NC0000272_MS Thesis Upper Pigeon_20070501

SPATIAL AND TEMPORAL PATTERNS OF DRIFTING FISH LARVAE IN THE
UPPER PIGEON RIVER, HAYWOOD COUNTY,NORTH CAROLINA
Michael J.LaVoie
A Thesis
Submitted to the
Faculty of the Graduate School
of
Western Carolina University
In Partial Fulfillment of
the Requirements for the Degree
of
Master of Science
Committee:
q Director
` � -- — Dean of the Graduate School
' Date:
Spring 2007
Western Carolina University
Cullowhee,North Carolina

SPATIAL AND TEMPORAL PATTERNS OF DRIFTING FISH LARVAE IN THE
UPPER PIGEON RIVER, HAYWOOD COUNTY, NORTH CAROLINA
A thesis presented to the faculty of the Graduate-School of Western Carolina University
in partial fulfillment of the requirements for the degree of Master of Science
By
Michael J. LaVoie
Director: Dr. Thomas H. Martin
Associate Professor
Department of Biology
May 2007

Acknowledgements
I would like to express my deep appreciation for all that made this project
possible. Dr. Tom Martin provided invaluable insight and guidance throughout the
process, for which I am grateful. The assistance provided by my committee members;
Dr. James Costa and Dr. Mack Powell, and readers; Dr. Sean O'Connell and Dr. Greg
Adkison, in the final draft preparation is greatly appreciated. I would like to thank the
North Carolina Wildlife Resources Commission and the Pigeon River Fund for funding.
Thank you to Blue Ridge Paper Products, Inc. for access to sampling sites and general
logistical assistance. Jeff Wesner, Matt Rajala, and Wes Cornelison provided assistance
in the field and laboratory, for which I am grateful. I would not be in this position
without the love and selfless support that my family has provided me throughout my life.
Finally,Erin McManus' unwavering patience, encouraging me through those sleepless .
nights at the paper mill and endless hours at the microscope,made all of this possible.
ii

Table of Contents
Page
Listof Tables -------- --------•----------------------•--- ---------------------------------- iv
Listof Figures-------------------------------------------------------------------------------------------------------------- vi
Abstract •••
Introduction I
Pigeon River Water Quality and Aquatic Biodiversity--------------------------------------- 1
Riverine Ecosystems and Fish Movement----------------------------------------------------------- 3
Fisheries Restoration----------------------------------------------------------------- -•-----• ------------------ 6
Methods---------------------------------------------------------------------------------------- ------ ------------------ 9
StudyArea----------------------------------------------- ---•--------------------------=------------------------------ 9
Sampling---------------------------------------------------------------------------------------------------------------- 9
SampleProcessing---------------------------------------------------------------- ------------------------------- 11
DataAnalysis.---------------------------------------------------------- ---------------------- ---------------------- 12
Results----------------------------------------------------------------------------------------------------- ------------------- 13
Discussion -------------------------------- ------------•---------------------------------------------•--------------- 39
Conclusions----------------------------------------------------------------------------------------------------------------- 55
LiteratureCited------------------------------ ----------------------------------------------------------------------------- 57
r,

List of Tables
Table Page
1. Larval fish collected in the Pigeon River at three sites during a March—
September2005 study------------------------------------------------------------------------------------------- 15
2. Summary from multivariate analysis of variance (MANOVA) of combined
Response of Percidae, Cyprinidae, Catostomidae Centrarchidae, Cottidae,
and Ictaluridae to location along the Pigeon River(Site),time of sampling
(Time),net location relative to the bank(Loc) and their 2-way interactions..-- 18
3. Summary of ANOVA for Catostomidae------------------------------------------------------------- 19
4. Summary of ANOVA for Centrarchidae------------------------------------------------------------- 19
5. Summary of ANOVA for Cottidae---------------------------------------------------------------------- 19
6. Summary of ANOVA for Ictaluridae------------------------------------------------------------------- 20
I
7. Summary of ANOVA for Cyprinidae------------------------------------------------------------------ 20
8. Summary of ANOVA for Percidae---------------------------------------------------------------------- 20
9. Diel patterns of larval fish drift density for six families of fishes collected in
the Pigeon River. Sampling block numeral refers to the time that the 4 hour
sampling period began. CA=Catostomidae, CE=Centrarchidae, CO=
Cottidae, CY=Cyprinidae, IC=Ictaluridae;and PE=Percidae---------------------- 22
10.Percent composition of two developmental phases and mean total length
for larval fishes collected in drift nets positioned at two instream locations
(N-B=near-bank,M-C=mid-channel) in the Pigeon River---------------------------- 25
11. Percent composition of two developmental phases and mean total length
of four families of larval fishes collected in drift nets set at two instream
Locations (N-B=near-bank,M-C=mid-channel) in the Pigeon River----------- 26
12. Mean densities (no./1000m'),total number, and percent composition(%) of
fish larvae collected at three sites in the Pigeon River--------------------------------------- 30
iv

Table Page
13. Relative abundance of adult fish (n=476) and larval fish (n= 1132)
Collected in two 2005 Pigeon River studies at Site 1 upstream of the
paper mill at Canton,NC (University of Tennessee unpublished 2005)___________ 42
v

List of Figures
Figure Page
1. Map of larval fish sampling sites on the Pigeon River near Canton,NC.......... 10
2. Mean densities of drifting larval fish compared to discharge collected at
three sampling stations on the Pigeon River in 2005------------------------------------------ 16
3. Diel drift patterns for larval fishes from April through September 2005 in
the Pigeon River from April through September 2005---------------------------------------- 21
4. Diel drift patterns for Catostomidae, Cyprinidae, and Percidae in the Pigeon
Rivernear Canton,NC------------------------------------------------------------------------------------------ 23
i
5. Larval drift density for Catostomidae and Cyprinidae relative to location
within the river channel for all sample sites on the Pigeon River______________________ 24
i
6. Length-frequency distribution of four families of drifting larval fishes
collected at three sites in the Pigeon River near Canton,NC____________________________ 27
I
7. Seasonal drift patterns for fish larvae collected from the Pigeon River at
Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of low-head),
and Site 3 (downstream of mixing zone)------------------------------------------------------------- 31
S. Seasonal drift patterns for Cyprinidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of
low-head), and Site 3 (downstream of mixing zone)------------------------------------------ 32
9. Seasonal drift patterns for Percidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of
low-head), and Site 3 (downstream of mixing zone)__________________________________________ 33
10. Seasonal drift patterns for Cottidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of
low-head), and Site 3 (downstream of mixing zone)__________________________________________ 34
11. Seasonal drift patterns for Ictaluridae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of
low-head), and Site 3 (downstream of mixing zone)__________________________________________ 35
A

Figure Page
12. Seasonal drift patterns for Centrarchidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of
low-head), and Site 3 (downstream of mixing zone)------------------------------------------ 36
13. Seasonal drift patterns for Catostomidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of
low-head), and Site 3.(downstream of mixing zone)------------------------------------------ 37
14. Water quality parameters measured at three larval fish sampling sites on the
Pigeon River near Canton,NC-------------------------------- --------------------------------------------- 38
vii

i
i
Abstract
SPATIAL AND TEMPORAL PATTERNS OF DRIFTING FISH LARVAE IN THE
UPPER PIGEON RIVER,NC
Michael J. LaVoie, M.S.
Western Carolina University (May 2007)
Director: Dr. Thomas H. Martin
The Pigeon River has undergone vast improvements in water quality in recent years
i
owing to upgrades in paper manufacturing and wastewater treatment in Canton. Species
richness of the fish assemblage has recently improved through both re-colonization from
tributaries and the reintroduction of previously extirpated species. Recent surveys have
indicated that certain fish species are present upstream of Canton but are absent below the
icity. Potential barriers to colonization exist in the main stem of the river at Canton in the
form of an impoundment and low-head dam, as well as thermal and chemical effluents.
The downstream drift of larval fishes has been shown to be an important dispersal and
j recruitment mechanism for some species. To examine the potential for colonization from
I
upstream and determine if barriers to dispersal exist, fish larvae were collected with drift
nets at three spatially separated sites to. Drifting fish larvae density directly below the
low-head dams was less than 50% of the density measured at the other two sites.
Relative abundance of larval fish taxa was also found to differ among sites with the
assemblage dominated by Percidae and Cyprinidae upstream of the paper mill and
k

Centrarchidae and Catostomidae downstream of the paper mill. It appears that barriers to
re-colonization exist that may hinder the lon.- term success of restoration efforts.
ix

Introduction
Pigeon River Water Quality and Aquatic Biodiversity
The southeastern U.S. contains the richest fish diversity and the highest number of
i
endemic fishes in North America. This region also contains the highest number of
i
imperiled freshwater fish species in the United States;(28%of the native fish fauna)
(Warren et al.2000). Threats to fishes and other aquatic organisms have been tied to
extensive habitat deterioration due to point and non-point source pollution and
channelization, as well as stream fragmentation and hydrologic alteration due to dams
(Neves and Angermeier 1990, Ricciardi and Rasmussen 1999). The degradation of lotic
ecosystems has been extensive throughout the last century in North America and
continues to threaten the persistence of many aquatic species (Benke 1990).
The Pigeon River (Haywood County,NC) is a prime example of an aquatic
system where these pervasive and complex threats have affected the native freshwater
fauna. As with many other rivers, the Pigeon River has served as an important source of
water for industry, as well as an outlet for waste disposal. In 1908 the Champion
International paper company began its operations on the Pigeon River in Canton,NC
(University of Tennessee Student Chapter of the Wildlife and Fisheries Society
(UTSCWFS)2007). Water was removed from the river to aid in paper manufacturing
and then discharged back to the channel as thermally polluted effluent contaminated with
toxic chemicals and tannins. Fish kills were observed immediately after the mill began
production and by the 1940's the fish assemblage was reduced to only pollution tolerant
species, with many reaches completely devoid of fish. The dramatic deterioration in
1

2
water quality resulted in the extirpation of all mollusks (40 species) and most of the
native fishes downstream of Canton (UTSCWFS 2007). In 1988 a consumption advisory
was instituted recommending that no fish caught between the mill and the Tennessee
state line should be eaten due to elevated dioxin levels (UTSCWFS 2007). Hydrologic
and habitat alterations from damming and channelization, as well as sedimentation from
poorly managed construction, agricultural and forestry practices have also contributed to
the degradation of the river.
In recent years the restoration of freshwater ecosystems altered by human
activities has received considerable attention. Improvements to degraded
habitats, including water quality problems, have been the primary focus of these
restoration efforts, with the presumption that habitat restoration will result in increased
aquatic biodiversity (Palmer et al. 1997). In the last ten years, habitat within the Pigeon
River has gone through a dramatic recovery due to improvements in pollution control
from paper manufacturing and upgrades in wastewater treatment by Blue Ridge Paper
Products Inc. (Coombs 2004). Toxic and colored effluents have been greatly reduced and
additional long term declines in conductivity, fecal coliform bacteria, and nutrient
concentrations downstream of the mill's discharge demonstrate additional improvements
in water quality (North Carolina Department of Environment and Natural Resources
(NCDENR) 2003). In January of 2007 the last remaining fish consumption advisory
below the Blue Ridge paper mill in Haywood County was removed (North Carolina
Department of Health and Human Services (NCDHHS) 2007). Although these events are
encouraging, recent studies of the Pigeon River fish fauna conducted by the North

3
i
I
i
Carolina Wildlife Resources Commission (NCWRC) and EA Engineering, Science, and
Technology, Inc. have consistently found species present in the main=stem of the Pigeon
River upstream of Canton that are absent or found in reduced abundance below the city
(EA 1996,EA 2001,NCWRC unpublished). These findings bring into question why re-
colonization of previously extirpated fish species has not contributed more to the
restoration of the native fish assemblage from upstream source populations now that the
physical habitat has improved.
Riverine Ecosystems and Fish Movement
Natural fluvial forms and processes have been found to be intimately linked to the
biological integrity of river systems. Geomorphologic and hydrological characteristics of
rivers and the effects of these physical forces greatly impact ecological processes related.
to aquatic biota. The human alteration of rivers and streams through the construction of
dams has altered the natural fluvial and geomorphic characteristics of many watersheds
throughout North America. Fishes have evolved a variety of lateral and longitudinal
movement patterns which have been disrupted by alterations in the hydrology and the
physiography of river channels (Schlosser 1991). Native fishes have been particularly
impacted due to this alteration in ecosystem functioning.
A major advance in theory concerning aquatic ecology and lotic systems occurred
with the development of the river continuum concept (Vanote et al. 1980). This theory
describes a continual change in the structure and function of the biological and physical
components of a river from upper to lower reaches due to changes in the input and

4
cycling of organic matter. This concept was a landmark step to advance the idea of the
physical and biological connectivity in riverine ecosystems, which refers to the open
longitudinal exchange of water, organisms, nutrients and energy between components of
the channel (Ward and Stanford 1995).
Rivers are dynamic systems with longitudinal and lateral changes in structure and
function. Patterns of pools, runs,riffles, braids, backwaters and oxbows are all various
components of lotic systems that vary in configuration and form a unique heterogeneous
riverine landscape (Wiens 2002). These and other spatial components of river systems at
finer scales can be considered as patches (Townsend 1989). These patches may have
unique physical attributes such as current velocities or substrate composition, as well as a
unique biotic assemblage. The transfer of resources and individuals that are contained
within patches are regulated by longitudinal, lateral, and vertical boundaries. The
downslope flow of water acts to transfer materials between boundaries. Because these
boundaries undergo dynamic alterations in their ability to be permeated, due mainly to
changes in hydrology and geomorphology, patches within rivers vary in their
connectedness. The connectivity of patches in rivers is especially important when
considering the ecology of aquatic organisms. Aquatic organisms vary in both their
ability to move between patches as well as their affinity for certain patches of riverine
landscapes. Patches differ in quality relative to an organism's needs and are selected
accordingly. Therefore the likelihood that an organism will move from one position to
another in a river system is related to both the array and quality of patch types and the
boundaries that separate the different locations (Wiens 2002).

i
5
j Many different movement patterns vital to distribution and persistence are
exhibited in aquatic organisms (Giller and Malmqvist 1998). These ecological attributes
have evolved in unison with natural riverine flow patterns and are closely related to an
organism's life history traits. Riverine fish species exhibit a variety of longitudinal and
latitudinal movement patterns throughout their life history that involve a variety of
physical, biological and chemical factors. Different life history stages such as the egg,
larval,juvenile, and adult phases of fishes also vary in their habitat requirements and their
ability to move among patches.
Upstream movements are associated with active migration in response to resource
abundance (Schlosser 1991). These migration patterns have drawn considerable attention
with studies directed at fishes moving upstream to reach spawning grounds: Three
principal forms of downstream migration also exist including 1) passive downstream
migrations in which the fish is not oriented relative to a water flow; 2) active downstream
migrations; and 3) active-passive downstream migrations where a fish's head points into
the current and is taken with the current with weak resistance (Pavlov 1994).
Hydrological conditions related to the flowing nature of water are the most important
factors influencing the downstream movement of fishes. These factors include water
velocity and discharge, as well as seasonal flood dynamics (Pavlov 1994). The
dislodgement, suspension and subsequent downstream transport of organisms in streams
and rivers can play a major role in the ecology of these habitats (Lancaster et al. 1996).
Downstream drift is a widely studied phenomenon, particularly for aquatic
invertebrates. Many fishes also exhibit drift patterns in the early stages of development

6
which can act as an important dispersal mechanism. The larval stage of fishes
encompasses a unique behavioral and ecological life history phase of the organism
(Snyder 1990). Because of their weak swimming ability, larval fishes are often entrained
and passively transported in downstream currents. It is also possible that larval fishes
actively enter the drift to procure resources or avoid predation (Brown and Armstrong
1985). These downstream drift patterns can act to transport fish to nursery habitats and
act to extend the range of populations. Dispersal and the colonization of newly available
habitats is important in maintaining connectivity in fish populations, adjusting the
distribution and abundance of taxa, and increasing genetic diversity within populations
(Jackson et al. 2001). Events occuring during early life history stages may largely control
recruitment and ultimately year-class strength (Schlosser 1985, Houde 2002).
Accordingly, the downstream drift of larval fishes is accepted as a major contributor to
recruitment of some riverine fishes (e.g., Muth and Schmulbach 1984; Reichard et al.
2004).
Fisheries Restoration
The ability of fishes to disperse downstream can be very important when
considering stream restoration. When a habitat recovers from a formerly degraded state,
the dispersal of fishes from upstream source populations may be important in restoring
ecological integrity by providing a source of colonists. Once the constraints of
degradation have been abated in an ecological community,restoration can occur on
passively through fish movement or actively through human intervention (Lake 2001).

i
7
i
The natural downstream flow of water in lotic systems most often assists the recovery of
riverine fish communities through dispersal in a passive manner(Allan and Flecker
1993). The recolonization and biotic recovery of disturbed areas often occurs rapidly in
streams and rivers (Gilley and Malmqvist 1998). The lack of this recovery in the Pigeon
River brings into question the ability of aquatic organisms to disperse from upstream of
Canton and re-colonize the restored habitat into question. When dispersal is impaired by
barriers to colonization, active restoration efforts through human management, such as
j
the direct manipulation of populations through stocking, are needed to assist recovery and
improve biotic integrity (Lake 2001, Angermeier 1997). In 2004 the NCWRC began
I
active restoration efforts through the reintroduction of five previously extirpated species
(Steve Frayley personal communication). Activities such as stocking can be important in
I
aiding the restoration of aquatic ecosystems,but the long term success of restoration
efforts often depends upon addressing critical riverine ecosystem processes (Bond and
Lake 2003).
In order to be successful, restoration projects require updated information
regarding the nature of potential threats that impinge upon a certain system (Soule and
Orians 2001). It is important to decipher the specific impediments to natural recovery
existing in the Pigeon River. Although the fragmentation of river systems by large dams
has been well documented, other anthropogenic alterations such as small dams and point
sources of pollution likely impact fish movements and deserve increased attention
(Dynessius and Nilsson 1994,Ward and Stanford 1979). Both the presence of two run-of
the-river low-head dams that create a small impoundment at the paper mill and the
i na gym"

8
continued release of heated effluent downstream from the paper mill may act as potential
barriers to colonization from upstream source populations. Low-head dams alter natural
hydrologic patterns and have been shown to affect the movement of aquatic biota and the
structure of riverine fish assemblages (Benstead et al. 1999,Porto et al. 1999, Gillette et
al. 2005). Additionally, effluents from industrial facilities can be extremely lethal to
larval fishes (Houde 2002). These potential barriers to re-colonization could inhibit the
natural recovery of the Pigeon River and threaten the long term success of restoration
efforts (Bond and Lake 2003). The long term viability of fish populations is also of
concern when connectivity is eliminated, resulting in biogeographical restriction (Jackson
et al. 2001).
I designed a study to assess if barriers to colonization exist in the main stem of the
river at Canton through a study of larval fish drift. Accordingly, my study aimed at
determining if larval fish are able to drift into the recovering section of the Pigeon River
in North Carolina below Canton. I made quantitative comparisons of the spatial
distribution of larval fishes throughout the fish assemblage's spawning season at three
sampling sites. These sampling sites were located in three different reaches of the Pigeon
River: 1) upstream of Canton; 2) downstream of the impoundment and two low-head
dams at Blue Ridge Paper Products Inc.; and 3)below the paper mill's wastewater outfall
and mixing zone. Larval fish taxonomic richness and relative abundance were expected
to be significantly lower downstream of the low-head dams and the mixing zone if these
structures were acting as barriers to dispersal from upstream source populations.

9
Methods
i
Study Area
The Pigeon River is a major tributary of the French Broad River in the upper
Tennessee River watershed, with its headwaters in the Blue Ridge physiographic
province of North Carolina. I studied the distributions and drift patterns of fish larvae in
the main channel of the upper Pigeon River near Canton,NC (Haywood County). The
hydrology of the upper Pigeon River is typical for the southern Appalachians, with
highest flows in spring and lowest flows in summer and fall. The Pigeon River at the
study area drains approximately 337 square kilometers and had a mean annual discharge
i
of 8.7 mYsec from 1984 to 2002 . Land use in the upper Pigeon River watershed is
primarily forested (84%) (NCDENR 2003).
i
iSampling
I collected drift samples at three spatially separated sites, in order to determine
spatial and temporal drift patterns of ichthyoplankton (Figure 1). The sampling locations
were chosen to distinguish the effects of two potential barriers to colonization. One
sampling station (Site 1) was located upstream of Blue Ridge Paper Products,Inc.,
approximately 100 meters below the NC 215 bridge. The second sampling station (Site
2) was located approximately 20 meters below two low-head run-of-the-river dams. The
third sampling station (Site 3) was located downstream from the Blue Ridge Paper
Product's waste water treatment outfall and mixing zone, approximately 10 meters
upstream of a second NC 215 bridge.
.gym

10
� Site 3 End of
p.
Canton,NC Mixing
B
Paper Mill
Outfall
Site 2
N Low-head
W + E dams
S
Flow
Direction
i
i
Site 1
0 .5 km
Figure 1.Map of larval fish sampling sites on the Pigeon River near Canton, NC.
i
Larval fish sampling was conducted over a 12 hour period on a weekly basis
between the first week of March 2005 until September 23, 2005 when larval fish were no
longer present in samples. Passive drift nets, which are commonly used to sample
drifting larval fishes in riverine environments were used in this study (Kelso and
Rutherford 1996). Rectangular nets ( 0.26 in x 0.45 in opening, 1 in long; 500-Om mesh)
were set at each site on each sampling date in areas of strong to moderate current to
I

! 11
capture drifting larvae. Nets were anchored to the substrate with rebar at four stations
i
distributed across the river channel at each site. Two nets were located.near each shore
(within 3 meters) and two nets were located in the mid-channel. The nets were set with
their long axes even with the surface of the water when the river depth exceeded the
height of the net opening. At water depths less than 0.26 in, the height of the water
column sampled was recorded. A Swoffer Model 2100 flow meter was used to measure
current velocity at the opening of each net immediately after deployment and before the
end of the net set. Mean velocity measurements and the area of net opening under the
river's surface were used to calculate the volume of water passing through the nets during
each sampling period. Nets were deployed for 30 minutes out of every four hours from
16:00 to 04:00 on each sampling date. Samples were collected over 24 hours on a bi-
monthly basis with the aforementioned sampling"regime to assess diel patterns in larval
drift. Sampling did not occur during extreme flow events due to the infeasibility of_
stabilizing the drift nets in fast currents.
Water temperature, conductivity, and dissolved oxygen were.measured with a YSI
I
meter during each sampling period at each site.-Discharge data were obtained from a
United States Geological Survey (USGS) gauging station (03456991) located at Canton,
NC.
Sample Processing
Each larval fish sample was immediately fixed in 3-5% formalin, stained with
I

12
rose bengal to accentuate larval myomere and fin ray elements, and returned to the lab for
jfurther processing. Fixatives were then removed by rinsing with water and fish were
sorted from debris. Larval fish were counted, measured (±0.5 mm total length), and
assigned to an appropriate developmental stage (Snyder 1976). Meristic, morphometric
and ecological characters were observed to identify the larvae collected to the lowest
' taxon feasible based on available taxonomic keys (Meyer 1970; Hogue et al. 1976; Auer
1982; Fuiman et al. 1983; McGowan 1988; Wallus et al. 1990; Kay et al. 1994; Simon
and Wallus 2003; Simon and Wallus 2006). Most fishes in the Centrarchidae, Cyprinidae
and Percidae families were not identified beyond the family level due to the lack of
taxonomic keys. Fishes in the Catostomidae, Cottidae and Ictaluridae families were
i
identified to genus and species. Juvenile fish (exhibiting full fin development but lacking
a median fanfold) collected in the drift nets were not included in the study.
I
Data Analysis
Larval fish drift densities were computed for each drift net set by dividing the
number of individual larvae collected by the volume of water filtered (number of
fish/1000 m3). Drift densities calculated for the entire sampling period include only those
t
sampling dates where larval fish were collected, thus excluding samples collected at the
beginning and end of the season that produced no fish. The total number of fish larvae
drifting through each site on each sample date was also calculated by multiplying the
I
mean drift densities by the total river discharge.
I

13
design with samples taken at three sites over
The study followed a 3x6x2 factorial
six time periods per day using nets stationed at two locations across the stream channel
(two nets near the banks and two near mid-channel). Data were analyzed at the family
level to accommodate the different degrees of classification accomplished within each
family. All experimental factors were fixed rather than random, and were treated as such
in analyses. Differences in larval drift among sites, sample times and stream channel
locations were tested for drift densities of larvae grouped into 6 families—Percidae,
Cyprinidae, Catastomidae, Centrarchidae, Cottidae, and Ictaluridae. Drift densities were
estimated as the number of larvae per unit volume of water sampled over the course of
i the study for each combination of river site, sample time, and net location within the
channel. Because I was testing using multiple dependent variables, I first tested main
effects and 2-way interactions using multiple analysis of variances (MANOVA) followed
by univariate analysis of variance (ANOVA) for those main effects and interactions
deemed statistically significant (a=0.05) in the MANOVA (Zar 1999). All statistical
I analyses were conducted using PROC GLM in SAS/STAT (v. 9, Copyright 2002, SAS
Institute Inc., Cary, NC, USA).
- - Results
I
A total of 928 drift samples were collected during the study with 42 % containing
i
larval fish. We obtained 2522 larval fish from these samples, representing 6 families
j (Table 1). Catostomidae accounted for approximately 33.6% of taxa collected at all three
f
sites and was comprised of three species: white suckers (Catostomus commersoni),
northern hogsuckers (Hypentelium nigricans), and black redhorse (Moxostoma

i 14
i
i
duquesnei). Centrarchidae was the second most abundant family (26.6%) found in drift
samples. Rock bass (Ambloplites rupestris) were found in low numbers within this
family. Lepomis spp, were also identified within the centrarchids collected but not
quantified at this time due to potential confusion with large mouth bass (Micropterus
salmoides) (Robert Wallus pers. comm.) Cyprinidae and Percidae accounted for 19 2%
and Percidae 19.4% of total larval fish collected,respectively. Two species of
i
Ictaluridae, channel catfish (Ictalurus punctatus) and white catfish (Ameirus catus), and
one species of sculpin (mottled sculpin, Cottus bairdi) were collected in low numbers.
Fourteen juvenile and adult fishes were also collected during the sampling period but
i
jwere not included in the analysis.
Larval fish were detected drifting in the study area from April 21 through
September 6 (Figure 2.) Mean drift densities for each sample date varied from 0 to
426.6 fish per 1000 cubic meters of water sampled. Based on USGS discharge data and
mean density calculations, approximately 138 million larval fish drifted through the study
area throughout the sampling period. Mean discharge was not directly correlated with
mean drift densities (Rz= 0.03, P = 0.45).
I Total drift densities were low at the beginning of the sampling period. Suckers and
darters were the first fishes to appear in samples. White suckers and northern hogsuckers
I
appeared first, followed by black redhorses later in the spring. High drift densities
occurred in mid-May and early-June due to an increasing abundance of suckers and
i
I

15
Table 1. Larval fish collected in the Pigeon River at three sites during a March -
September 2005 sampling period
Number of % of total Peak
Taxon fish number Abundance
sampled sampled Date
Catostomidae (suckers) 847 33.6
Catostomus commensori 182 7.2 5/5/05
Hypentelium nigricans 247 9.8 5/17/05
Moxostoma duquesnei 332 13.2 6/9/05
Undetermined sp. 86 3.4
Cyprinidae (minnows) 483 19.2 5/12/05
Centrarchidae (sunfishes) 670 26.6
Ambloplites rupestris 4 0.6 7/10/05
* Undetermined sp. 666 26.4 8/10/05
I
Cottidae (sculpins) 3 0.1
Coitus bairdi 3 0.1 5/12/06
Ictaluridae (catfishes) 8 0.3
Ictalurus punctatus 7 0.3 8/10/05
Ameiurus catus 1 0 8/26/05
Percidae (darters) 490 19.4 7/10/05
Undetermined sp.
(damaged) 20 0.8
* Specimens likely Lepomis sp.
r

16
0 Drift Density s Discharge
400
250
350
300 200
0 250 y
0 150
6 200
c
150 100
100
50
50
p 0
7 'd'
Figure 2. Mean densities of drifting larval fish compared to discharge collected at three
sampling stations on the Pigeon River in 2005.
I
darters, and the addition of members of Cyprinidae to the assemblage. Mottled sculpin
larvae were collected in low numbers on one date in early May. Low overall densities
were again observed in late June, followed by a spike in July with increasing numbers of
I
darters and cyprinids. Peak drift densities occurred in early August with 810 fish/1000
m', consisting mostly of fishes in the Centrarchidae family. Catfish drifted in low
numbers from early August to early September.
I
i
:1

17
An assessment of development revealed the majority of larval fishes collected
were in the post larval stage with yolk-sac absorption completed (69.3% of fish
collected). This dominant pattern was observed in all but one family. Approximately
96% of catostomids, 71% of ictalurids, 71% of catostomids, and 62% of cyprinids were
classified as post-yolk sac larvae. All cottid specimens collected were in a late post-
larval phase with some advanced fin ray development. Percids were the only taxon
whose composition was dominated by pro-larval specimens (62%).
Well defined size groups of certain taxa were found in the drift. For example,
greater than 99% of the centrarchids collected ranged from 6.2 to 8.7 mm total length.
The range of sizes for catostomids was similar for the three species found in drift samples
(11.1 mm to 20.6mm). The average size of black redhorses collected was 16.8 mm,
however, compared with 14.2 mm for northern hogsuckers and 13.2 mm for white
suckers. Mean total length of cottids and ictalurids collected was 10.8 mm and 16.8 mm,
respectively. Drifting percids and cyprinids exhibited similar mean sizes (8.6 mm and
8.3 mm respectively) and ranged in size from 5-17 mm.
I
i
All main effects examined by MANOVA for larval fish collected throughout the
sampling season were deemed significant as was the river-site by sample-time interaction
(Table 2). As interactions can significantly influence the proper interpretation of main
effects (Underwood 1997), we first examined the interactions of river-site by sample-
time.
The Site x Time interaction was found to be statistically significant for
i catostomids, centrarchids, cottids, and ictalurids (Tables 3, 4, 5, and 6). Examination of
i

18
I
means suggests that the significance of the interaction may be a trivial result of the
differences in abundance among sites. Three of the four families were relatively rare at
one or more sites resulting in a significantly different diel pattern of drift due to drift
densities being truncated at zero. The catostomids were relatively abundant at all sample
sites, but a larger proportion of these larvae were captured during daylight sample periods
' (08:00—20:00) at Site 2 than at either of the other two sample sites. There were at least
three species of catostomid present in the samples and small differences in diel drift and
abundance among species and sample site could have contributed to the observed
differences.
i
Table 2. Summary from multivariate analysis of variance (MANOVA) of combined
response of Percidae, Cyprinidae, Catastomidae, Centrarchidae, Cottidae, and
Ictaluridae to location along the Pigeon River(Site), time of sampling (Time), net
location relative to the bank (Loc) and their 2-way interactions.
Numerator Denominator
Source df df Wilk's X F P
Site 12 82 0.181 9.25 <0.0001
Time 30 166 0.093 4.48 <0.0001
Loc 6 41 0.671 3.36 0.0088
Site x Time 60 219.9 0.077 2.31 <0.0001
Site x Loc 12 82 0.808 0.77 0.6794
Time x Loc 30 166 0.551 0.89 0.6857
l

19
Table 3. Summary of ANOVA for Catostomidae
Source df SS MS F P
Site 2 8586 4293 10.66 0.0002
Time 5 22018 4404 10.93 <0.0001
Loc 1 4225 4225 10.49 0.0022
Site X Time 10 9436 944 2.34 0.025
Site X Loc 2 776 388 0.96 0.3891
Time X Loc 5 3942 788 1.96 0.1032
Error 46 18534 403
Total 71 67517
Table 4. Summary of ANOVA for Centrarchidae
Source df SS MS F P
I
Site 2 6835 3417 5.2 0.0092
Time 5 397825- 7956 12.11 <0.0001
Loc 1 5115 511 0.78 0.3824
Site X Time 10 13800 1380 _ 2.1 0.0438
Site X Loc 2 178 89 0.14 0.8739
Time X Loc 5 12888 258 0.39 0.8516
Error 46 302188 657
Total 71 926128
Table 5. Summary of ANOVA for Cottidae
Source. df SS ' MS F P
Site 2 0.491 0.24548836 2.42 0.1004
Time 5 1.227 0.24548836 2.42 0.05
Lac 1 0.009 0.00903311 0.09 0.7669
Site X Time 10 2.455 0.24548836 2.42 0.0209
Site X Loc 2 0.018 0.00903311 0.09 0.9151
Time X Loc 5 0.045 0.00903311 0.09 0.9936
Error 46 4.672 0.10155908
Total 71 8.917

20
Table 6. Summary of ANOVA for Ictaluridae
Source df SS MS F P
Site 2 1.321 0.661 3.79 0.0298
Time 5 4.049 0.81 4.65 0.0016
Loc 1 0.029 0.029 0.16 0.6866
Site X Time 10 4.304 0.43 2.47 0.0184
Site X Loc 2 0.333 0.167 0.96 0.3918
Time X Loc 5 0.565 0.113 0.65 0.6638
Error 46 8.01 0.174
Total 71 18.611
Table 7. Summary of ANOVA for Cyprinidae
Source df SS MS F P
Site 2 2733 1367 3.24 0.0482
Time 5 10573 2115 5.01 0.0009
Loc 1 2082 2082 4.94 0.0312
Site X Time 10 3974 397 0.94 0.5046
Site X Loc 2 873 437 1.04 0.3633
Time X Loc 5 2893 579 1.37 0.2523
Error 46 19399 422
Total 71 42529
Table 8. Summary of ANOVA for Percidae
Source df SS MS F P
Site 2 9871 4935 19.94 <0.0001
Time 5 5938 1188 4.8 0.0013
Loc 1 18 18 0.07 0.7902
Site X Time 10 4818 482 1.95 0.0626
Site X Loc 2 311 156 0.63 0.5378
Time X Loc 5 1318 264 1.06 0.3921
Error 46 11388 248
Total 71 33662
I

21
Overall drift densities varied with time of day and a significant diel periodicity was
observed for all taxa combined (P<0.0001). Approximately 88.5% of drifting larval fish
were collected between two sampling periods from 20:00 —04:00 (Figure 3). Highest
drift rates occurred during the 20:00-24:00 block with a mean density of 147.5 fish/1000
m3. Higher densities of larvae were also observed in the drift during the early morning
(04:00-08:00) when compared with daytime (08:00-20:00) values.
200
150
0
12 100
0
50
0
04:00 - 08:00 - 12:00 - 16:00 - 2000- 24:00 -
08:00 12:00 16:00 20:00 24:00 04:00
Figure 3. Diel drift patterns for larval fishes from April through September 2005 in the
Pigeon River near Canton, NC.

22
Members of the Catostomidae, Centrarchidae, Ictaluridae, Cyprinidae, and Percidae
were found to display a significant response to sample time (P <0.01) (Tables 3,4,6,7,8).
Drift predominantly occurred for these taxa during the night and early morning hours
(Table 9). Centrarchids, with the exception of one fish, and ictalurids were only collected
during the nighttime blocks of 20:00 and 24:00. Sculpin (Cottidae)were only captured
during the 20:00 block,but their response to sample time was less significant due to the
low density collected (P =0.05) (Table 9). Catostomids, cyprinids and percids exhibited
their highest drift densities during the nighttime and pre-dawn hours,with day to night
abundance ratios of 7:1, 13:1, and 3:1 for each taxon respectively (Fig. 4).
i
Table 9. Diel patterns of larval fish drift density for six families of fishes collected in
the Pigeon River. Sampling block numeral refers to the time that the 4 hour sampling
period began. CA= Catostomidae, CE= Centrarchidae, CO= Cottidae,
CY=Cyprinidae, IC=Ictaluridae, and PE=Percidae
Sampling no. /1000m'
Block CA CE CO CY IC PE
400 28.3 0 0 11 0 23
800 6.1 0 0 2.6 0 10.3
1200 4.9 0 0 1.2 0 8.1
1600 6.1 0.1 0 1.3 0 2.2
! 2000 45.2 45.2 0.4 27.3 0.6 27.5
2400 40.1 39.6 0 27 0.4 11.5
I

23
50
® Catostomidae
❑ Cyprmidae
40 p Percidae
30
iM
° 20
10
0 N
04:00 - 08:00 - 12:00 - 16:00 - 20:00 - 24 00 -
08:00 12:00 16:00 20:00 24:00 04:00
Figure 4. Die] drift patterns for Catostomidae, Cyprinidae, and Percidae in the Pigeon
River from April through September 2005.
Densities of drifting larvae in near-bank samples (102 fish/1000m') were
significantly greater than mid-channel samples (51 fish/1000m') (P<0.01),but the
catostomids and cyprinids were the only families that exhibited a statistically significant
difference in drift densities between near-bank and mid-channel samples (P < 0.05)
(Tables 3 and 7). Near-bank drift was approximately double that of near-mid-channel for
these two taxa (Figure 5).
Comparisons of larval fish size and developmental phase were made between
near-bank and mid-channel drift net collections for four families. Cottids and ictalurids
were not analyzed due to the low number collected. Pro-larvae and post-larvae
I RUN MEW IRMUMM,
i
I
ez.xv

24
i
40 Near-bank
Mid-channel
30
0
0
i o
20
c
10
hC. 'S3�n
�pp µ4 `� p� �S i•r�- � ..
Catastomidae Cyprinidae
i
j Figure 5. Larval drift density for Catostomidae and Cyprinidae relative to location within
the river channel for all sample sites on the Pigeon River.
i
composition, as well as mean total length for larval fish in near-bank nets and mid-
channel nets were very similar(Table 10).
This developmental and size pattern was also consistent when looking at the family
level for Catostomidae, Cyprinidae and Centrarchidae (Table 11). A t-test detected that
the mean total length of percids collected in near-bank and mid-channel sets was
significantly different, with slightly larger fish collected in mid-channel drift nets (P <
0.001). An assessment of length-frequency distribution between near-bank and mid-
channel samples also revealed a greater abundance of larger sized fish in mid-channel
collections for the Percidae (Figure 6). No consistent patterns of length-frequency
distributions were observed for the other families collected.

25
Table 10. Percent composition of two developmental phases and mean total length
for larval fishes collected in drift nets positioned at two instream locations (N-B=near
bank,M-C=mid-channel) in the Pigeon River.
N-B M-C
Pro-larva (%) 67.2 68.2
Post-larva(%) 32.8 31.8
Mean Total-length(mm) 10.7 10.4

Table 11. Percent composition of two developmental phases and mean total length of four families of larval fishes collected
in drift nets set at two instream locations (N-B =near-bank, M-C=mid-channel) in the Pigeon River near Canton,NC.
Percidae Centrarchidae Cyprinidae Catostomidae
N-B M-C N-B M-C N-B M-C N-B M-C
Pro-larvae 35.8 41.3 96.3 96.4 61.1 64.2 75.5 64.0
N
Post-larvae 64.2 58.7 3.7 3.6 38.9 35.8 24.5 36.0
N
Mean Total 8.3 9.1 7.6 7.5 8.1 8.5 15.1 15.1
Length (nun)
N
O�

PERCIDAE CENTRARCH DAE
0 Near-bank
Near-bank
El Mid-channel
40 El Mid-channel 70 m rv}
20 35 >..
O .,max 11 1
6 7 8 9 10 11 12 13 14 15 16 17 18 6 7 8 9
Total Length(mm) Total Length(mm)
CYPRINIDAE CATOSTOMIDAE
30 0 Near-bank 30 ®Near-bank
0 Mid-channel 0 Mid-channel
S ° T
w 15 w 15 ;
o 7 N
0 �ii I 1 0
5 6 7 8 9 10 11 12 13 14 15 16 17 11 12 13 14 15 16 17 18 19 20 21
Total Length(rum) Total Length(inn)
Figure 6. Length-frequency distribution of four families of drifting larval fishes collected at three sites in the Pigeon River
near Canton,NC
N
J

28
A total of 1,159, 380, and 940 fish were collected at Site 1, Site 2, and Site 3,
respectively (Table 12.). Sampling sites 1 and 3 produced the majority of overall larvae
collected (81.0%) with total mean drift density values of 97.5 fish/1000 m'and 81.3
fish/1000 m3,respectively. Larval fish drift density at sampling site 2 was 41.9 fish/1000
in which was significantly lower than the other two sites (P< 0.05).
The composition and relative abundance of larval fish taxa at each of the three sites
was found to be different throughout the sampling period (Table 12.). Drift densities in
April and early May were highest at Site 3 and dominated by catostomids (Figure 7).
Site 1 produced the greatest drift density of larval fish in mid-May and all of June and
July with cyprinids and percids dominating the drift. Drift densities spiked in early
August at site 2 with the samples being dominated by members of the Centrarchidae and
Ictaluridae families. Abundance estimates at each site declined dramatically thereafter
until no fish were collected in the drift.
Dominant taxa at each site also varied. Collections at site 1 were comprised of six
families with percids and cyprinids being the most abundant taxa (Table 12). In contrast,
site 2 consisted of five families and was dominated by the Centrarchidae family, with
Cyprinidae being the second most abundant famiiy collected. In site 3 samples, the
entrarchids and catostomids were the most abundant of the five families that were
recorded.
Cyprinids and percids were found to be more abundant upstream of the paper mill
when compared with the downstream sites (Figures 8 and 9). Site 1 produced the only
specimens of mottled sculpin (Coitus bairdi- Cottidae) collected in drift nets (Figure 10).

.. 29
i
Ictalurids were most abundant at site 3, with six channel catfish (Ictalurus punctatus)
' collected below the paper mill outfall mixing zone. (Table 12,Fig 11). A single white
bullhead(Ameirus catus) was collected at site 1 upstream of the mill (Table 12).
Centrarchids were also the most abundant at the downstream most site (Figure 12).
The catostomids demonstrated a more complicated response with higher densities at the
upstream most and downstream most sites and reduced drift at Site 2 (Figure 13). White
suckers appeared to be more abundant in samples from the downstream most site,while
black redhorse tended to be more abundant in the upstream site(Table 12). Black
redhorse and northern hogsucker tended to be less abundant in the samples taken
r
immediately below the low-head dam.
i Mean temperature and conductivity differed at the three study sites. The mean,
i
temperature at site 3 was approximately 3° C higher than the two sites upstream of the
mill outfall (Figure 14). More extreme differences were observed for conductivity with
an average site 3 value of 381.5 µs/cm compared to 20.0 µs/cm and 20.9 Ns/cm at site 1
and 2 respectively. No major differences in dissolved oxygen were observed among the
three study sites.
i

Table 12. Mean densities (no./1000m3), total number, and percent composition (%) of fish larvae collected at three sites in the Pigeon River.
Taxon Site 1 Site 2 Site 3
Density # % Density # % Density # %
Catostomidae (suckers)
Catostonuis commensori 2 24 2.1 2 17 4.6 12.3 141 14.6
Hypentelium nigricans 8.1 98 8.3 3.1 27 7.2 10.9 121 13.0
Moxostonta duquesnei 18.7 225 19.2 3 26 6.9 7.4 81 8.8
Undetermined spp. 3.9 47 4.0 0.9 8 2.1 2.8 31 3.3
Cyprinidae (minnows)
Undetermined spp. 25.6 309 26.3 9.3 81 21.5 8.5 95 10.1
Centrarchidae (sunfishes)
Ambloplites spp. 0.1 1 0.1 .0 0 0.0 0.3 3 0.4
Undetermined spp. 7.7 93 7.9 19.6 170 45.3 36.7 403 43.6
Cottidae (sculpins)
Cottus bairdi 0.2 3 0.2 0 0 0.0 0 0 0.0
Ictaluridae (catfishes)
Ictalurus punctahes 0 0 0.0 0.1 1 0.2 0.5 6 0.6
Ameiurus catus 0.1 1 0.1 0 0 0.0 0 0 0.0
Percidae (darters)
Undetermined sp. 29.7 378 30.6 5.2 49 12.0 4.6 63 5.5
Undetermined sp. 1.1 17 1.1 0.1 1 0.2 0.1 2 0.1
Totals 97.2 1196 43.3 380 84 946
w
0

,..1
31
Site 1
600
5 500
c 400
C> 300
200
a 100
0
°°�' o°"' o°�'
r\\� p\1 b\t \ti g\ti p\ti b\1 \�
h\L 5\ \1 \1 \1 p\ b\ \1 \`l, \`L ^\1
Site 2
600
E 500
0 400
0 300
200
a 100
0
05 05 05 05 05 05 05 05 05 p`' 05 O5 O5 05 05 05 05
01 01 \ryp \gyp \yo \,yo \yo \yo \tio \yo \tio \yo \yo \yo \yo \yo
p\tip 5\� 5\~� 5\�^ 5\L� O\� \� \,h (sp (.6'
Site 3
600
E 500
0 400
0 300
200
a 100
0
pZ pp5 pp5 pp5 pp5 pph pp5 pp5 °p5 pph pp5 pp5 pp5 pp5 pp5 pp�7 pp5
�f �\ti 'L\ry '\\O �\ti "�\ti 9\ti \ti \ti ^\ti °\ti b\ti \ti \ti °\ti b\ti b\ti
Figure 7. Seasonal drift patterns for fish larvae collected from the Pigeon River at Site 1
(upstream of mill at Canton,NC), Site 2 (downstream of low-head), and Site 3
(downstream of nixing zone).
�/31�l7He aaraffi

32
i
t
I
Cyprinidae (no./1000m3)
200 Site 1
100
I 0
Apr-05 May-05 Jun-05 Ju105 Aug-05 Sep-05
200 Site 2
100
0
Apr-05 May-05 Jun 05 Jul-05 Aug-05 Sep-05
I
200
Site 3
100
I
0
Apr-05 May-05 Jun-05 Ju105 Aug-05 Sep-05
Figure 8. Seasonal drift patterns for Cyprinidae larvae collected from the Pigeon River at
Site i (upstream of mill at Canton,NC), Site 2 (downstream of low-head), and Site 3
(downstream of mixing zone).
I

'33
i
i
Percidae (no.11000m3)
i
100
Site 1
50
i
0
Apr-05 May-05 Jun 05 Jul705 Aug05 Sep-05
100 Site 2
i
50
0
Apr-05 May-05 Jun-05 Jul05 Aug-05 Sep-05
100 Site 3
50
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
Figure 9. Seasonal drift patterns for Percidae larvae collected from the Pigeon River at
Site 1 (upstream of mill at Canton, NC), Site 2 (downstream of low-head), and Site 3
(downstream of nixing zone).
i

34
I
Cottidae (no./1000m3)
10
Site 1
5
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
10
Site 2
I
5
0
Apr-05 May-05 Jun-05 Ju105 Aug-05 Sep-05
10
Site 3
5
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
Figure 10. Seasonal drift patterns for Cottidae larvae collected from the Pigeon River at
Site 1 (upstream of mill at Canton, NC), Site 2 (downstream of low-head), and Site 3
(downstream of mining zone).
i

35
I
i
i
Ictaluridae (no./1000m3)
10
Site 1
5
0
Apr-05 May-05 Jun-05 Jul•05 Aug-05 Sep-05
10 Site 2
5
0
Apr-05, May-05 Jun-05 Jul705 Aug-05 Sep-05
10 Site 3
5
Az�
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
Figure 11. Seasonal drift patterns for Ictaluridae larvae collected from the Pigeon River at
Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of low-head), and Site 3
(downstream of mixing zone).

36
Centrarchidae (no./1000m3)
600 Site 1
300
I
I
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
i
I
600
Site 2
I
300
i
0
Apr-OS May-05 Jun-05 Jul-05 Aug-05 Sep-05
i
i
600
Site 3
300
I
I 0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
Figure 12. Seasonal drift patterns for Centrarchidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of low-head), and
Site 3 (downstream of mixing zone).
i
1

1 3
Catostomidae (no./1000m3)
400 Site 1
200
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
400
Site Z
200
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
400 Site 3
200
0
Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05
Figure 13. Seasonal drift patterns for Catostomidae larvae collected from the Pigeon
River at Site 1 (upstream of mill at Canton,NC), Site 2 (downstream of low-head), and
Site 3 (downstream of mixing zone).
6
I

i
38
;
i
Temperature(°C) ————Site 1
�—Site 2
30 Site 3
15
0
4/21/2005 5/19/2005 6/16/2005 7/14/2005 8/11/2005 9/8/2005
Dissolved Oxygen(mg/L) ———— Site 1
Site 2
10 — Site 3
5
0
4/21/2005 5/19/2005 6/16/2005 7/14/2005 8/11/2005 9/8/2005
Conductivity(µs/cm) ————Site 1
Site 2
800 Site 3
400
0
4/21/2005 5/19/2005 6/16/2005 7/14/2005 8/11/2005 9/8/2005
Figure 14. Water quality parameters measured at three larval fish sampling sites on the
Pigeon River near Canton,NC.
II
1

1
39
i
Discussion
The phenology of occurrence in the drift is likely a function of spawning
chronology. The appearance of cottids, catostomids,percids, cyprinids, centrarchids, and
ictalurids in the drift in sequential order was also observed in a study of larval fish in a
small Kentucky stream (Floyd et al. 1984). This pattern was consistent with the
spawning behaviors of fishes found in the Blue Ridge ecoregion(Etnier and Starnes
1993). Spawn timing is affected by a variety of factors including temperature,
photoperiod and flow. The rate of development of eggs and larvae is most dependent
upon temperature and peaks in drift abundance likely occurred in the first weeks after
hatching.
Some families showed distinct peaks in abundance over small time periods, while
other taxa were present at lower densities throughout the sampling season. For example,
approximately 99% of larval centrarchids were collected on August 10. Due to their late
summer spawn timing, these centrarchids were most likely Lepomis spp., but this could
not be verified at this time(Robert Wallus pers. comm.). Alternatively,the cyprinids and
percids exhibited pronounced multiple modes in their seasonal drift abundance patterns
from April through August, suggesting the presence of multiple species (Figures 8 and 9).
Unfortunately, larval fish taxonomy has not progressed to the point to allow accurate
separation of these species. Waters (1972) classified the drift of aquatic insects into three
categories. Constant drift referred to the inadvertent displacement of larvae at low
densities over the time period of active reproductive activity. High density drift resulting

40
from major disturbance events was considered as catastrophic drift. Behavioral drift was
i
catergorized as periodic movement patterns resulting from specific behavioral patterns
such as the avoidance of predators or the acquisition of resources. The drift of larval
fishes in the Pigeon River appears to fit the constant drift category. Although the large
spike in centrarchid abundance over a short time appears catastrophic, no extreme
discharge event occurred on that date and the high density is likely tied to a large
concentration of spawning fish.
The downstream migration of fish during early stages of development is most
heavily influenced by the reproductive life history traits of a particular organism (Pavlov
1994). Spawning placement in relation to downstream currents can directly influence
drift rates for certain fishes. The susceptibility of larval fish to initially drift is mainly
due to their location within the substrate, water velocity and light conditions.
Fluctuations in river velocity and discharge can act as disturbances that dislodge and
transport larval fishes. Initial displacement of a larval fish occurs when the velocity and
force of water overcomes the force of an organism to resist movement. This will occur
more readily with higher nearbed velocities and when the organisms are no longer
protected from small eddies associated with size variation in substrate particles. The
stability.of substrates is also important to consider as flow events associated with
substrate movement were associated with higher drift rates of some benthic insect
populations (Lancaster et al. 1996). The loss of visual orientation during low light
conditions may also influence accidental entrainment in river currents. Behavioral

41
f
i
mechanisms such as the acquisition of food,predator avoidance or nocturnal activity are
i
also potential explanations for drift.
The results.of this study indicate that some taxa may have a greater propensity to
drift than others. Percids, catostomids, cyprinids and centrarchids were the most
abundant taxa collected in the drift throughout all study sites. Assuming that
reproductive efforts were equal among taxa, adult fish surveys conducted by the
University of Tennessee in the summer of 2005 at the larval fish study site 1 provide
insightful information when considering the propensity of certain taxa to drift(University
of Tennessee unpublished 2005).
Larval percids were twice as abundant and larval catostomids were five times as
abundant when compared to adults in the site 1 reach, suggesting that these taxa have a
high tendency to drift during early developmental phases (Table 13). A high propensity
for larval darters and suckers to exhibit significant drift behavior was expected in this
study. Four species of darters (Percidae) are found in the upper Pigeon River as adults
including the greenfin darter(Etheostoma blenniodes), greenside darter(Etheostoma
chlorobranchium),tangerine (Percina aurantiaca), and olive darter(Percina squamata).
Darters demonstrate a wide variety of spawning behaviors including burying, clumping,
attaching, and clustering eggs in association with large rocks, gravel and algae. Many
species of Etheostoma exhibit a wide variety of drift patterns including pelagic, epi-
benthic and benthic forms (Turner 2001, Simon and Wallus 2006). Pelagic drift occurs
for almost all members of Percina. These drift patterns are believed to increase fitness
I

42
I
I
I
Table 13. Relative abundance of adult fish (n =476) and larval fish (n = 1132)
collected in two 2005 Pigeon River studies at Site 1 upstream of the paper mill at
Canton NC. (University of Tennessee unpublished 2005)
% % Adult/Larvae
Taxon' Composition Composition Abundance
of Adults of Larvae Ratio
Catostomidae (suckers) 5.9 30.7 0.2
Catostomus commensori 0.0 2.1 0.0
Hypentelium nigricans 1.5 8.7 0.2
Moxostoma duguesnei 4.4 19.9 0.2
Cyprinidae (minnows) 59.0 27.3 2.2
Centrarchidae
(sunfishes) 11.6 8.3 1.4
Ambloplites rupestris 9.9 0.1 111.8
*Lepomis spp. 1.3 8.2 0.2
Microptertis dolomeiu 0.4 0.0 na
Cottidae (sculpins) 5.7 0.3 21.4
Cottus bairdi 5.7 0.3 21.4
Ictaluridae (catfishes) 0.0 1.0 0.0
Ameiurus catus 0.0 1.0 0.0
Percidae (darters) 17.9 33.4 0.5
*Larval specimens very likely Lepomis sp.but cannot be confirmed.
l

43
I
by transporting larvae from turbulent spawning grounds in riffles to plankton rich nursery
habitats found in slower backwater sections of a river (Simon and Wallus 2006). The
lack of precise knowledge pertaining to reproductive biology, development and
taxonomy for many species of darters, including the olive darter and greenfin darter,
preclude further conclusions regarding species specific drift patterns at this time.
Catostomids have also been found to exhibit high'drift abundances in many
studies (Corbett and Powles 1986, Johnston et al. 1995,Kennedy and Vinyard 1997,
Marchetti and Moyle 2000,). The three species of catostomids found in the Pigeon River
seek access to open substrate spawning sites in areas of swift current where oxygen
supplies are adequate for egg development, and after hatching early are photophobic
(Balon 1975). The weak swimming ability associated with early hatching and the
presence of downstream currents in spawning locations make these taxa highly
vulnerable to drift,particularly at night.
In contrast, cyprinids comprised the largest proportion fishes collected in the
adult survey(59%),yet only 27% of the total larval assemblage consisted of cyprinids,
suggesting more limited drift tendencies. Cyprinid drift patterns have been demonstrated
in some studies (Marchetti and Moyle 2000, Muth and Schmulbach 1984). Reichard et
al. (2004) observed significant drift for a variety of cyprinids in the Czech Republic, and
attributed these patterns to active entrance into the current as a means of transport. The
presence of seven species of cyprinids which demonstrate a wide variety of spawning

44
behaviors, combined with the inability to identify fish beyond the family level limits any
futher conclusions regarding species specific drift patterns in this study.
The relative abundance of adult mottled sculpin (Cottus bairdi) at site 1 was much
higher than larval abundance estimates. Other drift studies have shown extremely high
drift abundances for Cottus spp. (Marchetti and Moyle 2000). Mottled sculpin spawn
beneath stones or ledges in riffle areas with males actively maintaining the nests after
courtship. As with other riffle spawning taxa, they are susceptible to being entrained in
currents after hatching. It should also be noted that spawning mottled sculpin are highly
adapted to swift waters with flattened bodies, large oblique pectoral fins, and a lack of a
swim bladder. These adaptations likely help to reduce accidental drift. The low
abundances of drifting mottled sculpin larvae in this study are probably attributed to the
sampling regime. Mottled sculpin can begin spawning during late winter in the
southeastern United States (Etmier and Starnes 1993). Drifting larvae were possibly
missed in the weeks prior to the date fish were first observed due to logistical constraints
regarding the seasonal timing of sampling.
Centrarchids exhibited higher drift abundances than expected in this study. Light
traps set in backwater areas are often the preferred sampling method for larval
centrarchids. These fishes are nest constructors, depositing eggs in cleaned areas of
substrate dug out of the river bed in slower currents, often along banks. Adults guard
embryos during development and larvae are thereafter pelagic. Spawning in areas of
slower currents and adult guarding would suggest a lower susceptibility to accidental
entrainment, yet free swimming pelagic larvae in search of food would be displaced by
i

45
i
strong currents. The relative abundance of adults was closely tied with the numbers of
drifting larvae at site 1 (Table 13). Therefore it is likely that the extremely high
abundances of drifting centrarcbid larvae at sites 2 and 3 were associated with large adult
populations.
Ictalurids,particularly channel catfish (Ictalurus punctatus),have been shown to
drift extensively in some studies,but usually in later developmental stages (Floyd et al.
1984). These fishes spawn in natural holes or cavities,with males guarding the nests as
development proceeds (Envier and Stames 1993). The low ictalurid densities observed in
this study are most likely a result of low denisities of these fishes in the study reach.
Other interesting patterns emerge when considering individual species. White
suckers(Catostomus commersont) and a single white catfish(Ameiurus catus)were
collected in the drift during the site 1 larval study,but were not observed in the adult
surveys. Thus, larval sampling may be used as a tool to detect the presence of a fish
species within a reach when adult numbers are difficult to detect due to low abundances
or sampling difficulty. Additionally,the detection of larvae in a reach signifies a self-
reproducing population, which may not necessarily be the case when small numbers of
adults are observed in a study. Rock bass (Ambloplites rupestris) exhibited the least
tendency to drift when compared with all other taxa in the surveys. One drifting larval
rock bass was collected at site 1,while their adult numbers comprised 10% of the entire
adult fish assemblage. Although more refined taxonomic identification was not verified
for the other centrarchids,it is very likely that the remainder were Lepomis sp.,which
drifted in a lower numbers when compared to adults in the reach. Smallmouth bass are a
i

46
sportfish of conservation interest to Pigeon River fisheries managers. Two adult
smallmouth bass (Micropterus dolomeiu)were collected, yet no larvae were found,
suggesting either the absence of drift behavior or the lack of successful reproduction in
this study reach.
Another objective of this study was to characterize the daily temporal patterns of
drift abundance. Knowledge pertaining to diel patterns in abundance is important when
designing sampling protocols. Benefits include the ability to obtain samples that are
representative of all larval fish taxa present within a river, as well as the minimization of
costs and sampling time. All families present as larvae in the study reach drifted
significantly more during the nighttime hours than during the daytime in this study,
particularly during the 2000-2400 time block. Carter et al. (1986) observed similar
patterns of overall diel abundance with the largest numbers found during the 20:00-24:00
time period. Overall peaks in nightime drift densities have also been observed in other
studies (Brown and Armstrong 1985, Corbett and Powles 1986, Elliot 1987, Johnston et
al. 1995, Gadomski and Barfoot 1998,Marchetti and Moyle 2000,Reichard et al. 2004).
When considering individual taxa,the centrarchids, ictalurids, and cottids drifted
exclusively during the nighttime and the catostomids, cyprinids, and percids exhibited
low levels of daytime drift. Other studies have shown mixed results in terms of diel
abundance patterns for different taxa. Muth and Schmulbach(1984) observed no overall
diel pattern,with ictalurids and catastomids more abundant at night, and cyprinids and
centrarchids exhibiting no pattern. Robinson et al. (1998) found significant diel drift
patterns for only one species of cyprinid. A study examining the downstream transport of
i

j
47
darter larvae (Etheostoma rubrum) found no significant patterns in drift over a 24-hour
period(Slack et al. 2004).
Diel periodicity in'drift is not a clearly understood phenomenon. Higher nighttime
drift densities are often primarily attributed to light intensity. It is likely that most larval
fishes are able to maintain their positions in slow water zones of the river using visual
and tactile senses when sufficient light is available. Groups of larval fishes were
observed along the Pigeon River banks in slower currents during the daytime. As light
diminishes,these fishes are more susceptible to losing orientation with respect to the
stream bank, and are more likely to become accidentally entrained in the current
(Northcote 1962, Harvey 1987, and Pavlov 1994). Additionally, many freshwater fishes
are photonegative at the larval stage and maintain positions near and within the bottom
substrate during daylight hours,thus avoiding downstream displacement (Balon 1976).
Elevated turbidity levels during periods of high discharge combined with associated high
flow rates would likely result in an increase in daytime drift rates due to the loss of visual
orientation. Turbidity was not measured in this study and drift samples were not
collected during high flow events due to the inability to stabilize drift nets. Therefore,
these potential trends could not be deciphered in this study.
Active entrance into the drift may also occur and act as an important predator
avoidance mechanism. Most fish in temperate river systems are visual predators and
drifting at night would provide a survival advantage (Gale and Mohr 1978, Armstrong
and Brown 1983, Corbett and Powles 1986,Harvey 1991). Studies involving aquatic
invertebrates have demonstrated an active response to the addition of predation,with

48
nocturnal drift patterns only developing after the addition of predators to predator free
systems (Giller and Malmqvist 1998). Brown and Armstrong (1985) also considered
entering currents in order to forage as a proximate cause of nighttime drift. This
mechanism seems unlikely due to a lack of visual acuity in low light conditions and poor
swimming abilities at the larval stage.
Muth and Schmulbach(1984) suggested that differences in daytime and nocturnal
densities could be related to changes in vertical distribution of larvae. All nets were not
anchored directly adjacent to the substrate and it is possible that some fish were missed
drifting below the net's surface. Although this scenario is plausible, diel changes in
vertical distribution have not been demonstrated for the taxa collected in the Pigeon River
and most of the drift nets sampled the entire water column. This likely had minimal
impact on differences in density. Increased efficiency at avoiding sampling gear during
the daytime could also be considered as a potential impact on decreased daytime drift
densities.
The lateral distribution of drifting larvae within the Pigeon River was determined
by species specific life history characteristics and spawning site selection, combined with
the physical impacts of a river's current (Pavlov 1994). The species composition of near-
shore drift and mid-channel drift samples was similar in this study. For all taxa pooled,
larval fishes were twice as abundant in near-bank samples than mid-channel samples.
This pattern is common and has been observed in other studies of larval drift (Brown and
Armstrong 1985, Carter et al. 1986,Harvey 1991, Robinson et al. 1998, Reichard et al.
2004). When all taxa were pooled there was no relationship between body size or

I
49
developmental stage with drift distance from the bank. Other studies have found all
body sizes in near bank samples (Carter et al. 1986,Pavlov 1994, Gadomski and Barfoot
1998).
Catostomids and cyprinids were the only taxa significantly more abundant in near-
bank samples. These patterns could be associated with the location of spawning sites
closer to nearshore areas. If spawning for catostomids and cyprinids took place
throughout the lateral channel, it is likely that either larvae actively sought these
nearshore areas to procure resources or were passively transported due to river
morphology and flow turbulence. Although there was no lateral pattern in size
distribution for catostomids,their mean total length was larger than that of all other taxa.
It is possible that these larger-bodied individuals had stronger swimming abilities than
did other larvae, and were able to reach nearshore refugia more easily after emergence.
Nearshore nets were also associated with shallower depths and slower currents. These
factors may have resulted in higher nearshore densities due to complete sampling of the
vertical water column and fish drifting near the substrate or from fish actively entering
into the nets.
Body size was positively correlated with distance from the bank for the Percidae
family. An interesting developmental pattern was also observed, as the percids were the
only family with the majority of individuals captured in the pro-larval stage. Reichard et
al. (2004)found older and larger larval fish in mid-channel drift samples compared with
nearshore samples. As the yolk-sac becomes depleted, fish must swim to procure food.
Their swimming ability still remains somewhat poor in the larval stage due to the lack of

I
50
complete fin development, such that they are still susceptible to entrainment in currents.
It is possible that larger larvae were collected drifting further from the shore due to an
increased swimming ability,but the predominance of pro-larval specimens suggests that
an increase in size may have not been correlated with an increased swimming ability.
Passive entrainment of weaker swimming pro-larvae may be more influential to
downstream dispersal for this taxa when compared with other families collected in the
drift that were comprised of free swimming post-larvae.
The primary objective of this study was to examine the potential for re-
colonization of drifting larval fishes to the Pigeon River below the city of Canton from
upstream source populations. Larval drift densities below the low-head dam were found
to be less than 30% of the natural drift densities found upstream for some taxa.
Significant difference in drift density observed between the upstream site and the site
directly downstream of the low-head dams suggests that dispersal is being impaired. In
particular, the two families of conservation concern (Percidae and Cyprinidae) exhibited
a five-fold and three-fold decline in abundance, respectively. Drift of ichthyoplankton
from the upper Pigeon River to the reach downstream of Blue Ridge Paper Products,Inc.
was substantially reduced due to a variety of potential factors.
The subsequent distance a fish is transported once it enters the drift and the
proximity of spawning sites is of primary importance when considering connectivity and
colonization dynamics. Ultimately the physical components of total drift distance can be
broken down into three interrelated components: 1) the frequency of drift events, 2) the
distance drifted per event, and 3) the duration of aquatic life vulnerable to drift (Lancaster

51
et al. 1996). The heterogeneity found in longitudinal, lateral and vertical components of
the velocity structure of a river channel will have the greatest influence on the distance a
larval fish drifts..The complexity of flow patterns and how drifting fish larvae are held in
moving water and hydraulic dead zones is very influential in determining drift distances.
Bond et al. (2000) concluded that the proportion of the stream with dead water zones as
well as the spatial arrangement of dead water zones was extremely influential in
determining drift distances of aquatic insects.
Low-head dams can alter the natural hydrological and geomorphologic conditions
of a river by creating areas of slower current and by altering substrate conditions (Gillette
et al. 2005). Variations in flow and discharge have also been demonstrated to affect the
permeability of boundaries created by small dams to upstream and downstream
movement in fishes (Schlosser 1995). Some larval fishes, such as darters, were collected
at site 2 downstream of the low-head dams that most likely drifted from upstream of the
backwater area created by the low-head dams. Therefore we can conclude that the Tow-
head dams did not completely block immigration from upstream and that drift distance is
not the primary factor limiting re-colonization. It is probable that the reduction in larval
fish abundance below Canton is partially attributed to drifting larvae settling out of the
current in unsuitable habitat above the dam due to lower flow velocities and an increase
in hydraulic dead zones. Parameters such as zooplankton density, substrate composition,
amount and type of cover, abundance and type of predators, and physiochemical
variables ultimately determine habitat quality (Robinson et al. 1998). Larval fishes have
evolved distinct drift patterns and nursery habitat requirements that must be met in a

i
52
timely manner for survival. When the acquisition of environmental resources is delayed
or prevented an increase in mortality rate is likely to occur.
Variability in the taxonomic composition of the larval fish assemblage between
study sites was likely influenced by differences in adult assemblages throughout the study
reach. Natural longitudinal changes in fish assemblages do occur in river systems that
relate to changes in habitat conditions and food availability. But these changes occur
naturally only over long reaches of river systems. Ward and Stanford (1983) put forth the
serial discontinuity concept to explain the disruption of natural physical and biotic
longitudinal patterns in river systems due to the presence of dams. The disparity between
larval assemblages at the three sites is also attributed to the low-head dams alteration of
riverine habitat and assemblage composition of adult fishes. Gillette et al. (2005)
demonstrated distinct patterns in small bodied stream fish assemblages in relation to Tow-
head dams, with lentic habitat adapted species more prevalent in upstream areas
containing slower velocities and increased siltation rates. It is likely that the reproductive
capability of many native fish species that spawn in riffles has been greatly reduced by
this alteration in the Pigeon River. This has allowed for the proliferation of fishes better
adapted to more lentic conditions, illustrated by the large numbers of larval centrarchids
collected downstream of the low-head dams. White suckers were found to be most
abundant at site 3 and black redhorse were most abundant at site 1. This pattern may also
be attributed to differences in habitat above the low-head dams, as black redhorse do not
occupy reservoir like habitats and white suckers are found in a variety of lotic and lentic
habitats throughout North America (Etnier and Starnes 1993).

53
Declining drift abundances for taxa drifting from upstream are also presumably
caused by mortality from increased predation rates. High centrarchid densities upstream
of the low-head dams likely produced an increase in predation on larvae drifting in
slowed currents. In particular, the diet of juvenile smallmouth bass is often dominated by
fish larvae (Etnier and Starnes 1993). Additionally when fish did pass over the low-head
dams, they were displaced into a small hydraulic with high turbulence. This likely
resulted in higher predation rates on larval fishes directly below the spillway due to an
increased concentration of larvae and lack of swimming proficiency in the strong circular
currents.
Another potential explanation for the observed abundance differences between sites
is the entrainment of larval fishes by Blue Ridge Paper Products Inc. cooling intake
system. Entrainment of fishes through intake valves of large power plants and industrial
facilities associated with lakes and reservoirs has received significant attention and has
been determined to negatively impact larval fish populations. (Travnichek et al. 1993).
Less is known regarding the impacts of water abstraction conducted by smaller industrial
facilities that use riverine systems. Benstead et al. (1999) studied the impacts of a low-
head dam and water intake on streams in Puerto Rico. Mortality of drifting
amphidromous freshwater shrimp larvae by entrainment into a water intake ranged from
42-100%. Therefore the cooling intake at Blue Ridge Paper Products Inc. could have a
significant impact on larval fish mortality,particularly during low flow conditions..
Some taxa exhibited increased densities at site 3 downstream of the effluent mixing
zone when compared with site 2 downstream of the low-head dams. For example,black

54
redhorse and northern hogsucker densities declined from site 1 to site 2, then increased
from site 2 to site 3. This pattern is likely attributed to the presence of suitable spawning
habitat downstream of the low-head dams and upstream of site 3. An increase in drift
densities at site 3 could also have resulted from higher turbidity levels. Although
turbidity measurements were not taken, an obvious difference in color was observed
downstream of the effluent release site due to the presence of tannins released from the
paper manufacturing process. Decreased visibility resulting from the presence of
suspended particulates could result in an increase in passive drift through a loss of visual
orientation. Another potential stimulus for higher drift densities at site 3 could be the
presence of chemical pollutants. Gale and Mohr (1978) suggested that larval fish may
actively enter the drift in response to pollutants, moving to creek mouths or further
downstream. The higher temperatures and conductivity values observed at site 3 did not
appear to hsve an impact on either temporal or spatial patterns of larval fish drift.
Management efforts focused on restoring biological integrity to lotic systems rely
on specific ecological knowledge of a particular system and its resident biota.
The body of knowledge concerning downstream movements of stream fishes is largely
incomplete(Schmutz and Jungwirth 1999). In particular there have been relatively few
investigations of larval fish ecology in rivers similar to the Pigeon River in size and
geographic location. In systems such as these,where fish diversity is often high,more
detailed taxonomic information is needed to refine study results. Detailed information
concerning the spatial dynamics of larval fish populations and early life will be of
increasing value to fisheries conservation and management with species specific

i
55
information. The collection, sorting and identification of larval fish samples can be
extremely labor intensive. More focused studies that take into account the temporal and
spatial patterns of species of conservation concern would be more cost effective.
Future studies that are able to link larval drift behavior and adult population
dynamics would be of value in better understanding larval dispersal and recruitment
mechanisms. An attempt to decipher the primary cause of lowered densities downstream
of the low-head dams should also be undertaken. Studies focusing on larval fish
predation (juvenile and adult diet studies),the relative abundance of larval fish entrained
by the paper mill, and flow dynamics throughout the study reach could potentially
provide management direction. If entrainment was found to be a major source of
mortality, a cessation of water extraction during peak flow periods (20:00-24:00)would
help to alleviate this issue.
Conclusions.
The abundance and survival of fish larvae are closely tied to environmental
changes, and studies focusing on this life history stage can be valuable in assessing
anthropogenic disturbance (Kelso and Rutherford 1996). Presently, it is believed that
fishes determined to be of conservation concern are drifting downstream in greatly
lowered densities to the recovering section of the Pigeon River from source populations.
Natural mortality rates are extremely high during the early life stages of fishes and any
increase in mortality, as observed in this study; could potentially result in recruitment
failure (Humphries and Lake 2000,Houde 2002,). The disruption of the natural physical,

56
thermal, chemical and biotic riverine environment in the study reach ultimately acts as a
population sink and likely eliminates the potential for re-colonization from upstream
source populations through larval drift.
River regulation in the form of dams of many sizes has resulted in greatly disrupted
natural flow regimes and has restricted the natural connectivity of riverine ecosystems.
Therefore fluvial systems must be viewed as spatially continuous mosaics where
organisms are highly dependent upon natural stream flow conditions in relationship to
both quantity and timing. Restoration goals will only be fulfilled when the maintenance
of natural flow patterns and disturbance regimes are achieved and natural physical and
biological processes are again intricately connected.

57
Literature Cited

58
Literature Cited
Allan, J.D. and A.S. Flecker. 1993. Biodiversity conservation in running waters.
Bioscience 43:32-43.
Angermeier,P. L. 1997. Conceptual roles of biological integrity and diversity. Pages 49-
65 in J. E. Williams, C. A. Wood, and M. P. Dombeck, editors. Watershed
Restoration: Principles and Practices. American Fisheries Society, Bethesda,MD
f
Auer,N.A. editor. 1982. Identification of larval fishes of the Great Lakes Basin with
emphasis on the Lake Michigan Drainage. Great Lakes Fishery Commission,
Special Publication 82-82,Ann Arbor,Michigan.
Balon,E.K. 1975.Reproductive guilds of fishes: a proposal and definition. Journal of the
Fisheries Research Board of Canada 32:821-864.
Benke, A.C. 1990. A perspective on America's vanishing streams. Journal of the North
American Benthological Society 91:77-88.
Benstead, J.P., J.G. March, C.M. Pringle, and F.N. Scatena. 1999.Effect of a low-head
dam and water abstraction on migratory tropical stream biota.Ecological
Applications 9(2): 656-668.
Bond,N.R., G.L.W. Perry and Barbara J. Downes. 2000. Dispersal of organisms in a
patchy steam environment under different settlement scenarios. Journal of Animal
Ecology. 69: 608-619
Bond,N.R. and P.S. Lake. 2003. Local habitat restoration in streams: Constraints on the
effectiveness of restoration for stream biota. Ecological Management and
Restoration 4(3): 193-198.
Brown, A.V. and M.L. Armstrong. 1985.Propensity to drift downstream among various
species of fish. Journal of Freshwater Ecology 3:3-17.
Carter, J.G.,V.A. Lamarra, and R.J. Ryel. 1986. Drift of larval fish in the upper Colorado
River. Journal of Freshwater Ecology 3:567-577.
Chick, J.H.,M.A. Pegg and T.M. Koel. 2006. Spatial patterns of fish communities in the
upper Mississippi River system: Assessing fragmentation by low-head dams.
River Research and Applications 22:413-427.
Coombs, J.A. 2004. Pigeon River Revival. Wildlife in North Carolina. December 2004.

59
Corbett,B.W., and P.M. Powles. 1986. Spawning and larva drift of sympatric walleyes
and white suckers in an Ontario stream. Transactions of the American Fisheries
Society. 115:41-46.
Dynesius,M. and C.Nilsson. 1994. Fragmentation and flow regulation of river systems
in the northern third of the world. Science 266:753-762.
EA. 1996. A study of the aquatic resources and water quality of the Pigeon River.
Prepared for Champion International Coporation, Canton,NC. Prepared by EA
Engineering, Science, and Technology,Inc.
EA. 2001. Canton Mill. Balanced and indigenous species study for the Pigeon River.
Prepared for Blue Ridge Paper Products, Canton,NC. Prepared by EA
Engineering, Science, and Technology,.Inc.
Elliot,J.M. 1987. The distances traveled by downstream-moving trout fry,Salmo trutta,
in a Lake District stream. 'Freshwater Biology 17:491-499.
Etnier, D.A. and W.C. Starnes. 1993. The fishes of Tennessee. I"edition.University of
Tennessee Press. Knoxville, TN.
Floyd,K.B., R.D. Hoyt, and S. Timbrook. 1984. Chronology of Appearance and Habitat
Partitioning by Stream Larval Fishes. Transactions of the American Fisheries
Society. 113:217-223.
Fuiman,L.A., J.V. Conner,B.F. Lathrop, G.L. Buynak, D.E. Snyder, and J.J. Loos. 1983,
State of the art identification for cyprinid fish larvae from eastern North America.'
Transactions of the American Fisheries Society 112:319-332.
Gadomski,D.M. and C.A Barfoot. 1998. Diel and distributional abundance patterns of
fish embryos and larvae in the lower Columbia and Deschutes rivers.
Environmental Biology.of Fishes. 51:353-368.
Gale,W.F. and H.W. Mohr. 1978. Larval fish drift in a large river with a comparison of
sampling methods. Transactions of the American Fisheries Society 107: 46-55.
Giller, S. and B. Malmgvist. 1998. The Biology of Streams and Rivers. Oxford
University Press, Oxford. Pp. 296.
Gillette,D.P., J.S. Tiemann, D.R. Edds, and M.L. Wildhaber. 2005. Spatiotemporal
patterns of fish assemblage structure in a river impounded by low-head dams.
Copeia 3:539-549.

60
Harvey, B.C. 1987. Susceptibility of young-of-the-year fishes to downstream
displacement by flooding. Transactions of the American Fisheries Society
116:851-855.
Harvey, B.C. 1991. Interactions of abiotic and biotic factors influencing larval fish
survival in an Oklahoma stream. Canadian Journal of Fisheries and Aquatic
Sciences 48:1476-1480.
Hogue, J.J.,Jr., R. Wallus, and L.K. Kay. 1976. Preliminary guide to the identification of
larval fishes in the Tennessee River. Tennessee Valley Authority Technical Note
B19. Norris, Tennessee.
Houde, E.D. 2002. Mortality. In Fishery Science: The Unique Contributions of Early Life
Stages (Fuiman, L.A. and R.G. Werner eds.),pp. 61-87. Oxford,UK: Blackwell
Science.
Humphries,P. and P.S. Lake. 2000. Fish Larvae and the Management of Regulated
Rivers. Regulated Rivers: Research and Management. 16:421-432.
Jackson,D.A, P.R. Peres-Neto, and J.D. Olden. 2001. What controls who is where in
freshwater fish communities—the roles of biotic, abiotic, and spatial factors.
Canadian Journal of Fisheries and Aquatic Sciences 58:157-170.
Johnston,T.A., M.N. Gadboury,R.A. Janusz, and L.R. Janusz. 1995. Larval fish drift in
the Valley River,Manitoba: influence of abiotic and biotic factors, and
relationships with future year-class strengths. Canadian Journal of Fisheries and
Aquatic Science 52: 2423-2431.
Kay,L.K.,R. Wallus and B.L. Yeager. 1994. Reproductive biology and early life history
of fishes of the Ohio River Drainage. Volume II— Catostomidae. Tennessee
Valley Authority, Chatanooga. 242 pp.
Kelso, W.E. and D.A. Rutherford. 1996. Collection, Preservation, and Identification of
Fish Eggs and Larvae.Pages 255-302 in B.R. Murphy and D.W. Willis, editors.
Fisheries techniques, 2Id edition. American Fisheries Society, Bethesda,
Maryland.
Kennedy,T.B and G.L. Vinyard. 1997. Drift ecology of western catostomid larvae with
emphasis on Warner suckers, Catostomus wamerensis (Teleostei). Environmental
Biology of Fishes 49: 187-195.
Lake,P.S. 2001. On the maturing of restoration: Linking ecological research and
restoration. Ecological Management and Restoration 2(2): 110-115.

61
Lancaster, J. A.G. Hildrew and Charlotte Gjerlov. 1996. Invertebrate drift and
longitudinal transport processes in streams. Canadian Journal of Fisheries and
Aquatic Science 53:572-582.
Marchetti,M.P. and P.B. Moyle. 2000. Spatial and temporal ecology of native and
introduced fish larvae in lower Putah Creek, California. Environmental Biology of
Fishes. 58:75-87.
McGowan,E.G. 1988. An illustrated guide to larval fishes from three North Carolina
piedmont impoundments. Carolina Power and Light Company,New Hill,N.C.
133pp.
Meyer,F.A. 1970. Development of some larval centrarchids. Progressive Fish Culturist.
32:130-136.
Muth,R.T. and J.C. Schmulbach. 1984. Downstream transport of fish larvae in a shallow
prairie river. Transactions of the American Fisheries Society 113:224-230.
Neves,R.J. and P.L. Angermeier. 1990. Habitat alteration and its effects on native fishes
in the upper Tennessee River system, east-central U.S.A. Journal of Fish Biology.
37 (Supplement A),Pp. 45-52.
North Carolina Department of Environment and Natural Resources (NCDENR). 2003.
Basinwide assessment report. French Broad River basin.North Carolina
Department of Environment and Natural Resources. Division of Water Quality.
Water Quality Section. Raleigh,NC.
North Carolina Department of Health and Human Services (NCDHHS). 2007. State
Removes Last Fish Consumption Advisory Below Canton Paper Mill,NC. Public
Affairs Office,Raleigh,NC.
Northcote,R.G. 1962. Migratory behavior of juvenile rainbow trout,Salmo gairdneri, in
outlet and inlet streams of Loon Lake, British Columbia. Journal of the Fisheries
Research Board of Canada 19:201-270.
Palmer,M.A., R.F. Ambrose, and L.N. Poff, 1997.Ecological theory and community
restoration ecology. Restoration Ecology 5:291-300.
Pavlov,D.S. 1994. The downstream migration of young fishes in rivers: Mechanisms and
distribution. Folia Zoologica 43:193-208.
Porto,L.M., R.L.McLaughlin, and D.L.G. Noakes. 1999. Low-head barrier dams restrict
the movement of fishes in two Lake Ontario streams.North American Journal of
Fisheries Management 19:1028-1036.

62
Reichard,M.,P. Jurajda, and C. Smith. 2004. Spatial distribution of drifting cyprinid
fishes in a shallow lowland river. Archiv fur Hydrobiologia (Supplement) 159(3):
395-407.
Ricciardi, A. and J.B. Rasmussen. 1999. Extinction Rates of North American Freshwater
Fauna. Conservation Biology. 13(5): 1220-1222.
Robinson,A.T.,R.W. Clarkson and R.E. Forrest. 1998. Dispersal of larval fishes in a
regulated river tributary. Transactions of the American Fisheries Society.
127:772-786
Schlosser,I.J. 1985. Flow regime,juvenile abundance, and the assemblage structure of
stream fishes. Ecology 66: 1484-1490.
Schlosser,I.J. 1991. Stream fish ecology: A landscape perspective. BioScience 41:704-
712.
Schlosser, I.J. 1995. Dispersal,boundary processes, and trophic-level interactions in
streams adjacent to beaver ponds. Ecology. 76:908-925.
Schmutz, S. and M. Jungwirth. 1999. Fish as indicators of large river connectivity: The
Danube and its tributaries. Archiv fur Hydrobiologia (Supplement) 115/3:329-
348.
Simon,TY & R.Wallus. 2003. Reproductive Biology and Early Life History of Fishes in
the Ohio River Drainage: Volume III Ictaluridae— Catfish and Madtoms. CRC .
Press, Boca Raton,FL. Pp. 232.
Simon,T.P &R.Wallus. 2006. Reproductive Biology and Early Life History of Fishes in
the Ohio River Drainage: Volume IV Percidae-Perch,Pikeperch, and Darters.
CRC Press,Boca Raton, FL. Pp. 617.
Slack, W.T., S.T. Ross and J.A Ewing, E I. 2004. Ecology and population structure of the
bayou darter,Etheostoma rubrum: disjunct riffle habitats and downstream
transport of larvae. Environmental Biology of Fishes 71:151-164.
Snyder, D.E. 1976.Terminologies for intervals of larval fish development. Pages 41-60
in J. Boreman, editor. Great Lakes fish egg and larvae identification;proceedings
of a workshop. U.S. Fish and Wildlife Service FWS/OBS-76/23.
Snyder,D.E. 1990. Fish larvae— ecologically distinct organisms.Pages 20-23 in M.B.
Bain, editor. Ecology and assessment of warm water streams: workshop synopsis.
U.S. Fish and Wildlife Service Biological Report 90(5).

63
Soule, M.E. and G.H. Orians. 2001. Introduction. Pages 1-10 in M.E. Soule and G.H.
Orians, editors. Conservation Biology: Research Priorities for the Next Decade.
Island Press,Washington,D.C.
Townsend, C.R. 1989. The patch dynamics concept of stream community ecology.
J.N.Am Benthol.Soc. 8:36-50.
Travnichek, V.H., A.V. Zale,W.L. Fisher. 1993. Entrainment of ichthyoplankton by a
warmwater hydroelectric facility. Transactions of the American Fisheries Society
122(5):709-716.
Turner,T.F. 2001. Comparative Study of Larval Transport and Gene Flow in Darters.
Copeia 3:766-774.
Underwood,A.J. 1997. Experiments in Ecology: Their logical design and interpretation
using analysis of variance. Cambridge University Press, Cambridge. 504 pp.
University of Tennessee. 2005. Unpublished study of adult fish populations in the Pigeon
River,NC.
University of Tennessee Student Chapter of the Wildlife and Fisheries Society
(UTSCWF). 2007 Pigeon River Recovery Project. 22 March 2007.
<http:fwf.ag.utk.edu/Sites/Pigeon/Webpages/history.asp>
Vannote,R.L., G.W. Minshall,K.W. Cummins, J.R. Sedell, and C.E. Cushing. 1980. The
River continuum Concept. Can. J. Fish. Aquat. Sci. 37:130-137.
Wallus,R. T.P. Simon, and B.L. Yeager. 1990.Reproductive biology and early life
history of fishes in the Ohio River drainage, Volume 1: Aciperseridae through
Esocidae. Tennessee Valley Authority, Chattanooga,Tennessee.
Ward, J.V. and J.S. Stanford. 1979. The Ecology of Regulated Streams,Plenum Press,
Boca Raton.
Ward,J. V. &J.S. Stanford. 1983. The serial discontinuity concept of lotic ecosystems.
Ch. 2 in T.D. Fontaine and S.M. Bartell (editors), Dynamics of lotic ecosystems.
Ann Arbor Science Publishers.
Ward,J.V. and J.A. Stanford. 1995. The serial discontinuity concept: extending the
model to floodplain rivers. Regulated Rivers:Research and Mangagement
10:159-168.

64
Warren,M.L.,B.M Burr, S.J.Walsh,H.L.Bart,R.C. Cashner,D.A.Etnier,B.J. Freeman,
B.R. Kuhajda,R.L. Mayden,H.W. Robison, S.T. Ross, and W.C. Starnes. 2000.
Diversity,Distribution, and Conservation Status of the Native Freshwater Fishes
of the Southern United States. Fisheries. 25: (10) 7-29.
Waters,T.F. 1972. The drift of stream invertebrates. Annual Review of Entomology
17:53-72.
Wiens,J.A. 2002. Riverine Landscapes: Taking landscape ecology into the water.
Freshwater Biology 47:501-515.
Zar, J.H. 1999. Biostatistical Analysis. 4th Edition. Prentice-Hall,Inc.,Upper Saddle
River,NJ. 931 pp.