 

 

’ LIBRARY
Mlciaigon State
‘ University

 

 

 

This is to certify that the
thesis entitled

DISTRIBUTION AND POPULATION DYNAMICS OF BEECH
SCALE (CRYPTOCOCUS FAG/SUGA) IN MICHIGAN

presented by

Daniel Wieferich

has been accepted towards fulﬁllment
of the requirements for the

MS. degree in Fisheries and Wildlife

 

 

Major Professor’s Signature
I Occ 296 0(

Date

 

MSU is an Afﬁnnative Action/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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 K:IProj/Acc&Pres/CIRCIDateDue.indd

DISTRIBUTION AND POPULATION DYNAMICS OF BEECH SCALE
(CR YPTOCOC US FAGISUGA) IN MICHIGAN

By

Daniel Wieferich

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE
Fisheries and Wildlife

2009

ABSTRACT

DISTRIBUTION AND POPULATION DYNAMICS OF BEECH SCALE
(CR YPTOCOC US FAGISUGA) IN MICHIGAN

By

Daniel Wieferich
Beech scale is a newly invading insect pest in Michigan, and consequently little is know
about their distribution and population dynamics. Beech scale densities were estimated
using qualitative visual assessments and a new quantitative method of digital
photography. Samples from 803 sites, ﬁ'om 2005-2009, across Michigan show that
beech scale has infested most of the distribution of beech in the Upper Peninsula, and
the north and western ranges of beech in the Lower Peninsula. In addition, beech scale
was found on several islands. Beech scale is distributed in 12 satellite populations,
which differed in density and spread rates. Also, some satellite populations are on
islands further from the mainland than wind dispersal is thought to carry scales. Four of
ﬁve satellites tracked from 2007-2008 showed increases in density whereas one satellite
declined. Within-year scale densities follow beech scale biology, showing slight
decreases in summer months when adults die, increases in fall when the new generation
of beech scale become visible, and slight decreases or no change in winter months when

scales are dormant.

ACKNOWLEDGEMENTS

This research would not have been possible without the generous ﬁnancial support by

the United States Forest Service PTIPS Program, thank you.

I would also like to recognize a few of the people who assisted with my research. My
advisors Daniel Hayes and Deborah McCullough, and my committee member Gary
Roloﬁ have helped me immensely throughout my graduate program. In addition, my
numerous ﬁeld/lab assistants have been crucial in collecting and analyzing my data.
These assistants included Jon Wagner, Christy Thomas, Shayna Hart, Justin Miller,

Justin Hodgins, Nicholas Wieferich and James Wieferich.

iii

TABLE OF CONTENTS

LIST OF TABLES .............................................................................. vi
LIST OF FIGURES ............................................................................. viii
CHAPTER 1

DISTRIBUTION OF AMERICAN BEECH (FAGUS GRANDIFOLIA) AND
BEECH SCALE (CR YPTOCOCC US FAGISUGA) IN MICHIGAN.

Introduction ................................................................................. 1
Methods ..................................................................................... 4
Results
Distribution of beech trees and evaluation of beech tree spatial data
layers ............................................................................... 7
Distribution of beech scale in the Lower Peninsula ................................. 8
Distribution of beech scale in the Upper Peninsula .......................... 9
Distribution of beech scale on islands ......................................... 10
Discussion ................................................................................. 11
CHAPTER 2

EVALUATION OF TECHNIQUES TO ESTIMATE BEECH SCALE
(CR YPTOCOCC US FAGISUGA) DENSITY.

Introduction ................................................................................ 27
Methods
Relationship between scale abundance and area of wax ................... 29
Photograph and qualitative data collection... ... .. .................................. 31
Photograph evaluation .......................................................... 32
Results
Relationship between wax and scale abundance ............................ 35
Digital photography ............................................................. 35
Qualitative results ............................................................... 36
Discussion ................................................................................. 36
CHAPTER 3

POPULATION DYNAMICS OF BEECH SCALE (CRYPT OCOCC US FAGISUGA) IN
MICI-HGAN.

Introduction ............................................................................... 46
Methods
Site establishment and sampling techniques ................................. 48
Site Revisits ...................................................................... 50

Population density over timeSO

iv

Results

Seasonal cycle of beech scale density on individual trees ................. 52
Annual beech scale dynamics ................................................. 53
Qualitative analysis of scale density on individual trees ................... 54
Plot level analysis of beech scale ............................................... 55
Discussion .................................................................................. 56
APPENDICES
Appendix A Site Coordinates ............................................................ 70
Appendix B IMAGEJ Photo Analysis Protocol ...................................... 96
Appendix C Beech Scale Photo Datasheets ........................................... 97
Literature Cited ................................................................................... 100

LIST OF TABLES

Table 1.1. Michigan islands sampled as part of our study. The years each island
was sampled follows the island name, showing the total number of plots and number of
plots infested in the given year. Residents indicates whether island has permanent
residence present. Distance represents shortest straight line distance to Michigan’s
mainland (measured in ArcGIS 9.2). Area (kmz) refers to island size ..................... 17

Table 1.2. Area (kmz) infested by beech scale in each year of sampling, as
determined by the minimum convex polygon method. Dashes (--) represent tirneﬁ'ame
without sampling. NA“ = Not enough points to create minimum convex polygon. The
same value was assigned to Cadillac and Ludington in 2009 because the two
populations coalesced .............................................................................. 18

Table 2.1. Comparison of the four digital photography analysis techniques are
summarized below. Time refers to the average time to analyze a single photograph,
mean scale and coefﬁcient of variation refer to the beech scale rating for each method.

The binary threshold method gives an area measure (cmz), while sub-sample methods
give qualitative ratings. Average sample size for sub-sampling techniques refers to
number of cells used in sub-sampling, where “no value” cells are not included .......... 40

Table 3.1. Number of plots visited in each region during each sampling time frame.
Annual density refers to the number of plots we revisited after one year (i 30 days) and
within year density refers to the number of plots we revisited in the given month and
year ................................................................................................ 61

Table 3.2. Results from annual photo analysis (% area infested) and visual
assessment of beech scale density, separated by region and sample timeframe. For
photo analysis, Year 1 and Year 2 represent mean % of bark covered by wax per tree.
For visual assessments, Year I and Year 2 were mean visual assessment. Visual
assessments were 0 = no beech scale, 1 = trace populations, 2 = patchy populations and
3= heavy populations. Time refers to sampling timeframe including March 2008 to
2009, July 2007 to 2008, and October 2007 to 2008. Number of trees per sample = 11,
Year 1 represents mean beech scale density per tree in initial visits, Year 2 represents
mean beech scale density per tree in revisits, CUP = Central Upper Peninsula, and EUP
= Eastern Upper Peninsula. Ratio represents scale density in Yearl/Year2. The
percentage of trees in the region with increasing beech scale abundance (Percent
Increase), decreasing abundance (Percent Decrease) and no change in abundance
(Percent Constant) are also reported .......................................................... 62

Table 3.3. General linear model results evaluating relationship between percent

change in beech scale on individual trees (n = 325), as measured by photographs. The
model was signiﬁcant (p < 0.001) and explained approximately 24 % of variation in

vi

change of scale density. Initial was Year 1 scale density measurements, from
photographs. Five regions were included (i.e. Cadillac, Emmet, Ludington, C.U.P, and
E.U.P.). Model included the two-way interaction between initial scale and region. . . ..63

Table 3.4. Annual change matrices, representing number of individual trees with
each initial (rows) and revisited (columns) qualitative ranking of scale density where 0
= no beech scale, 1 = trace populations, 2 = patchy populations and 3= heavy
populations ......................................................................................... 64

Table 3.5. General linear model results evaluating relationship between percentage
change in beech scale at the plot level (n = 19), as measured by photographs. Model 1
was signiﬁcant (p = 0.003) and explained approximately 81 % of variation in change of
scale density. Model 2 was signiﬁcant (p = 0.01) and explained approximately 32% of
variation in change of scale density. Initial was the mean scale density measurements
of each plot for Year 1 (calculated from photographs), other BA was the basal area of
all trees except beech in Year 1, and beech BA was the basal area of all beech trees.
The model also included the two-way interaction between initial scale and region. Five
regions were included (i.e. Cadillac, Emmet, Ludington, C.U.P, and E.U.P.) ........... 65

vii

LIST OF FIGURES

Figure 1.1. Sampling plot locations and results from 2005. Dark circles indicate
sites that were infested with beech scale. Open circles specify sites uninfested and
triangles indicate sites that were visited but had no beech trees ............................. 19

Figure 1.2. The northern hardwoods layer of Michigan, derived from the IF MAP
dataset ............................................................................................... 20

Figure 1.3. Sampling plot locations and results from 2006. Dark circles indicate
sites that were infested with beech scale. Open circles specify sites uninfested and
triangles indicate sites that were visited but had no beech trees ............................. 21

Figure 1.4. Sampling plot locations and results from 2007. Dark circles indicate
sites that were infested with beech scale. Open circles specify sites uninfested ......... 22

Figure 1.5. Sampling plot locations and results from 2008. Dark circles indicate
sites that were infested with beech scale. Open circles specify sites uninfested ......... 23

Figure 1.6. Sampling plot locations and results from 2009. Dark circles indicate
sites that were infested with beech scale. Open circles specify sites uninfested and
triangles indicate sites that were visited but had no beech trees ............................ 24

Figure 1.7. Sampling plot locations and results from 2005-2009. Dark circles
indicate sites that were infested with beech scale. Shaded circles specify sites
uninfested and triangles indicate sites that were visited but had no beech trees .......... 25

Figure 1.8. Beech scale satellite populations deﬁned in Michigan by sampling year.
Minimum convex polygons were used to identify boundaries of infestation ............. 26

Figure 2.1. Reference photos for the four point qualitative ranking system. A zero
was given to trees with no beech scale. (a) One was assigned to trees with a trace of
beech scale. (b) Two was assigned to trees with patches of beech scale. (c) Three was
assigned to trees with heavy infestations of beech scale .................................... 41

Figure 2.2. A total of 26 sites were established and revisited aﬁer one year (Annual

Sites). A subset of 14 sites was also revisited multiple times within a year (Within Year
Sites) ................................................................................................ 42

Figure 2.3. Eight grids were created for sub-sampling technique one and two. Each
grid cell was selected only once within the entire set of 8 grids. Technique I randomly
selected one grid to sub-sample a photo (total of 8 grid cells) while technique two used
two grids (total of 16 unique grid cells). ...................................................... 43

viii

Figure 2.4. Index of abundance relating number of beech scales to area (cm2) of
white waxy substance secreted by the insects. This index was created using samples
from the months of May to October. ........................................................... 44

Figure 2.5. Relationship between qualitative estimates of beech scale infestation
based on visual analysis and log quantitative measures as determined from binary

threshold methods .................................................................................. 45

Figure 3.1. Plot locations by region used for bimonthly and annual sampling of scale
density. Bimonthly plots (n=14) were used to determine bimonthly and annual changes
in scale density, while annual plots (n=12) were only used in annual analysis ........... 66

Figure 3.2. Within year change in beech scale densities by region from July 2007 to
June 2009. Percent area infested refers to average percent of bark infested with scale
per tree. Cadillac (n = 84 trees), Emmet (n = 36 trees), Ludington (n = 28 trees) ........ 67

Figure 3.3. Comparison of visual qualitative assessment and quantitative assessment
of change in beech scale density per tree. Change in qualitative scale density was a
difference of visual ranking from Year 1 to Year 2. Photos on each tree were averaged
for each year and then the difference between Year 1 and Year 2 was used to represent
quantitative scale density. ....................................................................... 68

ix

Introduction

Beech scale (Cryptococcusfagisuga Lind. )(Hemiptera; Coccidae) is a sap-
feeding insect native to western Asia and southeastern Europe (Gwiazdowski et a1.
2006) that is invasive throughout North America. In North America, beech scales are
host speciﬁc, feeding only on American beech (F agus grandifolia) (Ehrlich 1934).
Although beech scales reproduce annually, each parthenogenic adult can lay up to an

estimated 50 eggs (W ainhouse and Gate 1988), making populations capable of rapid

. . st .
expansron. Beech scales are only mobile as 1 instars, the stage also known as crawlers

(Ehrlich 1934). During this stage, crawlers move to a suitable location on their current
tree or disperse to another host tree. Wind is thought to be the main vector of dispersal
(Felt 1933; Wainhouse 1980), although migratory birds, wildlife and human movement
of infested logs may also play a role in dispersal (Ehrlich 1934; Houston 1994; Morin et
al. 2007). After dispersing to a suitable location, scales insert their stylets into the bark
and begin to feed on the sieve cells in the secondary phloem of the tree (Telewski
2009). Scales secrete white waxy wool over themselves as they feed. At this stage
scales molt and become immobile for the remainder of their lives.

When beech scales pierce the bark of beech trees they allow entry of two native
fungi, Neonectriafaginata Castleburry et a1. and Neonectria ditissima (Tul.& C.Tul.)
Samuels and Rossman, into the trees (Elrich 1934; Castlebury et al. 2006). Once
established on a tree, Neonectria will progress through and kill the living tissues of
bark, cortex, phloem, cambium, and sapwood (Ehrlich 1934). As lesions coalesce, the
tree looses capabilities to transport and store nutrients, which kills or reduces the health

of the tree (Ehrlich 1934).

Three phases of BBD have been deﬁned (Shigo 1972; Houston and O’Brien
1983; Houston 1994). The “advancing front”, refers to an area newly infested with
beech scale and without Neonectria infestation (Houston and O’Brien 1983). The
second phase is known as the “killing front”, which represents areas occupied by large
populations of beech scale and abundant Neonectria infection, causing beech mortality.
As the killing ﬁ'ont of beech bark disease passes through a landscape, an estimated 50 %
of beech trees over 25 centimeters in diameter are expected to die (Houston 1994). The
third and ﬁnal phase of BBD is the “aftermath forest”, referring to areas where the
disease is well established and continually affects current and regenerating beech trees.
In aftermath forests, beech scale populations decline because most surviving beech trees
are reduced in health from the Neonectria invasions (Shigo 1972; Houston 1994).

Beech scale, Neonectriafaginata and Neonectria ditissima were documented in
Michigan for the ﬁrst time in 2000 (O’Brien et a1. 2001; Castlebe et a1. 2006), placing
the state’s beech resource at risk from beech bark disease. Michigan has approximately
15 million beech trees over 22 centimeters in diameter at breast height (DBH) (Heyd
2005). These trees provide food and habitat for many species of wildlife such as black
bear, white tailed deer, and ruffed grouse (Burns and Honkala 1990). American beech
trees are also valued economically for use in pulpwood production, ﬂooring, furniture,
veneer, plywood, charcoal, baskets and other woodenware (Burns and Honkala 1990).
In addition, beech trees are valued aesthetically, both in natural settings and as
ornamental trees.

A critical step in the management of an invasive species such as beech scale is

to determine the distribution of the invader and its hosts. Morin et a1. (2005) mapped

the approximate beech distribution by interpolating the US. Forest Service forest
inventory analysis (F IA) (Hansen et a1. 1993) basal area data for the northeastern United
States, including Michigan. Although this dataset shows beech distribution in
Michigan, it was at a coarse 1 km resolution (Morin et al. 2005). In addition, FIA plots
are only located in forested areas, which would not detect beech trees in other land
types such as urban and agricultural areas.

The distribution of beech scale has not been fully documented for Michigan.
From 2001 to 2003, Petrillo et a1. (2005) monitored beech scale in Michigan to evaluate
impacts of beech bark disease. This study, however, focused on selected areas of the
state, and did not develop a complete state-wide map of the distribution of beech scale.
Similarly, from 2002 to 2003, Kearney et a1. (2005) evaluated the impact of beech bark
disease on wildlife resources, in a subset of established plots by Petrillo et a1. (2005).
Morin et a1. (2007) described the broad-scale distribution and predicted spread of beech
scale across the eastern United States, including Michigan. Their study, however, their
study provided only a coarse level of resolution at a county level. In this study, we
implemented an intensive survey to document the statewide distribution of beech scale
and beech trees spanning most of Michigan. Beech scale distribution was documented
for a ﬁve year timeframe, enabling us to assess the progression of the advancing front
over time. We mapped the annual distribution of beech scale in Michigan from 2005-
2009 and determined the annual change in beech scale distribution (area) for the

sampled years.

Methods

From May to August of 2005, we established 418 sites in 73 of Michigan’s 83
counties (Figure 1.1). Sites were primarily on forested lands, but beech trees in
agricultural and urban landscapes were also sampled. The beech basal area map
produced by Morin et a1. (2005) and a spatial land use layer provided by the Michigan
DNR (MDNR) assisted us in locating areas potentially containing beech trees. As
discussed the Morin et a1. (2005) map used F IA data to estimate beech basal area at a
1km resolution. The MDNR spatial layer (30 m resolution), known as Integrated Forest
Monitoring Assessment and Prescription (IFMAP) (MDNR 2003), allowed us to
identify areas designated as northern hardwood cover type, which often includes beech
(MDNR 2003) (Figure 1.2). Although these sources helped to initially guide sampling,
additional areas were also explored to determine whether beech trees and beech scale
were present. This ensured that a high proportion of the beech resource was included in
the study. In addition, this allowed us to evaluate the usefulness of the IF MAP layer for
locating beech trees. All of our sites containing beech trees were overlaid on the
IFMAP layer to determine the proportion of sites that corresponded with northern
hardwoods using ESRI® ArcMap.

Stands were initially examined for beech tree presence from the roadside. If the
habitat appeared suitable, but no beech was seen ﬁom the road, we walked through the
stand. Two types of sites were established, depending on accessibility of land. The ﬁrst
type of site was located on public land or on private land where we had permission to
access the property. In these sites three variable-radius plots were established using a

Panama 10 basal area factor (BAF) angle gauge (prism). Handheld GPS units (Garmin

Etrex), supporting 30 m accuracy, were used to collect coordinates of plot centers. A
center plot was identiﬁed as the ﬁrst area encountered with at least three large beech
trees. Then two additional plots, one established 100 m to the north or south of the
center plot, and the second 100 m to the east or west. The directions for the second and
third plots were randomly selected, unless property lines limited their directionality, in
which case we re-randomized and selected from assessable directions. At each site, the
basal area of the stand was recorded using the 10 BAF prism and each beech tree in the
three plots was evaluated for beech scale presence by visually examining the most
visible portion of the tree (below 4 meters). If the upper portions of the trunk or
branches were visibly infested, we recorded it.

The second type of site was established in areas with limited or minimal access
(e.g., private property) and in non-forested areas (e. g., roadside parks, campgrounds,
landscape trees). In areas with limited access, we sampled the ﬁrst ten beech trees
encountered along the road to assess presence of beech scale. Similarly in non-forested
areas, we sampled the ﬁrst ten beech trees encountered. On some occasions, less than
ten beech trees were available, in which case we sampled all available beech trees.
Similar to variable radius plots, beech scale infestation was determined by visually
examining each tree.

From May to August in 2006 to 2009, we continued to sample the areas across
Lower and Upper Michigan (Figures 1.4 — 1.7). When beech scale was detected,
additional sites were established in concentric circles approximately 1 km apart to
accurately delineate the advancing front every summer. Also, sites where beech scale

was absent in the previous year, but located in close proximity (< 5 km) to the

advancing front were revisited to assess the advancement of beech scale over time. In
situations where beech scale was present at revisited sites, the next closest, previously
uninfested site was revisited. This process continued until a buffer of uninfested sites
was established. If few sites with beech trees existed along the advancing front, an
adaptive sampling technique was used to ﬁll gaps in the sampling distribution. This
technique involved establishing new site locations within 10 km of sites known to be
infested and in between the advancing front and the closest uninfested sites. The
availability of these sites was limited by beech presence and land accessibility. Once
beech scale infestation was observed, a site was assumed to have beech scale through
the remainder of the study and therefore was not resampled.
In 2006, 48 sites that were uninfested in 2005 were revisited and 231 new sites

were established. In 2007, 43 sites that were uninfested in 2006 were revisited and 52
new sites were established. In 2008, 60 sites that were uninfested in 2007 were
revisited and 21 sites were created. In 2009, 103 sites that were uninfested in 2008
were revisited and 81 sites were created. Over the study as a whole, 803 sites were
established and 254 revisits were performed (Figure 1.7). The sites were distributed
across 11 of 15 counties in the Upper Peninsula and 62 of 68 counties in the Lower
Peninsula.

Beech trees on six of the state’s larger islands were examined, including Beaver,

Bois Blanc, Drummond, Mackinac, North Manitou and South Manitou Islands. These
islands differed substantially in size (roughly 14.5 to 645 kmz) and location (< 2 to 24

km from mainland) (Table 1.1). Similar techniques were used to sample islands as on

the mainland. Due to time and money constraints, islands were not sampled every year
(Table 1.1).

After sampling was completed, results were imported into ESRI® ArcGIS 9.2.
We deﬁned satellite populations, which represented beech scale infestations separated
geographically from one another by at least 20 km of uninfested stands with some beech
component. Using the Hawth’s Analysis Tools extention in ArcMAP (Beyer 2004), the
total infested area of each satellite population was calculated for each year using the
minimum convex polygon (MCP) method. Mob: (1947) deﬁnes a MCP as the complete
enclosure of all data points by connecting the outer locations in such a way as to create
a convex polygon. Areas of minimum convex polygons for consecutive years were

compared to evaluate annual change in beech scale distribution.

Results

Distribution of Beech Trees and Evaluation of Beech Tree Spatial Data Layers

Beech trees were located in 696 sites in 57 of the 73 sampled counties in
Michigan (Figure 1.7). Beech occurred across the majority of the Lower Peninsula,
including 49 of the 62 sampled counties (Figure 1.7). Although beech was widely
distributed, our data showed that most beech was concentrated in the north and western
parts of the Lower Peninsula. On the other hand, in the Upper Peninsula, our sampling
located beech trees across the eastern eight counties. In addition, beech trees were
present on all six of the islands we sampled.

The documentation of beech distribution allowed us to make comparisons with
other spatial data sets that can be used to predict the occurrence of beech. The northern

hardwoods layer of IFMAP was a poor indicator of beech presence. Of our 696 sites

with beech, only 140 sites overlapped with land classiﬁed as northern hardwoods. The
remaining 556 sites did not overlay areas classiﬁed as the northern hardwood type.
These results show that overall, the distribution of northern hardwood cover type
underestimates the range of beech trees in Michigan. In some cases, such as the
western portion of the Upper Peninsula, extensive areas classiﬁed as northern
hardwoods in IFMAP had no beech (Figure 1.2). Morin et al. (2005) mapped beech
basal area using F IA data, which we qualitatively compared to our observed occurrence
of beech. The datasets showed generally good correspondence between beech tree
presence and absence across the state, except in regions dominated by agricultural and

urban environments (e.g. southeastem Michigan).

Distribution of Beech Scale in the Lower Peninsula

In 2005, beech scale was distributed across substantial areas in the north and
western portion of Michigan’s Lower Peninsula (Figure 1.1, Figure 1.8 and Table 1.2).
Three distinct satellite populations of beech scale were identiﬁed in the Lower

Peninsula near Cadillac, Ludington and in Emmet counties (Figure 1.8), covering
approximately 2,667 km2 (Table 1.2).

From 2005 to 2009, beech scale distribution across the Lower Peninsula
substantially increased. Area infested increased every year (Table 1.2), but expanded
especially from 2006 to 2007, where it increased by 35%. Overall, the beech scale
distribution expanded by 146 %. A new satellite population was detected in Charlevoix
in 2006, where one isolated group of beech trees was infested. Two additional satellite
populations were discovered in 2009, one in Crawford county and one in Cheboygan

county (Figure 1.8). Also, the Cadillac and Ludington populations coalesced in 2009.

Even with the spread over the last ﬁve years, much of the forested area containing
beech in the eastern part of the Lower Peninsula remains uninfested.

There were large differences in overall expansion of beech scale distribution
between the different satellite populations in the Lower Peninsula from 2005 to 2009
(Table 1.7). The Ludington population spread slowly, only expanding by 40% before
coalescing with the Cadillac population. The Cadillac satellite expanded more rapidly,

increasing in area by 290% from 2005-2008. The Emmet population expanded most

rapidly with an initial area of 77 km2 in 2005 increasing to 841 km2 in 2008 and 1155

km2 in 2009, a change of 992 % and 1400 % respectively. The new satellite population
in Charlevoix found in 2006, did not appear to spread throughout the remainder of the
study.
Distribution of Beech Scale in the Upper Peninsula

In 2005, a majority of the Upper Peninsula’s stands of beech dominated forest
were infested by beech scale (Figure 1.1), with an infested area encompassing
approximately 6,214 kmz. Unlike the Lower Peninsula, there was only one contiguous
beech scale population in the Upper Peninsula until 2009, when beech scale was found
southwest of the main population in Menominee county (Figure 1.8). From 2006-2009,

beech scale distribution expanded east and west (Figure 1.8). The infested area

encompassed 8,203 km2 and 9,187 km2 in 2006 and 2007, respectively (Table 1.1). The

infested area then expanded to 10,373 km2 in 2008. By 2009, beech scale extended

throughout nearly all of the beech dominated forests in the Upper Peninsula, with

exception of the most western extent, expanding in area to 11,373 km2 (including the

Menominee satellite) (Table 1.1).
Distribution of Beech Scale on Islands
In 2005, beech scale was present in three of the 12 sites that were sampled

across Beaver Island. A total of 16 km2 was infested in 2005. Beaver Island was

resampled in 2006 using sixteen sites; ﬁve sites were infested with beech scale, seven
sites contained beech trees with no scale detected, and three sites contained no beech
trees. The total known infested area increased 125 percent from 2005 to 2006, when 36
km2 was infested.

Eight sites with beech trees were established on Drummond Island in 2006, but
no beech scale was detected. In 2008, these eight sites were revisited and one other site
was established. Beech scale was present in two of the nine sites, both located in the
south-central portion of the island. A sample of only two infested sites was inadequate
for an estimate of infested area using the minimum convex polygon method. In 2009,
the remaining seven sites were revisited. Beech scale was present at two more sites,
with a total infested area of 12 kmz.

Beech scale was present in 2006 on both Bois Blanc and Mackinac Islands. Of
the nine sites on Bois Blanc Island, the four western sites had beech scale present in an
area covering approximately 4 kmz. Seven of the 13 sites on Mackinac Island contained
beech scale. Unlike the other islands, Mackinac Island had infested sites scattered

throughout the island, with a total of 3 km2 infested.

10

In 2005, both the North and South Manitou Islands were sampled, but there was
no evidence of beech scale on either island. A total of 15 sites were established across
North Manitou Island, while time restraints limited sampling on South Manitou Island
to three sites. To be more thorough, we established six additional sites on South

Manitou Island in 2006, but did not detect beech scale.
Discussion

Our data allowed us to evaluate the effectiveness of the northern hardwoods
layer of IF MAP and the beech basal area map by Morin et al. (2005) for predicting
beech distribution in Michigan. We do not recommend using the northern hardwoods
layer in IFMAP to estimate beech distribution. Beech presence was over-estimated in
some regions and underestimated in other regions. The overestimation of beech likely
occurred in areas where northern hardwoods were present, yet without the beech
component. The underestimation of beech trees likely occurred because beech can be
found in more vegetation types than just northern hardwoods. On the other hand, beech
basal area mapped using FIA data (Morin et al. 2005) did provide useful predictions of
beech distribution in forested areas. We suggest users should be conscious of the coarse
resolution of FIA data, and that the data used to create the map only included samples
from forested areas. These trvo characteristics reduce accuracy in regions of the state
dominated by non-forested land types.

Our study expanded on previous efforts to describe the distribution of beech
scale. Morin et al. (2007) compiled beech scale presence in 2003 across the eastern
United States, including Michigan, at a county level resolution, which missed detailed

scale distribution. At the time of their study, Morin et a1. (2007) showed beech scale

11

occured primarily in the eastern Upper Peninsula and northwestern Lower Peninsula.
This study also modeled beech scale spread, predicting scale infestations in seven year
increments using 2003 as the initial year. Our 2009 data is similar to the model
predictions for 2010, but delineates the advancing front more clearly on an annual basis.
In addition, our results suggest the Morin et al. (2007) model over-predicted spread in
areas with limited distribution of beech (e. g. western Upper Peninsula and southern
Lower Peninsula). Using data from 2005 to 2007, Schwalm (2009) estimated beech
scale to spread 4 km/year and 1.5 km/year in the Upper and Lower Peninsulas of
Michigan, respectively. Our data suggest beech scale may be spreading more rapidly
in both peninsulas than Schwalm (2009) predicted. Similar to results noted by Schwalm
(2009), satellites spread more quickly in the Upper Peninsula than the Lower Peninsula.
Like many other states, studies of BBD in Michigan have largely focused on the
impacts of beech bark disease rather than the distribution of beech scale, i.e. the
advancing front (Kearny et al. 2005; Petrillo et al. 2005). Petrillo et al. (2005) provided
the most complete record of beech scale distribution, surveying 284 sites in 30 counties
and ﬁve islands from 2001-2003. Our ﬁve year study expanded on these projects with
803 total sites and 254 revisits across 73 counties and 6 islands, allowing us to deﬁne a
ﬁner scale and more complete distribution of beech scale. Our distribution results
detected nine satellite populations not documented by Petrillo et al. (2005), including
Charlevoix, Cheboygan, Crawford, Emmet, Menominee, Beaver Island, Mackinac
Island, Bois Blanc Island and Drummond Island. In addition, our data showed satellite
populations, deﬁned in previous studies, to be much larger in area. The differences

observed in beech scale populations from previous studies are largely due to their

12

continued spread since these studies, although our intense sampling helped identify
several new satellite populations. In addition, our data allowed us to estimate the area
of infestations, enabling us to monitor spread of beech scale over time.

Our data suggests that beech basal area may be one factor affecting initial beech
scale infestations. In 2005, most of the infested area in the Upper Peninsula had a

relatively high beech basal area, according to the beech basal area map created by

Morin et al. (2005) (0.2 — 7 mz/ha; 0.87 —- 30.5 ftz/acre). By 2009, beech scale had
spread to most of the remaining forests where beech basal area was > 0.2 mz/ha. Much

of the remaining beech areas had a low beech basal area (< 0.2 m2/ha), with exception

of a small area isolated southwest of the infestation (i.e. southern Menominee County).

Similarly, the satellites in the Lower Peninsula all started in areas with beech basal
areas > 0.2 m2/ha. Unlike the Upper Peninsula, there are still uninfested locations with

relatively high beech basal areas, although most of the uninfested regions are dominated
by agricultural and urban developments (MDNR 2003). These results suggest that
beech scale initially established in areas with relatively high beech basal area, spreading
to surrounding areas, including those of lower beech basal areas. We suggest that
managers consider beech basal area when studying and modeling spread of beech scale.
In addition to documenting the statewide distribution of beech scale, we
calculated minimum convex polygons (MCP) to infer spread rates of beech scale over
time. We used the MCP method because it is capable of using presence data to estimate
distribution. Several other techniques using presence and absence data were also
explored such as inverse distance weighting and kriging, but our sampling distribution

varied annually, making annual comparisons of these estimations inappropriate. One

13

disadvantage of the MCP technique is overestimation of distribution, especially in cases
where irregular shaped ranges are present (Burgman and Fox 2002). This was observed
in some satellite populations, where absence data was within the MCP estimation of
beech scale presence. We believe the MCP provided reasonable estimations of relative
annual range expansion of beech scale in our project.

Overall, satellite populations seemed to differ in spread rates. The Ludington
satellite population spread the slowest, which may be due to a more ﬁagrnented
population of beech in surrounding locations, because of areas with no beech and higher
densities of urban and agricultural lands. In contrast, the Cadillac and Emmet
populations increased rapidly. Unlike the Ludington satellite population, these
populations were in close proximity to a more continuous distribution of beech. In
addition to spatial differences in spread rates, beech scale has not spread evenly year to
year. This suggests that external factors such as weather may affect the spread of beech
scale.

The numerous satellite populations also allow us to infer mechanisms of
dispersal. Previous studies suggest that beech scale’s primary mode of natural dispersal
is wind (Felt 1933; Wainhouse 1980). Wainhouse (1980) showed that wind speeds of
less than two meters per second cause a mean dispersal of crawlers 10 m, limited by the
initial height of the insect and wind speed. Wainhouse and Gate (1980) also suggest
that scale crawlers above the canopy combined with high wind speeds may be
responsible for long-range dispersal of 6—1 8 kilometers per year occurring in North
America. Other forms of dispersal (e. g. migratory birds, wildlife and human movement

of infested logs) may also play a role in spreading beech scale, but no concrete evidence

14

has been published (Ehrlich 1934; Houston 1994; Morin et al. 2007). Of these modes
of dispersal, the most probable mode of long-range transport is movement of infested
logs by humans and the spread of scale via migratory birds. We detected the formation
of several satellite populations of beech scale that were > 20 km apart, suggesting that
wind dispersal was not responsible. Other modes of dispersal such as those discussed
above are likely playing large roles in the dispersal of beech scale in Michigan.

The unique opportunity to sample beech scale on six of Michigan’s islands also
provided insights into mechanisms of spread, especially anthropogenic actions.
Although all islands differed in size and location, the only islands without beech scale
were the Manitou Islands, which were unique because they have no permanent residents
(Table 1.1). In addition, Beaver Island is located farther from the mainland (24 km)
than any other island, making it unlikely that wind dispersal deposited scales on trees on
the island, yet it was infested. This again suggests that humans may have played a role
in spreading beech scale to Beaver Island too. Also, Beaver Island is located in ﬂyways
of several migratory bird species, so migratory birds may have transported beech scale
to the island (Lincoln 1939).

Knowing Michigan’s distribution of beech scale over time can be a valuable tool
for natural resource managers, allowing beech bark disease to be incorporated in
management plans. Current information about beech scale distribution can be used to
help reduce the spread and impacts of beech scale and BBD by reducing overstory
beech density and overmature beech in areas adjacent to infestations that are dominated
by beech trees (Mielke et al. 1987). In addition, the distribution of beech scale can help

managers determine when impacts of BBD are likely to occur. Distribution data can be

15

incorporated into a guidance plan for beech management and future timber sale activity
For example the restriction of timber and ﬁrewood harvest during the crawler stage
could likely minimize anthropogenic transportation of actively dispersing scales.

The fast paced spread of beech scale in Michigan, due to the combination of
local and long distance dispersal is similar to previous observations in the northeastern
United States. In the northeastern states beech scale was often ﬁrst documented in
isolated satellites, where it then expanded locally. For example, satellite populations of
beech scale were discovered in Massachusetts, Maine, New Hampshire, New York and
New Jersey in 1935, in West Virginia in 1990, in North Carolina and Tennessee in
1999, and in Ohio and Michigan in 2001. All of these satellite populations meet our
deﬁnition of being separated geographically from the established scale populations by
at least 20 km of uninfested stands with some beech component. Our data conﬁrms that
long distance dispersal is occurring throughout Michigan too, shown by the
development of several distinct satellite populations. With the ability of beech scale to
spread at variable rates, the remaining states with uninfested beech trees should be
monitored and preparations should be made for beech scale infestations in the near

future.

16

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e e meow
e m meow

Fw 5 oz amazes. 32:52 .w
e E. meow

ww :. 02 595:2 sauces. .z
N. or eeew

new v mm; :05: 09.285.
m e meow
w o eeew
e or eeew

eve w v me> :05: 302.820
v e eeew

nwr e mo> :05: 0.55 m_om
m or eeew
m we meow

mil. vw mm; 5920:). co>mom

:5: 23m some:
o
Aug: a h< cue—mama ﬂeet—mom 9.3 :23 Bar. % 30... u 9.2».

 

 

 

 

 

 

 

dim e53 8 when: Awe—xv «one. Awe mace E ooh—53.5 ena—EaE PoemEBE 8 853% on: 2988 $2.55
3:82AM“: 85.65 .283 2: co 3: $338k Evan—Eon .5533 m88_e=_ ﬁaoemwom use» :on 05 E eouomﬁ 82m no .383: 23 32a we
838:: :88 05 wEBoam 6:8: ecu—ﬂ 2: 95:8 BEES 83 one? some who» 2:. 39% So mo Ea we BEES €53 53:22 :4 ﬂash

l7

 

Table 1.2: Area (km2) infested by beech scale in each year of sampling, as determined by the
minimum convex polygon method. Dashes (--) represent timeframe without sampling. NA" =
Not enough points to create minimum convex polygon. The same value was assigned to
Cadillac and Ludington in 2009 because the two populations coalesced.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Year
Satellite 2005 2006 2007 2008 2009
Cadillac 51 136 144 199 5 200
Ludington 2,539 2,813 3,287 3,560 ’
Charlevoix 0 NA" NA" NA" NA*
Cheboygan 0 - -- - NA*
Crawford 0 - — - 196
Emmet 77 149 739 841 1,155
Total Lower
Peninsula 2,667 3,098 4,170 4,600 6,551
Menominee O - - - 255
Upper Peninsula 6,214 8,203 9,187 10,373 11,547
Total Upper
Peninsula 6,214 8,203 9,187 10,373 11,802
Beaver Island 16 36 - - -
Bois Blanc Island - 4 - - -
Drummond Island - - - NA* 12
Mackinac Island - 3 —- - -
North Manitou Island 0 - - -- -
South Manitou Island 0 0 - - -
Overall Total 8,897 11,344 13,357 14,973 18,365

 

18

 

 

 

 

Legend .°
. Scale Presence
0 Scale Absence . 0 at;
I No Beech

 

 

 

 

 

 

Figure 1.1: Sampling plot locations and results ﬁ'om 2005. Dark circles indicate sites that were
infested with beech scale. Open circles specify sites uninfested and triangles indicate sites that
were visited but had no beech trees.

     

 

Legend

- Northern Hardwoods

 

 

 

I:][Jlll::::lllle
03060 120 180 240

Figure 1.2: The northern hardwoods layer of Michigan, derived from the IFMAP dataset.

20

 

 

 

 

 

KI;
:gi‘e
Legend 0:, W? ‘
° Scale Presence °. 5:0 :
0 Scale Absence . o t: * j,’
A No Beech 8 o o

 

 

 

 

 

 

 

Figure 1.3: Sampling plot locations and results from 2006. Dark circles indicate sites that were

infested with beech scale. Open circles specify sites uninfested and triangles indicate sites that
were visited but had no beech trees.

21

 

 

 

Legend
0 Scale Presence
0 Scale Absence

 

 

 

 

 

 

Figure 1.4: Sampling plot locations and results from 2007. Dark circles indicate sites that were
infested with beech scale. Open circles specify sites uninfested.

22

 

 

 

Legend '
0 Scale Presence
0 Scale Absence

 

 

 

 

 

 

Figure 1.5: Sampling plot locations and results from 2008. Dark circles indicate sites that were
infested with beech scale. Open circles specify sites uninfested.

23

 

 

 

Legend to °~
0 Scale Presence °. 0
0 Scale Absence
4» No Beech

 

 

 

 

 

 

 

 

Figure 1.6: Sampling plot locations and results from 2009. Dark circles indicate sites that were
infested with beech scale. Open circles specify sites uninfested and triangles indicate sites that
were visited but had no beech trees.

24

 

   

 

Legend
0 Scale Presence
Scale Absence
A No Beech

 

 

 

0 80

 

 

 

Figure 1.7: Sampling plot locations and results from 2005-2009. Dark circles indicate sites that
were infested with beech scale. Shaded circles specify sites uninfested and triangles indicate
sites that were visited but had no beech trees.

25

 

 

Satellite Populations
1. Ludington 7. Beaver Island

2. Cadillac
3. Crawford

5. Emmet

8. Bois Blanc Island
9. Mackinac Island

4. Charlevoix 10. Drummond Island

11. Upper Peninsula

 

6. Cheboygan 12. Menominee

 

Year

W zoos
_ 2006

_ 2007
. 2008

2009

 

 

 

 

u:— km
0 40 80 120

 

 

 

Figure 1.8: Beech scale satellite populations deﬁned in Michigan by sampling year. Minimum
convex polygons were used to identify boundaries of infestation.

26

Introduction

Invasive forest insects cause substantial economic and ecological damage across
the world (Sakai et. a1 2001; NRC 2002). Recent costs of invasive forest insects in the
United States were estimated to be approximately $4 billion per year (USDA Forest
Service 2007). These insects also cause irreplaceable alteration and destruction to
forest systems by depleting biodiversity, altering unique forest systems and causing
species endangerment and extinction (Wilcove et al. 1998).

When studying invasive forest insects, scales (Hemiptera: Coccidae) are of
special concern. Scales are generally sap feeders and several species reproduce
parthenogenically (Miller and Kostarab 1979). Miller et al. (2005) found that roughly
25 % of scale insects in the United States are non-native, a high percentage compared to
other insect groups. Many of these non-native scales are major pests of crops, plants or
trees in the United States, causing economic and ecological damages (Miller and
Kosztarab 1979; Miller et al. 2005; Kondo et al. 2008).

Beech scale (Cryptococcusfagisuga Lind.) is one of the many invasive scales in
the United States. Originally from western Asia and southeastern Europe
(Gwiazdowski et al. 2006), beech scale was unintentionally transported to Nova Scotia
on imported nursery stock in 1890, where it spread to the United States sometime
thereafter (Ehrlich 1932). Beech scale. is host-speciﬁc to American beech (F agus
grandifolia) (Ehrlich 1934), and has spread through most beech dominated forests of
the United States (Morin et al. 2007).

Despite the widespread distribution and long-term presence of beech scale in the

United States (Morin et al. 2007), few quantitative descriptions of its population

27

dynamics are available (Houston et. a1 1979). Like most scale insects, beech scales
reproduce by parthenogenesis, allowing populations to build exponentially even from a
single individual (Ehrlich 1934). During mid-summer, each adult lays approximately
50 eggs before dying (Wainhouse and Gate 1988). Eggs hatch into ﬁrst instars, also
known as crawlers or nymphs, in late summer. Crawlers are equipped with legs and
antennae allowing for mobility (Ehrlich 1932). They may move to a suitable location
on their current tree or are carried by wind or wildlife, to nearby host trees (Ehrlich
1932; Felt 1933; Wainhouse 1980; Houston 1994; Morin et al. 2007). After moving to
a suitable location, crawlers pierce the outer bark with their stylets to feed on the
phloem of the tree. Crawlers molt into second instars, shedding their legs and becoming
immobile. As they feed, the scales secrete a white waxy coating over their bodies,
allowing for protection from weather and predators. The second instars overwinter and
molt into an adult stage the following spring (Ehrlich 1932).

Despite the univoltine life cycle and immobility of beech scale that make the
study of its population dynamics in a natural setting more feasible than that of most
invasive insects, little is known about beech scale population dynamics. Most previous
studies on beech scale population dynamics used a qualitative ranking system to
estimate scale density. Several authors, including Houston et al. (2005) and Kearny et
al. (2005), have visually ranked beech scale density. This is a quick and simple way of
assessing beech scale populations but is subjective and gives coarse results. In other
studies, individual scales on trees were counted (Wainhouse1980; Gardner 2005), but
the time required makes it difﬁcult to implement this approach for multiple trees and

sites. Further, while counting individual scales gives a true census within the sample

28

area, the sample area usually represents a small fraction of the total scale population.
For example, Wainhouse (1980) used four 1 cm2 samples to represent beech density on

an entire tree.

Digital photography can be a useﬁrl way to study beech scale because a high
contrast typically exists between the grey bark of beech trees and the white waxy
substance secreted over the immobile scales. We devised and evaluated a method of
quantifying beech scale density using digital photography. In addition, we compared
accuracy and efﬁciency of four techniques of assessing beech scale density on digital
photographs. Our goal was to determine if digital photography could provide
quantiﬁable results that were more time and cost efﬁcient than previous methods such
as individual scale counts. We also wanted to determine if this method would also
allow for reproducible results of beech scale density estimates. This could enable
scientists to monitor effects of weather or other factors affecting scale dynamics over

time.
Methods

Relationship between scale abundance and area of wax

We created an index to test the accuracy of using the percentage of bark area
covered by beech scales’ waxy coating to quantify the number of scales. From May to
October 2008, samples of infested beech tree bark were collected using a bark punch
and were brought back to the lab for processing. A total of 105 samples, each 2.14 cm
in diameter, were collected from ﬁve distinct beech scale populations throughout
Michigan; 22 samples were collected from 2 sites in Ludington’s scale populations in

May, 21 samples ﬁ'om 2 sites in Cadillac in May, 23 samples from 2 sites in Emmet in

29

June, 12 samples from 1 different site in Emmet scale populations in October, and 27
from 3 sites in the Upper Peninsula in July. No more than four samples were collected
from one tree and samples were from all aspects and heights less than 2.5 m of the tree.
Approximately one third of the samples had only a sparse amount of beech scale wax,
one third of the samples had an intermediate amount and one third appeared to be
heavily infested, with high amounts of scale wax. To collect each sample, the punch
was ﬁrst used to delineate the sample area on the tree. After sample areas were
delineated, they were photographed and then removed from the tree with the bark punch
(see Photograph collection below).

Photographs were analyzed using image analysis software called ImageJ
(available at http://rsb.info.nih.gov/ij; developed by Wayne Rasband, National Institutes
of Health, Bethesda, MD.), which allowed us to manually apply a binary threshold to
each photograph (Appendix B). Based on the brightness associated with each pixel, the
threshold was adjusted to select pixels that contained the white wax from beech scale.
Thresholds were different for each photograph, depending on photo quality, brightness
and contrast. The selected pixels were then used to calculate the percentage of pixels
covered with scale wax. After each photo was analyzed, the number of beech scales on
the sample of bark was counted using a microscope.

A linear regression was used to represent the relationship between the area
covered by wax and the number of insects in the sample. We used the equation y = a +
bx, where y represents the number of insects, x is the area of wax coverage, b is the

slope and a is the y-intercept.

30

Photograph and qualitative data collection

We established 26 eight meter ﬁxed-radius plots from July to August 2007. In
each plot, all beech trees over 6 cm in diameter at breast height (DBH) were tagged to
ensure unique identiﬁcation in subsequent visits. Individual plots contained two to 22
beech trees, with a mean number of 10 i 1 beech trees per plot. When less than eight
beech trees were present in one plot, beech trees closest to the plot were tagged until at
least 8 trees were included. There were two exceptions, where only ﬁve and seven
beech trees were tagged, due to low beech basal area.

A qualitative ranking of beech scale abundance was recorded for each tree to
allow for visual estimation methods used in previous studies to be compared to the
digital photography method. We used a four point qualitative ranking system using
reference cards (Figure 2.1), which help produce consistent rankings. Qualitative ranks
were 0 = no beech scale, 1 = trace populations, 2 = patchy populations and 3= heavy
populations. Qualitative ranking was performed by walking around the tree searching
for signs of beech scale (i.e. white wax) on the readily visible portion of the tree, < 4 m
above ground. If the upper portions of the tree were also visibly infested, this was
incorporated into the ranking.

After all beech trees in a plot were tagged and measured, three unique
photographs were taken of each tree. A stratiﬁed random sampling technique was used
to determine photo heights and aspects. One photo was taken at 0.9, 1.2 and 1.5 m
heights. Each photo height was assigned a random aspect before arriving at the site.
Photos were taken using a tripod with a built-in stabilizer to ensure the camera was the

same distance from the tree (30 cm) for each photograph. To reduce potential

31

problems, a duplicate photo was taken at each setting. If shadows were present within
the photo area, sunlight was blocked using a clipboard to reduce contrast within the
photo.

To study annual change in beech scale populations, all sites were revisited 365
days (i- 30 days) after establishment. In addition, a subset of 14 plots was revisited
more frequently to detect within-year changes of beech scale populations (Figure 2.2).
The 14 plots were revisited in October and December 2007, and January, March, May,
June, July and October 2008, and March and June of 2009. Due to a snowstorm, only
half of the plots were accessible in December 2007, but remaining sites were revisited
in early January 2008. During each revisit, qualitative and quantitative measures of
beech scale were again recorded. All photos were retaken on the same trees and at the
same heights and aspects of original visits to assess changes in beech scale density.
Photograph evaluation

Two general approaches were developed and tested to determine an efﬁcient
method of quantifying beech scale density through digital photograph analysis. The
ﬁrst approach used image processing software to calculate beech scale density on each
photo. In the second approach, we compared three sub-sampling techniques to estimate
scale density on each photograph. Thirty photographs that represented unique trees,
representing various sizes of beech trees and beech scale densities, were selected to test
both techniques. An equal number of photographs were selected from the following
classiﬁcations: large diameter beech trees (> 18 cm DBH) with heavy, moderate, and
low visual estimates of scale infestations and small diameter trees (< 18 cm DBH) with

heavy, moderate, and low infestations. Scale density, time required for analysis, and

32

problems encountered were recorded for each technique. Each technique was practiced
on ten different photos before the trial was conducted. This allowed the user to become
familiar with the protocol, making the time of analysis representative of the method in
routine practice.

The ﬁrst approach used ImageJ software as described in Relationship between
scale abundance and area of wax. This program allowed us to adjust a binary threshold
to select pixels with wax. Selected pixels were used to calculate density of infestation
within each photograph.

The second approach used Adobe Photoshop Elements ® software, which
allowed a system of random grids to be constructed and overlaid on each photo. Three
techniques were tested using the grids to sub-sample the images under different
constraints. For all three sub-sampling techniques, photographs were overlaid with an
8x8 grid (64 cells total).

In the ﬁrst sub-sampling technique, the 64 cells were used to randomly
predeterrnine eight grids, each containing eight transparent cells for photo evaluation
(Figure 2.3). Transparent grid cells were non-overlapping between grids (Figure 2.3).
One grid was randomly assigned for each photo, allowing us to visually estimate scale
coverage within the eight transparent grid cells. A qualitative rank of scale density was
assigned to each transparent grid cell. Zero represented no scale, one represented 1-
25% of the grid cell being infested, two represented 26-50%, three represented 51-75%
and four represented 76-100%. When the tree occupied less than half of a grid cell, the
cell was assigned a “no value” ranking, meaning it was not used in the analysis. This

was common in photos of trees with small diameters because smaller trees did not span

33

the entire photo. This occurrence reduced the number of cells evaluated in such cases.
After each grid cell was assigned a scale density, the eight cells were averaged to
represent beech scale density of the entire photograph.

The second technique used the same predetermined grid cells and evaluation
techniques as the ﬁrst technique. Unlike the previous technique, two predetermined
grids were overlaid on each photo, allowing for 16 grid cells to be evaluated.
Techniques 1 and 2 were expected to have low processing times for each photograph
because they used random subsarnples.

The third technique used a single grid approach (64 total cells). A total of 15
randomly selected grid cells were used to subsample the photo. Unlike techniques 1
and 2, this technique required each sampled cell to overlay the tree within the
photograph ensuring the total sample included 15 grid cells. In cases where a grid cell
did not overlay the photograph, it was discarded and another cell was randomly
selected. The 15 grid cells were ranked qualitatively, similar to the ﬁrst two techniques
and values were averaged for the photograph. This technique was expected to allow for
a randomization of sampling for each photo and a more consistent sample size than
methods 1 and 2, but require more time than the ﬁrst two methods because grids new
grids were created for each photo.

Some common rules were established with the three subsampling techniques
using Adobe Photoshop to maintain consistency throughout the analysis.

0 Grid cells were only used if more than half of the cell overlaid the tree;

0 Scale infestation on the grid line was not counted;

34

o If less than 4 grid cells overlaid the tree, then a new grid was randomly

selected.

Results
Relationship between wax and scale abundance
A positive linear relationship existed between percentage of the photographed
area covered by the white wax and the true number of beech scales (Figure 2.4). The

number of beech scales counted on each sample ranged from 0 to 1,768. The equation,
y = 869.02x + 45.798, yielded an adjusted r2 value of 0.796 (p < 0.0001), indicating that

one cm2 of white wax in the photo represented approximately 869 beech scales. As

expected an intercept close to zero was observed (Figure 2.4).

Digital photography

Digital photography produced more precise and less subjective results than
qualitative analysis, but required more time. This method took one hour to analyze all
30 photos, taking approximately two minutes per photo, or six minutes per tree. In
addition, 2.3 to 3.5 minutes per photo were required to process photos in the lab (Table
2.1). In contrast, the qualitative estimate of scale density only took two minutes per tree
and required no ﬁrrther processing. Overall, a tree sampled with digital photography
took greater than six times as long as a qualitative assessment of the same tree.

The binary threshold method was the most time consuming technique (Table

2.1). This approach, however, had the lowest coefficient of variation, indicating it was
more precise than other techniques. This technique resulted in a mean area of 0.13 cm2

of white wax per photo. The sub-sample 1 technique was least time consuming; sub-

35

sample 2 and 3 methods took approximately 1.5 times as long. The three sub-sampling
techniques had similar coefficients of variation and mean values of white wax per
photo, indicating similar accuracy and precision. Sub-sample 1 included an average
sample size of 5.7 cells, while sub-sample 2 and sub-sample 3 had average samples
sizes of 11.1 and 14.9 cells, respectively.
Qualitative results

The mean proportion of a tree that was infested, as measured by digital
photographs, varied strongly across categories assigned by visual assessment (Figure
2.5). In general, the central 50 % of digital measures for each category did not overlap
with adjacent categories (Figure 2.5). In addition, the qualitative method allowed for
complete assessment of each tree. This method, however, gave less precise data than

digital photographs and no form of visual documentation for later reference.

Discussion

Although we found no evidence that digital photos were used previously to
evaluate population dynamics of beech scale, similar techniques have been used for
applications such as quantifying percentage forest canopy cover, calibration of pesticide
spray from airplanes, and land-use classiﬁcation (Englund et al. 2000; Holowrricki et. a1
2002; Bruzzone and Fernandez Prieto 2000). Foresters often use hemispherical canopy-
photographs to quantify canopy cover and understory light availability (Englund et al.
2000; Nobis and Hunziker 2004). The photos in these studies are often processed in a
similar manner as the binary threshold method and are thought to provide advantages

such as versatility and lasting records (Englund et a1. 2000).

36

Our results demonstrate digital photography is a useful technique for capturing
beech scale density. Unlike most insects that are capable of moving throughout their
entire life, the immobility of beech scale makes it easier to study over time. In addition,
the white wax secreted by beech scale provides contrast from the bark of host trees,
allowing for identiﬁcation in photos, even when implementing a threshold technique.
These traits also allowed an index of abundance to be created using digital photography.
Unlike visual assessments, digital photography provides a lasting record of beech scale
abundance at a given time, allowing for more precise comparisons of density over time.
In addition, digital photography provides concrete and reproducible results, unlike the
subjective results of visual assessments.

Although our photographic technique was effective, there are cases where the
use of simple qualitative assessments of beech scale may be useful. Qualitative ranking
systems using reference photos gave comparable beech scale rankings as the binary
threshold technique, yet reduced time expenditures greatly. Also, several qualitative
assessments were accomplished in minutes in the ﬁeld, while photo analysis required
more time for ﬁeld collection plus hours of processing in the lab. With this in mind,
qualitative assessments may be more suitable for situations that do not require
quantitative data or when rapid assessment is desirable. Also, qualitative methods may
be more useﬁrl than photos when sampling is conducted in areas that contain a trace of
beech scale. Qualitative assessment of the entire visible portion of the tree increases the
likelihood of detecting small amounts of beech scale, while sub-sampling with digital

photography may fail to capture traces of scale by chance alone. In such situations, a

37

double-sampling approach using both qualitative and quantitative ranking may be
superior.

The binary threshold method was the most precise of the four techniques for
quantifying beech scale density from digital photography. The threshold data is most
precise because it directly measures scale density on the complete photograph while the
other methods only sub-sample the photo. Although the binary threshold method was
most precise, it was also the most time consuming of tested methods. We feel,
however, that the marginal cost in processing time is worth the additional precision if
studying population dynamics. Also, if a sub-sampling approach of photos is desired,
we recommend the sub-sampling 1 technique because it requires the least processing
time and produces similar results as other sub-sampling techniques.

Although data processing using binary threshold measurements had its
advantages, minor problems arose in this application. One of the largest challenges
involves the lack of consistency in contrast within and between photos. The most
ﬁequent contrast problem within a single photo was shadows covering portions of the
photo. This made automated detection of beech scale difﬁcult. We minimized this

problem in the ﬁeld by shading the entire photo frame with a clipboard. The most
common contrast issue among photos included differences in bark roughness, overall
photo brightness or bark coloration which made it unfeasible to use a set threshold for
all photographs. We had to set individual thresholds to best represent beech scale on
each photo, which required additional processing time. Another limitation of using
binary threshold measurements to assess scale density was misclassiﬁcation of other

white coloration on the bark. Examples of misclassiﬁcations included snow, spider

38

eggs, and other light-colored insects. Compared to beech scale, however, the surface
areas of these misidentiﬁed objects were minute. In addition, ImageJ encountered some
problems identifying beech scale when trees were covered with light colored moss and
lichens. Most of these problems were dealt with by manipulating images using paint
options in ImageJ to increase contrast of the wax and bark by changing color of infested
pixels, improving threshold sensitivity. Although paint options allowed us to cope with
problems, these issues still made the analysis more time consuming.

We recommend that new advances in digital photo analysis be explored in future
applications for assessing beech scale density and dynamics. Nobis and Hunziker
(2004) suggest automatic threshold algorithms may be less time consuming and more
objective, comprehensible and reproducible. Other image processing software exists
and could also be explored, to determine if they provide greater processing capabilities
and reduced processing time.

Given our success of quantifying beech scale density we feel this method could
be expanded to broader use. In particular, other insects with characteristics similar to
beech scale could be efﬁciently monitored using digital photography. Insects that are
mostly immobile during their life cycle and have high contrast with their host plant
would be most suitable. Some of these insects may include herrrlock woolly adelgid
(Adelges tsugae; Annand), balsam woolly adelgid (Adelges piceae; Ratzeburg), and

other scale species.

39

Table 2.1: Comparison of the four digital photography analysis techniques are summarized
below. Time refers to the average time to analyze a single photograph, mean scale and
coefﬁcient of variation refer to the beech scale rating for each method. The binary threshold

. 2 . . . . .
method gives an area measure (cm ), while sub—sample methods give qualitative ratings.

Average sample size for sub-sampling techniques refers to number of cells used in sub-
sampling, where “no value” cells are not included.

 

 

 

 

 

 

 

Time Mean Coefficient of Average Sample
Technique (min/photo) Scale Variation Size
Binary Threshold 4.0 0.13 0.40 Entire Photo
Sub-sample 1 2.3 0.92 0.60 5.7 grid cells
Sub-sample 2 3.5 0.96 0.54 11.‘Ijrid cells
Sub-sample 3 3.1 0.94 0.75 14.9 grid cells

 

 

 

 

 

 

40

.038 :83 we meeuﬂmomﬁ
4060; H23 805 8 Bewmmme was» 0045. A8 .038 £003 me 8:83 at? meet 9 “30:39.3 33 93. A5 .038 £0009 mo 02.5 a £15 80.: 9
Bowie 33 0:0 A3 .28a :83 on its 0.00.5 3 =0>_m 33 obn < .889? wet—Sc 0>sS=aec e58 See 05 Lee 88% 00:0c0e0M "—.w 0.5»:—

 

41

 

 

 

 

 

. _ S
.15
-. F \I‘ i rilim-k"
1'" ,~ u: g .‘ f‘
i”; Q _ * \
' Al.
I. “
‘. - .
(D
o , t" 5
. t, I../
iiv' ‘
Revisited Sites L x_ r
A Annual Sites , ‘ ; , ‘1 1;. 1'-
0 Within Year Sites . g
‘ rrrrrrrr . _ , . _, . .4 \h‘.
— J m ."
o 37.5 75 150 225

Figure 2.2: A total of 26 sites were established and revisited after one year (Annual Sites). A
subset of 14 sites was also revisited multiple times within a year (Within Year Sites).

42

 

Figure 2.3: Eight grids were created for sub-sampling technique one and two. Each grid cell
was selected only once within the entire set of 8 grids. Technique I randomly selected one grid
to sub-sample a photo (total of 8 grid cells) while technique two used two grids (total of 16
unique grid cells).

43

 

 

 

2000 - y = 869.02x + 45.798
1750 ~ ° r 2 = 0.796
,3 O
.5
'8
3
.0
E
3
2
U 0.5 l 1.5 2

Area of White Wax (cm?)

Figure 2.4: Index of abundance relating number of beech scales to area (cm2) of white waxy
substance secreted by the insects. Data were collected between May and October.

44

Quantitative Assessment of Beech Scale

(Log Proportion Infested)

 

 

I

I l_;*—Mean TI

1 ---cr- -- 25th Quantile;

“i' 0" ' — + — 75 Quantile
_+__ _*_L _, Wrench , __-__IL, _ _*u, ______ _-T_.a_.-______, .___,
0 1 2 3 4

Visual Assessment Ranking of Beech Scale

Figure,2.5: Relationship between qualitative estimates of beech scale infestation based on
visual analysis and log quantitative measures as determined from binary threshold methods.

45

Introduction

Beech scale (Cryptococcusfagisuga Lind. )(Hemiptera: Coccidae), an invasive
forest insect native to western Asia and southeastern Europe (vaiazdowski et al. 2006),
has been invading North American beech forests since its introduction in 1890 (Ehrlich
1934). Beech scale in North America is host speciﬁc to American beech trees (F agus
grandifolia). Beech scale adults are univoltine and each individual can lay
approximately 50 eggs (Wainhouse and Gate 198 8). Populations are capable of a rapid
increase because all individuals are females and they reproduce parthenogenically
(Ehrlich 1934). Beech scales are mobile only as 1St instars, referred to as crawlers.
Crawlers are active from late summer to early fall, and must ﬁnd a suitable location on
the bole or large branches of a beech tree where they can insert their stylets and feed on
the sieve cells of the secondary phloem (Ehrlich 1934; Telewski 2009). During
dispersal, crawlers sustain a high mortality estimated at 86% (Wainhouse and Gate
1988). Once feeding begins beech scale molt to the second instar and are immobile for
the remainder of their lives (Ehrlich 1934). Scales secrete a waxy substance over

themselves as they feed for protection from weather and predators (Ehrlich 1934). Even
with the wax, scales are subject to mortality when air temperatures drop below -3 7° C (-
35° F) and they are not protected by snow (Houston and O’Brien 1983). Several native

predators such as the twice-stabbed ladybird beetle (Chilocorus stigma Say),
cecidomyid ﬂies (Lestidiplosis sp.) and gall gnats (Diptera: Cecidomyidae,

Lestodiplosis spp.) feed on beech scale (Wainhouse and Gate 1988; Houston 1994).

46

Beech scale establishment represents the ﬁrst stage of beech bark disease
(BBD), referred to as the advancing ﬁ'ont. The second phase involves the infection of
trees by Neonectria spp. fungi (Ehrlich 1934; Houston 1994; Castlebury et al. 2006).
The minute wounds created when beech scales insert their stylets permit entry of the
fungi (Ehrlich 1934). The fungus invades and kills living tissues of the bark, cortex,
phloem, cambium, and sapwood (Ehrlich 1934). As lesions coalesce the tree looses
abilities to transport and store nutrients, which kills or reduces the health of the tree
(Ehrlich 1934). More than 50% mortality of trees over 25 cm in diameter may occur
(Houston 1994), altering forest composition and causing economic loss (Houston 1994).

Although beech scale was ﬁrst described in Eur0pe in the 18403 (Ehrlich 1934),
little is known about change in beech scale populations over time, especially on
American beech trees. This is surprising given the fundamental importance of
understanding the population dynamics of invasive species. The few studies that have
addressed beech scale dynamics relied on visual qualitative assessments of local beech
density on individual trees (Houston et al. 1979; Wainhouse and Gate 1988; Gora et a1.
1994; Houston 1994). Although this method is a quick and simple way of assessing
scale density, it is also subjective and yields coarse results. Also, all but one of these
studies were conducted in European beech stands, where beech scale are suspected to
interact with hosts differently than in American beech stands (Houston 1994). In
addition, studies by Gora et al. (1994) and Wainhouse and Gate (1988) assessed scale
density at ﬁve year intervals but did not monitor within year and annual population
changes. Houston et al. (1979) assessed annual change in beech scale densities for three

years, but only in a plantation of European beech trees. Houston (1994) compared

47

annual change of beech scale density on American beech trees in two different stages of
BBD. Houston (1994) found the stands newly infested with beech scale to generally
increase in scale density annually, while the long-infested stands sustained more
constant, lower beech scale densities.

We analyzed changes in beech scale densities on American beech trees in
Michigan using both qualitative and quantitative methods. We observed within-year
changes in scale density to understand the rate at which scale populations increase on
individual trees and whether detectable scale mortality occurs. Effects of tree and plot
characteristics on annual changes in populations of beech scale were evaluated at the

tree and plot levels.

Methods

Site Establishment and Sampling Techniques

We establishedia ﬁxed-radius plot, eight meters in diameter, in 26 locations
during July and August of 2007 (Figure 3.1). Plots were widely dispersed across the
distribution of beech scale at that time (Chapter 1). Plots were selected within each of
ﬁve satellite populations documented by Schwalm (2009), but the Upper Peninsula
(U .P.) was split into two regions due to the large size of this satellite population. Six
plots were located in the eastern U.P. (E.U.P.), four in the central U.P. (C.U.P.), four in
the Emmet satellite, three in the Ludington satellite, two in the Benzie satellite and
seven in the Cadillac satellite (Figure 3.1). Most plots were established in infested
areas, but two plots were uninfested, being 200 m from the nearest infested tree at the

time of establishment to observe spread from nearby populations.

48

All beech trees over 6 cm in diameter at breast height (DBH) were tagged to
ensure unique identiﬁcation in subsequent visits. Species and DBH were recorded for
all trees within the plot to allow for calculations of basal area. Individual plots
contained 2 to 22 beech trees, with a mean number of 9.92 (i 0.78) beech trees per plot.
When less then eight beech trees were present in a plot, the beech trees closest to the
plot were tagged until at least eight trees were included. There were two plots with low
beech density, where only ﬁve and seven beech trees were tagged.

Qualitative visual estimates and digital photography methods (Chapter 2) were
both used to assess beech scale density. Qualitative estimates were used to document
beech scale on entire trees. As described in Chapter 2, we used a four point qualitative
ranking system using reference cards, which helped produce consistent rankings.
Qualitative ranks were performed by walking around the tree searching for signs of
beech scale (i.e. white wax). Scale density was ranked as 0 = no beech scale, 1 = trace
populations, 2 = patchy populations and 3= heavy populations. The method focused on
the readily visible portion of the tree trunk, < 4 m aboveground. If only upper portions
of the trunk or branches were visibly infested, density was recorded in a similar manner.
The digital photography method was used to quantitatively monitor changes of beech
scale density within designated areas on each tree. Three unique photographs were
taken of each tree, using a stratiﬁed random sampling design to assign photo aspects at
selected heights. One photo was taken at each 0.9, 1.2 and 1.5 m above ground and
each of the three photos was assigned a random aspect before arriving at the site.
Photos were taken using a tripod with a built-in stabilizer to ensure the camera was the

same distance from the tree (30 cm) for each photograph. To reduce potential

49

problems, a duplicate photo was taken at each setting. If shadows were present within
the photo area, sunlight was blocked using a clipboard to reduce contrast within the
photo.
Site Revisits

After all plots were established, sites were revisited to assess change in beech
scale densities. During each revisit, qualitative and quantitative measures of beech
scale were recorded. All photos were retaken on the same trees and at the same heights
and aspects of original visits, allowing direct comparison of changes in beech scale
density at various levels (i.e. individual photo, tree, and plot). Annual change in scale
density involved a revisit after one year (i 30 days) and was observed at three sample
time ﬂames including July 2007- 2008, October 2007-2008, and March 2008-2009
(Table 3.1). In addition to annual revisits, a subset of 14 plots was revisited ten times to
document within-year changes in beech scale density from July 2007 to June 2009
(Table 3.1). Due to a snow storm, only half of the plots were accessible in December
2007, but the remaining plots were revisited in early January 2008 (Table 3.1).
Population Density Over Time

As described in Chapter 2, we created an index of scale abundance that showed
a linear relationship between the number of beech scales and the percentage of bark
covered with wax. This allowed us to quantify scale density using photographs of
beech scale wax. We used the binary threshold technique described in Chapter 2 to
analyze photos. ImageJ was used to apply a binary threshold, which identiﬁed pixels
with beech scale wax based on pixel brightness. The software was also used to

calculate the percentage of photographed bark that was infested. Thresholds were

50

manually set for each photograph, depending on photo quality, photo brightness and
photo contrast.

SAS 9.1 ® was used to analyze within-year and annual changes of scale
densities using both photo data and qualitative data. One tree ﬁ'om the Emmet satellite
was removed from the dataset because it had beech scale densities approximately 36
standard deviations above the mean. In addition, data from Cadillac in December 2007
and from Emmet in October 2007 and March 2009 were removed because less than 75
% of trees were sampled due to undesirable weather, and disproportional sampling of
uninfested and infested trees. Analyses of scale density at the tree level were conducted
by averaging the scale density of all photos from the same tree and with the qualitative
assessments. Fisher’s Exact test was used to determine if the proportions of trees with
increasing, decreasing and constant densities of beech scale differed among regions. A
general linear model was used to determine the relationship between tree DBH, initial
scale abundance and region with annual change of beech scale density on individual
trees. Annual change in beech scale density was calculated based on a tree’s average
scale density from the annual revisit (Year 2) minus the tree’s mean scale density from
the initial visit (Year 1). Similarly, a general linear model was used to determine the
signiﬁcance of sight variables including beech tree basal area, other tree basal area,
average beech tree DBH, average initial scale abundance and region for explaining the
annual change of beech scale density averaged for each plot.

The rate of change in population density was represented by dividing Year 1
population densities by Year 2 population densities. This ratio allowed us to minimize

the number of zero values in our dataset by assigning a 1 (no change) to trees with no

51

scale in Year 1 and Year 2. More speciﬁcally, there were only two trees where Year 2

was a zero and could not be included in the ratio.

Results
Seasonal Cycle of Beech Scale Density on Individual Trees

Beech scale density ﬂuctuated between July 2007 and June 2009. Overall mean
beech scale density decreased from 0.282% of bark infested in July 2007 to 0.129% in
June 2009. Between July and October of both years, the mean infested area increased
(Figure 3.2). Area infested decreased ﬁom October 2007 to March 2008, remained
relatively stable from March 2008 to June 2008, this showed a slight decrease from
June 2008 to July 2008. Population density remained stable ﬁ'om October 2008 to June
2009 (Figure 3.2).

Patterns of within year change in beech scale populations depended on region
and time (year and month) of sampling (p < 0.001) (Figure 3.2). The Cadillac and
Ludington regions had similar initial beech scale density, with 0.323% and 0.545 % of
bark infested, respectively. These sites also had similar trends in scale density
throughout the study (Figure 3.2). Scale density in both regions was constant between
July and October 2007 then decreased from October 2007 to March 2008. Both
populations remained constant until July 2008, where a slight decrease occurred (Figure
3.2). In October 2008, both populations showed slight increases. In March 2009, the
Ludington population decreased slightly while Cadillac remained constant. Densities at
both sites remained constant from March to June 2009.

The Emmet region started with a much lower mean scale density than Ludington

and Cadillac, with an average of 0.003% of sampled bark infested. The within year

52

trend of beech scale density also differed from the other regions (Figure 3.2). From
July 2007 to January 2008, the density increased, and decreased in density between
January and March 2008. The density then remained constant until it slightly decreased
in July 2008. The population increased substantially (four fold increase) ﬂom July to
October 2008. The population increased again from December 2008 to June 2009,
where a mean of 0.033% bark was infested, which was about 11-fold higher than at the
beginning of the study.
Annual Beech Scale Dynamics
Photo Analysis of Scale Density on Individual Trees

When digital photos were used to analyze scale densities on individual trees,
trends in beech scale density over time differed signiﬁcantly by region and sample
period (July 2007-2008, October 2007-2008, and March 2008-2009) (p < 0.001) (Table
3.2). Scale density increased across years in the Cadillac region during March and July
sample periods, but density decreased from October 2007 to 2008. During the March
sampling period, scale density increased on 44 % of trees, while density on 36 % of
trees decreased. In contrast, scale density on a majority of trees decreased during July
and October sample periods (Table 3.2). Mean scale density increased in the Emmet
region during all three annual sample time ﬂames. Scale density on 39 to 80% of trees
increased, while scale density only decreased on 3 to 22 % of trees. In contrast, the
Ludington region showed an overall decrease in scale density during every sample
period. In this region, density on 0-8% of trees increased while density on 92-96% of
trees decreased. The regions in the Upper Peninsula (U .P.) were only sampled annually

during July, and showed an overall increase in beech scale density from 2007 to 2008.

53

In the eastern and central U.P., scale density increased on 67% and 74% of trees,
respectively and decreased on 33% and 21% of trees, respectively.

Analysis of scale densities on individual trees using a general linear model
produced a signiﬁcant result (p < 0.001), explaining approximately 24% of the variation
in annual change (Table 3.3). Changes in beech scale abundance were affected by
region (p < 0.001), initial scale abundance (p = 0.01), and the interaction of region and
initial scale abundance (p = 0.033). The interaction reﬂected a positive relationship
between change in scale densities and initial scale abundance in the Emmet population,
while all other regions showed a negative relationship. DBH was not related to change
in scale density and was not included in the model.

Qualitative Analysis of Scale Density on Individual Trees

In most cases, qualitative estimates were similar to quantitative estimates, but
some differences occurred (Table 3.2; Figure 3.3). A majority of tree samples (64%)
resulted in constant qualitative rankings (Table 3.4). Of the samples that did change,
96% changed one visual rank, while the remaining four percent changed two ranks
(Table 3.4). In addition, 81% of samples that changed increased in scale density, while
only 19% decreased (Table 3.4). On several occasions (i.e. 324 of the 3,072 samples),
the qualitative technique showed beech scale presence when the quantitative rank did
not.

Trends from the qualitative assessment of beech scale density for each of the
three sample periods, March 2008 to 2009, October 2007 to 2008 and July 2007 to 2008
were similar, but changes in scale density varied among regions (p < 0.001) (Table 3.2;

Table 3.4). In the Cadillac region, the qualitative scale density remained the same for

54

72 to 86% of trees, increased for 10 to 25% of trees, and decreased for 3 to 4% of trees.
The net result of these changes was a slight overall increase in qualitative ranks for
scale density. Scale density in the Emmet region remained consistent on 29 to 69% of
trees, increased on 39 to 71% of trees, and decreased on only 3% of the trees. In the
E.U.P region, 58% of trees had no change in scale density but density increased on
42%, resulting in an overall increase in scale ranking. In the C .U.P region, we observed
constant rankings on 24% of trees while density on 76% of trees increased. Trends in
the Ludington region differed ﬂom other regions (p < 0.001); a nearly equal number of
trees increased and decreased in scale density (Table 3.2; Table 3.4). Qualitative
rankings for 50 to 78% of trees remained constant.

Plot Level Analysis of Beech Scale

The two plots that were initially uninfested in 2007 remained uninfested in 2008
despite the presence of infested trees within 200 m. These plots were not included in
further analyses.

Overall, the plot level general linear model was signiﬁcant (p = 0.002) and
explained approximately 81% of the variation in annual change of beech scale density.
Changes in beech scale populations varied among regions (p = 0.008) (Table 3.5).
Scale density changed more dramatically in the two regions in the Upper Peninsula than
in the other regions. Interactions of regions and other variables were not used as
independent variables due to the low number of samples per region. Initial scale
abundance and the log transformation of beech tree basal area and mean DBH were not
signiﬁcant (Table 3.5). Basal area of all non-beech trees was explored as an

independent variable but showed no relationship with change in beech scale density. A

55

new model was created without regions to focus on speciﬁc biological factors affecting
change in beech scale densities. This model was signiﬁcant (p = 0.04), explaining 41%
of variation in annual change of beech scale density. The log transformation of beech
basal area was the only signiﬁcant variable (p = 0.01), while initial scale abundance and
mean DBH of trees (log transformed) were not signiﬁcant (p = 0.74 and p = 0.17,
respectively). To reduce complexity, a ﬁnal model was selected with beech basal area
as the only signiﬁcant variable (p = 0.01). The model explained 32% of variation in

annual change of beech scale density (Table 3.5).

Discussion

Although our study design and sampling techniques differ from previous
studies of beech scale dynamics, our results showed similar trends. Houston (1994)
used a 40 point qualitative rank system to compare annual change of beech scale density
in a stand of newly infested American beech trees (ﬁrst decade of infestation) and a
stand where scale populations were present for over 50 years. He found that the newly
infested site generally increased in scale density annually over a ten year period, while
the long-affected stand had scale densities that remained lower and ﬂuctuated less than
those in the newly infested stand. Our data shows similar results, where the more
recently infested regions, Emmet, C.U.P., and E.U.P. showed large increases in scale
density, while the other regions showed lower and more stable densities of beech scale
(Table 2). Wainhouse and Gate (1988) used a 6 point qualitative rank system in their
study of beech scale populations over a seven year period. Their results showed that
beech scale populations on most European beech trees remained constant and those

populations that did change showed a relatively small difference in density. Gora et al.

56

(1994) also looked at long term changes in beech scale densities using a qualiative ﬁve
point ranking system (i.e. uninfested < very slight < slight < medium < severely). When
control trees were observed, a decrease of 96% of trees classiﬁed as very slight, medium
and no infestation occurred. A large increase in trees classiﬁed as slight infestations
also occurred (127 %). Our qualitative assessments over a two year period also showed
most populations of beech scale remained relatively constant from year to year. Also,
similar to Gora et a1. (1994), our results showed a large percentage of trees that were
uninfested or had a trace of scale in the ﬁrst year increased in scale density in the
second year. Houston et al. (1979) used a ﬁve point qualitative ranking system to
estimate scale density over three years following the inoculation of 40 of 80 European
beech trees in a plantation. Once infested, a majority of trees increased one ranking per
year. Of the trees that did not increase in scale density, most retained a constant rank
while one tree actually went ﬂom a small infestation to being uninfested. The year to
year results were similar to our ﬁndings; the trees in the Emmet, C.U.P. and E.U.P
regions, all started with low beech scale densities. Scale density on most trees in these
areas remained constant or increased one rank after one year. Also similar to Houston
et al. (1979), only two of the 275 trees we sampled were initially infested and became
uninfested after a year. This indicates that once infested with beech scale, a tree is
likely to remain infested.

Our sample design and sampling techniques allowed us to analyze with-in year
population dynamics of beech scale. The patterns we observed (Figure 3.2) appear to

be largely driven by beech scale reproduction. Adults lay eggs and then die in the early

summer (Ehrlich 1934). Their offspring do not secrete waxy coats until becoming 2"d

57

instars in late summer to early fall. Scales remain dormant through winter until early
summer, when they begin feeding again (Ehrlich 1934). These events should cause the
amount of wax generated by actively feeding to decrease during late summer and
increase during fall. Consistent with this, data ﬂom our photos showed a slight
decrease in scale density ﬂom June to July and a general increase in density during
October sampling (Figure 3.2). During the winter of 2007-2008, beech scale density
estimated by wax abundance decreased, suggesting that mortality occurred. Mortality
during these months is most likely ﬂom exogenous stresses such as weather. From
March 2008 to May 2008, a substantial increase in scale density occurred in the Emmet
region. This trend is unexplained by scale biology, although this result may be
inﬂuenced by scales losing wax during their dormant winter stage and replenishing wax
in the spring after breaking dormancy.

In addition to with-in year data, our technique of using digital photography to
measure scale density allowed us to more precisely quantify beech scale densities rather
than qualitative measures used in previous studies. Unlike visual assessments, digital
photography provided a lasting record of beech scale which allowed for more precise
comparisons of infestation over time. In addition, digital photography provided
concrete and reproducible results, unlike the subjective results of visual assessments.
These qualities allowed us to produce reliable with-in year estimates of scale density
and investigate more subtle changes in beech scale densities.

Our study is also the ﬁrst to sample beech scale dynamics in several regions.
Past studies have suggested beech scale dynamics are dependent on the stage of BBD

(Shigo 1972; Houston 1994). These studies proposed that in early stages, when only

58

beech scale is present, the populations of scale can reach higher densities. Once the
Neonectria fungus is established, the phloem dies and scale density decreases. Our
results suggest this may have occurred in our study in one region. Ludington has been
infested with scale and fungi since at least 2000 (O’Brien et al. 2001) and overall this
region showed decreases in annual scale density. Our other sample regions were in
areas where beech scale has more recently invaded (Schwalm 2009), and all of these
populations demonstrated annual increases in density. Different regions also have
different exogenous factors acting on scale populations. For example, weather patterns
and populations of insect predators may differ between regions. Gora et a1. (1994) also
suggested scale populations are sensitive to environmental conditions such as stand
light and climate.

The use of both qualitative and quantitative methods of estimating beech scale
density was valuable for detecting change in beech scale populations over time. In most
cases, the directionality of change in mean scale density of photos was similar to that of
qualitative mean scale density (Table 3.2), although the October sample period in
Cadillac showed a decrease in scale density with photos and an increase in scale density
with qualitative ratings. This difference is likely caused by the different precisions of
the two methods. The quantitative method was more precise so it detected changes in
beech scale densities that were missed with qualitative sampling (Figure 3.3). As a
result, the percentage of trees with constant beech scale density was higher when
reviewing qualitative results than quantitative (Figure 3.3). In contrast, qualitative

methods enabled us to sample the entire visible portion of the tree, increasing the

59

likelihood of detecting low densities or scattered scales, while photos sub-sampled only
the truck and could fail to capture low densities of scale.

Our data suggest that density dependence may occur in beech scale populations.
In particular, general linear models showed initial scale abundance to be a signiﬁcant
predictor variable of change in beech scale density. Also, on several occasions the
digital photographs yielded an overall decrease in population density ﬂom Year 1, yet
results ﬂom the qualitative method showed scale density on the same tree increased.
These situations show that the beech scale densities in speciﬁc locations of a tree may
decrease even though the tree’s overall level of infestation increases. Such scenarios
could be caused by density dependence, if crawlers move to areas of bark (not in the
photograph) where scale density was lower. Another likely explanation for such
scenarios involves Neonectria killing tree tissue in photographed areas, causing scale
mortality. In addition, it is possible that disproportional predation of scale occurred on
areas photographed and areas not photographed.

Our study increased knowledge of beech scale population dynamics, but several
questions still need to be addressed. A similar study conducted over a longer timeﬂarne
would allow comparison of within year and annual trends of beech scale populations.
Both qualitative and quantitative methods of estimating scale populations are useﬁrl to
ensure trends on both individual trees and plots are captured. In addition, several plots
within infested and uninfested areas of a single stand should be monitored over time to

assess stand level dynamics of beech scale.

60

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E o m e o o A 88 95..
2 o m m o o k 88 goes.
3 o m e o o a 88 .3900
S o m e o o k 88 2....
E o m e o o A 88 25..
E o m e o o s 88 was.
3 o m e o o k 88 8::
m o o e o o F 88 .5352.
w o m o o o m 38 .3588
E o m e o o k 38 .398
E o n v o o A B8 22.
. 33:8 50> 55.?
E o m e o o 8 88-38 .398
8 N m e e e A 88-38 22.
2 o m m o o 8 88-88 seas.
assessing.
.38 8:00 8.9.53 .eEEm .aj .n_.= .880 22:80
Eammm
85 a

 

 

80% one 5.8:. =0>_m 05 E @3639. 03 32m we 03:5: 05 8 80%: 36:30 80% 55:» “0:13.30 em 3 .80.» 0:0 coca
US$32 03 mac—a me 098:: 05 8 80m?— bmmeoe 13:5. .088.“ 0:5 wEEEsm :80 Mateo 5&2 :80 E BE.» $230 .3552 "ﬁn 033.

61

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm 0 o.. wow... mum... oo >30 00...u00
mm m 9. wow». .. mew. .. om £0.05. 00.....00
0:30:00 0000.000 0000.0... N 30> .. 30> : 0E. ... 003,00
«0003.. 0:00.00. 0:00.00 .
00.30 0:0—000002 .0:0.> 9:0: 3.03.. 0.000 .0:::<
o mm no 2.0 H moo... med Nmmd mm as... mam
m ..N 05 0N... H {0.0 Nmod mwod 0» >3... .50
n mo 0 00.. H ummd mmwd Sod KN 500.00 039.53..
0 oo o 00.8 H vode mood ommd mN >50 coat—=0...—
o No 0 00.0 H 50d owod NoNd mN £0.05. 039.53..
5. m on 00.0 H wNNd mood ..ood on 500.00 «oEEm
mm NN mm 0... H mwoN mood mood on >3... «05:5
9.0. v o... 00.0 H wand wood wood mN £0.05. aoEEm
5 m0 0m 8 .0 H BERN 0..de Nmmd mm .3200 00.5.00
or on .... 00.. H End oomd vmmd oo 3:... 00....000
cm on 00 . .N .0 H mom... .th mm ..d Nu £0.05. 00.....00
«00.0—.00 0000.000 0000.0:— 0300 N 30> .. 30> : 0...... 00301
0.00.0... 300.00. 0000.00.
8.2... 5.; 0305000 E80: 280 .852

 

 

 

02.000. 00.0 0.0 9:000:00 0:00.05

00:00:30 :. 0m:0£0 0: 0:0 0000.000 300.0... 00:00:30 w:.000.000 A8020... 0:020... 00:00:30 0.000 £000.. w:.000.0:. £005 :0.w0. 0£.

:. 000.. .0 0w0.:00.0n. 0£... .N.00>\.30> :. .9050 0.000 03000.00. 0.3.. 0.355.. .0003 0.0.00m. u A.Dm. 0:0 0.00500“. .0003 .0b:00 n .50
.0..0.>0. :. 00.. .00 30:00 0.000 £000.. :00... 03000.00. N 30> .0..0.> .0...:. :. 00.. .00 .5050 0.000 £000.. :00... 03000.00. . 30> .: u 0......00 .00
000.: .0 3.0.57. .wooN 0. nooN .3800 0:0 .wooN 0. hooN >0... .QooN 0. wooN £0.03. w:.0:.0:. 050.02.... w:..:E00 0. 0.0.0. 0E... 0:23.000:
>>00£ Hm 0:0 0:33.000. 30.0.. n N 05.3.2.2. 000.. n . 6.000 £000.. 0: H o 0.03 030800000 .0005 ..:0...000000 .0003 :00... 0.03

N .00 > 0:0 . 30> 0308000000 .000... .0 n. .00.... 0.. x03 .3 00.0>00 0.3.. .0 X. :00... 3000.00. N 30> 0:0 . .00 > 0.0.0.03 0.0%. .0 0. 08000:...
0.0500 0:0 :0.w0. .3 00.0.0000 50:00 0.000 £000.. .0 308000000 .000.> 0:0 00.00%. 00.0 $0 0.0.0.0:0 0.0%. .0:::0 80... 0.0.03. "Nd 0.00...

62

 

 

 

 

 

 

 

 

 

 

 

 

 

89o 8N Son? 538. v 2.053. 13:...

Sod v 3mm 38¢; F9089 v 2293.

So 8d Rmovmv Rmowwr P .525

882 wﬁmmomm m5 .otm

Sod v m9 BmNSN more x: o. .235.

u. A E o=_a> “— meuzwm mohmzum E239."— ouScw
:35. be 8295

 

 

dew“: 28 038 3:5 scion 55885 $3.93 2: cog—oi
$on Amid Ea $.30 .aonEEJ aBEEm 62—an dd vows—2: 89$ 22on 3E .Emﬁwouozm Sow “$562388 bane—V 28m _ 80>

33 3:5 .bmmcov 03% mo owﬁﬁo 5 count? mo .x. vm b03535“? 8:398 98 Good v & Enomemm 33 E55 2: .mnaﬁmowona E
35¢.on ma Ammm n 5 $3 .3232: :0 23m :83 E omega E023 5033 355322 manna—«>0 316.2 .258 Ron: .8250 "wen «Zak

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 O «V tum 8: mm a... .Sm
«N o 2 :5 «mod 8 3.. .50
on mm a :3 83 a 58% .6352...
S 3 3 mid 83 K 34 zen—.53
2 mm 0 $3 Ed mm 652 55:83
mm o E SE 39o mm .3200 552m
8 m 2. v2.0 $3 3 5. 355
B o 8 E: 83 8 news. 35.5
E m on a? 23 2 3&8 8:58

 

 

E8 «a 03¢

 

63

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 o o o n s F o o o F. s
2 NF 0 o N m v F o o N m
a A a A
F NF NF o F a a F 2 m o F a w
o o o o o w J o o a F a a 1
:r U Ir
F a x. F e u.
«cam 23% N 30> «cam 28% N 30>
5.: £28m c d: .228 3
NF NF F o N. s F o o o m s
m ON F F N m m o o o N m
o A a A
o v NF o F a w 0 NF 8 F F u n
o o F o c w 1 o N 8 2 c e 1
tr U Ir
F c x. F a x.
«tom 23% m 30> «cut 038 ~ 36>
.3552: 6 seem. a
NN n F o N s 8 FF N o F. s
m on o o N m 8 8 F F N m
o A a A
o NF 8 o F a m N 8 FmF F F a n
o 0 NF 8 o w 1 o N No 8 c a 1
Ir U ‘
F a x. F o x.
«cam 23% N Fug «cut 23% N .39
8:88 3 2.29: __< a

£53225: ~98..— um we: 255225: 383 n N £53230: comb n _ .038 :82. o: n a 20:? Diana 23m
.3 mac—:8 03:32.3: Gas—203 @8632 ca: @3on EEE some 5.3 $05 3:239: mo Fen—.5: was—.8952 .3052: own—Eu 3:52. 36 03.;

64

Table 3.5: General linear model results evaluating relationship between percentage change in
beech scale at the plot level (n = 19), as measured by photographs. Model 1 was significant (p
= 0.003) and explained approximately 81 % of variation in change of scale density. Model 2
was signiﬁcant (p = 0.01) and explained approximately 32% of variation in change of scale
density. Initial was the mean scale density measurements of each plot for Year 1 (calculated
from photographs), other BA was the basal area of all trees except beech in Year 1, and beech
BA was the basal area of all beech trees. The model also included the two—way interaction
between initial scale and region. Five regions were included (i.e. Cadillac, Emmet, Ludington,
C.U.P, and E.U.P.).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model 1

Degrees of Mean
Source Freedom Coefﬁcient Squares F Value Pr > F
Model 652,922 8.02 0.0115
Error 1 1 81,384
Log(10) Beech
Basal Area 1 -352.0 28214 0.81 0.388
Initial 1 0.24 0.2695 0.00 0.998
Logﬂ 0) DBH 1 315.0 28528 0.82 0.385
Locations 4 206462 5.92 0.009
Model 2

Degrees of Mean
Source Freedom Coefﬁcient Squares F Value Pr > F
Model 1 652922 8.02 0.01 15
Error 17 81384
Log(10) Beech
Basal Area 1 -1215.0 652922 8.02 0.0115

 

 

 

 

 

 

65

 

 

 

Revisited Sites .
0 Within Year Sites . Q

A Annual Sites ' ' i 3’ P

 

 

 

Regions

1. Central Upper Peninsula
2. Eastern Upper Peninsula
3. Emmet

4. Cadillac

5. Ludington

 

6. Benzie
— I m J
0 35 70 140 210

Figure 3.1: Plot locations by region used for bimonthly and annual sampling of scale density.
Bimonthly plots (n=14) were used to determine bimonthly and annual changes in scale density,
while annual plots (n=12) were only used in annual analysis.

66

 

   
 

 

g l +—---+.~1o
=.' §+—-—+\’\..‘.
§\ ~ —-
'5 —' 3 -“"=1—.‘ .‘—-‘ﬁ
3 0.1 +¥x .’/’+- .
in ,,
~‘-’ 4’
E
N .
2 '.
< 0.01 *1
\° -I-Cadillac
G
+Emmet
- . - Ludington
ooo1‘lLlllffflllll—lTTlTT yyﬁwlr
%s s as ﬁsh gag a 3°) 5°
\ C? “Q. o ’0 \ ('3' 0
5° 0 Qw3tb$ $30030 0 g 30°
Time

Figure 3.2: Within year change in beech scale densities by region ﬁ‘om July 2007 to June
2009. Percent area infested refers to average pencent of bark infested with scale per tree.
Cadillac (n = 84 trees), Emmet (n = 36 trees), Ludington (n = 28 trees).

67

 

 

a l
O
5 J
1: 3|
o O
o 4“ i
g
E 0* i l i i
a O
.E
O 4
E“ i
5
'6 a z
3
E
<
-12 - w _, ——%——~ ‘~——--#-~ **~——
-3 -2 -1 o 1 2 3

Annual change in qualitative scale density

Figure 3.3: Comparison of visual qualitative assessment and quantitative assessment of change
in beech scale density per tree. Change in qualitative scale density was a difference of visual
ranking from Year 1 to Year 2. Photos on each tree were averaged for each year and then the
difference between Year 1 and Year 2 was used to represent quantitative scale density.

68

APPENDICES

69

APPENDIX A: Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Appendix A : Site Coordinates
Site Date Longitude Latitude Scale County
1 1 7/1/2007 -86.51 18 43.639 Whitewashed OCEANA
25 8/8/2005 -86.4083 43.617 No Scale OCEANA
25 6/26/2007 -86.4083 43.617 Trace OCEANA
31 8/8/2005 -86.3826 43.6453 No Scale OCEANA
37 8/8/2005 -86.492 43.5413 No Scale OCEANA
40 8/8/2005 -86.4158 43.5448 No Scale OCEANA
43 8/8/2005 -86.4617 43.5312 No Scale OCEANA
44 7/10/2006 -86.4523 43.5449 Whitewashed OCEANA
49 8/8/2005 -86.3777 43.583 No Scale OCEANA
60 7/10/2006 -86.3745 43.7685 Trace OCEANA
76 8/9/2005 -86.3571 43.5101 No Scale OCEANA
76 7/10/2006 -86.3571 43.5101 No Scale OCEANA
83 8/9/2005 -86.3382 43.7979 No Scale OCEANA
83 7/10/2006 -86.3382 43.7979 No Scale OCEANA
115 5/30/2009 -85.7991 44.2012 No Scale WEXFORD
122 5/30/2009 -86.0143 44.0816 No Scale LAKE
123 6/19/2007 -86.1202 44.2697 No Scale MANISTEE
128 5/17/2005 -86.4127 43.6797 No Scale OCEANA
129 5/17/2005 -86.3694 43.6591 No Scale OCEANA
130 5/19/2005 -84.4771 42.7177 No Scale INGHAM
131 5/19/2005 -84.5112 42.6892 No Scale INGHAM
132 5/19/2005 -84.5912 42.6092 No Scale INGHAM
133 5/19/2005 -84.7582 42.5779 No Scale EATON
134 5/19/2005 -84.7598 42.7594 No Scale EATON
135 5/23/2005 -84.4714 42.5307 No Beech INGHAM
136 5/23/2005 -84.3637 42.5267 No Beech INGHAM
137 5/23/2005 -84.279 42.5964 No Beech INGHAM
138 5/23/2005 -84.3692 42.7054 No Scale INGHAM
139 5/23/2005 -84.4076 42.7552 No Beech INGHAM
140 5/23/2005 -84.3895 42.8125 No Beech CLINTON
141 5124/2005 -86.2667 43.1315 No Scale MUSKEGON
141 5/22/2008 -86.2667 43.1315 No Scale MUSKEGON
141 5/29/2009 -86.2667 43.1315 No Scale MUSKEGON
142 5/24/2005 -86.3586 43.2633 No Scale MUSKEGON
142 5/22/2008 -86.3586 43.2633 No Scale MUSKEGON
142 5/22/2008 -86.3586 43.2633 No Scale MUSKEGON
142 5/29/2009 -86.3586 43.2633 No Scale MUSKEGON
143 5/24/2005 -86.3961 43.3447 No Beech MUSKEGON
144 5/25/2005 -86.2662 43.4546 No Scale MUSKEGON
144 6/18/2007 -86.2653 43.4573 No Scale MUSKEGON

 

 

70

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
145 5125/2005 -86.3595 43.4854 No Beech OCEANA
146 5125/2005 -86.4395 43.5306 No Beech OCEANA
147 5125/2005 -86.4377 43.5348 No Scale OCEANA
148 5125/2005 -86.3589 43.5438 No Scale OCEANA
148 7110/2006 -86.3589 43.5438 Trace OCEANA
149 5125/2005 -86.4311 43.615 No Scale OCEANA
150 5125/2005 -86.4683 43.6144 No Scale OCEANA
150 7110/2006 -86.4683 43.6144 No Scale OCEANA
151 5125/2005 -86.4971 43.6608 Trace OCEANA
152 5125/2005 -86.4981 43.649 No Scale OCEANA
152 7110/2006 -86.4981 43.649 Trace OCEANA
153 5125/2005 —86.483 43.6641 Patchy OCEANA
154 5/2512005 -86.4681 43.6676 Whitewashed OCEANA
155 5125/2005 -86.4544 43.6963 No Scale OCEANA
156 5125/2005 -86.4874 43.6915 No Scale OCEANA
157 5125/2005 -86.4713 43.7335 No Beech OCEANA
158 5126/2005 -86.4963 44.0401 Whitewashed MASON
159 5126/2005 -86.4632 43.9931 Whitewashed MASON
160 5126/2005 -86.4583 43.9713 Whitewashed MASON
161 5126/2005 -86.3985 43.9447 No Beech MASON
162 5126/2005 -86.2849 43.8906 No Beech MASON
163 5126/2005 -86.3317 43.8765 Whitewashed MASON
164 5126/2005 -86.3683 43.8767 Patchy MASON
165 5126/2005 -86.4 43.8688 Whitewashed MASON
166 5126/2005 -86.3928 43.6972 Patchy OCEANA
167 5126/2005 -86.3715 43.4958 No Scale OCEANA
168 5131/2005 -86.2143 43.7588 No Scale OCEANA
169 5131/2005 -86.1286 43.7721 No Scale OCEANA
169 6119/2007 -86.1286 43.7721 Patchy OCEANA
170 5131/2005 -86.0871 43.8027 No Scale OCEANA
170 6119/2007 .86.0871 43.8027 No Scale OCEANA
170 5125/2008 -86.0871 43.8027 Trace OCEANA
171 5131/2005 -86.0773 43.8481 No Beech MASON
172 5131/2005 -86.1025 43.8719 No Scale MASON
173 5131/2005 -86.1107 43.8741 No Scale MASON
173 7113/2006 -86.1107 43.8741 No Scale MASON
173 5125/2008 -86.1 1 07 43.8741 Patchy MASON
174 5131/2005 -86.135 43.8737 No Scale MASON
175 5131/2005 -86.1909 43.8747 No Scale MASON
176 5131/2005 -86.2315 43.8754 Patchy MASON
177 5131/2005 -86.2277 43.8743 Patchy MASON
178 5131/2005 -86.2298 43.8899 No Scale MASON

 

 

 

 

 

 

71

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County

179 5131/2005 -86.2278 43.89 Patchy MASON

180 5131/2005 -86.2185 43.9045 Patchy MASON
181 5131/2005 -86.1033 43.9754 Patchy MASON

182 611/2005 -85.8945 44.3473 No Scale MANISTEE
182 61112009 -85.8945 44.3473 Whitewashed MANISTEE
183 611/2005 -85.9973 44.4387 No Scale MANISTEE
183 61112009 -85.9973 44.4387 Trace MANISTEE
184 611/2005 -86.2434 44.4792 No Scale MANISTEE
185 611/2005 -86.2304 44.4306 No Scale MANISTEE
186 611/2005 -86.2259 44.4037 Trace MANISTEE
187 61112005 -86.2274 44.4009 Whitewashed MANISTEE
188 61112005 861965 44.389 Patchy MANISTEE
189 61112005 -86.1648 44.3885 Trace MANISTEE
190 611/2005 -86.1673 44.4031 No Scale MANISTEE
191 611/2005 -86.1944 44.4064 Trace MANISTEE
192 61112005 -86.1451 44.3589 No Scale MANISTEE
193 611/2005 -86.125 44.3876 No Scale MANISTEE
194 611/2005 -86.1224 44.3714 No Scale MANISTEE
195 611/2005 -86.1477 44.3298 No Scale MANISTEE
196 61112005 -86.1727 44.3091 Whitewashed MANISTEE
197 612/2005 -86.1677 44.3158 Patchy MANISTEE
198 612/2005 -86.1622 44.3401 No Scale MANISTEE
199 612/2005 -86.1558 44.3813 No Scale MANISTEE
200 7114/2005 -86.0096 46.666 Patchy ALGER

208 7125/2006 -86.5464 46.4193 No Scale ALGER

214 7117/2009 -86.7166 46.3078 No Scale ALGER

222 6128/2006 -86.2718 46.5025 Patchy SCHOOLCRAFT
223 6128/2006 -86.3618 46.4282 No Scale SCHOOLCRAFT
223 ‘7123/2007 -86.3618 46.4284 Whitewashed SCHOOLCRAFT
224 7126/2006 -86.2914 46.3717 Trace SCHOOLCRAFT
224 712/2007 —86.2914 46.3717 Trace SCHOOLCRAFT
228 7126/2006 -86.2595 46.2777 No Scale SCHOOLCRAFT
229 7126/2006 —86.2299 46.2206 No Scale SCHOOLCRAFT
231 7126/2006 -86.228 46.1251 No Scale SCHOOLCRAFT
231 713/2007 -86.228 46.1251 No Scale SCHOOLCRAFT
231 7117/2009 -86.228 46.1251 No Scale SCHOOLCRAFT
236 71312007 -86.0826 46.0431 No Scale SCHOOLCRAFT
236 7120/2008 -86.0826 46.0431 No Scale SCHOOLCRAFT
236 7/1 912009 -86.0826 46.0431 Trace SCHOOLCRAFT
245 6127/2006 -85.9283 46.1858 No Scale SCHOOLCRAFT
257 711 812009 -86.8372 46.1419 No Scale DELTA

500 612/2005 -86.172 44.3585 No Scale MANISTEE

 

 

 

 

 

 

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Appendix A : Site Coordinates

Site 7 Date Longitude Latitude Scale County
501 612/2005 -86.2025 44.3497 No Scale MANISTEE
502 612/2005 -86.1979 44.3315 Patchy MANISTEE
503 61212005 -86. 1879 44.3313 No Scale MANISTEE
504 612/2005 -86. 1921 44.3168 Whitewashed MAN I STEE
505 612/2005 -86.1816 44.2627 No Scale MANISTEE
506 612/2005 -86. 1773 44.2671 Trace MANISTEE
507 616/2005 -86. 3585 43.8165 Trace 00 EANA
508 616/2005 -86.2982 43.7984 No Beech OCEANA
509 61612 005 -86.2922 43. 7877 Trace 0C EANA
510 616/2005 -86.2689 43.8175 No Scale OCEANA
511 616/2005 -86.2482 43.8721 No Scale MASON
512 616/2005 -86.2014 43.8389 No Scale MASON
512 7/1 312006 -86.2014 43. 8389 Trace MASON
513 616/2005 -86.1755 43.8357 No Scale MASON
513 7113/2006 -86.1755 43.8357 Trace MASON
514 6/6/2005 -86.1617 43.837 Whitewashed MASON
515 616/2005 -86.1749 43.7907 No Scale OCEANA
515 7110/2006 -86.1749 43.7907 No Scale OCEANA
516 616/2005 -86.2449 43.7934 Trace OCEANA
517 617/2005 -86.1531 43.9113 No Beech MASON
518 617/2005 -86.1634 43.9141 Trace MASON
519 617/2005 -86.1907 43.9038 Trace MASON
520 617/2005 -86.1751 43.9021 Trace MASON
521 617/2005 -86.05 43.9376 No Scale MASON
521 512512008 -86.05 43.9376 No Scale MASON
522 617/2005 -86.0546 43.9848 No Scale MASON
522 5125/2008 -86.0546 43.9848 No Scale MASON
523 617/2005 -86.0801 44.0178 Trace MASON
524 617/2005 —86.1159 44.0163 Trace MASON
525 617/2005 -86.0772 44.0322 No Scale MASON
525 6/1 912007 -86.0772 44.0322 Whitewashed MASON
526 617/2005 -86. 1204 44.041 8 Whitewashed MASON
527 617 12 005 -86. 1 324 44.0663 Trace MASON
528 617/2005 -86.1254 44.1049 Trace MASON
529 617/2005 -86.2211 44.1457 Trace MASON
530 617/2005 -86.2007 44.2509 Patchy MANISTEE
531 617/2005 -86. 1033 44.1707 No Scale MANISTEE
531 6/1 912007 -86. 1033 44.1707 No Scale MANISTEE
531 5126/2008 -86. 1033 44.1707 No Scale MANISTEE
532 618/2005 -86.141 1 44.3129 Trace MANISTEE
533 618/2005 -86.1023 44.305 Trace MANISTEE
534 618/2005 -86.2683 44.345 Patchy MANISTEE

 

 

 

 

 

 

73

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
535 61812005 -86.1054 44.5358 No Scale BENZIE
535 6127/2007 -86.1054 44.5358 No Scale BENZIE
535 611 812008 -86.1054 44.5358 No Scale BENZIE
535 611 112009 -86.1054 44.5358 Trace BENZIE
536 618/2005 -86.1024 44.5901 Trace BENZIE
537 618/2005 -86.1136 44.6833 No Scale BENZIE
538 618/2005 -86.0622 44.7207 No Scale BENZIE
538 6111/2009 -86.0622 44.7207 No Scale BENZIE
539 618/2005 -86.0746 44.764 No Scale BENZIE
539 6111/2009 -86.0746 44.764 No Scale BENZIE
540 618/2005 -86.0359 44.8459 No Scale LEELANAU
540 711712006 -86.0359 44.8459 No Scale LEELANAU
540 611 812008 -86.0359 44.8459 No Scale LEELANAU
540 611 112009 -86.0359 44.8459 No Scale LEELANAU
541 61812005 -86.021 44.897 No Scale LEELANAU
541 7/1 012007 -86.0198 44.897 No Scale LEELANAU
541 611 112009 -86.021 44.897 No Scale LEELANAU
542 618/2005 -85.9252 44.9354 No Scale LEELANAU
542 7117/2006 -85.9252 44.9354 No Scale LEELANAU
542 6111/2009 —85.9252 44.9354 No Scale LEELANAU
543 61812005 -85.9147 44.8771 No Scale LEELANAU
543 6126/2007 -85.9147 44.8771 No Scale LEELANAU
544 618/2005 -85.9698 44.8428 No Scale LEELANAU
545 619/2005 -85.9991 44.7575 No Scale BENZIE
545 6111/2009 -85.9991 44.7575 No Scale BENZIE
546 619/2005 -85.9572 44.8073 No Scale LEELANAU
547 619/2005 -85.8857 44.7182 No Scale BENZIE
548 619/2005 -85.9796 44.6449 No Scale BENZIE
548 6127/2007 -85.9796 44.6449 No Scale BENZIE
548 611 812008 -85.9796 44.6449 No Scale BENZIE
549 619/2005 -85.9579 44.6801 No Scale BENZIE
550 619/2005 -85.9091 44.6171 No Scale BENZIE
551 619/2005 -86.0463 44.6175 No Scale BENZIE
552 61912005 -86.2307 44.6958 No Scale BENZIE
553 619/2005 -86.13 44.531 Trace BENZIE
554 619/2005 -85.9588 44.5255 Patchy BENZIE
555 6113/2005 -84.356 44.5362 No Beech OSCODA
556 6113/2005 -84.5993 44.9391 No Scale OTSEGO
557 611 312005 -84.447 44.9712 No Scale OTSEGO
558 6113/2005 -84.5902 45.0157 No Scale OTSEGO
559 6113/2005 -84.3713 44.9602 No Scale MONTMORENCY
560 6/1 312005 -84.6779 44.8802 No Scale OTSEGO

 

 

 

 

 

 

74

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
560 611212009 -84.6779 44.8802 Trace OTSEGO
561 6113/2005 -84.1987 44.8682 No Scale MONTMORENCY
562 6114/2005 -85.8552 44.8104 No Scale LEELANAU
GRAND
563 6114/2005 -85.5197 44.5122 No Scale TRAVERSE
GRAND
563 611212009 -85.5197 44.5122 No Scale TRAVERSE
564 6114/2005 —85.8537 44.8788 No Scale LEELANAU
565 6114/2005 -85.6081 44.4971 No Scale WEXFORD
566 6114/2005 -85.8644 44.922 No Scale LEELANAU
566 7117/2006 -85.8644 44.922 No Scale LEELANAU
GRAND
567 6114/2005 —85.6775 44.5467 No Scale TRAVERSE
568 6114/2005 -85.7447 44.8916 No Scale LEELANAU
569 6114/2005 -85.8185 44.5473 No Scale BENZIE
570 6114/2005 -85.727 44.9822 No Scale LEELANAU
GRAND
571 6114/2005 -85.7979 44.6045 No Scale TRAVERSE
GRAND
571 611212009 -85.7979 44.6045 Trace TRAVERSE
572 6114/2005 -85.7599 44.9944 No Scale LEELANAU
573 6114/2005 -85.8419 44.6781 No Scale BENZIE
574 6114/2005 -85.7764 44.9805 No Scale LEELANAU
575 6114/2005 -85.8584 44.7399 No Scale BENZIE
575 611212009 -85.8584 44.7399 No Scale BENZIE
576 6114/2005 -85.7964 44.959 No Scale LEELANAU
GRAND
577 6114/2005 -85.6759 44.6681 No Scale TRAVERSE
578 6114/2005 -85.6331 45.0063 No Scale LEELANAU
GRAND
579 6114/2005 -85.599 44.6355 No Scale TRAVERSE
580 6114/2005 -85.6436 45.1028 No Scale LEELANAU
GRAND
581 6114/2005 -85.5567 44.6278 No Scale TRAVERSE
GRAND
581 611212009 -85.5567 44.6278 No Scale TRAVERSE
582 611 412005 -85.6745 44.8873 No Scale LEELANAU
GRAND
583 6114/2005 -85.4943 44.7124 No Scale TRAVERSE
584 6114/2005 -85.6522 44.8025 No Scale LEELANAU
585 611 512005 -85.423 44.8818 No Scale ANTRlM
GRAND
586 6/1512005 -85.4031 44.763 No Scale TRAVERSE
587 6/1 512005 -85.3532 44.8632 No Scale ANTRIM
GRAND
588 6115/2005 -85.3457 44.5878 No Scale TRAVERSE

 

 

 

 

 

 

75

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
589 611 512005 -85.1893 44.8577 No Scale KALKASKA
590 6115/2005 -85.3167 44.5267 No Scale KALKASKA
591 6115/2005 -85.1347 44.9898 No Scale ANTRIM
592 611 512005 -85.178 44.5839 No Scale KALKASKA
593 6115/2005 -85.0506 44.8394 No Scale KALKASKA
594 611 512005 -85. 1434 44.641 No Scale KALKASKA
595 611 512005 -85.1933 44.7925 No Scale KALKASKA
596 6115/2005 -85.1964 44.6549 No Scale KALKASKA
597 6115/2005 -85.2683 44.7829 No Scale KALKASKA
597 611212009 -85.2683 44.7829 No Scale KALKASKA
598 6115/2005 -85.1545 44.67 No Scale KALKASKA
599 611 512005 85.1347 44.9612 No Scale ANTRIM
600 6115/2005 -85.0738 44.6987 No Scale KALKASKA
601 6115/2005 -85.0714 44.7282 No Scale KALKASKA
602 611 512005 -85.2118 44.5726 No Scale KALKASKA
603 611 512005 -85.0727 44.7286 No Scale KALKASKA
604 6116/2005 -84.6458 44.8014 No Scale CRAWFORD
605 6116/2005 -84.6467 44.6805 No Beech CRAWFORD
606 6116/2005 -84.7067 44.9582 No Scale OTSEGO
606 611 312009 -84.7067 44.9582 No Scale OTSEGO
607 6116/2005 -84.7771 44.7539 No Scale CRAWFORD
608 6116/2005 -84.7767 44.9211 No Scale OTSEGO
609 6116/2005 -84.8845 44.7709 No Scale KALKASKA
609 611212009 -84.8845 44.7709 No Scale KALKASKA
610 611 612005 -84.7874 44.888 No Scale OTSEGO
611 611 612005 -84.976 44.7134 No Scale KALKASKA
61 1 612312008 -84.976 44.7134 No Scale KALKASKA
611 611212009 -84.976 44.7134 No Scale KALKASKA
612 6116/2005 -84.7811 44.7769 No Scale CRAWFORD
613 6116/2005 -84.9144 44.6353 No Scale KALKASKA
614 612012005 -84.7514 45.1867 No Scale CHARLEVOIX
615 6116/2005 -84.9558 44.5399 No Scale KALKASKA
616 612012005 84.9352 45.1467 No Scale CHARLEVOIX
617 6116/2005 -84.7657 44.5981 No Scale CRAWFORD
618 612012005 -84.9074 45.0646 No Scale ANTRIM
619 612012005 -84.5911 45.2052 No Scale CHEBOYGAN
620 612012005 -84.8164 45.113 No Scale OTSEGO
620 6/1 312009 -84.8164 45.113 No Scale OTSEGO
621 612012005 -84.6669 45.2444 No Scale CHEBOYGAN
621 6121/2008 -84.6669 45.2444 No Scale CHEBOYGAN
621 611 312009 -84.6669 45.2444 Trace CHEBOYGAN
622 612012005 -84.7556 45.0276 No Scale OTSEGO

 

 

 

 

 

 

76

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
623 612012005 -84.9356 45.2778 No Scale CHARLEVOIX
624 612012005 -84.699 45.0996 No Scale OTSEGO
625 612012005 -84.8166 45.3524 No Scale EMMET
626 612012005 -84.6235 45.0618 No Scale OTSEGO
627 612012005 -84.7418 45.3263 No Scale EMMET
628 6121/2005 -84.5762 45.505 No Scale CHEBOYGAN
629 6121/2005 -84.7666 45.4417 No Scale EMMET
630 6121/2005 -84.7159 45.5509 No Scale CHEBOYGAN
630 612012008 -84.7159 45.5509 No Scale CHEBOYGAN
630 611412009 -84.71 59 45.5509 Patchy CHEBOYGAN
631 6121/2005 -84.762 45.5161 No Scale EMMET
632 6121/2005 -84.6316 45.5749 No Scale CHEBOYGAN
632 611412009 -84.6316 45.5749 No Scale CHEBOYGAN
633 6121/2005 -84.7925 45.6287 No Scale EMMET
633 6125/2007 -84.7925 45.6287 Patchy EMMET
634 6121/2005 -84.7282 45.6916 No Scale CHEBOYGAN
634 6125/2007 -84.7282 45.6916 No Scale CHEBOYGAN
634 612012008 —84.7282 45.6916 No Scale CHEBOYGAN
634 6115/2009 -84.7282 45.6916 Patchy CHEBOYGAN
635 6121/2005 -84.7728 45.7192 No Scale EMMET
635 712712006 -84.7728 45.7192 No Scale EMMET
635 6125/2007 -84.7728 45.7192 No Scale EMMET
635 611 512009 -84.7728 45.7192 No Scale EMMET
636 6121/2005 -84.6681 45.7463 No Scale CHEBOYGAN
636 712712006 -84.6681 45.7463 No Scale CHEBOYGAN
637 6121/2005 -84.7695 45.7668 Whitewashed EMMET
638 6121/2005 -84.6507 45.6832 No Scale CHEBOYGAN
638 611 512009 -84.6507 45.6832 Whitewashed CHEBOYGAN
640 6121/2005 —84.6386 45.6646 No Scale CHEBOYGAN
640 6115/2009 -84.6386 45.6646 Patchy CHEBOYGAN
641 6121/2005 -85.0146 45.6507 No Scale EMMET
641 6125/2007 -85.0146 45.6507 Trace EMMET
642 6121/2005 -84.3282 45.5795 No Scale CHEBOYGAN
642 611 512009 -84.3282 45.5795 Trace CHEBOYGAN
643 6121/2005 -85.0851 45.6066 No Scale EMMET
643 6125/2007 -85.0851 45.6066 Trace EMMET
644 6121/2005 -84.3963 45.5387 No Scale CHEBOYGAN
645 6121/2005 -85.0161 45.5511 No Scale EMMET
645 712712006 -85.0161 45.551 1 Trace EMMET
646 6121/2005 84.4542 45.5692 No Scale CHEBOYGAN
647 6121/2005 -84.9358 45.5503 No Scale EMMET
647 6125/2007 -84.9358 45.5503 No Scale EMMET

 

 

 

 

 

 

77

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Appendix A : Site Coordinates

Site Date Longitude Latitude Scale County
648 6121/2005 -84.508 45.319 No Scale CHEBOYGAN
649 612112005 -84.9296 45.6138 No Scale EMMET
650 612112005 -84.4773 45.2542 No Scale CHEBOYGAN
650 611 512009 -84.4773 45.2542 No Scale CHEBOYGAN
651 612112005 -84.8495 45.6367 Trace EMMET
652 6122/2005 -85.1664 45.3495 No Scale CHARLEVOIX
653 612112005 -84.8457 45.5513 No Scale EMMET
654 6122/2005 -85.1781 45.3109 No Scale CHARLEVOIX
655 612112005 -84.8522 45.4711 No Scale EMMET
656 6122/2005 -85.0586 45.3123 No Scale CHARLEVOIX
657 612112005 -84.927 45.4568 No Scale EMMET
658 6122/2005 85.0429 45.2419 No Scale CHARLEVOIX
658 611 312009 -85.0429 45.2419 No Scale CHARLEVOIX
659 612112005 -84.7996 45.377 No Scale EMMET
660 6122/2005 -85.0933 45.2373 No Scale CHARLEVOIX
661 612112005 -84.9063 45.4078 No Scale EMMET
661 6125/2007 -84.9063 45.4078 Patchy EMMET
662 6122/2005 -85.131 45.1718 No Scale CHARLEVOIX
663 6122/2005 -85.2265 45.1685 No Scale ANTRIM
664 6122/2005 -85.0401 45.1282 No Scale CHARLEVOIX
665 6122/2005 -85.2017 45.2155 No Scale CHARLEVOIX
666 6122/2005 -85.154 45.0499 No Scale ANTRIM
666 6113/2009 -85.154 45.0499 No Scale ANTRIM
667 6122/2005 -85.2546 45.2793 No Scale CHARLEVOIX
668 6122/2005 —85.1387 45.0907 No Scale ANTRIM
669 6122/2005 -85.3218 45.2556 No Scale CHARLEVOIX
670 6122/2005 -84.5121 45.1591 No Scale OTSEGO
670 612112008 -84.5121 45.1591 No Scale OTSEGO
670 6115/2009 -84.5121 45.1591 No Scale OTSEGO
671 6122/2005 -85.3766 45.1822 No Scale ANTRIM
672 6123/2005 -84.4175 45. 1 562 No Scale OTSEGO
672 611 512009 -84.4175 45.1562 No Scale OTSEGO
673 6122/2005 -85.3612 45.0757 No Scale ANTRIM
674 6128/2005 -85.3596 44.2518 No Scale WEXFORD
675 6122/2005 -85.2998 45.1593 No Scale ANTRIM
676 6128/2005 -85.3245 44.1487 No Scale OSCEOLA
677 6122/2005 -85.2994 45.0487 No Scale ANTRIM
678 6128/2005 -85.3825 44.0734 No Scale OSCEOLA
678 5131/2009 -85.3825 44.0734 No Scale OSCEOLA
679 6122/2005 -85.2172 45.0297 No Scale ANTRIM
680 6128/2005 -85.1871 44.1 101 No Scale OSCEOLA
681 6122/2005 -85.1446 45.103 No Scale ANTRIM

 

 

 

 

 

 

78

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
682 612912005 -85.5007 44.2179 No Scale WEXFORD
682 5131/2009 -85.5007 44.2179 Patchy WEXFORD
683 612312005 -84.4708 45.4356 No Scale CHEBOYGAN
684 612912005 -85.6039 44.2225 No Scale WEXFORD
684 815/2007 -85.6035 44.2227 Trace WEXFORD
685 612312005 -84.4319 45.3439 No Scale CHEBOYGAN
686 612912005 -85.7037 44.2464 No Scale WEXFORD
686 7112/2006 -85.7037 44.2464 Trace WEXFORD
687 612312005 -84.2249 45.4358 No Scale PRESQUE ISLE
687 611 512009 84.2249 45.4358 No Scale PRESQUE ISLE
688 612912005 -85.7081 44.1856 No Scale WEXFORD
689 612312005 -84.1525 45.3567 No Scale PRESQUE ISLE
690 612912005 -85.5945 44.1322 No Scale LAKE

690 611 712008 -85.5945 44.1322 No Scale LAKE

691 612312005 -84.3077 45.274 No Scale CHEBOYGAN
691 6115/2009 -84.3077 45.274 No Scale CHEBOYGAN
692 612912005 -85.5031 44.0558 No Scale OSCEOLA
693 6127/2005 -84.9725 44.4634 No Scale MISSAUKEE
694 612912005 -85.6089 44.2658 No Scale WEXFORD
695 6127/2005 -84.9941 44.3058 No Scale MISSAUKEE
696 612912005 -85.6197 44.2759 No Scale WEXFORD
696 7112/2006 -85.6197 44.2759 Trace WEXFORD
697 6127/2005 -84.8695 44.0693 No Scale CLARE
698 612912005 -85.6403 44.2807 Trace WEXFORD
699 6127/2005 -84.6929 44.1067 No Scale CLARE

700 612912005 -85.6415 44.2841 Trace WEXFORD
701 6128/2005 -85.4071 44.3305 No Scale WEXFORD
702 612912005 -85.6099 44.3291 Trace WEXFORD
703 6128/2005 -85.4121 44.454 No Scale WEXFORD
704 6130/2005 -84.147 44.4457 No Scale OGEMAW
705 6128/2005 -85.3733 44.4252 No Scale WEXFORD
705 611/2009 -85.3733 44.4252 No Scale WEXFORD
706 6130/2005 -84.3072 44.2722 No Scale OGEMAW
707 6128/2005 -85.2401 44.4542 No Scale MISSAUKEE
707 611/2009 -85.2401 44.4542 No Scale MISSAUKEE
708 715/2005 -86.4477 46.2917 No Scale SCHOOLCRAFT
708 7123/2007 -86.4482 46.2915 Trace SCHOOLCRAFT
709 6128/2005 -85.3112 44.3599 No Scale MISSAUKEE
709 6117/2008 -85.3112 44.3599 No Scale MISSAUKEE
709 611/2009 -85.3112 44.3599 Trace MISSAUKEE
710 71612005 86.4682 46.0681 No Scale SCHOOLCRAFT
711 6128/2005 -85.3362 44.2602 No Scale WEXFORD

 

 

 

 

 

 

79

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County

711 5131/2009 -85.3362 44.2602 No Scale WEXFORD
712 716/2005 -86.364 45.9671 No Scale SCHOOLCRAFT
712 7126/2006 -86.364 45.9671 No Scale SCHOOLCRAFT
713 612812005 -85.135 44.295 No Scale MISSAUKEE
714 716/2005 -86.5997 45.7672 No Scale DELTA

715 612812005 -85.0749 44.445 No Scale MISSAUKEE
715 611/2009 -85.0749 44.445 No Scale MISSAUKEE
716 716/2005 -86.6639 45.7006 No Scale DELTA

717 612812005 -85.1354 44.4518 No Scale MISSAUKEE
718 71612005 -86.4699 45.7857 No Scale DELTA

719 612812005 -85.2034 44.3674 No Scale MISSAUKEE
719 6/112009 -85.2034 44.3674 No Scale MISSAUKEE
720 716/2005 -86.3689 45.8402 No Scale SCHOOLCRAFT
721 612812005 -85.0349 44.3076 No Scale MISSAUKEE
721 5131/2009 -85.0349 44.3076 No Scale MISSAUKEE
722 716/2005 -86.1364 45.9813 No Scale SCHOOLCRAFT
722 7126/2006 -86.1364 45.9813 Whitewashed SCHOOLCRAFT
723 612912005 -85.4894 44.3035 No Scale WEXFORD
724 716/2005 -86.0585 46.0717 No Scale SCHOOLCRAFT
724 7126/2006 -86.0585 46.0717 No Scale SCHOOLCRAFT
725 612912005 -85.536 44.3781 No Scale WEXFORD
726 7112/2005 -85.6004 46.5893 Whitewashed LUCE

727 612912005 -85.6191 44.4065 No Scale WEXFORD
727 611/2009 -85.6191 44.4065 No Scale WEXFORD
728 7112/2005 —85.745 46.6517 Whitewashed LUCE

729 612912005 -85.6975 44.4436 No Scale WEXFORD
730 7112/2005 -85.8316 46.6697 Whitewashed LUCE

731 612912005 -85.771 44.4691 No Scale WEXFORD
731 611/2009 -85.771 44.4691 No Scale WEXFORD
732 7112/2005 -85.929 46.6552 Whitewashed ALGER

733 612912005 -85.7388 44.3525 No Scale WEXFORD
733 7/1 112006 -85.7388 44.3525 Trace WEXFORD
734 7112/2005 -85.9279 46.165 Patchy SCHOOLCRAFT
735 612912005 -85.661 44.2515 Patchy WEXFORD
736 7112/2005 -85.7829 46.1008 Whitewashed MACKINAC
737 612912005 -85.6997 44.2512 No Scale WEXFORD
737 7129/2007 -85.7007 44.2511 No Scale WEXFORD
738 7112/2005 -85.6968 46.0355 No Scale MACKINAC
738 7126/2006 -85.6968 46.0355 Patchy MACKINAC
739 612912005 -85.6446 44.2517 No Scale WEXFORD
739 7112/2006 -85.6446 44.2517 Patchy WEXFORD
740 7112/2005 -85.572 46.2224 Whitewashed MACKINAC

 

 

 

 

 

 

80

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Appendix A : Site Coordinates

Site Date Longitude Latitude Scale County
741 6130/2005 -84.5205 44.0165 No Scale GLADWIN
742 7112/2005 -85.7831 46.3359 Whitewashed LUCE
743 716/2005 -86.5525 46.4673 No Scale ALGER
743 7125/2006 -86.5525 46.4673 No Scale ALGER
743 71212007 -86.5525 46.4673 No Scale ALGER
743 711 712008 —86.5525 46.4673 No Scale ALGER
744 7113/2005 -85.7073 46.4623 Whitewashed LUCE
745 71612005 -86.5744 46.4072 No Scale ALGER
746 7113/2005 -85.8014 46.4506 Whitewashed LUCE
747 71612005 -86.6157 46.3195 No Scale ALGER
749 716/2005 -86.6358 46.1574 No Scale DELTA
750 7114/2005 -86.3638 46.5573 No Scale ALGER
750 7124/2006 -86.3638 46.5573 Trace ALGER
751 716/2005 —86.5572 46.0739 No Scale DELTA
752 7112/2005 -85.9278 46.5641 Whitewashed ALGER
753 71612005 -86.8632 45.7883 No Scale DELTA
754 7112/2005 -84.254 42.5778 No Scale INGHAM
755 716/2005 -86.8571 46.036 No Scale DELTA
756 712012005 -87.3038 45.6258 No Scale DELTA
756 7118/2009 -87.3038 45.6258 No Scale DELTA
757 716/2005 -86.8376 46.1347 No Scale DELTA
757 7119/2008 -86.8376 46.1347 No Scale DELTA
758 712012005 -87.1111 45.8409 No Scale DELTA
758 7118/2009 -87.1111 45.8409 No Scale DELTA
759 716/2005 -87.3669 45.7718 No Scale MENOMINEE
759 7118/2009 -87.3669 45.7718 No Scale MENOMINEE
760 7125/2005 -86.7992 46.3333 No Scale ALGER
761 717/2005 -86.3635 46.5069 No Scale ALGER
761 7125/2006 -86.3635 46.5069 Trace ALGER
762 7126/2005 -85.5391 45.7479 Whitewashed CHARLEVOIX
763 717/2005 -86.2687 46.5208 Patchy ALGER
764 7126/2005 -85.5569 45.734 No Scale CHARLEVOIX
764 7118/2006 -85.5569 45.734 Trace CHARLEVOIX
765 717/2005 86.2525 46.5356 Whitewashed ALGER
766 7126/2005 -85.5643 45.7254 Whitewashed CHARLEVOIX
767 71712005 .86.1504 46.5524 No Scale ALGER

 

 

 

 

 

 

81

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
767 7125/2006 -86.1504 46.5524 Whitewashed ALGER
768 7126/2005 -85.5591 45.688 No Scale CHARLEVOIX
768 7118/2006 -85.5591 45.688 Trace CHARLEVOIX
769 717/2005 -86.0626 46.5475 Trace ALGER
770 712712005 -85.4914 45.6467 Patchy CHARLEVOIX
770 7118/2006 -85.4914 45.6467 Whitewashed CHARLEVOIX
771 71712005 -86.1709 46.4591 Trace SCHOOLCRAFT
772 712712005 -85.4963 45.606 No Scale CHARLEVOIX
772 7118/2006 -85.4963 45.606 No Scale CHARLEVOIX
773 717/2005 -86.1575 46.4194 No Scale SCHOOLCRAFT
773 7125/2006 -86.1575 46.4194 Whitewashed SCHOOLCRAFT
774 712712005 -85.5706 45.5755 No Scale CHARLEVOIX
774 7118/2006 -85.5706 45.5755 No Scale CHARLEVOIX
775 7/1 112005 -85.076 46.4314 Whitewashed CHIPPEWA
776 712712005 -85.5537 45.6597 No Scale CHARLEVOIX
777 711 112005 -85.1 153 46.6344 Whitewashed CHIPPEWA
778 712712005 -85.5794 45.6593 No Scale CHARLEVOIX
778 711 812006 -85.5794 45.6593 No Scale CHARLEVOIX
779 711 112005 -85.2524 46.5779 Whitewashed LUCE

780 712712005 -85.5831 45.6473 No Scale CHARLEVOIX
780 7118/2006 -85.5831 45.6473 No Scale CHARLEVOIX
781 7/1 112005 -85.3701 46.5537 Whitewashed LUCE

782 712712005 -85.5924 45.6087 No Scale CHARLEVOIX
782 7118/2006 -85.5924 45.6087 No Scale CHARLEVOIX
783 711 112005 -85.3076 46.6656 Whitewashed LUCE

784 712712005 -85.5965 45.5846 No Scale CHARLEVOIX
784 7118/2006 -85.5965 45.5846 No Scale CHARLEVOIX
785 7/1 112005 -85.5337 46.6572 Patchy LUCE

786 81112005 -86.1343 45.0035 No Scale LEELANAU
787 711 112005 -85.5973 46.454 Whitewashed LUCE

788 81212005 -85.986 45.1082 No Scale LEELANAU
789 7112/2005 -85.4269 46.4948 Whitewashed LUCE
790 813/2005 -86.0538 45.1199 No Scale LEELANAU
791 711 212005 -85.6342 46.4144 Whitewashed LUCE
792 813/2005 -86.0474 45.1381 No Scale LEELANAU
793 7112/2005 -85.1447 46.3496 Whitewashed CHIPPEWA
794 81312005 -86.0197 45.1429 No Scale LEELANAU
795 7112/2005 -85.2394 46.477 Whitewashed LUCE
797 7112/2005 -85.0123 46.2614 Whitewashed CHIPPEWA
798 813/2005 -85.9896 45.1228 No Scale LEELANAU
799 7112/2005 -85.1849 46.1745 Whitewashed MACKINAC
800 811/2005 -86.1395 45.0028 No Scale LEELANAU

 

 

 

 

 

 

82

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County

801 7112/2005 85.4465 46.1105 Whitewashed MACKINAC
802 811/2005 86.113 45.0057 No Scale LEELANAU
803 711 212005 85.1 123 46.0393 Patchy MAC Kl NAC
804 814/2005 85.0694 45.508 No Scale EMM ET

805 7112/2005 84.6688 46.4622 Trace CHIPPEWA
806 81412005 85.0154 45.5891 Trace EMMET

807 7114/2005 84.9256 45.9883 Whitewashed MACKINAC
808 812/2005 85.9805 45.1167 No Scale LEELANAU
809 7114/2005 84.8069 45.8928 No Scale MACKINAC
809 7124/2006 84.8069 45.8928 Patchy MACKINAC
810 812/2005 86.0062 45.0979 No Scale LEELANAU
811 7114/2005 84.7614 45.9665 No Scale MACKINAC
811 712712006 84.7614 45.9665 No Scale MACKINAC
81 1 71112007 84.7614 45.9665 Trace MACKINAC
812 813/2005 86.0592 45.111 No Scale LEELANAU
813 7125/2005 85.6046 44.2088 No Scale WEXFORD
813 711 212006 85.6046 44.2088 No Scale WEXFORD
814 813/2005 86.0474 45.0989 No Scale LEELANAU
815 7125/2005 85.6575 44.2028 No Scale WEXFORD
815 7112/2006 85.6575 44.2028 No Scale WEXFORD
815 815/2007 85.658 44.2029 No Scale WEXFORD
816 813/2005 86.0296 45.0809 No Scale LEELANAU
817 7125/2005 85.6405 44.206 No Scale WEXFORD
817 7112/2006 85.6405 44.206 No Scale WEXFORD
818 813/2005 86.0073 45.0734 No Scale LEELANAU
819 7125/2005 85.6233 44.237 No Scale WEXFORD
819 814/2007 85.6224 44.2367 Whitewashed WEXFORD
820 813/2005 85.9896 45.08 No Scale LEELANAU
821 7125/2005 85.6478 44.2404 Whitewashed WEXFORD
822 813/2005 85.9898 45.0909 No Scale LEELANAU
823 7125/2005 85.7075 44.2772 No Scale WEXFORD
823 7112/2006 85.7075 44.2772 No Scale WEXFORD
824 81312005 85.9851 45.1034 No Scale LEELANAU
825 7125/2005 85.6759 44.344 No Scale WEXFORD
825 711 112006 85.6759 44.344 Whitewashed WEXFORD
826 814/2005 85.067 45.4581 No Scale EMMET

827 7125/2005 85.6756 44.3441 Trace WEXFORD
828 814/2005 84.8929 45.6803 No Scale EMMET

828 712712006 84.8929 45.6803 No Scale EMMET

828 6125/2007 84.8929 45.6803 Trace EMMET

829 7125/2005 85.6164 44.3813 No Scale WEXFORD
829 7113/2006 85.6164 44.3813 No Scale WEXFORD

 

 

 

 

 

 

83

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
830 81412005 84.7741 45.6557 No Scale EMMET
831 7125/2005 85.567 44.3534 No Scale WEXFORD
831 6117/2008 85.567 44.3534 Trace WEXFORD
832 81812005 86.6256 41.8394 No Scale BERRIEN
833 7125/2005 85.5596 44.3035 No Scale WEXFORD
833 7112/2006 85.5596 44.3035 No Scale WEXFORD
834 818/2005 86.6018 41.9044 No Scale BERRIEN
835 8125/2005 85.6097 44.2227 Trace WEXFORD
836 81812005 86.3045 42.3306 No Scale VAN BUREN
837 81812005 86.3069 42.337 No Scale VAN BUREN
837 5128/2009 86.3069 42.337 No Scale VAN BUREN
838 818/2005 86.1972 42.703 No Scale ALLEGAN
838 5128/2009 86.1972 42.703 No Scale ALLEGAN
842 5129/2009 86.3573 43.5099 No Scale OCEANA
847 611/2005 86.2214 44.3962 No Scale MANISTEE
1000 51812006 84.5906 42.9144 No Beech CLINTON
1001 51812006 85.014 43.2902 No Scale MONTCALM
1002 51812006 85.0844 43.4275 No Scale MONTCALM
1003 51812006 85.4433 43.6048 No Scale MECOSTA
1004 51812006 85.2007 43.7035 No Scale MECOSTA
1005 51812006 85.2662 43.9706 No Scale OSCEOLA
1005 513112009 85.2662 43.9706 No Scale OSCEOLA
1006 51912006 85.7033 44.2468 No Scale WEXFORD
1007 51912006 85.6746 44.2481 Trace WEXFORD
1008 51912006 86.0371 44.1948 No Scale MANISTEE
1009 51912006 86.1783 44.2673 Whitewashed MANISTEE
1010 51912006 86.4185 44.1 1 18 Whitewashed MASON
101 1 51912006 86.4355 44.0855 Whitewashed MASON
1012 51912006 86.3872 44.0852 Whitewashed MASON
1013 51912006 86.3668 44.0921 Whitewashed MASON
1014 511 012006 86.4963 44.0425 Whitewashed MASON
1015 511 012006 86.4966 44.0447 Whitewashed MASON
1016 511 012006 86.5049 44.0378 Whitewashed MASON
1017 511 012006 86.5192 43.6482 Patchy OCEANA
1018 511 012006 86.3315 43.6301 Trace OCEANA
1019 511012006 86.2782 43.6147 Patchy OCEANA
1020 511012006 86.2664 43.5998 No Scale OCEANA
1020 611 812007 86.2664 43.5998 No Scale OCEANA
1020 5125/2008 86.2664 43.5998 No Scale OCEANA
1020 5129/2009 86.2664 43.5998 Trace OCEANA
1021 511 012006 86.1788 43.6206 Whitewashed OCEANA
1022 511 012006 86.1393 43.6728 No Scale OCEANA

 

 

 

 

 

 

84

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
1022 611 812007 86.1393 43.6728 No Scale OCEANA
1023 511012006 85.9112 43.7793 No Beech NEWAYGO
1024 5110/2006 85.8247 43.8499 No Beech LAKE
1025 511012006 85.7641 43.923 No Beech LAKE
1026 5110/2006 85.6431 43.8153 No Scale NEWAYGO
1026 513012009 85.6431 43.8153 No Scale NEWAYGO
1027 511 012006 85.6275 43.6638 No Scale NEWAYGO
1028 512312006 85.0139 43.408 No Beech MONTCALM
1029 512312006 84.8865 43.4607 No Scale MONTCALM
1030 5123/2006 84.9093 43.4808 No Scale ISABELLA
1031 512312006 85.0472 43.5246 No Scale ISABELLA
1032 5123/2006 85.0099 43.8427 No Scale CLARE
1033 512312006 85.4338 43.8948 No Scale OSCEOLA
1033 5131/2009 85.4338 43.8948 No Scale OSCEOLA
1034 512412006 85.8705 44.1405 No Beech LAKE
1035 512412006 86.0812 44.0804 No Scale MASON
1036 5124/2006 86.0156 44.0322 No Scale LAKE
1037 5124/2006 85.9852 43.8877 No Scale LAKE
1038 512412006 85.9705 43.8233 No Beech LAKE
1039 5124/2006 86.0188 43.7421 No Scale NEWAYGO
1039 611912007 86.0188 43.7421 No Scale NEWAYGO
1039 5125/2008 86.0188 43.7421 Whitewashed NEWAYGO
1040 512412006 86.1672 43.7318 No Scale OCEANA
1040 611912007 86.1672 43.7318 Trace OCEANA
1041 512412006 86.1263 43.6255 No Scale OCEANA
1041 611812007 86.1263 43.6255 No Scale OCEANA
1041 512412008 86.1263 43.6255 No Scale OCEANA
1042 5124/2006 86.153 43.5877 No Scale OCEANA
1042 6118/2007 86.153 43.5877 No Scale OCEANA
1043 512412006 86.1136 43.5322 No Scale OCEANA
1043 611812007 86.1136 43.5322 No Scale OCEANA
1043 512312008 86.1136 43.5322 No Scale OCEANA
1044 5124/2006 85.9 43.5883 No Scale NEWAYGO
1045 512512006 85.89 43.6691 No Beech NEWAYGO
1046 5125/2006 85.8128 43.6992 No Beech NEWAYGO
1047 512512006 85.7611 43.6107 No Scale NEWAYGO
1047 512312008 85.7611 43.6107 No Scale NEWAYGO
1048 512512006 85.6621 43.5588 No Scale NEWAYGO
1049 512512006 85.5228 43.558 No Scale MECOSTA
1050 5125/2006 85.4027 43.5156 No Scale MECOSTA
1051 5125/2006 85.3436 43.4323 No Scale MONTCALM
1052 5125/2006 85.5821 43.4589 No Beech NEWAYGO

 

 

 

 

 

 

85

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
1053 512512006 85.5353 43.3297 No Scale MONTCALM
1054 512512006 85.6973 43.2529 No Scale KENT
1055 5125/2006 85.7882 43.4802 No Scale NEWAYGO
1055 5123/2008 85.7882 43.4802 No Scale NEWAYGO
1056 5125/2006 86.0994 43.4318 No Scale MUSKEGON
1057 5125/2006 86.0823 43.3372 No Scale MUSKEGON
1058 512512006 86.189 43.3648 No Scale MUSKEGON
1058 5129/2009 86.189 43.3648 No Scale MUSKEGON
1059 512512006 86.0874 43.14 No Scale MUSKEGON
1059 5129/2009 86.0874 43.14 No Scale MUSKEGON
1060 5125/2006 86.1849 43.0019 No Scale OTTAWA
1061 5125/2006 86.1264 42.9355 No Scale OTTAWA
1062 512512006 86.1789 42.8752 No Scale OTTAWA
1063 5125/2006 85.8986 42.6472 No Beech ALLEGAN
1064 5125/2006 85.8988 42.636 No Scale ALLEGAN
1064 5128/2009 85.8988 42.636 No Scale ALLEGAN
1065 5126/2006 86.0105 42.4851 No Scale ALLEGAN
1066 5126/2006 85.9957 42.553 No Scale ALLEGAN
1067 512612006 85.4656 42.5947 No Scale BARRY
1068 512612006 85.4087 42.7855 No Beech KENT
1069 5126/2006 85.312 42.949 No Beech KENT
1070 5126/2006 85.4849 43.0436 No Beech KENT

1071 5126/2006 85.2832 43.1275 No Beech MONTCALM
1072 5126/2006 85.2558 43.2854 No Scale MONTCALM
1073 5129/2006 84.0553 45.3962 No Beech PRESQUE ISLE
1074 5129/2006 83.8863 45.4131 No Scale PRESQUE ISLE
1075 512912006 83.6219 45.2225 No Beech PRESQUE ISLE
1076 5129/2006 83.4835 45.2607 No Scale PRESQUE ISLE
1077 512912006 84.1211 45.1354 No Scale MONTMORENCY
1078 5129/2006 83.731 1 45.1762 No Scale ALPENA
1079 5129/2006 83.5633 45.0421 No Beech ALPENA
1080 5129/2006 83.617 44.9827 No Scale ALPENA
1081 5129/2006 83.6682 44.8986 No Beech ALPENA
1082 513012006 83.4451 44.8465 No Scale ALCONA
1083 513012006 83.404 44.7798 No Scale ALCONA
1084 513012006 83.4037 44.6838 No Scale ALCONA
1085 513012006 83.3682 44.5698 No Beech ALCONA
1086 513012006 83.592 44.5668 No Beech ALCONA
1087 513012006 83.666 44.6928 No Scale ALCONA
1088 513012006 83.7664 44.4563 No Beech IOSCO
1089 513012006 83.9449 44.4701 No Scale OGEMAW
1090 513012006 84.217 44.707 No Scale OSCODA

 

 

 

 

 

 

86

 

Aprgndix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
1091 513012006 84.1296 44.616 No Scale OSCODA
1092 513012006 83.9411 44.6323 No Beech OSCODA
1093 513012006 83.8321 44.7845 No Scale ALCONA
1094 513012006 84.0606 44.754 No Scale OSCODA
1095 513012006 84.0063 44.9572 No Beech MONTMORENCY
1096 513112006 84.738 45.2738 No Scale CHARLEVOIX
1097 513112006 84.7468 45.3132 No Scale EMMET
1098 513112006 84.8821 45.3103 No Scale EMMET
1099 5131/2006 84.8767 45.493 No Scale EMMET
1099 612012008 84.8767 45.493 Trace EMMET
1100 5131/2006 84.8559 45.5471 No Scale EMMET

1 101 615/2006 84.9134 45.9289 Whitewashed MAC KINAC
1102 61512006 84.9128 45.9286 No Scale MACKINAC
1 102 71112007 84.9128 45.9286 Patchy MACKINAC
1103 61512006 84.8983 45.9619 Patchy MACKINAC
1104 61512006 84.3697 46.086 No Scale MACKINAC
1105 61512006 84.1858 46.0036 No Scale CHIPPEWA
1106 61512006 84.0213 45.9618 No Scale CHIPPEWA
1106 712012009 84.0213 45.9618 Trace CHIPPEWA
1107 61612006 83.7247 46.0894 No Beech CHIPPEWA
1108 61612006 83.619 46.0724 No Beech CHIPPEWA
1109 61612006 83.6734 46.0312 No Scale CHIPPEWA
1110 61612006 83.7031 46.0131 No Scale CHIPPEWA
1110 7114/2008 83.7031 46.0131 No Scale CHIPPEWA
1110 712012009 83.7031 46.0131 Trace CHIPPEWA
1111 61612006 83.6675 45.9987 No Scale CHIPPEWA
1111 711412008 83.6675 45.9987 No Scale CHIPPEWA
1111 712012009 83.6675 45.9987 No Scale CHIPPEWA
1112 61612006 83.5455 45.9434 No Scale CHIPPEWA
1112 711412008 83.5455 45.9434 No Scale CHIPPEWA
1112 712012009 83.5455 45.9434 Trace CHIPPEWA
1113 61612006 83.5367 46.0024 No Scale CHIPPEWA
1113 711412008 83.5367 46.0024 No Scale CHIPPEWA
1113 712012009 83.5367 46.0024 No Scale CHIPPEWA
1114 61612006 83.6119 45.9599 No Scale CHIPPEWA
1114 711412008 83.6119 45.9599 Trace CHIPPEWA
1114 712012009 83.61 19 45.9599 Trace CHIPPEWA
1115 61612006 83.6973 45.9752 No Scale CHIPPEWA
1115 711412008 83.6973 45.9752 No Scale CHIPPEWA
1115 712012009 83.6973 45.9752 No Scale CHIPPEWA
1116 61612006 83.7876 45.9844 No Scale CHIPPEWA
1116 7114/2008 83.7876 45.9844 No Scale CHIPPEWA

 

87

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
1116 712012009 83.7876 45.9844 No Scale CHIPPEWA
1117 61712006 84.0726 45.9946 No Scale CHIPPEWA
1117 712012009 84.0726 45.9946 No Scale CHIPPEWA
1118 61712006 84.1751 46.159 No Beech CHIPPEWA
1119 61712006 84.2713 46.2129 No Beech CHIPPEWA
1120 61712006 84.3535 46.3758 No Beech CHIPPEWA
1121 61712006 84.7391 46.4096 No Scale CHIPPEWA
1 121 7124/2007 84.739 46.4102 Whitewashed CHIPPEWA
1 122 617/2006 84.6976 46.4318 Patchy CHIPPEWA
1 123 617/2006 84.6773 46.4507 Trace CHIPPEWA
1 124 61712006 84.9389 46.4037 Trace CHIPPEWA
1 125 61712006 84.9088 46.4229 Patchy CHIPPEWA
1126 61712006 84.8351 46.4135 Trace CHIPPEWA
1127 61712006 84.5917 46.3028 No Beech CHIPPEWA
1128 61712006 84.7871 46.1868 No Beech CHIPPEWA
1129 61812006 85.1466 46.0631 Whitewashed MACKINAC
1 130 61812006 85.0267 46.066 Trace MACKI NAC
1131 61812006 84.9189 46.0284 No Scale MACKINAC
1132 61812006 84.88 46.1023 No Scale MACKINAC
1133 61812006 84.7655 46.0736 No Scale MACKINAC
1 133 71112007 84.7655 46.0736 Trace MAC KINAC
1134 611212006 84.4092 43.2516 No Beech GRATIOT
1135 611212006 84.2099 43.9791 No Beech GLADWIN
1136 6113/2006 84.6053 45.8559 No Beech MACKINAC
1137 6113/2006 84.6241 45.8776 No Beech MACKINAC
1138 6113/2006 84.629 45.8795 No Scale MACKINAC
1 139 6113/2006 84.6242 45.871 1 Whitewashed MAC Kl NAC
1140 611 312006 84.6355 45.874 No Scale MACKINAC
1 141 6113/2006 84.6437 45.8729 Trace MAC KI NAC
1 142 611 312006 84.6453 45.8715 Patchy MAC KI NAC
1 143 611 312006 84.6462 45.8708 Whitewashed MAC KI NAC
1 144 6113/2006 84.645 45.8656 Whitewashed MAC KI NAC
1145 611312006 84.6376 45.8586 No Scale MACKINAC
1146 611312006 84.6335 45.8628 No Scale MACKINAC
1 147 611 312006 84.6254 45.8618 Patchy MAC KI NAC
1 148 611312006 84.61 45.8577 Trace MAC KINAC
1 149 611412006 84.571 45.809 Whitewashed MACKINAC
1 150 611412006 84.5364 45.7945 Trace MAC KINAC
1 151 611412006 84.5208 45.7906 Patchy MACKINAC
1 1 52 611412006 84.51 31 45.7707 Trace MAC KINAC
1153 611412006 84.4931 45.7524 No Scale MACKINAC
1154 611412006 84.3847 45.779 No Beech MACKINAC

 

 

 

 

 

 

88

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County

1 155 611412006 84.3859 45.7432 No Beech MAC Kl NAC
1 156 611412006 84.4244 45.7619 No Scale MACKINAC
1 157 611412006 84.4519 45.7728 No Scale MAC KI NAC
1158 612012006 86.1139 45.01 17 No Scale LEELANAU
1159 612012006 86.1138 45.0191 No Scale LEELANAU
1160 612012006 86.1 126 45.0327 No Scale LEELANAU

1 161 612012006 86.1 129 45.0342 No Beech LEELANAU
1162 612012006 86.1207 45.0255 No Scale LEELANAU
1 163 612112006 86.1399 44.9989 No Scale LEELANAU
1164 6122/2006 84.9847 45.5764 No Scale EMMET
1165 6122/2006 85.0572 45.5757 No Scale EMMET

1 165 6125/2007 85.0572 45.5757 No Scale EMM ET
1166 6126/2006 85.5945 46.0818 Patchy MACKINAC
1167 6126/2006 85.7001 45.9837 No Scale MACKINAC
1168 6127/2006 87.3756 45.4333 No Scale MENOMINEE
1169 6127/2006 87.7568 45.5153 No Beech MENOMINEE
1171 6127/2006 85.7952 46.1772 Whitewashed MACKINAC
1 172 612812006 86.6583 46.4368 No Scale ALGER

1 172 712/2007 86.6583 46.4368 No Scale ALGER

1 172 711 812008 86.6583 46.4368 Trace ALGER
1173 612812006 86.9282 46.396 No Scale ALGER

1 173 711 612009 86.9282 46.396 No Scale ALGER
1174 612812006 87.086 46.3813 No Scale ALGER
1174 711712009 87.086 46.3813 No Scale ALGER

1 175 71212 007 86.3618 46.4282 Whitewashed SCHOO LC RAFT
1 177 612812006 86.272 46.5022 Whitewashed SCHOOLC RAFT
1 178 612812006 86.2616 46.461 1 Whitewashed SCHOO LC RAFT
1 179 612812006 86.083 46.4267 No Beech SCHOOLCRAFT
1 180 612912006 84.8598 45.9459 Whitewashed MACKI NAC
1181 612912006 84.8413 45.6807 No Scale EMM ET

1 182 612912006 84.9101 45.7079 Trace EMMET

1 183 612912006 84.8683 45.7169 Patchy EMM ET

1 184 613012006 84.8905 44.2917 No Beech MISSAUKEE
1185 613012006 84.8722 44.205 No Beech MISSAUKEE
1186 612912006 84.9744 46.332 No Beech CHIPPEWA
1 187 71512006 84.9275 42.8204 No Beech IONIA

1 188 715/2006 85.1296 42.9363 No Scale IONIA
1189 71512006 85.7901 42.2984 No Scale VAN BUREN
1190 71512006 85.7695 41.9487 No Scale CASS

1 191 715/2006 85.3275 42.0861 No Beech KALAMAZOO
1192 71512006 84.4736 41.8333 No Beech HILLSDALE
1193 71612006 84.5867 43.6552 No Beech MIDLAND

 

89

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
1194 71612006 83.9089 43.6686 No Beech BAY
1195 71612006 83.3659 43.4599 No Beech TUSCOLA
1196 71612006 83.3789 43.1668 No Scale LAPEER ‘
1197 71612006 83.3462 42.9426 No Scale LAPEER
1198 71612006 83.5099 42.7943 No Scale OAKLAND
1199 71612006 83.5497 42.6447 No Scale OAKLAND
1200 71712006 83.5198 42.4317 No Scale WAYNE
1201 71712006 83.6958 41.8786 No Beech MONROE
1210 7110/2006 86.1789 43.7909 Trace OCEANA
121 1 711012006 86. 1 526 43.8253 Whitewashed MASON
1216 711 212006 85.6171 44.2372 Trace WEXFORD
1218 711212006 85.6227 44.2543 Patchy WEXFORD
1219 7112/2006 85.6035 44.2226 No Scale WEXFORD
1220 6117/2008 85.6046 44.2087 No Scale WEXFORD
1225 7112/2006 85.6878 44.2486 Trace WEXFORD
1226 711312006 85.8637 44.2841 No Scale MANISTEE
1226 612012007 85.8637 44.2841 No Scale MANISTEE
1226 5126/2008 85.8637 44.2841 No Scale MANISTEE
1227 711 312006 85.7065 44.3519 No Scale WEXFORD
1227 712812007 85.7064 44.3533 Whitewashed WEXFORD
1234 711 712006 85.8749 44.9176 No Scale LEELANAU
1234 611112009 85.8749 44.9176 No Scale LEELANAU
1236 711712006 85.3066 45.3077 No Scale CHARLEVOIX
1236 611912008 85.3066 45.3077 No Scale CHARLEVOIX
1237 711712006 85.31 1 1 45.3074 Whitewashed CHARLEVOIX
1237 611912008 85.31 1 1 45.3074 Trace CHARLEVOIX
1239 711812006 85.4914 45.6145 Whitewashed CHARLEVOIX
1247 711812006 85.539 45.7498 Whitewashed CHARLEVOIX
1249 711912006 85.5026 45.6942 No Beech CHARLEVOIX
1250 7119/2006 85.5138 45.6093 No Beech CHARLEVOIX
1251 7119/2006 85.5696 45.6339 No Scale CHARLEVOIX
1252 711912006 85.5279 45.6559 No Beech CHARLEVOIX
1257 7125/2006 86.4462 46.5399 Trace ALGER
1259 712512006 86.428 46.4745 No Scale ALGER
1259 711712008 86.428 46.4745 Trace ALGER
1260 7125/2006 86.1468 46.41 37 Whitewashed SCHOOLC RAFT
1262 712612006 86.2913 46.3716 No Scale SCHOOLCRAFT
1271 7126/2006 85.7492 45.9879 Patch; MACKINAC
1273 6115/2009 84.6704 45.7442 No Scale CHEBOYGAN
1277 712712006 85.083 45.6088 No Scale EMMET
2007011 611912007 86.1107 43.874 No Scale MASON
2007012 611912007 86.053 43.9868 No Scale MASON

 

9O

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
2007013 611912007 86.016 44.0325 No Scale LAKE
2007015 611912007 86.0811 44.0802 No Scale MASON
2007015 512612008 86.0811 44.0802 No Scale MASON
2007019 612012007 85.6956 44.2722 No Scale WEXFORD
2007020 612012007 85.6195 44.1866 No Scale WEXFORD
2007020 611612008 85.6195 44.1866 No Scale WEXFORD
2007020 513112009 85.6195 44.1866 Trace WEXFORD
2007022 612012008 84.7722 45.7197 No Scale EMMET
2007023 612512007 84.7122 45.7463 No Scale CHEBOYGAN
2007023 612012008 84.7122 45.7463 No Scale CHEBOYGAN
2007023 611512009 84.7122 45.7463 Whitewashed CHEBOYGAN
2007024 612512007 84.8434 45.6821 Patchy EMMET
2007029 612512007 84.9861 45.5768 Trace EMMET
2007030 612012008 84.9358 45.5506 No Scale EMMET
2007031 612512007 84.8536 45.55 Whitewashed EMMET
2007033 612512007 84.7162 45.5504 No Scale CHEBOYGAN
2007035 612612007 84.6877 45.4922 No Scale CHEBOYGAN
2007035 612012008 84.6877 45.4922 No Scale CHEBOYGAN
2007035 611412009 84.6877 45.4922 No Scale CHEBOYGAN
2007036 612612007 84.6502 45.3317 No Scale CHEBOYGAN
2007036 612012008 84.6502 45.3317 No Scale CHEBOYGAN
2007037 612612007 84.7427 45.3157 No Scale EMMET
2007038 612612007 85.3058 45.3073 No Scale CHARLEVOIX
2007039 612612007 85.8753 44.9171 No Scale LEELANAU
2007041 612612007 85.97 44.8426 No Scale LEELANAU
2007043 612612007 85.9683 44.6016 No Scale BENZIE
2007043 611812008 85.9683 44.6016 No Scale BENZIE
2007043 611 112009 85.9683 44.6016 No Scale BENZIE
2007044 612712007 86.0711 44.5735 No Scale BENZIE
2007046 612712007 85.9588 44.5288 No Scale BENZIE
2007046 611812008 85.9588 44.5288 No Scale BENZIE
2007051 71112007 84.7522 46.075 Trace MACKINAC
2007052 71112007 84.8799 46.1024 Trace MACKINAC
2007053 71112007 84.6966 46.0906 No Scale MACKINAC
2007053 711212008 84.6966 46.0906 Trace MACKINAC
2007054 71212007 86.4479 46.2918 No Scale SCHOOLCRAFT
2007057 71212007 86.4231 46.4762 No Scale ALGER
2007058 71212007 86.5503 46.4923 No Scale ALGER
2007059 71212007 86.5465 46.4736 No Scale ALGER
2007059 711 712008 86.5465 46.4736 No Scale ALGER
2007061 71212007 86.5871 46.4207 No Scale ALGER
2007062 71212007 86.5796 46.407 Trace ALGER

 

 

 

 

 

 

91

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
2007064 71212007 86.7595 46.397 No Scale ALGER
2007064 711812008 86.7595 46.397 No Scale ALGER
2007064 711612009 86.7595 46.397 No Scale ALGER
2007065 71212007 86.6155 46.3194 Whitewashed ALGER
2007066 71212007 86.7125 46.3028 No Scale ALGER
2007066 71212007 86.712 46.3044 No Scale ALGER
2007066 711812008 86.7125 46.3028 No Scale ALGER
2007067 71212007 86.6241 46.217 No Scale ALGER
2007067 711812008 86.6241 46.217 No Scale ALGER
2007069 71312007 86.3644 45.967 No Scale SCHOOLCRAFT
2007069 711812008 86.3644 45.967 Patchy SCHOOLCRAFT
2007070 711812008 86.0738 46.0434 No Scale SCHOOLCRAFT
2007073 612012007 85.701 44.2708 Patchy WEXFORD
2007083 711012007 84.7386 45.3126 Whitewashed EMMET
2007083 612512009 84.7386 45.3126 Whitewashed EMMET
2007084 711012007 84.7497 45.199 No Scale CHARLEVOIX
2007085 711012007 86.0365 44.8482 No Scale LEELANAU
2007086 711012007 86.101 44.6229 Trace BENZIE
2007088 711012007 85.9487 44.9347 No Scale LEELANAU
2007090 711612007 86.2075 44.7027 Patchy BENZIE
2007091 711612007 86.2091 44.7009 No Scale BENZIE
2007091 611812008 86.2091 44.7009 No Scale BENZIE
2007100 712212007 84.9752 45.9896 Whitewashed MACKINAC
2007108 712712007 86.1203 44.2701 Patchy MANISTEE
2007109 712712007 86.0356 44.2617 Patchy MANISTEE
2007114 71712008 85.7007 44.2511 No Scale WEXFORD
2007114 612512009 85.7007 44.2511 No Scale WEXFORD
20071 16 81312007 85.6857 44.2789 Trace WEXFORD
2007127 81612007 86.6445 46.3467 Trace ALGER
2007128 81612007 86.7854 46.4264 Trace ALGER
2008032 513012009 86.2664 43.5998 No Scale OCEANA
2008033 513012009 85.7879 43.4802 No Scale NEWAYGO
2008034 512312008 86.2124 43.4745 No Scale OCEANA
2008035 512912009 86.1132 43.5324 No Scale OCEANA
2008035 512912009 86.1132 43.5324 No Scale OCEANA
2008036 512312008 86.1194 43.5799 No Scale OCEANA
2008036 512912009 86.1194 43.5799 No Scale OCEANA
2008037 512912009 86.1263 43.6255 No Scale OCEANA
2008042 512512008 86.1299 43.7866 Trace OCEANA
2008044 512512008 86.0983 43.845 No Scale MASON
2008044 513012009 86.0983 43.845 No Scale MASON
2008046 513012009 86.05 43.9377 No Scale MASON

 

 

 

 

 

 

92

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
2008049 513012009 86.1033 44.1707 No Scale MANISTEE
2008052 513012009 85.8637 44.2849 Trace MANISTEE
2008057 512212008 86.4027 43.3443 No Scale MUSKEGON
2008057 511112009 86.4027 43.3443 No Scale MUSKEGON
2008100 513112009 85.658 44.2029 Trace WEXFORD
2008107 513112009 85.6046 44.2082 Trace WEXFORD
2008108 513112009 85.5953 44.1326 No Scale LAKE
2008111 611712008 85.795 44.4262 No Scale WEXFORD
20081 11 611/2009 85.795 44.4262 Patchy WEXFORD
2008118 611812008 86.0644 44.7128 No Scale BENZIE
2008120 611812008 85.9754 44.8452 No Scale LEELANAU
2008120 611 112009 85.9754 44.8452 No Scale LEELANAU
2008121 611412009 85.3066 45.3077 No Scale CHARLEVOIX
2008123 611912008 84.8465 45.2926 No Scale EMMET
2008123 611312009 84.8465 45.2926 No Scale EMMET
20081 32 612012008 84.714 45.2997 Trace CHEBOYGAN
2008135 81312008 87.011 46.4954 No Scale ALGER
2008302 711 312008 84.6914 46.061 1 Whitewashed MACKINAC
2008303 711312008 84.4354 46.0715 Whitewashed MACKINAC
2008304 711312008 84.441 46.0392 Whitewashed MACKINAC
2008309 711412008 83.6996 45.9813 No Scale CHIPPEWA
2008309 712012009 83.6996 45.9813 No Scale CHIPPEWA
2008312 711412008 83.6864 45.9955 Whitewashed CHIPPEWA
2008327 711812008 86.636 46.1465 Trace DELTA
2008328 711812008 86.6412 46.0725 No Scale DELTA
2008328 711 712009 86.6412 46.0725 Trace DELTA
2008330 711912008 86.8611 46.0307 No Scale DELTA
2008330 711812009 86.8611 46.0307 No Scale DELTA
2008338 711 512008 84.0919 46.0526 No Scale CHIPPEWA
2008339 711812008 86.719 46.358 No Scale ALGER
2009016 511 112009 85.78 43.025 No Scale KENT
2009017 512812009 84.73 42.8 No Scale CLINTON
2009018 512812009 86.1688 42.7054 No Beech ALLEGAN
2009022 512812009 86.1854 42.9775 No Scale OTTAWA
2009029 512912009 86.2862 43.5704 No Scale OCEANA
2009046 513112009 85.7126 44.1644 No Scale LAKE
2009050 513112009 85.3523 43.8845 No Scale OSCEOLA
2009053 513112009 85.3048 44.1314 No Scale OSCEOLA
2009061 61112009 85.56 44.3927 No Scale WEXFORD
2009067 61212009 85.2541 44.3612 No Scale MISSAUKEE
2009073 611 112009 86.0458 44.7991 No Scale LEELANAU
2009079 611 112009 85.9974 44.5291 Trace BENZIE

 

 

 

 

 

 

93

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale CounL

2009080 611212009 85.1574 44.7856 Trace KALKASKA
GRAND

2009086 611212009 85.3739 44.7322 NO Scale TRAVERSE
2009090 611212009 84.7867 44.7657 Trace CRAWFORD
2009092 611212009 84.7059 44.752 NO Scale CRAWFORD
2009093 611312009 84.6699 44.8877 NO Scale OTSEGO
2009096 611 312009 84.9392 45.0854 No Scale ANTRIM
2009101 611412009 85.3418 45.2835 NO Scale CHARLEVOIX
2009105 611412009 84.6863 45.5556 Whitewashed CHEBOYGAN
2009106 611412009 84.6418 45.5719 Trace CHEBOYGAN
2009114 611512009 84.5778 45.6921 No Beech CHARLEVOIX
2009117 611512009 84.0562 45.3951 No Scale PRESQUE ISLE
2009119 611 512009 84.4082 45.2462 NO Scale CHEBOYGAN
2009122 611512009 84.4723 45.2255 NO Scale CHEBOYGAN
2009124 611612009 84.6897 45.1069 NO Scale OTSEGO
2009137 612512009 85.6882 44.2485 Patchy WEXFORD
2009139 612612009 84.7578 44.5148 NO Beech CRAWFORD
2009140 612612009 84.675 44.423 No Beech ROSCOMMON
2009141 711612009 84.705 44.6124 No Beech CRAWFORD
2009144 711612009 86.7895 46.3419 No Scale ALGER
2009145 711612009 86.8658 46.3127 No Beech ALGER
2009147 711612009 86.931 46.4553 NO Scale ALGER
2009148 711712009 87.2215 46.4651 No Beech MARQUETTE
2009149 711712009 87.0428 46.3117 NO Scale ALGER
2009150 711712009 87.4077 46.3626 No Beech MARQUETTE
2009151 711712009 87.3613 46.2842 NO Beech MARQUETI'E
2009157 711812009 86.7304 46.1733 NO Beech ALGER
2009160 711812009 87.0764 45.7977 Whitewashed DELTA
2009161 711812009 87.1831 45.805 NO Beech DELTA
2009163 711812009 87.3667 45.7054 NO Beech MENOMINEE
2009164 711812009 87.3647 45.3952 No Scale MENOMINEE
2009165 711812009 87.5968 45.4059 NO Beech MENOMINEE
2009166 711812009 87.5238 45.6583 NO Beech MENOMINEE
2009167 711812009 87.603 45.7155 Trace MENOMINEE
2009168 711812009 87.8391 45.7762 No Beech DICKINSON
2009169 711812009 87.8746 45.7844 Patchy DICKINSON
2009170 711812009 87.9787 45.8061 No Beech DICKINSON
2009171 711812009 88.0385 45.9585 NO Beech DICKINSON
2009172 711912009 88.247 46.0844 No Beech IRON
2009173 711912009 88.4671 46.3042 No Beech IRON
2009174 711 912009 88.5329 46.5373 No Beech BARAGA
2009175 711912009 88.34 46.5751 No Beech BARAGA
2009176 711 912009 87.9571 46.5082 No Beech MARQUETTE

 

 

 

 

 

 

94

 

Appendix A : Site Coordinates

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Site Date Longitude Latitude Scale County
2009177 711912009 87.6273 46.5088 No Beech MARQUE‘I‘I’E
2009178 711912009 87.4615 46.565 NO Beech MARQUE‘I‘I’E
2009179 711912009 87.675 46.6602 No Scale MARQUETI'E
2009180 711912009 87.5481 46.3864 NO Beech MARQUE‘I‘I’E
2009181 711912009 87.2989 46.2553 No Beech MARQUE‘ITE
2009182 711912009 87.1993 46.1162 No Beech DELTA
2009183 711912009 87.0862 45.9783 No Beech DELTA
2009184 711912009 86.8913 45.8975 NO Beech DELTA
2009197 71312009 83.2702 44.7065 No Beech ALCONA
2009198 81112009 84.0075 43.3841 NO Beech SAGINAW
2009199 81112009 84.1036 43.4139 NO Scale SAGINAW
2009200 81412009 84.7555 42.348 NO Beech CALHOUN
2009201 81412009 84.2443 42.3194 NO Scale JACKSON
2009202 81412009 84.1982 42.3277 NO Beech JACKSON
2009203 81412009 84.0445 42.4187 No Beech WASHTENAW
2009204 81412009 83.9614 42.4071 NO Beech WASHTENAW
2009205 81412009 83.8643 42.506 NO Beech LIVINGSTON
2009206 811712009 82.7945 44.0182 No Scale HURON
2009207 811712009 82.5271 43.2692 No Beech SANILAC
2009208 811 712009 82.61 43.7166 No Beech HURON
2009209 811712009 82.4883 43.1078 NO Scale ST. CLAIR
2009210 811812009 84.3646 42.8071 No Scale CLINTON
2009211 811 812009 84.3563 42.8086 No Scale SHIAWASSEE

 

95

 

Appendix B: Image] Photo Analysis Protocol
Opening / Setting Up Program

1) Open a picture
a. Select File => Open
2) Setting the Scale by selecting Analyze => Set Scale
8. Set Distance in Pixels to 267.01
b. Set Known Distance to 1
-note must recalculate if new stabilizer is used (different distance
from the tree)
0. Change Unit of Length t0: Inch
(1. Check the Global Option

Analyzing Photos / Data Collection

1) Open Photo
2) Record Sampling Event, Tag Number and Photo Aspect
3) Visually asses photo for location of beech scale.
4) Select Image => Type => 8-bit
5) Select Image => Adjust => Threshold
Adjust the Threshold using the top bar (be sure the bottom bar is set to
255) *Note: do not hit apply“
6) If needed apply paint to increase contrast of scale from remaining photo

(Reapply Threshold)
7) Use polygon selection tool to select the tree
8) Select Analyze => Measure (Record Calculated Area)

9) Select Analyze => Analyze Particles (Uncheck Show results) (Record Total
Pixel Area)

10) Make notes if needed (e. g. used paint and cropped tree)

11) Save Changes Of new copy of photo

12) Repeat process

*Note: To minimize extra pixels when little scale is present, paint is usually needed.

96

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

97

 

 

 

 

 

a c..— 35:95 a 3.. 33:95 a 35 38:95
9.0an $82 832
6.7: 9:825 6.6 958.5 6.7: 9:525
298: 32: see: 32... Bus: 32:
a 39... a 32: a 32E
:5 33m 8 may :5 3.3 a m5. :5 33m 8 we...
a 3.. 38:95 a c: 35:95 a a: 38:55
882 Suez 852
6.28 9:825 $4.: 9:825 8.6 9:825
£95: 39... 2.98: 35:.— .auam 39:
a 355 a 39: a 39.:
:5 35m 8 was :5 35m 8 85. 55 3.3 a «5.
”55:36:52.5. .ﬁoamakeuﬁtﬂm £382— " N .32. n w .25: u a 3.50 28m :83— ESQ cash 89%” Ease—NEQ
:32: .3. ace 5 .3 .m 5:5
8:35 .353. .7: 98¢ .6 3.5 352 :8: a .351 :85 a .35. 5:. ac... 3.39:3 3.333
53— 3982.30 madam ram 865382
03:59:...“ 29:0 ”4mm ﬁeﬁuooom 25,—. .95 5:62 28> cone—=2 03m .53— ”5:55am

 

 

 

 

 

 

 

 

 

 

392—339 32: 23m Accum— "U EQZHmm<

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>926
:5
«2025
3 a. a. 2 2 a R o m n u e o. 9:: 850
52555.5. 08... .850
a 9.. 38:95 a 9.. 38:95 a 9.. 38:95
85 Z 852 852
6.5 9:825 3.5 9:825 5.5 9:825
9.98.. 39... 2:8: 39... 9.98: 39...
a 39:. a 39... a 39...
:5 98m ewe... 5:. 98m 9 may :5 98m :85...
e 9.. 38:95 a 9.. 38:95 a 9.. 38:95
852 857. 852
8.5 9:825 6.5 9:825 G5 9:825
9.98.. 39... 9.95.. 39... :88: 39...
a 39... a 39... a 39...
.55 98m II a we: .55 98: a we: .5: 98m 9 ma...

82.5500 ”0 9:525»:

98

Literature Cited

99

Literature Cited

Beyer, H. L. 2004. Hawth's analysis tools for ArcGIS. Available at
http://www.spatialecology.com/htools.

Burgman, MA. and J. F ox. 2003. Bias in species range estimates from minimum
convex polygons: implications for conservation and options for improved
planning. Animal Conservation 6: 19-28.

Bruzzone, L., D. Fernandez Prieto. 2000. Automatic analysis Of the difference image for
unsupervised change detection. IEEE Transactions on Geoscience and Remote
Sensing. 38.3: 1171-1182.

Burns, R.M., and B.H. Honkala, tech. coords. 1990. Silvics Of North America: 1.
Conifers; 2. Hardwoods. Agriculture Handbook 654.
U.S. Department of Agriculture, Forest Service, Washington, DC. vol. 2: 877.

Castlebury, L.A., Rossman, A.Y., and A. S. Hyten. 2006. Phylogenetic relationships
Of Neonectria/Cylindrocarpon on F agus in North America. Canadian Journal of
Botany 84: 1417-1433.

Ehrlich, J .. 1932. The occurrence in the United States Of Cryptococcus Fagi (Baer)
Dougl., the insect factor in a menacing disease of beech. Journal of the Arnold

Arboretum 13: 75—80.

Ehrlich, J. 1934. The Beech Bark Disease: A nectria disease of Fagus, following
Cryptococcus Fagi (Baer.). Canadian Journal of Research 10: 593-690.

Englund, S. R, J. J. O’Brien, and D. B. Clark. 2000. Evaluation Of digital and ﬁlm
hemispherical photography and spherical densiometry for measuring forest light
environments. Canadian Journal of Forest Research. 30: 1999—2005.

Felt, E.P. 1933. The beech scale and a fungus. Journal of Economic Entomology
26:510.

Gardner, G.. 2005. Population and large-scale dispersal dynamics in the beech scale
insect Cryptococcus F agisuga. Beech Bark Disease: Proceedings Of the Beech
Bark Disease Symposium. June 16-18, 2004. Ed. Evans, C. A., Lucas, J .A. and
Twery, M.J. Gen. Tech. Rep. NE-331. Newton Square PA, US. Department of
Agriculture Forest Service, Northern Research Station. 121—123.

Gora, V., R. Starke, M. Ziehe, J. Kdnig, G. Miiller-Starck and J. Lunderstﬁdt. 1994.
Inﬂuence of genetic structures and silvicultural treatments in a beech stand
(F agus sylvatica) on the population dynamics Of beech scale (Cryptococcus
fagisuga). For. Genet. 1 :157—164.

100

Gwiazdowski, RA. et al.. 2006. Possible geographic origin Of beech scale,
Cryptococcusfagisuga (Hemiptera: Eriococcidae), an invasive pest in North
America. Biological Control 39: 9—18.

Hansen, M.H., Frieswyk T., Glover J .F ., Kelly, J .E. 1993. The eastwide forest inventory
data base: users manual. Gen. Tech. Rep. NC-151. St. Paul, MN: U.S.
Department of Agriculture, Forest Service, North Central Forest Experiment
Station.

Heyd, R.L. 2005. Managing beech bark disease in Michigan. In: Evans, C.A., Lucas,
J .A. and Twery, M.J., eds. Beech Bark Disease: Proceedings Of the Beech Bark
Disease Symposium; 2004 June 16-18; Saranak Lake, NY. Gen. Tech. Rep.
NE-331. Newtown Square, PA: US. Department of Agriculture, Forest Service,
Northeastern Research Station: 128-132.

Holownicki, R. et al. 2002. Methods of evaluation of spray deposit and coverage on
artiﬁcial targets. Electronic Journal Of Polish Agricultural Universities. 5.1.

Houston, D. R. 1994. Major new tree disease epidemics: beech bark disease. Annual
Review Phytopathology 32: 75—87.

Houston, DR. and J. T. O’Brien. 1983. Beech bark disease: forest insect and disease
leaﬂet 75. U.S. Department of Agriculture: Forest Service 1983. Accessed on
October 13, 2004 http://www.na.fs.ged.us/spfO/pubs/ﬁdls/beechbark/fldl-
beech.htm. Houston, D.R., E.J. Parker, and D. Lonsdale. 1979. Beech bark
disease: patterns Of spread and development of the initiating agent Cryptococcus
fagisuga. Canadian Journal of Forest Research 9: 336-343.

Houston, D.R., B]. Parker, and D. Londsdale. 1979. Beech bark disease: patterns Of
spread and development of the initiating agent Cryptococcusfagisuga. Canadian
Journal of Forest Research 9: 336-344.

Houston, D.R., B.D. Rubin, M.J. Twery, and J .R. Steinman. 2005. Spatial and temporal
development of beech bark disease in the northeastern United States. Beech
Bark Disease: Proceedings of the Beech Bark Disease Symposium. June 16-18,
2004. Ed. Evans, C.A., Lucas, J .A. and Twery, M.J. Gen. Tech. Rep. NE-33l.
Newton Square PA, US. Department Of Agriculture Forest Service, Northern
Research Station. 43—47.

Kearney, A., McCullough, D.G., Walters, M. 2005. Impacts of beech bark disease on
wildlife resource abundance in Michigan. In: Evans, C.A., Lucas, J .A. and
Twery, M.J., eds. Beech Bark Disease: Proceedings of the Beech Bark Disease
Symposium; 2004 June 16-18; Saranak Lake, NY. Gen. Tech. Rep. NE-331.
Newtown Square, PA: US. Department of Agriculture, Forest Service,
Northeastern Research Station: 92-93.

101

Kondo, T., P.J. Gullan, and DJ. Williams. 2008. Coccidology The study of scale insects
(Hemiptera: Stemorrhyncha: Coccoidea). Ciencia y Technologia Agropecuaria 9
(2): 5581.

Lincoln, RC. 1939. “The migration of American Birds.” Doubleday Doran and
Company. New York.

McCullough, Deborah G., Robert L. Heyd, and Joseph G. O’Brien. Biology and
management of beech bark disease: Michigan’s newest exotic forest pest.
Michigan State University Extension Bulletin E-2746. 2004. rpt 2005.

Michigan Department of Natural Resources: Forest, Mineral and Fire Management
Division. 2003. IF MAP/GAP Land Cover.

<\\GPRPC 1 8\sippel_access\GIS_data\Michigan\CGI_data\Mich_Georef\
Michigan_entire state\land_use_cover\IFMAP_up_1andcover.img>.

Mielke, M.E., Houston, D.R., Bullard, A.T. Beech bark disease management
alternatives. 1987. USDA Forest Service. In: Proceedings, Integrated Pest
Management Symposium for Northern Forests, March 24-27, 1986. Cooperative
Extension Service, University Of Wisconsin- Extension, Madison, Wis.

Miller, D.R., G.L. Miller, G.S. Hodes, and J .A. Davidson. 2005. Introducing scale
insects (Hemiptera: Coccoidea) of the United States and their impact on U.S.
agriculture. Proc. Entomol. Soc. Wash 107(1): 123-158.

Miller DR. and M. Kosztarab. 1979. Recent advances in the study of scale insects.
Annual Rev. Entom0124: 1-27.

MOHR, C. 1947. Table of equivalent populations of North American mammals. Am.
Midland Nat. 37: 223-249.

Morin, R.S. et al. 2005. Mapping host-species abundance of three major exotic forest
pests. U.S. Department Of Agriculture Forest Service. Northeastern Research
Experiment Station. Research Paper NE-726.

Morin, R., Liebhold, A., Tobin, P., Gottschalk, K., and Luzader, E. 2007. Spread of
beech bark disease in the eastern United States and its relationship to regional
forest composition. Canadian Journal of Forestry Research 37: 726-736.

National Research Council. Predicting Invasions Of Nonindigenous Plants and Plant
Pests. Washington DC, National Academy Press, 2002.

Nobis, Michael, and Urs Hunziker. Automatic thresholding for hemispherical canopy-
photographs based on edge detection.” Agricultural and Forest Meteorology 128
(2005): 243—250.

102

O’Brien, J .G., M.E. Ostry, and ME. Mielke. 2001. First report of beech bark disease
in Michigan. Plant Disease 69: 905.

Petrillo, H., Witter, J .A., and Thompson, EM. 2005. Michigan beech bark disease
monitoring and impact analysis system. In: Evans, C.A., Lucas, J .A. and
Twery, M.J., eds. Beech Bark Disease: Proceedings Of the Beech Bark Disease
Symposilun; 2004 June 16-18; Saranak Lake, NY. Gen. Tech. Rep. NE-331.
Newtown Square, PA: US. Department Of Agriculture, Forest Service,
Northeastern Research Station: 48-51.

Pimentel, David, Rodolfo Auniga, and Doug Morrison. 2005. Update on the
environmental and economic costs associated with alien-invasive species in the
United States. Ecological Economics. 52.3: 273-288.

Rasband, W. “IMAGEJ 1.34S”. National Institutes of Health, USA.
http://rsb.info.nih.gov/ij.

Sakai, A., F. Allendorf, J. Holt, et. al. 2001. The population biology Of invasive
species. Ann. Rev. Ecol. Syst. 32: 305-326.

Schwalm, N. 2009. The beech scale (Crypotococcusfagisuga) in Michigan:
distribution, models Of spread and relation to forest and wildlife resources.
Michigan State University. East Lansing, MI 48824.

Shigo, AL. 1972. The beech bark disease today in the northeastern United States.
Journal of Forestry 70: 286-289.

Telewski, F. 2009. Michigan State University. pers. corn.

U.S. Department of Agriculture Forest Service. Forest insects. Forest Service Northern
Research Center. 30 Aug. 2007. United States Department of Agriculture. 6
Oct. 2007 <http://www.nrs.fs.fed.us/4506/>.

Wainhouse, D. 1980. Dispersal Of ﬁrst instar larvae of the felted beech scale,
Cryptococcusfagisuga. Journal of Applied Ecology. 17.3: 523—532.

Wainhouse, D. and I.M. Gate. 1988. The Beech Scale. In “Dynamics of
forest insect populations.” ed. AA Berryman. New York: Plenum 4: 76-85.

Wilcove, D.S., D. Rothstein, J. Bubow, A. Phillips and E. Losos. 1998. Quantifying
threats to imperiled species in the United States. Bioscience 48(8): 607-615.

103

|
I
|lll
Ill:
|
|
|

llllllllllll

III

'llllllllll

063 3501

1 9303

3