!

HI

 

WINNWWWWWMN

L

ll

   

141
241
.THS

«gut:

LIBRARY
Michigan §tate
University

This is to certify that the
thesis entitled

LONG-TERM IMPLICATIONS OF DAM REMOVAL FOR
MESOHABITAT AND MACROINVERTEBRATE
COMMUNITIES IN MICHIGAN AND WISCONSIN RIVERS

presented by
JONATHAN FORD HANSEN

has been accepted towards fulfillment
of the requirements for the

 

 

M. S. degree in lFISHERIES AND WILDLIFE

 

Major Professor’s Signature

'2 ”44)! Zd&<3’

 

Date

MSU is an affirmative-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

OCT 1 4 ME

101316

J."

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 ICIProj/Acd-PraleIRClDaleDue. indd

LONG-TERM IMPLICATIONS OF DAM REMOVAL FOR MESOHABITAT AND
MACROINVERTEBRATE COMMUNITIES 1N MICHIGAN AND WISCONSIN
RIVERS
By

Jonathan Ford Hansen

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

2008

ABSTRACT
LONG-TERM IMPLICATIONS OF DAM REMOVAL FOR MESOHABITAT AND
MACROINVERTEBRATE COMMUNITIES IN MICHIGAN AND WISCONSIN
RIVERS
By
Jonathan Ford Hansen

The removal of the numerous aging dams in the United States has recently been a highly
sought after stream restoration technique. The extent to which the damage done to
streams by dams is reversed upon removal is unknown, especially on decadal time scales.
The objectives of this study were to determine if mesohabitat heterogeneity and
macroinvertebrate assemblages within rivers recover following the removal of a dam, and
to estimate the time needed for recovery. A space-for—time substitution study was
employed on 8 rivers in various stages of recovery following a dam removal, ranging
from <1 - 40 years post—removal. Within each river, mesohabitat was mapped and
macroinvertebrates sampled in a zone unaffected by the dam removal and two zones
impacted by the dam removal (former impoundment and downstream zone), and
compared to evaluate the extent of recovery. Mesohabitat recovery was variable, with
some rivers recovering substantial heterogeneity within 7 years of removal while others
exhibited much lower levels than the reference zone even after 40 years following dam
removal. Generally, the macroinvertebrate community recovered 3 - 7 years following
removal both in terms of taxonomic similarity and richness, although densities took
decades to recover. Dam removal poses as a beneficial restoration technique yet the
recovery of important stream components can be variable and may take longer than

previous research has suggested.

This thesis is dedicated to my parents, Sandra Helpsmeet and David Hansen, who showed

me how to love the natural world.

iii

ACKNOWLEDGEMENTS

This work would not have been possible without funding provided by the
Michigan Department of Natural Resources, Schrem’s West Michigan Chapter of Trout
Unlimited, the Red Cedar Flyfishers, the Saginaw Bay Walleye Club, and the Paul H.
Young Chapter of Trout Unlimited. Thanks are also due to Mark Tonello and Tom
Rozich of Michigan Department of Natural Resources, Bob Stuber of the US. Forest
Service, Rob of Pappy’s, and Mike Metzelaars of Hamilton Road. My gratitude is
expressed to my superb crew of field assistants including Jon Wagner, Colby Bruchs,

J oel Berry, Marty Williams, and Firm the wonderdog.

The opportunity to tackle this graduate school endeavor was provided by Dan
Hayes, who guided my development as a student and a professional while simultaneously
providing a thoroughly enjoyable work environment. My thanks are extended to my
committee members, Michael Wagner and Rich Merritt. Bryan Burroughs provided
further guidance and is responsible for much of the formulation of the original ideas
behind this project.

I have to thank my lab comrades, especially Tracy Kolb and Amy Schueller, who
kept me sane, answered endless questions, and generated indispensable laughter. Thanks
go out to all the friends I’ve encountered over the last three years; you made the time
enjoyable. My heartfelt thanks and love go out to my family-no explanation needed.
Lastly, I express my gratitude and love for my wife, Gretchen, who is to owe for all that

is good in my life; you are my rock.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. vi
LIST OF FIGURES .......................................................................................................... vii
INTRODUCTION ............................................................................................................... 1
Study Area ................................................................................................................ 6
METHODS .......................................................................................................................... 8
Mesohabitat .............................................................................................................. 8
Macroinvertebrates ................................................................................................ 1 0
Data Analysis ......................................................................................................... l 1
RESULTS .......................................................................................................................... 14
Mesohabitat ............................................................................................................ 14
Macroin vertebrates ................................................................................................ 1 8
DISCUSSION .................................................................................................................... 37
CONCLUSIONS ................................................................................................................ 44
APPENDICES ................................................................................................................... 46
APPENDIX A. Maps of mesoscale habitat in 8 study rivers ............................... 46

APPENDIX B. Riffle assemblage macroinvertebrate data collected from 8 rivers
in various stages of recovery following dam removal. Heading numbers indicate
years since removal. Data are sum of three samples within each zone (Ref =
Reference; Imp = Former impoundment; DS = Downstream). Taxa sorted by
abundance for all rivers .......................................................................................... 50

APPENDIX C. Run assemblage macroinvertebrate data collected from 8 rivers in
various stages of recovery following dam removal. Heading numbers indicate
years since removal. Data are sum of three samples within each zone (Ref =
Reference; Imp = Former impoundment; DS = Downstream). Taxa sorted by
abundance for all rivers .......................................................................................... 56

LITERATURE CITED ...................................................................................................... 64

LIST OF TABLES

Table 1.1. Streams sampled and some key characteristics. Year removed and dam height
information from RAW 2005 and Hanshue 2006 ................................................................ 7

Table 1.2. Sampling dates for each study streams. Number in parentheses indicates
number of years since removal at time of sampling .......................................................... 10

Table 1.3. Percent composition of FFGs and total numbers within runs in the Hersey
River (<1 year post-removal) and the Pine River (3 years post-removal). CG = Collector-
Gatherer; CF = Collector-Filterer; Pr = Predator; Sc = Scraper; Sh = Shredder ............... 26

Table 1.4. Percent contributions of FF Gs to dissimilarity between runs in reference zones
and impacted zones within the Hersey River (<1 year post-removal) and the Pine River (3
years post-removal). CG = Collector-Gatherer; CF = Collector-Filterer; Pr = Predator; Sc
= Scraper; Sh = Shredder ................................................................................................... 28

Table 1.5. Taxa found in reference zones which are not found in impacted zones for the

two most recent dam removals. Ten most abundant taxa listed. (1) indicates larval stage,
(p) indicates pupa stage, and (a) indicates adult stage ....................................................... 34

vi

LIST OF FIGURES

Figure 1.1. a) Map of study streams within Michigan and Wisconsin. Numbers placed
approximately in location of former dam and indicate the number of years since removal.
b) Location of three zones within each study river .............................................................. 9

Figure 1.2. Percent difference in mesohabitat diversity for impacted zones compared to
the unimpacted reference zones. Shannon diversity index used ....................................... 16

Figure 1.3. Difference in percent run in impacted zones compared to reference zones.
Points above dashed zero line indicate zones with higher percent run than the reference
............................................................................................................................................ 17

Figure 1.4. Difference in mean percent gravel for impacted zones compared to the

reference zones. Solid line indicates regression for downstream zone excluding year 14
............................................................................................................................................ 19

Figure 1.5. Non-metric multidimensional scaling plot of all rivers, zones and habitat
types. Number in label indicates number of years since removal; Imp = former
impoundment zones, DS = downstream zone, Ref = reference zones; followed by habitat
type ..................................................................................................................................... 21

Figure 1.6. Non-metric multidimensional scaling plot of all rivers, zones and habitat
types. Number in label indicates number of years since removal; Imp = former
impoundment zones, DS = downstream zone; Ref = reference zones; followed by habitat
type. .................................................................................................................................... 22

Figure 1.7. Bray-Curtis similarity of taxonomic structure in impacted zones relative to the
reference zone over time since removal. Solid lines represent logarithmic relationships
between similarity and years since removal for both habitat types in the impoundment
zones (left panel) and downstream runs (right panel) ........................................................ 23

Figure 1.8. Bray-Curtis similarity of Functional Feeding Groups in impacted zones
relative to the reference zone over time since removal. Solid lines represent logarithmic
relationships between similarity and years since removal for impoundment runs (left
panel) and downstream runs (right panel). ........................................................................ 25

Figure 1.9. Difference in percent collector-gatherers between impacted zones and
reference zones over time since removal. Points below dashed zero line indicate areas
with lower percentage of scrapers than the reference zone. .............................................. 27

Figure 1.10. Difference in percent scrapers between impacted zones and reference zones

over time since removal. Points below dashed zero line indicate areas with lower
percentage of scrapers than the reference zone .................................................................. 29

vii

Figure 1.11. Difference in percent scrapers between impacted zones and reference zones
over time since removal with both habitat types and zones combined. Points below
dashed zero line indicate areas with lower percentage of scrapers than the reference zone..
............................................................................................................................................ 31

Figure 1.12. Percent difference in taxa richness between impacted zones and reference
zones over time since removal. Points below dashed zero line indicate areas with lower
percent EPT taxa than the reference zone. Solid lines represent logarithmic relationships
between richness and years since removal for impoundment runs and riffles (lefi panel)
and downstream runs (right panel). ................................................................................... 32

Figure 1.13. Difference in percent Ephemeroptera, Plecoptera, and Trichoptera (EPT)
taxa between impacted zones and reference zones over time since removal. Points below
dashed zero line indicate areas with lower percent EPT taxa than the reference zone ...... 35

Figure 1.14. Percent difference in densities between impacted zones and reference zones

over time since removal. Points below dashed zero line indicate areas with lower
densities than the reference zone ....................................................................................... 36

viii

INTRODUCTION

Dams provide numerous societal benefits including hydroelectric power
generation, agricultural irrigation, flood control, and creation of recreational activities.
Despite the many benefits dams have provided society, they have come at a cost to
natural resources. The effects of dams have been thoroughly studied and research has
highlighted a plethora of detrimental impacts to the ecological integrity of dammed rivers
(Baxter 1977; Petts 1980; Ward and Stanford 1987; Ligon et a1. 1995; Waters 1995;
Kondolf 1997; Graf 1999; Grubbs and Taylor 2004). With an estimated 2.5 million dams
in the United States (National Resource Council 1992), these detrimental effects are
widespread.

Dams disrupt the flow of rivers and alter physical attributes of the channel, the
thermal regime, and sediment transport (Heinz Center 2002; Poff and Hart 2002).
Irnpoundments formed by dams act as basins, filling in and restricting sediment and
nutrient cycling downstream (Waters 1995). Above the dam, impoundrnents transform
the physical habitat, shifting the system from an erosional, lotic environment to a
depositional, lentic enviromnent. Downstream of a dam, the river becomes sediment
starved due to the reduced sediment load thus increasing the erosive power and leading to
disproportionate hardening of the substrate (Ligon et a1. 1995). Dams warm impounded
reaches and, depending on the type of reservoir release, may substantially alter the
thermal regime, either warming or cooling downstream reaches. Thermal alterations
associated with dams can negatively impact downstream fish and macroinvertebrate

communities, especially in coldwater streams (Lessard and Hayes 2003).

The physical modifications associated with dams have significant implications for
biotic communities as well. A free flowing river has many characteristics associated
with its geomorphology that create a heterogenous and dynamic environment. An
unregulated river tends to exhibit a longitudinal profile characterized by a sequence of
alternating pools, riffles, and runs collectively referred to as mesohabitat. Mesohabitat
(i.e. pool-riffle-run sequence) heterogeneity regulates the energy expenditure of rivers
(Gordon et a1. 2004). This diversity of habitat is also important in maintaining the
biological integrity of a stream and can strongly affect macroinvertebrate assemblage
distributions (Pardo and Arrnitage 1997; Beisel et al. 1998; Rabeni 2000). Once
impounded, a stream shifts from a varying sequence of pool, riffle, run to primarily pool
and run habitat within the impoundment, which will subsequently alter the associated
macroinvertebrate community. Shifts from lotic to lentic systems associated with the
impoundment can cause changes in the fish and macroinvertebrate communities favoring
more tolerant organisms and reducing species richness (Tieman et a1. 2004; Santucci et
a1. 2005), mostly due to habitat transformation. The creation of an impoundment favors
fish and macroinvertebrates that prefer slower moving, warmer water. Modifications of
the macroinvertebrate community can alter various components of stream functioning,
including the aquatic and terrestrial vertebrate organisms which rely on aquatic
invertebrates for food.

Beyond their role in stream foodwebs, aquatic macroinvertebrates serve as
effective indicators of stream quality. Macroinvertebrates are widespread in all stream
habitats, show a gradient of responses to disturbance, and are easily sampled; therefore,

they have been widely used in biomonitoring efforts for nearly a century (Rosenberg and

Resh 1993). Numerous metrics have been developed to facilitate the analysis and
interpretation of biomonitoring studies. One of the most widely used classifications of
macroinvertebrates includes the designation of functional feeding groups (FF G)
(Cummins and Klug 1979; Merritt and Cummins 2006). Functional feeding groups are
categories of invertebrates based on their morphological attributes associated with
resource acquisition. The river continuum concept (V annote et al. 1980) suggests that the
distribution and ratios of FFGs varies predictably along the longitudinal gradient of
rivers. Thus, shifts in FF G composition are indicative of changes to the ecological
functioning of streams.

The disruption of stream functioning from dams coupled with the aging state of
dams in the United States has facilitated a recent movement supporting dam removal.
Dams generally have a life expectancy of 50 years, and current estimates predict that
more than 80% of the documented dams in the United States will exceed this age by 2020
(USACE 1996). Given these estimates, the fate of many dams will soon be in question.
To date, removal is an extremely popular option with an estimated 654 dams removed
from the nation’s rivers (American Rivers 2007). Overall, darn removal appears to be a
promising restoration technique beneficial to the integrity of aquatic systems. Yet, as
with any restoration technique, the scientific community needs to evaluate all potential
effects of dam removal, both negative and positive. By assessing the effects of dam
removals on stream geomorphology and biota, restoration efforts can be more efficiently
and effectively administered to maximize ecological benefits.

A dam removal can be viewed as a significant ecological disturbance and

therefore presents a unique experimental research opportunity. Recently, this opportunity

has been recognized and a surge of research has been conducted on the ecological effects
of dam removal (e. g. Kanehl et al. 1997; Stanley et al. 2002; Doyle et al. 2003; Pollard
and Reed 2004; Doyle et al. 2005; Thomson et al. 2005; Stanley et al. 2007; Cattalano et
al. 2007), yet the body of technical literature demonstrating the effectiveness of removal
as a restoration technique is still relatively small. Most studies have focused on short-
term (<5 years) changes in the river geomorphology (Doyle et al. 2003; Stanley et al.
2002), macroinvertebrate response (Stanley et al. 2002; Pollard and Reed 2004; Thomson
et a1. 2005), and fish response (Kanehl et al. 1997). Stanley et a1. (2002) observed
significant erosion and sediment transport in the former impoundment, resulting in rapid
channel development close to equilibrium. Sedimentation was observed downstream but
was characterized as small and transient. Bushaw-Newton et al. (2002) discovered
similar erosion in the former impoundment, yet ten months following removal, coherent
riffle-pool formation had failed to emerge. Additionally, they observed downstream
aggradation covering coarse cobble dominated riffles. Kanehl et al. (1997) found
improved habitat, abundance, and recruitment of desirable fish species following the
removal of a dam on the Milwakee River, Wisconsin.

Studies focused on the response of macroinvertebrate communities following dam
removals have yielded mixed results. Bushaw-Newton et al. (2002) noted a shift in
macroinvertebrate taxa from lentic to lotic in the former impoundment, and seemingly
transient declines in sensitive taxa following a Pennsylvania dam removal. Within one
year of dam removal, Stanley et al. (2002) found macroinvertebrate communities in both
the former impoundment and downstream similar to an upstream reference site. Pollard

and Reed (2004) found varying responses in the upstream reference and limited

taxonomic changes near the dam and downstream sites. Downstream of dam removal,
Thomson et al. (2005) observed no changes in macroinvertebrate assemblage structure
and declines in density yet suggest that the negative impacts are relatively minor and
temporary.

One recurring message throughout the peer-reviewed and gray literature is the
need to evaluate long-term consequences of dam removal (e. g. Bednarek 2001; The
Heinz Center 2002; Stanley and Doyle 2003). Few published studies have examined
impacts beyond three years, with the exception of Cattalano et al. (2007) and Burroughs
(2007) which focused on fish and fluvial geomorphological changes. Despite this
acknowledged gap in understanding, some literature on dam removals has suggested time
frames for some of these responses (Hart et al. 2002; Doyle et al. 2005). Major
geomorphic adjustments including the sediment regime and channel form are suggested
to take the longest to return to pre-dam condition. Conversely, many believe that
macroinvertebrates will quickly respond to any newly available habitat and recolonize the
impacted areas, due to their mobility and short life—cycles. However, studied responses
of macroinvertebrate communities to dam are somewhat inconclusive and any observed
decline in macroinvertebrate community metrics has been attributed to transient
processes and suggested to be short-term. Before this study, no research has been
conducted focused on evaluating the long-term effects of dam removal on
macroinvertebrate communities.

Connections between macroinvertebrate community composition and mesohabitat
(i.e. pool-riffle-run sequence) (Pardo and Armitage 1997; Beisel et al. 1998; Rabeni

2000) suggest that a dearth in available heterogenous pool-riffle-run habitat will have

significant implications for macroinvertebrates. Considering the predicted and observed
long time frame needed for recovery of key physical attributes such as mesohabitat
(Bushaw-Newton et a1. 2002) following dam removal, the recovery of the
macroinvertebrate community could take more time than previous researchers have
predicted. The goal of this study was to evaluate the long—term restorative capacity of
dam removal. My objectives were to determine if mesohabitat heterogeneity and
macroinvertebrate assemblages within rivers recover following the removal of a dam, and
to estimate the time needed for recovery. This research serves as a valuable tool for
managers when faced with questions regarding the long-term capabilities of dam removal
as a restorative technique. Understanding long-term responses of community and
physical dynamics of a river following dam removal helps to clarify expectations,
determe reasonable objectives, and thus, paramount in conducting effective and
responsible stream restoration.
STUDY AREA

I used a space-for-time substitution approach to quantify changes in mesohabitat
and macroinvertebrate communities following the removal of dams from eight rivers in
Michigan and Wisconsin (Figure 1.1a). Considering the lack of long-term datasets on
dam removal, this approach addresses long—term questions by incorporating rivers in
various stages of recovery following dam removal. Wadable rivers were selected from
lists of all documented dam removals in Michigan and Wisconsin (Table 1.1; RAW 2005;
Hanshue 2006) to maximize the temporal range of time since removal. Time since
removal ranged from less than 1 year to 40 years, mean width ranged from 9.0 m to 28.2

m, and darn height ranged from 1.4 m to 5.2 m.

.88 E @8238”. use 33 E wagmmon 3588 568m 332?: p

 

 

m. S. omow «A 82 82.8.35 :2 £8250 53m 320 wcflooq
adv mnmo 8522:: $3 «sebum 34 £82 33¢ 8E
od mmnm mé $3 DEESEZ :5 .owmtom 53¢ 38.888.
4.2 8% 3 SS cease E .86; see 8%on
0.2 omow Nd g2 wfizmm H2 “28330 SSE 22mm :<
Gwm when 5m ooom 823nm :5 .xoom x080 2:3.
mom was 2 ~88 :oaaoam =2 semen: 35 BE
v.3 mmom mmd ooom >83: :2 £830 :52 >853
AEV 82? SSE 0:8 33m Ema 30> BENZ Eda 38m 5850 62%

doom
0:55: cam moom 22M Sow 832585 Emma: 83 use @2688 80> .moumtowofimco >03 08% 98 3383 952% .3 2an

METHODS

Each study river was divided into three roughly equivalent sized zones relative to
the location of the former dam: a reference zone assumed to not be impacted by the dam
or subsequent dam removal (reference), the formerly impounded area immediately
upstream of the former dam (impoundment), and a zone downstream of the dam site
(downstream) (Figure 1.1b). Study zone lengths ranged from 5625 m to 8650 m, and
were constrained by the estimated size of the impoundment and logistical constraints
such as stream confluences and downstream impoundments. Within each stream, I
compared the response variables in zones affected dam removal (impoundment and
downstream) to those in the zone not impacted by dam removal (reference).
Mesohabitat

Mesohabitat was mapped for each river following the general guidelines of Hicks
and Watson (1985), where visual assessment while walking the entire study section was
used to differentiate between habitat types. The extent of the stream was categorized into
a minimum of 25 m units each assigned one the following habitat categories: pool, riffle,
run, rapids, or complex. Pools were slower, deeper (> 1 In) sections of the river in which
water velocity was obviously reduced. Riffles were higher gradient, shallow (< 1 m)
sections which displayed water surface disruption. Runs varied from shallow (< 1 m) to
deep (> 1.5 m), and exhibited an undisturbed water surface while maintaining steady
water velocities. Rapids were marked by sudden increases in gradient resulting in high
water velocity usually displaying some form of “whitewater.” Complexes consisted of

more than one habitat type.

DEE 32m some 553 3:3 025 we cosmos
3 .3388 35m E8» mo .885:— ofi 8865 can Emu 82:8 mo .8382 E bBuEioaaa

coon—n 83:52 32885, 28 5&322 52.23 2:35 33m .«o 32 An .2 053m

 

_ 7 _ . _ . . _ d
89080:! ON». 0 FN m9. 0

 

 

ocON Z
383E. N.
$352260 0 v

ocoN
8685.
-EmEccsoag fl

Lento”. 26E

 

mcoN
umfimaEE:
-oocouoamm

 

3

 

 

 

 

 

 

 

At the upstream extent of each habitat type unit (i.e. pool, riffle, etc...) the length and
width of each unit were measured using a Nikon® laser rangefinder (+/- 0.5 m accuracy).
Substrate composition within each habitat type was assessed visually and percentages
were qualitatively assigned to each substrate type (clay, silt, sand, gravel, cobble,
boulder, or vegetation). Mesohabitat mapping occurred between June 2005 and October
2007 (Table 1.2).

Table 1.2. Sampling dates for each study streams. Number in parentheses indicates
number of years since removal at time of sampling.

 

 

River Mesohabitat Macroinvertebrates
Hersey River 6/20/2007 (<1) 6/21/2007 (<1)
Pine River 6/18/2007 (4) 7/7/2006 (3)
Turtle Creek 5/30/2007 (7) 5/31/2007 (7)

Au Sable River 7/5-6/2006 (14) 7/6/2006 (14)
Kickapoo River 6/2/2007 (15) 6/3/2007 (15)
Tomorrow River 6/5/2007 (20) 6/6/2007 (20)

Flat River 6/7/2005 (32) 6/28/2006 (33)

Looking Glass River 10/30-31/2007 (41) 6/23/2006 (40)

 

Macroinvertebrates

A stratified random sampling approach was used to characterize the
macroinvertebrate assemblages (Cummins 1962). Sample sites were selected following
the mesohabitat mapping by randomly selecting three riffles and three runs in each zone.
Within each riffle and run a random distance from the upstream extent of the habitat unit
and a random distance from shore were selected. Riffle sampling was conducted with a
modified Hess sampler (0.086 m2 area, 500 um mesh) and run sampling was conducted
using one minute timed intervals with a triangular dip net (500 pm mesh) (Merritt and

Cummins 1996). At both run and riffle sample sites, if the depth was greater than the

10

sampling equipment allowed, the sample was moved laterally until an adequately shallow
depth was encountered. Macroinvertebrate sampling occurred in June and July 2006 and
May and June 2007 (Table 1.2).

Macroinvertebrate samples were preserved in 80% ethanol and separated from the
sediment in the laboratory via a sucrose floating method (Anderson 1959). Samples
estimated to contain more than 400 individuals, based on preliminary sorting, were
subsampled using a plankton splitter and taxa counts were multiplied according to the
number of times the sample had been split (i.e. split once = N * 2, split twice = N * 4).
Insects were identified to the family level and placed into fimctional feeding groups
according to Bouchard Jr. (2004) and Merritt and Cummins (2006), while non-insects
were identified to the lowest practical taxonomic level. Insects in the family
Chironomidae were divided between the FFG collector-gatherer and predator, 90% and
10% respectively (R. W. Merritt, Michigan State University, personal communication).
Data Analysis

Patterns in mesohabitat changes over time since dam removal were evaluated
multiple ways. Meshohabitat maps for each river were visually examined for spatial
patterns. Trends in time following dam removal were evaluated by looking for
differences in habitat heterogeneity or homogeneity between the reference zones and the
two impacted zones (impoundment and downstream). Due to variations in field
determinations I combined rapids with riffles and complexes with pools. Mesohabitat

diversity (H) was determined using the Shannon diversity index (Krebs 1999) as:

s
H = .lez' log2 pl.
1 :

11

where s is the number of habitat types and p,- is the proportion of total sample belonging

to the ith habitat type. Percent differences in mesohabitat diversity (PD A) between the

reference zone and the impacted zones were calculated as:

r

PD 2 x100

A

where D,- is the diversity in the impacted zone i, and D, is the diversity in the reference

zone r. The amount of run habitat was expected to dominate zones impacted recently by
a dam removal, therefore differences in percent run between the reference zone and the
impacted zones were examined.

Responses of substrate were evaluated using the qualitative assessments within
each zone. Percentage of gravel was used to indicate the extent of substrate coarsening,
operating under the assumption that the amount of gravel will increase in impacted zones
with time since removal. The percentage of gravel was averaged over all habitat types
for each zone and the differences between the reference zones and the impacted zones
were calculated. Patterns in differences in mesohabitat diversity, percent run habitat, and
mean percent gravel between both the impoundment and downstream zones and the
reference zones were examined for trends over time since dam removal using linear
regression.

Responses of macroinvertebrate communities over time following dam removal
were evaluated using multivariate based comparisons of similarities and other metrics
plotted over time since dam removal including functional feeding group proportions,

percent Ephemeroptera, Plecoptera, and Trichoptera (EPT) taxa, and percent differences

12

in densities. Multivariate approaches included initial exploration of between zone and
river similarities using classification and ordination techniques based on Bray-Curtis
similarities (Bray and Curtis 1957) using the software package PRIMER (Clarke and

Warwick 1994). Given the dominance in abundance of a few families, a one was added

to all macroinvertebrate data which were then loge transformed. The CLUSTER routine

within PRIMER classified samples based on their Bray-Curtis similarities and produced a
dendrogram illustrating the grouping of certain samples and their associated levels of
similarity. Non-metric multidimensional scaling (NMDS) was also used to portray the
grouping of samples in ecological space. Non-metric multidimensional scaling addresses
the inherent non-normal distribution of macroinvertebrate count data and ordinates
samples based on the ranking of Bray-Curtis similarities and plots them according to
those ranks. A stress value is calculated with NMDS which measures the level of
agreement between the rank order of distances calculated from the data versus the rank
order of distances from the ordination, thus indicating the degree of true representation of
the ordination. Typically, a stress level <0.20 is considered acceptable (Clarke and
Warwick 1994).

Trends in macroinvertebrate similarity of the impacted zones to the reference over
time since removal were evaluated using linear regression. Using the Bray-Curtis
similarities generated from PRIMER, I plotted similarities of the macroinvertebrate
community in riffles and runs in the impoundment zone and downstream zone to the
associated reference habitat types and zones, which then created a temporal pattern of
change in similarity. Similarities based on family taxa and FF Gs were both evaluated to

determine if taxonomic structure responded differently than function. Habitats and zones

13

of rivers which were markedly dissimilar to the reference were firrther examined to
determine which functional feeding groups were driving the differences. The highest
contributing fimctional feeding groups were determined using the SIMPER routine in
PRIMER which identifies the percent contribution of each taxonomic group to the
average dissimilarity between samples.

General linear models were used to evaluate differences in percents of the
collector-gatherer and scraper F FGs, percent differences in taxa richness, differences in
percent EPT taxa, and percent differences in densities, all of which were differences
relative to the reference zone. Each model followed the design of:

y = t + h + t*h + e
where y was the parameter of interest, t was a continuous variable representing the effect
of time since removal, h was a categorical variable indicating the effect of habitat (riffle
or run), t*h was the interaction between time and habitat, and e was the error term.
Separate models were run for each parameter of interest and each zone. General linear
models were run using SAS (version 9.1, SAS Institute, Inc., Cary, North Carolina).

RESULTS

Mesohabitat

Visual examination of mesohabitat maps revealed mixed and variable results
(Appendix A). Obvious differences in mesohabitat heterogeneity existed within the
impoundment of the Hersey River, after less than a year post-removal, yet the
downstream zone exhibited higher heterogeneity than the reference zone. Other rivers
exhibited similarly mixed results with variable indications of recovery. Surprisingly

some rivers, such as the Looking Glass River, experiencing removals even up to 40 years

14

ago, showed some indication of limited mesohabitat recovery within the former
impoundment, while Turtle Creek (7 years post removal) appears to exhibit higher
mesohabitat heterogeneity in both impacted zones when compared to the reference zones.

Mesohabitat diversity appeared to show no clear trend in recovery over time
(Figure 1.2). Overall, many of the impacted zones were lower in diversity than the
reference zone, even 40 years afier dam removal. Some zones within rivers, however,
exhibited high diversity compared to the reference zone soon after dam removal. Within
1 year of removal, the Hersey River displayed heterogenous habitat downstream of the
former dam, 8% higher diversity than the reference, yet the impoundment showed very
limited habitat diversity, 59% lower than the reference. After 7 years following dam
removal, both the downstream and impoundment zones of Turtle Creek exhibited 56%
higher diversity than the reference zone. The Looking Glass River displayed slightly
higher diversity of mesohabitat in the downstream zone (7%) after 40 years, while the
impoundment was 20% lower in diversity than the reference.

The difference in percent run between the impoundment and the reference zone
also showed no clear pattern over time since removal (Figure 1.3). Half of the
impoundments contained proportionately higher amounts of run habitat than the reference
zones, even afler 40 years. Similarly, downstream of the dams the difference in percent
run from the reference Showed no patterns. Run type habitat in downstream zones was
proportionally higher than the reference zone in 62.5% of cases. Overall, 56% of the
impacted zones exhibited proportionately higher amounts of run type habitat when

compared to the reference zones.

15

68: 52: 3536 co>83éoc=£m .mocon oocucofit 38388:
05 8 Ban—Eco 8:8 383:: 8m gauze “852808 E 855va Bony—om .N._ DSwE

_m>oEo._ ooEm 23>

av cm on cv

p h _ —

 

 

anbmcioo RV D
EOE—E3095 .

 

 

 

 

I oer.

L

l
o
"3

O

I
O
In

 

 

ArtsJeAgp u! eoueraggp ruemed

l6

.85me 2: ~85 8: Benton 8&5 £5 mecca 8865 on:
Sou 35% 26% £50m .mocoN 355%: 8 359:8 3:8 cBomQE_ E E: Eugen E 35:55 .mA PSME

_m>oEm._ 005m mam;

 

 

3. on ow 3 o
_ p _ r fin?
. No.
0
O
.......... D 3
D D
. .»
D - No
5352260 D
EaEucsong .

 

 

 

 

 

 

v6

aouaJapJ won um % u! eouaJayga

l7

Qualitative substrate assessments showed an increase in mean percent gravel
within the impacted zones over time (Figure 1. 4). By the 15th year following removal,
the impoundments of all streams had the same or higher levels of mean percent gravel
than the reference zone. Within downstream zones, mean percent gravel levels did not
exceed that of the reference until 33 years post-removal. Linear regression indicates a

weak positive trend of increasing mean percent gravel over time for both impacted zones

7
(R‘ = 0.14, F1,14 = 2.35, p = 0.15). The Au Sable River ( 14 years post-removal) differed

drastically in mean percent gravel between the reference zone and both the impoundment
and downstream zones, -50.8% and -58.9%, respectively. When the Au Sable River was

removed from the regression, the strength of the trend in gravel recovery over time
became significant (Figure 1.4; R2 = 0.61, Fug = 18.34,p < 0.01).

Macroinvertebrates

A total of 39,986 macroinvertebrates were collected from the 8 study rivers. The
majority of individuals (30,046) were found in riffle habitat (Appendix B) compared to
run habitat (9940) (Appendix C). The most dominant taxon in both habitat types was
Chironomidae, which accounted for 29% of the individuals from riffle samples and 42%
of the individuals from run samples. Within riffles, Hydropsychidae caddisflies were the
second most dominant group, accounting for 16% of the individuals, while Ehnidae
larvae were also common, representing 10% of the individuals. Within run habitat, the
two families most prevalent after Chironomidae were Caenidae and Baetidae mayflies

which accounted for 10% and 6% of the individuals, respectively.

18

.3 So» mics—88 econ anmesoc com .868“on 83065 0:: Bow
.828 858%: on: 8 BBQ—:8 meson 388:: com 139% :5ch :88 E oocoeoba a: charm

_N>OE0._ ¢Uc_m mhm¢>

cw cm on E. c

anzmcioo my
EoEucaoac: .

 

 

D
.

 

Rod u q .50 "mm .33 + «.2- n >

|6AEJB % ueaw u! aouaiamq

 

 

 

 

l9

Bray-Curtis classification of the macroinvertebrate communities of all zones and
habitat types indicated a strong divergence in similarity of run communities in the
impacted zones of the Hersey River (< 1 year post-removal) and run and riffle
communities in the Pine River (3 years post-removal) relative to the reference zones
(Figure 1.5). In contrast, most of the rivers in later stages of the recovery grouped
together, indicating strong similarities between zones of the same river. Non-metric
multidimensional scaling displayed the same relationships with an acceptable stress level
(Figure 1.6; stress = 0.1).

Macroinvertebrate assemblages were more similar between impacted zones and
reference zones after longer amounts of time since removal, as supported by the
multivariate classifications and ordinations (Figure 1.7). Within the impoundment, riffle
assemblages exhibited lower similarity to the reference zone within the first 3 years than
the later years (> 7). The most extreme difference was n'ffle assemblages in the Pine
River’s impoundment which were markedly different than those of the reference zone,
showing only 30% similarity. Run assemblages within the impoundment showed similar
patterns as the riffles over time but were more dissimilar. Initially, run assemblages in

impoundments of the Hersey River and the Pine River displayed limited similarity to the

reference zone, 32% and 27%, respectively. By the 7th year following dam removal, the

riffle and run assemblages appeared to be relatively similar to those of the reference,
showing at least 70% and 66% similarity, respectively. With riffle and run samples

grouped together, the taxonomic structure of the macroinvertebrate assemblage within the

20

Bray-Curtis similarity

~r 00L

+ 09

- 09

- or

1- oz
0

20-DS-Run
20-Ref-Run
ZO—I mp—Run
zoos-Rm
20—RefRiff
20-I mp—Riff
B-Ref—Riff
3-Ret—Run
1-Ref-Run I
1-1mm.“ '
14—DS-Run T
14-Ref-Run #‘4

14-1 mp-Run
7—Ref—Run
15—l mp-Run H
15-DS—Run
15-Ref—Run
1~DS Riff
1—Ref—Riff }_
40Ref—Run

4008“" :—
40-Ref-lef

40-DS—Run
40-1 mp—Run

334 mp Run .3
33~DS—Run
33-Ref—Run
15—I mp—Riff
15-08 Riff
15ReH-1'iff
33-DS-Riff
33-Ref-Riff
33—1 mp~Riff
7-DS—Run
7~| mp—Run
7Ref-Riff
IDS—Riff
7-1 mp—Riff
3-1 moRun
3~DS~Run
3-I mp—Riff
1—DS-Run
1—I mp-Run

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.5. Clustering dendrogram of Bray-Curtis similarities. Number in label indicates
number of years since removal; Imp = former impoundment zones, DS = downstream
zone, Ref = reference zones; followed by habitat type.

21

dab “553 .3 330:8 ”mecca confluence n and ”0:3 Ewobmgov
H mm 633 Eofiwgomfi_ 58.5% H 95 ”—3088 025. E3» mo “38:: mBaomuE :53 E
598:2 .893 «83m: 98 mocou £83.— :a mo poi wfifiom Enemy—086298 0505.82 .0; 053m

 

 

Sade;

csmtwn;

 

tide;

Side:
gamma;
to ”$96

 

 

22

dong Ewtv wee Emubmgov 98 20:3 do: meson Eoficgomfi_ on“ 5 3ng “332 53 e8 3588
85m 88% can bum—EH6 5933 3230928 BEEtawS 28858 8:: 23m ._m>oEoc 85m 2:: .8>o
0:3 856%: 05 8 9529— mocou @8895 5 2803.3 02:28me me bra—«EB £90435 .5; “35me

_m>oE2 oufim 23>

G? en ON 2. c ov

p

 

 

_m>oEa.. 35m 33>
on cu or c

 

mod n Q .33 u mm .32: + 2N u .2

D 55 v q .30 H mm .Eswfi + Sm u >

 

 

 

 

 

. . _ o

Mueuwis Slum-£918

1
O
Q

 

 

m: :1 D
Emohmcaon 35$. 0

c3.
EmEuczoag

23

impoundment followed an asymptotic trajectory of recovery (Figure 1.7; R2 = 0.64, F1,13

= 23.37,p < 0.01).

The zones downstream of the former darn exhibited similar patterns of
macroinvertebrate assemblage recovery as the former impoundment in runs yet not in
riffles. Taxonomic structure of riffle macroinvertebrates within the downstream zone
appeared to recovery immediately, returning to > 65% similarity within 1 year of dam
removal (Figure 1.7). Run assemblages downstream of the dam appeared to show limited
recovery within the first 3 years of dam removal. Downstream run assemblages within
the Hersey River were 41% similar to those of the reference zone while downstream runs
in the Pine River were only 15% similar. Following the patterns of the impoundment,

similarity of the downstream zone to the reference zone asyrnptotes around 7th year

following the dam removal (Figure 1.7; R2 = 0.64, 131,6 = 10.63, p = 0.02).

Functional feeding group composition of the macroinvertebrate communities
appeared to be less affected by dam removal than taxonomic structure, exhibiting overall
higher similarities across all time steps (Figure 1.8). Specifically, riffle assemblages in
both the impoundment and downstream zones returned to at least 83% similarity within 1
year of dam removal, with the exception of riffles in the former impoundment of the Pine
River, which showed only 42% similarity after 3 years. Run assemblages within both

impacted zones displayed noticeably lower similarities within 3 years of dam removal,

while runs within rivers in the 7th year or more following removal exhibited high levels

of similarity. Functional recovery within runs in both zones follows the same asymptotic

24

.9053 :93 m5: Emgaaou 23 90:3 «05 m5: EDEESOQE 8m 3682 35m
memo» was 353:6 c8253 man—80:32 ogfitmwB 582%: 8:: 23m .3582 35m 08: .85 28s
85.53“ 05 8 9322 mecca @2895 E 8:95 wagon Enoaocsm mo bra—Eta flwBUJAEm .mg 233m

 

 

 

 

 

 

 

 

_m>OE9_ qum 2m0> _m>0:._0.. 005m ”30>
8 S 8 2 c 8 8 8 2 c
p . . . p . . . o
:3 v q .86 u & .Ssu. : + 4.5 u > 8.0 v q .86 "mm .359 + 4.8 u >
- 1 8 a
m
.A
a
r n
m...
S
S
D r w.
w
- m.
a w 0_
mczm D

Emobmcion «$.82:an

moat”. O

25

trajectory as taxonomic structure (Figure 1.8; R2 = 0.80, Impoundment: F1,6 = 24.20,

Downstream: F1,6 = 25.30,p < 0.01).

The composition of the F F Gs within Hersey River and Pine River runs varied by
zone (Table 1.3). The FF Gs contributing to most of the dissimilarities were predators,
scrapers, and collector-gatherers (Table 1.4). The dissimilarities between runs in both the
impacted zones of the Hersey River was primarily attributed to predators. Additionally,
scrapers contributed to the dissimilarities in both the impoundment and downstream
zones, 18% and 22%, respectively. Within the Pine River, collector-gatherers accounted
for the majority of the differences between runs in the impacted zones and runs in the
reference zones.

Table 1.3. Percent composition of FFGs and total numbers within runs in the Hersey

River (<1 year post-removal) and the Pine River (3 years post-removal). CG = Collector-
Gatherer; CF = Collector-Filterer; Pr = Predator; Sc = Scraper; Sh = Shredder

 

 

 

Hersey River Pine River
Ref Impound DS Ref Impound DS
% CF 3 7 3 2 0 0
% CG 85 89 94 51 56 10
% Pr 10 4 2 25 22 57
% Sc 1 O 0 3 22 33
% Sh l 0 0 18 0 0
Total 875 27 123 175 9 30

 

General linear models failed to reveal any significant patterns in the difference in

percent collector-gatherers between the reference and the impoundment or downstream
zones (Figure 1.9; Impoundment: F1,13 = 2.38, p = 0.15; Downstream: F1’12 = 2.62,p =

0.13). Visual examination of the relationships within the impoundment corroborates the

26

.OGON OOGGHDM OH

2.: :9: £388 90 owficoocom 832 5:? £on 8865 0:: Son @2748 323 wagon .3588 35m
08: 8>o mecca cow—8&8 28 3:3 @8895 59,33 muocofimwdouoozg «scuba E 35.8me d; 8:me

.9082 35m 93>

 

 

 

 

S. 8 ca 2. o
.I .I
D D DD DD-
-. .I .U a!
vltIlIIIIOIIIIIIllllllllIIUDIOOtIIlIIIIIIIIIUIIIDI IIIIIII I!
D
«cam
Emahmcion 85E

OV

_a>oEo.. 35m 23>
an ON 2.

- p p

c
o
‘3'

 

 

I
O
F

I

;

 

 

01>

D b
0B?
9%
é
SJaJeqiefi-Joroauoo % u! aouamgga

O
N

EwEucsan

27

lack of any pattern; certain zones and habitat types had higher numbers of collector-
gatherers relative to the reference while others have lower numbers. However, the
downstream zone appeared to exhibit overall lower numbers of collector-gatherers than
the reference. Riffles within the downstream zone of all rivers displayed lower levels of
collector-gatherers than the reference zone, with the largest difference being 20% in the
Tomorrow River, 20 years post-removal. Similarly, 6 out of the 8 rivers maintained
lower levels of collector-gatherers in run habitat, including rivers 33 and 40 years post-
removal.

Table 1.4. Percent contributions of F FGs to dissimilarity between runs in reference zones
and impacted zones within the Hersey River (<1 year post-removal) and the Pine River (3

years post-removal). CG = Collector-Gatherer; CF = Collector-Filterer; Pr = Predator; Sc
= Scraper; Sh = Shredder.

 

 

 

 

Hersey River Pine River
% F FG Impoundment Downstream Impoundment Downstream
CF 15 16 14 19
CG 24 14 24 36
Pr 26 26 24 1 1
Se 18 22 21 13
Sh 17 21 17 22

 

Patterns in the difference between percent scrapers in the reference zone and the
impacted zones appeared to occur as rivers progressed through the stages of recovery.
Habitat factors and the interaction between time since removal and habitat were removed
from the model because they were not significant, which resulted in significant negative
relationship between the time since removal and the difference in percent scrapers

between the reference zone and both the impoundment and downstream zones (Figure

1.10; Impoundment: R2 = 0.30, F1,13 = 6.68,p = 0.03; Downstream: R2 = 0.29, F1,” =

28

.OCON

8:888 05 55 8388 mo owficoohoa 826— 5; 38¢ 88%.: on: 88 35% 323 $58 .3388
85m 08: B>o mecca 8:888 28 meson 368:: 5833 8388 8083 E oozes-ca .3 ._ 858m

.3058 35m 83>

 

 

 

 

O? on ON Or 0
D
O 5
D
C D C
..................................... PD.-
.
D G
.l
D
mczm
Emomeson

mop—hm

_m>oEm.. 35» 83>

 

 

 

 

ow on cm 2. c
. _ p . ON!
0
D
o 5 2.
0 Q
D D
............................... D5555: 5 o
O D O
D
5 3.
on
D
O «cope—.588.

SJSdBJOS % u! aouaiayga

29

4.79, p = 0.05). When both impacted zones and habitat types were combined, a clear
negative trend existed (Figure 1.11; R2 = 0.34, F137 = 13.77, p < 0.01), indicating a

proportionate decrease in the amount of scrapers in the impoundment and downstream
zones as these zones recovered.

Percent differences in taxa richness between impacted zones and reference zones
generally decreased as time since removal passed (Figure 1.12). Within the former
impoundment, richness in both riffle and run assemblages was lower than the reference
zone for the rivers 1 and 3 years post-removal. Percent differences in richness followed a

logarithmic trajectory for assemblages in both habitat types within the former

impoundment (R2 = 0.54, F1,13 = 14.88, p < 0.01), with virtually no difference in

richness by the 7‘h year post-removal. No clear trend in differences in richness existed

with riffle assemblages within the downstream zone (Figure 1.12). Run assemblages in
the downstream zone within the first 3 years following removal exhibited differences in
richness of more than 60% relative to the reference zone, yet taxa richness was virtually
the same by the 7th year post-removal. Similar to the former impoundment, recovery of

richness within run assemblages in the downstream zone followed a logarithmic pattern
(Figure 1.12; R2 = 0.65, Fm = 10.99, p = 0.02).

Within the former impoundment of the Hersey River (< 1 year post-removal), run
assemblages and riffle assemblages exhibited 7 and 2 fewer taxa, respectively, than the
reference zone. Within the downstream zone of the Hersey River, run assemblages
exhibited 7 fewer taxa than the reference zone. Selected families found within the

reference zone but not within the impacted zones of the Hersey River included Elmidae

3O

0:3 8:888 8: :8: 80:88 mo $8580: 8032
5:5 38: 833:5 0:: 80: 3:3: 323 mwfiom 80:58:”. 8:0: E: .085 “85:: £3 55, .9688
3:6 0:5 80>: 3:0: 3:888 98 8:8 383:: 5833 80:88 8080: E 85:55 A _.~ PEME

_m>oE8 35m 83>

3. an ON 2. c
b L . . ON:

 

.

l
O
‘1-

 

C
I
O
F

O
SJGdBJOS % u! aouaJama

S: v q .43 u mm 5: 5 ma .5. > O

 

 

 

ON

31

.905: ES: 0:2 8080056: 0:0 20:0: O03 005: 0:0 0:3 0:083:26:

Sm 06:88 00:6 800% 0:0 00053: :003000. 0:306:38 0::fit0w2 0:00.898 00:: “om—ow 0:00
8:888 05 :05 00:3 .Em 0:080: 8032 ES» 0080 80065 0:: 800 00:00: 26—09 8:6m .3688 3:8
0:5 :06 00:00 00:80.3: 0:0 00:00 0800::: :002509 80:58 003 5 00:88.50 0:080m .3 ._ 8amE

 

 

—N>OEGL 00C.” flflw> _N>°EO.— GOG—m fififi>
O? O” ON O—. O 0% On ON Ow O
D DD
. Q n . . . . > I . I O9-
NC 0 I 00 O INT-s COG—N mN + V DO: I P0.0 V Q .mm.o "mm .AXvC_N.Q—. + ”mel H >

 

not

 

 

 

 

 

 

85¢ D
Emmbmcioo 02E: 0 0:05—0:3an—

SSOUlIO!J 2X21 U! GOUOJGJJ") lUBDJOd

32

larvae, Tricorythidae, Perlidae, Glossosomatidae, Ephemerellidae, and Tipulidae (Table
1.5). Within the former impoundment of the Pine River (3 years post-removal), run
assemblages and riffle assemblages exhibited 6 and 8 fewer taxa, respectively than the
reference zone. Within the downstream zone of the Pine River, run assemblages
exhibited 6 fewer taxa than the reference. Selected families found within reference zone
but not within the impacted zones of the Pine River included Brachycentridae,
Heptageneidae, Elmidae larvae, Tipulidae, Ephemerellidae, and Athericidae (Table 1.5).
Differences in percent EPT taxa between impacted zones and reference zones do

not follow any significant pattern as time since removal increases (Figure 1.13;

Impoundment: F1,13 = 0.67, p = 0.43; Downstream: F1,” = 0.00, p = 0.95). Within the

impoundment any differences in percent EPT taxa did not exceed 20% and the majority
of habitat types exhibit higher levels of percentage of EPT taxa than the reference zone.
Within the downstream zone the majority of habitat types of all rivers displayed higher
percentages of EPT taxa than the reference as well. The notable exception being runs in
the Pine River which had a difference of 40% EPT taxa from runs in the reference zone.
Percent differences in macroinvertebrate densities between impacted zones and

the reference zones revealed mixed patterns. Within the impoundment initial linear

models were not significant (Figure 1.14; F3,“ = 0.36, p = 0.78), yet visual analysis

indicated Turtle Creek (7 years post-removal) greatly deviated from an apparent trend in
the other rivers. When Turtle Creek was removed from the analysis, both

macroinvertebrates in both habitat types exhibited a strong positive relationship of

. . . . . . 2
1ncreasmg densmes over tlme when compared the reference zone (F 1gure 1.14; R = 0.70,

33

 

 

 

085520: 822080 03808 8508: 088808
000:0:80zm 0005808085 00330808 0005880.»: 00058080an
00068002 00068003. 0000088m 00058080088 0000886
08:208an 8888: 08:28.08: 30008 8380 80800: 8388:
000306000F 00088:. A8 000.85 0006:03 002000800 000:8:- 0000800
00:85. 0000:0880: 0800833 00050800 00280???” 000008 0002095.
88380 82806 888? 080508: 382288: 9028020 3808000
0008:00308m 0005508005 000_0:0w0a0m 08000803; 000008000805 000530008H 0005800ch
8 0805: e 2850 :0 03060: 8208 580200: 88282 008382
0820088: 288: 8088308: 88:88... 8200: 8 3280 e 80.80
005E 80% 80M 0083 mama 80% 80M
0869:: 8000830Q 08085 80808309 08085 8080:30Q 80808685

 

m

_

.0w000 2000 080808 AS 0:0 .0w000 00:: 0800008 Ev .033 030— 0800008 8 .0000: 0000 800850

HmOE GDP .m~N>OE®.— Gav HGDUDH “mos 025 03a wow wee—ON flow—OMAHA“: Gm Una—Om HOA— DHO SOP—3 mDfiON UOGOHOMOH a: Ugom «Nah. .2 ofifiwrfi

34

.0800 008080008 05 8080 0000 Pam 0800808 8030— 23
00080 0000202 08: 0800 008000 3209 0020A ._0>0808 0020 088 33 00800 008080008 080 00800 0000088
800380; 0800 Ahmmv 0800808089 080 “80080005 800808080805 80808 8 008080.20 .24 088me

 

 

 

 

 

 

 

 

.0580. 0050 £00> .0580. 0050 £00>
00 on an 2. o 00 an an 0.. o
8 p . . 8 . r _ cm!
0 - - 3..
I I 00.
I I ONI
m
D I 0....
O
............. O’D.s DD--.JH O
D B
O D
O O D 1 u r or
D J o D
p, - 00
on
08:”. D
500808300 0:08.08:an—

meE 0

exe; .ldEl % u! aouaJagga

35

\P

“

 

.0800 008080.008 08 808 0020800 8030— .23 00080 00002082 08: 0800 000000 090—00 020m ._0>0808
00820 08: 896 00800 008080.008 080 00800 00000082 800300; 0020800 82 008080.20 8080A .3 A 088wE

_0>oE0.. 0050 £00>
av on ca 3.

D b b I

 

00.0 n q .000 H mm 00.0 + 0.00- n >

0

L

i\
\\

 

 

 

80030832...

.0580. 0050 200>
00 cm ON 2.

D D b P

 

5.0 v 0 .000 "mm 00.0 + 0.00. n 0

D

\\

A:
o
o
‘7

Mgsuap u! aouaJaggp weaned

I.
O
In

I

l
O

ram

“03

 

 

08:”. 9
005?. 0

0808080008.

36

F 3,9 = 8.80, p < 0.01). Within the downstream zone, the full model was significant (R2 =
0.56, F110 = 4.33, p = 0.03) which indicated that habitat and year since removal by

habitat interaction were also significant (F 1 = 10.27, p < 0.01 and F1 = 8.78, p = 0.01,

respectively). The significant habitat and year since removal interaction indicated
different relationships of density recovery over time for runs and riffles downstream of

the former dam. Linear regression revealed a significant relationship between time since

removal and percent difference in density of the downstream runs (Figure 1.14; R2 =
0.55, F1,6 = 7.38, p =0.03) yet the relationship in riffles was not significant (R2 = 0.48,

F1,4 = 3.68, p =0.13).

DISCUSSION

In this study, I examined the ability of mesohabitat and macroinvertebrate
communities to recover following dam removal. Patterns in mesohabitat recovery failed
to emerge as a function of time since removal and suggest that responses may depend on
numerous situation specific attributes. Dissimilarity of the macroinvertebrate community
and differences in richness compared to the reference zone were relatively high for 3
years following dam removal, yet were minimal by the 7th year. Some components of the
macroinvertebrate community followed linear trends in recovery yet time frames appear
to be much longer than other research has suggested, specifically densities and the
proportion of scrapers. Macroinvertebrate assemblages appeared to recover at slower

rates in run habitat than riffle habitat, which has significant implications for run

37

dominated systems. Generally, important attributes of the macroinvertebrate community
appear to recovery 3- 7 years following darn removal while still others remain low after
decades have passed.

Pool-riffle-run habitat showed varying levels of recovery without a discemible
trend. Mesohabitat diversity appeared to be low in many of the impacted zones even
after 40 years with the exception of Turtle Creek that showed high diversity after 7 years.
Additionally, along with the downstream zone of Hersey River (< 1 year post-removal),
Turtle Creek appeared to be less dominated by run type habitat relative to the
impoundment than many of the other study rivers. The low diversity and extended
dominance of run type habitat within many of the impacted zones suggest that some
components of habitat only partially recover following dam removal. Failure to recover
in some downstream zones could be attributed to the presence of impoundments
downstream, which was the case for the Pine River (3 years post-removal), the Au Sable
River (14 years post-removal), and the Flat River (33 years post-removal). If the
downstream zone of the Flat River was excluded fiom consideration, the difference in
diversity appears to be greatly reduced after 20 years.

With the exception of the Au Sable River, substrate coarseness eventually
recovered, yet not until 17 years following removal (Figure 1.4). The Au Sable River
was much more dominated by sandy substrates throughout the former impoundment and
downstream zone than the reference zone. This could be attributed to natural geological
differences or possible differences in gradient within the impacted zones compared to the
reference zone. If the impacted zones have relatively low gradient the ability of the river

to transport sediment may be inhibited. Additionally, the Salling Dam on the Au Sable

38

River, was 5.2 m (Table 1.1), which was the tallest dam within our study. The impacts
due to the original dam construction could have hindered the river’s ability to recover
more so than the other study rivers.

Generally, the macroinvertebrate taxonomic structure and differences in taxa
richness in impacted zones appeared to become similar to unaffected areas 3 - 7 years
following dam removal. These findings are contrary to some previous studies which
found assemblages within the former impoundment and downstream zones
indistinguishable from those in a reference site after only 1 year following a dam removal
(Stanley et al. 2002). Similarly, Thomson et al. 2005 found assemblage structure
downstream of a dam removal to be similar to an upstream reference site afier less than 1
year, yet sampling was conducted only in riffles. Within this study, riffle habitat
appeared to recover more rapidly than run habitat, especially downstream of the dam
removal where no differences were obvious in less than one year following removal.
Run habitat showed lower similarities within both impacted zones and appeared to
require more time for macroinvertebrate assemblage structure to recover.

Differences in taxonomic richness were also most prevalent in run type of
impacted zones compared to reference zones. The Hersey River (<1 year post-removal)
and the Pine River (3 years post-removal) exhibited the largest differences in taxa
richness, indicating that dam removal does not immediately restore families lost either by
the damming or the removal itself. However, by the 7th year after dam removal, richness
differences between the reference zone and the impacted zones in Turtle Creek were
negligible. It appears that lower richness in impacted zones soon after dam removal

could be attributed to the lack of more sensitive families. Some of the more common

39

families found within the reference zone yet not in the impacted zones associated with
recent removals included Brachycentridae, Ephemerellidae, Athericidae, Perlidae, and
Glossosomatidae which are all considered relatively sensitive taxa to numerous forms of
pollution (Carter et al. 2006). These more sensitive taxa may not appear for up to 7 years
following removal especially within runs, which has important implications for
monitoring efforts focused on sensitive groups of macroinvertebrates.

The discrepancy between recovery rates for macroinvertebrates between riffles
and runs is interesting and noteworthy. Run communities appeared to recover from
removal disturbance slower than riffle communities. Most of the limited research
regarding macroinvertebrates response to removal has focused on cobble and riffle
habitats or used sampling equipment restricted to shallow depths (Pollard and Reed 2004;
Thomson et al. 2005). Stanley et al. (2005) sampled impounded reaches, presumably in
run type habitat, yet only investigated coarse level assemblage and biotic integrity
responses and found that communities become more lentic than lotic following removal.

Run type habitat is often overlooked in sampling because of the inherently higher
levels of biomass and richness associated with riffles. Given the predominance of run
habitat in rivers within our study region, the composition of macroinvertebrates within
runs is important to consider. High water velocities within riffles facilitate rapid removal
of deposited sediment while runs are subjected to more deposition due to lower
velocities. While macroinvertebrates may be immediately displaced from riffles
following the pulse of sediment from a dam removal downstream, they appear to be able
to return shortly. This research suggests the runs may recover at slower rates than higher

velocity habitats.

40

The negative trend in percent of scrapers over time was interesting in that it
applied to both habitat types and zones. Initially following dam removal, the proportion

of scrapers appeared to be higher than that of the reference. These differences declined

. . . th .
over time, becomrng the same as the reference zone 1n the 12 year, and continued to

lower levels of scrapers than the reference zones in the following years. Scrapers feed on
algae which prosper in open canopies, similar to what would be found within a recently
drained impoundment. As the extent of riparian shading develops from vegetation
growth the dominance of scrapers would decline in the former impoundment. Orr and
Stanley (2006) showed that former impoundments become revegetated rapidly after dam
removal yet the frequency of trees remains low for decades. Mechanisms explaining this
decrease in scrapers over time since removal in downstream zones are not the same as the
former impoundment and the cause is unclear. The role of riparian plantings should be
considered to facilitate more rapid recovery of the stream-side canopy. Accelerating the
riparian development of a formerly impounded zone could additionally buffer against
excessive warming and enhance bank stability, thus reducing the influence of slumping
banks contributing to downstream sediment transport.

With the exception of Turtle Creek, macroinvertebrate densities within the former
impoundment did not recover for nearly 33 years, much longer than anticipated by other
studies (Doyle et al. 2005). Downstream of former dams, runs responded similarly to the
former impoundment, recovering after 25 years. This extended time may be
overestimated due to the strong influence of the two most recent removals, which show

greatly reduced densities. Riffles downstream of the dam returned to densities at or

41

above the reference zone within a year of dam removal, similar to the predictions by
Thomson et al. (2005).

Answering complex ecological questions on long-term time scale is difficult due
to the logistical constraints of funding and foresight and the current need for management
recommendations. The space-for-time approach I used was valuable, yet has limitations.
Pickett (1989) describes the past successes of space-for-time substitution studies
especially when attempting to describe general patterns and generate working
hypotheses. I believe this study has done just that, in demonstrating general patterns of
recovery of mesohabitat and macroinvertebrates following dam removal. The largest
limitation to the approach employed here is the assumption that all of the study rivers
have been influenced equally over time by external sources. We have attempted to
minimize the role of external influences here by only comparing within rivers to generate
the points in time and not treating the rivers as replicates. We also assumed that any
external influences on the river acted equally on the reference zone and the impacted
zones. The role of external influences carmot be removed, and is an inherent problem in
space-for-time substitutions.

A river’s ability to recover from massive disturbances like past land use and dam
removal depends on numerous factors. Variation within and between rivers is a major
concern when making general statements about macroinvertebrate recovery following
darn removal. The rivers within this study vary substantially in size, location, gradient,
dam size and type, geology, and watershed characteristics. The number of rivers which
met the numerous logistical criteria and desired gradient of time since removal was

surprisingly small. Ideally, our study would have incorporated numerous replicates with

42

equally, well distributed times between removals, yet this was not possible. Specifically,
further investigation between the years 4 and 13 years following removal is worth
consideration. I recognize the localized differences between rivers yet our results should
hold true as long as within river variation is not greater than between. Multivariate
ordination and clustering supported the premise of our approach by revealing the tight
grouping of samples by river except those which had been recently affected by dam
removal.

Considering the natural variability of rivers is an integral component of accurately
interpreting results. How a river responds to large-scale disturbances such as darn
removal will depend on the specific situation. Within this study, we documented some
surprising responses which can be partially attributed to river variation. Downstream of
the former Hersey Dam on the Hersey River, the mesohabitat was highly heterogenous
with more riffle habitat than the reference zone, even after only 8 months following
removal. This section of the river happened to have high gradient and comprised the last
2 km of the river before entering the much larger Muskegon River. Alternatively, the Au
Sable River maintained relatively homogenous habitat even afier 14 years. The former
impoundment and downstream zones appeared to be wider and lower gradient than the
reference zone. Natural longitudinal changes in ecological parameters of rivers (V annote
et al. 1980) must be acknowledged as potentially muddling the interpretation of our
results although, the majority of longitudinal changes occur on a much larger scale than

the 5-9 km reaches this study focused on.

43

CONCLUSIONS

Despite some of the limitations involved with the approach employed by this
study, some informative long-term patterns of recovery were clarified. Before this study,
research on macroinvertebrate and physical habitat responses had only been conducted
focused on shorter time frames, usually one river, and limited sampling effort restricted
to riffle habitats. This work is the first to address long-term questions, including
numerous rivers, and incorporating multiple habitat types. Doyle et al. (2005) pose two
models of ecological recovery following dam removal (full or partial) and predict the
general time frame needed for various parameters to recover. I believe
macroinvertebrates may only partially recover and require much longer than anticipated,
on the order of decades for certain parameters. Mesohabitat recovery seems to vary
greatly by river and may largely depend on site characteristics like dam size, gradient,
and the presence of other dams downstream. River variation will play a vital role in
forecasting a river’s response to dam removal. Certain rivers have much higher
restorative capacity than others and a valuable research approach would attempt to
identify those differences. Dam removal can not be prescribed for all rivers as an
effective and responsible restoration approach, especially considering the plethora of
social and economic issues involved (Born et al. 1998).

Considering the long-term ecological implications of dam removal is necessary
for responsible and successful management. Managers will continue to face difficult
decisions regarding the fate of aging dams and having realistic expectations of recovery
levels and rates will ensure the long-term success of removal as a viable restoration

technique. As with any restoration, the need for some basic level of evaluation is

44

paramount. Given the dearth of funding for stream restoration, it is understandable that a
decision must be made between monitoring and additional restoration. This situation is
unfortunate, yet must be addressed. Easy approaches to evaluation exist and should be
explored. Simple efforts such as community-based volunteer macroinvertebrate
monitoring or site visits with visual documentation are quick and economically accessible
approaches to coarse levels of evaluation. If removals continue with little to no
evaluation, nearly all of our region’s rivers could only partially recover while the
possibility of full recovery may exist. Given the variability of rivers, we will never be
able to perfectly predict how a river will respond to any restoration technique yet if we
learn from removals with monitoring and adjust approaches accordingly, we can

continually improve our restoration efforts and strive to reach full recovery.

45

APPENDIX A. Maps of mesoscale habitat in 8 study rivers. Number of years since
removal in parentheses next to river name. Lengths start at upstream extent of
reference zone.

 

 

 

 

    

Hersey River (< 1) $223.5...
Reference Impoundment Downstream 2::
l e
2 . . ,1, DPool
" III III
I I ‘ I
g1 I. I IIIIIIIIIII ‘IIIIIIIIIIIII IIIIII IIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIII
: I I‘ "IIIIIIIIIIIIIIIIIIIIIIII III "III II
§ 1 [I , I IIIIIIIIIIIIIIIIIIIIII IIIIIIIIIII
3 I II I III. g IIIIIIIIIII IWIII III
E 5 I: .' I II III I "III "IIIIIIII IIIIIIII II
, ., I,“ . , . I , 'IIIIIIIIIIIIIIIIIIIIIII
0 , | . I III I" -.
o 1250 2500 3750 5000
Length (m)

Pine River (3)

 

 

 

 

  
     
 
 
  

  
        

   

 

 
         

    

I Rapids Downstream
50 ‘ IICompIex 'mpwndment IIIiIIIIIIIIIIIIIIIIIII
45 ‘ Run Reference "t “HI“IIIIIIIIIIIIIIII
40 J :Ruffle iIIII' Iii r‘flllIIIIIIIIII
A _ P I I
g 35 .. I IIIIIII.
a 3°" IIIIIIII
0 ‘ I IIIII
E 25 I ,IIImI WiIIIIIIIIIII"
5 20 II: III" I I - III "I- ’" I I “Wm""m'
£ 15 ‘ ""I'IIIIIiIi""'IIIIIII" III" I III Iiim"
10 _ IiIII'I'IIiIIIiIIIIIIII I "I“ I I "I" III “ iiiIiiIiI
I ‘III II I I ' I. ' . I
g "I" IIIIII"I"' III I II""" 1 I i" III II III ‘IIIIIIIII
0 1 250 2500 3750 5000 6250
Length (m)

46

APPENDIX A (cont’d).

 

Turtle Creek (7) I Complex
'II' Run

   

 

 

 

I R11!
45 Reference Impoundment :1 pI 'e
40 0°
35

    

Mean width (m)
N
o

 

0 I I
0 1250 2500 3750 5000 6250
Length (m)
Au Sable River (14)
30 III Run
Impoundment I Riffle
25 Reference El Pool
A D tree
5 20 owns m
i .5
c
g 10
5
0
0 1250 2500 3750 5000 6250 7500

Length (m)

47

APPENDIX A (cont’d).

 

Kickapoo River (15) I Rapids
l Complex
20 Reference Impoundment Downstream i“ Run

 

 

 

Mean width (m)

 

0 1250 2500 3750 5000 6250

Length (m)

Tomorrow River (20)

  
    

Impoundment

Mean width (m)

0 1250 2500 3750 5000
Length (m)

48

APPENDIX A (cont’d).

   
 
   

 

 

 

 

 

 

 

 

 

    
    

Flat River (33) ' C°mp'°x
Run
I Riffle
60 Reference Impoundment Downstream DPOOJ JJJJJJJJJJJJJJJJJJ
A :3 JJJJJJ 111111 JJJJJJJJJJJJJJJIJJI‘ JJJJJJ 1.. JJJJJJJJJJJJJJJJJJJJJJJJI111|J|1J11111111111 J J JJJ
E. J... 1.111111111111111111 JJJ JJJ111111111111111 JJ 11 1111111111 JJJJJJJJ JJJJ 1 11‘ ‘
g 30 11111111111111111111111111111111 JJJ 11111111111111! JJJJJJJ J111111JJ' JJJ JJJJ1JJ1 J ' 'J
;- 20111111111111111 1 1111111111111111111'1JJ111111 1“ 1 J J
g JJJ. J111‘11111J1J1111J1J1JJJ1JJJ JJJJ JJJJJJJJJJJJJJJJ 'JJJJJIJIIIJ IJJJJJJ 1JJJJJJJJJJ JJJJJ J11
: 10 JJJ JJJJJJJJJJ11111111 11 11111111111111 JJJJJJJ JJJJJJJJJ J JJJJJJJJJJJJJJJJ11
0 11111 1111111111111111111111 111 JJJJJJJJJJJJ JJIJIJJJ JJJ 1|
0 1250 2500 3750 5000 6250
Length (m)
Looking Glass River (40)
[Complex
40 Reference Impoundment Downstream 11111 Run
35 '
.RIffle
A 30 [JPool
E 25 "' 1J1|11J1""""JJ11|1111' J “J
5 2011111'11111111111J11JJ1| JJJJJJJJIJJ JJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJJ
E i11J1111J1JJJJJJJJJJ ll JJiiJJ JJJJJJ JJJ] JJJ JJJJJ JJJJJJJJJJJJJJJ
3 15 .J‘JJJJJJJJJJJJJJJJ'JJ111J-JJJ1111JJJ 11JJ1J111JJJJJJ11JJJJJJJJJ
: 1111111JJ1JJJJJJJ ll JIJJJJJJJ IIJJ JJJJ JJJJJJJJJJJ J
:10 1111111J11 J" 11 1J'1111111 "JJJ 1"" JJJ JJJ" JJ'J1 'J'JJJJ
: itJJJJJIJJJ 1111 JJJ'J ' J
3 11.11.1111 11, 1'11 1111 , 111111111J
0 1 250 2500 3750 5000 6250 7500
Length (m)

49

APPENDD( B. Riffle assemblage macroinvertebrate data collected from 8 rivers in
various stages of recovery following dam removal. Heading numbers indicate years
since removal. Data are sum of three samples within each zone (Ref = Reference; Imp =
Former impoundment; DS = Downstream). Taxa sorted by abundance for all rivers.

 

 

 

l 3 7
Taxon Name Ref Imp DS Ref Imp Ref Imp DS
Chironomidae 468 470 456 206 16 566 817 886
Hydropsychidae 490 32 687 4 0 82 269 195
Elmidae (larvae) 78 7 143 69 0 23 1 17 75
Baetidae 40 7 121 74 8 203 304 506
Brachycentridae 0 0 0 158 0 0 0 0
Caenidae 0 1 4 O 0 121 48 37
Amphipoda 21 3 0 0 0 6 5 0
Heptageniidae 0 4 34 0 l 14 20 69
Polymitarcyidae 0 0 O 0 0 89 146 161
Unknown pupae 6 20 18 29 1 59 43 77
Simuliidae 292 4 18 0 91 4 23
Tipulidae 5 53 109 O 3 8 0
Tricorythidae 12 2 l3 1 0 0 16 26
Elmidae (adults) 42 5 1 15 13 4 7 5 29
Oligochaeta 27 8 2 0 0 l 1 2
Isopoda 12 6 6 0 0 0 0 8
Trichoptera pupae 8 l 4 6 0 O 8 12
Perlidae 4 0 0 0 O 32 72 21
Psychomyiidae 0 24 107 0 0 0 0 O
Hydracarina 2 0 0 24 l 0 1 O
Isonychiidae 0 0 1 2 O O 0 0 5
Glossosomatidae 16 0 74 0 0 0 0 0
Ephemerellidae 0 8 l9 0 9 0 5
Athericidae l 1 66 O 0 0 0
Simuliidae (pupae) 24 1 5 O O O 0 0
Philopotamidae O 0 0 O 0 0 0 0
Hydroptilidae 0 0 20 8 0 0 4 O
Psephenidae (adults) 0 0 0 0 0 0 O 0
Phryganeidae 4 0 0 0 0 0 0 4
Unknown adult 0 0 0 0 0 0 0 0
Cambaridae 2 0 0 0 0 4 0 0
Uenoidae 12 0 0 0 0 0 0 0
Corydalidae 0 0 0 0 0 0 O 0
Hirudinea 0 0 0 0 O 0 0 0
Potamanthidae 0 0 0 0 0 4 8 9
Leptophlebiidae 0 O 0 0 0 2 6 0

 

50

APPENDIX B (cont’d).

Taxon Name
Tabanidae
Empididae
Leptoceridae
Pyralidae
Pteronarcyidae
Siphlonuridae
Ceratopogonidae
Capniidae
Chrysomelidae (larvae)
Culicidae

Sialidae

Corixidae
Lepidostomatidae
Nemouridae
Stratiomyidae
Coenagrionidae
Haliplidae (larvae)
Hydrophilidae (larvae)
Limnephilidae
Rhyacophjlidae
Sisyridae

Total of zone

Ref

D
co

Ref

5

OOOOOOOOONOOOOOOOOOOO

H
OOOHOOOOOOOOOOOOOOOOOQF‘
OOOOOOOOOOOOOOOOOOOO-fi
OOOOOOOOOOOOOOOOHOOOO
OOOOOOOOOOOOOOOOOOOOO

DJ
.—

1571 650 1961 685

51

Ref

OOOOOOOOOOOOOONOOOOOO

1318

Ex,

ooooooooooooooo-hoooo

O

1906

U
m

OOOOOHOOOOOOOOOOOOOOO

2151

APPENDIX B (cont’d).

 

15 20

Taxon Name Ref Imp DS Ref Imp DS
Chironomidae 5 8 1 396 690 800 679 604
Hydropsychidae 138 12 150 60 12 2
Elmidae (larvae) 211 13 142 292 52 93
Baetidae 250 95 166 136 351 111
Brachycentridae 1 16 10 46 272 181 1 150
Caenidae 17 18 100 O l 16
Amphipoda 60 37 140 600 161 271
Heptageniidae 10 6 18 0 0 8
Polymitarcyidae 0 0 O 0 O 0
Unknown pupae 50 6O 3 8 44 50 80
Simuliidae 38 4 14 20 35 16
Tipulidae 28 3 0 164 52 19
Tricorythidae 0 1 0 0 O O
Elmidae (adults) 0 2 16 20 16 5
Oligochaeta 33 39 26 12 0 O
Isopoda 3 2 0 8 21 61
Trichoptera pupae 0 2 0 52 99 12
Perlidae 10 4 4 0 0 O
Psychomyiidae 12 1 O 4 0 0
Hydracarina 6 O 12 52 8 32
Isonychiidae 12 1 104 0 0 O
Glossosomatidae 0 0 O 1 2 8 0
Ephemerellidae l l 4 4O 2 1
Athericidae 4 0 0 0 0 4
Simuliidae (pupae) 8 0 2 4 44 2
Philopotamidae 0 0 0 0 0 O
Hydroptilidae 2 0 O 0 0 0
Psephenidae (adults) 1 O O O O O
Phryganeidae O 1 0 0 0 0
Unknown adult 4 O 0 0 0 l
Cambaridae 0 0 0 O 0 O
Uenoidae 0 O 0 0 O O
Corydalidae O O 0 O 0 O
Hirudinea 0 0 0 O 0 O
Potamanthidae O 0 O 0 O 0
Leptophlebiidae 7 1 0 O 0 0

 

52

APPENDIX B (cont’d).

H
kl!
N
CD

Ref

8
U
a)
3
D
on

Taxon Name Ref
Tabanidae

Empididae
Leptoceridae

Pyralidae
Pteronarcyidae
Siphlonuridae
Ceratopogonidae
Capniidae
Chrysomelidae (larvae)
Culicidae

Sialidae

Corixidae
Lepidostomatidae

OOOOOOOOONOOOO

Nemouridae
Stratiornyidae
Coenagn'onidae
Haliplidae (larvae)
Hydrophilidae (larvae)
Limnephilidae
Rhyacophilidae
Sisyridae

Total of zone

Soooo~oooooo>oo~o~oooo
SoooooCOOOOAOAhoooooooo
joooooooooooooooooooao

Ov—OOOOONOOOOOOOQOOOOOO

goooocooooocoooooooooo
N
—n

~OO~OOOO
N
N

fl
\1
N
p—n

6 2505

~

53

APPENDIX B (cont’d).

 

 

33 40
Taxon Name Ref Imp DS Ref Imp DS Total
Chironomidae 257 275 406 135 106 33 8814
Hydropsychidae 5 54 306 954 2 89 608 93 4844
Elmidae (larvae) 365 200 499 221 291 110 2891
Baetidae 106 77 132 38 108 11 2833
Brachycentridae 19 29 0 0 2 O 1983
Caenidae 200 162 94 95 65 128 979
Amphipoda 12 26 1 1 15 2 1359
Heptageniidae 117 131 114 267 53 101 866
Polymitarcyidae 293 96 73 12 7 10 877
Unknown pupae 20 22 44 5 14 4 680
Simuliidae 26 1 3 l O 0 O 603
Tipulidae 9 20 62 2 4 2 544
Tricorythidae 17 19 0 14 34 18 155
Elmidae (adults) 11 38 31 22 6 13 387
Oligochaeta 2 5 0 35 23 34 216
Isopoda 4 2 0 1 1 74 24 2 18
Trichoptera pupae 8 12 13 4 O O 241
Perlidae 5 0 0 4 38 9 194
Psychomyiidae 0 0 22 5 3 0 178
Hydracarina 13 4 13 0 l 0 169
Isonychiidae 2 2 24 O 0 3 1 62
Glossosomatidae 5 9 O 8 2 3 134
Ephemerellidae 0 0 24 O 0 0 l 18
Athericidae 0 O 0 0 0 0 76
Simuliidae (pupae) 0 0 0 0 O 0 9O
Philopotamidae 0 8 47 O 4 l 59
Hydroptilidae 0 10 3 0 12 0 59
Psephenidae (adults) 0 6 4 20 18 8 49
Phryganeidae 0 O 0 0 0 0 9
Unknown adult 23 4 4 2 0 O 3 8
Cambaridae O 7 4 10 3 6 3O
Uenoidae O 0 0 0 4 20 16
Corydalidae 5 5 6 O 0 O 1 6
Hirudinea 0 0 O 30 O 0 3O
Potamanthidae 0 0 0 O 0 0 2 1
Leptophlebiidae 0 O 0 O O O 16
Tabanidae 0 1 O 0 0 O 13
Empididae O 0 0 O O 0 12

 

54

APPENDIX B (cont’d).

 

Taxon Name

Ref

fit»)
DJ
8

U
03

Ref

A
O

DS

Total

 

 

Leptoceridae
Pyralidae
Pteronarcyidae
Siphlonuridae
Ceratopogonidae
Capniidae
Chrysomelidae (larvae)
Culicidae

Sialidae

Corixidae
Lepidostomatidae
Nemouridae
Stratiomyidae
Coenagrionidae
Haliplidae (larvae)
Hydrophilidae (larvae)
Limnephilidae
Rhyacophilidae
Sisyridae

OOOOOOOOOOOOOOOOJ’BH

O

OOOOOOOOOOOOOOOOONDJ

OOOOONOOOOOOOOOOAO

O

OOOOOOOOOOOOOOOOOO

O

H
ooooooomooooooooohg

p—s

OOOOOOOOOOOOOO—‘OOOO

H
COO

~H-HNNNNA-Ahb50\\o

n

 

Total of zone

2078

1494

2581

1230

1502

634

30046

 

55

 

HM-

 

 

 

APPENDIX C. Run assemblage macroinvertebrate data collected from 8 rivers in various
stages of recovery following darn removal. Heading numbers indicate years since
removal. Data are sum of three samples within each zone (Ref = Reference; Imp =
Former impoundment; DS = Downstream). Taxa sorted by abundance for all rivers.
1 3
Taxon Name Ref Imp Ref Imp
Chironomidae 763
Caenidae 0
Baetidae 1
Elmidae (larvae) 9
Amphipoda 5
Hydropsychidae 23
Brachycentridae 0
Tricorythidae 3
Oligochaeta 6
Unknown pupae 10
Polymitarcyidae O
Isopoda 22
Heptageniidae 0
Tipulidae
Corixidae
Hydracarina
Simuliidae
Elmidae (adults)
Athericidae
Ephemerellidae
Ceratopogonidae
Empididae
Psephenidae (adults)
Phryganeidae
Perlidae
Isonychiidae
Potamanthidae
Cambaridae
Leptoceridae
Psychomyiidae
Tabanidae
Leptophlebiidae
Unknown adult
Baetiscidae
Glossosomatidae
Uenoidae
Aeshnidae
Ephemeridae
Hydroptilidae
Lepidostomatidae
Polycentropodidae
Trichoptera pupae

U
m
U
m

 

oooo~oo-oo~o$
Kit
00
N

I—it—i

{J‘v—fi

\1

so
Soogooooooor—ooooooocw

OOOOOOOOOOOOOOOOOOOOOHO—‘OOOOOOOOOHOOOONNO

coooooooooowoooo—ooohoog~o~oo-oo§wouooggo
oooooooooooooooocoooo

OOOOOOOOOO—‘OOOOOOOOOOOOOOOOOOONONOOONOO—‘OB

OOOWOOOOOOOOOOOOOOOOOO—‘OOMN
OOOOOOOOOOOOOOOOOOOOOOOOOHO

56

 

APPENDIX C (cont’d).

 

Taxon Name

Ref

U
m

Ref

U
m

 

Corydalidae
Hydr0philidae (adults)
Perlodidae
Pteronarcyidae
Gomphidae

Hirudinea

Saldidae

Sialidae
Coenagrionidae
Dytiscidae (larvae)
Helicopsychidae
Hydrophilidae (larvae)
Libellulidae
Mesoveliidae
Notonectidae
Philopotamidae
Simuliidae (pupae)
Siphlonuridae
Sisyridae

Total of zone

OOOOOOOOOOOOWNOOOOO

885

gooooooooooooooooooogfl

57

OOOOOOOOOOOOOOOOOOO

H
N
b

OOOOOOOOOOOOOOOOOOO

p—a
\O
O\

\oooooooocooooooooooogw

OOOOOOOOOOOOOOOOOOO

U)
.—l

APPENDIX C (con’d).

 

 

7 l4

Taxon Name Ref Imp DS Ref Imp DS
Chironomidae 166 584 134 145 168 257
Caenidae 16 47 20 321 48 2
Baetidae 55 39 37 31 70 18
Elmidae (larvae) 18 31 34 63 5 8
Amphipoda 1 0 O 0 O O
Hydropsychidae 3 10 l 1 1 18 9 4
Brachycentridae 0 O 0 26 0 0
Tricorythidae O O 21 136 41 13
Oligochaeta l l 9 0 24 7 17
Unknown pupae 15 34 14 4 8 11
Polymitarcyidae 27 8 8 7 1 1 3 O
Isopoda l 2 2 1 7 0
Heptageniidae 4 22 1 0 O 0
Tipulidae O 5 O 17 1 28
Corixidae 5 0 0 0 0 2
Hydracarina 0 O 1 1 l 1 2
Simuliidae 4 0 4 6 1 10
Ehnidae (adults) 0 0 6 3 2 0
Athericidae 0 0 0 1 0 O
Ephemerellidae 2 3 1 8 0 0
Ceratopogonidae 0 0 O 0 O 1
Empididae 1 O O 4 O O
Psephenidae (adults) 0 O 0 O O O
Phryganeidae l 1 4 3 l 7
Perlidae 1 5 3 0 O 0
Isonychiidae O 1 3 4 l 0
Potamanthidae O 5 7 0 O 0
Cambaridae O 0 0 0 O O
Leptoceridae 0 0 0 1 1 l
Psychomyiidae 0 0 0 0 0 0
Tabanidae 0 O O O 0 6
Leptophlebiidae 4 1 0 3 O 0
Unknown adult 2 0 O 0 O 0
Baetiscidae 0 0 O 2 2 3
Glossosomatidae 0 O l 0 0 0
Uenoidae O 0 O 0 0 O
Aeshnidae O 0 0 2 O 1
Ephemeridae 1 O 0 O 0 0
Hydroptilidae 1 O 1 O 0 0
Lepidostomatidae 0 0 0 0 0 0
Polycentropodidae O O 0 0 0 0
Trichoptera pupae O 2 0 0 0 1
Corydalidae 0 O O 1 0 0
Hydrophilidae (adults) 0 O 0 0 0 0
Perlodidae 0 O O 0 O O

 

58

 

APPENDIX C (cont’d).

Taxon Name
Pteronarcyidae
Gomphidae
Hirudinea
Saldidae

Sialidae
Coenagrionidae
Dytiscidae (larvae)
Helicopsychidae
Hydrophilidae (larvae)
Libellulidae
Mesoveliidae
Notonectidae
Philopotamidae
Simuliidae (pupae)
Siphlonuridae
Sisyridae

Total of zone

Ref

OOOOOOOOOOHOOOOO

340

oooooooooooooooogu

889

59

U
m

OOOh-‘OOOOOOOOOOOO

377

Ref

OOOOOOOOOOOHOOOO

937

Hr—
ocoooooooooooo—‘ogb

377

U
m

OO-‘OHOOOOOOOHOOO

395

APPENDIX C (cont’d).

 

20
DS Ref Imp DS

g;

Taxon Name Ref

 

135 173 254 216
88 0 9 4
l4 13 59 28
7 25 21 46
27 122 132 162

\I
A

Chironomidae 21 1
Caenidae
Baetidae

Elmidae (larvae)
Amphipoda
Hydropsychidae
Brachycentridae
Tn'corythidae
Oligochaeta
Unknown pupae
Polymitarcyidae
Isopoda
Heptageniidae
Tipulidae
Corixidae
Hydracarina
Simuliidae
Elmidae (adults)
Athericidae
Ephemerellidae
Ceratopogonidae
Empididae
Psephenidae (adults)
Phryganeidae
Perlidae
Isonychiidae
Potamanthidae
Cambaridae
Leptoceridae
Psychomyiidae
Tabanidae
Leptophlebiidae
Unknown adult
Baetiscidae
Glossosomatidae
Uenoidae
Aeshnidae
Ephemeridae
Hydroptilidae
Lepidostomatidae
Polycentropodidae
Trichoptera pupae
Corydalidae
Hydrophilidae (adults)
Perlodidae

085368
©3035
A
0

33¢

33...

219

so—
com.—
gow
NO
\IO
Ob)

p.‘
y—s
p—:
\O
p—i
O
N
N
b»)
p—n

OOOOOOONOOOOWOOOOOOHOOOOOOOOOOOOWNO
COOOOOOOOOOOOHOOOOOOOOOHHHOOOOOOQON
OOOOOOOONOOOOOOOOOOt-‘OHF‘OHHOOONOOQ—‘ON
OOOHOWOOOOOOOOOOOOOOOOONOHA-fiOU-IOONOOOO
CHOOOOOOOO-‘OOONOOOOOOOOQObONOMOOO—I

O0OOOOOOOOOOOOOOOOOOOOONOMt—Iv-‘OUJO

6O

APPENDIX C (cont’d).

p—A
kit
N
O

U

m

Ref

H
C
CD
H

Taxon Name Ref

 

Pteronarcyidae
Gomphidae

Hirudinea

Saldidae

Sialidae
Coenagrionidae
Dytiscidae (larvae)
Helicopsychidae
Hydrophilidae (larvae)
Libellulidae
Mesoveliidae
Notonectidae
Philopotamidae
Simuliidae (pupae)
Siphlonuridae
Sisyridae

Total of zone 338

ocoé

oooooooooooooooo
oooooo~oooooooow
ooooooooooocoooo

Hoooooooo~ooooooo§
oooooooooooooooo

OOHOOOHOOOOOOO

b—l
O
O‘
|—d

343 619 597

61

APPENDIX C (cont’d).

 

 

33 4O

Taxon Name Ref Imp DS Ref Imp DS Total
Chironomidae 213 144 172 103 l 18 74 4153
Caenidae 75 97 71 70 71 218 971
Baetidae 22 73 15 6 4O 18 606
Ehnidae (larvae) 42 10 15 125 66 1 14 578
Amphipoda 7 28 2O 5 5 10 553
Hydropsychidae 40 131 32 29 13 45 432
Brachycentridae 4 7 25 0 0 0 3 9 1
Tricorythidae 1 8 1 1 1 8 6 61 107 333
Oligochaeta 7 3 26 46 81 58 327
Unknown pupae 24 4 24 5 1 1 279
Polymitarcyidae 47 7 7 2 1 1 14 266
Isopoda 1 O O 3 l 53 14 187
Heptageniidae 9 12 15 31 43 32 154
Tipulidae O 2 O O 0 1 1 1 8
Corixidae O O O O 0 O 101
Hydracarina 8 6 5 3 1 l 59
Simuliidae 0 7 7 0 0 0 49
Elmidae (adults) 0 2 0 5 8 3 44
Athericidae 0 0 0 0 0 0 37
Ephemerellidae 0 O 0 0 0 1 33
Ceratopogonidae 1 1 1 2 0 O 26
Empididae 0 O 0 0 O 0 20
Psephenidae (adults) 0 0 0 8 11 11 20
Phryganeidae 0 0 0 0 0 2 18
Perlidae 1 0 2 1 3 6 16
Isonychiidae 0 0 0 0 0 0 12
Potamanthidae O O 0 0 O 0 12
Cambaridae 0 O 2 2 7 l 1 1
Leptoceridae 0 O 0 0 8 4 1 1
Psychomyiidae 0 0 1 7 3 0 1 l
Tabanidae 0 O 0 O O 1 1 1
Leptophlebiidae 0 0 0 0 0 O 10
Unknown adult 2 3 2 0 O O 10
Baetiscidae 0 1 O 0 0 0 8
Glossosomatidae 0 l 1 1 2 O 7
Uenoidae O 0 0 3 4 0 7
Aeshnidae 0 0 0 O 0 0 5
Ephemeridae 0 O 0 2 O l 5
Hydroptilidae 0 0 O 0 O O 5
Lepidostomatidae O O 0 O 2 0 5
Polycentropodidae 0 O 0 2 2 O 4
Trichoptera pupae 0 0 0 O 0 2 4
Corydalidae 1 O l 0 0 0 3
Hydrophilidae (adults) 0 0 0 0 2 O 3
Perlodidae 0 O 3 0 O 0 3
Pteronarcyidae 0 0 O 0 O 0 3
Gomphidae 0 O l O 0 0 2

 

62

APPENDIX C (cont’d).

 

 

 

33 40
Taxon Name Ref Imp DS Ref Imp DS Total
Hirudinea O 0 0 O O O 2
Saldidae O 0 0 0 O 0 2
Sialidae 0 0 l 0 0 0 2
Coenagrionidae 0 0 O O O 0 1
Dytiscidae (larvae) 0 0 O 1 O 0 1
Helicopsychidae 0 0 0 O 0 0 1
Hydrophilidae (larvae) O 0 O 1 0 0 1
Libellulidae 0 0 O 0 O 0 1
Mesoveliidae 0 0 0 O 0 O 1
Notonectidae O 0 0 O 0 O 1
Philopotamidae 0 0 0 O O 5 l
Simuliidae (pupae) 0 0 0 0 O O l
Siphlonuridae 0 O 0 0 O 0 1
Sisyridae 0 0 O O l 1 1
Total of zone 522 550 467 497 617 745 9940

 

63

 

LITERATURE CITED

American Rivers. 2007. http://www.americanrivers.org/ [Accessed 4/21/08].

Anderson, R. O. 1959. A modified flotation technique for sorting bottom fauna samples.
Limnology and Oceanogrophy 4: 223-225.

Barbour, M. T., J. Genitsen, B. D. Snyder, and J. B. Stribling. 1999. Rapid bioassessment
protocols for use in streams and wadeable: periphyton, benthic macroinvertebrates
and fish, 2nd ed. EPA. 841-B-99-002. U. S. Envrionmental Protection Agency;
Office of Water, Washington, DC. 202 pp.

Baxter, R. M. 1977. Environmental effects of dams and impoundments. Annual Review
of Ecology and Systematics 8: 255-283.

Bednarek, A. T. 2001. Undamming rivers: a review of the ecological impacts of dam
removal. Environmental Management 27: 803-814.

Beisel, J .-N., Usseglio-Polatera, P., Thomas, S. , and Moreteau, J .-C. 1998. Stream
community structure in relation to spatial variation: the influence of mesohabitat
characteristics. Hydrobiologia 389: 73—88.

Bouchard Jr., R. W. 2004. Guide to Aquatic Invertebrates of the Upper Midwest.
University of Minnesota, St. Paul, MN.

Born, S. M., Genskow, K. D., Filbert, T. F ., Hemandez-Mora, N., Keefer, M. L., White,
K. A. 1998. Socioeconomic and institutional dimensions of dam removals: the
Wisconsin experience. Environmental Management 22: 359-370.

Bray, J. R. and J. T. Curtis. 1957. An ordination of the upland forest communities of
Southern Wisconsin. Ecological Monographs 46: 327-354.

Burroughs, B. 2007. Effects of dam removal on fluvial geomorphology and fish. Ph. D
dissertation. Michigan State University, Department of Fisheries and Wildlife.

Bushaw-Newton, K. L., Hart, D. D., Pizzuto, J. E., Thomson, J. R., Egan, J ., Ashley, J.
T., Johnson, T. E., Horwitz, R. J ., Keeley, M., Lawrence, J ., Charles, D., Gatenby,
C., Kreeger, D. A., Nightengale, T., Thomas, R. L., and Velinsky, D. J. 2002. An
integrative approach towards understanding ecological responses to dam removal:
the manatawny creek study. Journal of the American Water Resources
Association 38: 1581-1599.

Carter, J. L., Resh, V. H., Hannaford, M. J ., and Myers, M. J. 2006. Macroinvertebrates
as biotic indicators of environmental quality, pp. 805-834. In Methods in stream
ecology. F. R. Hauer and G. A. Lamberti, eds. Elsevier, Amsterdam, Netherlands.
877 pp.

Cattalano, M. J ., Bozek, M. A., and Pellett, T. D. 2007. Effects of dam removal on fish

64

 

assemblage structure and spatial distributions in the Baraboo river, Wisconsin.
North American Journal of Fisheries Management 27: 519-530.

Clarke, K. R. and R. M. Warwick. 1994. Change in marine communities: an approach to
statistical analysis and interpretation. Plymouth Marine Laboratory, Plymouth.

Cummins, K. W. 1962. An evaluation of some techniques for the collection and analysis
of benthic samples with special emphasis on lotic waters. American Midland
Naturalist 67: 477-504.

Cummins, K. W., and M. J. Klug. 1979. F eeding ecology of stream invertebrates. Annual
Review of Ecological Systematics 10: 147-172. I

Doyle, M. W., Stanley, E. H., and Harbor, J. M. 2003. Channel adjustments following
two dam removals in Wisconsin. Water Resources Research 39: 1-15.

Doyle, M. W., Stanley, E. H., Orr, C. H., Selle, A. R., Sethi, S. A., and Harbor, J. M.
2005. Stream ecosystem response to small dam removal: lessons from the ~
heartland. Geomorphology 71: 227-244. E

 

Gordon, N. D., McMahon, T. A., Finlayson, B. L., Gippel, C. J ., and Nathan, R. J. 2004.
Stream hydrology: an introduction for ecologists, ed. Wiley & Sons. England.

Graf, W. L. 1999. Damn nation: a geographic census of American dams and their large-
scale hydrological impacts. Water Resources Research 35: 1305-1311.

Grubbs, S. A. and Taylor, J. M. 2004. The influence of flow impoundment and river
regulation on the distribution of riverine macroinvertebrates at Mammoth Cave
National Park, Kentucky, USA. Hydrobiologia 520: 19-28.

Hanshue, S. 2006. Michigan dam removal case studies. Michigan Department of Natural
Resources, Fisheries Division, Habitat Management Unit.

Hart, D. H., Johnson, T. E., Bushaw-Newton, K. L., Horwitz, R. J ., Bednarek, A. T.,
Charles, D. F., Kreeger, D. L., and Velinsky, D. J. 2002. Dam removal:
challenges and opportunities for ecological research and river restoration.
Bioscience 52: 669-681.

Heinz Center, The. 2002. Dam removal: science and decision making, ed. The H. John
Heinz III Center for Science, Economics, and the Environment. Washington, D.
C.

Hicks, B. J ., and Watson, N. R. N. 1985. Seasonal changes in abundance of brown trout
and rainbow trout assessed by drift diving in the Rangitkei River, New Zealand.
New Zealand Journal of Marine and Freshwater Research 19: 1-10.

Kanehl, P. D., Lyons, J ., and Nelson, J. E. 1997. Changes in the habitat and fish
community of the Milwaukee River, Wisconsin, following removal of the Woolen

65

Mills Dam. North American Journal of Fisheries Management 17: 387-400.

Kondolf, G. M. 1997. Hungry waters: effects of dams and gravel mining on river
channels. Environmental Management 21 : 533-55 1.

Krebs, C. J. 1999. Ecological Methodology. Addison-Welsey Educational Publishers.
Menlo Park, CA.

Lessard, J. L. and Hayes, D. B. 2003. Effects of elevated water temperature on fish and
macroinvertebrate communities below small dams. River Research and
Applications 19: 721-732.

Li gon, F. K. and Dietrich, W. E. 1995. Downstream ecological effects of dams.
Bioscience 45: 183-192.

Lydy, M. J ., Crawford, C. G., and Frey, J. W. 2000. A comparison of selected diversity,
similarity, and biotic indices for detecting changes in benthic-invertebrate
community structure and stream quality. Archives of Environmental
Contamination and Toxicology 39: 469-479.

Merritt, R. W. and Cummins, K. W. 1996. An Introduction to the Aquatic Insects of
North America, Third Edition, ed. Kendall/Hunt Publishing Company. Dubuque,
IA.

Merritt, R. W., and K. W. Cummins. 2006. Trophic relations of aquatic invertebrates, pp.
585-609. In Methods in stream ecology. F. R. Hauer and G. A. Lamberti, eds.
Elsevier, Amsterdam, Netherlands. 877 pp.

National Research Council. 1992. Resotration of Aquatic Ecosystems, ed. National
Academy Press. Washington, DC.

Orr, C. H. and Stanley, E. H. 2006. Vegetation development and restoration potential of
drained reservoirs following darn removal in Wisconsin. River Research and
Applications 22: 281-295.

Pardo, I. and Arrnitage, P. D. 1997. Species assemblages as descriptors of mesohabitats.
Hydrobiologia 344: 111-128.

Petts, G. E. and Foster, 1. 1985. Rivers and Landscapes, ed. Edward Arnold. London,
England.

Pickett, S. T. A. 1989. Space-for-time substitution as an alternative to long-term studies.
pp. 110-135. In Long-term studies in ecology: Approaches and alternatives. G. E.
Likens (ed.). Spring-Verlag: New York.

Poff, N. L. and Hart, D. D. 2002. How dams vary and why it matters for the emerging
science of dam removal. Bioscience 52: 659-668.

66

Pollard, A. I. and Reed, T. 2004. Benthic invertebrate assemblage change following dam
removal in a Wisconsin stream. Hydrobiologia 513: 51-58.

Rabeni, C. F., Doisy, K. E., and Galat, D. L. 2002. Testing the biological basis of a
stream habitat classification using benthic invertebrates. Ecological Applications
12: 782-796.

River Alliance of Wisconsin (RAW). 2005. List of Wisconsin dams intentionally
removed since 1950. www.wisconsinrivers.org. [Accessed 11/20/2006].

Rosenberg, D. M., and Resh, V. H. (editors). 1993. Freshwater biomonitoring and benthic
macroinvertebrates. Chapman and Hall, New York. 488 pp.

Santucci Jr., V. J ., Gephard, S. R., and Pescitelli, S. M. 2005. Effects of multiple low-
head dams on macroinvertebrates, habitat, and water quality in the F ox River,
Illinois. North American Journal of Fisheries Management 25: 975-992.

SAS. 2002-2003. SAS Institute, Inc. Cary, North Carolina, USA.

Stanley, E. H., Luebke, M. A., Doyle, M. W., and Marshall, D. W. 2002. Short-terrn
changes in channel form and macroinvertebrate communities following low-head
dam removal. Journal of the North American Benthological Society 21: 172-187.

Stanley, E. H., Catalano, M. J ., Mercado-Silva, N., and Orr, C. H. 2007. Effects of dam
removal on brook trout in a Wisconsin stream. River Research and Applications
23: 792-798.

Thomson, J. R., Hart, D. D., Charles, D. F., Nightengale, T. L., and Winter, D. M. 2005.
Effects of removal of a small dam on downstream macroinvertebrate and algal
assemblages in a Pennsylvania stream. Journal of the North American
Benthological Society 24: 192-207.

Tiemann, J. S., Gillette, D. P., Wildhaber, M. L., and Edds, D. R. 2004. Effects of
lowhead dams on riffle-dwelling fishes and macroinvertebrates in a Midwestern
River. Transactions of the American Fisheries Society 133: 705-717.

United States Army Corps of Engineers (USACE). 1998. United States Army Corps of
Engineers: National Inventory of Dams.
http://crunch.tec.army.mil/nidpublic/webpages/nid.cfm. [Accessedz 11/14/2006].

Vannote, R. L., Minshall, G. W., Cummins, K. W., Sedell, J. R., and Cushing, C. E.
1980. The river continuum concept. Canadian Journal of Fisheries and Aquatic
Sciences 37: 130-137.

Ward, J. V. and Stanford, J. A. 1983. The serial continuity concept of lotic ecosystems.
Pages 29-42 in T. D. Fontain and S. M. Bartell, editors. Dynamics of Lotic
Ecosystems. Ann Arbor Science, Ann Arbor, MI USA.

67

 

Waters, T. E. 1995. Sediment in Streams: Sources, Biological Effects, and Control, ed.
American Fisheries Society. Bethesda, MD.

68

IIIIIIIIIIIII U VERSITY LIBRARIES

Ill/11WIIfill/1111111111llWl/llllfll/l/Hl

3 1293 02956 8817