundance or occurrence (Bonin et al. 1997, Hecnar 1997, Knutson et
al. 1999). Landscape-level features such as the density and arrangement of roads,
forested areas surrounding breeding ponds, and proximity of habitat patch to water bodies
are important in structuring amphibian communities (Knutson et a1. 1999). Species
richness of herpetofauna in wetlands has been shown to decrease with increasing road
density on adjacent lands (Findlay and Houlahan 1997). Local vegetation and soil
conditions can be affected by land management practices and in turn, have the potential
to influence the occurrence and relative abundance of herpetofauna.

The landscape configuration in southern Michigan (i.e., a patchwork of SGWA in
a matrix of private lands) makes understanding the contribution of private lands to
SGWA conservation objectives a priority. With the majority of land in southern
Michigan not being specifically managed for wildlife, it is essential to quantify the
impact these lands have on SGWA management goals like conserving and enhancing
wildlife and wildlife habitat. To properly understand the ecological importance of

SGWA to the overall landscape, and in turn to the herpetofauna communities,

12

herpetofaunal species richness, occurrence, and abundance at sites of similar vegetation

type and soil association were compared on multiple scales of patch and landscape.

l3

nut-I.

Ther
Michigan's 1

Objectix'es 0

 

 

 

GOAL AND OBJECTIVES

The overall goal of this project was to evaluate the ecological contribution of

Michigan’s SGWA in providing herpetofaunal habitat to the landscape. The specific

objectives of this research were to:

1)

2)

3)

4)

survey the amphibian and reptile communities in selected Michigan
SGWA and privately-owned lands to deterrrrine the occurrence, relative
abundance, and diversity of these species,

compare the amphibian and reptile communities between SGWA and
privately-owned lands to ascertain the importance of the SGWA in the
overall landscape, and determine if micro-site conditions such as
vegetation, soil type, rainfall, and distance to nearest water body influence
the occurrence and relative abundance of these species,

assess landscape scale influences on the occurrence and relative
abundance of amphibians and reptiles on SGWA and private lands, and
make management recommendations for herpetofaunal diversity and

conservation on SGWA.

l4

(199:5) .
totaling
a broad
relatl‘. c
10‘) d3)

lflzgus
0th h;
and “vi ct
link 3 lk’
51*.ng
hcax } a
Lie Far:
L‘ili‘onu

83*?crn\

(bar;

(“M-at.

STUDY AREA

The study area was located within the Lansing sub-subsection of Ionia (IV. 1)
subsection of Region 1 in the southern Lower Peninsula of Michigan, based on Albert’s
(1995) ecoregions. The Lansing sub-subsection is the largest in the Lower Peninsula,
totaling 13,092 krnz. The entire state of Michigan was glaciated and the sub-subsection is
a broad till-plain characterized by rich, loamy soils. It has uniform annual precipitation,
relatively mild climactic conditions contributing to a growing season ranging from 140 to
160 days, and gently rolling topography. Three large rivers occur in this sub-subsection:
the Maple, Grand, and Thomapple, as well as other smaller rivers.

Pre-settlement plant vegetation in this sub-subsection was characterized by beech
(F agus grandzfolia)-sugar maple (Acer saccharinum) forests, oak-hickory forests, and
other hardwoods such as black maple (Acer nigrum), as well as lowland swamp forests
and wet prairie. Current land use is dominated by agriculture. The majority of uplands
have been converted to row-crop agriculture, and pastureland has replaced most of the
swamp forests. Urbanization has occurred in this sub-subsection, and combined with
heavy agricultural use, few large tracts of forest remain. This area also contains some of
the rarest plant communities in Michigan: inland salt marsh and wet prairie.
Unfortunately, only one salt marsh remains, and all of the wet prairies have either been
drained for agricultural purposes or badly degraded (Albert 1995).

Site locations were selected from map data using ArcGIS 9.1 (Environmental
Systems Research Institute 2004). Specific patch sites were chosen based on three
characteristics: 1) land ownership (SGWA vs. private), 2) vegetation type, and 3) soil

association. The Integrated Forest Monitoring, Assessment, and Prescription (IFMAP)

15

 

QanXM
Mhumg\
pmflHXl
Mhumg:
flmdeL
Mmdwf'
Edit 3rd ( ,\

En
afiqtnts
Of land rm
nature Ccn
Wnklj
Created an
l““lnl SCH:
S} slt‘m\

“n m .-

 

‘- A Cr]
GWVQ I

md75

 

 

system (Michigan Department of Natural Resources 2001b) was used to select sites in the
following vegetation types: lowland hardwoods (LHM, northern hardwoods (N HW),
pine (PIN), upland hardwoods (U HW) (Table 1.1). Sites were also chosen in the
following soil associations using US. General Soil Map (STATSGO) for Michigan data
(United States Department of Agriculture 1994): Houghton—Carlisle-Adrian (HCA),
Marlette-Capac-Parkhill (MCP), Marlette-Capac-Spinks (MCS), Miami-Hillsdale-
Edward (MHE), and Spinks-Houghton-Boyer (SHB) (Table 1.2).

Eighty-two sampling sites were located on 24 different private parcels and 7
different SGWA. Private lands ranged from 3 to 105 hectares (Table 1.3) with the type
of land owner varying from private individual to nature preserve closed to the public to
nature center opened to the public. SGWA ranged in size from 82 to 1,921 hectares
(Table 1.3). All of the SGWA were accessible to the public. Sampling points were
created within a patch comprised of a particular soil and vegetation type using a random
point generator in the Animal Movement extension of ArcView 3.2 (Environmental
Systems Research Institute 1999). Due to limitations of finding cooperative land owners
on particular soil and vegetation types, replicate sites were often located on SGWA after
a site on private land was selected. Of the 13 different SGWA in the Lansing sub-
subsection, 7 were used in the study: Dansville, Grand River, Middleville, Muskrat, Oak
Grove, Portland, and Rose Lake. Twenty-four private land owners located in 7 counties

and 7 SGWA located in 6 counties were used in this study (Figure 1.1, Table 1.4).

16

_3 3:12.29.— ._.4..$3._ 2.722237. .2: 2.. 2.4.2.227. 23,292.; 1.2.7.. >727. 2,. 7.7.: A: :CCr. 7.3.3.2...732 22.222.22.22 22.222.22..1..A\ :...:.:\...C z;
2747): 7.832;: 2.2722929... 12: 4.5.—_xzbzzzx .1_._._.::._2C—.< 22.92.; 74:21.42: 5:: .2 12:12.49... .1..1..2U..Z2U 22:22.34...’ .\ .\ LEE:

 

mm .Nmogo me $00 @888 much 32663 mo :09?“on SEQ 803253 98:5 @822
2 n .3058 05 we o\ooo meoooxo mafia (Ho 5quon A27: 0:5

.3088
2: mo $8 $085.. Ememwegwmze 338$ :83 30:?» Adam
“.3525 Ego Aeeeotmfie 32398.25 :3 323 .Aencotcfie 3:5

 

hm @0958: £82. 58584 Aim». .393 839: .3 28:33:80 ABEZV wooaecmm c.8582
.mzom 2.6%; so “898 9.600: 33062.5.“ mm 35
on 9:2 so 3980 2: mo $8 25088 305 3.036% no 2232305 952.: 3onch 98.33
ax; >0a§oo<
cosmowmmmflo eomcmtomoa xcowouau :o_§owo>

 

.2833 Bee ooom was moom 2.: macaw cmwfioaz
mo «_smficom .533 50:38 05 5 Bang 5:028 mozm beam 5 com: 383 80.50QO 3.53 Z we “sugaoa ”83:33:
80?? $92sz 553835 98 .EoEmmomm< .wctozcoE 580m wBfiwBE on: 8 wEEooom motowgmo cosfiowo> .2 03a...

17

     

n . 1. n— i. q. l\.\yu.-. 1 .u.ru . . —r 1 aw .-1.1 1 ..1|-..... . n In . .. ~.-..H‘ u\—\—\h. Nun-a. ll—\-\P .‘a~s nuau...--\\ —~:-—.-\.u_-N (..\An ~..\s~. . A

.533.— 2._...:::32 9.: 2. 23233747. 1.42.1. .3227. 2.. Cut: A.a.__.::.a..._l<..: 22.422.222.42 75:27. 2.42:: 5.2.53.1. 2222...»-..4222. v 2.49.22.27.32 E....:.2..../\
..::..r.. >..J.2-_3V. EST; .22.. 22.1.2.2: 23.. AA..V..VHJ.._.<._.T.» 2.32 :37. .2...Z..4..v .712 . .: 2.23.2.sz «7.333.; :22» 7.23.33.21.22: 52.7.. .rl.~ .4425

 

363 .0w0306
:00: 30> .w:0_ 60>

 

36.3 :0: 6060.56 6:0: 60306:: 3.6.3 :0:
60306 :03 6:00 3:03 630: 333 .302: 6030.6 :03 6:00 .303 mmm :03m-:06w:0m-m6am
3.6.3 .w:0_ 30> 606066 36.3 6:
6:0: 43:30:: 3600 60306 :03 .603 6:00 60306 :03 .3003 5:2 6:036m-0_063=m-_632
2.0.3
6: 6:00 6m 5:03 0.06.3 :0: $603 36.3 6: 5:03 m02 0633030006302
363 .w:0_ 606066 6:00
60306 .3000 $603 36.3 :0: $603 363 6: .3000— m02 23360300000063.3004
3.6.3 .0w0306 3.6.3 .0w0306 3.6.3 .0w0306
.80: 30> .w:0_ 60> :00: 60> .w:0_ 30> :00: 60> .w:0_ 60>
606066 6:00 60306:: 606066 6:00 60306:: 606066 6:00 60306::
.53 002 .62.. as: 002 .62.. as: 00: .62.. <0: 0.000-268-0268:
0600
m 60:00:60 N 60:09:00 3 60:09:00 =0m :06300000. 60m

 

60:3— 000 06600600 OOmH<Hm 03.06 00:6 :06 06 .6 6363:000Q 0:00000 E06 ooom 6:0 moom 06 36:6 53:32 00 06660:
.833 E03600 06 :3 :060300 0030 .630 :3 600: 0:363:30. 6 60:60:09 0306 603:3 .03.C0m :06030600 00060003 0602

.030 32.50 :00: :00 000232 80 603.500 002 new 0.580 0.: 9 050.88 0008 0:3 8002830 :00 .0: 0300

18

Table 1.3. Total hectares (acres) of study sites sampled in the southern Lower Peninsula
of Michigan during the 2005 and 2006 field seasons.

 

 

 

SGWA ha (acres) Private ha (acres)
Dansville 1,921 (4,746) BEN 105 (259)
Grand River 330 (816) BOY 47 (116)
Middleville 1,841 (4,548) CAR 19 (57)
Muskrat 82 (202) CARL 12 (30)
Oak Grove 815 (2,013) COO 16 (40)
Portland 806 (1,991) F LEI 7 (18)
Rose Lake 1,526 (3,770) FLE2 32 (80)
FLO 59 (146)

GIL 16 (40)

HAY 3O (74)

KRI 32 (80)

MIL 36 (90)

PAS 16 (40)

RED 4 (10)

RBI 8 (20)

RIC 30 (73)

SEL 28 (70)

STE 16 (40)

STEW 109 (27)

VAN 16 (40)

WEL 32 (80)

WIL I6 (40)

WIN 3 (8)

WOL 61 (150)

Total 7,319 18,086 659 (1,628)
Average 1,046 2,584 28 (68)

 

19

 

 

 

 

 

Counties located in the Lansing
D sub-subsection of the Ionia subsection

of Region lecoregion where study
sites were evaluated

 

 

 

 

 

 

 

 

The Lansing sub-subsection of the :
Ionia subsection of Region 1 {(
ecoregion ,7

A

Fig. 1.1. Location of study sites in the Lansing sub-subsection of Ionia
subsection of Region 1 ecoregion within Michigan’s Lower Peninsula, 2005 and
2006. Numbers represent private land owners and SGWA surveyed in each
county. One SGWA was located in 2 different counties.

 

 

 

 

 

 

 

 

 

 

 

 

 

20

Table l
wumc
field 5:

BENI
BEND
BEN}
BEN4
BEXS
BU\l
BOY:
CARI
CAR:
CARS
CAR4
CARL
C(K)1
DAN]
DAX}
DAN:
BM
[)\\4
DAN:
Dix,
DAX'
Dix;
[Jigs
FLE;
LP:
Flffi
GM}
(ka3
GPLX
Hay
KR: '
SUD
\Ub
MID
"mi
mm
Xu-i
Hi:

I

‘51:“

Table 1.4. Site information for 82 survey sites located on 7 different SGWA and 24 different
private lands in the southern Lower Peninsula of Michigan sampled during the 2005 and 2006
field seasons.

 

 

 

Site ID County Veg * Soil “ Owners Site ID County Veg . Soil " Owners
Type Type Type Type
BEN] Clinton PIN SHB Private OAKZ Livingston UHW MHE SGWA
BEN 2 Clinton LHW SHB Private OAK3 Livingston UHW SHB SGWA
BEN3 Clinton NHW SHB Private OAK4 Livingston LHW SHB SGWA
BEN4 Clinton UHW SHB Private OAKS Livingston NHW SHB SGWA
BENS Clinton LHW HCA Private OAK6 Livingston PIN SHB SGWA
BOY] Ingham UHW MCS Private PAS] Livingston PIN MHE Private
BOY2 Ingham NHW MCS Private PASZ Livingston UHW MHE Private
CAR] Barry PIN MCS Private POR] Ionia NHW MCP SGWA
CAR2 Barry LHW MC S Private POR2 Ionia UHW MCP SGWA
CAR3 Barry UHW MC S Private POR3 Ionia LHW MCP SGWA
CAR4 Barry NHW MCS Private POR4 Ionia NHW MCP SGWA
CARL] Shiawassee LHW HCA Private PORS Ionia UHW MCP SGWA
COO] Clinton UHW SHB Private RED] Clinton PIN SHB Private
DAN ] Ingham PIN MC S SGWA REI] Eaton LHW MCP Private
DAN 10 Ingham PIN MCS SGWA RIC] Ingham LHW MCS Private
DANZ Ingham PIN MCS SGWA ROS] Clinton LHW SHB SGWA
DAN3 Ingham LHW MCS SGWA ROSIO Clinton PIN SHB SGWA
DAN4 Ingham UHW MCS SGWA ROS] 1 Clinton LHW SHB SGWA
DAN 5 Ingham NHW MC S SGWA ROSZ Clinton NHW SHB SGWA
DAN6 Ingham LHW HCA SGWA ROS3 Clinton UHW SHB SGWA
DAN 7 Ingham LHW HCA SGWA ROS4 Clinton PIN SHB SGWA
DANS Ingham LHW MC S SGWA ROSS Clinton NHW SHB SGWA
DAN9 Ingham PIN MC S SGWA ROS6 Shiawassee PIN SHB SGWA
FLE] Clinton LHW MCP Private ROS7 Shiawassee LHW SHB SGWA
FLE2 Clinton UHW MCP Private ROS8 Clinton NHW SHB SGWA
FLO] Ingham PIN MCS Private ROS9 Clinton UHW SHB SGWA
GIL] Shiawassee LHW HC A Private SEL] Clinton UHW SHB Private
GRA] Ionia LHW HCA SGWA SELZ Clinton NHW SHB Private
GRAZ Ionia LHW HCA SGWA SEL3 Clinton PIN SHB Private
HAY] Ingham UHW MCS Private STE] Clinton PIN SHB Private
KRll Shiawassee LHW HCA Private STEZ Clinton LHW SHB Private
MID] Barry NHW MCS SGWA STEW] Clinton LHW SHB Private
MID2 Barry NHW MCS SGWA VAN] Ionia UHW MCP Private
MID3 Barry LHW MCS SGWA VANZ , Ionia NHW MCP Private
MIL] Clinton NHW MCP Private WELZ Clinton NHW MCP Private
MIL4 Clinton UHW MCP Private WEL3 Clinton LHW MCP Private
MUS] Clinton LHW MCP SGWA WIL] Shiawassee LHW HCA Private
MUSZ Clinton LHW MCP SGWA WIN] Clinton NHW SHB Private
MUS3 Clinton UHW MCP SGWA WOL] Eaton UHW MCS Private
MUS4 Clinton NHW MCP SGWA WOLZ Eaton PIN MC S Private
OAK] Livingston PIN MHE SGWA WOLS Eaton NHW MC S Private

 

jCodes explained in Table 1.], " Codes explained in Table 1.2.

2]

Amine;

Albt’l’].

Alford.

Blaustt

Bldugk

8mm

BUTTQU:

 

LITERATURE CITED

Ankney, R. M. 1988. The history of the Rose Lake Wildlife Research Center. Wildlife

Division Report No. 3054. Michigan Department of Natural Resources. Lansing,
Michigan, USA.

Albert, D. A. 1995. Regional landscape ecosystems in Michigan, Minnesota, and
Wisconsin: a working map and classification. General Technical Report NC-178.
US. Department of Agriculture, Forest Service, North Central Forest Experiment
Station, St. Paul, Minnesota, USA.

Alford, R. A., and S. J. Richards. 1999. Global amphibian declines: a problem in applied
ecology. Annual Review of Ecology and Systematics 1999(30): ] 33-165.

Bean, M. J ., and D. S. Wilcove. 1997. Editorial: the private land problem. Conservation
Biology ]l(]):]-2.

Blaustein, A. P., and J. M. Kiesecker. 2002. Complexity in conservation: lessons from
the global decline in amphibian populations. Ecology Letters 52597-608.

Blaustein, A. P., Wake, D. B., and W. P. Sousa. 1994. Amphibian declines: judging
stability, persistence, and susceptibility of populations to local and global
extinctions. Conservation Biology 8(1):60-71.

Bonin, J. J ., L. DesGranges, J. Rodrique, and M. Ouellet. ]997. Anuran species richness
in agricultural landscapes of Quebec: foreseeing longterm results of road call
surveys. Pages 1-15 in D. M. Green, editor. Amphibians in decline: Canadian
studies of a global problem. Society for the Study of Amphibians and Reptiles,
St. Louis, Missouri, USA.

Burroughs, R. D. 1946. Game refuges and public hunting grounds in Michigan. The
Journal of Wildlife Management ]0(4):285-296.

22

BUNC-

11‘
Com

C mm

0.161%}

Dunnin

 

Burton, T. M., and G. E. Likens. 1975. Energy flow and nutrient cycling in salamander
populations in the Hubbard Brook Experimental Forest, New Hampshire.
Ecology 56:1068-1080.

Collins, J ., and A. Storfer. 2003. Global amphibian declines: sorting the hypotheses.
Diversity and Distributions 2003(9):89-98.

Cummins, C. P. ]989. Interaction between the effects of pH and density on growth and
development in Rana temporaria L. tadpoles. Functional Ecology 3:45-52.

Dempsey, D. 200] . Ruin and recovery. The University of Michigan Press, Ann Arbor,
Michigan, USA.

Duellman, W. E. 1999. Global distribution of amphibians: patterns, conservation, and
future challenges. Pages 1-30 in W. E. Duellman, editor. Patterns of distribution
of amphibians. The John Hopkins University Press, Baltimore, Maryland, USA.

Dunning, J. B., B. J. Danielson, and H. R. Pulliam. ]992. Ecological processes that affect
populations in complex landscapes. Oikos 65:169-175.

Dunson, W. A., and J. Connell. 1982. Specific inhibition of hatching in amphibian
embryos by low pH. Journal of Herpetology 16:314-316.

Eagle, A. C., E. M. Hay-Chmielewski, K. T. Cleveland, A. L. Derosier, M. E. Herbert,
and R. A. Rustem, eds. 2005. Michigan’s Wildlife Action Plan. Michigan
Department of Natural Resources. Lansing, Michigan, USA.

Environmental Systems Research Institute. 1999. ArcView 3.2. Redlands, California,
USA.

Environmental Systems Research Institute. 2004. ArcMap 9.]. Redlands, California,
USA.

Findlay, C. S., and J. Houlahan. 1997. Anthropogenic correlates of species richness in
southeastern Ontario wetlands. Conservation Biology 11(4):1000-]009.

23

Gordin

Guenjx'

Hecnar

Hilty. J

Houlah

Kats. L

 

Kl 63% '_

 

Knjg'r.

Kiln-r.

 

Gordinier, E. J. 1960. A place to hunt. Michigan Department of Conservation, Lansing,
Michigan, USA.

Guerry, A. D., and M. L. Hunter, Jr. 2002. Amphibian distributions in a landscape of
forests and agriculture: an examination of landscape composition and
configuration. Conservation Biology 16(3):745-754.

Hecnar, S. J. 1997. Amphibian pond communities in southwestern Ontario. Pages ]-15
in D. M. Green, editor. Amphibians in decline: Canadian studies of a global
problem. Society for the Study of Amphibians and Reptiles, St. Louis, Missouri,
USA.

Hilty, J ., and A. M. Merenlender. 2003. Studying biodiversity on private lands.
Conservation Biology ]7(1):132.

Houlahan, J. E., and C. S. Findlay. 2003. The effects of adjacent land use on wetland
amphibian species richness and community composition. Canadian Journal of
Fisheries Aquatic Sciences 60:]078-1094.

Kats, L. B., and R. P. F errer. 2003. Alien predators and amphibian declines: a review of
two decades of science and the transition to conservation. Diversity and
Distributions 9:99-110.

Kiesecker, J. M., A. R. Blaustein, and L. K. Belden. 2001. Complex causes of
amphibian population declines. Nature (London) 410:681-684.

Kiesecker, J. M., and A. R. Blaustein. 1995. Synergism between UV-B radiation and a
pathogen magnifies amphibian embryo mortality in nature. Proceedings of the
National Academy of Science 92:]1049-1105].

Knight, R. L. 1999. Editorial: Private lands: the neglected geography. Conservation
Biology ]3(2):223-224.

Knutson, M. G., J. R. Sauer, D. A. Olsen, M. J. Mossman, L. M. Hemesath, and M. J.
Lannoo. 1999. Effects of landscape composition and wetland fragmentation on

frog and toad abundance and species richness in Iowa and Wisconsin, USA.
Conservation Biology 13(6): ] 43 7- 1 446.

24

Kutk;

Latin.

Mac.j

Marc

McDi;

.\lCGar

Mats. |

 

 

Kutka, F. J. 1994. Low effects on swimming activity of Ambystoma salamander larvae.
Environmental Toxicology and Chemistry 13(11):]821-1824.

Lannoo, M. J. 1998. Status and conservation of midwestem amphibians. University of
Iowa Press, Iowa City, Iowa, USA.

MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography.
Princeton University Press, Princeton, New Jersey, USA.

Maxell, B., and G. Hokit. 1999. Amphibians and reptiles. Pages ]8-29 in G. Yoslin and
H. Youman, coordinators. Effects of recreation on Rocky Mountain wildlife: a
review for Montana. Committee on Effects of Recreation on Wildlife, the
Montana Chapter of the Wildlife Society.

McDiarmid, R. W. 1994. Diversity and natural history. Pages 5-15 in W.R. Heyer, M.
A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington DC, USA.

McGarigal, K. 2002. Landscape pattern metrics. Pages 1135-1142 in A. H. El-Shaarawi
and W. W. Piegorsch, editors. Encyclopedia of Environmentrics Volume 2.
Wiley and Sons, Sussex, England.

Metts, B. S., W. A. Hopkins, and J. P. Nestor. 2005. Interaction of insecticide with
larval density in pond-breeding salamanders (Ambystoma). Freshwater Biology
50(4): 685-696.

Michigan Department of Natural Resources. 2000. Michigan GAP state land ownership
(stewardship) digital geospatial data. Department of Natural Resources, Land and
Minerals Services Office, Lansing, Michigan, USA.

Michigan Department of Natural Resources. 2001a. The history of the organization;
landmark dates for the Department of Natural Resources. Department of Natural
Resources, Human Resources Division, Lansing, Michigan, USA.

25

P:

Pr);

Rtgt'

Michigan Department of Natural Resources. 2001b. IF MAP Southern Michigan land
cover digital geospatial data. Department of Natural Resources, Forest, Mineral,
and Fire Management Division, Lansing, Michigan, USA.

Michigan Department of Natural Resources. 2004. Wildlife Program Description.
Department of Natural Resources, Lansing, Michigan, USA. <http://www.midnr.
com/Publications/pdfs/InsideDNR/ProgDesc/newfolder/wi]dpro__2004.pdf.>.
Accessed: 4 September 2005.

Michigan Department of Natural Resources. 2005. Biodiversity Conservation Planning
Process. Department of Natural Resources, Forest, Mineral, and Fire
Management Division, Lansing, Michigan, USA.

Morrison, J. L., and S. R. Humphrey. 2001. Conservation value of private lands for
crested caracaras in Florida. Conservation Biology 15(3):675-684.

Myers, N. 2001. Protecting endangered species in the United States. Cambridge
University Press, New York, New York, USA.

Perkins, D. W., and M. L. Hunter. 2006. Effects of riparian timber management on
amphibians in Maine. Journal of Wildlife Management 70(3):657-670.

Pough, F. H. 1979. Acid precipitation and embryonic mortality of spotted salamanders,
Ambystoma maculatum. Science 192168-70.

Primack, R. B. 2004. A primer of conservation biology. Third edition. Sinauer
Associates, Inc., Sunderland, Massachusetts, USA.

Reaser, J. K., and A. Blaustein. 2005. Repercussions of global change. Pages 60-63 in
M. Lannoo, editor. Amphibian declines: the conservation status and of United
States species. University of California Press, Berkeley, California, USA.

Regosin. J. V., B. S. Windmiller, R. N. Homan, and J. M. Reed. 2005. Variation in
terrestrial habitat use by four pool-breeding amphibian species. Journal of
Wildlife Management 69(4): 148 l -1493.

26

Sauns:

Saundc

SnuflL

C (1

Saterson, K. A., N. L. Christensen, R. B. Jackson, R. A. Kramer, S. L. Pirnm, M. D.
Smith, and J. B. Wiener. 2004. Disconnects in evaluating the relative
effectiveness of conservation strategies. Conservation Biology 18(3):597-599.

Saunders, D. A., R. J. Hobbs, and C. R. Margules. 199] . Biological consequences of
ecosystem fragmentation: a review. Conservation Biology 5(1):]8—32.

Small, R. J ., J. C. Holzwart, and D. H. Rusch. 1991. Predation and hunting mortality of
ruffed grouse in Central Wisconsin. Journal of Wildlife Management 55(3):512-
520.

Spies, T. A., W. J. Ripple, and G. A. Bradshaw. 1994. Dynamics and pattern of a
managed coniferous forest landscape in Oregon. Ecological Applications 43(3):
555-568.

Stebbins, R. C., and N. W. Cohen. 1995. A natural history of amphibians. Princeton
University Press, Princeton, New Jersey, USA.

Thompson, K. F. 2004. Evaluation of partners for fish and wildlife wetland restoration
efforts in the Saginaw Bay watershed. Master’s Thesis, Michigan State
University, East Lansing, Michigan, USA.

United States Department of Agriculture. 1994. US. General Soil Map (STATSGO) for
Michigan. Natural Resources Conservation Service. Fort Worth, Texas, USA. <
http://www.fiw.nrcs.usda.gov/stat_data.html>. Accessed 10 Jan 2005.

University of New Mexico Center for Wildlife Law and R. S. Musgrave. 1998. Federal
wildlife law of the 20'“ century. Center for Wildlife Law at the Institute of Public

Law, School of Law, University of New Mexico, Albuquerque, New Mexico,
USA.

Vitt, L. J ., and J. P. Caldwell. 1990. Viewpoint: amphibians as harbingers of decay.
BioScience 40:418

Welsh, H. H., Jr., and S. Droege. 2001. A case for using Plethodontid salamanders for
monitoring biodiversity and ecosystem integrity of North American forests.
Conservation Biology 15(3):558-569.

27

Welsh. ‘.

Wilson.

I

Welsh, H. H., Jr., and L. M. Ollivier. 1998. Stream amphibians as indicators of
ecosystem stress: a case study from California’s redwoods. Ecological
Applications 8(4):] 1 18-1 132.

Wilson, E. O. 1988. The current state of biological diversity. Pages 3-20 in E. O.
Wilson, editor. Biodiversity. National Academy Press, Washington, DC, USA.

28

Cl

Kid:
ulrr:
glob

infer

€0.95
CCQS
0th
19%):
a,

and }

PF ll I

CHAPTER 2: DESCRIPTION OF THE HERPETOFAUNAL COMMUNITIES
IN STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

INTRODUCTION

During recent decades, much attention has been focused on the global decline of
herpetofauna, particularly amphibian populations (Kiesecker et al. 2001). Several factors
have been implicated in this recent decline including habitat loss and fragmentation
(Blaustein et a] 1994, Blaustein and Kiesecker 2002), introduced predators (Blaustein and
Kiesecker 2002, Kats and Ferrer 2003), pollution (Alford and Richards 1999), increased
ultraviolet B (UV-B) radiation (Alford and Richards 1999, Reaser and Blaustein 2005),
global climate change (Alford and Richards 1999, Reaser and Blaustein 2005), and
infectious disease (Alford and Richards 1999, Reaser and Blaustein 2005).

Concern about amphibian declines is most likely due to their potential role as
indicators of environmental stress (Halliday 2000, Blaustein and Kiesecker 2002, Reaser
and Blaustein 2005). Amphibians use both aquatic and terrestrial habitats, and therefore
come into contact with both aquatic and terrestrial stressors (Stebbins and Cohen 1995,
Blaustein and Kiesecker 2002). Amphibians are believed to play a critical role in
ecosystem dynamics because they are the most abundant vertebrate in many forested
ecosystems (Burton and Likens 1975). Their decline could have a substantial impact on
other organisms because it could disrupt ecosystem functioning (Blaustein and Wake
]995, Blaustein and Kiesecker 2002). A broad scale and severe population decline could
have considerable and lasting effects on ecosystems potentially resulting in adjustment
and restructuring of food webs of amphibian invertebrate prey and vertebrate predators.

The recent global amphibian decline demonstrates the need to conserve and

protect these species (Alford and Richards 1999); however, 23.4% of amphibians are

29

"Data De‘
deficient 1
nmdmc
extinct (l
protectim
be llSlL‘d

ornnern;

character

Sllccessr~
Despite
Ol‘ll’lcsc
esSé'mla
@3011.) gjk
[filmm-
beCCiUsc

”3118 cr

lhlt‘e lc
‘nll ha
(Hiitr
“ale, Ct»

.V‘."
rm'l-‘L‘x

“Data Deficient” (IUCN, Conservation International, and NatureServe 2006). Data
deficient means that the information available on particular species’ distribution,
abundance, and status is not sufficient to properly assess the potential of the species to go
extinct (Hilton-Taylor 2000). Thus 23% of all amphibians worldwide may need
protection, but the information to make informed decisions is unavailable. For species to
be listed as a species of special concern, threatened, or endangered at the state, federal,
or international level, adequate and accurate information based on life history
characteristics and/or field ecology studies is needed (Bury 2006).

Knowledge of distributions and abundances of herpetofauna is critical for the
successful management and protection of these declining organisms (Bury 2006).
Despite the need for this kind of information, there has been a decline in recent decades
of these types of studies (Bury et al. 2006, McCallum and McCallum 2006). It is
essential to determine life history traits when a species is still common and has an
ecologically functioning population (McCallum and McCallum 2006). Life history
requirements should not be determined when species are on the verge of extinction,
because environmental stressors may have influenced these traits, thereby altering the
traits consistent with healthy environments (McCallum and McCallum 2006).

The continuing loss of biological diversity (the variety of life that encompasses
three levels of biological organization: genetic, species, and ecosystem (Primack 2004))
will have an effect on our understanding of species and the function of ecosystems
(Heyer et al. 1994). Knowledge of basic life history traits, such as species distributions,
also contributes to the understanding of biological diversity (Greene 2005). Biodiversity

resides in a diverse landscape of private and public lands (Saterson et al. 2004) and

30

species «
BL’C 311M
Chdilt'll ;
adjucen
1101 cnr
necess.

11568.

OR]

her

species do not recognize legal boundaries between land ownership types (F orrnan 1995).
Because of biological diversity’s complex nature, natural resource managers are often
challenged with ways to conserve it on both protected and/or managed areas and the
adjacent matrix of multiple-use lands (deMaynadier and Hunter 1995). Therefore, it is
not enough to conduct research solely on protected or managed land, but rather it is
necessary to study biological diversity on an assortment of land ownership types and
uses.

To assess biological diversity on the species level, the distribution and abundance
of herpetofauna in the southern lower peninsula of Michigan were sampled on
Michigan’s State Game and Wildlife Areas (SGWA) as well as on privately-owned
parcels in the spring and summer of 2005 and 2006. Eighty-two sites were sampled on a
patch scale, and sites were selected based on three characteristics: 1) land ownership, 2)
vegetation type, and 3) soil type. The southern Michigan landscape has been
manipulated and transformed, much like the rest of the agricultural dominated Midwest,
and over 75% of wetland habitats in the Midwest have been lost to agricultural,
industrial, and development (Detenbeck et a]. 1999). Because herpetofauna can be
indicators of environmental quality and contribute to overall biological diversity, and
because the southern lower peninsula of Michigan landscape is a mosaic of land
ownership types and land uses, the study area was an acceptable location to research
herpetofaunal distribution and abundance, as well as the contribution of land ownership
types to the conservation of herpetofaunal diversity.

The objectives of this study were to:

I) generate a species inventory,

3]

B} in If
vs ithin a
begin to
decisions
A]

the Michig

 

2) describe herpetofaunal richness, occurrence, relative abundance, and
diversity in forested patches,

3) compare these factors by differing land ownership types, and

4) compare these factors within study years.
By investigating the herpetofaunal communities within different sampling years and
within a patchwork of SGWA in a matrix of private lands in southern Michigan, we can
begin to build a foundation of knowledge upon which to base future conservation
decisions.

All capture, handling, and marking protocols used in this study were approved by

the Michigan State University Animal Care and Use Committee (AUF# 07/03-082-00).

METHODS

Drift Fence Arrays

Drifi fence arrays made from 60 cm high aluminum flashing with pitfall and
fiinnel traps were used to capture herpetofauna (Corn 1994, Enge 1997). Drifi fences
intercept herpetofauna moving on the ground and re-direct them into a pitfall or funnel
trap. Drift fences with pitfall and funnel traps were installed in April and early May prior
to opening the traps in mid-May. They were opened for 5 consecutive nights each month
in 2005 and 4 consecutive nights each month in 2006 from May through August (Table
2.1) and were checked once daily. Three 5 m long sections of aluminum flashing were

installed in a Y arrangement and 4 pitfall traps and 6 funnel traps were placed within the

32

Tab}
oper

51-1”:
_' M.

Table 2.]. Dates drifi fence arrays, pitfall traps, funnel traps and coverboards were
opened in southern Lower Peninsula of Michigan on SGWA and private lands in summer
2005 and 2006.

 

 

Date Opened Date Closed Number of Sites
05/17/2005 05/22/2005 20
05/24/2005 05/29/2005 22
06/06/2005 06/1 ]/2005 20
06/14/2005 06/ 19/2005 22
07/ 1 ]/2005 07/16/2005 20
07/19/2005 07/24/2005 2]
08/05/2005 08/09/2005 20
08/10/2005 08/14/2005 2]
05/08/2006 05/12/2006 17
05/15/2006 05/19/2006 14
05/22/2006 05/26/2006 9
06/06/2006 06/ 16/2006 1 7
06/12/2006 06/ 16/2006 20
06/ 19/2006 06/23/2006 3
07/05/2006 07/ 09/2006 1 7
070/9/2006 07/13/2006 3
07/12/2006 07/16/2006 20
07/31/2006 08/04/2006 20
08/07/2006 08/1 ]/2006 20

 

33

array (see

frum 18.9

to allow f

below gm

placed in z

1994). Th

Traps wen
Fun

in the am}-
bOd} comp:
Screening

c
\—

diameter ll

hmn'els ar

Coverho

the dri‘
Ema}- ll
1 m 3e

mlii 3L

Sim/la; ;

 

array (see Fig.1 Enge 1997). Arrays were oriented to the north. Pitfall traps were made
from 18.9 L buckets. Holes were drilled approximately 2 cm'from the bottom of the trap
to allow for drainage of rainwater (Enge 1997). The pitfall traps were buried slightly
below ground level, allowing animals to drop into the bucket. Moistened sponges were
placed in all pitfall traps to prevent desiccation of captured animals (Greenberg et al.
1994). The sponges were remoistened as needed (Enge 1997, Richter and Seigel 2002).
Traps were closed by placing lids over the buckets.

Funnel traps were placed at the midpoint of each wing of the aluminum flashing
in the array (Greenberg and Tanner 2005). Funnel traps were double entry with the main
body comprised of aluminum window screening and the funnels of flexible fiberglass
screening. Traps had 20 cm openings at both ends with funnel openings of 5 cm in
diameter (Corn 1994). When not in use, fimnels traps were closed by inverting the
funnels and clipping them shut.

Coverboards

Coverboards (Fellers and Drost 1994, Davis 1997) were placed within. 1m from
the drift fence array in the four cardinal directions at least one week prior to drift-fence
array installation. Coverboards were made of untreated birch plywood and cut into 1 m x
l m sections and were placed on bare ground. Coverboards are designed to provide
moist, cool refuge for herpetofauna (Houze and Chandler 2002) and create a microhabitat
similar to a downed log. Coverboards were checked every day that the drift fence arrays

were open.

34

Area Time-l
Area
Cnrmp and
30")6: 23 \1
l1 August.
3"- m area (
time-const
ATC sum
directions.
and \Eggr
3133 Was I
alld Start
the US. l
CidSSlfigd

Anuran

Area Time-Constrained Surveys

Area time-constrained surveys (ATC surveys) (Campbell and Christrnan 1982,
Crump and Scott 1994) were conducted on each site once a month from May to August in
2006: 23 May - 3] May, 1 June— 2 June, 19 June — 27 June, 19 July — 26 July, 1 August —
12 August; and in June and July in 2005: 22 June — 30 June, 22 July — 31 July. A 2 m x
37 in area (the same total area as the drift fence array) was delineated to conduct the area
time-constrained surveys, and this same area was used for surveys throughout the season.
ATC survey areas were selected near the drift fence array in one of the four cardinal
directions, starting east of the array. When a 2 m x 37 m area fit in the designated soil
and vegetation type, it was georeferenced with a GPS unit and flagged. This designated
area was hand-searched for a period of 20 minutes by overturning downed logs and rocks
and searching through leaf litter. Decay classes of the logs were recorded according to
the US. Forest Service’s FIA Field Methods for Phase 3 Measurements, five decay
classifications (USDA Forest Service 2004) (Table 2.2).
Anuran Call Surveys

Frog and toad song counts were conducted between April and August in 2005 and
2006 using a method similar to the North American Amphibian Monitoring Program
(NAAMP) (Weir and Mossman 2005). Frog call surveys differed from NAAMP protocol
by the number and location of survey points. NAAMP protocol requires 10 listening
locations along a road near potential amphibian breeding habitat. The number of
listening locations during my surveys varied, as well as the location where the calling
survey was conducted. During the 2005 season, surveys were conducted in different

habitat types (lowland hardwoods, northern hardwoods, upland hardwoods, and pine) at

35

Table 2.2.
area time-c
and 13er

 

Class

Ll

 

14

IS

 

 

Table 2.2. Five decay classifications of downed logs (US. Forest Service 2004) used in
area time-constrained surveys in the southern Lower Peninsula of Michigan on SGWA
and private lands in summer 2005 and 2006.

 

 

Class Description

L1 Bark intact, twigs present. Texture is intact. Wood is original in color.
Log is elevated on supported points above ground.

L2 Bark intact, twigs absent. Texture is intact to partially soft. Wood is
original in color. Log elevated on support points but sagging slightly.

L3 Trace of bark. Twigs are absent and texture is hard large pieces. Color of
wood is original to faded. Log is sagging near ground.

L4 Bark and twigs are absent. Texture of wood is small, soft, blocky pieces.
Wood is light brown to faded brown or yellowish. All of the log is on the
ground.

L5 Bark and twigs are absent. Texture of wood is soft and powdery. Color of

wood is faded to light yellow or gray. The diameter of the log is
attainable. Wood and log debris is not spread out in a flat manner. If a
diameter is not attainable, then it is not considered a log, but a pile of
debris.

 

36

'1!

'l;
H

'IJ

the center of the drift fence array; however in 2006, surveys were conducted from the
edge of the water body in closest proximity to the drift fence array. All sites were
surveyed for five minutes one night in May, June, and July, and 33 of the 82 sites were
surveyed one night in April. Surveys began 30 minutes after sunset and were completed
by midnight. They were conducted if the air temperature was greater than the minimum
allowable temperature for each sampling period (Sargent 2000) (Table 2.3) and during
little or no wind (at a level three or less on the Beaufort scale) (Berry et al. 1945) (Table
2.4). Surveys conducted outside of the Michigan Frog and Toad Survey periods (July)
used the minimum air temperature for the early summer time period. The intensity of
calling males were rated 0 through 3, with 0 = no individuals calling; ] = few individuals,
but with non-overlapping calls; 2 = overlapping of calls, but individuals can be identified;
3 = full chorus with indistinguishable individual calls (Weir and Mossman 2005).
Mark-Recapture Techniques

All animals captured by pitfall traps, funnel traps, cover boards, and area time-
constrained searches were processed before being released. Herpetofauna were identified
to species, sexed (when possible), and marked. All amphibians were measured (snout to
vent length, mm); and northern leopard frog (Rana pipiens), American toad (Bufo

americanus), and green frog (Rana clamitans) with a snout to vent length measurement >
50 mm were PIT tagged (passive integrated transponders; AVID®). Salamanders and all

other Anurans were marked by toe-clipping except treefrogs and relatives, but not for
individual recognition. Measurements and weights for Anurans and salamanders are
listed in Appendices 2.1 and 2.2, respectively. Snakes and turtles were measured and

marked for individual recognition. All snakes were marked by clipping half of one

37

Table 2.3.
Peninsula t
protocol (S

 

Surrey Per

 

E851) Spfll
late Spring

Earl} Sunrr

 

Table 2.3. Survey periods for conducting fi'og call surveys in the southern Lower
Peninsula of Michigan, using Michigan Department of Natural Resources call survey
protocol (Sargent 2000).

 

 

Survey Periods Range of Dates Minimum Air Temperature
Early Spring 25 March — 30 April 7.2 C (45 F)
Late Spring 1 May — 31 May 12.8 C (55 F)
Early Summer 1 June — 30 June 18.3 C (65 F)

 

38

Table 2.4. Windspeed classifications used to determine if frog call surveys should be
conducted based on the Beaufort scale. Surveys were not conducted if windspeeds were
>3 on the Beaufort scale (Berry et al. 1934).

 

 

Scale Description

0 Calm, (<1.6 kph) smoke rises vertically

1 Light Air (1.6-4.8 kph) smoke drifts, weather vane inactive

2 Light Breeze (6.4-1 1.3 kph) leaves rustle, can feel wind on face

3 Gentle Breeze (12.9-19.3 kph) leaves and twigs move around small flag
4 Moderate Breeze (20.9-28 kph) moves thin branches, raises loose papers
5 Fresh Breeze (30.6 kph or greater) small trees begin to sway

 

39

ventral scale (Brown and Parker 1976) and turtles were marked by notching marginal
scutes (Cagle 1939). Measurements, weights, and sexes are listed for individual snakes
and turtles in Appendices 2.3 and 2.4, respectively. Individuals were releaSed at least 5
m fiom the point of capture on the opposite side of drift fence to minimize the probability
of immediate recapture.
Data Analysis

Data were analyzed in three different ways to meet each of the study objectives.
First, I examined the data across all sites by pooling both years of data collection to
generate a species inventory and to describe herpetofaunal richness, occurrence, relative
abundance, and diversity for the study region. To determine how land ownership could
affect these factors, data from SGWA were pooled and compared to pooled data from
private lands. Last, because of potential annual affects of temperature and precipitation
on herpetofaunal, I also examined species richness, occurrence, abundance, and diversity
data within years regardless of land ownership type.

Species richness is defined as the number of species present in a specified area
(Bolen and Robinson 1995). Total species richness was calculated by tallying the total
number of species detected for each ownership type (i.e., SGWA or private) using all
trapping methods as well as incidental sightings. Mean species richness was calculated
by summing the total number of species trapped at each survey site and dividing by the
number of trap sites.

Species frequency of occurrence is a count of the number of sites out of the total
number of sites sampled where a species was observed (Lancia et al. 1994). Species

frequency of occurrence was calculated by tallying the total number of sites where

40

Re

{1ft

Ill:

Re

uh

tr h

Ian

BIT.

particular species were detected using all trapping methods as well as incidental
sightings.

Species abundance refers to the number of individual animals (Lancia et al. 1994)
and relative species abundance adjusts for differences in raw abundances among sites.
Relative species abundance was calculated by standardizing array captures by trap nights
and standardizing area time-constrained captures by number of surveys and combining
them. Relative species abundance only included ATC captures from June and July.
Recaptures of individuals were excluded when evaluating species abundance.

Species diversity considers the species richness and relative abundance in a
specified area (Bolen and Robinson 1995). Species diversity for each survey site was
calculated using Shannon’s Diversity Index (Shannon 1949), as well as the Simpson

Index of Diversity. Shannon’s Diversity Index was calculated as follows:
H' = - 2 (1):) (ln 1):)
where p,- denotes the proportion of the sample belonging to the ith taxon (Shannon 1949).

There is no definitive range for the Shannon Diversity Index. The Simpson Index of

Diversity (Simpson 1949) was calculated as follows:

D = 1 - z (pf)
where p,- denotes the proportion of the sample belonging to the ith taxon. The index
ranges from 0 to 1, with values near 1 indicating a highly diverse or heterogeneous site
and values near 0 corresponding to a more homogeneous site.

Herpetofaunal community similarity among survey sites was calculated using the

Bray-Curtis coefficient (Bray and Curtis 1957), which standardized the City-block

41

(Manhattan) metric. It has a range of 0 (similar) to 1 (dissimilar), where distance
measures (proximity of sites in species space) are expressed as proportions of the
maximum possible distances between sites (Krebs 1999). The Bray-Curtis is a
proportional distance coefficient, which is the most commonly used distance metric in
species abundance data, and is often used with community ecology studies (McGarigal et
al. 2000). The Bray-Curtis coefficient was calculated as follows:

_ 22m1n(xij,xhj)
Bjk —1—
loci]. +xhj)

 

where min is the smaller of the two values (Mche and Grace 2002), xi} is the predicted
probability of species j at site I, and xhj is the predicted probability of species h at site j.
To define this as a measure of similarity, the complement of the Bray-Curtis measure was
used: (l-B) (Wolda 1981). Two sites were removed before calculating the resemblance
matrix because they both did not have any captures; therefore, I was unable to generate
an index for these two sites. All analyses were done using R 2.4.0 (R Development Core
Team 2006).
RESULTS

Species Inventory

Twenty-two species of amphibians and 29 species of reptiles have been
documented to occur in the Lower Peninsula of Michigan (Conant and Collins 1991,
Harding 1997) (Table 2.5). In the Lansing sub-subsection of the Ionia subsection of
Region One, 1] species of Anurans (frogs and toads), 7 species of Caudates
(salamanders), 9 species of Testudines (turtles and tortoises), and 17 species of

Squamates (lizards, Amphisbaenians, and snakes) are expected to occur (44 total;

42

9:. E z: 2.4.5 7.: .::1:_.J;£.~3 EDI—JED.— ._..._$C ~ 0.: :._ -2743: C. 7.9.2.1223... 7.3:.ng 7:... 7.2.....£.:7:.:...\2 1.9.9.317. Vim. 2.72:.

 

 

88:82 0033 38:00 0:0 805.82 00.0.: 0003 88000.8 8.800905 000m0>8m
X 0058022 0030; 88003 0:0 808005 008: 008000 883% 08805 000_0>8m
X 88:82 0033 008: 8008 :08800 0500000 82080505 00280000x8v~
X 0230005 .0008: 580005 008800 02.220800. 0.008000 00200325
X 0230005 mow 6000—08 0.50800. 88% 002:5.”
X 0230005 awob 08000. 505.0: 0.208% 88% 002:0m
X 0230005 .0508 0003 0000380 020% 002:0M
X 0230005 .308 :008 88.880 88% 002:0M
X 0230005 308:3 028000080 050% 0028.3
X 0230005. 0.000000 98.80 00.8030 8.800000% 00233
X 0230005 050.0 80:0 800003 80.2088 88008.: 80000=0£ 002—32
X 0230005 500008 .08 0.0000 800000000 003 00233
X 023805 30.0008 >08 800000 838.80; 8.0% 00233
X 85882 00304 805005 m2.“ 0002.8 805.5: 8830.8 8.0V 0023:
88:82 00304 E00003 0:0 808005 000. 9.830.: t8§0\0\:m 00200.05
X 0230005 *0000 580:2. 800000 82002080 0.03 00200.05
X 0230005 00288800 008-88 888000 883000.80m 0028000503
X 0230005 “0000008800 000—000000 80080 2000803 0028000805
X 0230005 0100:0888 00w: E00000 88.2%: 0880380. 00200890302
008882 0033 8000038005 000880000 00308 53000.0 usemésw 002008903840.
88882 00304 8000008005 000008800 0050088080 8:20.000 0588.080. 000008009380.
X 0230005 0100008800 003000080 08088 08803800 00200890380.
X 0230005 00:08.88 003000 5803008 0580.350. 002008030380.
X 0230005 30: 80.000 8000.00.83 0380038002 00200008805
«08:82 00304 8000038005 :080 00008 803.03 .8280: 800.588 000.5 0008005
X 0230.05 3000008 8003008 0.55.002 00208.5
803.805
00.2~ 005 885805 0802 008800 080 Z 02:85 30:0,.“

 

.8000Z030 _080208 880808 0.3200 00808800 080000 0:0 0000000080
0:0 .0008 30:3 .0008 20.20 .23 003000 $0.00 00:00 200 >9 0080000 :0 0003 008 0:0 moom 008800 8 00000 >23 08
8 003030 003 002005... .9005 880082 0:0 200: 08:00 0:0 80:00 808 00880000 0003 00800 0020005 .000 080

08 8 00 :03 00 805882 m0 0308:; 00304 05 8 .8000 8 008080000 00:80.— 0:0 0:038:80 mo 0020005 .m.m 030B

43

 

00000000000

 

X 000882 0030.0 0w00000008 0000000 000000000 0800080 0000.005 002008>
X 0030002 0033 02000 00000-w00 0000000 00030003 00000000: 00200800
X 0230.05 08000 002000-080 00000300 000000000 0.88005 00200800
X 0230005 0000000 0:00 0000000 880800.20 580000.000 030.00.080.00 00200800
0080002 0030.0 0000000 08000 0.0.0 008000 .0088 080.000.0000. 00200800
X 008.0002 030.0 00000005 02000 00.. _000000 00000000 080.000.0002 00200800
X 008022 0033 00000005 000 0000003 000000 080 000x0008000000 00080 0 00200800
X 0008002 0033 000.0005 000 0000003 00000 0.080 000.0000 0000000000 0000000000 00800 00200800
X 0230005 02000 00000 80080 0.00000: 0000000006 00200800
00800080008000
X 0230005 *02000 000000-000 0000000: 00800000008000 0.20.005 00200800
X 0230005 02000 03000 0000000 0.20005 00200800
X 000882 00300 0.00.000 00000.. 00000000 0000000000000 000.2000 0.030080: 00200800
X 000.882 0030..— 0000000 0.02000 00000w 0.0000m .2008 080.00.00.00 00200800
X 0230005 u0000000 00.000 0000000 0.8030 0.0030 001000.000 00200800
0008002 00301— 0008005 08000 0.0000002 00000000 030.0080 00200800
X 008022 0033 0000003 000 000.0005 000000 00000 808080300 000002 00200800
X 0080002 0030.0 000000-505 02000 0003 0050000008 0000000: 00000800000000 0.000002 00200800
X 0230005 0.000000 00003 00000000 000080 000080 0.000002 00200800
X 0230005 00:20 0087030 0000.0000x0000000i 0020805
X 0008002 .0300 0:000 =000¢00 0.000 000.08% 0000050. 002008800.
X 0080002 033 000.0005 00000 0000000.. 0000000 88000 0080000000 0020.080
X 0230005 n000000 000800 0003 08000080 002085
X 00.0.0022 0030.0 0000003 000 0005005 0:000 008 008800 003%00M00M 008008.00 0020080
X 0230005 0.0008 0080005 00.80003 000000AEN 00238.0
X 0080002 0033 000003 000 00000005 0,0008 0.00 0000000 00.80000 00.80000 0000.05.00 0020080
0000000005
00000 >005 0280.805 0802 008800 0802 0.0000005 b08000

 

.0008 .00. 2000

44

Hardin
familic
indnid

Species

study ("
0.23 SE
pm are .
(“Bragg
richness
WT Sam;
lands (T;
(Emzfldo
(CM dr.
bcHicd Sr
Pm are 13‘
(.ngdi-a A.
’nbngulu
\\

heme?”

from 0 h

”Chum

Harding 1997). Twenty-two different species of amphibian and reptiles from 8 different
families were observed during the surveys (Table 2.5). Specific methods used to observe
individual species ranged from one technique to all six different methods (Table 2.6).
Species Richness

Twenty-two species were observed on both SGWA and private lands during the
study (Table 2.5). Overall species richness among the 82 sites ranged from O to 9 (4.8 i
0.23 SE). Nineteen species were observed on SGWA, and 17 species were observed on
private lands (Table 2.7). At SGWA sites, species richness ranged from 1 to 9, with an
average of 4.6 species (:1: 0.36 SE) captured per sampling site. At private sites, species
richness ranged fiom zero to nine, with an average of 4.4 species (i 0.28 SE) captured
per sampling site. Mean species richness was greatest in July for both SGWA and private
lands (Table 2.8). Five species were found solely on SGWA: Blanding’s turtle
(Embydoidea blandingii), painted turtle (Chrysemys picta), common snapping turtle
(Chelydra serpentina), Butler’s garter snake ( T hamnophis butleri), and northern red-
bellied snake (Thamnophis sauritus septentrionalis) and 3 species were only found on
private land: eastern box turtle (Terrapene carolina carolina), northern water snake
(Nerodia sipedon sipedon), and eastern milk snake (Lampropeltis triangulum
triangulum).

When examining a potential year effect, I saw differences in species richness
between 2005 and 2006. Overall species richness among the 42 sites in 2005 ranged
from 0 to 9 (5.3 :i: 0.32 SE) with 22 species observed (Table 2.9). Overall species

richness among the 40 sites in 2006 ranged from 0 to 9 (4.2 i 0.32 SE) with 15 species

45

 

 

X 32 8:25 .833 Eonto:
$2 X 2:8 @853
X $2 255 flwcmvcflm
X $2 2:3 xon 8830
X $2 X 0E3 wcammcm 5888
X X X X X X wow 2302 805.5:
X X X X X X mob coo?
X X X X X wofi 5on
x 85:5
X X X X X X common wctam
X X X wofi 35% £283
X X X X X weave: haw 88mm»
X X X X X X 38 5858.4 5830
.32 X X X X covgfiafim 9.66362
32 X X HousmEmBm how: 88me
$2 X X X X BecaESmm 3:25-095
nosetomno hogm .1635
35205 :50 wen UH< _Eti _oSSm “6.80%?on 88on

 

.3033 Eon coom can moom 2: macaw
23322 mo Bnmficom .833 50558 2: E 86on 3:339: 83% 8 wow: mwofiofi >955 can mosvmqfioa 8395 .o.m 2an

46

.up ~: 3.4

.C .rl 371.;

Lou—wow _uocwmwaOo Demaamouw H UH<*

 

 

32 X 88.8 VEE 5888

X $2 .822 833

32 X X 8.28 35362 Boston

«:2 X X 835 coast F855:

<\Z X 8:28 88% P835

X <\Z X X X 3.58 89% F883

5:83.830 anagram 33.8w

38385 zoo moi UH< 2835 855...— Eoo98>o0 88on

 

is 8.235

47

Table 2
the sou‘
capture
cox erb<

blue-5p
eastern

red-haul
eastern .
eastern .
WESICI'n
Spring p
bUllfrth
136m frr
“00d in
nonhcm
mmmOn
mm] b
Blanding
Painted 1'.
“(inhem

eaSicm g-

 

Table 2.7. Overall herpetofaunal species richness on SGWA and private lands in
the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drifi fence arrays coupled with pitfall traps, fimnel traps, and
coverboards and area-time constrained survey including incidental observations.

 

Species SGWA Private

 

X
X

blue-spotted salamander
eastern tiger salamander
red-backed salamander
eastern American toad
eastern gray treefro g
western chorus flog
spring peeper

bullfrog

green frog

wood frog

XXXXXXXXXX

northern leopard fio g

XXXXXXXXXXX

common snapping turtle

X

eastern box turtle

X

Blanding’s turtle

X

painted turtle
northern water snake X
eastern garter snake
Butler’s garter snake
northern ribbon snake
northern red-bellied snake

blue racer

XXXXX
X

eastern milk snake X

 

48

Table 2.8. Mean (SE) herpetofaunal species richness of SGWA and private lands
by month in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drifi fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

Month SGWA Private
(n = 41) (n = 41)
May 1.2 (0.16) 1.4 (0.18)
June 2.1 (0.23) 1.2 (0.21)
July 3.0 (0.33) 2.8 (0.26)
August 2.5 (0.21) 2.6 (0.23)

 

49

C00

”(If
llor
blu,

e351

Table 2.9. Overall herpetofaunal species richness by year in the southern Lower
Peninsula of Michigan in summer 2005 and 2006 based on captures by drifi fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area—time
constrained survey including incidental observations.

 

é

Species 2005

 

X

blue-spotted salamander
eastern tiger salamander
red-backed salamander
eastern American toad
eastern gray treefro g
western chorus frog
spring peeper

bullfi'og

green frog

wood frog

northern leopard frog

XXXXXXXXXXXX

common snapping turtle
eastern box turtle
Blanding’s turtle

painted turtle

northern water snake
eastern garter snake
Butler’s garter snake
northern ribbon snake
northern red-bellied snake

blue racer

XXXXXXXXXXXXXXXXXXXXX

eastern milk snake

_

50

observed (Table 2.9). Seven species observed during the field season in 2005
were not observed in 2006, and all were reptiles: Blanding’s turtle, painted turtle,
common snapping turtle, Butler’s garter snake, northern water snake, eastern milk
snake and northern red-bellied snake.

Species Frequency of Occurrence

The eastern American toad was the most ubiquitous species, occurring at
71 out of 82 sampling sites (87%) (Figure 2.1). The next most frequently
occurring species was the wood frog, observed in 69 out of 82 sites (84%),
followed by the green frog, observed in 58 out of 82 sites (71%) (Figure 2.1).
Nine species occurred at less than 15 of the sampling sites (Figure 2.2) across the
study area. The least frequent species included northern water snake, eastern milk
snake, Butler’s garter snake, painted turtle, eastern box turtle, and Blanding’s
turtle; each occurred in only one out of 82 sites (<2%).

When evaluated by land ownership type, I saw a similar pattern of species
occurrence: eastern American toad occurred most frequently in SGWA (34 out of
41 sites (83%)) and private lands (37 out of 41 sites (90%)), followed by wood
frogs in SGWA (33 out of 41 sites (80%)) and private lands (36 out of 41 private
sites (88%)) (Figure 2.3).

When examining the year effect, I again saw a similar pattern of species
occurrence. The eastern American toad occurred most frequently in 2005 (39 out
of 42 sites (93%)), followed by wood frogs and green frogs (37 out of 42 sites

(88%)) (Figure 2.4). In 2006, both eastern American toads and wood frogs were

51

.mcozotomno 3:822: @622: @833 ooEmbmcoo 25388
was mocmon8>oo was .38: ESE .mmmb =m§a :23 83:8 $8.8 828% are c3 8:598 no woman ooom use moom _oEEsm
E 5wE22 mo «Sitcom 8304 E8538 2: E 8%. @2958 mm L8 oocotzooo 88on 3550898: :m8>O ._.m .mE

 

88on
.000 ca 09 l9
by 0 Q 6 9.-
9 o 0 o 06 o a tea 0%
$9 onv vobuv 9&0 9&0 96% 9 $09 00% 96% 1091 «9.0% Ova) mambonvoIIOve 20%.
J. v / w v a e I
vs befeovvoezoo re A a
e o. v o v e V o oo o» o
e o, 9’9 9o». «94% 91. ea. 0% 90 av cos/a. a on. o» a... «9% A... e». or
.160/96091991913.04»...o.e¢§voi.o!e/e/e/
so vvvoo.+§vooe:oaveeoflavo¢
9% a.» 9% 9% 9% 9% 9% #9 «0; v9; 1%. o 90 9 .999» W 9 a; oo 90% 0%
9.7 v9 9.7 a? 9.7 9.7 9.7 all 80 e/ o/ 9009 990 900 900 990 09 vs v9
to
OF
ON
Yon N
n
w
M
ov J
0
I.
S
on w
.oo
.65
8

 

 

52

62:88 .88on 85 8% mo 8555

585 86 $8858 8:: 5388 85 use ”Egooo 888% 85 88 mo 895:: 52555 85 $8838 85— 359%
Eotop 2: ”3888 888% 85 85 go .8555 5:5on 85 38.852 85— “8.8% no“ at. .8858 358828
853-88 28 mo8o€o>oo 58 .88“ 85.3 .38“ :55.” :53 33:8 835 828.“ are 3 8538 no woman coom one
moom 8558 5 53:52 mo 5358a 83o.._ F8538 85 5 oocogooo 888% 32288888: mo 5838““ .N.~ .wE

.88on mo 8:5: Z

©_. 3. Nr or w o V N o

 

 

 

l
C)
0')

sans JO JaqlunN

 

 

 

53

.0833 00:85:00 055.88
58 8800830 58 .38: 8:03 .38.: zaasfi £88 0080 $8 50¢ 880 00380 :0 0080 coom 08 moom 8553
5 58552 90 530.5me 8304 08608 05 5 0.58— 0838 E8 <>>Om :0 000008000 880% Ecsflouoacom .m.m .wE

      

880.5
a o 9
o 06 a wee 09¢ 91%. 96/
o t .
99.0 “.047 010 v0! 009 16¢ 000
v a; v o .7 9o! 6.
9 VJ 6V ”6 %b .00 v0 09
v a o w a
a? 0990 $0! v09 re 040 ea. ole 99
oo 0 so 90 9.. 6% %0 age
900 «as «J m. o)
90 99 0V0 §0 09 #90 V90 V90

283.8 n:
, <30m I

sang Jo Jaqum N

 

 

 

 

CV

54

0000003030 00002000 @6205 9033 0000000000
005-0000 000 00830300 000 .300 #0003 £000 _305 00>» 003000 @3000 00000 $00 3 0003000 00 00mg ooom
000 moon 0000050 5 gwfioaz mo 2000003 00304 E05000 05 E 000% 3 0000000000 860% _0000m000EOI .vN .mE

8.0on
owooo 9% 9 av we» 9/9
a o 0 06 9 o b a) ‘9
e 6v 96 #10 me.» be 06¢ )v 00. 1&0 010 e9 0v 00
9%.. 09v 090 a! vs) vs? 99 2009 00 v00, 9:0 6% Lo.» 9% 9%.. 069
s 0: 9 0 0 9 9 f V 9
a $000fi%%0%% %u%%%%%
00 o. 9/ 9 av 2v a. lo 9% 9 i a m. «a o. a? . 6 x x /
II. 0/ 9. o 9 a a 9 o 0. O o 0 § v I o. A. O 0
.7 o o o v v v o f .1 so 0 o a d 0 av 0o ,0 Q Q
9 0.. 0.. ea 0.. 0o 0; 91 o; o: o o o as .. a 60 60 oo
o90.4%..“a¢oaovovovoovoveolooo
o o a a a! o/ s/ 0/ 9o 0o 9e «a so 00 o 99 v9 90

N

n

w

88. E , m
88 I m.
s

m

 

 

 

mv

 

55

observed most frequently (32 out of 40 sites (80%)), followed by green frogs (21 out of
40 sites (53%) (Figure 2.4).

When considering data from the fi'og call surveys, frequency of occurrence for
Anuran species differed dramatically. Overall, the most frequently heard species of
Anurans were gray treefrogs followed by green fi'ogs (Figure 2.5). Bullfrogs and wood
frogs were the most infrequently heard species during the call surveys (Figure 2.5). In
both 2005 and 2006, the most fi'equently heard species of Anurans were gray treefi'ogs
followed by green frogs and spring peepers (Figure 2.6). In 2005, the most infrequently
heard species were wood fi'ogs and bull fiogs, and in 2006 the most infrequently heard
species were bullfrogs and western chorus frogs (Figure 2.6).

Species Abundance

A total of 4,410 individuals were captured during the 2005 and 2006 seasons,
resulting in 4,379 individual amphibians and 31 individual reptiles. Wood frogs were
captured in greatest numbers and comprised 53% of all captures (2,342 individuals
captured), followed by the eastern American toad (870 individuals captured (20%)) and
green frog (560 individuals captured (13%)) (Table 2.10). Captures of wood frogs and
green frogs were greatest in July (1,302 and 298 individuals, respectively) accounting for
56% of all wood frogs captured and 53% of all green frogs captured (Table 2.11). The
most captures of eastern American toads occurred in August (339 individuals) accounting
for 39% of captures (Table 2.11).

Species capture abundances on land ownership types varied considerably from

month to month. On SGWA, captures of wood frogs were the greatest in July (Figure

56

.3303 :00 wot
00 00000 coom 000 moon 0000800 00 00wfioaz 00 20000000 0033 0.005000 05 E 000000000 0050< 00 30000000 .m.~ .wE

00.025

.38
no... 9:300. no...— 03000 wok—00... 03:005.
0.35.3: MP: 0003 M0...— 0000» wet—=5 02—02— 953 583.: Era 5823 58:8

 

Vow

row

row

roc

rom

sans :0 non-nu

 

Tow

 

.om

 

Tom

 

 

 

57

000300 :00 mob 00 00000 ooom
000 Sam 008003 E 00manz 00 23.00000 .0304 0.00038 05 00 000% .3 000000000 000:0».~ 00 000000000 .o.m .wE
m0_000m

000.
we: 0.3002 000000 M0...— 000000 00:00.: 00020:;
0.5500: M0...— 0003 M0,... 0000M went—=0 u0t0m 00800.: 00..“ 0.5300 0.53.00

 

 

ON

88 n_,
88 I0

 

ymN

says )0 JaqlunN

 

 

 

 

 

 

 

58

Table 2.10. List of herpetofaunal species captured and their combined capture abundance
in the southern Lower Peninsula of Michigan in summer 2005 and 2006 fiom drift fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area-time constrained
surveys.

 

 

Common Name Total Captured Number of Sites
where Captured
blue-spotted salamander 57 22
eastern tiger salamander 87 5
red-backed salamander 67 22
eastern American toad 870 71
eastern gray treefrog 38 20
western chorus frog ll 8
spring peeper 215 42
green frog 560 58
wood frog 2,342 69
northern leopard frog 132 28
common snapping turtle 3 2
painted turtle l l
eastern garter snake l6 l4
Butler’s garter snake l l
northern ribbon snake 7 4
northern red-bellied snake 2 2

eastern milk snake

 

59

Table 2.11. Total herpetofaunal species captured and their monthly capture abundance in
the southern Lower Peninsula of Michigan in summer 2005 and 2006 from drift fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area-time constrained
surveys.

 

 

Common Name May June July August
blue-spotted salamander 28 7 15 7
eastern tiger salamander 0 O 83 4
red-backed salamander 23 16 23 5
eastern American toad 64 112 335 339
eastern gray treefrog 9 3 12 14
western chorus frog 0 2 6 3
spring peeper 9 26 132 48
green frog 10 117 298 135
wood frog 134 230 1302 676
northern leopard frog 1 4 85 42
common snapping turtle l 2 0 O
painted turtle 0 O 1 0
eastern garter snake 0 3 7 6
Butler’s garter snake O 0 1 0
northem ribbon snake 6 0 0 l
northern red-bellied snake l O l 0
eastern milk snake 0 0 l 0

 

60

2.7). However, on private lands, captures of wood frogs were the greatest in August
(Figure 2.8). Captures of eastern American toads were greatest in August for private
lands and in July for SGWA (Figures 2.7 and 2.8). Captures of green frogs were greatest
on both land ownership types in June (Figure 2.7 and 2.8). On SGWA, captures of blue-
spotted salamanders were high in both May and July, while they peaked on private lands
in May (Figure 2.7 and 2.8). Eastern tiger salamanders were primarily captured on
SGWA in July (83 out of 87 captures) (Figure 2.7).

Total captures in 2005 were greater than in 2006. A total of 2,530 individuals
were captured in 2005 resulting in 2,507 individual amphibians and 23 individual
reptiles. A total of 1,880 individuals were captured in 2006, resulting in 1,872 individual
amphibians and 8 individual reptiles. As with the overall captures, wood frogs were
captured in the greatest numbers in both 2005 and 2006. In 2005, wood frogs made up
45% of all captures (1,145 individual captured) followed by green frog (493 individuals
captured (20%)), and eastern American toad (430 individuals captured (17%)) (Table
2.12). In 2006, wood frogs made up 64% of all captures (1,880 individuals captured)
followed by eastern American toad (440 individuals captured (23%)), and green frog (67
individuals captured (4%)) (Table 2.12).

When eliminating trap data influenced by temporary individual trap closures due
to flooding and differences in trap nights in a given year, there was a standardized catch
per array effort of 263 animals per 100 array nights (when pooling years). When
examining within years, there was a standardized catch per array effort of 285 animals
per 100 array nights in 2005, and a standardized catch per array effort of 240 animals per

100 array nights in 2006. In determining relative abundance, only area-time constrained

61

000300 0000000000 050.0000
000 0000000030 000 .0000 0000.0. .0000 20,300 00000 00000 £00 00000 0000 008000 00 00000 ooom 000 moom 000E030
E 00wEBE .30 00000000 003010 0000050 08. E 5008 .3 <me 00 0000000 0000000000000 00 0000000 00000. EN. .wE

 

 

 

 

 

 

02000w
e
o a o.
10. 00 v o I
9 0 Q o 09 9 [J o
a eve #0 09 009 0 be cos 91.. 9% 0% ob
.. . ..... .7, .. o. .9 o. a. a. 0...
yo; be e o. 09 0 00.. 00 9.. o b no «9 no
0 o9 94 no he %+ V? oo/ o. a on» 9% e: 0.. 00
0% 99/ o? 9 91¢ a) Q. av ow. We 99 . o9 f 00 Oi Oi o!
0.7 ..o oo 5 2 o a e no as» an. a» o. o0 % .0
90 00 06 00 ob owe) “9 a?» a? a». v 09 099 ea) 09 06 be
$$e§$$e9§§§vo§soavevavo
0 0 . 0 0 _ _IN 0 I. ,I0 0 1 O
CON
yoov
I0
0
4000 N
002. 0 m
255 8 m
A: a
in .89 w
000w=<fl .A
W
.82 m
M.
.oovv
vooww

 

 

 

cow F

62

.3038 005000000 080-0000
000 0000000950 000 .0000 _0003 .0000 =0§0 .0000 0000.0 «00 0000.3 0000 000000 00 00000 coon 000 moom 0000000 E
000000002 00 03000000 0304 0.005000 05 E 5000 .3 0003 000200 00 0000000 0000000000000 00 0000000 _0000. .w.m .wE

00.000w
o
0qu 6 ea 9 av 09» 0/9
a 60... 00 00 09 666 v96 009 9100 one 0091 v...»
4.. a. «so. a 91.. oo o,» «a, z a. 99 lo
9..» 9» «o so... 9% a 9% 00 0.. ao e.» a... v... a.»
a 9’. o e rm. 0 s/ 9 a 90 a w r 0
0 // o, v e v a so . e. r v e e e
I a]. 0 JV I I 9!. V 47 v §¢ 0 I. 9/ AV 6/
# e vs a e e o o a v 1 00 av av av
em 0% 0.. ea 6% in I 0.. on 09 9? we see 6 o o e
0.0.40.0.00,259....»«.09090100900
9 e, o a a 0/ a/ $0 00 9o v0 9a cc 0 «0 vs vs

_ _ 0 0 0 _ _ 0

l - p - h m — I — I w o
. A
% - _ V o?

 

 

 

 

74 80
u
80 m.
0
8». .m.
_ m
M a con m
w 0 w
W 80 w
W, u.
W 2:
W
W _ 80
7

 

63

Table 2.12. List of herpetofaunal species captured by year and their combined capture
abundance in the southern Lower Peninsula of Michigan in summer 2005 and 2006 from
drift fence arrays coupled with pitfall traps, fimnel traps, and coverboards and area-time
constrained surveys.

 

 

Common Name 2005 2006
blue-spotted salamander 39 18
eastern tiger salamander 77 10
red-backed salamander 37 3O
eastern American toad 430 440
eastern gray treefrog 25 13
western chorus frog 9 2
spring peeper 150 65
green frog 493 67
wood frog 1145 1197
northern leopard frog 102 30
common snapping turtle 3 0
painted turtle 1 0
eastern garter snake 9 7
Butler’s garter snake l 0
northern ribbon snake 6 l
northern red-bellied snake 2 0
eastern milk snake l 0

 

64

search captures from June and July were used to eliminate any bias from difference in
number of months the searches were conducted in a given year. When eliminating trap
data influence from the differences in area time constrained searches, there was a
standardized catch per search effort of 154 animals per 100 searches (when pooling
years). When examining within years, there was a standardized catch per search effort of
186 animals per 100 searches in 2005, and a standardized catch per search effort of 120
animals per 100 searches in 2006.

Overall, mean relative abundance (proportion of captures adjusted for effort) was
greatest for the wood frog (0.56 :l: 0.04 SE), followed by eastern American toad (0.35 :1:
0.04 SE), spring peeper (0.15 :t 0.03 SE), and green frog (0.15 d: 0.03 SE) (Figure 2.9).
Mean relative abundance for SGWA was greatest for wood flog (0.58 i 0.06 SE), eastern
American toad (0.33 i 0.06 SE), spring peeper (0.17 i 0.04 SE), and red—backed
salamander (0.13 d: 0.04 SE) (Figure 2.10). Mean relative abundance for private lands
was greatest for wood frog (0.53 i 0.06 SE), eastern American toad (0.38 :t 0.06 SE),
green frog (0.17 i 0.04 SE), and northern leopard frog (0.07 d: 0.02 SE) (Figure 2.10).

Mean relative abundance by year was similar to the overall trends observed. In
2005, mean relative abundance was greatest for wood frog (0.52 i 0.06 SE) followed by
eastern American toad (0.33 :1: 0.05 SE), spring peeper (0.24 d: 0.05 SE), and green frog
(0.23 i 0.04 SE), and in 2006, mean relative abundance was greatest for the wood frog
(0.60 i 0.07 SE), followed by eastern American toad (0.38 i 0.06 SE), red-backed
salamander (0.09 d: 0.04 SE), and spring peeper (0.06 :t 0.03 SE) (Figure 2.11).

Mean species abundance was related with frequency of occurrence. Species with

high frequency of occurrences had higher mean abundances (Figure 2.12).

65

000300 0000000000 080-0000 000
00000050 000 .0000 00003 .0000 20.000 003 0000000 0000.0 0000.0 0.00 0000.0 0000 000000 00 0000 08m 000 moom
0000050 5 00w000=>0 .00 03000000 0030.0 0.0000000 00“ 00 0000000 0000000000000 00 000000000 0200000 0002 .0.m .wE

 

00.000w
«We? 6 $00 0 69 av 00.9 0/9
9 o 9 .0 9 ‘9
a o. a. «a o... go. we 9., I... e. e9 a. e.
we 9v v9 9/ 9) 9 90. 10 v9 00. 9 o. 10
vol .0 e f 09 a 00 9.. 0 v... e b 07 no a;
o 99 9% we 09 20 V». 9/ «0 90 no 96v 00 we 00
1% III 09 JV IV .10 . IV V0 47 W0 900/. 00' Th. Iv 0’0 0/0 9/9
b.os.o «9 «a 99?... go as". v w 99% e e
o o w o o 0 I o a e 9 we 9 o e e e
00 90 $0 00 #0 #09 1v»: 0%. av). 09 V9 a? cow 092 06 06 900
070799.799/9/999099099990
l_ .l 0 p I» r _ n C
To
m
- Nd .-
u
“a
P
B
u.
. 0.0 M
V
q
n
u
. w.
, no u
a
9
- m.o

 

 

 

0d

66

000300 0000000000 005-0000 000 m000on00>00 000 £000
000000 £000 000000 003 0000000 0000.00 00000 $00 08¢ 0000 000000 00 00000 ooom 000 moom 0000000 00 00w2002 0o
230.0000 003010 000808 05 00 000.2 000300 000 <>>Om 00 0000000 000000800000 00 000000000 030200 0002 60:0. .wE

02025

or
a»?

3020 n
<38 I

aannpunqv ”papa uuaw

 

 

 

 

No

67

.mzoasm 0050009000 05:

.0000 000 0000000300 000 .0000 000000 .0000 000000 003 0000000 000.00 00000 $00 05.0 0000 0000000 00 00000 000m 000 moom

00000.50 00 00w0000>0 00 000000000 00300 0.0000000 05 0_ 000% 00 0000000 000000800000 00 000000000 0300000 0002 .0 0m .wE
00_000w

09 04 0.0 To «a

9 v 90 9V0 v] 900’. 6v
9 we» 90 909 or

QVO Q0 §0 00

 

 

I‘L‘LlllulLln-J, , o
to
, No W
3
B
u
U
0 md m
IO
M.
3
0 . v
00 m.
u
D.
B
u
. no .3.
0.0

 

 

 

 

68

0000000000

00000 0wE 000 000 5000000000 00000000 0000? 00000000000 000000-000 000 00000000000 00w0 0000000

.00 000000000 000 0:3 AmdAv 0000000000 00000 0&3 000 AmmAv 0000000000 .00 0000000000 0E0 003 00000 000
0000000 000000m 00000000 000000000 0000-00.00 000 0000000030 000 .0000 000000 .0000 00000 003 000000
00000 500 0000.0 0000 0000000 00 00000 000m 000 moom 00000000 00 0030002 00 000000000 00300 00000000 000
00 000.000 000 000 0000000005 000000000 00000 0000000 m0m00> 0000000000 0000000 00 000000000 .N0.N .wE

0000000000 .00 00000—0000

on we ow mm om mvowmm om mm ON .39 m
0___0 _ _ h. _ _ _ 0.

 

web 000.0% x A

0000000200 000.000-00.00 I o o
v 00 00000000 0.00 .000 r wd
0 0 . < 0 0000000200 00w: 0.0000000

0
000000 E00000 F m. _‘

aaucpunqv uuaw

 

 

web 0003

 

69

Widespread species such as the wood frog, eastern American toad, and green fi'og were
more abundant. Less frequently occurring species such as Butler’s garter snake, northern
red-bellied snake, eastern milk snake, and common snapping turtle were less abundant.
Red-backed salamanders and especially eastern tiger salamanders were exceptions to the
rule: they had high mean abundances, but occurred infrequently.
Species Diversity

The Shannon Index of Diversity ranged from an index value of 0 (low diversity)
to an index value of 1.74 (higher diversity) and the Simpson Index of Diversity ranged
from 0 (low diversity) to an index value of 0.80 (higher diversity) (Table 2.13). The
mean Shannon Index of Diversity for the 82 sites was 0.83 (i 0.05 SE) and the mean
Simpson Index of Diversity was 0.43 (d: 0.03 SE). Species diversity was higher on
private lands (mean Shannon Index of 0.86 i 0.06 SE; mean Simpson Index of 0.45 i
0.03 SE) than SGWA (mean Shannon Index of 0.79 i 0.08 SE; mean Simpson Index of
0.4] :t 0.04 SE).

Species diversity differed between years and was significantly higher during
2005. In 2005, the Shannon Index of Diversity ranged from an index value of O to 1.74
and the mean Shannon Index of Diversity for the 42 sites was 1.04 (i 0.06 SE). In 2005,
the Simpson Index of Diversity ranged from 0 to 0.80 and the mean Simpson Index of
Diversity was 0.53 (:t 0.03 SE). In 2006, the Shannon Index of Diversity ranged from an
index value of 0 to an index value of 1.32 and the mean Shannon Index of Diversity was
0.61 (i 0.06 SE). In 2006, the Simpson Index of Diversity ranged from ranged from an
index value of O to an index value of 0.67 and the mean Simpson Index of Diversity was

0.32 (3: 0.04 SE).

70

Table 2.13. Species diversity indices in the southern Lower Peninsula of Michigan based
on capture data fi'om drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

Site Shannon Simpson Site Shannon Simpson
1 1.17 0.57 42 1.12 0.52
2 1.52 0.75 43 0.73 0.36
3 1.30 0.66 44 1.33 0.72
4 1.32 0.68 45 1.59 0.76
5 1.15 0.62 46 0.86 0.54
6 0.76 0.37 47 0.00 0.00
7 1.02 0.58 48 0.84 0.46
8 1.02 0.56 49 0.33 0.18
9 1.36 0.74 50 0.00 0.00

10 1.31 0.67 51 0.00 0.00
1 1 1.52 0.75 52 0.00 0.00
12 0.95 0.53 53 0.00 0.00
13 0.80 0.32 54 0.28 0.15
14 1.38 0.66 55 0.07 0.03
15 0.51 0.25 56 0.46 0.29
16 1.26 0.69 57 0.95 0.50
17 0.91 0.48 58 0.84 0.49
18 1.18 0.63 59 0.56 0.26
19 1.32 0.59 60 0.75 0.42
20 1.11 0.60 61 1.39 0.70
21 1.74 0.80 62 0.60 0.28
22 1.02 0.61 63 0.65 0.33
23 1.32 0.66 64 1.29 0.65
24 0.00 0.00 65 0.61 0.25
25 0.87 0.42 66 0.99 0.60
26 0.35 0.20 67 1.04 0.55
27 1.15 0.65 68 0.92 0.42
28 0.61 0.32 69 0.82 0.37
29 0.00 0.00 70 0.80 0.47
30 0.63 0.34 71 0.87 0.46
31 0.79 0.44 72 0.39 0.17
32 0.77 0.37 73 1.15 0.67
33 0.20 0.08 74 0.82 0.41
34 0.11 0.04 75 0.37 0.17
35 0.94 0.54 76 0.90 0.49
36 0.84 0.44 77 1.06 0.55
37 0.00 0.00 78 0.82 0.49
38 0.80 0.46 79 0.92 0.44
39 1.16 0.55 80 1.36 0.70
40 0.80 0.52 81 0.54 0.27
41 0.59 0.30 82 1.16 0.63

 

71

Herpetofaunal Community Similarity

Herpetofaunal community similarity indexes among sites ranged from 0.04 (no
similarities) to 1 (absolute similarity), however none had absolute dissimilarity (with an
index equal to 0. The mean similarity index for 80 sites was 0.70 (:1: 0. 004 SE),
indicating that on average the sites were more similar than dissimilar to each other.

I also examined community resemblance of sites within years. One site was
removed from each year before calculating the resemblance matrix because they did not
have any captures. In 2005, the herpetofaunal community similarity indexes ranged from
0.15 to 1 and the mean similarity index was 0.69 (:t < 0. 01 SE). In 2006, the
herpetofaunal community similarity index ranged from 0.04 to 1 and the mean was 0.69

(i < o. 01 SE).

DISCUSSION

Species inventories are necessary for conserving and managing for biological
diversity (Oliver and Beattie 1993 ), and acquiring accurate data on species richness and
abundance is also an important component of conservation efforts (Gibbons et al. 1997).
Data on both rare and common species should be used in making management decisions.
The 22 species documented from the 82 study sites in this project represented 51% of the
known Lansing sub-subsection of Ionia subsection of Region One amphibian and reptile
fauna (Harding 1997). Seventy-three percent (8 of 1 1) of Anurans, 43% (3 of 7) of
salamanders, 44% (4 of 9) of turtles, and 44% (7 of 16) snakes expected to be found in

this area were documented.

72

I surveyed all of the Anurans that I expected. The three species that I did not
observe tend to have low detection probabilities in my study area. Pickerel frogs are
generally uncommon in the study area, which may be attributed to its preference for cool
clear waters (Harding 1997). It is rarely detected in the volunteer-based Michigan Frog
and Toad Survey (MFT S) conducted throughout southern Michigan (Genet 2004).
Cope’s gray treefrog is morphologically identical to the eastern gray treefrog, and their
calls can be difficult to distinguish under certain temperature conditions (Harding 1997).
In my study area and throughout southern Lower Michigan, the eastern gray treefrog is
more common than the Cope’s gray treefrog (Harding 1997). The northern cricket frog is
very rare in Michigan, and it is the only Anuran species to have protected status in the
state (Genet 2004). Severe declines in the northern extent of their range have resulted in
scattered and isolated populations (Lannoo 2005). Formerly healthy populations of
northern cricket fiogs in both Wisconsin and Michigan have been drastically reduced or
extirpated (Harding 1997).

1 surveyed all of the salamanders that 1 expected. Life history characteristics not
conducive to terrestrial sampling contributed to lack of detection for four species not
observed. Mudpuppies are strictly an aquatic species that occur in permanent waters
including rivers, reservoirs, inland lakes, and Great Lake bays and shallows (Harding
1997). Spotted salamanders spend most of their time in burrows underground and are
rarely seen after their spring breeding season which goes from late March until mid-April
(Harding 1997). Four-toed salamanders are generally rare, which is most likely attributed
to the lack of suitable habitat (Harding 1997). They prefer moist woodlands in close

proximity to spring-fed creeks or bogs (Harding 1997). Four-toed salamanders are

73

believed underrepresented in inventories because of their short larval period, secretive
nature, scattered isolated populations, and the tendency of adult females to skip years of
reproduction (Lannoo 2005). The eastern newt is aquatic as an adult and inhabits pools,
ponds, wetlands, sloughs, and canals (Lannoo 2005). However, juvenile efis are
terrestrial using wooded areas and movement occurs when the ground is moist, especially
during rain or humid days and nights (Lannoo 2005).

More species of turtles were not detected than were observed, but this was not
surprising. Life history characteristics not conducive to terrestrial sampling were
responsible for the lack of detection at the study sites. Common musk turtles are aquatic
and rarely leave the water (Conant and Collins 1991). They inhabit a wide array of
permanent water bodies including but not limited to ponds, lakes, marshes, and rivers
(Harding 1997). Spotted turtles use shallow ponds, bogs, fens, tarnarack ponds, and
sphagnurn seepages, and reduce their movements and basking activity during the summer
months (Harding 1997). They are rare in the Great Lakes region (Harding 1997) and are
listed as threatened in Michigan (Michigan Nature Features Inventory 2002). Common
map turtles are highly aquatic, inhabiting and feeding in larger lakes, rivers, open
marshes, and reservoirs (Harding 1997). Red-eared sliders inhabit mostly permanent
bodies of water including ponds, lakes, reservoirs, swamps, and slower sections of rivers,
and they rarely travel far from water bodies (Harding 1997). Spiny sofishell turtles are
highly aquatic. They inhabit rivers, lakes, reservoirs, and streams; bask and forage in the
water; and bury themselves in silt or soft mud (Conant and Collins 1991, Harding 1997).

Lack of detection for several species of snakes can be attributed to habitat

requirements and declining populations. Copper-bellied water snakes are aquatic and use

74

shrub swamps, ponds, lakes, fens, and slow moving streams, and mating takes place at
basking sites at edges of ponds and swamps (Harding 1997). Copper-bellied water
snakes are listed as state endangered and federally threatened (Michigan Nature Features
Inventory 2002). Queen snakes have specialized habitat requirements including warm,
shallow streams with rocky bottoms and an abundance of crayfish (Harding 1997). They
are declining in the Great Lakes region due to their sensitivity to stream pollution, habitat
modification, and siltation and agricultural runoff effects on crayfish populations
(Harding 1997). Brown snakes are very secretive (Conant and Collins 1991) and spend
the majority of their time underground, making detection difficult (Harding 1997). Green
snakes prefer moist grassy areas such as prairie remnants, meadow and fields. They do
occur in open deciduous and pine stands, corresponding to vegetation types sampled in
this study, but they are declining and locally extirpated from much of the southern Lower
Peninsula (Harding 1997). Central rat snakes also suffer from declining populations;
they are uncommon to rare, and populations have been drastically reduced or extirpated
in the Great Lakes region (Harding 1997). They are listed as a species of special concern
in Michigan (Michigan Nature Features Inventory 2002). Ring-necked snakes have also
suffered from population declines. This decline coupled with its secretive nature of
remaining below ground or under cover makes it a difficult species to survey (Harding
1997). Hog-nosed snakes occur in a wide variety of terrestrial habitat types and oflen
remain underground when they are not foraging (Harding 1997). They too suffer from
declining numbers and are uncommon to rare in much of the Great Lakes region (Harding

1997). The eastern massasauga rattlesnake has been extirpated through the majority of its

75

range (Prior and Weatherhead 1994). Massasaugas prefer lowland habitats and they are
listed as a species of special concern in the state of Michigan (Harding 1997).

To ensure accurate and all-encompassing species inventories, more extensive
sampling should employed to increase the likelihood of capture for the approximately
50% of species not detected by my surveys. To ensure accurate inventories, a variety of
sampling methods in varying habitats needs to be employed because effectiveness of
different methods varies even between closely related herpetofauna species (Gibbons et
al. 1997). To accurately assess the herpetofaunal community, it is important to use
several different survey techniques to take advantage of the various life history strategies
and detection phenologies. In particular, I observed different survey results between frog
call surveys and pitfall traps even though they were conducted in the same vicinity.
Wood frogs were underrepresented during the call surveys because they have a short
calling window during their 6 to 14 day breeding season in late March to early April
(Harding 1997) and many sites were not sampled during this period. However, they were
the most abundant species captured (2,342 individuals captured). Eastern gray treefrogs
were heard more frequently than captured because of their climbing ability which enables
them to escape from pitfall traps and drift fences (Dodd 1991). Because wetlands were
not specifically targeted during captures and because bullfrogs are closely associated with
larger, more permanent bodies of water (Kolozsvary and Swihart 1999), bullfrogs were
only encountered during the frog call surveys that sampled larger permanent wetlands.

Because my study focused on terrestrial sampling and aquatic sampling
techniques such as hoop traps and minnow traps were not utilized, most aquatic species

of amphibians and reptiles were not observed or were underrepresented in this study.

76

Managers should employ target searches for species that are declining or naturally rare or
they may need to develop long-term monitoring programs. The scale of sampling that I
undertook required a considerable amount of effort, time, and funding. Most managers
do not have the fimding or the personnel to ensure comprehensive detection of all
species. Species-specific detection probabilities most likely vary by month and
ownership which further complicates survey efforts. Understanding species-specific
detection probabilities would help managers estimate the amount of effort required to
document a species occurrence. Therefore, my results should be considered a snapshot of
the potential species in the ecoregion. Future studies could focus on the species that were
not detected and on calculating detection probabilities.

It is common for species with high mean abundances to occur at high frequencies
of occurrence (Brown 1984). I found evidence to support this relationship with the
commonly occurring species like wood frogs, eastern American toads, and green frogs.
Eastern tiger salamanders were found at relatively high abundances, but only occurred at
five sites. Out of those five sites, 82% (71 of 87 captures) were from one site. The
sampling site was randomly placed between two vernal ponds and was likely located in
optimal habitat for eastern tiger salamanders which depend on access to permanent and
semi-permanent water bodies (Harding 1997). The high abundance of Eastern tiger
salamanders at relatively few sampling sites suggests that some populations are highly
localized and as such, true random sampling may not be the most efficient sampling
approach for some species. Due to their aquatic habitat requirements, secretive nature, or
declining population numbers, other species with low frequencies of occurrence such as

Butler’s garter snake, northern red-bellied snake, northern water snake, and eastern milk

77

iii 4

 

snake were probably anecdotally sampled and therefore their population-habitat
relationships are not accurately described by the data.

The herpetofaunal species pool is similar throughout the ecoregion sampled in
this study. The study was purposefully designed to reside in one ecoregion so that the
herpetofaunal communities would be similar; and thereby giving me a better chance to
detect ownership or habitat effects. Therefore it is not surprising that herpetofauna
community similarity among sites was more similar than dissimilar. My results confirm
that I was successful in removing across ecoregion variability from analyses. Similarities
occur because I had the potential to observe the same amphibian and reptile species
across all of my sites. Although some sites were more dissimilar fiom one another, they
most likely reflect the instances when rare or under-sampled species occurred.
Comparisons between Land Ownerships

Although, the herpetofaunal species pool is the same throughout the ecoregion,
my results revealed different patterns of abundance and occurrence between land
ownerships. Overall, more species were detected on SGWA including Blanding’s turtle,
painted turtle, common snapping turtle, Butler’s garter snake, and northern red-bellied
snake. Three species were located solely on private land including the eastern box turtle,
northern water snake, and eastern milk snake. However, all of these unique species were
sampled in low numbers, regardless of land ownership. The disparity between species
observed on SGWA- versus private lands is most likely attributed to life history
requirements, and the difficulty of efficiently sampling these species. Herpetofaunal

communities as a whole were similar between ownerships with average richness and

78

amphibian composition the same for SGWA and private lands (4.6 (:1: 0.36 SE) vs. 4.4 (:1:
0.28 SE)).

Both SGWA and private lands had similar species occurrence patterns with
eastern American toad being the most ubiquitous followed by wood frogs; however, both
of these species occurred more frequently at private sites. Mean abundance of individual
species differed on SGWA and private lands. For both land ownership types, wood fiogs
followed by eastern American toad were the most abundant. Wood frogs were more
abundant on SGWA and eastern American toads were more abundant on private lands.
On SGWA, spring peepers and red-backed salamanders were more abundant than the
next most abundant species on private lands (green fi'ogs and northern leopard frogs).
Differences in abundances of individual species between SGWA and private lands are
most likely related to species-specific life history requirements, habitat conditions and
availability, micro-site conditions as well as environmental conditions. For example,
wood frogs and blue-spotted salamanders are highly dependent on vernal ponds in
wooded areas of Michigan. If one of the ownership types tended to support more vernal
ponds in the vicinity of the sampling sites, it could have an effect on wood frog and blue-
spotted salamander relative abundances.

Comparisons between Years

Although, the herpetofauna species pool is similar throughout the ecoregion, my
survey also revealed different patterns of incidence and occurrence among years.
Temperature and precipitation are critical to amphibians (Carey et al. 2001) and these
weather factors vary from year to year. Herpetofaunal breeding is initiated by a response

to increasing temperatures and rainfall (Stebbins and Cohen 1995), and temperature and

79

rainfall also affect the hydroperiod of a wetland (Pechmann et a1. 1989). The number of
herpetofauna using a particular wetland in a given year is a function of the wetland
hydroperiod which can vary annually (Pechmann et al. 1989). Therefore, the number of
species occurring at a given site is partially dependent on the type of wetlands in the
surrounding landscape and the hydroperiods of these wetlands. For example, wood fro gs
are explosive breeders. If certain areas within the ecoregion during a given year received
more early season rainfall that resulted in a longer hydroperiod for breeding ponds, this
would create conditions favorable for tadpoles to completing metamorphosis. As a result
chances for individual survival are increased.

In 2005, I observed warmer temperatures at my sampling sites, which could have
contributed to the greater species richness. The average temperature for the 4 sampling
periods was obtained for each sample site location from the closest weather station
distributed through the ecoregion. Seven species were observed only in 2005 and all
were reptile species (Blanding’s turtle, painted turtle common snapping turtle, Butler’s
garter snake, northern water snake, eastern milk snake, northern red—bellied snake). The
discrepancy between species observed in 2005 versus 2006 is most likely a combination
of factors including temperature and precipitation differences, life history requirements,
and the difficulty of efficiently sampling these species in terrestrial habitats. My results
suggest climatological differences played a role in my characterization of southern
Michigan herpetofaunal communities.

Overall, individual species abundances were greater in 2005, and 658 more
individuals were captured in 2005. The majority of captures were Anurans, and of these

Anurans, most were juveniles. Here again, more favorable weather conditions could

80

have triggered earlier breeding seasons for these species resulting in more tadpoles
completing metamorphosis before vernal pools dried and as juveniles, they dispersed
through terrestrial habitat. Species diversity was greater in 2005, but this was not
surprising. Diversity indices take into account species richness and abundance. Because
I observed more species in 2005 and abundances of these species were greater, the
average diversity index was higher.

My study was successful in sampling the targeted terrestrial species and
documented the majority of terrestrial species expected to occur in the study region.
Because accurate information on basic life history characteristics are needed to conserve
species, my research demonstrates the difficulty in surveying complete species
assemblages, particularly for uncommon species or species with specialized habitat
requirements. There is variation in herpetofauna populations fiom year to year and
therefore basing management policies on short-term studies could result in misleading
conclusions (Dodd 1994). Amphibian abundance fluctuates with regard to environmental
conditions such as rainfall and hydroperiod (Pechmann et a1. 1989). Long-term studies at
Savannah River Site National Research Park (SRS-NEP) of turtle populations in wetlands
have shown that population rates changed annually in relation to fluctuating water levels
(Gibbons et al. 1997). Turtle densities were high in years of high water levels and wet
conditions and densities were low in years with limited water (Gibbons et al. 1997).
Therefore, it is important to present study results in the context of environmental
conditions over which the data were collected, especially since funding and personnel are
usually only available for short-term studies. During the two year sampling period that I

conducted my research, I documented over half of all the species known to historically

81

occur in the ecoregion. My research will give the MDNR a solid knowledge base of the
herpetofaunal community in the study area and can be used as a foundation for further
research.

It is essential to look at the local and spatial environment associated with
herpetofaunal communities to determine underlying relationships among sites, land
ownership types, and between years. Describing herpetofaunal communities in this
context is necessary to guide natural resource management. Also, understanding the
distribution and abundance of herpetofauna in the context of site and landscape-level
variables is fundamental to conserving and managing for biological diversity. The lack
of basic understanding in abundance and distribution of herpetofauna, as well as of many
other taxa, complicates management and conservation efforts. Further coordinated
efforts such as Partners for Amphibian and Reptile Conservation (PARC) and long-term
studies such as the ecological studies conducted since 1951 on the Savannah River Site
National Enviromnental Research Park (SRS-NERP) (Gibbons et al. 1997) should be

initiated to properly manage for herpetofaunal diversity.

82

LITERATURE CITED

Alford, R. A., and S. J. Richards. 1999. Global amphibian declines: a problem in applied
ecology. Annual Review of Ecology and Systematics 1999 (30):133-165.

Berry, F. A., E. Bollay, and N. R. Beers, editors. 1945. Handbook of meteorology. Mc-
GrawHill Publishing, Chicago, Illinois, USA.

Blaustein, A. P., and J. M. Kiesecker. 2002. Complexity in conservation: lessons fiom
the global decline in amphibian populations. Ecology Letters 52597-608.

Blaustein, A. P., and D. B. Wake. 1995. The puzzle of declining amphibian populations.
Scientific America 272(4):52-58.

Blaustein, A. P., D. B. Wake, and W. P. Sousa. 1994. Amphibian declines: judging
stability, persistence, and susceptibility of populations to local and global
extinctions. Conservation Biology 8(1):60-71 .

Bolen, E.G., and W. L. Robinson. 1995. Wildlife ecology and management. Prentice
Hall, Inc., Englewood Cliffs, New Jersey, USA.

Bray, J. R., and J. T. Curtis. 1957. An ordination of the upland forest communities of
southern Wisconsin. Ecological Monographs 27:325-349.

Brown, J. H. 1984. On the relationship between abundance and distribution of species.
The American Naturalist 124(2):255-279.

Brown, W. S., and W. S. Parker. 1976. A ventral scale clipping system for permanently
marking snakes (Reptilia, Serpentes). Journal of Herpetology 10:247-249.

Burton, T. M., and G. E. Likens. 1975. Energy flow and nutrient cycling in salamander
populations in the Hubbard Brook Experimental Forest, New Hampshire.
Ecology 56:1068-1080.

83

Bury, B. 2006. Natural history, field ecology, conservation biology and wildlife
management: time to connect the dots. Herpetological Conservation and Biology
1(1):]01-108.

Bury, B., M. L. McCallum, R. A. Saumure, D. J. Germano, and S. E. Trauth. 2006. A
revisit of a classic paper on amphibians and reptile community ecology.
Herpetological Conservation and Biology 1 (2): 101 -1 08

Cagle, F. R. 1939. A system of marking turtles for firture identification. Copeia.
1939:170-172.

Campbell, H. W., and S. P. Christrnan. 1982. The herpetological components of Florida
sandhill and sand pine scrub Associations. Pages 163-171 in N. J. Scott, editor.
Herpetological communities. Fish and Wildlife Service Wildlife Resources
Report 13. Washington, DC, USA.

Carey, C., W. R. Heyer, J. Wilkinson, R. A. Alford, J. W. Amtzen, T. Halliday, L.
Hungerford, K. R. Lips, E. M. Middleton, S. A. Orchard, and A. S. Rand. 2001.
Amphibian declines and environmental change: use of remote sensing data to
identify environmental correlates. Conservation Biology 15(4):903-913.

Conant, R., and J. T. Collins. 1991. Field guide to reptiles and amphibians eastern and
central North America. Houghton Mifflin Company, New York, New York,
USA.

Corn, P. S. 1994. Straight-line drift fences and pitfall traps. Pages 109-117 in W.R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek, and M. S. Foster,
editors. Measuring and monitoring biological diversity: standard methods for
amphibians. Smithsonian Institution Press, Washington DC, USA.

Crump, M. L., and N. J. Scott. 1994. Visual encounter surveys. Pages 84-92 in W.R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington DC, USA.

Davis, T. M. 1997. Nondisruptive monitoring of terrestrial salamanders with artificial
cover objects on southern Vancouver Island, British Columbia. Herpetological
Conservation 1:161-174.

84

deMaynadier, P. G., and M. L. Hunter, Jr. 1995. The relationship between forest
management and amphibian ecology: a review of North American literature.
Environmental Review 3 2230-261 .

Detenbeck, N. E., Galatowitsch, S. M., Atkinson, J. and H. Ball. 1999. Evaluating
perturbations and developing restoration strategies for inland wetlands in the
Great Lakes Basin. Wetlands 19:789—820.

Dodd., K. C., Jr. 1991. Drift fence-associated sampling bias of amphibians at a Florida
sandhills temporary pond. Journal of Herpetology 25:296-301.

Dodd, K. C., Jr. 1994. Monitoring and protecting biotic diversity. Pages 1-11 in S. K.
Majumdar, F .J . Brenner, J. E. Lovich, J .F. Schalles, and E. W. Miller, editors.
Biological diversity: problems and challenges. Pennsylvania Academy of
Sciences, Eaton, Pennsylvania, USA.

Enge, K. M. 1997. A standardized protocol for drift-fence surveys. Florida Game and
Fresh Water Fish Commission. Technical Report No.14. Tallahassee, Florida,
USA.

Fellers, G. M., and C. A. Drost. 1994. Sampling with artificial cover. Pages 146-150 in
W. R. Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster,
editors. Measuring and monitoring biological diversity: standard methods for
amphibians. Smithsonian Institution Press, Washington DC, USA.

Forman, R. T. T. 1995. Land mosaics: the ecology of landscapes and regions.
Cambridge University Press, Cambridge, UK.

Genet, KS. 2004. Status and distribution of fiogs and toads in southern Michigan:
population trends and the influence of habitat and landscape characteristics.
Dissertation, Michigan State University, East Lansing, Michigan, USA.

Gibbons, J. W., V. J. Burke, J. E. Lovich, R. D. Semlitsch, T. D. Tuberville, J. R. Bodie,
J. L. Greene, P. H. Niewiarowski, H. H. Whiteman, D. E. Scott, J. H. K.
Pechmann, C. R. Harrison, S. H. Bennett, J. D. Krenz, M. S. Mills, K. A.
Buhlman, J. R. Lee, R. A. Seigel, A. D. Tucker, T. M. Mills, T. Lamb, M. E.
Dorcas, J. D. Congdon, M. H. Smith, D. H. Nelson, M. B. Dietsch, H. G. Hanlin,
J. A. Ott, and D. J. Karapatakis. 1997. Perceptions of species abundance,

85

distribution, and diversity: lessons from four decades of sampling on a
government-managed reserve. Environmental Management 2:259-268.

Greenberg, C. H., D. G. Neary and L. D. Harris. 1994. A comparison of herpetofaunal
sampling effectiveness of pitfall, single-ended, and double-ended funnel traps
used with drift fences. Journal of Herpetology 28(3):319-324.

Greenberg, C. H., and G. W. Tanner. 2005. Spatial and temporal ecology of eastern
spadefoot toads on a Florida landscape. Herpetologica 61(1):20-28.

Greene, H. W. 2005. Organisms in nature as a central focus for biology. Trends in
Ecology and Evolution 20:23-27.

Halliday, T. 2000. Do frogs make good canaries? Biologist 47: 143-146.

Harding, J. H. 1997. Amphibians and reptiles of the Great Lakes Region. The
University of Michigan Press, Ann Arbor, Michigan, USA.

Heyer, W.R., M. A. Donnelly, R. W. McDiarmid, L. C. Hayek, and M. S. Foster, editors.
1994. Measuring and monitoring biological diversity: standard methods for
amphibians. Smithsonian Institution Press, Washington DC, USA.

Hilton-Taylor, C., compiler. 2000. 2000 IUCN red list of threatened species.
International Union for Conservation of Nature, Gland, Switzerland.

Houze, C. M., and C. R. Chandler. 2002. Evaluation of coverboards for sampling
terrestrial salamanders in South Georgia. Journal of Herpetology 36:75-81.

IUCN, Conservation lntemational, and NatureServe. 2006. Global Amphibian
Assessment. <www.globalamphibians.org >. Accessed 27 Jan 2007.

Kats, L. B., and R. P. Ferrer. 2003. Alien predators and amphibian declines: a review of
two decades of science and the transition to conservation. Diversity and
Distributions 9:99-1 10.

86

Kiesecker, J. M., A. R. Blaustein, and L. K. Belden. 2001. Complex causes of
amphibian population declines. Nature 410:681-684.

Kolozsvary, M. B., and R. K. Swihart. 1999. Habitat fragmentation and the distribution

of amphibians: patch and landscape correlates in farmland. Canadian Journal of
Zoology 77:1288-1299.

Krebs, C. J. 1999. Ecological methodology. Addison-Welsey Educational, Publishers
Inc., New York, New York, USA.

Lancia, R. A., J. D. Nichols, and K. H. Pollock. 1994. Estimating the number of animals
in wildlife populations. Pages 215-253 in T. A. Bookhout, editor. Research and

management techniques for wildlife habitats. Allen Press, Inc., Lawrence,
Kansas, USA.

Lannoo, M. L., editor. 2005. Amphibian declines: the conservation status of United
States species. University of California Press, London, England.

McCallum, M. L., and J. McCallum. 2006. Publication trends of natural history and
fields studies in herpetology. Herpetological Conservation and Biology 1(1):62-
67.

McCune, B., and J. B. Grace. 2002. Analysis of Ecological Communities. MjM
Software Design, Glenedon Beach, Oregon, USA.

McGarigal, K., S. Cushman, and S. Stafford. 2000. Multivariate Statistics for Wildlife
Ecology Research. Springer-Verlag, New York, New York, USA.

Michigan Natural Features Inventory. 2002. Michigan’s special animals. Michigan
Natural Features Inventory, Lansing, MI, USA. http://www.web4.msue.msu.edu/
mnfi/data/specialanimals.cfm>. Accessed 26 Mar 2007.

Oliver, 1., and A. J. Beattie. 1993. A possible method for the rapid assessment of
biodiversity. Conservation Biology 7:562-568.

87

Pechmann, J. H., K., D. E. Scott, J. W. Gibbons, and R. D. Semlitsch. 1989. Influence of
wetland hydroperiod on diversity and abundance of metamorphosing juvenile
amphibians. Wetlands Ecology and Management 1:3-11.

Primack, R. B. 2004. A primer of conservation biology. Third edition. Sinauer
Associates, Inc., Sunderland, Massachusetts, USA.

Prior, K. A., and P. J. Weatherhead. 1994. Response of free ranging eastern massasauga
rattlesnakes to human disturbance. Journal of Herpetology 28(2):255-257.

R Development Core Team 2006. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-
900051-07-0, URL http://www.R-proj ect.org.

Reaser, J. K., and A. Blaustein. 2005. Repercussions of global change. Pages 60-63 in
M. Lannoo, editor. Amphibian declines: the conservation status and of United
States species. University of California Press, Berkeley, California, USA.

Richter, S. C., and R. A. Seigel. 2002. Annual variation in the population ecology of the

endangered Gopher Frog, Rana sevosa Goin and Netting. Copeia 2002(4):962-
972.

Sargent, L. G. 2000. Frog and toad population monitoring in Michigan. Journal of the
Iowa Academy of Sciences 107(3):]95-199.

Saterson, K. A., N. L. Christensen, R. B. Jackson, R. A. Kramer, S. L. Pirnm, M. D.
Smith, and J. B. Wiener. 2004. Disconnects in evaluating the relative
effectiveness of conservation strategies. Conservation Biology 18(3):597-599.

Shannon, C. E. 1949. The mathematical theory of communication. Pages 29-125 in C.
E. Shannon and W. Weaver, editors. The Mathematical Theory of
Communication. University of Illinois Press, Urbana, Illinois, USA.

Simpson, E. H. 1949. Measurement of diversity. Nature 163:688

88

Stebbins, R. C., and N. W. Cohen. 1995. A natural history of amphibians. Princeton
University Press, Princeton, New Jersey, USA.

USDA Forest Service. 2004. Forest inventory and analysis national core field methods
for phase 3 measurements. USDA Forest Service, Inventory and Analysis,
Arlington, VA, USA. < http://www.fia.fs.fed.us/ library /field-guides-methods-
proc/docs/p3__2-0_secl4_ 3_04. pdf.>. Accessed 3 Sep 2005.

Weir, L. A., and M. J. Mossman. 2005. North American monitoring program
(NAAMP). Pages 307-313 in M. Lannoo, editor. Amphibian declines: the
conservation status and of United States species. University of California Press,
Berkeley, California, USA.

Wolda, H. 1981. Similarity indices, sample size and diversity. Oecologica 50:296-
302.

89

 

 

 

mo 030520.“ 2253 2.55900 05 E 303900 050.890.. mo cams? EB 3>mv 03:2 05> 9 28:3 2588:0002 ._.N 022505.00

00.0 2 002 08502 00.0 2 009 08:25. 00.0 2 009 5255. 2 .0 2 002 0888.0
00.0 2 009 505:2 20 2 002 0858.0 2 ._ 2 08. 08:25. 002 2 009 085.5.
00.0 2 009 08:05. 2 .2 2 002 0858.0. 02 .2 2 08. 08:82 00.2 2 009 08:05.
00.0 2 009 08502 00.0 2 009 0888.0 2 .2 2 08. 0855. 00.0 2 009 08502
00.0 2 002 08.55 00.0 2 009 0358.0 00.2 2 08. 5058.0 00.0 2 002 08580.
00.0 2 002 805.5. 00.0 2 009 085:2 00.0 2 009 0855. 00.0 2 002 08502
00.0 2 009 085:2 00.0 2 009 08:25. 00.0 2 08. 0858.0. 00.0 2 009 03:25.
00.0 2 002 80.0.82 00.0 2 009 03:05. 00.0 2 009 08580. 00.0 2 009 085.50
00.0 2 002 0855. 00.0 2 009 085.5. 00.0 2 009 0858.0 00.0 2 002 085.5.
00.0 2 002 8058.0 00.0 2 08. 0855. 00.0 2 009 08:25. 00.0 2 009 08:08.0
00.0 2 08. 0858.0. 00.0 2 009 085:2 00.0 2 002 08:22 00.0 2 002 08584
00.0 2 002 50502 2.0 2 002 0888.0 00.0 2 009 5058.0 00.0 2 002 08:25.
00.0 2 009 08:25. 00.0 2 002 0885. 00.0 2 002 08502 00.0 2 002 08:20.0
00.2 2 002 Sofia... 00.0 2 009 0858.... 00.0 2 08. 035.5. 00.0 2 002 085:2.
02 2 009 5055. 00.0 2 009 08:82. 00.0 2 009 08:82. 00.0 2 009 0853
00.0 2 002 505:2 2 ._ 2 002 08:25. 00.0 2 08. 08502 00.0 2 08. 085:2
00.0 2 002 0888.0 00.2 2 08. 08:03. 00.0 2 000. 08:82 00.0 2 009 08:25.
00.0 2 002 085:2 00.0 2 002 08502 00.0 2 009 08:82 00.0 2 009 08:82
00.0 2 009 805.50 00.0 2 009 0855 00.0 2 002 0858.0 00.0 2 002 0858.0
00.0 2 002 8058.0 00.0 2 009 085:2 000 2 08. 08550 00.0 2 009 0855
00.0 2 002 585.5. 00.0 2 009 505.5. 00.0 2 08. 5055.0 00.0 2 009 08:82
00.0 2 009 0855. 00.0 2 009 505.5. 00.0 2 08. 035.5. 00.0 2 08. 0852.0
00.0 2 003 8053. 00.0 2 002 0855. 00.0 2 08. 08:82 00.0 2 009 08:25.
00.0 2 009 80:25. 00.0 2 009 5055. 00.0 2 002 088:2 00.0 2 002 0855.
00.0 2 002 08:25. 00.0 2 002 08:82 00.0 2 08. 50:05... 00.0 2 002 08:08.0.
00.0 2 009 80:82 00.0 2 009 50:08.0. 00.0 2 08. 500.25. 00.0 2 009 085:2
00.0 2 39 9886.0 00.0 2 08. 500.05. 00.0 2 08. 08502 00.0 0 08. 000802
00.0 2 009 82082 00.0 2 002 5055. 2 .0 2 08. 0855. 00.0 0 009 505:2
00822 35 208200 2880 @5000 250 208800 25
2063 .50 0802 88800 2203 .50 0802 8588 £033 .50 camz 00888 2053 .50 0802 02:88
.0000 000 0000 2 000222

90

 

 

002 2.0 082 802208... 02 2 2.0 082 080.082.. 00. 2 00 082 08:08... 00. 2 00 002 08:08...
002 2.0 082 80.2205. 022 2.0 082 .2858... 002 00 082 53220222 00. 2 00 082 08:08...
002 2.0 2.82 502.25.. 022 2.0 082 2.82223. 002 00 2.82 082.025. 002 00 082 08.22222
02.2 2.0 082 2282.250 002 2.0 082 08.22050 02 2 00 082 5.02225. 002 00 0002 082.0250
002 2.0 2.82 8022082.. 002 2.0 082 2282220... 02 2 00 2.82 08.0220... 02.2 00 2.82 5022025..
002 2.0 2.82 802225.. 002 2.0 082 802225.. 02 2 00 082 5.05.5. 02.2 00 08. 082205...
002 2.0 082 22822250 00.0 2.0 082 2282205. 02 2 00 082 802225.. 02.2 00 082 082205...
002 2.0 082 082208... 00.0 2.0 082 2282.25. 02 2 00 082 5022250 002 00 082 082225..
002 2.0 082 500.082. 00.0 2.0 082 2282203... 02 2 00 082 500225.. 002 00 082 0822025..
002 2.0 2.82 22002208... 00.0 2.0 082 08.58... 002 00 082 5022025. 002 00 08. 082208...
02.2 2.0 2.82 2282225.. 00.0 2.0 082 22822082. 002 00 08. 5.02.2.2 002 00 082 085522
02.2 2.0 2.82 2.82.25. 00.0 2.0 082 2282.220... 002 00 08. 2.82205... 00. 2 00 082 082222....
02.2 00 082 2282208.... 00.0 00 082 08:08... 002 00 082 502225. 002 00 08. 2.82.25.
02.2 2.0 082 802282... 002 00 082 0822025.. 002 00 082 082.208... 00. 2 00 2.82 0822052.
02.2 2.0 082 802208... 002 00 082 22855... 002 00 2.82 5.02.05... 00. 2 00 082 082.228..
02.2 2.0 082 502.225.. 002 00 2.82 08.55. 002 00 082 50:25. 002 00 082 2.82.25.
02.2 2.0 2.82 2.82225. 002 00 0002 8022222 002 00 082 80220222 002 00 082 22822222
002 2.0 082 80.58.... 002 00 0002 082208... 00.0 00 082 502225. 02 2 00 082 532225.
002 2.0 2.82 802225. 002 00 082 802208... 00.0 00 2.82 50:08.... 022 00 082 22822222
002 2.0 082 802.25. 002 00 082 228.2208... 00.0 00 082 80220252 022 00 082 228.008...
002 2.0 082 805a... 002 00 082 082208... 00.0 00 082 2282208... 022 00 082 082208...
002 2.0 082 2.82225. 002 00 082 502.25.. 00.0 00 082 502225. 02 2 00 08. 08220.5.
002 2.0 082 802.208... 02.2 00 082 228200820 00.0 00 0002 5.02025. 00. 2 00 08. 22822250
002 2.0 082 8058.2. 02.2 00 082 2282208... 00.0 00 082 502208... 002 00 08. 0002.205.
002 2.0 003 2282.205. 02.2 00 082 082225. 00.0 00 082 502225.. 002 00 082 22822025.
02 2 2.0 082 802208... 002 00 0002 2282208.... 002 00 082 502.225. 00.0 2 082 2282225.
02 2 2.0 082 802225. 002 00 2.82 0822290 002 00 2.82 228525.. 00.0 2 082 2.82225.
022 2.0 082 802.25. 002 00 082 50220222.... 002 00 082 502.205. 00.0 2 08. 08:25.
022 2.0 082 802203. 002 00 082 5022082.. 002 00 082 082208... 00.0 2 082 22002205...
00801202 2225 208202 222280 208800 22520 20882 2525

2220203 2>0 0802 25888 2220203 2>0 0802 8858 2.20203 2>0 0802 0052228 2220203 2>0 252 22022288

 

2.2200 2.0 20220022....

91

 

 

00.2 00 082 2282225.. 00.0 00 082 082.208... 00.2 00 2.82 2282225. 00.2 00 082 2282.25.
00.2 00 2.82 2.80208... 00.0 00 082 22822082.. 00.2 00 2.82 2282223. 00.2 00 2.82 2282025.
02.2 00 2.82 2282.08... 02 .0 00 2.82 0822222.... 00.2 00 2.82 2282.225. 02.2 00 082 22822222
00.2 00 2.82 2282208... 00.0 00 2.002 2282.208... 002 00 082 2282.25. 02.2 00 2.82 2282.222...
002 00 2.82 228220222. 002 00 2.82 22822230 00.2 00 082 2282208... 02.2 00 082 2282225.
00.2 2.0 082 2.82225. 00.2 00 082 2282225. 002 00 2.82 22822222 02.2 00 2.82 2282205<
00.2 00 2.82 802208... 00.2 00 2.82 22822002... 002 00 2.82 2282225. 02.2 00 082 22822022....
002 00 2.82 2282208... 002 00 2.82 2282205... 002 00 2.82 082.225. 00.2 00 082 082225..
00.2 00 2.82 802.208... 00.2 00 082 802208... 00.2 00 082 5022023.. 00.2 00 082 22822222....
00.2 00 082 2232223. 00.2 00 2.82 22822222... 00.0 00 2.82 2282225. 00. 2 00 082 2282225..
002 00 2.82 22822025. 002 00 2.82 2282225.. 02 .0 00 2.82 2282225. 00. 2 00 082 2.82.205...
002 00 2.82 22822222 00.2 00 082 502208... 00.0 00 082 502208... 002 00 08. 2.82.2290
00.2 00 2.82 2282205. 00.2 00 082 228.2205... 0 2 .0 00 2.82 2282.205. 02 .2 00 082 2282.25.
02.2 00 2.82 22822082.. 002 00 2.82 2282223.. 02 .0 00 082 2282202222. 0 2 .2 00 082 228220222...
02.2 00 2.82 000.2208... 00.2 00 082 2282282 00.0 00 2.82 228220222... 02 .2 00 082 220022025.
02.2 00 082 802208... 00.2 00 2.82 2282206.... 00.0 00 0282 2.8.2.208... 0 2 .2 00 2.82 2282225.
02.2 00 082 22822023.. 00.2 00 082 2282.208... 002 00 2.82 502.0222... 00.2 00 082 22822222
00.2 00 2.82 802223.. 00.2 00 082 2282282... 00.2 00 2.82 502208... 00.2 00 082 2282.05...
002 00 2.82 22002225. 002 00 2.82 2282208... 002 00 082 5022250 00.0 00 2.82 2282225..
02.2 00 2.82 80220222 00.2 00 2.82 228220820 00.2 00 082 8022082.. 00.0 00 082 082.2025.
02 .2 00 2.82 2282282. 00.2 00 082 22822222... 002 00 082 22822222 00.0 00 082 22822290
02.2 00 2.82 2282208... 002 00 082 22822222... 002 00 2.82 502208... 00.0 00 2.82 2282205...
002 00 2.82 2282205. 00.2 00 082 502208.... 00.2 00 2.82 802208... 00.0 00 2.82 22822222...
002 00 2.82 2280208... 002 00 0002 2282.205. 002 00 082 2282.205... 00.0 2.0 082 2282.222...
000 00 082 2280.08... 2.2 00 082 228238.... 00.2 00 082 2282205... 02.0 2.0 082 2282208...
00.0 00 082 802208... 02.2 00 082 228.2208... 00.2 00 082 2282.25. 00.0 2.0 082 228205...
00.0 00 2.82 2282.282 02.2 00 082 082.208... 00.2 00 082 22822222.... 00.0 2.0 2.82 5.022222
02 .0 00 2.82 22822222... 02.2 00 082 30202222.. 00.2 00 082 502225. 00.2 2.0 0002 2.82225.
00.0 00 2.82 802225.. 02.2 00 082 22820222... 002 00 2.82 2200220222... 002 2.0 002 22822222
208022 022282 2022.00 2825 20822 02.25 0250200 222225

2220203 .50 22202 825228 2220203 2>0 08072 22082228 2220203 2>0 0802 22022288 2220203 2>0 082 2202222228

 

2.0000 2.0 020222222...

92

 

 

and 2m 2002 5002.203 oo.~ on 20002 22002022. ovd om 2009 500202.52. oo.~ mm 009 5002208222
end 5 2008 2302.58.42 end On 203 2200220250 omd om 39 22.002.208.00 oo.~ mm 2002 28.220250
end 5 2009 0200208222 ofiu on v09 2802.205 S .N am 2002 280202.220 82 mm 2009 22002208220
om.~ : 2009 9002208202 own on 2009 28020252. 2 .N on 008 22002208220. 9: mm 0002 2200220822..
omd 2m 909 50208.0. ou.~ on 209 502.2092. cod am 009 2200202322 00.2 mm 2002 220020820.
0 2 .N 2 m B28 2002.—0822.. end on 209 500208.22. 82 am .009 2802.20.50. om .2 mm 282 220022022.
2 .N 220. 20002 5002.20.02. and on “208 2802.20.22. 8.. on 22.52 200220820. om; 0m 209 2282.20.25.
oo.~ 2m 209 220020822. ofim on 2082 2200220820.. 00.2 am 20002 280220820. 02... mm 2008 2200220222..
cod 5 2202 080.2022. 2 .N on 202 200203 ow; mm 2082 280202922. on; mm @002 220022022220.
om; 2m 2009 2200202220 oo.~ on 008 200208.22. ow; on 2009 2802208222. o~._ mm 209 222002.223.
022 R 2009 502,208.20 oo.~ on 2008 502,208.02 on.— om 2209 50:08.2... om; mm 2082 22002208222.
2: 2m 39 500220820. so; on 229 5002082.. 2..— o~ 2009 502202.52. oo.~ 0m 209 500220252.
8.2 2.0. 2008 230208222 cm; on 2009 802.2023. 2... am 2008 9002208420. Sum 0m @209 220.02.208.02
and cm 003 22002023.. or: on 909 500220290. 022 am 2003 2200220252. onN R 208 2200.20.90.
8% cm 2008 502.2022. 00.2 on 2208 2200220290. om; am 20208 22002208222. mod 0m 202 50208.4.
and on 909 50.208222. 8.2 on 2082 280220222. om; am 209 2200220250. cod pm 20208 500220820.
and on 39 90020250 00.2 on 209 50022025.. omd 0.0. E02 2282.20.52. cod 0m 208 28020820..
om.m om 208 982.2084 c: om 082 2200220820.. ow.~ mm 2009 228220250. om.~ R 209 28020822..
cud om 38 3020820. 9.6 am 2082 22.002.208.20 owd mm 2009 280220280. Pad 2 209 22820822..
2 .n on 2008 22.002.208.20. o_.m on 2008 500220222. Sum mm 002 28020050.. Sam R 20002 50208.0.
cod cm 0202 0200202220 cod am 008 2200220822.. oo.~ mm 2082 220020820. ovd R 0002 22002208222.
cad on 20229 0802.20.02. cod am 2008 2200220822.. 8.“... 0.0. 2009 22002208202 omd R 2002 2200220250
oa.~ on 2009 220020252. cod an 2082 0200220820. end mm 0002 2200220222. omd 0m 982 280220820.
cad om 39 280202.922. cod on 0002 220020222 ov.~ mm c208 220020222. 2 .N R 0002 228220222.
8N on was 9002.20.32. cad mm @208 2220020250 ovd mm 982 50020252 no.2 R 082 502.2022.
end on 2009 502.2084 mud mm 2082 220020252. omd mm 202 22200220050. co; 0m 209 500208.22.
oo.~ cm 2008 50208.00 cod am 2009 0200208220 c 2 .N mm 009 500220222. ow; R 202 5002.20.22.00
cod on 209 280208200 end am 008 500220820. 8:0. mm 2009 52.202520 ow; 0.0. 2008 2220020822..
cod on 2003 520208.22. ovd am 2009 800208.00 oo.~ mm 2209 50020822.. 2..— nm 2002 2200220820.
208022 222222 208022 222280 2022222 222225 208022 02225

3303 A>m 0802 288800 320? 15m 0802 2202:2280 222203 2>m 0802 2208800 2.2203 A>m 0802 2882.280

 

2.228 2.0 020522222.

93

 

 

26 mm 009 809.2080. cad 0m 009 800208.220 Sam 0m 009 8000208220 ohm mm 009 2200020820..
2 .n mm 009 800208.20 36 on 009 800208200 0nd 0m 009 809208.20. own mm 009 809208.00
omé mm 009 80920820. cm...” on 009 800020820 om.m 0m 009 80020820. om.m mm 009 800.2080.
and. mm 009 800208.00 Sam on 009 80020822.. 00m 0m 009 809208.00 o_ .m mm 009 2200208222.
3.0 mm 009 800208.00 ova 0m 009 800208220 end 0m 009 8002208200 ood mm 009 200208.22.
8.0 mm 009 80020822.. and on 009 809208222. own 0m 009 220002080. cod mm 009 2002080.
on;. mm 009 8002080. 3.6 m m 009 2209.88.20 on. m 0m 009 000208.20 cod mm 009 220920820.
ové mm 009 809.2080. cod m m 009 80020822.. 2 .m 0m 009 800208.220 an .N mm 009 220020820.
omé mm 009 8002080.. cod mm 009 2200020820 oo.m 0m 009 22002082220 omd mm 009 2200220822.
8.0 mm 009 22.002.208.20 8.0 mm 009 80020822.. owd 0m 009 22002208402 om.m mm 009 2209208400
ova mm 009 809208.220 90.0 m m 009 2200020820.. on .20 m m 009 22000208420 oo.~ mm 009 2200220822..
9.6 mm 009 809.2080. omé mm 009 800208.220 ooé mm 009 0002080.. 00.2 mm 009 209208.00
and 0m 009 80020820. A: .0 mm 009 0002.205 cod mm 009 8000208200 own 3. 009 200.2080.
O06 0m 009 800020844. ooé mm 009 809.2080. 02m mm 009 2200208220 c— .m _m 009 220022080.
om.n 0m 009 809.2080. 22.0“ mm 009 800208222V and mm 009 800208.00 cod 3 009 220022080.
cod 0m 009 2200208202 Sum mm 009 800208.02 o~.m mm 009 800208.202 own 20. 009 220002080.
2.0 hm 009 809.2080. cod mm 009 8002208220 cod mm 009 800220800 00.0. S 009 8002082..
90.0 nm 009 809.2082 2mm mm 009 2200220820. om.~ mm 009 220020820. 02m 3 009 22002084..
8.0 hm 009 809.2080. oc.w 0m 009 800208220 00.2 mm 009 2209208220 oflm 2m 009 22002208200
22m 0m 009 809208202 2% 0m 009 220020800 end mm 009 800208.222 02 .m _m 009 809.2080.
02m 0m 009 809.2080. 02 .m 0m 009 8092082220 ooé mm 009 2209208220 0 2 .m 2 009 2200220820..
cs. 0.0. 009 809208.222 2 .m 0m 009 2209208220 cad mm 009 80920820. 02 .m _m 009 809.2080.
8.0 on 009 800208.00 cod 0m 009 220920802 owd mm 009 809208.00 cod 3 009 2200208202
and on 009 80920800 co... 0m 009 2.09.2080. ow...” mm 009 220920824. oo.~ 2m 009 8002080.
00.0 cm 009 809208.00 om... 0m 009 800208.220 own mm 009 809208.202 2am _m 009 2200220820.
Om.m 0m 009 2200208222. 3.0 0m 009 200020250. ova mm 009 2200020820. O00 2m 009 =0ot08<
own on 009 80920820. ooé 0m 009 200208220 00.20. mm 009 2200208220 cod 2m 009 220920800
36 cm 009 800208.00 cod 0m 009 80020800 02m mm 009 800208.00 oflm _m 009 200208.20.
00.2. on 009 8002080. 2.0. 0m 009 200208.220 02m mm 009 800208.40 ovd 2m 009 220920820.
A0883 A882 008023 4882 008023 A882 A0883 A805

2203 A>w 0802 258800 3203 A>m 0802 2208800 222203 4>m 0802 288800 222203 2>m 080 2 258800 .

 

2.2.8 2.0 0202.22.22.

94

 

 

8.2 2 O0 009 8008080. 8.8 00 009 8008082 8.0 m0 009 8008080. 8.0 20 009 800808<
8d O0 009 800808< O06 00 009 8008080. 8.0 m0 009 800808< 8.x O0 009 800808<
Ond Q0 009 800808< 8.0 00 009 800808< O00 m0 009 800808< O0.0 O0 009 800808<
8.8 00 009 800808< 8.8 00 009 800808< On .0 m0 009 80080820 O2 .0 O0 009 800808<
8.2 80 009 800808< O00 00 009 800808< 8.0 m0 009 800808< O2 .0 O0 009 800808<
On .02 80 009 8008080. On .2 2 m0 009 800808< 8.2 m0 009 800808< O 2 .0 O0 009 800808<
8.2 80 009 800808< On .2 2 m 0 009 83808.0. O06 N0 009 800808< On .0 O0 009 800808<
8.0 2 80 009 800808< 8.2 2 n0 009 800808< O06 N0 009 800808< On .0 O0 009 800808<
8.2 80 009 800808< OO. 2 2 m 0 009 800808< On .0 N0 009 800808< 8.0 O0 009 800808<
8.2 m0 009 800808< 8.3 m0 009 800808< O06 m0 009 800808< 8.0 O0 009 800808<
8.2 80 009 800808< 8.2 m0 009 800808< O2 .0 N0 009 800808< O0.m O0 009 800808<
8.2 80 009 80080880. 8d 00 009 800808< O08 N0 009 8008084 O00 O0 009 800808<
On .22 80 009 800808< 8.0 m0 009 800808< 8.0 m0 009 800808< 8.0 O0 009 800808<
8.2 2 80 009 8008082 8.x 0 0 009 8008080. 8.0 N0 009 800808< O2.m O0 009 8008080.
8.2 80 009 8008080. 8.2 00 009 8008080. O2 .0 m0 009 8008084 Oo.m O0 009 800808<
8.2 00 009 800808< 8.2 00 009 800808< 8.0 20 009 800808< 2.0 on 009 800808<
8.2 00 009 800808< 8.0 00 009 8008080. 8.0 20 009 800808< 8.0 mm 009 800808<
On .2 2 00 009 8008080. 8.8 00 009 800808< On .8 20 009 800808< 8.0 8 009 800808<
8.2 2 00 009 800808< O0.» 00 009 800808< 8.8 20 009 800808< 8.0 on 009 800808<
8.22 00 009 800808< 8.8 00 009 800808< 8.0 20 009 800808< 8.0 on 009 800808<
O0.@ 00 009 80080820 8.8 00 009 8008080. 8.0 20 009 800808< O 2 .0 on 009 800808<
82¢ 00 009 8008080. On .0 00 009 8008080. O0.0 20 009 800808< O0.m am 009 800808<
O0.” 00 009 800808< 8.0 00 009 800808< O0.0 20 009 800808< Om.m 8 009 800808<
O08 00 009 800808< 8.0 00 009 800808< 8.0 20 009 800808< 8.0 mm 009 800808<
O28 00 009 800808< 8.0 00 009 800808< O0.0 20 009 8008080. 8.0 mm 009 800808<
On .0 00 009 800808< O08 m0 009 800808< 8.0 20 009 800808< 8.0 mm 009 800808<
8.02 00 009 800808< 8.8 m0 009 800808< O00 20 009 800808< Om .0 mm 009 800808<
8.2 00 009 800808< O28 m0 009 888080. Om .0 20 009 800808< 8.0 mm 009 800808<
8.3 00 009 800808< 8.8 m0 009 800808< O2 .0 20 009 800808< O2 .0 w m 009 800808<
88003 2880 88va 8880 208080 2880 208082 2880

82803 12>m 0802 808800 8303 .2>m 0802 808800 820203 2>m 08072 808800 820203 .2>m 0802 808800

 

0.80 .20 805090

95

 

 

00.0 00 #000 2:220 00.00 00 080 50825 00.02 00 080 088250 00.: 00 080 08582
00.0 08 000 2:000 00.00 00 080 508080. 00.: 00 080 508080 00.08 00 08. 0882.2
00.0 08 000 2:000 00.00 00 080 80882 00.: 00 080 80882 00.08 00 080 08882
00.0 8 000 2:000 00.3. 00 08. 30808.0 00.00 00 080 08882 00.02 00 08. 08882.
00.0 t 000 08220 00.00 00 080 50825. 00.2 00 080 08808.0 00.08 00 080 08580
00.0 t 008 202.0 00. _ 0 00 080 508080 00.: 00 08. 508080 00.: 00 080 08805.
08 .0 02 000 03220 00. 2 0 00 080 50825. 00.02 00 080 50808.0 00.02 00 080 508080
00.; 00 080 50805. 00.00 00 080 03502 00.2 00 080 50806.0 00.02 _0 080 08580
00.00 00 080 808080 00.00 00 08. 80.580 00.0_ 00 080 50808.0 00.0_ _0 08. 08805..
00.00 00 080 508200 00.00 00 080 08882 00.00 00 080 08882 00.2 _0 080 08882
00.00 _0 080 80808.0 00.00 00 080 08825. 00.00 00 080 5082.2 00.02 00 080 088200
00.00 20 080 508290 00.00 00 08. 50082. 00.00 00 080 $0825. 00.02 00 080 0855.
00.00 00 080 5058.0 00.00 00 08. 08825. 00.00 00 08, 0882.2 00.0. 00 08. 08805.0
00.3. E 080 508080 0080 00 080 50808.0 00.: 00 080 085:2 00.: 00 080 085:2
00.00 00 080 088200 00.00 00 080 50502 00.02 00 080 50805. 00.: 00 080 0858.0
00.00 00 080 50.825. 00.00 00 08. 08882 00.02 00 08. 038080 00.2 00 080 08082
00.00 00 080 088082. 00.00 20 08. 088080 00.: 00 080 0008080 00.8 00 080 088050
00.00 00 080 505:2 00.00 s 080 085:2. 00.: 00 080 03882 00.: 00 08. 08805.0
00.00 00 080 808080 00.00 80 08. 08550 00.: 00 080 0856.0 00.8 00 08. 0882.2
00.00 00 080 808002 00.2 s 080 085.5. 00.00 00 080 088002 00.02 00 080 08:82
00.00 00 080 08882 00.00 00 080 08808.0 00.: 00 080 08822 008 _ 00 080 08550
00.:. 00 080 08825. 00.00 00 08. 0882.2 00.0_ 00 080 085:2 00.: 00 08. 085:2
00.00 00 009 50825. 00.00 00 080 08808.0 00.2 00 080 508002 00.2 00 080 08825..
00.00 00 080 80882 00.02 00 08. 08808.0 00.: 00 080 50825. 00.2 00 08. 088050
00.00 00 080 808290 00.00 00 08. 50808.0 00.08 00 080 08502 00.2 00 080 0855.0
00.00 00 080 508080 00.00 00 080 088080 00.00 00 080 08805.0 00.: 00 08, 08882.
00.00 00 080 53825.. 00.00 00 080 50882. 00.8 00 080 08882. 00.: 00 080 088290
00.00 00 080 308082 00.08 00 080 508080 00.: 00 08. 508080 00.: 00 080 088050
00.00 00 080 80825. 00.02 00 08. 508080 00.2 00 08. 80082 00.08 00 080 50.580
0588 055 83800 055 88.200 055 05200 05.0

8003 0>0 252 88800 8083 .50 252 88800 8083 .50 252 008500 8003 .50 252 88800

 

0.000 8.0 800003

96

 

 

97

ovd 2 833 watam end 2 .833 watam omd 2 838Wm=23m 86 2 833 wctmm
ovd 2 833 38% end S 833 wctam 2 .o 2 833 watnm end S 833 wctnm
end 2 833 wntqm omd S 833 wctqm etc 2 833 wntam and E 833 wztaw
end 2 833 matqm end 2 833 wctqm ovd 2 833 3er end 3 833 3ch
end 2 833 3.3m end 2 833 wfiam ovd 2 833 macaw and E 833 wstqm
oflo 2 833 wqtam end 2 833 wEEm ovd 2 833 395 omd E 833 3.55
end 2 833 macaw end 2 833 mctnm ovd 2 833 mctqm end 3 833 wstam
end 2 .0383 watnm omd E 833 macaw end 2 833 wctam end 3 833 wctqm
omd 2 833 watam omd 2 833 mctmm omd 2 833 wctqm omd E 833 wctam
cod 2 833 watmm omd 2 833 wctmm end 2 833 wctam omd E 833 wctam
0nd 2 833 wctam cod 2 833 wctam end 2 833 wE3m omd 3 833 watam
end 2 833 watqm ovd 2 833 matam end 2 833 watam omd E 833 wctqm
2.6 2 833 323m ovd 2 833 wztam omd 2 833 wctnm o _ .o E 833 watam
ovd 2 833 323m ovd 2 833 matam omd 2 833 wstam o _ .o E 833 watum
ovd 2 833 wctqm ovd 2 833 mctqm 00.6 2 833 wctqm owd 2 833 33m
ovd 2 833 3.03m ovd 2 833 watam end 2 833 wctmm ovd 2 833 watqm
ovd 2 833 wctqm ovd 2 833 wctam omd 2 833 wctqm end 2 833 macaw
ovd 2 833 wctam 9.6 2 833 matam 00.6 2 833 3th omd 2 833 wctam
90.0 2 833 macaw mmd 2 833 mctmm 36 2 833 w53m omd 2 833 macaw
ovd 2 833 383m mmd 2 833 wctmm and 2 833 watqm omd 2 833 33m
ovd 2 833 wfi3m end 2 .833 macaw and 2 833 watmm omd 2 833 wctam
ovd 2 833 wfi3m omd 2 833 wctqm end 2 833 matqm m _ .c 2 833 wctam
ovd 2 833 warn—m omd 2 833 m53m end 2 833 watqm ovd Z 833 watqm
ovd 2 833 wntam omd 2 833 watam 00.6 2 833 wctqm Omd : 833 watam
omd 2 833 watam omd 2 833 wctam omd 2 833 watam and _ 2 833 watqw
omd 2 833 wfi3m end 2 833 wztam end 2 833 watam 002 pm wow 0325
end 2 833 watam end 2 833 macaw omd 2 833 watam owd mm wofi 038:0
00.0 t 3080 00:00 00.0 8 5030 0580 00.0 0_ 5080 00:00 00.0 00 000 0398
end 2 833 watqm omd 2 833 wstmm omd 2 833 macaw 00.0 _N wo¢ 038:0
000:va A805 8ng G55 00883 $55 20:8qu A805

2383 A>m 8802 38800 2383 q>m 8802 38800 3383 A>m 8802 38:80 :38? A>m 8802 88800

 

0.08 0.0 80503

 

00.0 00 I083.“ 80 00.0 02 00030 .80 00.0 00 50808800 00.0 00 5030 00:00

 

00.0 G 00080 080 00.0 S 00080 080 8.0 00 5080 0800 02 .2 02 3080 8:00
and 2,. 38.880 >80 2.6 2.2 30.880 >80 2.6 mm 833 323m end 2 833 323m
CNN on 30.880 >80 end 2 wodoob >80 2.0 mm 833 wctqw 2.6 2 833 323m
2 .N am 3.2.288 >80 end 2 30.28.: >80 006 mm 833 wctmm cod 2 833 matam
2 .N am 3.2.2888 >80 2 .2 am 833 323m omd mm 833 323m cod 2 833 323m
8:0. mm 30.288 >80 2 .2 mm 833 323m omd mm 833 323m cod 2 833 88%
cod mm 38.28.: >80 00.2 R 833 323m 32 2m 833 38% omd 2 833 32.3w
32 mm 3.0.2888 >80 2 .2 mm 833 98% o 2.2 22.. 833 323m omd 2 .833 2.23m
20.2 2.. 3.2.2888 >80 cad on 833 323m 2 .2 2m 833 32.3w ovd 2 833 wctqm
22.2 cm 30.2on >80 oo.~ mm 833 323m 82 2N 833 323m ovd 2 833 38%
8.2 mm 30.880 >80 N222 mm 833 323m 8.2 2m 833 323m 2.6 2 833 wc23m
2 .2 2.. 30.2888 >80 82 mm 833 323m 226 2m 833 32.3w ovd 2 833 323m
82 mm 3.2.2880 >80 2 .2 mm 833 wqtam Ed 2N 833 wctqm mmd 2 833 38%
who mm 3.2.28.5 >80 22.2 mm 833 323m cod 2N 833 w=23m end 2 833 wctqm
cod mm womoob >80 05o mm 833 323m cod 2m 833 323m end 2 833 323m
cod 202 3.2.2880 >80 00.2 g 833 323m mmd 2.02 833 323m end 2 833 323m
cod 2m 30.2880 >80 2 .2 3 833 323m omd 2m 833 38% 00o 2 833 323m
nmd 2m 3.2.2.88 >80 25.2 vm 833 war—am 2.0 2m 833 3th end 2 833 watqm
and 2N woboob >80 cad mm 833 323m 8.2 2.. 833 323m o~.o 2 833 323m
cod Om 30.888 >80 cod mm 833 323m 3.0 on 833 323m 2.6 . 2 833 3th
cod on 30,880 >80 36 mm 833 88% and cm 833 32.3w cod 2 833 323m
00.0 2 woaoob >80 02; mm 833 323m 006 on 833 323m omd 2 833 323m
cod 2 388.3 >80 owd mm 833 323m omd ON 833 323m 006 2 833 323m
006 2 woaoob >80 oed mm .833 323m omd on 833 323m end 2 833 watqm
cod 2 woboob >80 2 .2 mm 833 323m omd on 833 323m 206 2 833 28%
00.0 02 0008b 80 00.0 00 3080 0800 00.0 00 588 8:00 00.0 02 3080 0800
end 2 30.2on >80 owd mm 833 323m ovd cm 833 323m ovd 2 833 323m
and 2 30.2888 >80 owd mm 833 323m ovd cm 833 323m ovd 2 833 323m
20883 2882 288% >882 888% 2882 20882 2882

82883 .2>m 8802 088800 22983 5.6 8802 38800 82883 .2>m 88072 88800 022M283 .2>m 8:82 88:80

 

:50 ._.0 £502

98

 

00.0. 00 000 0085 00 .0 0.0 000 008.5 00.0 00 000 520 00.0 00 000 520

 

00.0. 00 000 820 00.0 0.0 000 =38 00.0 00 000 085 00.0 00 020 085
00.0 00 000 0085 00.0. 0.0 02.0 08.6 00.0 00 0000 52c 00.0 _0 020 085
00.0. 00 000 820 00.0. 0.0 000.0 =88 00.0 00 000 085 00..0 00 000 0800
00.0. 00 000 085 00..0. 0.0 000 0085 00.0 00 020 820 00.0 0 0 020 085
00.0. 00 000 085 00.0. 0.0 000 0800 00.0 00 080 085 00.0 00 020 085
00.0. 00 000 0088 00 .0. 0.0 000 520 00.0 00 000 520 00.0 _0 020 085
00.0. 00 000 520 00.0. 0.0 020 0085 00.0. 00 000 :85 00.0 0 0 0000.0 086
00.0. 00 000 0086 00.0 0.0 000 0085 00.0. 00 000 8200 00.0 _ 0 000 520
00.0. 00 020 085 00.0 0.0 000 82o 00.0. 00 020 .085 00.0 _0 000 5.08
00.0. 00 000 82o 00.0 0.0 000 52o 00.0. 00 000 0085 00.0 00 000 0085
00.0. 00 08.0 520 00.0 0.0 020 088 00.0. 00 000 :85 00.0 00 020 086
00.0. 00 000 088 00.0 0.0 000 0085 00.0. 00 02.0 80.5 00.0 00 080 820
00.0. 00 000 085 00.0 0.0 000 085 00.0. 00 000 820 00..0 00 000 085
00.0. 00 020 82o 00.0 0.0 020 0085 8.0 00 020 0086 00.0 00 020 :85
00.0 00 000 088 00.0 0.0 000 085 00.0 00 020 :85 00.0 00 00.0 085
00.0 00 000 820 00.0 0.0 080 0085 00.0 00 000 520 00 .0 00 00.0 0.85
00.0 00 000.0 008.5 00.0 0.0 000 0800 00.0 00 000 :88 00 .0 00 000 085
00.0 00 000 .085 00.0 00 000 085 00.0 00 020 820 00.0 00 080 520
00.0 00 000 0085 00.0 00 000 820 00.0 00 000 0.800 00.0 00 02.0 0800
00.0 00 020 sec 00.0. 00 020 0085 00..0 00 000 0035 00..0 00 020 0.0.0.6
00..0 00 000 008.6 00.0. 00 000 820 00..0 00 000 0085 00.0 00 000 =85
00..0 00 000 086 00.0. 00 080 0085 00.0 00 000 520 00.0 00 000 085
00.0 00 000 0085 00.0. 00 09.0 520 00.0 00 000 00.020 00.0 00 0000 085
00.0 00 000 :85 00.0 00 080 085 00.0 00 0000 520 00.0 00 000 820
00.0 00 000 820 00.0 00 080 820 00.0 00 020 .085 00.0 0.0. 08.080 020
00.0 00 080 8200 00..0 00 we... 085 00.0 00 000 0088 00.00 00. 02.08.00 0005
00 .0 00 0.000 0085 00..0 00 020 805 00.0 00 02.0 0086 00.00 00. 020800 0006
00.0 00 000 820 00.0 00 00.0 :85 00.0 00 000 0085 00.00 00. 020800 0.05
2.088% 2882 20883 2882 20883 2805 20883 2880

0.2883 2>m 8802 088800 002883 .2>m 88072 88800 022883 2>m 88072 0888000 00.2883 .2>m 88072 0888000

0.08 .00 00002000.

99

 

 

100

00.0 00 W08 :85 00.0 00 080 08.5 00.0 00 02.0 085 00..0 00 080 085
00.0 00 020 085 00.0 0.0 000 0085 00 .0 00 080 0085 00.0 00 020 :85
00.0 00 000 085 00.0 0.0 000 :85 00 .0 00 020 :85 00.0 00 00.0 520
00.0 00 000 0085 00.0 00 02.0 085 00 .0 00 020 0085 00.0 00 000 520
00.0 00 020 808 00.0 0.0 080 088 00.0 00 000 :85 00.0 00 000 085
00.0 00 0000 0800 00.0 0.0 000 0080 00.0 00 08.0 :86 00.0 00 000 000.000
00.0 00 000 038 00 .0 00 0B0 880 00.0. 00 080 520 00.0 00 000 085
00.0 00 000 0085 00.0 00 080 0086 00.0 00 020 520 00.0 00 08.0 086
00.0 00 020 826 00.0 00 020 :85 00.0 00 000 085 00.0 00 000 8.05
00.0 00 020 820 00.0. 00 000 820 00.0. 00 020 820 00.0 00 020 520
00.0 00 08.0 085 00.0 00 000 520 00.0 00 000 0.85 00.0 00 000 :85
00 .0 00 080 820 00.0 0.0 000 820 00.0. 00 020 085 00.0 00 080 :85
00.0 00 000 820 00.0. 00 02.0 0085 00.0 00 000 52o 00 .0 00 000 520
00.0. 00 000.0 085 00.0 0.0 000 0085 00.0. 00 020 088 00.0 00 000.0 805
00.0. 00 020 820 00.0 00 000 820 00.0 00 000 :85 00.0 00 000 :85
00.0 00 080 0085 00.0 00 000 82o 00.0 00 020 520 00.0 00 02.0 520
00..0 00 080 0086 00.0 00 020 :86 00.0 00 000 :85 00.0 00 080 085
00.0. 00 08.0 085 00.0 00 020 8200 00.0 00 080 0085 00.0 00 00.0 820
00.0 00 020 088 00.0 00 000.0 5200 00.0. 00 000 52o 00.0 00 02.0 085
00.0 00 020 0085 00.0 00 02.0 0088 00.0. 00 000 088 00.0. 00 000 0020
00.0 00 00.0 :85 00.0 0.0 020 820 00.0 00 000 085 00.0 00 02.0 00020
00.0 00 000 0085 00.0 00 000 820 00.0 00 000 0085 00.0 00 020 :35
00.0. 00 000 085 00.0 00 000 085 00 .0. 00 020 820 00.0 00 020 505
00.0 00 000 82¢ 00.0 00 020 820 00 .0 00 000 520 00.0. 00 020 085
00.0 00 000 820 00.0 00 000 820 00.0 00 000 0088 00.0 00 000 520
00.0. 00 0000 880 00.0 00 000 820 00.0 00 000 08.0 00.0 00 000 520
00.0 0.0 080 820 00.00 00 000 520 00.0 00 020 0.85 00.0 00 020 0088
00.0. 00 0000 8200 00.0 00 08.0 008.5 00.0 00 000 820 00.0 00 000 :85
00.0 00 020 085 00.0 00 000 085 00.0 00 000 820 00.0. 00 020 0088
008821 2882 4088.3 2882 20883 2882 20883 2882

002883 12>w 88072 088800 002883 A>m 8802 088800 002883 2>m 88072 808800 022883 .2>m 8802 088800

 

0.008 .00 00050000.

 

00.0 00 0.000 0088 00.0 00 020 :85 00.0 00 000 00905 00.0 00 000.0 085

 

00 .0 00 02.0 820 00.0 00 080 085 00.0 00 000 085 00.0 00 08.0 085
00.0 00 000 0035 00.0 00 020 0086 00.0 00 020 085 00.0 00 020 035
00 .0 00 020 0085 00.0 00 000 0800 00.0 00 020 83o 00.0 00 020 085
00.0 00 000 008.00 00.0 00 080 00020 00.0. 00 000 82o 00.0. 00 000 0085
00.0 00 000 820 00.0 00 000 085 8.0 00 080 520 00.0 00 00.0 086
00.0 00. 000 085 00 .0 00 0e... 8200 00.0 00 02.0 820 00.0 00 02.0 :88
00.0 00 000 820 00 .0 00 000 820 00.0 00 000.0 085 00.0 00 020 830
00.0 00. 000 820 00 .0 00 08.0 085 00..0 00 080 820 00.0 00 020 :85
00.0 00 0000 085 00.0 00 000 0006 00.0 00 08.0 00.05 00.0 00 000 820
00.0 00 000 0085 00.0 00 020 0085 00.0 00 080 0.800 00.0. 00 000 08.5
00.0 00. 0000 820 00.0. 00 020 :85 00.0 00 000 0085 00.0 00 0000 085
00.0 00. 000 820 00.0 00 020 085 00.0 00 020 085 00..0. 00 020 820
00.0 00. 0000 820 00.0. 00 000 :85 00.0 00 020 085 00.0. 00 0000 085
00.0 00. 0000 820 00.0 00 08.0 820 00.0 00 080 820 00.0 00 000 520
00.0 00. 000 0085 00.0 00 020 088 00.0 00 000 820 00.0 00 000 :85
00.0 00 000.0 52o 00.0. 00 000 008000 00.0 00 080 0085 00.0. 00 000 085
00.0 00 0000 82o 00.0 00 020 085 00.0 00 02.0 086 00.0 00 020 :85
00.0 00 020 82o 00.0 00 02.0 085 00.0 00 000 0085 00 .0 00 020 5200
00.0 00 020 0085 00.0 00 000 0085 00.0 00 000 0006 00.0 00 080 085
00.0 00 000 0085 00.0 00 000 820 00..0 00 000 8.0.6 8.0. 00 020 08.5
00.0 00 020 .085 00.0. 00 02.0 085 00.0 00 000 820 00.0 00 020 520
00.0 00 020 0085 00.0 00 000 88o 00.0 00 000 820 00.0 00 000 085
00.0 00 000 82o 00.0 00 02.0 0085 00.0 00 02.0 0088 00.0 00 020 085
00.0 00 020 00020 00.0 00 02.0 880 00.0 00 000 035 00.0 00 000 088
00.0 00 000 0020 00.0 00 000 0088 00.0 00 000 0085 00.0 00 000 085
00.0 00 0000 0085 00.0 00 08.0 0088 00.0 00 02.0 0085 00.0 00 000 820
00.0 00 020 820 00.0. 00 080 0020 00 .0 00 02.0 085 00.0 00 020 520
00.0 00 080 820 00.0 00 000 085 00.0 00 080 085 00.0 00 020 0086
208800 0.80 4080000 2080 2000800 080 28023 2080

0020283 2>m 88.02 088800 022883 12>w 88072 88800 0.2883 .2>m 88.02 088800 002883 .2>m 8802 058800

 

0.08 0.0 00058.0

101

 

00.0 00 |000 0085 00.0 00 80 086 00.0 00 090 .085 00.0. 00. 000 085

 

00.0 00 000 085 00.0 00 000 820 00.0 00 000 820 00.0 00 000 :85
8.0 00. 000 085 00.0 00 000 0085 00.0 00 000 0088 00.0 00. 000 :85
00.0 00. 000 0088 00.0 00 000 820 0E 00. 000 085 00.0 00. 000 085
00.0 00. 000 0085 00.0 00 000 .085 00.0 00 000 0005 00.0 00 000 :85
00.0 00. 000 0085 8.0 00. 000 00005 00.0 00 000 085 00.0 00 000 :85
00.0 00. 000 0088 00.0 00 000 0800 00.0 00 000 520 00.0 00. 000 :88
00.0 00. 000 820 00.0 00. 000 0085 8.0 00 000 :85 00.0 00. 000 820
00.0 00 02.0 085 00.0 00 000 820 00.0 00 000 085 00.0 00 000 :85
00.0 00. 000 085 00.0 00 000 085 00.0 00 000 0085 00.0 00 000 :85
00.: 0.0. 000 82o 00.0 00. 000 520 00.0 00 000 0085 00.0 00 000 .085
00.0 00 000 085 00.0 00 000 880 00.0 00 000 0085 00.0 00. 02.0 :85
00.0 00. 000 820 00.0 00 000 08.5 00.0 00. 000 :85 00.0 00. 000 0085
00.0 00 000 08.5 00.0 00 000 520 00.0 00. 000 085 00.0 00 000 0085
00.0 00 000 808 00..0 00 000 820 00.0 00. 000 085 00.0 00 000 .085
00.0 0.0. 000 0085 00 .0 00. 000 820 00 .0 00. 000 085 00.0 00 000 085
00.0 00 000 820 00.0 00 000 085 00 .0 00. 000 82o 00.0 00 000 0085
00.0 00 000 0085 00.0 00 000 820 00.0 00 000 085 00.0 00 000 820
00.0 0.0 000 880 00.0 00. 000 0085 00.0 00 000 820 00.0 00 090 :85
00.0 00 000 088 00.0 00. 000 :85 00.0 00 000 0085 00.0 00 000 82o
00.0 0.0. 000 820 00..0 00 000 00020 00.0 00. 000 :86 00.0 00. 000 820
00.0 00 000 820 00.0 00 000 00020 00.0 00. 000 82o 00.0 00. 000 520
00.0 00 000 086 00.0 00 000 085 00.0 00 000 820 00.0 00. 000 :85
00.0 00. 000 085 00.0 00 000 805 00.0 00 000 0085 00.0 00 000 085
00.0 00. 000 00020 00.0 00. 08.0 085 00.0 00 000 0800 00.0 00. 000 .085
00.0 00 000 820 00.0 00. 000 00020 00.0 00 000 0005 00.0 00. 000 085
00.0 00. 000 8200 00.0 00 000 505 00.0 00 000 :88 00.0 00. 000 088
00.0 00. 000 0085 00.0 00 000 820 00.0 00. 000 820 00.0 00 000 :85
00.0 00 000 820 00.0 00 000 0088 00.0 00 000 086 00.0 00 000 506
20:83 080 0000800 2880 2080000 085 205800 055

002883 a2>m 8802 38800 002883 12>m 88072 38800 002083 2>m 88.02 088800 0.2083 x2>m 88072 38800

0.008 0.0 0000200...

102

 

 

00 .0 00 000 E0300 .2 00.0 00 0.00 E082 .2 00.0 00 000 000080 .2 00.00 00 0.00 085
00.2 00 000 000082 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.00 00 000 :85
00.0 00 000 E0002 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.: 00 000 0088
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.0 00 000 0085
00.0 00 000 0.00080 .2 00.0 00 000 00008. .2 00.0 00 000 000080 .2 00.00 00 000 085
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.0 00 000 0.00080 .2 00.00 00 000 08.5
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.: 00 000 :85
00.0 00 000.0 00.0080 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.00 00 000 008.6
00.0 00 020 000002 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.0 00 000 820
00.0 00 000 00.0080 .2 00.0 00 000 000080 .2 00.0 00 000 00008. .2 00 .0 00 000 8200
00.0 00 000 E0002 .2 00.0 00 000 000080 .2 00.0 00 000 E0080 .2 00.00 00 02.0 0020
00.0 00 000 E0082 .2 00.0 00 000 000080 .2 8.0 00 000 0000000 .2 00.0 00 000 085
00.0 00 000 E0080 .2 00.0 00 080 000080 .2 00.0 00 000 000002 .2 00.00 00 000 085
00 .0 00 000 0000000 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.: 00 000 086
00 .0 00 000 000080 .2 00.0 00 000 000080 .2 00.00 00 000 085 00.0 0 00 000 :85
00.0 00 000 000002 .2 00.0 00 000 000080 .2 00.00 00 000 805 00.00 00 000 820
00.0 00 000 E0080 .2 00.0 00 000 E0080 .2 00.00 00 000 088 00.0 00 000 00020
00.0 00 000 E0002 .2 00.0 00 000 0.0080 .2 00.00 00 000 805 00.: 00 000 0085
00.0 00 000 000080 .2 00..0 00 000 0000000 .2 00.00 00 000 085 00 .0 00 000 505
00.0 00 000 000080 .2 00.0 00 000 000002 .2 00.00 00 000 088 00 .0 00 000 0005
00.0 00 000 000082 .2 00.0 00 000 0000000 .2 00.00 00 000 820 00.0 00 020 088
00.0 00 000 0.00002 .2 00.0 00 000 000080 .2 00.00 00 000 0088 00.0 00 000 505
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.00 00 000 8200 00 .0. 00 000 520
00.0 00 000 0.0080 .2 00.0 00 020 000080 .2 00.00 00 020 820 00.0 00 000 085
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.0.0 00 000 520 00.0 00 000 00800
00.0 00 000 E0080 .2 00.0 00 000 0.0002 .2 00.00 00 000 82o 00.0 00 000 820
00.0 00 000 000080 .2 00.0 00 020 000080 .2 00.: 00 000 82o 00.0 00 000 820
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.00 00 000 0088 00.0 00 000 520
00.0 00 000 000080 .2 00.0 00 000 000080 .2 00.: 00 000 085 00.0 00 000 086
2% 20800 005200 4080 2080080 2080 205800 080

002883 ..2>m 88072 088800 002883 .2>m 88072 38800 002883 .2>m 8802 38800 002883 2>m 880 Z 088800

 

0.000 .00 0005000.

103

 

2.2 2 wow 20003 cad 2.2 mob 2303 00.0 20 onb 208302 .2 end mm wot 283082 .

2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 E0080 .2 00.0 00 000 E0080 .2
00.0 00 000 083 00.0 t 000 0083 00.0 00 000 000080 .2 00.0 00 000 E0080 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 08.0 000080 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 t 000 083 00.0 00 000 000080 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 t 000 083 00.0 00 000 000080 .2 00.0 00 000 0000000 .2
00.0 00 000 0083 00.0 t 000 083 00.0 00 000 000080 .2 00.0 00 000 000080 .2
00..0 00 000 083 00.0 t 000 083 00.0 00 000 000080 .2 00.0 00 000 E0080 .2
00.0 00 000 083 00.0 t 000 083 00.0 00 000 0.0080 .2 00.0 00 000 0.0080 .2
00.0 00 000 083 00.0 0.0 000 083 00.0 00 000 000080 .2 00.0 00 000 000080 .2
00.0 00 000 0003 00.0 S 000 083 00.0 00 000 000080 .2 00.0 00 02.0 0000000 .2
00.0 00 000 08>» 00.0 t 000 083 00.0 00 000 000002 .2 00.0 00 000 0.0080 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 E0002 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 020 0.0080 .2 00.0 00 000 E0080 .2
00.0 00 000 083 00.0 t 000 083 00.0 00 000 000002 .2 00.0 00 000 000080 2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 000080 .2 00.0 00 000 0:008. .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 000080 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 E0080 .2 00.0 00 000 E0080 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 000080 .2 00.0 00 000 E0080 .2
00.0 00 000 0.83 00.0 00 000 083 00.0 00 000 0.0080 .2 00.0 00 000 00000.... .2
00.0 00 000 083 00 .0 00 000 083 00.0 00 000 E0080 .2 00.0 00 000 000002 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 000080 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 E0002 .2 00.0 00 000 E0080 .2
00.0 00 02.0 0.83 00.0 00 000 083 00.0 00 000 000080 .2 00.0 00 000 000002 .2
00.0 00 000 083 00.0 00 000 083 00.0 00 000 0.0080 .2 00.0 00 000 00008. .2
00.0 00 000 083 00.0 00 000 000080 .2 00.0 00 000 E0002 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 00. 000 000080 .2 00.0 00 000 000080 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 00 000 000082 .2 00.0 00 000 E0080 .2 00.0 00 000 000080 .2
00.0 00 000 083 00.0 00 000 E0080 .2 00.0 00 000 0.0080 .2 00.0 00 000 000080 .2

 

2801.03 2882 A832 8082 $883 4882 20883 2880
002883 12>m 8802 808800 002883 A>m 88.02 808800 002883 .2>m 88072 38800 002883 .2>m 880072 088800

 

0.0000 0.0 0005000.

104

 

00.0 00 new 083 00.0 00 wow 083 8.0 00 wé 083 8.0 2 88 803

 

00.0 0N men 083 00.0 00 wow 083 8.0 00 ”08 083 8.0 2 mg 083
00.0 00 m3 083 8.0 00 me“ 083 8.0 00 wow 803 8.0 2 mg 083
00.0 00 woa 083 8.0 00 web 803 8.0 0N wow 083 8.0 2 m8 083
00.0 00 New 083 8.0 00 me“ 083 8.0 00 won 803 8.0 2 25 083
00.0 00 m8 083 8.0 00 m8 083 8.0 00 web 803 8.0 2 000 803
00.0 00 wow 083 8.0 00 25 083 8.0 00 wow 083 8.0 2 000 083
00.0 00 wow 08? 8.0 00 mob 803 00.0 00 mew 083 8.0 2 008 083
00.0 00 m8 083 8.0 00 020 083 82 2 ”08 083 8.0 2 0.08 803
00.0 00 web 083 8.0 00 «08 083 02 2 008 083 8.0 2 ”80 803
00.0 00 mob 083 8.0 00 wow 083 02 2 2:0 083 8.0 2 25 083
00.0 00 way 083 8.0 00 mew 083 0: 2 we“ 08? 8.0 2 me“ 0003
00.0 00 not 083 8.0 0N mew 08B 002 2 mew 083 8.0 2 28 083
00.0 om wow 08? 8.0 00 new 083 00.0 2 wow 08? 8.0 2 mew 083
00.0 00 m8 803 8.0 00 mg 803 00.0 2 25 083 8.0 2 W088 083
00.0 00 me... 083 8.0 0N 25 083 8.0 2 wow 083 8.0 2 me“ 803
00.0 0.0. mow 803 8.0 00 0.08 083 00.0 2 me... 083 8.0 2 25 083
00.0 00 008 083 8.0 00 mom 083 00.0 2, web 083 8.0 2 ”08 083
00.0 00 m8 083 8.0 00 we... 803 00.0 2 me.“ 083 8.0 2 25 083
00.0 00 not 083 8.0 00 web 083 00.0 2 wow 083 8.0 2 25 803
00.0 00 woa 083 8.0 00 m8 803 00.0 2 me“ 803 8.0 2 ME 083
00.0 00 08 083 8.0 00 25 083 00.0 2 wow 083 8.0 2 m3 083
00.0 00 wow 083 8.0 00 web 083 00.0 2 mg 083 8.0 2 web. 083
00.0 00 wow 803 8.0 00 wow 083 00.0 2 28 083 8.0 2 wow 083
00.0 00 M08 083 8.0 00 mg 0003 00.0 2 wot 803 8.0 2 mg 803
00.0 00 new 083 8.0 on «08 083 00.0 2 020 083 8.0 2 ”08 803
00.0 00 m3 083 8.0 00 808 083 00.0 2 wow 083 8.0 2 M00 083
00.0 00 web 803 8.0 00 25 083 00.0 2 30 083 8.0 2 25 083
00.0 00 wow 083 8.0 00 m8 083 00.0 2 wow 083 00.0 2 88 083
808% 880 8080 88 482.80 800 $880185

£303 A>m 0802 808800 8903 4>m 0882 8800 Emma? A>m 0802 808800 5803 a>m 0802 808800

 

0.000 .3 8503

105

 

 

106

00.0 _m Jmob 083 00.0 5 #ob 083 8.0 _m 8b 083 00.0 00 wob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 00.0 00 mob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 00.0 00 mob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 00.0 00 mob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 80 00 mob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 8.0 0m mob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 8.0 00 mob 083
00.0 a mob 083 00.0 a mob 083 8.0 a mob 083 8.0 om mob 083
00.0 a mob 083 00.0 a mob 083 0b .0 om mob 083 8.0 00 mob 083
00.0 a mob 083 00.0 a mob 083 0: 0m mob 083 8.0 0m mob 083
00.0 a mob 083 00.0 a mob 083 8.2 om mob 083 00.0 00 mob 083
00.0 a mob 083 00.0 a mob 083 02 om mob 083 8.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 00.2 00 mob 083 00.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 03 om mob 083 8.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 00.2 om mob 083 8.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 02 00 mob 083 8.0 0m mob 083
00.0 a mob 083 8.0 a mob 083 02 00 mob 083 00.0 om mob 083
00.0 a mob 083 8.0 a mob 083 02 00 mob 083 8.0 0N mob 083
00.0 a mob 083 8.0 a mob 083 02 0m mob 083 80 00 mob 083
00.0 a mob 083 8.0 a mob 083 03 00 mob 083 8.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 0: 0m mob 083 8.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 02 om mob 083 3.0 00 mob 083
00.0 a mob 083 8.0 a mob 083 002 00 mob 083 00.0 om mob 083
00.0 a mob 083 8.0 a mob 083 002 0m mob 083 00.0 00. mob 083
00.0 a mob 083 8.0 a mob 083 002 om mob 083 00.0 0m mob 083
00.0 a mob 083 8.0 a mob 083 00.0 om mob 083 00.0 om mob 083
00.0 a mob 083 8.0 a mob 083 00.0 om mob 083 00.0 0m mob 083
00.0 a mob 083 8.0 a mob 083 00.0 00 mob 083 00.0 0m mob 083
00.0 a mob 083 8.0 a mob 083 00.0 00 mob 083 00.0 00 mob 083
@80wa A805 888% A88V €80wa A88V $8va 0805

obmbo3 b>m 882 8888 obmbo3 b>m 282 08:80 obm_o3 b>m ooaz ooaeoo 0883 b>m oeoz ooeaou

 

room .00 80083

 

8.0 mm imob 083 002 20 mob 083 00.0 a mob 083 00.0 8 mob 083

 

8.0 mm mob 083 82 a mob 083 00.0 a mob 083 00.0 20 mob 083
8.0 mm mob 083 82 20 mob 083 00.0 20 mob 083 8.0 20 mob 083
8.0 mm mob 083 002 20 mob 083 00.0 20 mob 083 8.0 a mob 083
8.0 mm mob 083 002 20 mob 083 00.0 a mob 083 8.0 20 mob 083
0mm 20 mob 083 002 20 mob 083 00.0 20 mob 083 8.0 2m mob 083
00.0 20 mob 083 022 a mob 083 00.0 R mob 083 8.0 S mob 083
002 20 mob 083 022 a mob 083 00.0 20 mob 083 8.0 20 mob 083
002 20 mob 083 022 20 mob 083 00.0 a mob 083 00.0 :2 mob 083
002 20 mob 083 022 2b mob 083 00.0 20 mob 083 8.0 a mob 083
82 20 mob 083 002 20 mob 083 00.0 a mob 083 00.0 a mob 083
002 20 mob 083 002 a mob 083 80.0 20 mob 083 8.0 20 mob 083
82 20 mob 083 002 20 mob 083 8.0 2.0. mob 083 8.0 20 mob 083
82 a mob 083 002 20 mob 083 00.0 20 mob 083 8.0 20 mob 083
82 20 mob 083 002 a mob 083 8.0 20 mob 083 8.0 a mob 083
82 a mob 083 002 20 mob 083 8.0 20 mob 083 8.0 a mob 083
82 a mob 083 002 20 mob 083 8.0 a mob 083 8.0 a mob 083
82 20 mob 083 002 20 mob 083 8.0 20 mob 083 00.0 20 mob 083
82 a mob 083 002 20 mob 083 8.0 a mob 083 80.0 a mob 083
82 a mob 083 002 20 mob 083 8.0 20 mob 083 3.0 a mob 083
82 20 mob 083 002 20 mob 083 8.0 a mob 083 P0 20 mob 083
002 20 mob 083 002 a mob 083 8.0 8 mob 083 00.0 a mob 083
002 a mob 083 80.0 20 mob 083 00.0 20 mob 083 00.0 a mob 083
002 5 mob 083 00.0 20 mob 083 8.0 20 mob 083 00.0 20 mob 083
002 20 mob 083 00.0 20 mob 083 00.0 20 mob 083 00.0 20 mob 083
002 20 mob 083 00.0 20 mob 083 8.0 a mob 083 8.0 E mob 083
002 20 mob 083 00.0 20 mob 083 8.0 a mob 083 00.0 3 mob 083
002 20 mob 083 00.0 20 mob 083 8.0 20 mob 083 8.0 a mob 083
002 20 mob 083 00.0 a mob 083 8.0 20 mob 083 00.0 20 mob 083
2080.23 28v 8883 2825 208003 288V 880%2 28V

3803 A>m 0802 808800 3303 A>m 0802 808800 5803 A>m 0802 288800 2303 A>m 0802 808800

 

080 2.0. 8088.0

107

 

002 mm Jmob 083 00.0 mm mob 083 8.0 mm mob 083 80.0 mm mob 083

 

002 mm mob 083 00.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
002 mm mob 083 00.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
80.0 mm mob 083 00.0 mm mob 083 8.0 00 mob 083 08.0 mm mob 083
00.0 mm mob 083 00.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 2.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 88.0 mm mob 083 08.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083
00.0 mm mob 083 00.0 mm mob 083 8.0 mm mob 083 00.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 80 mm mob 083 8.0 mm mob 083
8.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 00.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 08.0 mm mob 083 80 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 00.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 00.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 00.0 mm mob 083 00.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083 8.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 08.0 mm mob 083 08.0 mm mob 083
00.0 mm mob 083 8.0 mm mob 083 80.0 mm mob 083 08.0 mm mob 083
8.8080 boob 28.880 080 288mb 055 28882 25:0

8303 A>m 0802 288800 5303 A>m 0802 808800 8303 A>m 0802 008809 8803 A>m 0802 808800

0.80 2.0 80080.0~

108

 

 

109

002 88 mob 083 00.0 88 mob 083 002 88 mob 083 022 88 mob 083
002 88 mob 083 00.0 88 mob 083 002 88 mob 083 022 88 mob 083
002 88 mob 083 00.0 88 mob 083 82 88 mob 083 022 88 mob 083
002 88 mob 083 00.0 88 mob 083 082 88 mob 083 02 2 88 mob 083
002 8.8 mob 083 00.0 88 mob 083 082 88 mob 083 02 2 88 mob 083
002 88 mob 083 80.0 88 mob 083 082 88 mob 083 02 2 88 mob 083
002 88 mob 083 8.0 88 mob 083 082 88 mob 083 002 88 mob 083
002 88 mob 083 08.0 88 mob 083 82 88 mob 083 002 88 mob 083
002 88 mob 083 00.0 88 mob 083 82 88 mob 083 002 88 mob 083
002 88 mob 083 8.0 88 mob 083 82 88 mob 083 002 88 mob 083
002 88 mob 083 8.0 88 mob 083 82 88 mob 083 002 88 mob 083
002 88 mob 083 8.0 88 mob 083 82 88 mob 083 002 88 mob 083
002 8.8 mob 083 00.0 88 mob 083 082 88 mob 083 002 88 mob 083
80.0 88 mob 083 8.0 88 mob 083 002 88 mob 083 002 88 mob 083
00.0 88 mob 083 88.0 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 88.0 88 mob 083 82 88 mob 083 002 88 mob 083
00.0 88 mob 083 08.0 88 mob 083 82 88 mob 083 002 88 mob 083
8.0 88 mob 083 08.0 88 mob 083 82 88 mob 083 002 88 mob 083
00.0 88 mob 083 08.0 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 8.0 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 8.0 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 08.8 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 8.8 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 08.8 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 02 .8 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 00.8 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 00.8 88 mob 083 082 88 mob 083 002 88 mob 083
00.0 88 mob 083 00.8 88 mob 083 022 88 mob 083 002 88 mob 083
00.0 88 mob 083 002 88 mob 083 022 88 mob 083 002 88 mob 083
88823 080 800mb 00.5 2882 85 280me 055
08203 2>8 802 8880 .8203 .58 282 8088 bbm2o3 2>m 282 8880 .bm2o3 2>8 802 08080

 

0.80 2.8 8088.8

 

02 2 08 Jmob 083 002 08 mob 083 82 88 mob 083 02 2 88 mob 083

 

022 08 mob 083 00.0 08 mob 083 82 88 mob 083 022 88 mob 083
02 2 08 mob 083 00.0 08 mob 083 002 88 mob 083 022 88 mob 083
022 08 mob 083 00.0 08 mob 083 002 88 mob 083 02 2 88 mob 083
022 08 mob 083 00.0 08 mob 083 002 88 mob 083 02 2 88 mob 083
022 08 mob 083 00.0 08 mob 083 002 88 mob 083 02 2 88 mob 083
022 08 mob 083 00.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
022 08 mob 083 00.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
022 08 mob 083 00.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
02 2 08 mob 083 00.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
022 08 mob 083 00.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
802 08 mob 083 88.0 08 mob 083 082 88 mob 083 022 88 mob 083
002 08 mob 083 08.0 08 mob 083 882 88 mob 083 022 88 mob 083
002 08 mob 083 08.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
002 08 mob 083 88.0 08 mob 083 082 _ 88 mob 083 02 2 88 mob 083
002 08 mob 083 08.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
002 08 mob 083 8.0 08 mob 083 082 88 mob 083 02 2 88 mob 083
002 08 mob 083 08.8 88 mob 083 082 88 mob 083 022 88 mob 083
002 08 mob 083 08.8 88 mob 083 082 88 mob 083 02 2 88 mob 083
002 08 mob 083 00.8 88 mob 083 082 88 mob 083 02 2 88 mob 083
002 08 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
002 08 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
002 08 mob 083 082 88 mob 083 082 88 mob 083 002 88 mob 083
002 08 mob 083 082 88 mob 083 082 88 mob 083 002 88 mob 083
002 08 mob 083 082 88 mob 083 082 88 mob 083 002 88 mob 083
002 08 mob 083 82 88 mob 083 022 88 mob 083 002 88 mob 083
002 08 mob 083 82 88 mob 083 022 88 mob 083 002 88 mob 083
002 08 mob 083 82 88 mob 083 022 88 mob 083 002 88 mob 083
002 08 mob 083 82 88 mob 083 022 88 mob 083 002 88 mob 083
Amaflmvl 285 @019 20:5 288% 080 8883 4005

8303 A>m 0802 008800 5303 A>m 0802 808800 8803 A>m 0802 008800 8803 A>m 0802 008800

 

r80 28 80080.8

110

 

 

1'11

082 88 Mob 083 002 88 m8 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 . 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 002 88 mob 083 082 08 mob 083 082 08 mob 083
082 88 mob 083 80.0 88 mob 083 82 08 mob 083 082 08 mob 083
082 88 mob 083 00.0 88 mob 083 002 08 mob 083 082 08 mob 083
082 88 mob 083 00.0 88 . mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 00.0 88 mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 00.0 88 mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 00.0 88 mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 00.0 88 mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 00.0 88 mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 08.0 88 mob 083 002 08 mob 083 082 . 08 mob 083
02 2 88 mob 083 08.0 88 mob 083 002 08 mob 083 082 08 mob 083
022 88 mob 083 00.8 08 mob 083 002 08 mob 083 02 2 08 mob 083
02 2 88 mob 083 00.8 08 mob 083 002 08 mob 083 02 2 08 mob 083
022 88 mob 083 002 08 mob 083 002 08 mob 083 02 2 08 mob 083
002 88 mob 083 002 08 mob 083 882 08 mob 083 022 08 mob 083
002 88 mob 083 082 08 mob 083 082 08 mob 083 02 2 08 mob 083
002 88 mob 083 082 08 mob 083 082 08 mob 083 022 08 mob 083
002 88 mob 083 082 08 mob 083 082 08 mob 083 02 2 08 mob 083
002 88 mob 083 082 08 mob 083 082 08 mob 083 022 08 mob 083
80.8800 080 288mb 8255 2200801850 28883 055

0883 2>8 282 8880 bbm2o3 .58 282 8880 obm2o3 2>m 8.02 8880 28203 2>8 8.02 8.8.8

 

0.80 2.8 8208098

 

on.— on Job 0003 omd mm womb 0003 on; mm wofi 0003 90.— mm wow 0003

 

082 08 mob 083 02 .8 88 mob 083 82 88 mob 083 002 88 mob 083
082 08 mob 083 00.8 88 mob 083 82 88 mob 083 002 88 mob 083
082 08 mob 083 00.8 88 mob 083 82 88 mob 083 002 88 mob 083
082 08 mob 083 00.8 88 mob 083 82 88 mob 083 002 88 mob 083
082 08 mob 083 00.8 88 mob 083 82 88 mob 083 002 88 mob 083
082 88 mob 083 00.8 88 mob 083 82 88 mob 083 882 88 mob 083
082 08 mob 083 002 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 002 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 002 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 002 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 002 88 mob 083 082 88 mob 083
082 08 mob 083 082 88 mob 083 002 88 mob 083 082 88 mob 083
022 08 mob 083 082 88 mob 083 002 88 mob 083 082 88 mob 083
022 08 mob 083 082 88 mob 083 002 88 mob 083 082 88 mob 083
022 08 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
002 88 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
002 08 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
00.0 08 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
00.0 08 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
88.0 08 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
00.8 88 mob 083 82 88 mob 083 002 88 mob 083 082 88 mob 083
00.va 00.5 28.8000 85 288mb 2800 80.68001 800

8303 128/m 0802 808800 8303 A>m 0802 808800 8903 15m 0802 008800 8203 A>m 0802 008800

 

0.80 28 80080.8

112

 

 

113

082 88 |mob 083 00.8 88 Job 083 082 88 88 083 82 88 mob 083
082 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
082 88 mob 083 002 88 mob 083 082 88 mob 083 082 88 mob 083
082 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
082 88 mob 083 002 88 mob 083 082 88 mob 083 082 88 mob 083
022 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
022 88 mob 083 002 88 mob 083 882 88 mob 083 82 88 mob 083
022 88 mob 083 002 88 mob 083 82 88 mob 083 82 88 mob 083
022 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
022 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
002 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
08.0 88 mob 083 882 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 082 88 mob 083 082 88 mob 083 802 88 mob 083
08.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
08.8 88 mob 083 082 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 082 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
08.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
02.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
02.8 88 mob 083 082 88 mob 083 082 88 mob 083 002 88 mob 083
02.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
00.8 88 mob 083 882 88 mob 083 82 88 mob 083 002 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 002 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 882 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
288080 882 08.882 8800 28080 288 080080 88

88203 2>m 2002 8880 .8203 2>8 282 8080 02203 .58 802 8088 02283 2>8 282 8880

 

2.80 2.8 8200800.

 

082 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083

114

 

082 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 802 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
00.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
00.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
00.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 82 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 82 88 mob 083 002 88 mob 083
08.8 88 mob 083 002 88 mob 083 82 88 mob 083 002 88 mob 083
02.8 88 mob 083 002 88 mob 083 082 88 mob 083 882 88 mob 083
02.8 88 mob 083 002 88 mob 083 082 88 mob 083 882 88 mob 083
02.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
02.8 88 mob 083 082 88 mob 083 082 88 mob 083 082 88 mob 083
02.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
02.8 88 mob 083 082 88 mob 083 082 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 082 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
00.8 88 mob 083 082 88 mob 083 82 88 mob 083 082 88 mob 083
488083 2882 882va A825 2880wa 2882 2880Wv 2882

8803 A>m 0802 008800 8303 A>m 080 2 008800 80203 q>m 0802 008800 20203 A>m 0802 008800

 

2.80 2.8 8208000

 

022 08 lmob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083

 

002 08 mob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083
002 08 mob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083
00.8 88 mob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083
08.8 88 mob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083
08.8 88 mob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083
08.8 88 mob 083 02.8 88 mob 083 002 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 002 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 002 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083
00.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083
88.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 082 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 082 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 82 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 082 88 mob 083
08.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
02.8 88 mob 083 002 88 mob 083 082 88 mob 083 002 88 mob 083
02.8 88 mob 083 002 88 mob 083 082 88 mob 083 082 88 mob 083
880480 2888 0.80me 8880 280me 055 285mb 2005

8303 A>m 0802 008800 8303 A>m 0802 808800 8303 128/m 0802 008800 8903 A>m 0802 008800

8.80 2.8 8208828

115

 

08.8 08 Jmob 083 08.8 08 8ob 083 00.8 08 mob 083 082 08 mob 083

 

08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083
80.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083
00.8 08 mob 083 08.8 08 mob 083 802 08 mob 083 082 08 mob 083
00.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
00.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 82 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 82 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 082 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 082 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 82 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 82 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 82 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 82 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 002 08 mob 083
88.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 082 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 082 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 082 08 mob 083 082 08 mob 083
A; 8.5 803mb 080 285mb 88 288mb 055

8303 A>m 0802 808800 8303 15m 0802 808800 8203 A>m 0802 808800 8203 A>m 0802 808800

8.80 2.8 8200800.

116

 

 

02.8 08 Jmob 083 08.8 08 mob 083 02 .8 08 Lmob 083 082 08 8b 083
00.8 08 mob 083 00.8 08 mob 083 02.8 08 mob 083 082 08 mob 083
00.8 08 mob 083 00.8 08 mob 083 02.8 08 mob 083 082 08 mob 083
00.8 08 mob 083 00.8 08 mob 083 02.8 08 mob 083 082 08 mob 083
08.8 08 mob 083 00.8 08 mob 083 02 .8 08 mob 083 082 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 02 .8 08 mob 083 82 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 02.8 08 mob 083 082 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 82 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 002 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 00.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 00.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 00.8 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 02 .8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 00.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
88.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 08.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 82.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 002 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 082 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 082 08 mob 083 08.8 08 mob 083
08.8 08 mob 083 02.8 08 mob 083 082 08 mob 083 08.8 08 mob 083
8880880 055 285002 080 288mm 055 288800 055
02203 08 282 8.080 82883 08 282 8.8.8 028203 28 282 88080 828203 23 802 8.080

 

0.80 2.8 82008088

117

 

002 88 |mob 083 00.8 28 mob 083 08.8 28 8b 083 02 .8 28 mob 083

 

082 88 mob 083 00.8 28 mob 083 08.8 28 mob 083 02.8 28 mob 083
082 88 mob 083 00.8 28 mob 083 08.8 28 mob 083 02 .8 28 mob 083
82 88 mob 083 00.8 28 mob 083 00.8 28 mob 083 02 .8 28 mob 083
08.2 88 mob 083 00.8 28 mob 083 00.8 28 mob 083 00.8 28 mob 083
02.8 28 mob 083 08.8 28 mob 083 00.8 28 mob 083 00.8 28 mob 083
08.0 28 mob 083 08.8 28 mob 083 00.8 28 mob 083 00.8 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 00.8 28 mob 083 00.8 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 00.8 28 mob 083 002 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 00.8 28 mob 083 00.2 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 00.8 28 mob 083 002 28 mob 083
08.8 28 mob 083 88.8 28 mob 083 00.8 28 mob 083 00.2 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 002 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.2 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 082 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.2 28 mob 083
08.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 082 28 mob 083
02.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.2 28 mob 083
02.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.2 28 mob 083
00.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.2 28 mob 083
00.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 082 28 mob 083
00.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.2 28 mob 083
00.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 082 28 mob 083
00.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.0 28 mob 083
00.8 28 mob 083 08.8 28 mob 083 08.8 28 mob 083 08.8 08 mob 083
00.8 28 mob 083 08.8 28 mob 083 02 .8 28 mob 083 08.8 08 mob 083
00.8 28 mob 083 08.8 28 mob 083 02 .8 28 mob 083 08.8 08 mob 083
00.8 28 mob 083 08.8 28 mob 083 02 .8 28 mob 083 08.8 08 mob 083
00.8 28 mob 083 08.8 28 mob 083 02.8 28 mob 083 02.8 08 mob 083
28280.88 0.000 28:82 2280 288mb €25 288mb 0025

£303 A>m 0802 808800 8303 A>m 0802 808800 80203 q>m 0802 808800 8903 A>m 0802 2208800

8.80 2.8 8200808.

118

 

 

08.8 08 mob 083 08.0 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 08.2. 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
02.8 08 mob 083 00.2. 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
02.8 2.8 mob 083 00.2. 88 mob 083 02 .8 88 mob 083 08.8 88 mob 083
00.8 08 mob 083 00.8 88 mob 083 02 .8 88 mob 083 00.8 88 mob 083
00.8 2.8 mob 083 08.8 88 mob 083 02 .8 88 mob 083 00.2 88 mob 083
00.8 08 mob 083 08.8 88 mob 083 02 .8 88 mob 083 08.8 88 mob 083
80.8 2.8 mob 083 08.8 88 mob 083 02 .8 88 mob 083 08.8 88 mob 083
00.8 08 mob 083 08.8 88 mob 083 02 .8 88 mob 083 08.0 88 mob 083
00.8 08 mob 083 08.8 88 mob 083 02 .8 88 mob 083 00.0 88 mob 083
00.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 00.0 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 8.8 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 00.8 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 00.8 88 mob 083
08.8 2.8 mob 083 08.8 88 mob 083 00.8 88 mob 083 00.8 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 00.8 88 mob 083 08.8 88 mob 083
08.8 2.8 mob 083 08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
00.8 2.8 mob 083 08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
00.8 08 mob 083 08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
02.8 2.8 mob 083 00.8 88 mob 083 08.8 88 mob 083 02 .8 88 mob 083
00.8 08 mob 083 08.8 88 mob 083 08.8 88 mob 083 02.8 88 mob 083
082 08 mob 083 08.8 88 mob 083 08.8 88 mob 083 00.8 88 mob 083
08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083 00.8 88 mob 083
08.0 88 mob 083 08.8 88 mob 083 08.8 88 mob 083 00.8 88 mob 083
00.2. 88 mob 083 08.8 88 mob 083 00.8 88 mob 083 002 88 mob 083
08.0 88 mob 083 08.8 88 mob 083 00.8 88 mob 083 002 88 mob 083
2880.23 2882 2880.23 A882 8180wa A882 28800.0: A882

8903 15m 0802 808800 8303 A>m 0802 008800 8803 4>m 0802 808800 8803 q>m 0802 008800

 

0.80 2.8 8.2022qu08

119

 

 

120

02.0 88 mob 083 08.0 88 Job 083 08.8 88 mob 083 08.8 08 mob 083
00.2. 88 mob 083 8.0 88 mob 083 08.8 88 mob 083 00.8 08 mob 083
00.0 88 mob 083 08.2. 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
00.0 88 mob 083 8.0 88 mob 083 08.8 88 mob 083 00.0 08 mob 083
00.0 88 mob 083 82. 88 mob 083 08.8 88 mob 083 08.2. 2.8 mob 083
00.2. 88 mob 083 08.2. 88 mob 083 08.8 88 mob 083 8.0 08 mob 083
00.8 88 mob 083 00.2. 88 mob 083 08.8 88 mob 083 00.0 08 mob 083
00.8 88 mob 083 00.0 88 mob 083 08.8 88 mob 083 08.0 08 mob 083
00.8 88 mob 083 00.0 88 mob 083 08.8 88 mob 083 02 .2. 2.8 mob 083
08.8 88 mob 083 08.0 88 mob 083 00.8 88 mob 083 00.8 08 mob 083
08.8 88 mob 083 08.0 88 mob 083 00.8 88 mob 083 00.8 2.8 mob 083
08.8 88 mob 083 08.0 88 mob 083 00.8 88 mob 083 00.8 2.8 mob 083
08.8 88 mob 083 08.0 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
08.8 88 mob 083 88.2. 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
08.8 88 mob 083 08.2. 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
08.8 88 mob 083 08.2. 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
08.8 88 mob 083 02 .0 88 mob 083 08.8 88 mob 083 08.8 2.8 mob 083
08.8 88 mob 083 00.2. 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
00.8 88 mob 083 00.2. 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
00.8 88 mob 083 00.0 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
08.8 88 mob 083 00.2. 88 mob 083 80.8 88 mob 083 08.8 2.8 mob 083
08.8 88 mob 083 00.8 88 mob 083 00.8 88 mob 083 08.8 2.8 mob 083
08.8 88 mob 083 00.8 88 mob 083 00.8 88 mob 083 08.8 08 mob 083
02.8 88 mob 083 08.8 88 mob 083 00.8 88 mob 083 08.8 08 mob 083
02.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083 08.8 08 mob 083
08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083 00.8 08 mob 083
08.8 88 mob 083 08.8 88 mob 083 08.8 88 mob 083 00.8 08 mob 083
08.8 88 mob 083 08.8 88 mob 083 08.2 88 mob 083 08.8 2.8 mob 083
02.8 88 mob 083 08.8 88 mob 083 08.8 08 mob 083 08.8 08 mob 083
0880.23 0882 Gav A882 8883 2882 8&3 2882

8303 A>m 0802 008800 8203 A>m 0802 808800 8203 A>m 0802 808800 8303 4>m 0802 008800

 

8.80 2.8 8200898

 

08.2. 08 ab 083 00.8 88 mob 083 00.8 88 mob 083 00.2. 88 Lmob 083

 

08.2. 08 mob 083 00.2. 88 mob 083 00.8 88 mob 083 00.2. 88 mob 083
08.2. 08 mob 083 00.2. 88 mob 083 08.8 88 mob 083 00.8 88 mob 083
8.0 08 mob 083 00.2. 88 mob 083 02.8 88 mob 083 88.8 88 mob 083
8.2. 08 mob 083 88.0 88 mob 083 02.8 88 mob 083 08.8 88 mob 083
8.2. 08 mob 083 08.0 88 mob 083 08.8 88 mob 083 08.8 88 mob 083
08.2. 08 mob 083 08.2. 88 mob 083 02.8 88 mob 083 02.8 88 mob 083
8.2. 08 mob 083 08.0 88 mob 083 02.8 88 mob 083 00.8 88 mob 083
02.2. 08 mob 083 8.2. 88 mob 083 02 .8 88 mob 083 08.8 88 mob 083
02.2. 08 mob 083 08.2. 88 mob 083 02 .8 88 mob 083 00.8 88 mob 083
02.2. 08 mob 083 08.2. 88 mob 083 00.8 88 mob 083 00.88 88 mob 083
08.2. 08 mob 083 08.2. 88 mob 083 00.2. 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 02.2. 88 mob 083 00.2. 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 02.2. 88 mob 083 00.0 88 mob 083 08.8 88 mob 083
08.8 08 mob 083 82. 88 mob 083 08.2. 88 mob 083 02 .8 88 mob 083
08.8 08 mob 083 08.2. 88 mob 083 08.0 88 mob 083 00.2. 88 mob 083
08.8 08 mob 083 02.2. 88 mob 083 08.2. 88 mob 083 00.2. 88 mob 083
08.8 08 mob 083 00.0 88 mob 083 08.2. 88 mob 083 08.2. 88 mob 083
00.8 08 mob 083 00.2. 88 mob 083 8.2. 88 mob 083 8.2. 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 8.0 88 mob 083 08.2. 88 mob 083
02.8 88 mob 083 08.8 88 mob 083 8.2. 88 mob 083 08.2. 88 mob 083
08.8 88 mob 083 02.8 88 mob 083 02.2. 88 mob 083 8.2. 88 mob 083
08.8 88 mob 083 02 .8 88 mob 083 00.0 88 mob 083 08.2. 88 mob 083
08.8 88 mob 083 00.8 88 mob 083 08.2. 88 mob 083 08.2. 88 mob 083
02.8 88 mob 083 08.8 88 mob 083 08.2. 88 mob 083 08.2. 88 mob 083
08.8 88 mob 083 08.8 88 mob 083 08.2. 88 mob 083 08.2. 88 mob 083
02.8 88 mob 083 08.8 88 mob 083 08.2. 88 mob 083 08.2. 88 mob 083
02.8 88 mob 083 02.8 88 mob 083 00.2. 88 mob 083 02 .2. 88 mob 083
02 .8 88 mob 083 00.8 88 mob 083 00.0 88 mob 083 02 .2. 88 mob 083
238mb 055 885880 8280 288mb 800 288882 02225

3303 A>m 0802 008800 8303 15m 0802 808800 8303 4>m 0802 008800 228203 A>m 0802 008800

8.80 2.8 8200800.

121

 

 

122

08.8 82. mob 083 08.8 22. mob 083 08.8 02. mob 083
08.8 82. mob 083 08.8 22. mob 083 08.8 02. mob 083
02.8 82. mob 083 2.2.8 22. mob 083 08.8 02. mob 083
08.8 82. mob 083 00.8 22. mob 083 00.8 02. mob 083
08.8 82. mob 083 00.8 22. mob 083 00.2. 02. mob 083
08.8 82. mob 083 00.8 22. mob 083 00.2. 02. mob 083
08.8 82. mob 083 00.8 22. mob 083 08.2. 02. mob 083
08.0 2.0 mob 083 08.8 82. mob 083 00.8 22. mob 083 8.2. 02. mob 083
08.8 2.2. mob 083 02 .8 82. mob 083 08.8 22. mob 083 82.2. 02. mob 083
02.8 2.2. mob 083 00.8 82. mob 083 08.8 22. mob 083 00.2. 02. mob 083
00.8 2.2. mob 083 08.8 82. mob 083 08.8 22. mob 083 08.8 02. mob 083
08.8 2.2. mob 083 08.8 82. mob 083 08.8 22. mob 083 08.8 08 mob 083
02.8 2.2. mob 083 08.8 82. mob 083 02.8 22. mob 083 08.8 08 mob 083
02 .0 82. mob 083 08.8 82. mob 083 02.8 22. mob 083 08.8 08 mob 083
00.0 82. mob 083 02.8 82. mob 083 08.8 22. mob 083 08.8 08 mob 083
08.8 82. mob 083 08.8 82. mob 083 08.8 22. mob 083 08.8 08 mob 083
02.8 82. mob 083 00.0 22. mob 083 00.8 22. mob 083 00.8 08 mob 083
08.8 82. mob 083 00.8 22. mob 083 00.8 22. mob 083 08.8 08 mob 083
08.8 82. mob 083 02.8 22. mob 083 00.8 22. mob 083 08.8 08 mob 083
08.8 82. mob 083 08.8 22. mob 083 00.8 22. mob 083 02.8 08 mob 083
08.8 82. mob 083 08.8 22. mob 083 08.2. 22. mob 083 08.8 08 mob 083
00.8 80 mob 083 02 .8 22. mob 083 02.8 02. mob 083 02 .8 08 mob 083
00.8 82. mob 083 02 .8 22. mob 083 08.8 02. mob 083 02 .8 08 mob 083
08.8 82. mob 083 00.8 22. mob 083 00.8 02. mob 083 02 .8 08 mob 083
08.8 82. mob 083 08.8 22. mob 083 08.8 02. mob 083 02.8 08 mob 083
00.8 82. mob 083 08.8 22. mob 083 02.8 02. mob 083 02 .8 08 mob 083
00.02 82. mob 083 08.8 22. mob 083 08.8 02. mob 083 00.8 08 mob 083
08.0 82. mob 083 08.8 22. mob 083 02.8 02. mob 083 00.8 08 mob 083
08.0 82. mob 083 08.8 22. mob 083 08.8 02. mob 083 00.2. 08 mob 083
282282 2280 282582 00202 @882 22282 282823 202202
2228203 2>8 2002 88.08 222m2o3 .58 8282 88280 222m2o3 .58 8282 82880 222803 2>8 282 820280

 

2.80 2.8 8202280....

 

8.0 on 20022080208 .8»: 8080.2 3.0 we 2028080208 002208280825 oo.~ 82. 2008080208 2802208280825

 

 

who on 2028080208 20w: 8080.2 08.8 00 2028080208 00220080225 omgm 3. 2028080208 002008-028.—
ond om 0028080208 20w: 883m 38. 2.0 2000080208 003008-005 omd 9. 20022080208 002208280225
22 0.0. 2028080208 20w: 80280m 08....” 80 202228080208 002208280825 08.2 N2. 20022080208 00220080225
82 am 2028080208 20w: 80280m 2.8 we 2028080208 2802208280225 om.~ 22. 2028080208 00220080225
00.0 mm 2028080208 bowa 80283 omd Nb 20022080208 00220080225 00.2 22. 2028080208 002.008-0225
00.0 mm 2028080208 20»: 80280m 8.8 20 2028080208 2802208280225 02.8” 02. 2028080208 00220080225
$6 08.. 8028080208 20w: E02m0m one o0 2028080208 00220080825 cod o0 2000080208 00220080225
owd 0N 8028080208 2082 8080m— omé 08 2028080208 00220080225 00.2 9. 2028080208 2002208280225
08.0 mm 0028080208 2082 802800 cod 00 202822080208 002208280225 9.2 9. 20022080208 00220080225
owd om 2028080208 Emu E0280m 8.2. cc 2028080208 008008-005 00.2 on 2028080208 002008-005
owd 0.0. 2002080208 2032 8080”.“ 00.0 08 2000080208 00220080225 cod mm 0008080208 000008-005
o8..o om 8008080208 20w: 80280m 00.2. on 2000080208 00220080225 08.2 8.2.” 20022080208 00220080225
88.0 0.0. 2028080208 20w: 80280m om.m mm 20022080208 00220080225 02.2 88 2028080208 2002208280225
086 mm 20022080208 20w: 80200.82 2: .2. 88 2008080208 00220080225 0: 8.8 20022080208 200220080225
86 mm 2028080208 .2082 80200m cod mm 2028080208 00220080225 08.0 88 2000080208 2202208280225
owd mm 2028080208 20w: 8200mm om.m mm 2008080208 00220080225 08.2 08. 2028080208 00220080225
oho mm 2028080208 20w: 8080.2 00.2. 28 2028080208 00220080225 02: mm 2000080208 2802208280225
006 mm 20022080208 20»: 80280m 03. 28 2028080208 2002208280225 22.022 mm 2000080208 2002208280225
Euo 8.8 2008080208 20»: E0280m 00.2 3 2028080208 00220080225 08.2 2.8 2028080208 00220080225
cod 80. 20228080208 20»: 8280”,. own 08 2028080208 2002208280225 32 28m 2028080208 200220080225
00..0 80. 2028080208 20m: 80280m 08.2. on 2028080208 00220080825 02 .2 on 2028080208 00220080225
00.0 mm 2028080208 2032 80280.82 00:0. 08 2028080208 00220080225 00.8 0N 2028080208 00220080225
08.6 mm 8028080208 2082 80280882 088 02. 2028080208 00220080225 8.0 mm 2028080208 00220080225
00.2 mm 2000080208 20w: 8080M o2..m 82. 2028080208 0022008005 cod R 2028080208 00220080225
cod on 20022080208 008008-005 end 82. 2028080208 00220080225 08.0 mm 2008280208 00220080225
ohm 2 8 2028080208 0080080225 cod 5. 20022080208 002.008.0225 end S 2028080208 2002208280225
cod 00 2028080208 008008-005 00.8 82. 2028080208 2802208280225 00.8 n 2000080208 2002208280225
Ang 2882 288023 2882 2880.23 2882
223203 A>m 0802 008800 282303 4>m 0802 008800 222303 q>m 0802 8208800
.8008 000 8008 02 00822222

20 0282888082 0304 802.2808 08 8 00.8800 82028080208 20 22.8203 280 3.2/m2 Ewan: 80> 02 80:82 8808088002 .N.N 8820220098..

123

 

owd mm 202022080208 200020022602 om2 02 2028080208 2200200822202 cod mm 2028080208 20w: E0280m

 

on .2 2.8 2028080208 2000200226022 02.0 02 2028080208 000200822032 36 mm 2028080208 20022 80280.82
8.2 mm 2028080208 2008200822832 9:. 2. 2028080208 2200200822032 0 2 .2 28 2028080208 20m: 80280m2
00.0 mm 802822080208 2800200222202 222 .2 2. 2028080208 2008200222002 o 2 .2 28 2028080208 20022 802800
08.0 88 202022080208 280020082602 00.82 80 2028080208 20022 802800 00.2 28. 2028080208 0022 80280.82
omd 2m 2028080208 2208200822032 228.2 cm 202222080208 20022 8080M 00.2 28 2028080208 20022 820m
00.2 28 202022080208 22002008222002 228.2 on 2028080208 20022 8080M cod 2 202022080208 20022 80280m2
00.0 2m 202022080208 2800200822832 3.2 R 202022080208 20022 8080M cod 2 2028080208 20822 80280m2
08.0 2 n 2028080208 2200200822002 08.2 on 202222080208 80022 80280m2 cod 2 n 2028080208 20022 80280m2
and 2m 20200080208 2000200222802 082 on 2028080208 20w: 80280m2 00.0 28 202822080208 208: 8080.2
end 08 2028080208 2200200222002 002 mm 2028080208 0022 80280m2 cod 2 2028080208 20022 80280m
9.6 on 2028080208 2800200222032 2282 2.8 2028080208 20022 80280m cad 28 2028080208 20022 838m
end cm 2028080208 2200200822002 32 3“ 2028080208 20w: E02832 mwd 2m 2028080208 20»: 802800
2.0 0.0. 2028080208 2000200822232 o 8..o 2.8 2028080208 0022 80280.22 owd 2 2028080208 20w: 80280m2
00.0 R 2028080208 280020082632 00.0 2.8 2028080208 0022 80280.82 08.0 28 2028080208 20022 8080M
02 .0 08 2028080208 2800200222832 08.2 mm 2028080208 20w: 8080M and 2 2028080208 20w: 80280m2
00.0 mm 2028080208 2002008260282 82 mm 2028080208 0022 80280.82 00.0 28 2028080208 0022 80280.82
omd mm 2028080208 2208200822032 8.2 mm 2028080208 20»: 802800 08.0 28 2028080208 20022 802800
omd mm 2028080208 2800200260282 no.2 mm 2028080208 20m: 80280m cod 28 2028080208 20022 8080M
omd 2.8 2028080208 2000200222232 002 mm 2028080208 80022 802882 and 28 2028080208 20022 80283
222 .o 2.8 2028080208 2000200222032 002 mm 2028080208 20w: 8080.2 08.0 28 2028080208 20822 80280182
omd mm 2028080208 2000200222832 002 mm 2028080208 20w: 802800 82 08 2028080208 20022 80280m2
omd mm 2028080208 288200822232 cod mm 2028080208 .2082 8080.2 8.2 om 2028080208 20w: 80280m2
omd mm 202082080208 200020082632 08.0 mm 2028080208 20m: 80280m2 00.2 on 2028080208 20w: 80280.22
02 .0 mm 2028080208 2800200822232 owd mm 2028080208 20w: 802m0m2 00.2 on 2028080208 20022 802800
00.0 mm 2028080208 2000200822832 88.6 mm 2028080208 20w: 80283 00.0 on 0028080208 2002 820m
and 2N 2028080208 20082002632 00.0 mm 2028080208 20022 8200M 08.0 on 8028080208 20022 802m0m
cod on 2028080208 20002008260282 cod mm 2028080208 200.22 802m0m 08.0 on 2028080208 20022 888m
28808222 2222022 282220.202 222282 282882 222225
8303 2.2/m 0802 2208800 222203 2>m 080 2 008800 222303 2>m 080 2 008800

 

2.80 .8.8 8200802.

124

 

 

08.2 82. 28.202202222222228. 082822-000
08.2 82. 2002202220208 082822000

02 .2 NV uoCCMEMEm voxomnufiom

Owd NV hoUGmEQEm UmXownufiomm

on; —V 202082922228 Coxoanévomm

0N2 2V 209205223. onomfiuvvm

00.2 _V 202088222me onomnu—vom

o0; OV 206220222323 fioxomnufiom

0V2 OV hofiGQEBQm CoxomDAuvM

Om; OV 220508me Coxownufiom

08.2 28 20022022202228 082822-032 08.2 08 2002282220208 082822000
082 02. 28222222222208 082822-032 02.2 08 2002202228288 002822-032
on .2 8.2. 2028080208 2200200822202 002 on 2028080208 2000200826002
082 82. 20022222220208 092822-88 00.0 08 28022222222282 082822-032
ON; 0V hoUGdEw—mm fioxomnufiom 00.0 On 20202.88me fioxomnufiud
08.2 82. 20022022202228 082822000 002 88 20022022222208 082822-088
08.8 2.2. 20022222220208 082822-032 08.2 88 2002282220208 082822-008
08.2 2.2. 2002202228208 032822032 002 88 20022222220208 082822-080
Om; VV uowfimeEm toxoannfium God mm 202356—280. quomnufivm
OMA VV 2Dficmegwm fioxomnufiom Omd mm newsman—mm 2082028822001
082 2.2. 200222222202228 082822-002 08.2 88 200220222222228 082822-008
ONA VV 200888228 fioxomnnfiod 02 .2 5m HOUGQEEMm fivfiomnnvom
82 2.2. 2028080208 2800200222802 2.6 nm 20022080208 20002008220002
00.0 2.2. 200222222282228 082822-032 00.0 88 2002202228222.2 0022822008
00.0 VV uofifimamfiw @oxowaumfim 05.0 cm 2302228222030. woxomnnwom
08.2 82. 200222222222208 082822032 002 88 20022222220202 082822-032
08.2 82. 2002202220208 082822-032 082 88 2002202220208 082822032
02 .2 82. 222022822222288 082822-032 08.2 88 2002202220208 082822-032

28222082 22222222 28222882 22222222 2822202822 22222222
222m203 2>8 0222222 8222280 222m2o3 2.8 022202 2282280 2228203 2>8 8282 8.22280

 

2.80 .88 82022022222.

125

Appendix 2.3. Measurements (snout to vent length (SVL), tail length, weight, and sex)
of snakes captured in the southern Lower Peninsula of Michigan in 2005 and 2006.

 

 

Common Name SVL Tail Weight Sex* Comments
(mm) (mm) (grams)

Eastern milk snake 442 59 26.00 U
Red-bellied snake 102 26 1.60 U
Red-bellied snake 180 45 4.00 U
Eastern garter snake 135 25 2.00 U
Eastern garter snake 145 30 2.20 U
Eastern garter snake 155 28 2.50 U
Eastern garter snake 160 50 2.80 U
Eastern garter snake 220 56 4.00 U
Eastern garter snake 350 100 3.10 U
Eastern garter snake 360 100 27.00 U
Eastern garter snake 375 110 ------- U
Eastern garter snake 410 120 32.00 U
Eastern garter snake 420 110 31.50 U
Eastern garter snake 435 103 63.00 M
Eastern garter snake 470 145 57.00 U
Eastern garter snake 480 150 36.00 M
Eastern garter snake 494 116 74.00 U
Eastern garter snake 500 128 59.00 U
Eastern garter snake 523 124 92.00 F
Butler’s garter snake 620 1 15 1 10.00 U
Northern ribbon snake 195 83 2.50 U

Northern ribbon snake 195 80 3.40 U Postnatal scar

Northern ribbon snake 215 87 4.20 U Postnatal scar

Northern ribbon snake 225 85 3.80 U Postnatal scar
Northern ribbon snake 252 75 6.50 U
Northern ribbon snake 330 145 14.00 U
Northern ribbon snake 345 145 13.00 U

 

*F = Female, M = Male, U = Unknown

126

2285222: u 22 .0202 u 2 022222202 u .2.

 

 

822222222222 88 88 28 88 22 0222222 02222022
022020 088 802 088 82.8 2 222222 822222222222m 8222280
02.22228 82 002 82 02 888 22 0222222 822222220228 8222280
082 882 82.8 888 2 0222222 2222200228 22822280
22222222 22222222 22222222 22222222
8808800 2.2220285 2202280222 282202 2202280222 22220285 000820.200 220022012 000820.200 880m 080 2 2208800

 

.8008 02222 8008 222 2208222022 2o 222228222022 20282 22822288
0222 8 220288200 8022.28 .20 22222223 220880282 280 282202 2202280222 22223 000820.200 .222w2202 000820.200 208082 82820802228002 .2..~ 8220220828220.

127

CHAPTER 3: COMPARISON OF THE HERPETOFAUNAL COMMUNITIES ON
MICHIGAN ’S STATE GAME AND WILDLIFE AREAS AND PRIVATE LANDS

INTRODUCTION

In recent decades, much attention has focused on the global decline of
herpetofauna, particularly amphibian species (Kiesecker et a1. 2001). Several factors
have been implicated in recent amphibian population declines including habitat loss and
fragmentation (Blaustein et a1 1994, Blaustein and Kiesecker 2002), introduced predators
(Blaustein and Kiesecker 2002, Kats and Ferrer 2003), pollution (Alford and Richards
1999), increased ultraviolet B (UV-B) radiation (Alford and Richards 1999, Reaser and
Blaustein 2005), global climate change (Alford and Richards 1999, Reaser and Blaustein
2005), and infectious disease (Alford and Richards 1999, Reaser and Blaustein 2005).

Concern about amphibian declines is most likely due to their potential role as
indicators of environmental stress (Halliday 2000, Blaustein and Kiesecker 2002, Reaser
and Blaustein 2005). Amphibians use both aquatic and terrestrial habitats, and therefore
come into contact with stressors in both environments (Stebbins and Cohen 1995,
Blaustein and Kiesecker 2002). Amphibians have moist permeable skin and their semi-
permeable eggs are directly exposed to sunlight, water, and soil (Blaustein and Kiesecker
2002). Amphibians are believed to play a critical role in ecosystem dynamics because
they are the most abundant vertebrate in many forested ecosystems (Burton and Likens
1975). Their decline could have substantial impacts on other organisms by disrupting
ecosystem function (Blaustein and Wake 1995, Blaustein and Kiesecker 2002). A severe
population decline could have considerable and lasting effects on ecosystems potentially
resulting in adjustment and restructuring of food webs of amphibian invertebrate prey and

vertebrate predators.

128

Herpetofaunal communities are influenced by local habitat conditions,
environmental factors, and landscape features. Local terrestrial habitat conditions that
have been shown to affect amphibian species include litter depth, coarse woody debris (of
various sizes and decay classes), overstory canopy cover, and soil moisture (deMaynadier
and Hunter 1995). For example, amphibian species richness was positively correlated
with local habitat conditions such as variables related to vegetation cover and negatively
correlated with water depth in southwestern Ontario (Hecnar and M’Closkey 1998).
Salamander abundance was found to be negatively correlated with increases in percent
bare ground as a result of leaf litter reductions due to prescribed burns in bottomland
hardwood stands in Georgia (Moseley et al. 2003). Environmental factors such as
ambient temperature and water availability are critical to amphibians (Carey et al. 2001).
The rate of water loss across permeable amphibian skin is dependent upon temperature,
and most amphibians need a freshwater source to rehydrate (Carey et a1 2001).
Amphibian’s breeding activities are initiated by response to increasing temperatures and
to rainfall (Stebbins and Cohen 1995). On a landscape level, amphibians and reptiles are
dependent on juxtaposition of wetlands and adjacent terrestrial habitats, as well as
connectivity between the two (McDiarmid 1994). The proximity of forested patches and
ponds to the next nearest pond can influence herpetofaunal communities (Knutson et a1.
1999)

The majority of the southern Michigan landscape is privately owned. State Game
and Wildlife Areas (SGWA) are found primarily in the Lower Peninsula of Michigan and
have been designated for natural resource conservation. They are publicly held lands

managed by the Michigan Department of Natural Resources (MDNR) to conserve and

129

restore the State’s wildlife resources (Application for Federal Assistance for the
Michigan Department of Natural Resources-Statewide Land Acquisition Grant Oct. 16,
1998). SGWA and private lands are configured as a patch—matrix of land ownership
throughout the Lower Peninsula, with private lands as the matrix and SGWA as the
patches.

Monitoring herpetofauna can be an effective way to understand the responses of
wildlife to varying land management regimes. Herpetofauna, particularly amphibians,
may provide ecological insights to enviromnental stressors such as habitat modifications
and environmental change. Detecting declines in herpetofauna at the population level,
given their environmental sensitivity, may indicate a loss of ecosystem integrity.
Monitoring herpetofauna populations and identifying their responses to land management
regimes can shape the way lands are managed. For SGWA, knowledge of herpetofaunal
community composition can provide a broader context for understanding how MDNR
activities, such as land acquisition and forest management, contribute to herpetofaunal
conservation in southern Michigan. Preserving and managing adequate and essential
habitat can limit reductions in not only herpetofaunal populations, but in biological
diversity as a whole.

The primary objective of this study was to describe the important determinants of
herpetofaunal community structure to see if differences between land ownerships existed.
This information will help the MDNR understand how their SGWA contribute to
landscape-level diversity. Because SGWA tend to be comprised of larger, contiguous
blocks of forested land in southern Michigan, 1 expected to find more forest dependent

herpetofaunal species and greater abundances of these species on SGWA than on private

130

lands. Independent of ownership, I expected areas that received greater rainfall with
warmer temperatures to contain more amphibian species and greater abundances of these
species. I also expected areas in closer proximity to water bodies to have greater
herpetofaunal diversity.

All capture, handling, and marking protocols used in this study were approved by

the Michigan State University Animal Care and Use Committee (AUF# 07/03-082—00).

METHODS

To assess differences in herpetofaunal communities in the southern Lower
Peninsula of Michigan, sites were sampled on SGWA as well as on privately-owned
parcels in the spring and summer of 2005 and 2006. Eighty-two sites were sampled on a
patch level scale, and sites were selected based on three characteristics: 1) land
ownership, 2) vegetation type, and 3) soil type.
Site Stratification Variables
Soil Associations

Site locations were selected from map data using ArcGIS 9.1 (Environmental
Systems Research Institute 2004). The soil information used for this map was Natural
Resources Conservation Service 1994 STATSGO data for Michigan (United States
Department of Agriculture 1994). STATSGO was compiled at 1:250,000, was designed
to be used primarily for regional, multi-state, State, and river basin resource planning,
management and monitoring, and is not detailed enough for interpretation at the county
level (United States Department of Agriculture 1995). As such, 1 used STATSGO to

coarsely stratify my study area into patches. STATSGO soil maps were compiled by

131

generalizing more detailed Soil Survey Geographic (SSURSGO) maps. The following
soil associations were used in this study: Houghton-Carlisle—Adrian (HCA), Marlette-
Capac-Parkhill (MCP), Marlette-Capac-Spinks (MCS), Miami-Hillsdale-Edward (MHE),
and Spinks-Houghton-Boyer (SHB) (Table 3.1).
Vegetation Types

Site locations were selected fiom map data using ArcGIS 9.1 (Environmental
Systems Research Institute 2004). I selected sites using the Integrated Forest Monitoring,
Assessment, and Prescription (1F MAP) system developed by the MDNR (Michigan
Department of Natural Resources 2001). The map was fiom derived fi'om Landsat TM 5
and 7 imagery from 1999 — 2001, had a 30 pixel size, and was ground verified for
accuracy. The IFMAP system contained 7 major classes (level one) of land use (Table
3.2). Sites were selected in four level 3 classes of the main forested landscape class:
lowland hardwoods, northern hardwoods, pine, and upland hardwoods (Table 3.2).
Accuracy assessments varied considerably for the subclasses that I studied ranging from
25% to 87% (Table 3.2).
Herpetofauna
Drift Fence Arrays

Drift fence arrays made from 60 cm high aluminum flashing with pitfall and
funnel traps were used to capture herpetofauna (Corn 1994, Enge 1997). Drift fences
intercept herpetofauna moving on the ground and re-direct them into a pitfall or funnel
trap. Drift fences with pitfall and funnel traps were installed in April and early May prior
to opening the traps in mid-May. They were opened for 5 consecutive nights each month

in 2005 and 4 consecutive nights each month in 2006 from May through August (Table

132

 

02.22823 .0w080220
20082 22.2028 $2202 80>

 

022282822 2022 8022022220 2022082 2280222022282 022202222 2022
22202220820 2203 .2808 288002 2222082 2.2322 .2008 0080222 2203 .2808 288002 9.5 20.208280288002802825
022222222 .w2202 22.2028 80220820 022222882 2082
2222082 220220208 02.28208 6080.220 :03 .8002 22222208 002220222 2203 228002 E222 0030m-0202082222.2-2802§
o220822
2022 .2808 02222 2.88002 022222222 2022 $8002 022222222 2022 228002 m02>2 8820m00820002202202
0.828.822 2w2202 222022022228 28022
22202220222 28220022 2.28002 02.220822 2082 $8002 02.202222 2022 $8002 2202 22222022002200-02202202
022222222 09820222 02.220222 .0w0222022o 022222222 008222020
20082 80> .w2202 80> 20022 80> .w2202 80> 20082 80> .w2202 80>
80220820 28082 280222022282 222022028222 2222022 200222022822 8022022222 28082 2200222022082
82022 22w282 2020228 28082 2.2822 .2008 82022 22822 20228 <02 220222220702822200220222msoz
02000
m 222082082800 N 222022082800 2 222022082800 220m 2202202008820. 220m

 

133

2202822 0.20 8222022082800 00mH<Hm 202.08 008222 28.282 0222 20 80022822208082 82208008 22202.2 boom 28220 moom 0222 822.820 2203220222 .20 0222822222022
0302 802222208 0222 8 2202200208 80228 22202228 222 200882 202822202232. .20 222082208200 80205 002222: .002>.20m 22022032088200 8002220832 20.280 2
E06 808m 220m2 02022 2203220232 20.2 503.09% 8202 220m 20202200 .m.D 02 w2222o20000 2802000 2802 8220220200880 220m .2.m 02820.2.

Table 3.2. Landscape categories (and codes) according to the Integrated Forest
Monitoring, Assessment, and Prescription (IFMAP) system (Michigan Department of

 

 

Natural Resources 2001).
Landscape Categories and Associated Class Descriptions Classification
IFMAP Class Names and Grid Values Rate
Urban
Low Intensity Urban Land area >10% and <25% manmade 84
(Residential) (1) structures, including paved and gravel roads
and parking lots.
High Intensity Urban (2) Land area >2 5% solid impervious cover made 83
from manmade materials, other than airports,
roads, or parking lots.
Airports (3) lmpervious land within airport grounds, 99
including runways.
Roads / Pavement (4) Roads or parking lots 96
Agricultural
Non-vegetated agriculture (5) Land area tilled for crop production with <25% 55
currently vegetated.
Row Crops (6) Vegetation is annual crops planted in rows 95
(e.g. corn, soybeans).
Forage Crops (7) Vegetation used for fodder production (e.g. 100
alfalfa, hay). Also includes land used for
pasture, or non-tilled herbaceous agriculture.
Orchards/Vineyard/Nurseries(9) Woody trees not grown for Christmas trees. 95
Openland Upland
Herbaceous Openland (IO) <25% of land area is covered by woody cover. 73
Low Density Trees (12) The combination of woody shrubs and trees is 88
>25% of the land area and >25% of the woody
cover is trees. None mapped in SLP.
Parks, Golf Courses (13) Upland open land maintained for recreational 96
purposes.
Forested
Deciduous Forest
Northern Hardwoods (14)* Combination of maples, beech, basswood, 87
white ash, cherry, and yellow birch >60% of
the canopy.
Oak Type (15) Proportion of oaks >60% of the canOpy. 36
Aspen Type (16) Proportion of aspen >40% of the canopy. 45
Other Upland Deciduous (l7) Proportion of any other single species >60% of
the canopy.
Mixed Upland Deciduous (1 3)* Proportion of deciduous trees >60% of the 25
canopy.
Lowland Deciduous Forest Proportion of deciduous trees >60% of the 56
(24)* canopy.
Coniferous Forest (CF)
Pines (19)* Proportion of pines >60% of the canopy. 71
Other Conifers (20) Proportion of non-pine upland conifers >60% 36
of the canopy.
Mixed Upland Conifers (21) Proportion of coniferous trees >60% of the 0
canopy. None mapped in SLP.
Lowland Coniferous Forest (25) Proportion of coniferous trees >60% of the 80

canopy.

 

l34

Table 3.2. Con’t

 

Upland Mixed Forest (22) Mixed forest not falling into any other 9
category. Proportion of conifers: deciduous
ranges between 40%:60% to 60%:40%.
Lowland Mixed Forest (26) Mixed forest not falling into any other 0
category. Pr0portion of conifers:
deciduous ranges between 40%:60% to
' 60%:40%.
Nonforested Wetland
Floating Aquatic (27) Proportion of floating aquatic vegetation 76
>60%of non-water cover.
Lowland Shrub (28) Proportion of lowland shrub >60% of 96
non-water cover.
Emergent Wetland (29) Proportion of emergent wetland >60% of 81
non water cover.
Mixed Non-forest Non-forested wetlands not falling into 85
Wetland (30) any other category.
Water
Water (23) Proportion of open water >75% of the 75
land area.
Bare/Sparsely Vegetated
Sand, Soil (31) Land cover is formed primarily of sand or 80
bare soil.
Exposed rock (32) Land cover is formed of solid rock. None 100
mapped in SLP.
Other Bare\Sparsely None. 92
Vegetated (35)

 

*Categories were used in study site selection in Michigan during the 2005 and 2006 field seasons.

135

3.3) and were checked once daily. Three 5 m long sections of aluminum flashing were
installed in a Y arrangement and 4 pitfall traps and 6 funnel traps were placed within the
array (see Fig.1 Enge 1997). Arrays were oriented to the north. Pitfall traps were made
from 18.9 L buckets. Holes were drilled approximately 2 cm fi'om the bottom of the trap
to allow for drainage of rainwater (Enge 1997). The pitfall traps were buried slightly
below ground level, allowing animals to drop into the bucket. Moistened sponges were
placed in all pitfall traps to prevent desiccation of captured animals (Greenberg et al.
1994). The sponges were remoistened as needed (Enge 1997, Richter and Seigel 2002).
Traps were closed by placing lids over the buckets.

Funnel traps were placed at the midpoint of each wing of the aluminum flashing
in the array (Greenberg and Tanner 2005). Funnel traps were double entry with the main
body comprised of aluminum window screening and the funnels of flexible fiberglass
screening. Traps had 20 cm openings at both ends with funnel openings of 5 cm in
diameter (Corn 1994). When not in use, funnels traps were closed by inverting the
funnels and clipping them shut.

Coverboards

Coverboards (Fellers and Drost 1994, Davis 1997) were placed within Im from
the drift fence array in the four cardinal directions at least one week prior to drift-fence
array installation. Coverboards were made of untreated birch plywood and cut into 1 m x
l m sections and were placed on bare ground. Coverboards are designed to provide
moist, cool refiige for herpetofauna (Houze and Chandler 2002) and create a microhabitat
similar to a downed log. Coverboards were checked every day that the drifi fence arrays

were open.

136

Table 3.3. Dates drift fence arrays, pitfall traps, funnel traps and coverboards were
opened in the southern Lower Peninsula of Michigan on SGWA and private lands in
summer 2005 and 2006.

 

 

Date Opened Date Closed Number of Sites
05/17/2005 05/22/2005 20
05/24/2005 05/29/2005 22
06/06/2005 06/1 1/2005 20
06/14/2005 06/19/2005 22
07/1 1/2005 07/16/2005 20
07/19/2005 07/24/2005 2]
08/05/2005 08/09/2005 20
08/10/2005 08/14/2005 21
05/08/2006 05/12/2006 1 7
05/15/2006 05/19/2006 14
05/22/2006 05/26/2006 9
06/06/2006 06/1 6/2006 1 7
06/ 12/2006 06/ 16/2006 20
06/19/2006 06/23/2006 3
07/05/2006 07/09/2006 1 7
070/9/2006 07/13/2006 3
07/12/2006 07/16/2006 20
07/31/2006 08/04/2006 20
08/07/2006 08/1 1/2006 20

 

137

Area T ime-Constrained Surveys

Area time-constrained surveys (ATC surveys) (Campbell and Christrnan 1982,
Crump and Scott 1994) were conducted on each site once a month from May to August in
2006: 23 May — 31 May, 1 June- 2 June, 19 June —- 27 June, 19 July — 26 July, 1 August —
12 August; and in June and July in 2005: 22 June — 30 June, 22 July — 31 July. A 2 m x
37 in area (the same total area as the drift fence array) was delineated to conduct the area
time—constrained surveys, and this same area was used for surveys throughout the season.
ATC survey areas were selected near the drift fence array in one of the four cardinal
directions, starting east of the array. When a 2 m x 37 m area fit in the designated soil
and vegetation type, it was georeferenced with a GPS unit and flagged. This designated
area was hand-searched for a period of 20 minutes by overturning downed logs and rocks
and searching through leaf litter. Decay classes of the logs were recorded according to
the US. Forest Service’s F IA Field Methods for Phase 3 Measurements, five decay
classifications (USDA Forest Service 2004) (Table 3.4).
Mark-Recapture Techniques

All animals captured by pitfall traps, funnel traps, cover boards, and area time-
constrained searches were processed before being released. Herpetofauna were identified
to species, sexed (when possible), and marked. A11 amphibians were measured (snout to
vent length, mm); and northern leopard frog (Rana pipiens), American toad (Bufo

americanus), and green frog (Rana clamitans) with a snout to vent length measurement >
50 mm were PIT tagged (passive integrated transponders; AVID®). Salamanders and all

other Anurans were marked by toe-clipping except treefro gs and relatives, but not for

individual recognition. Snakes and turtles were measured and marked for individual

138

Table 3.4. Five decay classifications of downed logs (U .S. Forest Service 2004) used in
area time-constrained surveys in the southern Lower Peninsula of Michigan on SGWA
and private lands in summer 2005 and 2006.

 

 

Class Description

L1 Bark intact, twigs present. Texture is intact. Wood is original in color.
Log is elevated on supported points above ground.

L2 Bark intact, twigs absent. Texture is intact to partially soft. Wood is
original in color. Log elevated on support points but sagging slightly.

L3 Trace of bark. Twigs are absent and texture is hard large pieces. Color of
wood is original to faded. Log is sagging near ground.

L4 Bark and twigs are absent. Texture of wood is small, soft, blocky pieces.
Wood is light brown to faded brown or yellowish. All of the log is on the
ground.

L5 Bark and twigs are absent. Texture of wood is soft and powdery. Color of

wood is faded to light yellow or gray. The diameter of the log is
attainable wood and log debris is not spread out in a flat manner. If a

diameter is not attainable, then it is not considered a log, but a pile of
debris.

 

139

recognition. A11 snakes were marked by clipping half of one ventral scale (Brown and
Parker 1976) and turtles were marked by notching marginal scutes (Cagle 1939).
Individuals were released at least 5 m from the point of capture on the opposite side of
drift fence to minimize the probability of immediate recapture.
Vegetation Variables
Ground Vegetation Percent Cover

Vegetation cover was collected at each drift fence array and corresponding ATC
area using a Daubenmire frame (Daubenmire 1959) modified in size from 1,000 cm2 to
2,500 cmz. Measurements were taken within the drift fence arrays (5 fi'ames: 2 on either
side of the north-south arm of flashing; one on east side of southeast arm of flashing; one
on west side of southwest arm of flashing; and one south of center bucket) and ATC
search areas (5 frames at 7 m, 14 m, 21 m, 28 m, and 35 m along the transect) once in
July. Vegetation variables included percent live cover (fern, forb, grass, moss, and
woody), standing dead cover, as well as estimates of bare ground, litter cover, and
average litter depth. In 2005, vegetation variables were measured at 42 different sites,
and in 2006, they were measured at 40 different sites. Vegetation measurements fiom the
drift fence arrays and ATC survey locations were examined for correlations using
Spearrnan Rank R statistics to determine if like measurements from individual study sites
could be pooled resulting in one measurement per variable for the sampled patch. Each
variable was correlated (P < 0.05) between the drift fence and ATC locations, thus they

were averaged together to obtain a single measurement per variable.

140

Canopy Percent Cover and Composition

Overstory vegetation was identified to species for all tree species that overlapped
the drift-fence arrays and area time-constrained survey sites. At each site, canopy cover
estimates were obtained with a spherical densiometer. At each drift fence array, canopy
cover was estimated at the pitfall traps and cover board locations and at each ATC site,
canopy cover was estimated along a line transect at 0 m, 6 m, 12 m, 18 m, 24 m, 30 m,
and 36 m. For each location, four readings were taken in the four cardinal directions, and
averaged to obtain on estimate per location. Because the canopy cover estimates were
not independent of each other, the median of all measurements from both drift fence
arrays and ATC sites was taken and the canopy cover measurement closest to the median
and therefore, most representative of the study site, was used during analyses.
Coarse Woody Debris

Coarse woody debris > 10 cm in diameter that either intersected or were
contained within the drift fence array and ATC sites were recorded. If coarse woody
debris intersected both the drift fence array and ATC site, it was only recorded once.
Length, width, and number of pieces were recorded for downed wood at each sample
patch. A count of the number of coarse woody debris pieces, average width, and average
length were used in the analysis.
Weather Variables
Precipitation

A rain gauge was installed at each sampling site. Gauges were checked and
emptied every day that the drift fence arrays were surveyed, as well as the day of the

ATC survey. The average precipitation for the sampling period was used in the analysis.

141

The first day of the sampling period, corresponding to opening the traps, was not
included in the average, as the gauge was not emptied from previous days. Some days
also were censored fiom the average precipitation calculation because the rain gauge had
been knocked over.
Temperature

Temperature was obtained from the state Climatologist in the Geography
Department at Michigan State University as well as through National Climatic Data
Center (National Ocean and Atmospheric Administration 2007) for weather stations
distributed through the ecoregion. Sample site locations were matched to their nearest
climate station using a GIS overlay of study site locations and climate station locations.
Daily maximum and minimum temperatures were used to obtain an average daily
maximum and minimum temperature for each site during the days that the drift fence
arrays were surveyed, as well as the day the area time-constrained survey was conducted.
Spatial Feature
Proximity to Water Body

Distance to the nearest water body (wetland, lake, stream, vernal pool, wet
meadow) was measured from the center bucket of all pitfall traps at the drift fence array
in May, June, and July. The type of water body, width and depth was recorded, and the
location of the closest water body edge to the bucket of the drift fence array was
georeferenced with a GPS unit. The mean distance, width, and depth of each recorded

water body for the three sampling periods was used in the analyses.

142

Statistical Analyses
Mixed Model with Repeated Measures — Considering Weather Eflects

After examination of descriptive statistics and normal probability plots for
original variables and potential data transformations, all climatology variables were log
transformed and column standardized prior to analysis. Certain criteria were used to

select variables for inclusion in potential models including R squared, Adjusted R
squared, Cp (Mallow) statistics (Mallows 1973), forward selection, and backward

selection. Outliers, collinearity, residual plots, meaningfulness and interpretation were
examined for candidate models. To address differences between years (a < 0.05), a
mixed model with repeated measures was used to determine which climatological
variables were most influential in determining amphibian richness and herpetofaunal
species richness and diversity. The treatment design consisted of three fixed factors
including land ownership with two levels, vegetation type with four levels, and soil
association with five levels and one to three potential random factors including minimum
temperature (with 4 repeated measures), precipitation (with 4 repeated measures), and
survey time (with 4 repeated measures). Models were ranked by how well they explained

amphibian richness and herpetofaunal species richness and diversity data using Akaike’s

information criterion corrected for small sample sizes (AICC) (Bumham and Anderson
2002).AICc was calculated as follows:
AICC = -2*ln(likelihood) +2*K + (2*K*(K+1))/(n-K-l ),

where In is the natural logarithm, (likelihood) is the value of the likelihood, and K is the

number of parameters in the model, and n is the sample size (Burrrharn and Anderson

2002). Models were also evaluated using AIC weights (wi), which can be treated as the

143

weight of evidence in favor of a particular model, given the data set and the set of
candidate models (Bumham and Anderson 2002). The factor main effects and the
interaction effects were analyzed using ANOVA. When main effects were significant (a
< 0.05), the treatment comparison among the marginal means was conducted. The data
analysis was conducted using PROC MIXED (SAS Institute 2002).

Mixed model — Focus on All Variables (Across Sample Periods)

After examination of descriptive statistics and normal probability plots for
original variables and potential data transformations, all habitat variables and climate
variables (using the average of the 4 sampling periods for all climate variables, instead of
repeated measures, as in the prior analyses) were log transformed and column
standardized prior to analysis. Certain criteria were used to select candidate variables for
inclusion in the models including R2, Adjusted R2, Cp (Mallow) statistics (Mallows
1973), forward selection, and backward selection. Outliers, collinearity, residual plots,
meaningfulness and interpretation were examined for candidate models. A mixed model
was used to determine which variables (climatological, habitat, and one spatial) were
most influential in determining amphibian richness and herpetofaunal species richness
and diversity. The treatment design consisted of three fixed factors including land
ownership with two levels, vegetation type with four levels, and soil association with five
levels, and 21 potential random factors. Models were ranked by how well they explained

amphibian richness and herpetofaunal species richness and diversity data using Akaike’s

information criterion corrected for small sample sizes (AICC) (Bumham and Anderson
2002). AlCc was calculated as follows:

AICC = -2*1n(1ikelihood) +2*K + (2*K*(K+1))/(n-K-l),

144

where In is the natural logarithm, (likelihood) is the value of the likelihood, and K is the

number of parameters in the model, and n is the sample size (Bumham and Anderson
2002). Models were also evaluated using AIC weights (wi), which can be treated as the

weight of evidence in favor of a particular model, given the data set and the set of
candidate models (Bumham and Anderson 2002). The factor main effects and the
interaction effects were analyzed using ANOVA. When main effects significant ((1 <
0.05), the treatment comparison among the marginal means was conducted. The data
analysis was conducted using PROC MIXED (SAS Institute 2002).

Constrained Ordination

A constrained ordination was used to examine the relationship between
herpetofaunal species and habitat and weather variables (collectively called
environmental data). Herpetofaunal species abundance data were the response variables
and the environmental data were the explanatory variables. The goal of the constrained
ordination was to infer patterns in herpetofaunal species composition from patterns in the
environmental variables.

I removed species from my herpetofaunal community data set with less than 3
occurrences because they were most likely not sufficiently sampled and would not be
accurately placed in ecological space. Of 18 species sampled, 12 were retained (Table
3.5). The species abundance data was highly skewed and heteroskedastic, so the data
was log transformed as well as a column normalized using R 2.4.0 (R Development Core
Team 2006).

Twenty-one explanatory variables were included in the environmental data set

145

tot0 wfififig 8o 03682.80. 3 02:00:23.W

 

 

mood wood 0. «ammo. 0038 sonar 505.8: amueomhemummw 06.23% 030402305
mood oood 3 Hmmh 003mm 009% E0600 £030 £030 wEmozsefi
:00 :50 mm :20 080 0382 52:8 3&3 230
35.0 $3 % >m§ 080 803 8:35. sex
Sod 3.3 mm .658 020 :80 E30550 930
omod on o .o mv MUmm 803a wctom $5050 wtoehzmwm
Sod mood w Mkmm woo 080:0 50603 30.20%: wi0§30£
mood mood om mi: 8500.: ham 50500 333.20; Sat
ovod mmmd K 2<Dm 009 28058.8. E00000 $200,580 003%
mmod omod mm 54m 80988200 “000—03608 0300050 «$8505
wood wood m :32. 80988200 80m: E00000 335%.: eESnEEV
oood omod mm <43? 802880—00 3:80-035 0303‘ chfiaoEV
$059580
Bum “00593930 E0005 0900
38:05 5002 00:0 we a 0060mm 080 Z :08800 0802 05625

 

.3088 “00:85:00 08:-008 98 020098050 28 .30.: #053 .80: 50.33 53, 33:00 0.80.8 000$ £6 .3 0003900 no

@003 ooom one moom 08830 E gwfiom—Z o0 235:0; 8033 50538 05 5 20:0— 8035 98 <>>Om :0 380:8 800 208050
E03368 3:85:00 05 E Boa—0E 0200mm 35030080: m~ 05 mo Ammo 00:09:50 85309800 2008 one 00:82 .m.m 030...

146

that included all of the habitat variables, weather variables, and proximity of the sampling
site to water, as well as water depth and width (Table 3.6). The data set was assessed for
multicollinearity issues by computing Pearson product-moment bivariate correlations
between pairwise combinations of variables (Ott and Longnecker 2001). Variables with
high correlations > :1: 0.7 indicated potential multicollinearity problems. Variance
inflation factors (VIF) were also assessed for each variable. A deviation of VIF from 1
indicates a tendency toward collinearity and VIF can be used to measure multiple
correlations between an individual variable and the other variables in the data set
(Chatterjee et a1. 2000, Ott and Longnecker 2001). A high VIF (> 10) indicates that the
variable is highly correlated with other variables (Chatterjee et a1. 2000). Two variables
showed problems with multicollinearity and were removed: percent total cover and
percent live cover. It was expected that these two variables would be highly correlated
with each other and with other habitat variables because they nested (Table 3.6). To
achieve a parsimonious suite of explanatory variables to include in the constrained
ordination, I subjected the environmental variables to forward stepwise variable selection
and backward elimination using Akaike’s information criterion (AIC) (Akaike 1987).
The variables chosen for inclusion in the model were from the three sets of data: habitat
variables (percent canopy cover, distance to water, and litter depth), climate variables
(average maximum daily temperature, average daily precipitation), and categorical
variables (soil association, and ownership type).

To determine whether a linear response model (Redundancy Analysis (RDA),
Constrained Analysis of Principle Coordinates (CAPSCALE)) or unimodal response

model (Canonical Correspondence Analysis (CCA)) was the most appropriate for the

147

Table 3.6. Mean, standard error (SE), and range of habitat and weather variables
recorded in the southern Lower Peninsula of Michigan on SGWA and private lands in
summer 2005 and 2006 and used in model selection for repeated measures, mixed

models, and constrained ordination.

 

 

Variable Model Code Mean (SE) Range
Habitat Variables
Total Cover (%) totalcover 95.0 (0.84) 57.5-100.0
Live Cover (%) totallive 36.0 (2.09) 2.5—78.5
Grass Cover (%) grass 0.9 (0.3 5) 0—21.0
Fern Cover (%) fern 3.3 (0.62) 0—21.5
Forb Cover (%) forb 10.9 (1.30) 0—66.5
Moss Cover (%) moss 1.6 (0.34) 0—18.5
Woody Cover (%) woody 19.3 (1.50) 0—56.5
Standing Dead (%) dead 0.] (0.07) 0—5.0
Litter Cover (%) litter 58.8 (2.47) 5.5—97.5
Bare Ground (%) bare 5.0 (0.84) 0—42.5
Litter Depth (cm) litterdepth 2.6 (0.15) 0.6—5.4
Canopy Cover (%) canopy 94.9 (0.48) 0—99.2
Coarse Woody Debris (# counted) cwd 4.0 (0.37) 0—15.0
Width (cm) width 18.3 (0.83) 10.75—58.0
Length (m) length 8.5 (0.46) 0.5—22.9
Distance to water (m) waterdist 95.7 (1 1.04 0—407.2
Water depth (cm) waterdepth 21.8 (1.89; 2.5—89.0
Water width (m) waterwidth 37.9 (6.16) O.8—340.0

 

148

Table 3.6. Con’t.

 

 

Variable Model Code Mean (SE) Range
Weather Variables

Minimum Daily Temperature (°C) mintemp 14.3 (0.26) 12.5—16.7
Maximum Daily Temperature (°C) maxtemp 25.9 (0.27) 24.1-28.5
Precipitation (mm) rain 4.73 (0.28) 1.5—1 1.1

 

1. 49

 

 

C0115

a ri
The
lin<
joir
We.
loc
be I
fall
the:
1 lo

and

 

 

constrained ordination, I conducted a Detrended Correspondence Analysis (DCA) (Hill
and Gauch 1980). First, I calculated the gradient lengths of the DCA axes, which are the
lengths of the ordination axes in terms of standard deviations in the rate of change in
species composition (Knut et a1. 2003). A length greater than four is indicative of a
unimodal response (bell-shaped response curve) because it allows species to demonstrate
a rise and decline in abundance. Shorter DCA lengths (< 2) indicate a linear response.
The gradient lengths for my community data set were between 1 and 3, indicating that a
linear response model was most likely appropriate (Table 3.7). Second, I produced a
joint plot of sites and species (Figure 3.1). In my joint plot, I found that most species
were located outside of the distribution of my sites, indicating that the species optima was
located outside the range of my sites and that the maximum abundance was predicted to
be outside of the range of my sampled sites. Because the species optima did not rise and
fall within the range of my sites, a unimodal response model was not appropriate, and
therefore a linear response model best described the response model (Figure 3.1). Third,
I looked at the species’ fitted curve on each axis. These showed a mixture of unimodal
and monotonic responses indicating that linear or unimodal responses were appropriate
depending on individual species response. Based on the above criteria, 1 used a
Redundancy Analysis (Rao 1964), which is a linear response model, in combination with
chord distance (row normalization). Standardizations such as chord distance are
recommended for use with RDA to circumvent problems associated with using Euclidean
distance for the analysis of community data sets (Legendre and Gallagher 2001).

To interpret species’ responses to major environmental gradients, 1 produced a

150

Table 3.7. Gradient lengths of the ordination axes from the Detrended Correspondence
Analysis (DCA). Herpetofaunal community data collected on SGWA and private lands
in the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and

area-time constrained surveys.

 

 

DCAl DCA2 DCA3 DCA4
Eigenvalues 0.358 0.234 0.190 0.144
Decorana Values 0.461 0.222 0.135 0.1 10

Axes lengths 2.276 2.438 1.654 2.050

 

151

 

THSA

DCA2

HI)!

 

 

-1
1

 

 

 

 

DCA1

Fig. 3.1. Joint plot of species and sample sites. Species shown in red letters and
represented by codes (see Table 3.5) and sites shown in black numbers. Species fall
outside of sites indicating that a linear response model is appropriate. Herpetofaunal
community data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled
with pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

152

biplot of the significant model. The biplot demonstrated the relationship between
species, the environmental variables, as well as the ordination axes derived from the
species data (ter Braak and Smilauer 1998). Species are represented by points and
environmental variables by vectors, which point toward rate of maximum change and
extend in both directions (ter Braak 1986). The length of the vector indicates its
importance to the constrained ordination (ter Braak 1986). Perpendiculars drawn from
species to vectors give the approximate ranking of that species’ response to the
environmental variable and indicate the species’ optimum on that variable (ter Braak
1986). A smaller angle between the vector and the ordination axis indicates a greater
relationship of the variable to the derived constrained ordination gradient (Grand and
Mello 2004).

To aid in interpretation, generalized additive models (GAMs) were used to fit
surfaces of explanatory variables to the ordination scores. A surface of parallel, straight
contours is indicative of a variable that lies linearly across the ordination space and also
indicates that the relationship between the variable and ordination space is effectively
represented by the linear biplot vector (Oksanen 2007). All data analysis was conducted

with R 2.4.0 (R Development Core Team 2006).

RESULTS
Eighteen species were observed during the study with average species richness of
sites ranging from 0 to 4. Sixteen species were observed on SGWA and 14 on private
lands (Table 3.8). None of the five soil types had all 18 species present (Table 3.9) and

none of the four vegetation types had all 18 species present (Table 3.10).

153

Table 3.8. Species observed by land ownership on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by
drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time

constrained survey including incidental observations.

 

 

Scientific Name Species SGWA Private
Ambystoma laterale blue-spotted salamander X X
Ambystoma tigrinum eastern tiger salamander X X
Plethodon cinereus red-backed salamander X X
Bufo americanus americanus eastern American toad X X
Hyla versicolor eastern gray treefrog X X
Pseudacris triseriata triseriata western chorus frog X X
Pseudacris crucifier spring peeper X X
Rana clamitans green frog X X
Rana sylvatica wood frog X X
Rana pipens northern leopard frog X X
Chelydra serpentina common snapping turtle X

T errapene carolina carolina eastern box turtle X
Chrysemys picta painted turtle X
Thamnophis sirtalis sirtalis eastern garter snake X X
Thamnophis butleri Butler’s garter snake X
Thamnophis sauritus septentrionalis northern ribbon snake X X
Storeria occiptomaculata northern red-bellied snake X
Lampropeltis triangulum triangulum eastern milk snake X

 

154

 

X 0x000 0:00 E00000
X 00000 002—00-000 E05000
X 00000 0000: E05000

0x000 00t0w 0.00—00m

XXXX

X X X X 00000 00000 E00000

0:02 00000

X
X

0:000 00 E00000
0E0. $000000. 0000000
moo E00000 E05000
woo 0003

web 00000

XXXXX

000000 @0000.
woo 00.0000 E00003

w00o000 0% 0007.00

XXXXXXX

n50“ CMUCDE< Evade

XXXXXXXXX

0000000200. 000—00060.

0000000200. 00w: 0000000

XXXXXXXXXX
XXXXXXXXXXX

XXXXX

000008200 000000-003

155

 

mam 5.32 x52 mUE 00603

<
L)
:13

 

000008030 00000205 w0_00_00_ 00800
0050508 0000.000 000 0000000000 000 .0000 0000.0 .0000 =0.t_0 003 020000 080.00 00000 $00 >0 0000000 00 0000p doom 000
moom 0000000 5 0080002 .00 20000000 0300 0.0000000 05 E 0000— 000300 000 <>>Om 00 008 :00 .3 0080000 00603 dd 030%

 

XX

XXXXXXXXXX

x
000300000
05000

XXXXXXXX

X

080

X

XXXXXXXXX

X
00030003
00000002

XXXXXXXXXX

X
00030002
000.300

00.000. 0:00 0000100
000000 00500-000 0000000
00.007. 000000 00000000
000000. 000000 0.00005
0x000. 00000w 00000.00

00000 0000000

0003 x00 0000000

0:08 w0000000 00000000
0000 000002 00000000

wow 0003

M000 0000w

000000 900000

w000 000000 0000003
w0000000 >00w 0000000

0000 00000080. 0000000
000008300 000—000-000
0000000200 00w: 0000000

0000080000 000000-020

000000m

00000300000 00000000000 w0002000 >0>0=m 00000000000
0000-0000 000 0000000030 000 .0000 000000 .0000 00.000 003 0000000 m>0000 0000.0 «00 >0 0008000 00 00000 000m 000 moom
00000000 00 00w00002 .00 03000000 00301— 00000000 000 00 0003 000300 000 <>>Om 00 00>0 000000w0> >0 0030000 m00000m 60m 0300.

156

Mixed Model with Repeated Measures — Considering Weather Effects
Amphibian Species Richness

The mixed model with repeated measures for amphibian richness with the best fit
(lowest AICc value) included the three fixed effects (owner, soil, and vegetation, as well

as all interactions between these terms), and three random effects (minimum temperature,
precipitation, and time period) (Table 3.11). The effect of ownership and vegetation type
was not significant (P = 0.60), however the main effect of soil was significant (P < 0.01)
(Table 3.12). The random effects of minimum temperature (P < 0.01) and precipitation
(P < 0.01) were significant as well (Table 3.13). Comparisons were performed among
the marginal means of the soil associations and among the soil associations, MCP was
found to have a lower mean response than the other 4 soil associations (P < 0.01) (Table
3.14).

Herpetofaunal Species Richness

The mixed model with repeated measures for herpetofaunal species richness with
the best fit (lowest AICc value) included the same variables as the model for amphibian

species richness (Table 3.15). The effect of ownership and vegetation type was not
significant (P = 0.65), however the main effect of soil were significant (P < 0.01) (Table
3.16). The random effects of minimum temperature and precipitation were also
significant (P < 0.01 for both) (Table 3.17). Comparisons were performed among the
marginal means of soil associations and among the soil associations, MCP was found to

have a lower mean response than the other 4 soil associations (P < 0.01) (Table 3.18).

157

 

20000.0 000 00.0

06 0000080000 .00 0000000 000 .030 00303 0000000. A023 02 0300000 A0020 00000000 :00; 00.0 00000000 00000000 0000800000 0.0000004. 0000000 00000.0 0

 

 

000.0

N0 mood 0. mm 0.003 +0000+0§00EE+M003 0.000.... 00:30+M0>* 0.003%?! 00030+>000r 00=3Q+M00+200+00§x0

000.: +000:

2 m2 .o 0.2 0&2: S00+0§00§E+M03 :00... 00:30+M0>* 0.0030000... 00:30i0000. 00:30+M00+000+00030

000.00 +

N0 mm 0 .o m0 meme SE+0§00§E+M03 >000... 00:30+M0>* 000+??? 00:30i000... 00=E0+M00+m00+00030

0000:0000.

m0 0mmd Wm QMNOH + SE+QE$EE+M0$ 0.000.... 00:30+M0>* 0.003%?! 00:30i000... 000030+M0>+200+00=30

000.0

2 03.0 0.0 wde +0000+0§00EE+M03 20010030+M0f 000+M0>f00=30+00$ 00:30+w00+000+00=30
00 _3 0:2. .0? 00oz

 

.m>0>00m 00000000000 0000-0000 000 0000000030 000 .0000 000000

.0000 00000 003 020000 m>0000 0000.0 500 >0 00000000 00 00000 000m 000 Sam 00000000 00 00w00002 00 000000000 00300 00000000 000
00 00000 000300 000 <30m 00 0000000000 00000000 003 0200000 000000 300000 000000. 0000000000 0.000 m 00.0 00000.0 02 .0 0m 200%

158

Table 3.12. Type 3 fixed effects for the amphibian species richness mixed model with

repeated measures (a = 0.05). Data collected on SGWA and private lands in the southern
Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by drift fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area-time constrained

 

 

surveys.
Type of Variation Numerator Denominator P Value Pr>F
DF DF

Owner 1 48.1 2.04 0.16
Soil 4 48.9 4.55 <0.01
Vegetation 3 48.1 1 .93 0.14
Owner*Soil 4 48.5 2.32 0.07
Owner*Vegetation 3 48.1 0.63 0.60
Soil*Vegetation 6 48.1 0.41 0.87
Owner*Soil* Vegetation 5 48.1 0.63 0.68

 

159

 

Table 3.13. Solution for random effects for the amphibian species richness mixed model
with repeated measures (a = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by
drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

 

 

Effect Time DF t Value Pr>|t|
mintemp 55.2 5.96 < 0.01
rain 234 5.74 < 0.01
Time May 4.52 1.92 0.12
Time June 2.77 -2.05 0.14
Time July 3.02 0.21 0.85
Time Agust 3.45 -0.44 0.69

 

160

Table 3.14. Mean comparisons by soil for the amphibian species richness mixed model
with repeated measures. Data collected on SGWA and private lands in the southern
Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by drift fence
arrays coupled with pitfall traps, funnel traps, and coverboards and area-time constrained
surveys.

 

 

Soil Type Mean Standard Error
HCA -2395“ 4.14
MCP -25_o7b 4.14
MCS -24303 4.15
MHE .2356a 4.13
SHB -24, 1 9a 4.15

 

*means within a column with the same letter are not significantly different (a=0.05).

161

0.0008 000

000 0: 00000000000 .00 0000000 000 .030 00w003 0000000.. .6003 02 0300000 .CUEV 00000000 00000 00.0 00000000 00000000 000000000— 0.00=0x< 000.000 000000 0

 

 

0 0 306 ohm 0.02: 0500+RE00EE+M0320P 00:30+M0>* 00:00.03. 00:30i000... 00:30+M0>+~.000+00=30

000.00 +0000+ 2.8.0

2 00 0 .o 0.0 N680 +05008E+M0>0 >000... 00:30+m0>* hoe-+0000... 00030+N000r 000030+M0>+200+00=30

000.0 +500

N0 mmmd 0. m 0.080 +Q500§E+M0320§ 00:30+M0>* 0.00330... 00:30+~.000* 00:§0+M0>+>000+00030

0:00:00?

m0 086 wd m. mmS EE+RESSE+M0>LBP 00:30+M0>* 0.0030000... 00030+000* 00:30+M0>+>000+00=30

0000+

.2 thd od mdmg 2000+S§0EE+M00100P 00:30+M0>* 0.00.330... 00030+2000 00:30+M0>+>80+00=30
00 _a 0000 000.. 0002

 

.0>0>000 00000000000 0000-0000 000 00000000060 000 .00000 00003 .0000

00000 003 0000000 0>0000 0000.0 0000 >0 00000000 00 00000 002.. 000 moom 0000000 00 0030002 00 000000000 0033 00000000 000 00
00000 000300 000 <>>Om 00 00000008 00000000 003 00000000 000000 0000000 0000000 0000000000000 0000 m 00.0 000000 0?. .m 0.0. 2000.

162

Table 3.16. Type 3 tests of fixed effects for herpetofaunal species richness mixed model
with repeated measures (a = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by
drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

 

 

Type of Variation Numerator Denominator F Value Pr>F
DF DF
Owner 1 49.4 1 .64 0.21
Soil 4 50.1 3.70 0.01
Vegetation 3 49.5 1 .72 0.1 8
Owner* Soil 4 49.8 1.64 0.18
Owner*Vegetation 3 49.5 0.55 0.65
Soil*Vegetation 6 49.4 0.37 0.89
Owner*Soil* Vegetation 5 49.5 0.55 0.74

 

163

Table 3.17. Solution for random effects for herpetofaunal species richness mixed model
with repeated measures (a = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in smnmer 2005 and 2006 based on captures by
drifi fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

 

 

Time DF t Value Pr>|t|
mintemp 45.5 5.50 < 0.01
rain 240.0 6.23 < 0.01
May 4.8 1.69 0.16
June 2.9 -2.25 0.11
July 3.2 0.42 0.70
August 3.7 -O.22 0.83

 

164

Table 3.18. Mean comparisons for soil types for herpetofaunal species richness mixed
model with repeated measures. Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by
drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

 

 

Soil Type Mean Standard Error
HCA -22,123* 4.12
MCP -2123" 4.12
MCS .2140a 4.13
MHE -21.78“' 4.11
SHB .223421 4.13

 

*means within a column with the same letter are not significantly different (0L=0.05).

165

 

Herpetofaunal Species Diversity

The mixed mode] with repeated measures for herpetofaunal species diversity with
the best fit (lowest AICc value) included the fixed effects of soil and vegetation, as well

as the random effects of minimum temperature, precipitation, and time period (Table
3.19). The main effects of soil and vegetation type were not significant (P = 0.06 and P =
0.23, respectively) (Table 3.20). The random effects of minimum temperature and
precipitation were significant (P < 0.01 for both) (Table 3.21).

Mixed model — Focus on All Variables (Across Sample Periods)

Amphibian Species Richness
The mixed mode] for amphibian richness with the best fit (lowest AICC value)

included the three fixed effects of owner, soil, and vegetation, as well as all interactions
between these terms, and only the random effect of maximum temperature (Table 3.22).
The effect of ownership was not significant (P = 0.08), however the main effects of soil
and vegetation were significant (P < 0.01 and P = 0.02, respectively), as well as the
interaction between soil and ownership (P < 0.01) (Table 3.23). The random effect of
maximum temperature was significant as well (P < 0.01)) (Table 3.24). Comparisons
were performed among the marginal means of the soil and vegetation types. Among the
soil associations, MCP had a lower mean response than the other 4 soil associations (P <
0.01) (Table 3.25). Among vegetation types, upland hardwood had the greatest mean
response, but none of the 4 vegetation types differed from each other (P > 0.05) (Table

3.26).

166

00008 05

8.0 0: 0000080000 .00 000800 000 6.5 3203 0008? .623 0?. 0350.. A0003 003800 2080 .80 000000.80 00000.5 00008085 0.000000. 00205 0:000m 0

 

 

2 w~ mo 93 Wmmm +E§+QSO0SE+M03 200* 00=§Q+M0f norm?! 002300.580... x0=§o+M0>+2000~M~MMMHW

05.:
N" on fl .o 0.3 0.0 M m +53+Q80NEE+M03200100030+M0358.3%?»x0~§0+>80§00=§Q+M0>+>80+00S$c
a 0m H .o 0.2 m. m H m 0E§+0n0+ SE+R§0§85+ M0>+>80+x0=§o
0 000.0 0.0 0.80 050203300035 +00>+200+0§0
n 2 md 06 Name». 050+SE+QE£SE+M0>+NS0
0 E 020 002 0.002

 

000200 00500008 080-080 0:0 00.00.0050 000 .0000 _0003 .0000

€03 .005 003000 000.30 0000» £00 03 0003000 00 00000 coca 0:0 moom 08800 E 003022 mo 0305:0m 0033 E00000 05 5
00:0— 000309 000 <>>Om co 00.30008 00000900 505 000008 00me 5000030 006000 0000090900 0000 m 08 000000 02 .0 fl .m 030...

167

Table 3.20. Type 3 tests of fixed effects for herpetofaunal species diversity mixed mode]
with repeated measures (a = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by
drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

 

 

Type of Variation Numerator DF Denominator F Value Pr>F
DF

Soil 4 69.1 2.36 0.06

Vegetation 3 68.1 1 .44 0.24

 

168

Table 3.2]. Solution for random effects for herpetofaunal species diversity mixed model
with repeated measures (a. = 0.05). Data collected on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures by
drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

 

 

Effect Time DF t Value Pr>|t|
mintemp 44.2 4.66 < 0.01
rain 147.0 3.99 < 0.01
Time May 4.5 1.91 0.12
Time June 2.8 -1.91 0.16
Time July 3.0 -O.67 0.55
Time August 3.5 0.27 0.81

 

169

 

0000000 05

.80 Cc 0000080000 .00 000800 000 A05 Ems? 0008?. .623 02 0300—00 .Aoogv 003800 :00; 08 000000.80 0200000 000000£E 0.000002 002000 0:000M 0

 

 

0 0 m0 _ .o 00 0&2 000\+0.E00=.§+M0>* >80... 00=§0+w0>s 0030003. 00=§0+mo0r L0=§0+M0>+000+00=§0

£3 +

Na mm _ .0 Wm 0. fl 2 RESRQE+£Q+M0>£NQ00 00=§0+M0>1000+M03 meioimcr. x0=§0+m§+200+00§8

2 00 _ .0 Wm ~62 QE0NEE+M0>020P 00=§0+M0$200+M03 x0=§0+200* x0030+m§+200+00=§0

$00085

_ ~ womd 0.N N62 +0.5? +M0fmofl. 00=§0+M0>LB0+M03 xmnicimcur x0=§0+M0>+200+00=§0

3 nmmd 0.0 0.03 Q500§S+M032000 00:§0+M0>L.80+M0>* gmnioimof 00=§c+m§+000+00=§0
0 _3 020 002 0002

 

000200 0000000000 080-0000 000 00000030

000 .0000 _0003 .0000 =0.t_0 0005 003000 00000 00000 E00 3 00.00000 00 0000p coom 000 moon. 0050.50 0_ 00w332 00 0305000
00.504 0.005000 20 E 0000— 000300 000 <30m 00 0200000 00028 0000000 020000 0053900 0000 m 00.0 000000 02 .mm.m 030...

170

Table 3.23. Type 3 fixed effects for amphibian species richness mixed model (a = 0.05).
Data collected on SGWA and private lands in the southern Lower Peninsula of Michigan
in summer 2005 and 2006 based on captures by drift fence arrays coupled with pitfall
traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Type of Variation Numerator Denominator F Value Pr>F
DF DF
Owner 1 51.1 3.12 0.08
Soil 4 51.1 9.96 < 0.01
Vegetation 3 5 l 3.74 0.02
Owner*Soil 4 51 3.92 <0.01
Owner*Vegetation 3 51 1 .54 0.21
Soil*Vegetation 6 51 0.61 0.72

Owner*Soil* Vegetation 5 51 1.83 0.12

 

171

Table 3.24. Solution for random effects for amphibian species richness mixed model

(a = 0.05). Data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled
with pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

Effect DF t Value Pr > M

 

maxtemp 51.7 5.78 < 0.01

172

Table 3.25. Mean comparisons by soil for amphibian species richness mixed model.
Data collected on SGWA and private lands in the southern Lower Peninsula of Michigan
in summer 2005 and 2006 based on captures by drifi fence arrays coupled with pitfall
traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Soil Type Mean Standard Error
HCA 2323* 0.27
MCP 123,1) 0. 16
MCS 207a 0.14
MHE 2,743 0.39
SHB 2213' 0.13

 

*means within a column with the same letter are not significantly different (0t=0.05).

173

Table 3.26. Mean comparisons by vegetation for amphibian species richness mixed
model. Data collected on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Soil Type Mean Standard Error
Lowland Hardwood 1983* 0.15
Northern Hardwood 2.22a 0.18
Pine 1.91a 0.19
Upland Hardwood 2,358 0.17

 

*means within a column with the same letter are not significantly different (0t=0.05).

174

Herpetofaunal Species Richness
The mixed model for herpetofaunal species richness with the best fit (lowest AICc

value) included the three fixed effects of owner, soil, and vegetation, as well as all
interactions between these terms, and only the random effect of maximum temperature
(Table 3.27). The effect of ownership was not significant (P = 0.15), however the main
effects of soil and vegetation were significant (P < 0.01 and P = 0.02, respectively)
(Table 3.28). The random effect of maximum temperature was significant (P < 0.01)
(Table 3.29). Comparisons were performed among the marginal means of the soil and
vegetation types. Among the soil associations, MCP had a lower mean response than the
other 4 soil associations (P < 0.01) (Table 3.30). Among vegetation types, upland
hardwood had the greatest mean response and pine had the lowest mean response (Table
3.31). Upland hardwoods differed from lowland hardwoods and pine (P = 0.05 and P =
0.03, respectively) (Table 3.31 ).

Herpetofaunal Species Diversity

The mixed model for herpetofaunal species diversity with the best fit (lowest
AICc value) included only soil and maximum temperature (Table 3.32). The effect of

soil and maximum temperature were significant (P = 0.02 and P < 0.01, respectively)
(Table 3.33 and 3.34). Comparisons were performed among the marginal means of the
soil associations. Among the soil associations, MCP had a lower mean response than soil
associations MCS and SHB (P < 0.01 for both) (Table 3.35).
Constrained Ordination

Seven environmental variables accounted for 28.2 % of the total variation in the

herpetofaunal community data set (Table 3.36). The full model was significant

175

200000 000 00.0

C: 0000080000 .00 0000000 000 .033 ES?» 8092 .623 0?. 0300000 A023 0030000 :20 000 08000000 0000000000 00000000000 0.0000010. 002000 000000 0

 

 

0 0 mm 0 .0 N0 m6? EE+Q500EE+M03 0000000=§0+M03 ~000+M0300=§0+2000 00:30+M0>+200+00=§0
00 am 0 .o Wm w.wm_ QE00=NE+M0>0 0.000... 00=§0+M0>*200+M0>* 00=§0+~000* 00=§0+M0>+200+000§0

0800+
N0 mm 0 .0 Wm wme 500+0§00005+M0>0 200... 00=§0+M00f ~000+M0300=§0+~0000 00=§0+M0>+~000+00030
0 0 N. 0 Nd 0.0 o0»; SE+0§00§E+M0$~0000 00:§0+M0>*N.000+M0>* 00=§0+\.000* 00=§0+M0>+200+00=§0
c0 vmmd od mam— Q000008=0+M0>0 0000000=§0+M03 203mg... 00=§0+~000* 00=§0+M0>+200+00=§0
.0 E 0000 .000 0082

 

0.000000 0000000000 00000000 000 00000000>00

000 .0000 0000000 .0000 20,0000 0005 0000000 0.0000 0000.0 500 .00 0008000 00 00000 000m 000 38 00000000 00 0030002 .00 000000000
00.53 0.0000000 05 S 0000— 000300 000 <>>Om 00 00000000 000000 00000000 0000000 0000000000000 0000 m 00.0 000000 02 .nmxr. £000.

176

Table 3.28. Type 3 Analysis of Variance for herpetofaunal species richness mixed model
(a = 0.05). Data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled
with pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Type of Variation Numerator Denominator P Value Pr>F
DF DF

Owner 1 51.1 2.18 0.15
Soil 4 51.1 8.22 < 0.01
Vegetation 3 51.0 3.43 0.02
Owner* Soil 4 51.1 2.47 0.06
Owner*Vegetation 3 51 .0 1 .39 0.26
Soil*Vegetation 6 51.0 0.62 0.72
Owner*Soil* Vegetation 5 51.0 1.79 0.13

 

177

Table 3.29. Solution for random effects for herpetofaunal species richness mixed mode]
(a = 0.05). Data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled
with pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

Effect DF t Value Pr > |t|

 

maxtemp 51.5 5.52 < 0.01

 

178

Table 3.30. Mean comparisons by soil for herpetofaunal species richness mixed model.
Data collected on SGWA and private lands in the southern Lower Peninsula of Michigan
in summer 2005 and 2006 based on captures by drift fence arrays coupled with pitfall
traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Soil Type Mean Standard Error
HCA 143* 0.27
MCP 1,51) 0.16
MCS 2,2a 0.14
MHE 2_4a 0.40
SHB 2,3,a 0.13

 

*means within a column with the same letter are not significantly different (0t=0.05).

179

Table 3.31. Mean comparisons by vegetation for herpetofaunal species richness mixed
mode]. Data collected on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled with
pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Vegetation Type Mean Standard Error
Lowland Hardwood 203* 0.19
Northern Hardwood 2,3ab 0.18
Pine 1,88' 0.23
Upland Hardwood 2,5b 0.16

 

*means within a column with the same letter are not significantly different (a=0.05).

180

0000.0 000 00.0

03 0000080000 .00 000800 000 .033 00303 000002 .623 02 0300—00 $023 0009000 =080 000 08000000 000000000 0000000000000 0.000002 002000 000030 0

 

 

0. 0w 0.0 m.o_ 0. T SE+0§0008E+M0>+200+00=§0
0 am 0 .o 0.00 ON- Q500§E+M0>+mc0+00=§0
m mm 0 .0 5m v.0- 02500005+M0>+200
v 0NN.O Nam 0.x- Q500=05+200
0 mead od #5.. S§08E+200
0 .3 020 .000 0022

 

0.00000 00000000000 080-0000 000 00000000>00

000 .0000 00000.0 .0000 00.000 0005 0000000 0.000 00000 $00 .00 0003000 00 00000 000m 000 meow 008800 00 00300.02 00 03000000
00.500 00000000 000 00 0003 000.000 000 350.0. 00 0200000 0808 30000.00 0000000 0000000000000 0000 m 000 000000 02 .mm.m 030.0.

181

Table 3.33. Type 3 analysis of variance for herpetofaunal species diversity mixed model
(a = 0.05). Data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled
with pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Type of Variation Numerator Denominator F Value Pr>F
DF DF
Soil 4 73.3 3.18 0.02

 

182

Table 3.34. Solution for random effects for herpetofaunal species diversity mixed model
(a = 0.05). Data collected on SGWA and private lands in the southern Lower Peninsula
of Michigan in summer 2005 and 2006 based on captures by drift fence arrays coupled
with pitfall traps, funnel traps, and coverboards and area-time constrained surveys.

 

Effect DF t Value Pr > |t|

 

maxtemp 55.7 3.54 < 0.01

 

183

Table 3.35. Mean comparisons by soil for herpetofaunal species diversity mixed mode].
Data collected on SGWA and private lands in the southern Lower Peninsula of Michigan
in summer 2005 and 2006 based on captures by drift fence arrays coupled with pitfall
traps, funnel traps, and coverboards and area-time constrained surveys.

 

 

Soil Type Mean Standard Error
HCA 2.2231” 0.25
MCP 1.40b 0.16
MCS 2.19a 0.15
MHE 2.90ab 0.40
SHB 2.30a 0.14

 

*means within a column with the same letter are not significantly different (a=0.05).

184

Table 3.36. Summary of the results of the constrained ordination explained by the
environmental variables. Herpetofaunal community data and environmental data
collected on SGWA and private lands in the southern Lower Peninsula of Michigan in
summer 2005 and 2006.

 

 

Inertia Proportion of Variance
Total 0.397 1.000
Constrained 0.1 12 0232*
Unconstrained 0.285 0.71 8

 

*Proportion of variance explained is calculated by dividing the sum of canonical eigenvalues by the total
inertia (sum of unconstrained eigenvalues).

185

(P <0.005) according to the Monte Carlo permutation test, which indicates that the
proportion of species variance explained by the environmental constraints was greater
than expected by chance. Monte Carlo permutation tests also showed that the first 2
canonical axes (RDAl and RDA2) were significant (P < 0.005) (Table 3.37) and all of
the variables except litter depth were significant (P 50 .05) (Table 3.38).

The first axis explained 13% of total species variance (eigenvalue/total
inertia*100). This axis was a gradient of patches with increasing canopy cover and
increasing distance to the nearest water body. The second axis explained 7.5% of the
total species variance. This axis was a gradient of patches with increasing temperature,
overstory canopy cover, and increasing leaf litter depth.

In the biplot, perpendicular projections of species points to the enviromnental
gradients explain the associations of individual species. For example, eastern American
toads were associated with patches that occurred at greater distance to the nearest water
body and less overstory canopy cover (Figures 3.2, 3.3, 3.4); wood frogs were associated
with patches that contained greater overstory canopy cover and leaf litter depth, and
shorter distances to the nearest water body (Figures 3.2, 3.5, 3.6); spring peepers were
associated with patches that contained increasing litter depth (Figures 3.2, 3.7); red-
backed salamanders were associated with patches that contained denser canopy cover and
higher maximum temperatures (Figures 3.2, 3.8); and green frogs were associated with
patches that contained less canopy cover and higher maximum temperatures (Figures 3.2,
3.9).

In the biplot, it is more appropriate to represent categorical variables such as land

ownership type (SGWA and private) and soil association (HCA, MCP, MCS, MHE,

186

Table 3.37. Summary of Monte Carlo permutation tests showing significance of each
axis in the constrained ordination (a = 0.05). Data collected on SGWA and private lands
in the southern Lower Peninsula of Michigan in summer 2005 and 2006.

 

 

Axes Df Variance F # Permutations Pr(>F)
RDA] l 0.05 2.18 200 < 0.01
RDAZ l 0.03 1.23 700 0.03
RDA3 1 0.02 0.80 100 0.23
RDA4 1 0.01 0.32 100 0.96
RDAS 1 < 0.01 0.07 100 1.00
RDA6 1 < 0.01 0.06 100 1.00
RDA7 1 < 0.01 0.02 100 1.00
RDA8 1 < 0.01 0.01 100 1.00
RDA9 1 < 0.01 0.01 100 1.00

RDAlO 1 < 0.01 0.00 100 1.00
Residual 12 0.29

 

'187

 

Tabl
variz
in th

1336
wate
rain
littel
mat
own

soil

Resi

Table 3.38. Summary of Monte Carlo permutation tests showing significance of
variables in explanatory data set (a = 0.05). Data collected on SGWA and private lands
in the southern Lower Peninsula of Michigan in summer 2005 and 2006.

 

 

Variable Df Variance F # Permutations Pr(>F)
canopy 1 0.023 0.98 100 < 0.01
waterdist 1 0.018 0.77 100 0.01
rain 1 0.012 0.49 100 0.01
litterdepth 1 0.008 0.34 100 0.12
maxtemp 1 0.009 0.3 8 100 0.05
owner 1 0.01 1 0.45 100 0.05
soil 4 0.032 0.33 100 < 0.01
Residual 1 2 0.285

 

188

 

 

 

 

 

0. _
‘_
soilMCP
PLCI
‘0. _ '
O
D
(I
O. a
O
HYVE waterdist
SGT/IMHEE BUAM
litter epth SOIIMCS

to. _, t '

max emp :
C? soilSHBE

g RACL
PSCR '
l l i r l
-1.0 -0.5 0.0 0.5 1.0
RDA1

Fig. 3.2. Constrained ordination diagram (biplot). Species are represented by codes (see
Table 3.5), the proximity of species in ordination space indicates occurrence in similar
environmental conditions. Environmental variables are represented by vectors, which
point toward rate of maximum change and extend in both directions. The length of the
vector indicates its importance to the constrained ordination (ter Braak 1986).
Perpendiculars drawn from species to vectors give the approximate ranking of that
species response to the environmental variable and indicate the species optimum on that
variable (ter Braak 1986). A smaller angle between the vector and the ordination axis
indicates a greater relationship of the variable to the derived constrained ordination
gradient (Grand and Mello 2004).

189

R DA2

 

 

RDA2

 

 

 

 

 

 

-1.0 -0.5 0.0 0.5 1.0
RDA1
Fig. 3.3. Plot of eastern American toad abundance in a Generalized Additive Model
(GAM) surface for distance to water (m) on SGWA and private lands in the southern

Lower Peninsula of Michigan in summer 2005 and 2006. Increased size of circles
indicates greater abundance.

190

 

 

 

 

RDA2

Fig. 3
(GM
South
Circle

 

 

RDA2

 

 

 

 

l 1 i I l

 

-1.0 -0.5 0.0 0.5 1.0
RDA1

Fig. 3.4. Plot of eastern American toad abundance in a Generalized Additive Model
(GAM) surface for overstory canopy cover (%) on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006. Increased size of
circles indicates increase in abundance.

191

 

RDA2

 

 

 

 

 

-1.0 -0.5 0.0 0.5 1.0
RDA1
Fig. 3.5. Plot of wood frog abundance in a Generalized Additive Model (GAM) surface
for overstory canopy cover (%) on SGWA and private lands in the southern Lower

Peninsula of Michigan in summer 2005 and 2006. Increased size of circles indicates
increase in abundance.

192

 

RDA2

Fig. 3
for 11'
Mich
abUn

 

 

 

RDA2

 

 

 

 

-1.0 -O.5 0.0 0.5 1.0
RDA1
Fig. 3.6. Plot of wood frog abundance in a Generalized Additive Model (GAM) surface
for litter depth (cm) on SGWA and private lands in the southern Lower Peninsula of

Michigan in summer 2005 and 2006. Increased size of circles indicates increase in
abundance.

193

 

 

RDA2

 

 

 

 

 

 

-1.0 -0.5 0.0 0.5 1.0
RDA1
Fig. 3.7. Plot of spring peeper abundance in a Generalized Additive Model (GAM)
surface for leaf litter depth (cm) on SGWA and private lands in the southern Lower

Peninsula of Michigan in summer 2005 and 2006. Increased size of circles indicates
increase in abundance.

194

 

 

10

 

RDA2

W/////j
//////////°

-LO -05 00 05 1D

 

OD

 

 

 

 

RDA1

Fig. 3.8. Plot of red-backed salamander abundance in a Generalized Additive Model
(GAM) surface for overstory canopy cover (%) on SGWA and private lands in the
southern Lower Peninsula of Michigan in summer 2005 and 2006. Increased size of
circles indicates increase in abundance.

195

 

 

1.0

0.5

RDA2

 

0.0

-0.5

 

 

 

 

 

-1.0 -0.5 0.0 0.5 1.0
RDA1
Fig. 3.9. Plot of green frog abundance in a Generalized Additive Model (GAM) surface
for overstory canopy cover (%) on SGWA and private lands in the southern Lower

Peninsula of Michigan in summer 2005 and 2006. Increased size of circles indicates
increase in abundance.

196

SHB) as points in the data space because they do not have a gradient of change. Classes
containing sites with high values for a particular species will tend to lie in close
proximity to that species in the data space. For example, red-backed salamanders were
associated with patches found in MCP soils and green frogs were associated with patches
found in MCS soil associations (Figure 3.2).

The variance explained by the full model, which consisted of three sets of
environmental factors combined including habitat, weather, and site stratification was
decomposed. Habitat variables accounted for slightly more variance (12.4%) than the
categorical variables (soil type and ownership combined, 10.7%), however climate

related variables accounted for considerably less (4.7%) (Figure 3.10).

DISCUSSION

Results indicated that amphibian and herpetofaunal species richness and diversity
were related to a combination of factors with temperature and soil association as the most
important variables. However when interpreting individual species habitat requirements
and the factors that can influence community composition, a combination of
environmental variables (overstory canopy cover, litter depth, and distance to the nearest
water body), were most important. Land ownership did not play a significant role in my
findings, suggesting that at least in my study area, SGWA and private lands support
similar herpetofaunal communities. These findings are encouraging for herpetofaunal
conservation, as private lands form the landscape matrix in most regions of the Midwest
United States.

Temperature and precipitation are critical to amphibians (Carey et al. 2001).

197

 

Categorical Habitat

 

 

Residual = 71.8

 

 

Fig. 3.10. Percent of total variance in the herpetofaunal community data set in the
southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys explained by habitat variables, site
stratification variables, and weather variables, as well as variance explained by
each combination of factors. Of the variation in the herpetofaunal community,
71.8% of the variance was left unexplained.

198

Breeding is initiated by a response to increasing temperatures and rainfall (Stebbins and
Cohen 1995), and temperature and rainfall also affect the hydroperiod of a wetland
(Pechmann et al. 1989), which in turn affects the rate at which tadpoles complete
metamorphosis and emigrate from ponds. Univariate analyses, suggested that
temperature, whether average minimum or average maximum, played a role in
herpetofaunal species richness, amphibian species richness, and herpetofaunal species
diversity. Warmer temperatures were indicative of greater species richness and greater
species diversity. In all of the mixed models with repeated measures, minimum daily
temperature, precipitation, and time period were included in the best models, but only
minimum daily temperature and precipitation were significant. However, in all of the
mixed models, which included habitat variables in addition to the weather variables, only
temperature was found to affect species richness or diversity, with greater richness and
diversity related to warmer temperatures. These results reinforce the importance of
climatological variables in describing herpetofaunal diversity patterns across the
landscape.

Although microsite conditions such as overstory canopy cover, leaf litter, coarse
woody debris (deMaynadier and Hunter 1995) and distance from the site to the nearest
wetland (Knutson et al. 1999) are known to affect particular species occurrences and
abundances, when focusing on average richness or average diversity, the requirements of
individual species are not taken into account. For species richness, a species observed
once in the sampling period (and that may have been dispersing through the area) has
equal weight to a species that was actually using the area and observed every day during

the sampling period. Individual species abundances and requirements are not considered

199

when looking at richness and diversity, however, when looking at individual species
simultaneously, the constrained ordination effectively showed the importance of selected
microsite conditions on individual species. Thus, it seems more appropriate to consider
individual species abundances simultaneously to make the most effective management
decisions. These two different types of analyses provided different results and when
interpreting results solely on species richness or diversity, critical habitat requirements of
individual species could be overlooked.

Soil association contributed to all of the models but one (mixed model with
repeated measures for herpetofaunal diversity). The soil information used to select sites
for this research projected was Natural Resources Conservation Service 1994 STATSGO
data. Although STATSGO soil associations were assessed on a broad scale, soil
association MCP had an almost consistent lower mean response than the four other soil
associations (MCS, MHE, SHB, and HCA), indicating lower richness and diversity.
Because the soil associations are described on such a broad scale using multiple
components and layers, it is difficult to quantify differences between them, particularly at
the scale of herpetofaunal habitat response. Soil association MCP was dominated by
fine, loamy soils that were not hydric and were well drained. However, components of
the soil association contained loamy, poorly drained soils that had long pond duration
(from November to May). Soil association HCA differed from MCP in that it had muck
soils that were poorly drained. The soils were hydric and had very long pond duration
(from September to June). Soil association MCS had fine loamy soils that were not
hydric, but were somewhat poorly drained. Soil association MHE was dominated by

loamy soils that were not hydric and were well-drained. Components of this soil

200

 

association had hydric soils that had very poor drainage and very long pond duration
(from September to June). Finally, soil association SHB had soils that were dominated
by loamy sand that were not hydric and were well drained. Components of this soil
association were hydric, poorly drained, and had very long pond duration (from
September to June).

Most likely amphibians and reptiles were seeking out areas that contained these
poorly drained soils (HCA, MCS, and SHB) because they were more conducive to
supporting wetlands and the potential to hold water in vernal ponds throughout the
breeding season. All soil associations contained areas with the potential to hold water
throughout the breeding season; however, fine scale components of the soil association
could not be mapped. In general, MCP soils tend to consist of drier soil components,
most likely corresponding to lower herpetofaunal richness and diversity responses.

The fixed effect of vegetation type did not contribute to any of the mixed models
with repeated measures. Although wetlands are the primary breeding habitats for
amphibians, forested areas are also an important component of amphibian life cycles.
Many spend all or most of the non-breeding season in forested areas in shrubs, trees, or
leaf litter and these forested areas can regulate temperature, as well as the rate of
evaporation of surrounding wetlands (Knutson et al. 1999). My results suggest that forest
type is not the important characteristic, but rather the amount of canopy cover regardless
of forest type is important. Amphibian response to even subtle variations in average
canopy cover were identified in this study reinforcing the observation that the effects of
canopy cover on temperature and precipitation are probably important considerations for

herpetofaunal diversity.

201

Vegetation type however did contribute to the mixed models for both amphibian
and herpetofaunal species richness. For herpetofaunal species richness, the mean
responses of upland hardwoods differed from pine stands and lowland hardwood stands.
Mean herpetofaunal species richness was greater in the upland hardwoods. Amphibians
tend to prefer mixed aged forests to even aged forests (deMaynadier and Hunter 1998)
and most pine stands in my study area on both land ownership types were even-aged
plantations. Amphibians have also been found to prefer deciduous forests over
coniferous forests (Strijbosh 1980), and the pine stands where I sampled tended to be dry
with very little to no understory or leaf litter. Five species that were found in upland
hardwoods were not found in pine stands, but lowland hardwoods only differed from
upland hardwoods by two species (Table 3.8). However, my analyses did not focus on
total species richness, but on average species richness per sampling period. On average,
upland hardwoods had more species observed per sampling period than either lowland
hardwoods or pine stands. Although the main effect of vegetation type was different for
amphibian species richness, the mean responses did not differ for any of the vegetation
types. Therefore, the observed effects of vegetation type were more likely determined by
reptile species and less by commonly occurring amphibian species such as wood frogs
and eastern American toads.

When conducting the constrained ordination and analyzing the herpetofaunal
community data set in its entirety, I discovered that habitat variables were more important
in explaining the variance in the data set than the most significant variable, temperature,
in the univariate analyses. The most likely reason for this is that in the constrained

ordination, individual species and their abundances are taken into consideration

202

simultaneously, whereas in the univariate analyses, the community data set is given a
presence absence signature and regarded in terms of number of species present, but not
which species are present. The constrained ordination demonstrates the potential bias
that can occur when using richness or diversity metrics to manage for herpetofaunal
species and the potential ramifications that could result when ignoring specific habitat
and individual species requirements. It is clear that temperature affects species richness
and abundance; however, my results suggest that when investigating factors that structure
herpetofaunal community composition, individual habitat requirements must be
considered.

The landscape of southern Michigan is a mosaic of land ownership and land uses
with differing degrees of naturalness. Land ownership did not show up as a significant
descriptor of the herpetofaunal community; however land use should be taken into
account when deciding and implementing management objectives, because SGWA do
not exist in isolation. SGWA are affected by activities on adjacent or surrounding land
which could alter habitat conditions and impact the effectiveness of management
techniques. My study demonstrates that the southern Michigan landscape should not be
managed as a monoculture, but should contain a diverse array of habitat types. If species
need forested habitat and short distances to wetlands, areas adjacent to ponds or vernal
pools should be left intact to act as buffers around the water bodies. For species like
wood frogs to thrive, they require habitat with denser canopy cover and increased litter
depth with vernal pools in close proximity. Spring peepers also need areas to be
managed for increasing litter depth, while red-backed salamanders need habitat with

denser canopy cover. Green frogs on the other hand require habitat with less canopy

203

cover and habitat generalists like eastern American toads require less canopy cover and
tend to be more terrestrial using areas further away from water bodies.

Herpetofauna have individual life history requirements and these requirements
must be taken into consideration when managing for herpetofaunal diversity. To ensure
conservation of herpetofauna on SGWA requires a broader consideration of the southern
Michigan landscape. My results reinforce the ramifications of managing for
herpetofauna solely within a stand and the how the juxtaposition of patch types can
impact the herpetofaunal community. To preserve herpetofaunal communities in the
ecoregion, the MDNR needs to maintain existing habitat or manage for needed habitat
requirements with regard to the broader southern Michigan landscape to ensure the

integrity of the herpetofaunal community will persist throughout the ecoregion.

204

LITERATURE CITED

Akaike, H. 1987. Factor analysis and AIC. Psychometrika 52(3):317-332.

Alford, R. A., and S. J. Richards. 1999. Global amphibian declines: a problem in applied
ecology. Annual Review of Ecology and Systematics 1999 (30):]33-165.

Blaustein, A. P., and J. M. Kiesecker. 2002. Complexity in conservation: lessons from
the global decline in amphibian populations. Ecology Letters 5:597-608.

Blaustein, A. P., and D. B. Wake. 1995. The puzzle of declining amphibian populations.
Scientific America 272(4):52-58.

Blaustein, A. P., D. B. Wake, and W. P. Sousa. 1994. Amphibian declines: judging
stability, persistence, and susceptibility of populations to local and global
extinctions. Conservation Biology 8(1):60-7l.

Brown, W. S., and W. S. Parker. 1976. A ventral scale clipping system for permanently
marking snakes (Reptilia, Serpentes). Journal of Herpetology 10:247-249.

Bumham, K. P., and D. R. Anderson. 2002. Model selection and inference: a practical
information-theoretic approach. Second edition. Springer-Verlag, New York,
New York, USA.

Burton, T. M., and G. E. Likens. 1975. Energy flow and nutrient cycling in salamander
populations in the Hubbard Brook Experimental Forest, New Hampshire.
Ecology 56:1068-1080.

Cagle, F. R. 1939. A system of marking turtles for future identification. Copeia.
1939:170-172.

Campbell, H. W., and S. P. Christman. I982. The herpetological components of Florida
sandhill and sand pine scrub Associations. Pages 163-171 in N. J. Scott, editor.

205

Herpetological Communities. Fish and Wildlife Service Wildlife Resources
Report 13. Washington, D.C., USA.

Carey, C., W. R. Heyer, J. Wilkinson, R. A. Alford, J. W. Arntzen, T. Halliday, L.
Hungerford, K. R. Lips, E. M. Middleton, S. A. Orchard, and A. S. Rand. 2001.
Amphibian declines and environmental change: use of remote sensing data to
identify environmental correlates. Conservation Biology 15(4):903-913.

Chatterjee, S., A. S. Hadi, and B. Price. 2000. Regression analysis by example. Third
edition. John Wiley & Sons, Inc, New York, New York, USA.

Corn, P. S. 1994. Straight-line drift fences and pitfall traps. Pages 109-117 in W.R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington D.C., USA.

Crump, M. L., and N. J. Scott. 1994. Visual encounter surveys. Pages 84-92 in W.R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington D.C., USA.

Davis, T. M. 1997. Nondisruptive monitoring of terrestrial salamanders with artificial
cover objects on southern Vancouver Island, British Columbia. Herpetological
Conservation 12161-174.

Daubenmire, R. F. 1959. A canopy coverage method of vegetational analysis. Northwest
Science 33:43—64.

deMaynadier, P. G., and M. L. Hunter, Jr. 1995. The relationship between forest
management and amphibian ecology: a review of North American literature.
Environmental Review 3 :230-261 .

deMaynadier, P.G., and M. L. Hunter, Jr. 1998. Effects of silvicultural edges on
distribution and abundance of amphibians in Maine. Conservation Biology
12(2):340-352.

206

 

Enge, K. M. 1997. A standardized protocol for drift-fence surveys. Florida Game and
Fresh Water Fish Commission Technical Report No.14. Tallahassee, Florida,
USA.

Environmental Systems Research Institute. 2004. ArcMap 9.1. Redlands, California,
USA.

Fellers, G. M., and C. A. Drost. 1994. Sampling with artificial cover. Pages 146-150 in
W. R. Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster,
editors. Measuring and monitoring biological diversity: standard methods for
amphibians. Smithsonian Institution Press, Washington D.C., USA.

Grand, J ., and M. Mello. 2004. A multi-scale analysis of species-environment
relationships: rare moths in a pitch pine-scrub oak (Pinus rigida-Quercus
ilicifolia) community. Biological Conservation 119:495-506.

Greenberg, C. H., D. G. Neary and L. D. Harris. 1994. A comparison of sampling
effectiveness of pitfall, single-ended, and double-ended funnel traps used with
drift fences. Journal of Herpetology 28(3):319-324.

Greenberg, C. H., and G. W. Tanner. 2005. Spatial and temporal ecology of eastern
spadefoot toads on a Florida landscape. Herpetologica 61(1):20—28.

Halliday, T. 2000. Do frogs make good canaries? Biologist 47:143-146.

Hecnar, S. J ., and R. T. M’Closkey. 1998. Species richness patterns of amphibians in
southwestern Ontario ponds. Journal of Biogeography 25:763-772.

Hill, M. 0., and H. G. Gauch, Jr. 1980. Detrended Correspondence Analysis: an
improved ordination technique. Vegetatio 42:47-58.

Houze, C. M., and C. R. Chandler. 2002. Evaluation of coverboards for sampling
terrestrial salamanders in South Georgia. Journal of Herpetology 36:75-81.

207

 

Kats, L. B., and R. P. Ferrer. 2003. Alien predators and amphibian declines: a review of
two decades of science and the transition to conservation. Diversity and
Distributions 9:99-110.

Kiesecker, J. M., A. R. Blaustein, and L. K. Belden. 2001. Complex causes of
amphibian population declines. Nature 410:681-684.

Knut, R., R. H. Okland, and T. Okland. 2003. Species response curves along
environmental gradients: a case study from SE Norwegian swamp forests.
Journal of Vegetation Science 14:869-880.

Knutson, M. G., J. R. Sauer, D. A. Olsen, M. J. Mossman, L. M. Hemesath, and M. J.
Lannoo. 1999. Effects of landscape composition and wetland fragmentation on
frog and toad abundance and species richness in Iowa and Wisconsin, USA.
Conservation Biology 13(6):]437-1446.

Legendre, P. and E. D. Gallagher. 2001. Ecologically meaningful transformations for
ordination of species data. Oecologica 129:271-280.

Mallows, C. L. 1973. Some comments on Cp. Technometrics 15:661-675.

McDiarmid, R. W. 1994. Amphibian diversity and natural history. Pages 5-15 in W. R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington D.C., USA.

Michigan Department of Natural Resources. 2001. IFMAP Southern Michigan land
cover digital geospatial data. Department of Natural Resources, Forest, Mineral,
and Fire Management Division, Lansing, Michigan, USA.

Moseley, K. R., S. B. Castleberry, and S. H. Schweitzer. 2003. Effects of prescribed
burning on herpetofauna in bottomland hardwood forests. Southeastern Naturalist
2(4): 475-486.

National Ocean and Atmospheric Administration [NCAA]. 2007. National Weather
Service intemet services team. Monthly temperature for Michigan. <http://

208

www4.ncdc.noaa.gov/cgi-win/wwcgi.dll?wle~SelectStation~USA~MI>.
Accessed 13 January 2007.

Oksanen, J. 2007. Multivariate analysis of ecological communities in R: a Vegan
tutorial. httpzl/cc.oulu.fi/~jarioksa/opetus/metodi/vegantutorpdf. Date accessed:
5 Feb 2007.

Ott, R. L., and M. Longnecker. 2001. An introduction to statistical methods and data
analysis. Fifth edition. Wadsworth Group, Pacific Grove, California, USA.

Pechmann, J. H., K., D. E. Scott, J. W. Gibbons, and R. D. Semlitsch. 1989. Influence of
wetland hydroperiod on diversity and abundance of metamorphosing juvenile
amphibians. Wetlands Ecology and Management 1:3-11.

R Development Core Team 2006. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-
900051-07-0, URL http://www.R-project.org.

Rao, C. R. 1964. The use and interpretation of principal component analysis in applied
research. Sankhya A 26:329-358.

Reaser, J. K., and A. Blaustein. 2005. Repercussions of global change. Pages 60-63 in
M. Lannoo, editor. Amphibian declines: the conservation status and of United
States species. University of California Press, Berkeley, California, USA.

Richter, S. C., and R. A. Seigel. 2002. Annual variation in the population ecology of the
endangered Gopher Frog, Rana sevosa Goin and Netting. Copeia 2002(4):962-
972.

SAS Institute. 2002. Version 9.1. SAS Institute, Cary, North Carolina, USA.

Stebbins, R. C., and N. W. Cohen. 1995. A natural history of amphibians. Princeton
University Press, Princeton, New Jersey, USA.

Strijbosch, H. 1980. Habitat selection by amphibians during their terrestrial phase.
British Journal of Herpetology 6:93-98.

209

ter Braak, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector
technique for multivariate direct gradient analysis. Ecology 67:1167-1179.

ter Braak, C. J. F., and P. Smilauer. 1998. CANOCO reference manual and user’s guide
to Canoco for Windows, software for canonical community ordination (version
4). Microcomputer Power, Ithaca, New York, USA.

USDA Forest Service. 2004. Forest inventory and analysis national core field methods
for phase 3 measurements. USDA Forest Service, Inventory and Analysis,
Arlington, VA, USA. < http://www.fia.fs.fed.us/ library /field-guides-methods-
proc/docs/p3_2-0_secl4_ 3_04. pdf.>. Accessed 3 Sep 2005.

United States Department of Agriculture. 1994. US. General Soil Map (STATSGO) for
Michigan. Natural Resources Conservation Service. Fort Worth, Texas, USA. <
http://www.ftw.nrcs.usda.gov/stat__data.html>. Accessed 10 Jan 2005.

United States Department of Agriculture. 1995. State soil geographic (STATSGO) data
base: data use information. United States Department of Agriculture, Natural

Resources Conservation Inventory, and National Soil Survey Center
Miscellaneous Publication 1492, Washington, D.C., USA.

210

CHAPTER 4: EVALUATION OF THE EFFECTS OF LANDSCAPE PATTERN
METRICS ON HERPETOFAUNA

INTRODUCTION

During recent decades, much attention has focused on the global decline of
herpetofaunal populations, particularly amphibians, throughout the world (Kiesecker et
al. 2001). Although all causes of amphibian declines have not been clearly determined,
anthropogenic habitat modification including habitat loss and fragmentation from
conversion of lands to agriculture, urbanization, and development is one of the best-
documented causes of amphibian declines (Alford and Richards 1999). For example,
both habitat loss and habitat fragmentation are considered to be main causes of
amphibian declines in the Midwest United States (Lannoo 1998).

Amphibians are perceived to exhibit a metapopulation structure (Skelly et a1.
1999, Hecnar and M’Closkey 1996) with several subpopulations occupying optimal
habitat patches and using intervening habitat patches for movement between these high
quality areas. Reptiles and amphibians (collectively called herpetofauna) use several,
distinct habitats during varying parts of their life cycle and as a result may be particularly
vulnerable to anthropogenic changes in the landscape fiom habitat modification,
fragmentation, and loss (Regosin et al 2005). The term landscape complementation was
first used by Dunning (1992) to describe the ability of species to link critical resources
located in different habitat patches to complete their life cycles. Landscape
complementation refers to the proximity of these species’ critical habitats to each other
(Dunning et al. 1992) and connectivity refers to the ease of species’ movement between

these habitats (Taylor et al. 1993).

211

Many herpetofaunal species depend on wetlands such as wooded vernal pools,
streams, and river floodplains; adjacent terrestrial habitat; and connections between the
two (Semlitsch and Bodie 2003, Calhoun et al. 2005). Ephemeral pools are considered
important habitat for several herpetofaunal species such as wood frogs (Rana sylvatica)
and blue-spotted salamanders (Ambystoma laterale) (Calhoun et al. 2005). These species
prefer isolated wetlands that have seasonal hydroperiods and lack predatory fish (Morin
1983, Russell et al. 2002). Fragmentation of forested habitat can degrade habitat quality,
especially for edge-sensitive species, and can eliminate continuous blocks of habitat,
thereby reducing between patch access (Primack 2004). F ragrnentation can also lead to
the wetland isolation which can firrther breakdown critical metapopulation processes.
Maintaining connectivity between wetlands and terrestrial habitat is essential for
herpetofaunal species that breed or lay eggs in wetlands and emigrate to terrestrial
habitats to forage and overwinter (McDiarmid 1994).

Research has been conducted at the landscape-scale to determine the relationship
between herpetofaunal community compositions and the surrounding landscape or land-
use patterns (e.g., Hecnar and M’Closkey 1996, Knutson et al. 1999, Hermann et al.
2005). Land uses such as agricultural or urban development reduces availability of
optimal habitat, tends to reduce habitat suitability and landscape connectivity, and can
have a negative impact on amphibian abundance and occurrence (Bonin et al. 1997,
Hecnar 1997, Knutson et al. 1999). These land uses can also result in isolation of
wetlands, which can lead to reduced species richness at isolated sites due to reduced
interactions with other populations (Findlay and Houlahan 1997, Semlitsch and Bodie

1998, Lehtinen et al. 1999). Agriculture and urban land uses can also act as movement

212

barriers between populations reducing recolonization rates and reducing local species
richness (Findlay and Houlahan 1997, Lehtinen and Galatowitsch 2001, Parris 2006).

Development of road systems throughout the human-dominated landscape can
result in the direct loss of habitat, fragmentation of habitat, and the degradation of habitat
quality (Jackson 2000). It can also result in modification of an animal’s behavior such as
altered movement patterns (Trombulak and Frissell 2000). Roads cause direct mortality
which can reduce population sizes, as well as reducing movement between populations of
species and their required resources (Carr et al. 2000). Road systems can act as
movement barriers which can reduce access to high quality habitat and further isolate
populations (Jackson 2000).

In much of the Midwestern United States, the majority of historic land conversion
resulted from draining wetlands and clearing land for agricultural practices (Brinson and
Malvarez 2002). In southern Michigan, the landscape has been anthropogenically
modified and presently consists of a mosaic of land uses including agriculture, forests,
and urban development in both metropolitan and rural areas. Currently, Michigan is
dominated by private lands (79%) with only 21% being publicly held (Michigan
Department of Natural Resources 2000). Some portions of southern Michigan are being
managed to conserve and restore wildlife resources by the Michigan Department of
Natural Resources (MDNR) such as Michigan’s State Game and Wildlife Areas
(SGWA). SGWA have been established and maintained to improve and restore wildlife
populations and habitat and these areas are presumed to be essential components in the
conservation of biological diversity throughout the state. Managing for species like

amphibians that require the use of several distinct spatially discrete habitats can be

213

challenging in a landscape that is human dominated, particularly in areas that are
urbanized (Gibbs 2000). This study provides an opportunity to investigate an
anthropogenic modified mosaic of land ownerships and land uses, and to determine their
influence on herpetofaunal communities. The objective of this study was to assess how
patch characteristics over multiple scales influences (including land cover type, distance
to roads, and distance to surface water) the occurrence and relative abundance of
amphibians and reptiles on SGWA and private lands in the southern lower peninsula of
Michigan.

All capture, handling, and marking protocols were approved by the Michigan

State University Animal Care and Use Committee (AUF# 07/03-082-00).

METHODS

To assess the influence of anthropogenic landscapes on herpetofaunal
communities in the southern lower peninsula of Michigan, 82 sites were sampled on
SGWA and on privately-owned parcels in the spring and summer of 2005 and 2006.
Drift Fence Arrays

Drift fence arrays made from 60 cm high aluminum flashing with pitfall and
funnel traps were used to capture herpetofauna (Corn 1994, Enge 1997). Drift fences
intercept herpetofauna moving on the ground and re-direct them into a pitfall or funnel
trap. Drift fences with pitfall and funnel traps were installed in April and early May prior
to opening the traps in mid-May. They were opened for 5 consecutive nights each month
in 2005 and 4 consecutive nights each month in 2006 from May through August (Table

4.1) and were checked once daily. Three 5 m long sections of aluminum flashing were

214

Table 4.]. Dates drift fence arrays, pitfall traps, funnel traps and coverboards were
opened in the southern Lower Peninsula of Michigan on SGWA and private lands in
summer 2005 and 2006.

 

 

Date Opened Date Closed Number of Sites
05/ l 7/2005 05/22/2005 20
05/24/2005 05/29/2005 22
06/06/2005 06/1 1/2005 20
06/14/2005 06/19/2005 22
07/1 1/2005 07/16/2005 20
07/19/2005 07/24/2005 21
08/05/2005 08/09/2005 20
08/10/2005 08/14/2005 21
05/08/2006 05/12/2006 1 7
05/1 5/2006 05/19/2006 14
05/22/2006 05/26/2006 9
06/06/2006 06/ 16/2006 1 7
06/12/2006 06/1 6/2006 20
06/ 1 9/2006 06/23/2006 3
07/05/2006 07/09/2006 1 7
070/9/2006 07/ 1 3/2006 3
07/12/2006 07/16/2006 20
07/31/2006 08/04/2006 20
08/07/2006 08/1 1/2006 20

 

215

installed in a Y arrangement and 4 pitfall traps and 6 funnel traps were placed within the
array (see Fig.1 Enge 1997). Arrays were oriented to the north. Pitfall traps were made
from 18.9 L buckets. Holes were drilled approximately 2 cm from the bottom of the trap
to allow for drainage of rainwater (Enge 1997). The pitfall traps were buried slightly
below ground level, allowing animals to drop into the bucket. Moistened sponges were
placed in all pitfall traps to prevent desiccation of captured animals (Greenberg et al.
1994). The sponges were remoistened as needed (Enge 1997 , Richter and Seigel 2002).
Traps were closed by placing lids over the buckets.

Funnel traps were placed at the midpoint of each wing of the aluminum flashing
in the array (Greenberg and Tanner 2005). Funnel traps were double entry with the main
body comprised of aluminum window screening and the firnnels of flexible fiberglass
screening. Traps had 20 cm openings at both ends with funnel openings of 5 cm in
diameter (Corn 1994). When not in use, funnels traps were closed by inverting the
funnels and clipping them shut.

Coverboards

Coverboards (F ellers and Drost 1994, Davis 1997) were placed within Im from
the drift fence array in the four cardinal directions at least one week prior to drift-fence
array installation. Coverboards were made of untreated birch plywood and cut into 1 m x
l m sections and were placed on bare ground. Coverboards are designed to provide
moist, cool refuge for herpetofauna (Houze and Chandler 2002) and create a microhabitat
similar to a downed log. Coverboards were checked every day that the drift fence arrays

were open.

216

 

Area Time-Constrained Surveys

Area time-constrained surveys (ATC surveys) (Campbell and Christrnan 1982,
Crump and Scott 1994) were conducted on each site once a month from May to August in
2006: 23 May — 31 May, 1 June- 2 June, 19 June — 27 June, 19 July — 26 July, 1 August —
12 August; and in June and July in 2005: 22 June — 30 June, 22 July — 31 July. A 2 m x
37 m area (the same total area as the drift fence array) was delineated to conduct the area
time-constrained surveys, and this same area was used for surveys throughout the season.
ATC survey areas were selected near the drift fence array in one of the four cardinal
directions, starting east of the array. When a 2 m x 37 m area fit in the designated soil
and vegetation type, it was georeferenced with a GPS unit and flagged. This designated
area was hand-searched for a period of 20 minutes by overturning downed logs and rocks
and searching through leaf litter. Decay classes of the logs were recorded according to
the US. Forest Service’s FIA Field Methods for Phase 3 Measurements, five decay
classifications (USDA Forest Service 2004) (Table 4.2).
Mark-Recapture Techniques

All animals captured by pitfall traps, funnel traps, cover boards, and area time-
constrained searches were processed before being released. Herpetofauna were identified
to species, sexed (when possible), and marked. All amphibians were measured (snout to
vent length, mm); and northern leopard frog (Rana pipiens), American toad (Bufo

americanus), and green frog (Rana clamitans) with a snout to vent length measurement >
50 mm were PIT tagged (passive integrated transponders; AVID®). Salamanders and all

other Anurans were marked by toe-clipping except treefrogs and relatives, but not for

individual recognition. Snakes and turtles were measured and marked for individual

217

Table 4.2. Five decay classifications of downed logs (US. Forest Service 2004) used in
area time-constrained surveys in the southern Lower Peninsula of Michigan on SGWA
and private lands in summer 2005 and 2006.

 

 

Class Description

L1 Bark intact, twigs present. Texture is intact. Wood is original in color.
Log is elevated on supported points above ground.

L2 Bark intact, twigs absent. Texture is intact to partially soft. Wood is
original in color. Log elevated on support points but sagging slightly.

L3 Trace of bark. Twigs are absent and texture is hard large pieces. Color of
wood is original to faded. Log is sagging near ground.

L4 Bark and twigs are absent. Texture of wood is small, soft, blocky pieces.
Wood is light brown to faded brown or yellowish. All of the log is on the
ground.

L5 Bark and twigs are absent. Texture of wood is soft and powdery. Color of

wood is faded to light yellow or gray. The diameter of the log is
attainable wood and log debris is not spread out in a flat manner. If a

diameter is not attainable, then it is not considered a log, but a pile of
debris.

 

218

recognition. All snakes were marked by clipping half of one ventral scale (Brown and
Parker 1976) and turtles were marked by notching marginal scutes (Cagle 1939).
Individuals were released at least 5 m from the point of capture on the opposite side of
drift fence to minimize the probability of immediate recapture.

Landscape Analysis

To quantify and characterize the landscape on SGWA and private lands, three
different analysis distances (AD) were generated around each survey site at 100 m, 200
m, and 1000 m radii. The center of each AD was the center pitfall trap of each drift fence
array. ADs of 100 m and 200 m were chosen to quantify and characterize the area in
close proximity the survey point. An AD of 1000 m was chosen because it is a
biologically relevant estimate of dispersal and migration distances for the species in my
study area (Berven and Grudzien 1990, Stebbins and Cohen 1995, Lannoo 2005) and has
been used in several other amphibian landscape pattern metric studies (e. g., Knutson et
al. 1999, Guerry and Hunter 2002, Genet 2004, Price et al. 2004).

Drift fence arrays sites were georeferenced in the field and the coordinates were
displayed in ArcGIS (Environmental Systems Research Institute 2004). I analyzed
landscape composition with a vegetation land cover data map developed by Michigan
Department of Natural Resources (Michigan Department of Natural Resources 2001).
The map was created from 1999-2001, had a 30 m pixel size and contained 7 level I
classes of land use (Table 4.3). The land use classes were reclassified to include six land
use types: agriculture (row crops, tilled crops, fields, orchards, vineyards, herbaceous
Openland, low density trees, and parks, golf courses), forest (all forest types), wetlands

(non-forested wetlands), water (ponds, lakes, and rivers), and urban (commercial,

219

Table 4.3. Landscape classes and IFMAP categories developed by Michigan Department

of Natural Resources 1999-2001.

 

Landscape Categories and Associated
IFMAP Class Names and Grid Values

Class Descriptions

 

Urban

Low Intensity Urban (Residential) (1)

Intensity Urban (2)

Airports (3)

Roads / Pavement (4)
Agricultural
Non-vegetated agriculture (5)

Row Crops (6)

Forage Crops (7)

Orchards/Vineyard/Nurseries(9)
Openland Upland

Herbaceous Openland (10)

Low Density Trees (12)

Parks, Golf Courses (13)

Forested
Deciduous Forest
Northern Hardwoods (14)

Oak Type (15)
Aspen Type (16)
Other Upland Deciduous (l 7)

Mixed Upland Deciduous (18)

Lowland Deciduous Forest (24)
Coniferous Forest (CF)

Pines (19)

Other Conifers (20)

Mixed Upland Conifers (21)
Lowland Coniferous Forest (25)

Upland Mixed Forest (22)

Lowland Mixed Forest (26)

Landareatrlledforcropproducnonwrth<2596

Land area >10% and <25% manmade structures,
including paved and gravel roads and parking lots.
Land area >25% solid impervious cover made from
manmade materials, other than airports, roads, or
parking lots.

Impervious land within airport grounds, including
runways.

Roads or parking lots

   

 

 

currently vegetated.

Vegetation rs annual crops planted rri rows (e. g.j_"
corn, soybeans).

Vegetation used for fodder production (6. g. alfalfa,
hay). Also includes land used for pasture, or non-
tilled herbaceous agriculture.

Woody trees not grown for Christmas trees.

<25% of land area is covered by woody cover.
The combination of woody shrubs and trees is
>25% of the land area and >25% of the woody
cover is trees. None mapped in SLP.

Upland open land maintained for recreational
purposes.

Combination of maples, beech, basswood, white
ash, cherry, and yellow birch >60% of the canopy.
Proportion of oaks >60% of the canopy.
Proportion of aspen >40% of the canopy.
Proportion of any other single species >60% of the
canopy.

Proportion of deciduous trees >60% of the canopy.
Proportion of deciduous trees >60% of the canopy.

Proportion of pines >60% of the canopy
Proportion of non-pine upland conifers >60% of the
canopy.

Proportion of coniferous trees >60% of the canopy.
None mapped in SLP.

Proportion of coniferous trees >60% of the canopy.
Mixed forest not falling into any other category.
Proportion of conifers: deciduous ranges between
40%:60% to 60%:40%.

Mixed forest not falling into any other category.
Proportion of conifers: deciduous ranges between
40%:60% to 60%:40%.

 

Table 4.3. Con’t

 

Nonforested Wetland
Lowland Shrub (28) Proportion of lowland shrub >60% of non-water
cover.
Emergent Wetland (29) Proportion of emergent wetland >60% of non water
cover.
Mixed Non-forest Wetland (30) Non-forested wetlands not falling into any other
category.
Water
Water (23) Proportion of open water >75% of the land area.
Bare/Sparsely Vegetated
Sand, Soil (31) Land cover is formed primarily of sand or bare soil.
Exposed rock (32) Land cover is formed of solid rock. None mapped in
SLP.
Other Bare\Sparsely Vegetated (3 5) None.

 

221

industrial, and residential), and other. The road network data and depth to ground water
data used in the analysis were obtained from the Michigan Geographic Data Library

(Michigan Center for Geographic Information 2007).
At each spatial scale, the following metrics were calculated: total area (m2) of

each of the six land cover categories, total road length (m), mean depth to ground water

(m), distance to nearest road (m), distance to nearest urban (m), and distance to nearest
surface water (m). The total area (m2) of each of the six land cover categories was

calculated at each scale using the focal sum firnction under spatial analyst tools. The total
road length (m) was calculated by clipping the road layer with each buffer and the length
of individual road arcs were recalculated using Hawth’s tools (Beyer 2004). The clipped
road layer was intersected with the respective buffer and assigned to a sample site. The
road lengths were summed by sample site to get the total road length within each buffer
for each sample site. The mean depth to ground water (m) was calculated using the
landscape characterization function under Hawth’s tools. In addition, the distance to
nearest road (m) and distance to the nearest urban area (m) were calculated using the near
command under spatial analyst tools. The distance to the nearest surface water (m) was
calculated by extracting surface water <= 0 and creating a new layer fi'om this data. This
layer was converted from raster to points and this created a point at that center of each
30m pixel on the surface water raster. The near command was used to measure the
distance from each sample site point to the nearest surface water point.

AD data were analyzed separately to determine landscape effects at the three
different scales. Because of the potential for spatial overlap among sites and the ensuing

lack of independence, only non-overlapping sites were included in the analysis. A

222

random selection process was used to select the non-overlapping sites fiom the site
population.
Statistical Analysis
Mixed Model

After examination of descriptive statistics and normal probability plots for
original variables and potential data transformations, all variables were log transformed

to normalize the data and homogenize variance. Certain criteria were used to select
variables for inclusion in the model including R2, Adjusted R2, Cp (Mallows 1973)

statistics, forward selection, and backward selection. Outliers, collinearity, residual plots,
meaningfulness and interpretation were examined for candidate models. A mixed model
was used to determine which variables (land ownership and landscape variables) were
most influential in determining amphibian richness, herpetofaunal species richness, and
herpetofaunal species diversity (collectively called herpetofaunal species abundance data)
at the 100 m, 200 m, and 1000 m scale. The treatment design consisted of one fixed
factor: land ownership with two levels (SGWA vs. private) and 11 potential random
factors. Models were ranked by how well they explained amphibian richness,

herpetofaunal species richness, and herpetofaunal species diversity data using Akaike’s

information criterion corrected for small sample sizes (AICc) (Bumham and Anderson
2002). AICC was calculated as follows:
AICc = -2*ln(likelihood) +2*K + (2*K*(K+l))/(n-K-l),

where In is the natural logarithm, (likelihood) is the value of the likelihood, and K is the

number of parameters in the model, n is the sample size (Bumham and Anderson 2002).

Models were also evaluated using AIC weights (wi), which can be treated as the weight

223

 

of evidence in favor of a particular model, given the data set and the set of candidate
models (Bumham and Anderson 2002). The factor main effects and the interaction
effects were analyzed using ANOVA. When main effects were found to be significant (a
< 0.05), a treatment comparison among the marginal means was conducted. The data
analysis was conducted using PROC MIXED (SAS Institute 2002).

Constrained Ordination

To examine the relationship between herpetofaunal species and the environment
using landscape variables, a constrained ordination was used. The constrained ordination
was conducted using landscape variables at all 3 spatial scales for one independent data
set. The herpetofaunal species abundance data were the response variables and the
environmental data were the explanatory variables. The goal of the constrained
ordination was to infer patterns in herpetofaunal species composition from patterns in the
environmental variables.

I removed species from my herpetofaunal community data set with less than 2
occurrences because they were most likely not sufficiently sampled and would not be
accurately placed in ecological space. Of 18 species sampled, 12 species were retained
(Table 4.4). The species data were highly skewed and heteroskedastic, so the data was
log transformed as well as column normalized using R 2.4.0 (R Development Core Team
2006)

Twelve explanatory random variables were included in the environmental data set
for each of the 3 different spatial scales including distance to the nearest surface water,
distance to the nearest road, distance to the nearest urban area, and landscape variables

total road length, mean ground water depth, and the total area of land class (agriculture,:

224

:83 329:8 no.“ 3~€8v=8m 2 00cnv==n<a

 

 

wood wood N «swig. 0x05 coast 805.8: $303333 marge maemeesefi
mood mood m 53H 0035 Stem E0300 33.5 302.2 flamoefiefi
:oo mmoo w a; wet 3302 505.5: Enema 38%
mwoo mmoo em >m<m woo 303 3.203% 38%
God 2 3 2 SE woe :80 assesses use
mmoo 3 _.o ofi MUmm n0m00o macaw Shoat miefiexam
flooov ~oo.o N Mhmm mot maono E0503 SEEMS wroehxak
wood wood 5 mar woo00b 3% E0300 330.50; 03E
omod o g mo NN 2<Dm 39 323:2. E030 .32001053 exam
wmoo emoo n 5.5 3988030 303368 $3.050 293303
omoo mmoo m 2.32 3988200 Ewe E0300 532.2%: 02838wa
nooo omoo o <43? “0288030. 3325.035 03.03. eESbEEV
“0050953
“ohm 3&3355 0:032“ 030
330.8% 502 v.05 we a 0060mm 080 Z :oEEoU 0:8 Z 0EEOBm

 

 

.9028 350.538 095-003 30 0202330 30 £93 L053 .805 33% 5:5 33:00 .983 00:0o 5.6 E 3538 :o

303 ooom one moom .5883 E 93332 no 335:3 0030..— 50538 0E 5 3:2 03>?» 98 <>>Om co 380:8 800 .mmmbmem
83530 3:85:00 05 5 39:05 360% 3503330: .2 0% mo Ammo 0020953 3N€So§am .808 one 380 Z .vé 030,—.

225

 

forest, open water, wetland, urban, and other). The data set was assessed for
multicollinearity issues by computing Pearson product-moment bivariate correlations
between pairwise combinations of variables (Ott and Longnecker 2001). Variables with
high correlations > :t 0.7 indicated potential multicollinearity problems. Variance
inflation factors (V IF) were also assessed for each variable. A deviation of VIF from 1
indicates a tendency toward collinearity and VIF can be used to measure multiple
correlations between an individual variable and the other variables in the data set
(Chatterjee et al. 2000, Ott and Longnecker 2001). A high VIF (> 10) indicates that the
variable is highly correlated with other variables (Chatterjee et a1. 2000). To achieve a
parsimonious suite of explanatory variables to include in the constrained ordination, I
subjected the environmental variables to forward stepwise variable selection and
backward elimination using Akaike’s information criterion (AIC) (Akaike 1987). The
random variables chosen for the model were 4 land class categories: urban at a 100 m
scale, wetland at a 200 m scale, forest at a 200 m scale, and wetland at a 1000 m scale.
The fixed effect of ownership was also included in all models.

To determine whether a linear response model (Redundancy Analysis (RDA),
Constrained Analysis of Principle Coordinates (CAPSCALE)) or unimodal response
model (Canonical Correspondence Analysis (CCA)) was the most appropriate for the
constrained ordination, I conducted a Detrended Correspondence Analysis (DCA) (Hill
and Gauch 1980) to evaluate three items. First, I calculated the gradient lengths of the
DCA axes, which are the lengths of the ordination axes in terms of standard deviations in
the rate of change in species composition (Knut et al. 2003). A length greater than four is

indicative of a unimodal response (bell-shaped response curve) because it allows species

226

to demonstrate a rise and decline in abundance and shorter lengths (< 2) indicate a linear
response (a straight line response). The gradient lengths for my community data set were
between 1 and 3, indicating that a linear response model was most likely appropriate
(Table 4.5). Second, I produced a joint plot of sites and species (Figure 4.1). In my joint
plot, I found that most species were located outside of the distribution of my sites,
indicating that the species optima was located outside the range of my sites and that the
maximum abundance was predicted to be outside of the range of my sampled sites.
Because the species optima did not rise and fall within the range of my sites, a unimodal
response model was not appropriate, and therefore a linear response model best described
the response model (Figure 4.1). Third, I looked at the species’ fitted curve on each axis.
These showed a mixture of unimodal and monotonic responses. Based on the above
criteria, I used a Redundancy Analysis (Rao 1964), which is a linear response model, in
combination with chord distance (row normalization). Standardizations such as chord
distance are recommended for use with RDA to circumvent problems associated with
using Euclidean distance for the analysis of community data sets (Legendre and
Gallagher 2001).

To interpret species’ responses to major environmental gradients, I produced a
biplot of the significant model. The biplot demonstrated the relationship between
species, the environmental variables, as well as the ordination axes derived from the
species data (ter Braak and Smilauer 1998). Species are represented by points and
environmental variables by vectors, which point toward rate of maximum change and
extend in both directions (ter Braak 1986). The length of the vector indicates its

importance to the constrained ordination (ter Braak 1986). Perpendiculars drawn from

227

Table 4.5. Gradient lengths of the ordination axes from the Detrended Correspondence
Analysis (DCA). Herpetofaunal community data collected on SGWA and private lands
in the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on

captures by drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys.

 

 

DCA1 DCA2 DCA3 DCA4
Eigenvalues 0.372 0.204 0.131 0.143
Decorana Values 0.379 0.155 0.097 0.048
Axes lengths 2.517 1.460 1.515 1.425

 

228

 

AMT]

‘— .. 3 THSA

 

 

 

Sf , 3 1 10
U 2 E
1 : BUAM
o ‘_ 2 16 I
O" _ PSCR
"2 -
PSTR
1 l I i l I l 1
-3 -2 -1 0 1 2 3 4
DCA1

Fig. 4.1. Joint plot of species and sample sites. Species shown in red and represented by
codes (see Table 4.4) and sites shown in black numbers. Species fall outside of sites
indicating that a linear response model is appropriate. Herpetofaunal community data
collected on SGWA and private lands in the southem Lower Peninsula of Michigan in
summer 2005 and 2006 based on captures by drifi fence arrays coupled with pitfall naps,
funnel traps, and coverboards and area-time constrained surveys.

229

 

 

species to vectors give the approximate ranking of that species’ response to the
environmental variable and indicate the species’ optimum on that variable (ter Braak
1986). A smaller angle between the vector and the ordination axis indicates a greater
relationship of the variable to the derived constrained ordination gradient (Grand and
Mello 2004).

To aid in interpretation, generalized additive models (GAMs) were used to fit
surfaces of explanatory variables to the ordination scores. A surface of parallel, straight
contours is indicative of a variable that lies linearly across the ordination space and also
indicates that the relationship between the variable and ordination space is effectively
represented by the linear biplot vector (Oksanen 2007). The data analysis was conducted

with R 2.4.0 (R Development Core Team 2006).

RESULTS

Three land cover types dominated the landscape at all three spatial scales: forest,
agriculture, and wetland. In both the 100 m and 200 m AD, forest was the dominate land
cover type around the sampling sites, comprising 77% of the total AD area and 64% of
the total area, respectively (Table 4.6). This is consistent with how site selection was
conducted as the study was designed to characterize herpetofauna of forested
communities. However, in the landscape 1000 m from the survey site, agriculture was
the most dominate land cover type, comprising 44% of the total AD area (Table 4.6). At
the 100 m scale, two other land cover types comprised 11% each of the total AD area:

agriculture and wetland (Table 4.6). Water and urban made up less than <0.01% of

230

 

3.0 50.8 92.2 adv 3: Ma adv 2: o merge

m; 88.3 53%. Ed 8%.: 33; :.o 23 $3 $52253
Nod :33 83.: 5o 3: E adv as me 96:35
.86 gas: 31.8 5.0 8me 5w adv a: .2 mavens, :26
one $3.8: 26.02; 3o sons $5.0 Rd 5.3 32.8 $8528

36 23.3.: $0.83 85 Azod was :.o $3 $3 £5 ea_8&<
.85 GE 502 .85 Q8 502 can Ame 50:
e 82 a com a 2: 0385/

 

 

 

.25 H260 2:: some 3 vomtmfioo 88 Q< 33 2: mo :oEomoa 05 manomoaoa

.995 98 A3" ZV E 83 ”Gm" Zv 8 com x3. .I. 7c 8 o2 new 8% :08 5 8% 33.0. _E 88m Ammv £88 333% @8308?
:2: £538 v.8 mos—Hg .8ng 38228 :03 55:5 ANEV new :39 on. 380358 $95 530 9:2 =< doom can moon

E §wEom2 mo 235:3 533 c.8538 05 E 838 332mm 085 05 “a 38.0.38 893 .530 98— mo QEQSm .3“ 28d.

231

the total AD area within 100m of the survey sites and the other landscape category was
not present (Table 4.6). In the landscape up to 200 m from the survey site, a similar
pattern was observed with agriculture covering 20% and wetland covering 14% of the
total AD area (Table 4.6). Open water and urban both comprised 1% each of the total
AD area (Table 4.6). In the landscape up to 1000 m from the survey site, forest
comprised 39% and wetland comprised 13% of the total AD area. Open water and urban
both comprised 2% and other comprised less than <0.01% (Table 4.6).

There was an average road length of 18.9 m (5.8 SE) within the 100m AD area;
an average road length of 111.5 m (24.4 SE) within the 200m AD area; and an average
road length of 4,575 m (372.9 SE) within the 1000 m AD area (Table 4.7). There was an
average ground water depth of 1.9 m (0.3 SE) within 100m of survey sites; an average
ground water depth of 1.8 m (0.4 SE) within 200m of survey sites; and an average ground
water depth of 2.5 m (0.4 SE) within 1000 m of survey sites (Table 4.7).

Because data sets differed (75 sites at 100m; 55 sites at 200m; 25 sites at 1000m),
the average distances to the nearest road, urban area, and surface water differed among
landscape scales. The average distance to road was the greatest for the 200 m data set
(Table 4.7) and the average distance to the nearest urban area was greatest for the 100 m
data set (Table 4.7). The average distance to the nearest surface water was greatest for
the 1000 m data set, followed by the 100 m data set (Table 4.7).

Mixed Model - 100 m Scale

Amphibian Species Richness

The mixed model for amphibian richness with the best fit (lowest AICc value)

included one fixed effect (owner) and one random effect (forest (m2) (Table 4.8). The

232

Table 4.7. Summary of landscape variables measured at the three spatial scales in the
southern Lower Peninsula of Michigan in 2005 and 2006. Water depth represented the
average of ground water depth (m) within each AD; surface water represented the
distance from survey site (m) to nearest surface water; distance- road represented the
distance from survey site (m) to the nearest road; road length represented the total length
of roads (m) contained within each AD; distance-urban represented the distance from
survey site (m) to the nearest urban area. Values are means and their associated standard
errors (SE) from all study sites in each data set: 100 m (N = 75); 200 m (N =55); 1000 m

(N =25).

 

 

 

Variable 100 m 200 m 1000 m
(n=75) (n=55) (n=25)

Mean (SE) Mean (SE) Mean (SE)

Water depth (m) 2 (0) 2 (0) 2.0 (0)
Surface water (m) 52 (7) 50 (7) 58.3 (12)
Distance-road (m) 286 (20) 289 (24) 244.1 (3 5)
Road length (m) 19 (6) 1 12 (24) 4575.9 (373)
Distance-urban (m) 288 (23) 273 (23) 258.6 (39)

 

233

20.00:: 05 :8

05 £208.80: me 0:83: 0:0 .95 2303 00:82 .623 0?. 0.620.— .Aoozv 00388 :35 Sm 00:00:00 :otoEU 3:08:85 0.00:8? 0:205 230% a

 

 

o mofio wmd $.th SEEN hue: + $003-00:st + BBQ + .052:
m nmfio SN ovSmm SRO: .033 + Bazox. + 0:30
m 83 SN Elma $00.3 :88 + 203:. + 02:8
m oemd 3.0 2.6mm 3:300:83: + ESQ + 00:08
v 03.0 00.0 mmfiwm 203\ + :33:
0: _3 024 .02 082

 

.m>0>.8m 00:35:00 083-003 0:0 $808030 0:0 .80: 3:5:

.30: 20.3: 53> @0338 $3.5 00:0» at: .3 8:398 :0 00mm: ooom :5 moom 8883 :m SwwEom—z mo 335:0: .8304 :0 E0538
0:: 5 £09: 033:: 98 <>>Om :0 038 E 2: 8 «£08.: 005:: $0.33.. 860% 5529:: 50: m :8 8:03. 01¢. .3. 030:.

234

model with owner, forest, and distance to nearest road had a similar AICc value, but I did

not select it because distance to the nearest road did not contribute to the model (P >
0.05). The effect of ownership was not significant (P = 0.23) (Table 4.9), but the random
effect of forest was significant (P < 0.01) (Table 4.9). Amphibian species richness had a
positive relationship with forest (P < 0.01) (Table 4.10).

Herpetofaunal Species Richness

The mixed model for herpetofauna] species richness with the best fit (lowest AICc
value) included one fixed effect (owner) and one random effect (forest (m2) (Table 4.11).

The model with owner, forest, and distance to nearest road had a similar AICC value, but I

did not select it for the same reason as with amphibian species richness, because distance
to the nearest road did not contribute to the model (P > 0.05). The effect of ownership
was not significant (P = 0.11) (Table 4.12), but the random effect of forest was significant
(P < 0.01) (Table 4.12). Herpetofaunal species richness had a positive relationship with
forest (P < 0.01) (Table 4.10).

Herpetofaunal Species Diversity

The mixed mode] for herpetofaunal diversity with the best fit (lowest AICc value)
included one fixed effect (owner) and one random effect (forest (m2) (Table 4.13). There

were several other models with similar AICc values, but these models were not selected

because the other variables included in these models: distance to the nearest surface water
(P = 0.50) and total road length (P >0.05) did not contribute to the model. The effect of

ownership was not significant (P = 0.21) (Table 4.14), but the random effect of forest was

235

Table 4.9. Type 3 fixed effects and solution for random effects for amphibian species
richness mixed model at 100 m scale (a = 0.05). Data collected on SGWA and private
lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 1.45 0.23
Type of Variation t Value Pr>F
Forest 3.73 < 0.01

 

236

Table 4.10. Correlations between landscape variables and amphibian richness,

herpetofaunal species richness, and herpetofauna] species diversity at the 100 m scale
using Pearson’s Correlation Coefficients. Data collected on SGWA and private lands in
the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures
by drift fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-

time constrained surveys.

 

 

Variable Amphibian Herpetofaunal Herpetofaunal
Richness Richness Diversity
R2 (P-value) R2 (P-value) R2 (P-value)
Distance to surface water (m) 0.02 (0.88) -0.05 (0.65) 0.01 (0.92)
Mean ground water depth (m) -0.17 (0.14) -0.22 (0.06) -0.24 (0.04)
Distance to nearest road (m) 0.0] (0.94) 0.06 (0.60) 0.15 (0.21)
Total road length (m) 0.17 (0.14) 0.14 (0.25) 0.03 (0.83)
Distance to nearest urban (m) 0.09 (0.45) 0.03 (0.82) 0.08 (0.48)
Total agriculture area (m2) -0.22 (0.06) -0.11 (0.35) -0.15 (0.19)
Total forested area (m2) 0.42 (< 0.01) 0.38 (< 0.01) 0.35 (< 0.01)
Total urban area (ml) -0.03 (0.79) -0.05 (0.65) -0.03 (0.80)
Total open water area (m2) N/A N/A N/A N/A N/A N/A
Total wetlands (m2) -0.09 (0.45) -0.05 (0.66) -0.05 (0.67)
Total other (m2) N/A N/A N/A N/A N/A N/A

 

237

.2808 2:

Ba 06 820823 go 838:: 98 63v 5203 8:82 .023 02 3:22 501$ $388 :25 Ba cocooboo :otoEU 838885 Poo—EVE. own—2: 338m a

 

 

m 8; EN 2:3 Swag. 32 + ~85: 52:8
0 wfld we; 2.3m hoormoacfin + Summificamxé + 3.on + .5230
m ago E; 3.2% Snowbauimsu + asst: bake
m Emd 3:0 233 Egmczswfi + “853+ 88.8
v ommd 86 3.8». EEQ+ base
2 E 0:2 002 382

 

.9333 3.55:8 06:-83 28 "6309630 can £995 355%

.895 zflzm :23 33:8 993.3 85% are .3 $538 no woman woom 28 moom 5883 E gwfioaz MO 23595 .533 c.5558
2.: E mun—.2 033a Ea <>>Om co 2an E 2: 8 «£288 8me mmofiot 86on _ncséowofion “was m 8m $83. 02 A 3“ 2an

238

Table 4.12. Type 3 fixed effects and solution for random effects for herpetofaunal
species richness mixed model at 100 m scale (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 2.67 0.11
Type of Variation t Value Pr>F
Forest 3.73 < 0.01

 

239

£088 2:

com CC £80888 «0 398:: 98 65 Emma», 8:82 .623 02 03:28 .Aoozv 8388 :95 8.“ 388.50 cor—Btu song—:85 98.52 3205 £8an a

 

 

m memo cwd wmdw Sawhxm~c§M>a+ >83K+ $238
0 v5.0 nod 2.0m 5M5: has + .333 8683“. + ESQ? .6838
m Sad one :3 53$ ES + $3? $238
w Smd Omd whww $3.: 8655. + BEQP .5830
v mvwd ood wiww 3.me + .558
M :5 024 00?. 3on

 

.9333 8:85:00 Bureau 98 $53838 was .395 353%

£95 anti 53, @2950 992.8 85m $8 3 8593 no woman coom 28 moon 8883. E gwfioaz mo 335:3 533 c.6558
05 5 $52 Bu>tm Ea <3Om co 28m E o3 3 «20on 3me bacozw 860% “scammeofio; «won m com 888 02 .m _ .v 2an

240

Table 4.14. Type 3 fixed effects and solution for random effects for herpetofaunal
species diversity mixed model at 100 m scale (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 1.62 0.21
Type of Variation t Value Pr>F
Forest 3.31 < 0.01

 

241

significant (P < 0.01) (Table 4.14). Herpetofaunal diversity had a positive relationship
with forest (P < 0.01) (Table 4.10).

Mixed Model — 200 m Scale

Amphibian Species Richness

The mixed model for amphibian richness within 200 m of the survey site with the

best fit (lowest AICc value) included one fixed effect (owner), and two random effects

(water (m2) and forest (m2)) (Table 4.15). The effect of ownership was not significant (P

= 0.82) (Table 4.16), but the random effects of water (P < 0.01) and forest (P = 0.03)
were significant (Table 4.16). Amphibian species richness had a negative relationship
with water (P < 0.01) and a positive relationship with forest (P < 0.01) (Table 4.17).
Herpetofaunal Species Richness

The mixed mode] for herpetofaunal species richness within 200 m of the survey

site with the best fit (lowest AICc value) included one fixed effect (owner), and two

random effects (water (m2) and wetland (m2)) (Table 4.18). The effect of ownership was
not significant (P = 0.12) (Table 4.19), but the random effects of water (P < 0.01) and
wetland (P < 0.01) were significant (Table 4.19). Herpetofaunal species richness had a
negative relationship with water (P < 0.01) and a positive relationship with wetland (P <
0.01) (Table 4.17).

Herpetofaunal Species Diversity

The mixed model for herpetofaunal species diversity within 200m of the survey site

with the best fit (lowest AlCC value) included one fixed effect (owner), and two

242

20008 08

.8.“ O: 008080.88 mo 00388 .80 63V 5%?» 00:82 .623 02 0.8200 .AQUZV 008800 :80 8.“ $00,000.80 :20in 8:05.85 0.0v=§< 09:05 0:300m a

 

 

0 «~80 00; 30mm 353.: + 2.03.: $3: + 00.38
0 $3 cm; 9.08 25302.00 + $3.: + Essa
m mid 3; mmémm 8520.: + .03.: + 0058
v oomd wwd ooémm .68.: + 8:30
0 ommd cod Edam 0003.: $3... + 00.38
m .3 0:3 002 322

 

0.880 008800800 083-0000 can 080000080 cam .098 #088

.098 €0.38 53» @0838 0.308 080m «30 ‘3 00.8800 so 0008. ooom 80 meow $8830 8 53:22 mo «80:83 0033 808300
05 8 00:0— 0:03.8 88 <>>Om co 0300 8 com 00 “20008 @058 00083.. 006080 8053880 0009 m How 00800 0?. .m _ .v 033.

243

Table 4.16. Type 3 fixed effects and solution for random effects for amphibian species
richness mixed model at 200 m scale (a = 0.05). Data collected on SGWA and private
lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drifi fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys.

 

 

 

 

Type of Variation P Value Pr>F
Owner 0.05 0.82
Type of Variation t Value Pr>F
Water -2.99 < 0.01
Forest 2.3 1 0.03

 

244

Table 4.17. Correlations between landscape variables and amphibian richness,

herpetofaunal species richness, and herpetofaunal species diversity at the 200 m scale
using Pearson’s Correlation Coefficients. Data collected on SGWA and private lands in
the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures
by drifi fence arrays coupled with pitfall traps, funnel traps, and coverboards and area-

time constrained surveys.

 

 

Variable Amphibian Herp Richness I-lerp Diversity
RiChDESS R2 (P-value) R2 (P-value)
R2 (P-value)
Distance to surface water (m) 0.07 (0.61) -0.01 (0.92) 0.08 (0.59)
Mean ground water depth (m) 0.04 (0.78) -0.01 (0.95) -0.07 (0.64)
Distance to nearest road (m) -0.07 (0.62) -0.01 (0.96) 0.23 (0.10)
Total road length (m) 0.10 (0.47) 0.08 (0.58) -0.17 (0.22)
Distance to nearest urban (m) 0.18 (0.20) 0.10 (0.45) 0.18 (0.20)
Total agriculture 31330112) -0.01 (0.96) <-0.01 (0.98) -0.09 (0.51)
Total forested area (m2) 0.37 (<0.01) 0.38 (<0.01) 0.37 (<0.01)
Total urban area (m2) 0.03 (0.85) 0.09 (0.50) -O.ll (0.44)
Total open water area (m2) -0.40 (<0.01) -0.44 (<0.01) —O.48 (<0.01)
Total wetlands (m2) 0.28 (0.04) 0.36 (<0.01) 0.33 (0.01)
Total other (m2) -0.07 (0.61) 0.04 (0.80) <0.0] (0.99)

 

24S

20008 05

no.“ 06 000008083 mo 0038:: 080 63v 3303 00:82 A023 0?. 006200 A003; 003800 :80 Ba 0080800 82002.5 803088.8— 0.0V=0x< 00208 0.0—30¢ a

 

 

K. 2:6 mm: 006mm “00.8? 0.5030.th + 08203 + .033 + .0238
m 0.1.0 om; ~5me N00xo\+ 008.: + 00830
0 mid m: 00.03 3,sz + 050203 + .03: + 0238
o bmmd mmd mnhmm 00530th + 38:03 + 008.: + 00:30
0 oomd cod mmdmm 02020.: + .03.: + 0020
m _3 03a .02 082

 

033:0 008080800 083-0000 080 08008030 080 .088 #083

.0808 :88 53> 008000 0380 00.0% «£0 E 008800 so 00009 003 080 moom 008830 8 803:022 Mo 0308:?“ 00304 805300

05 8 0083 0003.8 080 <>>Om so 2000 8 com 00 a20008 0058 0008520 006080 3800.08080: 0000 m 08 00800 02 .w _ .v 030%

246

Table 4.19. Type 3 fixed effects and solution for random effects for herpetofaunal
species richness mixed model at 200 m scale (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drift fence arrays coupled with pitfall traps, fimnel traps, and
coverboards and area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 2.53 0.12
Type of Variation t Value Pr>F
Water -3.85 < 0.01
Wetland 2.86 < 0.01

 

247

random effects (wetland (m2) and water (m2) (Table 4.20). The model with owner,

water, wetland, and average ground water depth had a similar AICc value, but I did not

select it because average ground water depth did not contribute to the model (P > 0.05).
The effect of ownership was not significant (P = 0.16) (Table 4.21), but the random
effects of water (P < 0.01) and wetland (P 0.02) were significant (Table 4.21).
Herpetofaunal species diversity had a negative relationship with water (P < 0.01) and a
positive relationship with wetland (P = 0.01) (Table 4.17).
Mixed Model — 1000 m Scale

Due to the high correlation between forest and agriculture (R2 = 0.84, P = <0.0l)
in the 1000 m data set, it was difficult to choose between them in defining the best model.
Therefore, I ran two separate analyses: one using agriculture in the model and the other
using forest in the model. I selected the best five models fiom the two different sets of
analyses for each of my three dependent variables: amphibian species richness,
herpetofaunal species richness, and herpetofaunal species diversity.
Amphibian Species Richness

The mixed model for amphibian species richness within 1000 m of the survey site

with the best fit (lowest'AICC value) included one fixed effect (owner), and two random
effects (agriculture (m2) and water (m2)) (Table 4.22). There were several competing

models with similar AICC value, and the model with forest and water could most likely be

used interchangeably. The effect of ownership was not significant (P = 0.77) (Table

4.23), but the random effects of agriculture (P < 0.01) and water (P 0.02) were

248

0.002: 0:: :8

0: 00808000: 00 50:3: 0:0 A05 Ems? 8:002 A023 0?. 03003: 60?; 003800 :20 :8 00:00:00 :otato 0000808:— 0.0x_0x< 00205 0:000m 0

 

 

0 mm mo 0N; owfim SM:& .003 + 050203 + .0009: + 00530
0 02.0 «N: 00.00 000kox+ 000002 + 0000.: + .0238
0 300 00.0 00.00 0000Q+ 0000.: + 00020
0 2N6 vmd 36m fin§§3§®§ + .3820? + 003: + 00:30
0 mwmd ocd 00.00 02000;» + 00003 + 00030
0 _3 020 .02 .0032

 

000350 0050:0900 080-0000 0:0 0000000030 0:0 .30.: _0::£

.000: =0.ta 0:3 003:8 000000 00:00 £00 >0 0003000 :o 0000: 000w 0:0 meow 008850 E :0w3032 mo 00:05:90 0033 E0538
0.0 E 00:0— 000>wa 0:0 <>>Om :o 2000 8 com 00 00300:: 0002:: 300030 006000 0050009080: 0000 m :8 00:80 02 .om.v 030...

249

Table 4.21. Type 3 fixed effects and solution for random effects for herpetofaunal
species diversity mixed model at 200 m scale (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 2.05 0.16
Type of Variation t Value Pr>F
Water -4.15 < 0.01
Wetland 2.59 0.02

 

250

 

$0008 0::

:8 Co 8808803 :0 :0083: 0:0 6:3 Ems? 00:00? A023 0?. 03:20: A025 003800 =080 :0: 00:00:00 :otBtU 80088:: 0.00000? 0020:: 0:003: 0

 

 

n. momo nmo oo.wo~ =00§00=8§0 + 53$ :88: + 0083 + 000.:o\ + .6208
v momo nmo ow.mo: :00:0\+ 00:30
v 08.0 00.0 00.8: 0:539:00 + 002.30
m ammo moo om.moH :03: + ESQ? 50:30
0 Ed 00.0 00.00: 0000:: + 0:330:00 + 00030
0: :3 020 00?. 0002

 

.0>0>::0 0080:0800 080-080 0:0 00:08:38 0:0 .030: 385.:

.030: =0.t3 £05 003:8 0%080 00:0,: 5:0 :3 008300 :0 0000: 0oo~ 0:0 moom :088s0 8 58:32 «:0 00:08:03 5301: 805300
0:: 8 00:0— 0:033 0:0 <>>Om :o 0:000 8 ooo: :0 020008 00:28 00080:: 00:00:00 8050380 :00: m :0: 00:80 02 .NNé 030B

251

IT—¥—

 

Table 4.23. Type 3 fixed effects and solution for random effects for amphibian species
richness mixed model at 1000 m scale (a = 0.05). Data collected on SGWA and private
lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on
captures by drifi fence arrays coupled with pitfall traps, funnel traps, and coverboards and
area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 0.08 0.77
Type of Variation t Value Pr>F
Agriculture -3.36 < 0.01
Water -2.56 0.02

 

252

 

significant (Table 4.23). Amphibian richness had a negative relationship with both
agriculture (P <0.01) and water (P = 0.17) (Table 4.24).
Herpetofaunal Species Richness

The mixed model for herpetofaunal species richness within 1000 m of the survey

site with the best fit (lowest AICc value) included one fixed effect (owner), and three

random effects: forest (m2) and water (m2), and road length (m2) (Table 4.25). The effect

of ownership was not significant (P = 0.06) (Table 4.26), but the random effects of forest,
water, and road length were all significant (P < 0.01) (Table 4.26). Amphibian richness
had a positive relationship with forest (P = 0.01) and road length (P = 0.12), and a
negative relationship with water (P = 0.46) (Table 4.24).

Herpetofaunal Species Diversity

The mixed model for herpetofaunal species diversity within 1000 m of the survey

site with the best fit (lowest AICC value) included one fixed effect (owner), and one
random effect: agriculture (m2) (Table 4.27). There were several competing models with

similar AICc value, including a model with ownership and forest and a model with

ownership and road length. The effect of ownership was not significant (P = 0.06) (Table
4.28), but the random effect of agriculture was significant (P = 0.02) (Table 4.28).
Herpetofaunal diversity had a negative relationship with agriculture (P = 0.02) (Table
4.24).
Constrained Ordination

The five landscape variables selected accounted for 36.4 0/o of the total variation

in the herpetofaunal community data set (Table 4.29). The full model was significant

253

Table 4.24. Correlations between landscape variables and amphibian richness,

herpetofaunal species richness, and herpetofaunal species diversity at the 1000 m scale
using Pearson’s Correlation Coefficients. Data collected on SGWA and private lands in
the southern Lower Peninsula of Michigan in summer 2005 and 2006 based on captures
by drift fence arrays coupled with pitfall traps, fimnel traps, and coverboards and area-

time constrained surveys.

 

Variable

Amphibian
Richness
R2 (P-value)

Herp Richness
R2 (P-value)

Herp Diversity
R2 (P-value)

 

Distance to surface water (m)
Mean ground water depth (m)
Distance to nearest road (m)
Total road length (m)

Distance to nearest urban (m)

Total agriculture area (m2)
Total forested area (m2)
Total urban area (m2)
Total open water area (m2)
Total wetlands (m2)

Total other (m2)

0.18
0.07
-0.31
0.09
0.22

-O.52
0.51
-0.2l)
-O.29
0.49

0.03

(0.41)
(0.76)
(0.15)
(0.68)
(0.30)

(0.01)
(0.01)
(0.36)
(0.17)
(0.02)

(0.90)

0.13

0.08

-O.32

0.32

0.07

-0.54

0.54

0.03

-0.15

0.49

0.25

(0.55)
(0.71)
(0.12)
(0.12)
(0.73)

(0.01)
(0.01)
(0.89)
(0.46)
(0.01)

(0.24)

0.24

0.05

-O.ll

0.39

0.08

-0.47

0.45

0.06

-0.03

0.43

0.24

(0.24)
(0.80)
(0.62)
(0.06)
(0.72)

(0.02)
(0.03)
(0.77)
(0.90)
(0.03)

(0.25)

 

254

80008 08 :8

Co 0.8008800 .80 .8080: 0:0 630 80803 088?. A023 02 0800—0: A0020 008800 =080 :8 00:08:00 08:08.5 8:08:88 0.000002 000.08 000008 0

 

 

0 bmoo oow :0: 000200 000.: + 00000: + 000000.200 + 00030
0 :00 00.0 00.0: 000000200 + 000.00
v Nmoo mmd ofim: 00000\+ 00:30
N. mm _.o No.m 00.02 2000000830 + 000000 0000 + 00003 + 000:0\+ 00:30
0 20.0 00.0 00.00: 000000 0000 + 00000 + 000% + 00000
0: _a 020 00?. 0002

 

00000.80 00808080 080-080 0:0 00.008950 0:0 .0008 0:88

.0008 :88: :83 003000 000::0 00:8 8:0 3 00.8800 :0 00000 woom 0:0 moom :08800 8 :00 882 .80 00:08:00 :0304 808000
08 8 00:0— 0:03.8 0:0 <>>Om :0 2000 8 83 :0 080008 0088 00080:: 0080000 380000080: 0000 m :8 00:000 0000 .30 050,—.

255

Table 4.26. Type 3 fixed effects and solution for random effects for herpetofaunal
species richnws mixed model at 1000 m scale (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drift fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

 

 

Type of Variation P Value Pr>F
Owner 41.5 0.06
Type of Variation t Value Pr>F
Forest 4.55 < 0.01

Road length 4.70 < 0.01
Water -4.66 < 0.01

 

256

20008 05

:8 Co 80080.00 00 :0980: 0:0 A03 8003 00000000 A023 02 08:20: A0083 003800 :85 :8 00:00:00 :8:08:0 8:08:88 0.080000. 00808 000008 0

 

 

0 0.8.0 00.: 00.00 000000 00000 + 000000 + 000000
0 00:0 8.: 00.00 00000.0 + 00000
0 000.0 00.0 00.00 00000: 0000 + 00.00
.0 000.0 00.0 00.00 00000x+ 000000
0 000.0 00.0 00.00 000000.000 + 000.00
M :3 0830 0030 00002

 

00380 00808800 080080 0:0 000080060 0:0 0000:: 6:88

.0008 :88: 805 008000 000::0 00:8 8:0 03 008800 :0 00000 ooom 0:0 moom :08800 8 :00882 00 0:08:00 :033 E08000
08 8 00:8 008:: 0:0 «.300 :0 2000 8 ooo: :0 020008 0088 080880 0080000 3880:0800 0000 m :8 00:000 030 .000 030,—.

257

Table 4.28. Type 3 fixed effects and solution for random effects for herpetofaunal
species richnass mixed model at 1000 m scale (a = 0.05). Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006
based on captures by drifi fence arrays coupled with pitfall traps, funnel traps, and
coverboards and area-time constrained surveys.

 

 

 

 

Type of Variation F Value Pr>F
Owner 0.27 0.61
Type of Variation t Value Pr>F
Agriculture -2.38 0.03

 

258

Table 4.29. Summary of the results of the constrained ordination for herpetofaunal
community data explained by the landscape variables. Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006.

 

 

Inertia Proportion of Variance
Total 0.396 1 .000
Constrained 0.144 0364*
Unconstrained 0.252 0.636

 

*Proportion of variance explained is calculated by dividing the sum of canonical eigenvalues by the total
inertia (sum of unconstrained eigenvalues).

259

(P < 0.005) according to the Monte Carlo permutation test, which indicates that the
proportion of species variance explained by the environmental constraints was greater
than expected by chance. Monte Carlo permutation tests also showed that the first
canonical axis (RDA1) was significant (P = 0.01) (Table 4.30). Three of the variables
(urbanlOO, forestZOO, and wetlanleOO) were significant (P < 0 .05), but land ownership
and wetland200 were not significant (P > 0.05) (Table 4.31).

The first axis explained 26% of the total species variance (eigenvalue/total
inertia*100). This axis was a gradient of increasing total area of forest land cover at a
200 m scale, increasing total area of wetland land cover at a 1000 m scale, and increasing
total area of urban land cover at a 100 m scale. The second axis explained 5% of the total
species variance. This axis was a gradient of increasing total area of wetland land cover
at a 1000 m scale.

In the biplot, perpendicular projections of species points on the environmental
gradients explain the associations of individual species. For example, eastern American
toads were associated with landscapes that contained increasing amounts of urban areas
at a 100 m scale and less total forest area at a 200 m Scale (Figures 4.2 and 4.3); spring
peepers were associated with landscapes that were comprised of increased wetland land
class at a 1000 m scale (Figures 4.2 and 4.4); and wood frogs were associated with
landscapes that contained greater amounts of forest and wetlands at a scale of 200 m and
less urban land class at a scale of 100 m (Figure 4.2 and 4.5); green fi'ogs were associated
with landscapes that contained greater amounts of forest and wetlands at a scale of 200 m
(Fig. 4.2).

In the biplot, it is more appropriate to represent categorical variables such as land

260

Table 4.30. Summary of Monte Carlo permutation tests for landscape data showing
significance of each axis in the constrained ordination (o. = 0.05). Data collected on
SGWA and private lands in the southern Lower Peninsula of Michigan in summer 2005
and 2006.

 

 

Axes Df Variance F # Permutations Pr(>F)
RDA1 1 0.104 4.95 200 < 0.01
RDA2 1 0.020 0.94 100 0.80
RDA3 1 0.01 l 0.52 100 0.94
RDA4 1 0.008 0.38 100 0.90
RDAS 1 0.002 0.08 100 1.00

Residual l 2 0.252

 

261

Table 4.3]. Summary of Monte Carlo permutation tests for landscape scale data showing
significance of variables in explanatory data set (a = 0.05). Forest200 represented by the
total area of forest (m2) in the 200 m AD; urban100 represented the total area of urban
(m2) in the 100 m AD; wetland1000 represented the total area of wetland (m2) in the
1000 m AD; ownership represented the land ownership where the survey site was located
(SGWA vs. private); wetland200 represented the total area of wetland (m2) in the 200 m
AD. Data collected on SGWA and private lands in the southern Lower Peninsula of
Michigan in summer 2005 and 2006.

 

 

Variable Df Variance F # Permutations Pr(>F)
Forest200 1 0.042 2.02 100 < 0.01
Urban100 1 0.034 1 .60 100 0.01
Wetland 1 000 1 0.041 1.93 100 0.04
Ownership 1 0.013 0.62 100 0.41
Wetland200 1 0.014 0.68 100 0.44
Residual 12 0.252

 

262

 

0.8
J

wetland1000

0.6

/

 

   
 

 

 

 

urban100
AMLA
N -
2’ ° \
0 HYVE THSA
(I forest200 \
O
- —+ --------------------- _ -------------------------------------------- BUAM
0 RACL wetlandn
THSI
N_ PLCI .
OI RAPI
V
c? ‘RASY
ownershipPrivate
l i I
-0.5 0.0 0.5

RDA1

Fig. 4.2. Redundancy Analysis ordination diagram (biplot). Species are represented by
codes (see Table 4.4), the proximity of species in ordination space indicates occurrence in
similar environmental conditions. Environmental variables are represented by vectors,
which point toward rate of maximum change and extend in both directions. The length of
the vector indicates its importance to the constrained ordination (ter Braak 1986).
Perpendiculars drawn from species to vectors give the approximate ranking of that
species response to the environmental variable and indicate the species optimum on that
variable (ter Braak 1986). A smaller angle between the vector and the ordination axis
indicates a greater relationship of the variable to the derived constrained ordination
gradient (Grand and Mello 2004). Forest200 represented by the total area of forest (m2)
ata200mscale;mban100representedthetota1areaofm'ban(m2)ata100mscale;
wetlarrleOO represented the total area of wetland (m2) at a 1000 m scale; ownership
represented the land ownership where the survey site was located (SGWA vs. private);

wetland200 represented the total area of wetland (m2) at a 200 m scale.

263

 

15

1D

RDA2
05

OD

 

05

 

 

 

 

 

-LO

-1.0 —0.5 0.0 0.5 1.0 1.5
RDA1
Fig. 4.3. Plot of eastern American toad abundance in a Generalized Additive Model
(GAM) surface for urban land class (m2) at a scale of 100 m. Data collected on SGWA

and private lands in the southern Lower Peninsula of Michigan in summer 2005 and
2006. Increased size of circles indicates increase in abundance.

264

 

 

 

 

 

 

 

l0. _
l
o, _ /
to, _ //
O
N
<
o
n: /
Q
o /
l0
Cé ‘ @539
o- _
‘7 l I r I I
-1.0 -O.5 0.0 0.5 1.0 1.5
RDA1

Fig. 4.4. Plot of spring peeper abundance in a Generalized Additive Model (GAM)
surface for wetland land class (m2) at a scale of 1000 m. Data collected on SGWA and
private lands in the southern Lower Peninsula of Michigan in summer 2005 and 2006.
Increased size of circles indicates increase in abundance.

265

 

1.5

1.0

RDA2

 

0.0

 

 

 

 

 

-1.0

-1.0 -0.5 0.0 0.5 1.0 1.5
RDA1
Fig. 4.5. Plot of wood frog abundance in a Generalized Additive Model (GAM) surface
for forest land class (m2) at a scale of 200m. Data collected on SGWA and private lands

in the southern Lower Peninsula of Michigan in summer 2005 and 2006. Increased size
of circles indicates increase in abundance.

266

ownership type (SGWA and private) as points in the data space because they do not have
a gradient of change. Classes containing sites with high values for a particular species
will tend to lie in close proximity to that species in the data space. For example, eastern
American toads were associated with patches found on private lands and spring peepers
were associated with patches found on SGWA (Figure 4.2).

The variance explained by the fill model, which consisted of four sets of
environmental factors combined including urban land class at 100m, land classes at 200
m, and wetland land class at 1000 m, and land ownership was decomposed without
including land ownership. Land ownership was represented as a categorical variable and
was assessed as a site specific variable, as opposed to across the landscape, so I chose to
exclude it for the variance decomposition. Land classes at a 200 m scale (forest and
wetland) accounted for more variance (12.0%) than wetland land class at a 100 m scale

(5.5%), and urban land class at a 100 m scale accounted for considerably less (3.9%)

(Figure 4.6).

DISCUSSION
Results indicated that amphibian species richness, herpetofaunal species richness,
herpetofaunal species diversity, and herpetofaunal community assemblages exhibited
significant relationships with landscape attributes at differing spatial scales, particularly
with land cover classes, regardless of analyses (univariate and multivariate analyses).
Species richness and diversity were related to a combination of factors that depended on
the spatial scale studied. At a 100 m scale, forest cover was the most important variable

for amphibian and herpetofaunal richness, and species diversity. At a 200 m scale, open

267

 

urban100 wetland1000
3.9 A 5.5
12

scale200m

 

 

 

 

Fig. 4.6. Percent of total variance in the herpetofaunal community data set
explained by urban land class within 100m of survey site, wetland land class
within 1000 m of the survey site, and scale200 (comprised of forest and wetland
land classes within 200 m of survey sites), as well as variance explained by the
combination of factors. Data collected in the southern Lower Peninsula of
Michigan in summer 2005 and 2006 based on captures by drift fence arrays
coupled with pitfall traps, funnel traps, and coverboards and area-time
constrained surveys.

268

water was associated with richness and diversity, however, forest cover also exhibited a
significant relationship with amphibian richness and wetland exhibited a significant
relationship with herpetofaunal species richness and diversity. At a 1000 m scale,
amphibian species richness was associated with agriculture and open water;
herpetofaunal species richness was associated with forest cover, open water, and total
road length; and herpetofaunal species diversity was associated with agriculture.
However, when interpreting individual species requirements and factors that can
influence community composition, a combination of landscape attributes at differing
spatial scales (urban at a 100 m scale, forest at a 200 m scale, and wetland at a 1000 m
scale), were the most important variables. Land ownership did not play a significant role
in my findings, which suggests that at least in my study areas herpetofauna benefit from a
mosaic of land use classes and that both land ownerships contain these desired land use
classes.

I expected to find a positive association between forest cover and species richness
and species diversity. Amphibians breed in wetlands and emigrate to surrounding
terrestrial habitats to forage and overwinter during the non-breeding season (Richter et al
2001). Forests provide essential habitat for these amphibians that spend all or a portion
of their non-breeding seasons in trees, shrubs, or heavy leaf litter (Knutson et al. 1999).
The importance of forest cover in structuring amphibian richness and abundance has been
widely documented (e.g., Bonin et al. 1997, Hecnar I997, Hecnar and M’Closkey 1998).
Reptiles differ from amphibians in that they emigrate to terrestrial areas to nest or
overwinter, but they tend to live and forage in aquatic habitats (Semlitsch et al. 1988).

Houlahan and Findlay (2003) found higher herpetofaunal species richness with high

269

forest cover within 2 km of wetlands. Maintaining connectivity between wetlands and
terrestrial habitat is essential for herpetofauna species that use multiple, distinct habitats
during their life cycle. Guerry and Hunter (2002) found species richness increased when
the amount of upland forest cover adjacent to wetlands increased.

Forests may be important to herpetofaunal communities because they provide
relatively undisturbed habitat in the landscape, as compared to land uses like agriculture
and urban development that are frequently modified or disturbed (Knutson et al. 1999).
Forests moderate temperature and evaporation rates of adjacent aquatic habitats,
influence pH levels in soil, and contribute to increased leaf litter and coarse woody debris
(W aldwick 1997). Rubbo and Kiesecker (2005) found that anurans associated with forest
cover when it was relatively common in the landscape (in Wisconsin) and when it was
rare (in Iowa).

I found a consistent positive relationship between amphibian species richness,
herpetofaunal species richness, and herpetofaunal species diversity and forest cover. At a
100 m scale, it was the most important factor in determining richness and diversity. It
was important to amphibian species richness at a 200 m scale and herpetofaunal species
richness at a 1000 m scale.

I expected to find a negative association between open water and species richness
and species diversity. Open water had a negative effect on amphibian species richness,
herpetofaunal species diversity, and herpetofaunal species diversity at a 200 m scale, as
well as to amphibian species richness and herpetofaunal species diversity at a 1000 m
scale. Open water is indicative of permanent water, suggesting that the hydroperiod (the

amount of time it constantly holds water) is long and does not result in drying of the

270

water body. Fish predation most likely plays a valuable role in structuring amphibian
distribution and amphibian community composition (Heyer et al. 1975). Previous studies
have also found a negative relationship between open water bodies and species richness
and diversity. Predatory fish occur in permanent water bodies, and thus there is a higher
risk of predation by fish on amphibians in permanent bodies of water (Hecnar 1997).
Previous studies have also found a negative relationship between open water bodies and
species richness and diversity. Hecnar (1997) observed significant decreases (P < 0.01)
in amphibian species richness at ponds that had predatory fish versus ponds that either
did not have any fish or did not have any predatory fish. Hecnar and M’Closkey (1998)
found species richness to be negatively correlated with water depth (indicating pond
permanence) and fish predators.

Open water is also negatively associated with amphibian larvae. Larval
amphibians are vulnerable to several predators including vertebrate and invertebrate
predators (Alford 1999). Ponds with a hydroperiod greater than 2 years are considered
detrimental to amphibian species because they have a wide array of predators including
invertebrates and fish (Semlitsch 2002). Certain species have adapted to more permanent
waters and they can coexist with fish because tadpoles possess antipredator behavior and
skin toxins (Semlitsch 2002). Species in my study area that have evolved antipredator
behavior include: green frog and bullfrog (Semlitsch 2002). Studies have also shown that
adult fi'ogs exhibit avoidance behavior to waters with predatory fish (e. g., Kats and Sih
1992, Hopey and Petranka 1994).

Larval amphibians that have evolved a range of antipredator defense mechanisms

have done so with native species that coexisted in ponds. The widespread introduction of

271

predatory fish and bullfrogs has exposed native amphibians to an entirely new array of
predators with which they most likely had no prior interactions. For example, the yellow-
legged frog (Rana muscosa) in the Sierra Mountains of California and Nevada were
negatively affected by exposure to non-native predators. Yellow-legged frogs were
exposed to non-native trout that had been stocked in ponds since 1800’s. By 1910,
populations had been decimated by trout predation (Collins and Storfer 2003). In
addition to site-specific population declines, the effects of introduced predatory fish can
reduce recolonization rates of amphibians and alter metapopulation structure.

I expected to find a positive association between wetlands and amphibian species
richness, herpetofaunal species richness, and herpetofaunal species diversity. Wetlands
have been positively associated with amphibian species richness (Knutson et al. 1999). I
observed a positive relationship between wetlands and herpetofaunal richness and
diversity at a 200 m scale. Amphibians use wetlands as breeding habitats (Richter et al.
2001) and for larval development, and some reptiles ofien live and forage in wetlands
during the majority of the year (Semlitsch et al. 1998). Herpetofauna are influenced by
wetland hydroperiod, and it is necessary to maintain wetlands with diverse hydroperiods
(Pechmann et al. 1989). The positive relationship between wetlands and species richness
and diversity reflects the requirement for this habitat type during certain phases of their
life cycle, but also demonstrates the importance of maintaining connectivity among
various wetland habitats to support a metapopulation structure (Semlitsch 2002).

I expected to find a negative relationship between agriculture and amphibian and
herpetofaunal species richness, and herpetofaunal species diversity. Amphibian diversity

has been shown to decrease in association with intensive agricultural landscapes (Bonin

272

et al. 1997). I observed a negative association between agriculture and amphibian species
richness and herpetofaunal species diversity at a 1000 m scale. Farming practices modify
the landscape resulting in the loss of forest habitat, the draining, filling, and degradation
of wetlands, the conversion of natural habitats to managed monocultures, and soil
compaction (Bonin et al. 1997). In addition, the use of herbicides and pesticides in
support of agricultural practices can potentially affect herpetofauna by degrading habitats
and reducing plant and invertebrate diversity (Sotherton et al. 1988). Runoff from
agricultural lands has the potential to impact aquatic habitats by modified nutrient
loadings, sediment accretion, increased pollutants, and altered water temperatures
(Saunders et al. 2002). Aquatic organisms’ mortality rates, reproductive success, and
behavior and growth can be negatively affected by these alterations to aquatic systems
(Abramovitz 1996). Pesticides can also negatively impact herpetofaunal species through
direct toxicity resulting in sublethal effects on behavior (Bridges 1997) and development
(Bridges 2000, Metts et al. 2005) and negative impacts can be intensified by interactions
between sublethal effects and environmental factors (Metts et al. 2005).

Fewer species tend to be found in urban areas (Rubbo and Kiesecker 2005).
Urbanization negatively affects herpetofauna through changes in land use including the
loss of natural habitats and increase in development, fragmentation of previously
continuous populations, and the increased exposure to pollutants and contaminants
(Genet 2004). Knutson et al. (1999) found a negative association between urban lands
and species richness in Iowa and Wisconsin. However, the total area of urban land class
was not a significant factor in characterizing species richness and diversity in my 3

spatial scale analyses. I did however observe its influence on specific species in my

273

constrained ordination. The majority of species had a negative association with urban at
a 100 m scale, however, eastern American toad had positive association with urban at a
100 m scale. Eastern American toads are commonly found in fragmented and
anthropogenically modified landscapes (Kolozsvary and Swihart 1999). Guerry and
Hunter (2002) showed a negative association between eastern American toads and forest
cover; they prefer warmer, shallow waters for breeding, which may occur more often in
human-modified landscapes (Houlahan and Findlay 2003). Additionally, they are a
habitat generalist (Hecnar and M’Closkey 1998) which most likely is why they are not
affected by changes in the landscape due to urban development (Rubbo and Kiesecker
2005)

Although it is clear that land use affects species richness and diversity, my results
suggest the importance of species specific conservation strategies. When investigating
factors that structure herpetofaunal community composition, individual requirements
must be taken into consideration. Although the best model for the herpetofaunal
community data set only explained 36.4% of the variation, I conclude that when
managing for particular species these factors need to be considered at different spatial
scales.

Land cover classes have a significant impact on amphibian species richness,
herpetofaunal species richness, herpetofaunal species diversity, and herpetofaunal
community composition. Herpetofauna have different life history requirements and these
requirements must be taken into consideration when managing for specific herpetofaunal
species. The associations I found between richness, diversity, and community

composition at the landscape level can provide the MDNR with valuable information for

274

management and conservation strategies to ensure that the integrity of the herpetofaunal

community will persist throughout the ecoregion.

275

LITERATURE CITED

Abramovitz, J. N. 1996. Irnperiled waters, impoverished firture: the decline of
freshwater ecosystems. Paper 128. Worldwatch Institute, Washington, D.C.,
USA.

Akaike, H. 1987. Factor analysis and AIC. Psychometrika 52(3):317-332.

Alford, RA. 1999. Ecology: resource use, competition and predation. Pages 240-278 in
R. W. McDiarmid, R. Altig, editors. Tadpoles: the biology of anuran larvae.
University of Chicago Press, Chicago, Illinois, USA.

Alford, R. A., and S. J. Richards. 1999. Global amphibian declines: a problem in applied
ecology. Annual Review of Ecology and Systematics 1999 (30):]33-165.

Berven, K. A., and T. A. Grudzien. 1990. Dispersal in the wood fi'og (Rana sylvatica):
implications for genetic population structure. Evolution 44:2048-2056.

Beyer, H. L. 2004. Hawth’s Extension Tool for ArcGIS. <http://spatialecology.com
/htools>.

Bonin, J ., J. DesGranges, J. Rodrique, and M. Ouellet. 1997. Anuran species richness in
agricultural landscapes of Quebec: foreseeing long-term results of road call
surveys. . Pages 141-48 in D. M. Green, editor. Amphibians in decline:
Canadian studies of a global problem. Society for the Study of Amphibians and
Reptiles, St. Louis, Missouri, USA.

Bridges, C. M. 1997. Tadpole swimming performance and activity affected by acute
exposure to sublethal levels of carbaryl. Environmental Toxicology and
Chemistry 16:1935—1939.

Bridges, C. M. 2000. Long-term effects of pesticide exposure at various life stages of
the southern leopard frog (Rana sphenocephala). Archives of Environmental
Contamination and Toxicology 39:91—96.

276

Brinson, M. M. and A. I. Malvarez. 2002. Temperate freshwater wetlands: types, status,
and threats. Environmental Conservation 29(2)]15-133.

Brown, W. S., and W. S. Parker. 1976. A ventral scale clipping system for permanently
marking snakes (Reptilia, Serpentes). Journal of Herpetology 10:247—249.

Bumham, K. P., and D. R. Anderson. 2002. Model selection and inference: a practical
information-theoretic approach. Second edition. Springer-Verlag, New York,
New York, USA.

Cagle, F. R. 1939. A system of marking turtles for future identification. Copeia.
1939:170-172.

Calhoun, A. J. K., N. A. Miller, and M. W. Klemens. 2005. Conserving pool-breeding
amphibians in human-dominated landscapes through local implementation of Best
Development Practices. Wetlands Ecology and Management 13:291-304.

Campbell, H. W., and S. P. Christman. 1982. The herpetological components of Florida
sandhill and sand pine scrub Associations. Pages 163-171 in N. J. Scott, editor.
Herpetological Communities. Fish and Wildlife Service Wildlife Resources
Report 13. Washington, D.C., USA.

Carr, L.W., S.E. Pope, and L. F ahrig. 2000. Impacts of landscape transformation by
roads. Pages 225-243 in K. J. Gutzwiller, editor. Concepts and applications of

landscape ecology in biological conservation. Springer-Verlag, New York, New
York, USA.

Chatterjee, S., A. S. Hadi, and B. Price. 2000. Regression analysis by example. Third
edition. John Wiley & Sons, Inc, New York, New York, USA.

Collins, J ., and A. Storfer. 2003. Global amphibian declines: sorting the hypotheses.
Diversity and Distributions 2003 (9)289-98.

Corn, P. S. 1994. Straight-line drift fences and pitfall traps. Pages 109-1 17 in W.R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington D.C., USA.

277

Crump, M. L., and N. J. Scott. 1994. Visual encounter surveys. Pages 84-92 in W.R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.

Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington D.C., USA.

Davis, T. M. 1997. Nondisruptive monitoring of terrestrial salamanders with artificial

cover objects on southern Vancouver Island, British Columbia. Herpetological
Conservation 1 :161-1 74.

Dunning, J. B., B. J. Danielson, and H. R. Pulliam. 1992. Ecological processes that affect
populations in complex landscapes. Oikos 65:169-175.

Enge, K. M. 1997. A standardized protocol for drift-fence surveys. Florida Game and
Fresh Water Fish Commission. Technical Report No.14. Tallahassee, Florida,
USA.

Environmental Systems Research Institute. 2004. ArcMap 9.1. Redlands, California,
USA.

Fellers, G. M., and C. A. Drost. 1994. Sampling with artificial cover. Pages 146-150 in
W. R. Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster,

editors. Measuring and monitoring biological diversity: standard methods for
amphibians. Smithsonian Institution Press, Washington D.C., USA.

Findlay, C. S., and J. Houlahan. 1997. Anthropogenic correlates of species richness in
southeastern Ontario wetlands. Conservation Biology 11(4):1000-1009.

Genet, KS. 2004. Status and distribution of frogs and toads in southern Michigan:
population trends and the influence of habitat and landscape characteristics.
Dissertation, Michigan State University, East Lansing, Michigan, USA.

Gibbs, J. P. 2000. Wetland loss and biodiversity conservation. Conservation Biology
l4(l):3l4-3l7.

Grand, J ., and M. Mello. 2004. A multi-scale analysis of species-environment

relationships: rare moths in a pitch pine-scrub oak (Pinus rigida-Quercus
ilicifolia) community. Biological Conservation 119:495-506.

278

Greenberg, C. H., D. G. Neary and L. D. Harris. 1994. A comparison of herpetofaunal
sampling effectiveness of pitfall, single-ended, and double-ended funnel traps
used with drifi fences. Journal of Herpetology 28(3):319-324.

Greenberg, C. H., and G. W. Tanner. 2005. Spatial and temporal ecology of eastern
spadefoot toads on a Florida landscape. Herpetologica 61(1):20—28.

Guerry, A. D., and M. L. Hunter, Jr. 2002. Amphibian distributions in a landscape of
forests and agriculture: an examination of landscape composition and
configuration. Conservation Biology 16(3):745-754.

Hecnar, S. J. 1997. Amphibian pond communities in southwestern Ontario. Pages 1-15
in D. M. Green, editor. Amphibians in decline: Canadian studies of a global
problem. Society for the Study of Amphibians and Reptiles, St. Louis, Missouri,
USA.

Hecnar, S. J., and R. T. M’Closkey. 1996. The effects of predatory fish on amphibian
species richness and distribution. Biological Conservation. 79:123-131.

Hecnar, S. J., and R. T. M’Closkey. 1998. Species richness patterns of amphibians in
southwestern Ontario ponds. Journal of Biogeography 25:7 63-7 72.

Hermann, H. L., K. J. Babbitt, M. M. Baber, and R. G. Congalton. 2005. Effects of
landscape characteristics on amphibian distribution in a forest-dominated
landscape. Biological Conservation 123:] 39-149.

Heyer, W. R., R. W. McDiarmid, D. L. Weigrnann. 1975. Tadpoles, predation, and pond
habitats in the tropics. Biotropica 7:100-100.

Hill, M. O., and H. G. Gauch, Jr. 1980. Detrended Correspondence Analysis: an
improved ordination technique. Vegetatio 42:47-58.

Hopey, M. E., and J. W. Petranka. 1994. Restriction of wood fiogs to fish free habitats:
how important is adult choice? Copeia l994(4):1023:1025.

279

Houlahan, J. E., and C. S. Findlay. 2003. The effects of adjacent land use on wetland
amphibian species richness and community composition. Canadian Journal of
Fisheries and Aquatic Science 60:1078-1094.

Houze, C. M., and C. R. Chandler. 2002. Evaluation of coverboards for sampling
terrestrial salamanders in South Georgia. Journal of Herpetology 36:75-81.

Jackson, S. D. 2000. Overview of transportation impacts on wildlife movements and
populations. Pages 7-20 in T. A. Messamer, and B. West, editors. Wildlife and
highways: seeking solutions to an ecological and socio-economic dilemma. The
Wildlife Society, Nashville, Tennessee, USA.

Kats, L. B., and A. Sih., 1992. Oviposition site selection and avoidance of streamside
salamanders (Ambystoma barbouri). Copeia 1992(2):468-473

Kiesecker, J. M., A. R. Blaustein, and L. K. Belden. 2001. Complex causes of
amphibian population declines. Nature 410:681-684.

Knut, R., R. H. Okland, and T. Okland. 2003. Species response curves along
enviromnental gradients: a case study from SE Norwegian swamp forests.
Journal of Vegetation Science 14:869-880.

Knutson, M. G., J. R. Sauer, D. A. Olsen, M. J. Mossman, L. M. Hemesath, and M. J.
Lannoo. 1999. Effects of landscape composition and wetland fragmentation on

frog and toad abundance and species richness in Iowa and Wisconsin, USA.
Conservation Biology 1 3(6): 143 7-1446.

Kolozsvary, M. B., and R. K. Swihart. I999. Habitat fragmentation and the distribution
of amphibians: patch and landscape correlates in farmland. Canadian Journal of
Zoology 77:1288-1299.

Lannoo, M. L. 1998. Status and conservation of midwestem amphibians. University of
Iowa Press, Iowa City, Iowa, USA.

Lannoo, M. L., editor. 2005. Amphibian declines: the conservation status of United
States species. University of California Press, London, England.

280

 

Legendre, P. and E. D. Gallagher. 2001. Ecologically meaningful transformations for
ordination of species data. Oecologica 129:271-280.

Lehtinen, R. M., S. M. Galatowitsch, and J. R. Tester. 1999. Consequences of habitat
loss and fiagmentation for wetland amphibian assemblages. Wetlands l9(1):l-12.

Lehtinen, R. M., and S. M. Galatowitsch. 2001. Colonization of restored wetlands by
amphibians in Minnesota. American Midland Naturalist 145(2):388-396.

Mallows, C. L. 1973. Some comments on Cp. Technometrics 15:661-675.

McDiarmid, R. W. 1994. Amphibian diversity and natural history. Pages 5-15 in W. R.
Heyer, M. A. Donnelly, R. W. McDiarmid, L. C. Hayek and M. S. Foster, editors.
Measuring and monitoring biological diversity: standard methods for amphibians.
Smithsonian Institution Press, Washington D.C., USA.

Metts, B. S., W. A. Hopkins, and J. P. Nestor. 2005. Interaction of an insecticide with
larval density in pond breeding salamanders (Ambystoma). Freshwater Biology
50:685-696.

Michigan Center for Geographic lnforrnation. 2007. Michigan Geographic Data Library.
<http://www.mcgi. state.mi.us/mgdl/>. Accessed 10 Jan 2007.

Michigan Department of Natural Resources. 2000. Michigan GAP state land ownership
(stewardship) digital geospatial data. Department of Natural Resources, Land and
Minerals Services Office, Lansing, Michigan, USA.

Michigan Department of Natural Resources. 2001. IF MAP Southern Michigan land
cover digital geospatial data. Department of Natural Resources, Forest, Mineral,
and Fire Management Division, Lansing, Michigan, USA.

Morin, P. J. 1983. Predation, competition, and the composition of larval anuran guilds.
Ecological Monographs 53:] 19-138.

281

Oksanen, J. 2007. Multivariate analysis of ecological communities in R: a Vegan
tutorial. <http://cc.oulu.fi/~jarioksa/opetus/metodi/vegantutor.pd£>. Date
accessed: 5 Feb 2007.

Ott, R. L., and M. Longnecker. 2001. An introduction to statistical methods and data
analysis. Fifth edition. Wadsworth Group, Pacific Grove, California, USA.

Parris, K. M. 2006. Urban amphibian assemblages as metacommunities. Journal of
Animal Ecology 75:757-764.

Pechmann, J. H., K., D. E. Scott, J. W. Gibbons, and R. D. Semlitsch. 1989. Influence of
wetland hydroperiod on diversity and abundance of metamorphosing juvenile
amphibians. Wetlands Ecology and Management 1:3-11.

Price, S. J ., D. R. Marks, R. W. Howe, J. M. Hanowski, and G. J. Niemi. 2004. The
importance of spatial scale for conservation and assessment of anuran populations

in coastal wetlands of the western Great Lakes, USA. Landscape Ecology
20:441—454.

Primack, R. B. 2004. A primer of conservation biology. Third edition. Sinauer
Associates, Inc., Sunderland, Massachusetts, USA.

R Development Core Team 2006. R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-
900051-07-0, URL http://www.R-project.org.

Rao, C. R. 1964. The use and interpretation of principal component analysis in applied
research. Sankhya A 26:329-358.

Regosin. J. V., B. S. Windmiller, R. N. Homan, and J. M. Reed. 2005. Variation in
terrestrial habitat use by four pool-breeding amphibian species. Journal of
Wildlife Management 69(4): 1481-1493.

Richter, S. C., J. E. Young, R. A. Seigel, and G. N. Johnson. 2001. Post-breeding
movement of the dark gopher frog, Rana sevosa Goin and netting: implications
for conservation and management. Journal of Herpetology 35:316-321

282

Richter, S. C., and R. A. Seigel. 2002. Annual variation in the population ecology of the
endangered Gopher Frog, Rana sevosa Goin and Netting. Copeia 2002(4):962—
972.

Rubbo, M. J., and J. M. Kiesecker. 2005. Amphibian breeding distribution in an
urbanized landscape. Conservation Biology 19(2):504-511.

Russell, K. R., D. C. Guynn, Jr., and H. G. Hanlin. 2002. Importance of small isolated
wetlands for herpetofaunal diversity in managed, young growth forests in the
Coastal Plain of South Carolina. Forest Ecology and Management 163:43-59.

SAS Institute. 2002. Version 9.1. SAS Institute, Cary, North Carolina, USA.

Saunders, D. L., J. J. Meeuwig, and A. C. J. Vincent. 2002. Freshwater protected areas:
strategies for conservation. Conservation Biology I6(l):30-41.

Semlitsch, R. D. 2002. Critical elements for biologically based recovery plans of aquatic
—breeding amphibians. Conservation Biology l6(3):619-629.

Semlitsch, R. D., and J. R. Bodie. 2003. Biological criteria for buffer zones around
wetlands and riparian habitats for amphibians and reptiles. Conservation Biology
17(5):1219-1228.

Semlitsch, R. D., J. H. K. Pechmann, and J. W. Gibbons. 1988. Annual emergence of
juvenile mud snakes (F arancia abacura) at aquatic habitats. Copeia 1988(1 ):244-
246.

Skelly, D.K., E. E. Werner, S. A. Cortwright. 1999. Long term distributional dynamics
of a Michigan amphibian assemblage. Ecology 8(7):2326-2337.

Sotherton, N. W., J. W. Dover, N. R. W. Rands. 1988. The effects of pesticide exclusion
strips on fauna] populations in Great Britain. Ecology Bulletin 39:197-199.

Stebbins, R. C., and N. W. Cohen. 1995. A natural history of amphibians. Princeton
University Press, Princeton, New Jersey, USA.

283

Taylor, P. D., Fahrig, L., Henein, K., and G. Merriam. 1993. Connectivity is a vital
element of landscape structure. Oikos 68:571-573.

ter Braak, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector
technique for multivariate direct gradient analysis. Ecology 67:]167-1179.

ter Braak, C. J. F., and P. Smilauer. 1998. CANOCO reference manual and user’s guide
to Canoco for Windows, software for canonical community ordination (version
4). Microcomputer Power, Ithaca, New York, USA.

Trombulak, S. C., and C. A. Frissel. 2000. Review of ecological road effects on
terrestrial and aquatic communities. Conservation Biology 14:18-30.

USDA Forest Service. 2004. Forest inventory and analysis national core field methods
for phase 3 measurements. USDA Forest Service, Inventory and Analysis,
Arlington, VA, USA. < http://www.fia.fs.fed.us/ library /field-guides-methods-
proc/docs/p3_2-0_secl4_ 3_04.pdf.>. Accessed 3 Sep 2005.

Waldwick, R. 1997. Effects of forestry practices on amphibian populations in eastern
North America. Pages 191-205 in D. M. Green, editor. Amphibians in decline:
Canadian studies of a global problem. Society for the Study of Amphibians and
Reptiles, St. Louis, Missouri, USA.

284

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

WWWWMWW