ABSTRACT
Michael D. Walker. FISH UTILIZATION OF AN INUNDATED SWAMP-STREAM
FLOODPLAIN. (Under the direction of Dr. Robert P. Sniffen and Dr.
Charles W. O'Rear) Department of Biology, July 1984.
The inundated floodplain of Creeping Swamp (Pitt Co., N.C.) was
sampled weekly for fish from November 1979 through May 1980. The fish
community was dominated by pirate perch and redfin pickerel. Other
frequently occurring species included flier, mud sunfish, eastern
mudminnow, American eel, banded sunfish, creek chubsucker, bowfin and
swampfish. Two-way fixed weir traps were the most effective fish
collection devices for both the inundated floodplain and the stream.
Three environmental factors affected fish occurrence and movement:
water level, water temperature and dissolved oxygen. Small water level
fluctuations caused large changes in the amount of aquatic habitat.
Fish movements were minimal when water temperature was below 60C.
Dissolved oxygen was not a limiting factor on fish occurrence during the
wet season (typical ly December through Apri 1 ) except in shal 1 ow edge
areas of the floodplain.
Abundant standing crops of aquatic invertebrates were available for
fish foraging early in the wet season. Crayfish were present in large
numbers and represented an important food source for larger (>10 cm)
fish in the swamp. Fish fed almost exclusively on invertebrates,
especially the dominant crustaceans (e.g. amphipods, isopods, copepods,
crayfish) and chironomids. Crayfish and other relatively large
invertebrates appeared to have replaced fish in the diet of the larger
piscivorous fish species (e.g. redfin pickerel and bowfin).
Oligochaetes constituted a large portion of the benthic invertebrates in
the swamp, but these rapidly-digested soft-bodied organisms were seldom
found in fish stomachs.
The inundated floodplain/stream system serves as a spawning and
nursery area for several swamp-stream fish species during the wet
season. Fifty-one percent of fish collected in February and March were
found to be in spawning condition.
FISH UTILIZATION
OF AN
INUNDATED SWAMP-STREAM FLOODPLAIN
A Thesis
Presented to
the Faculty of the Department of Biology
East Carolina University
In Partial Fulfillment
of the Requirements for the Degree
Master of Science in Biology
by
Michael D. Walker
July 1984
T. JOYNr:x
mast CASOLIKA UNíVSi.r'íTT
FISH UTILIZATION
OF AN
INUNDATED SWAMP-STREAM FLOODPLAIN
by
Michael D. Walker
COMMITTEE MEMBER
CHAIRMAN OF THE DEPARTMENT OF BIOLOGY
n
Charles E. Bland, Ph.D.
DEAN OF THE GRADUATE SCHOOL
J'oseph G. Boyetté, Ph.D.
ACKNOWLEDGMENTS
I wish to express my deep appreciation to my graduate committee,
Dr. Robert P. Sniffen (supervisor of the research project). Dr. Charles
W. O'Rear (committee co-chairman). Dr. Susan J. McDaniel and Dr. William
H. Queen, for their guidance and encouragement. I want to thank Dr.
Sniffen for his persistence and willingness to assist me in all phases
of this study. I would also like to acknowledge the generous aid of Dr.
Donald W. Stanley in reviewing the manuscript.
I am especially grateful to Loede Harper and Ken Sholar for their
assistance in the field work. Carl Hayes also provided frequent
guidance on data processing and computer applications.
Mrs. Sandra Tomlinson professionally typed the numerous drafts and
final manuscript. Her encouragement and attention to detail made
completion of the thesis easier and faster.
The research was supported by grants to Dr. William H. Queen,
Director of the Institute for Coastal and Marine Resources, East
Carolina University and Dr. Robert P. Sniffen from the Environmental
Protection Agency (R-806988010 and R-806988020).
Finally, and most importantly, I want to thank my wife, Cindy, for
the unending love and support she has given me during my graduate study.
DEDICATION
This thesis is dedicated to the memory of my father, Edgar D. Walker.
TABLE OF CONTENTS
PAGE
LIST OF TABLES vi
LIST OF FIGURES vili
INTRODUCTION 1
Study Area 3
METHODS 9
MOVEMENT STUDY 9
FOOD HABITS STUDY 12
Invertebrate Standing Crop 12
Fish Gut Analysis 14
PHYSICAL AND CHEMICAL MEASUREMENTS 17
SPAWNING/NURSERY AREA OBSERVATIONS 17
ROTENONE STUDY 18
RESULTS 20
MOVEMENT STUDY 22
Fish and Crayfish Catch 22
Diurnal and Nocturnal Habits 26
Seasonal Occurrence and Movement 26
Environmental Effects on Fish Occurrence and Movement .... 30
Water Level 30
Water Temperature 30
Dissolved Oxygen 35
FOOD HABITS STUDY 35
Invertebrate Standing Crop 35
Floodplain Taxa 35
Stream Taxa 40
Comparison of Collection Techniques 40
Fish Gut Analysis 45
Collection Techniques 45
Stomach Fullness 45
Stomach Contents . 48
SPAWNING/NURSERY AREA OBSERVATIONS 48
Fish Spawning Condition 48
Young-of-the-Year 52
ROTENONE STUDY 52
DISCUSSION 56
Species Composition of Fish Occupying the Floodplain 56
Diurnal and Nocturnal Habits 57
Seasonal Occurrence and Movement 58
Comparison of Fish Collection Techniques 58
Environmental Effects on Fish Occurrence and Movement 60
Water Level 60
Water Temperature 61
Dissolved Oxygen 62
TABLE OF CONTENTS (continued)
PAGE
Food Habits of Fish Occupying the Floodplain 63
Invertebrate Food Supply 63
Food Habits 65
Spawning/Nursery Area Observations 66
Importance of Creeping Swamp to Downstream Systems 67
Ecological Importance 67
Economic Importance 68
SUMMARY 69
LITERATURE CITED 71
LIST OF TABLES
TABLE PAGE
1. Stream Water Quality at the Creeping Swamp
Study Site 5
2. Invertebrates Length versus Dry Weight Relationships 15
3. Fish and Crayfish Catch in Creeping Swamp from November
1979 Through May 1980 21
4. Fish and Crayfish Catch by Collection Device 23
5. Number of Sets, Total Number of Set Days, Fish and
Crayfish Catch per Set Day by Collection Device 24
6. Comparison of Floodplain and Stream Fish Catches by
Using Fixed Weir Traps and Fyke Nets 25
7. Comparison of Floodplain and Stream Fish and Crayfish
Catches Using Fixed Weir Traps and Fyke Nets 27
8. Monthly Catches of Fish and Crayfish in Creeping Swamp
During the 1979-80 Wet Season by Month 28
9. Characteristics of Major Spates for the 1979-80 Wet
Season in Creeping Swamp During Rising and Falling
Water Levels 32
10. Average Fish Catches on Rising and Falling Water
Levels During Major Spates 33
11. Floodplain-Stream Comparison of Average Dissolved
Oxygen Levels from February 1980 through April 1980 37
12. Water Temperature-Dissolved Oxygen Relationship to
Fish Catch 38
13. Invertebrates Collected on the Inundated Creeping
Swamp Floodplain Using Drift Nets, Lift Nets and
Core Samplers 39
14. Invertebrates Collected in the Creeping Swamp Stream
Using Drift Nets and Core Samplers 41
15. Average Number of Invertebrates Sampled by
Collection Device 42
16. Day-Night Comparison of Drift Net Collections 44
VI
LIST OF TABLES (continued)
TABLE PAGE
17. Floodplain-Stream Comparison of Fish Collected for
Gut Analysis 46
18. Floodplain-Stream Comparison of Average Estimated
Stomach Fullness of Fish Collected for Gut Analysis 47
19. Stomach Contents of 140 Floodplain-Caught Fish Collected
for Gut Analysis 49
20. Stomach Contents of 60 Stream-Caught Fish Collected
for Gut Analysis 50
21. Number of Male and Female Fish for Each Species in
Spawning Condition by Month 51
22. Fish Catch from Rotenone Sampling of a Section of the
Creeping Swamp Stream on May 13, 1980 53
23. Fish Length versus Weight Relationships 55
vi i
LIST OF FIGURES
FIGURE PAGE
1. Map of Location of Creeping Swamp in North Carolina
Coastal Plain 4
2. Study Site in the Creeping Swamp Floodplain 7
3. Floodplain and Stream Sampling Areas Within
Creeping Swamp 8
4. Average Combined Fixed Weir and Fyke Net Catch per
Set Day in the Floodplain and Stream by Month 29
5. Hydrograph of Water Level Fluctuations at the Creeping
Swamp Study Site During the 1979-80 Wet Season 31
6. Thermograph of Water Temperature Fluctuations at the
Creeping Swamp Site During the 1979-80 Wet Season 34
7. Average Number of Fish Caught in Fixed Weir Traps
and Fyke Nets per Collection Device per Set Day
versus Water Temperature 36
vi 1 i
INTRODUCTION
Stream floodplains in eastern North Carolina are often inundated
with flowing water during winter and early spring, creating seasonal
swamp-stream environments. This inundation may increase the stream
system aquatic area by one to two orders of magnitude (Huish and Pardue
1978). Hardwood forests cover most of the floodplains and contribute
large amounts of organic matter to the floodplain swamps during the
autumnal leaf fall just prior to seasonal flooding. The accumulated
litter is a rich organic substrate for microbial colonization and
subsequent development of an abundant aquatic invertebrate community
(Sniffen 1981). Consequently, a plentiful food source is available to
swamp-stream fish during floodplain inundation. Since the spawning
season of most swamp-stream fish in eastern North Carolina occurs during
floodplain inundation (Huish and Pardue 1978), the floodplain may
function as an important spawning and nursery area.
Although few studies have been made of fish utilization of
temperate stream floodplains, the importance of these areas has been
recognized for several decades. Forbes and Richardson (1920), in their
survey of the Illinois River and its fishery, noted that a "wider range
and breeding ground and a greater food supply are afforded to the fishes
of the stream" during times of flood. Crossman (1962) reported that the
redfin pickerel (Esox americanos a.), one of the dominant predatory
fishes of North Carolina stream swamps, spawns on floodplains in water
sometimes less than one foot deep. He also found young-of-the-year on
the inundated floodplain in water only 3-4 inches deep. Guillory (1979)
studied the movements of large numbers of fish that migrate from the
2
Mississippi River to adjacent floodplains during spring floods. He
suggested that these lateral movements are spawning and feeding
migrations, resulting in large numbers of juvenile fish that contribute
to the recruitment of fish populations in the river channel.
Most of the research on how fish utilize inundated floodplains has
been conducted in the tropics, where environmental factors affecting
fish populations are relatively stable (but with distinct wet and dry
seasons). Probably the most comprehensive publication resulting from
these studies is Welcomme's (1979) Fisheries Ecology of Floodplain
Rivers. Wei comme states that "flood rivers are complex ecosystems which
behave in an essentially similar manner in both the tropical and
temperate zone." Even though winter climates establish seasonality in
temperate rivers, tropical floodplain rivers endure 'physiological
winters' during the dry season when fish growth rates are suppressed
(Wei comme 1979).
Channelization of swamp streams has considerable effect upon the
stream fauna. Tarplee e_t (1971) found from a study of 46
channelized and 28 natural swamp streams in eastern North Carolina that
the channelized streams contained 75% less fish biomass than the natural
streams. The decrease was thought to have resulted from habitat changes
within the channelized stream channels. However, it might have been
caused by elimination of floodplain inundation and the resultant loss of
most of the aquatic habitat and invertebrate food supply.
The overall goal of this study was to determine to what degree
swamp-stream fish utilize the adjacent floodplains during the wet
season. The following specific questions were posed:
3
1. What fish species inhabit the inundated floodplain?
2. What environmental factors affect their occurrence and
movement?
3. What are the food habits of fish occupying the floodplain?
4. Does the inundated floodplain function as a fish spawning
and/or nursery area?
In order to answer these questions, a fish collection regime had to
be developed to determine fish presence and spawning condition and to
monitor fish movements. A food habits survey was conducted including
measurement of invertebrate standing crop and gut analyses. Physio-
chemical data was collected to determine environmental effects. The
developmental nature of the fish sampling regime and data collection in
the 1979-80 wet season has produced semi-quantitative results.
Study Area
The Creeping Swamp watershed is located in the North Carolina
Coastal Plain at the junction of Pitt, Beaufort and Craven Counties
(Figure 1). Comprised of 70% woodland, the 80 km2 watershed northeast
of N.C. Highway 43 has some (30%) agricultural development, periodic
lumbering activities in scattered pine plantations and no major sources
of pollution (Kuenzler £t 1977). Elevations vary from 6-18 meters
above sea level with slight slopes producing a flat to gently rolling
topography (Yarbro 1979).
This research was conducted in the floodplain of Creeping Swamp, a
second or third order stream (average current velocity = 10-15 cm/sec).
The stream water has dark color, moderate acidity, low conductivity and
turbidity, and low plant nutrient concentrations (Table 1). The
4
Figure 1. Map of Location of Creeping Swamp in North Craolina Coastal
Plain. Dashed lines indicate watershed boundaries (from
Sniffen 1981).
Table 1. Stream Water Quality at the Creeping Swamp Study Site (from Sniffen 1981).
Parameter Wet Season (Dec. 15-April 30) Dry Season (May 1-Dec. 14)
Mean S.D. Range n Mean S.D. Range n
Conductivity (pmho cm"^) 55.5 13.8 38-81 21 80.4 48.3 36-217 29
Color (Std, Pt. units) 56.4 22.5 20-100 21 160.0 183.8 40-1100 25
pH 5.3 0.3 4.4-5.9 20 5.8 0.4 4.8-6.5 23
Turbidity (JTU) 3.8 1.6 2-7 20 11.3 11.3 2-40 29
Nitrate Nitrogen (nig«l"M .027 .032 .005-. no 24 .094 .256 .005-1.27 24
Ammonium Nitrogen (mg-l"M .027 .043 .005-.227 24 .475 1.08 .012-3.78 24
Total Organic Nitrogen (mg-l"M .225 .091 .108-.473 24 .629 .805 .119-3.93 24
Filterable Reactive Phosphorus (mg-1"^) .003 .001 .001-.006 24 .005 .004 .001-.015 24
Filterable Unreactive Phosphorus {mg-l“l) .006 .004 .002-.015 24 .014 .014 .002-.057 25
Particulate Phosphorus (mg'l"M .006 .005 .000-.016 24 .045 .074 .004-.315 24
Total Phosphorus (mg’l"M .015 .007 .006-.031 24 .063 .088 .007-.370 24
cn
6
floodpl-ain is covered by a mixed hardwood forest with a wel 1-developed
canopy shading the forest floor. The resulting low light intensity
limits plant growth to various vines and sparse herbaceous vegetation.
Creeping Swamp streamf 1 ow is intermittent, with floodplain
inundation normally occurring from December to April. Streamflow
continues until June, followed by a dry season when stream flow ceases
for 20 - 130 days (Sniffen 1981). In late autumn, forest transpiration
decreases, allowing rainfall to re-establish the shallow ground water
and stream flow. Although current velocities are normally low during
the wet season, the swamp is usually a wel 1-oxygenated lotie system
which flows over the forest leaf litter substratum (Sniffen 1981). Even
though floodplain widths vary from approximately 100 to 500 meters
(Figure 2), water levels and percent swamp inundation exhibit relatively
rapid rises and slow declines following rainfall events (spates) during
the wet season (Mul hoi land 1979).
Sampling was performed in a remote area of Creeping Swamp
approximately 300 meters from the end of a private unpaved road
adjoining S.R. 1800 (Figure 2). Two floodplain areas were sampled, each
approximately two hectares in size and located about 200 meters apart
along the Creeping Swamp primary stream channel (Figure 3).
Figure 2. Study Site in the Creeping Swamp Floodplain (from
Sniffer 1981).
8
Figure 3. Floodplain (Fl, F2) and Stream (SI, S2) Sampling
Areas Within Creeping Swamp (refer to Figure 2).
Two-way fixed weir traps indicated at sampling
sites. High ground within floodplain inundated
only during times of extremely high water levels.
METHODS
MOVEMENT STUDY
In order to determine when and to what extent fish inhabit the
fl oodpl ai n areas, all fi sh movements i nto and out of two fl oodpi ai n
study sites (Fl, F2; Figure 3) were monitored weekly during the wet
season from November 1979 through May 1980. For comparison, fish
movements in the adjacent stream channel (Sl-primary, S2-secondary) were
monitored simultaneously.
The topographic relief within the floodplain boundaries shown in
Figure 3 is only about 1.2 meters. When the water level is at a
relative elevation of 17 centimeters, stream flow ceases and only stream
pools contain any water. At 40-50 cm the stream channels are full and
the floodplain sloughs contain some non-flowing water. At 30 cm all of
the floodplain except the high ground (Figure 3) is inundated with
slowly (0-15 cm*sec-l) flowing water. At 120 cm the entire floodplain
becomes inundated. Water levels above 120 cm only occur infrequently.
Proven methods for sampling fish populations on inundated
floodplains similar to Creeping Swamp could not be found in the
literature. In a Mississippi River floodplain study, Guillory (1979)
used seines, trammel nets and experimental gill nets to monitor fish
movements from the river laterally onto the adjoining floodplain.
Creeping Swamp, however, has a heavily wooded floodplain with an
abundance of underwater snags and obstructions that would severely
hinder the use of such seines and nets. Consequently, this study
required that various methods be tested until the most efficient
collection techniques were determined.
10
Low fish densities were anticipated because of the greatly expanded
aquatic habitat during floodplain inundation. Therefore, most collection
devices were placed in primary floodplain drainage sloughs to take
advantage of natural topographic barriers that concentrated fish into
smaller and deeper locations. Collection devices were placed near the
confluence of primary floodplain drainage sloughs and the stream to
monitor fish movements into and out of floodplain areas (Figure 3).
Stream sampling sites were located just upcurrent of the slough-stream
channel confluence (Figure 3).
Collection devices used to monitor fish movement included fyke
nets, two-way fixed weir traps, minnow traps, baited minnow traps
(baited with fish and crayfish scraps) and small, one-way, movable weir
traps. Fyke nets (0.64 cm mesh) were used only early in the wet season
until two-way fixed weir traps could be constructed and set. Wings of
the weirs were constructed of 0.64 cm mesh plastic netting placed at 45-
degree angles to the trap opening and extended to high ground. Two-way
trap boxes (1.5 m long, 0.6 m wide and 0.3 m deep) had 0.64 cm mesh
plastic netting with entrance apertures 10 cm in diameter. Entrance
apertures of this size allowed several size classes of fish to enter the
trap, incl uding 1 arger predator species (e.g. Esox americanos a.. Ami a
cal va). Leaf 1 itter pi aced in the bottoms of the weir traps provided
shelter for small fish and crayfish, reducing within-trap predation.
Smal 1 one-way movable weir traps (0.45 m long, 0.30 m wide and 0.15 m
deep) that had 0.32 cm mesh plastic netting with entrance apertures 2-3
cm in diameter were also used at the fixed weir sites to help detect
movements of small fish and prevent in-trap predation by larger fish.
Standard minnow traps (0.64 cm mesh wire, 2-3 cm entrance apertures)
11
were used to determine the presence of small fish within the floodplain,
while baited minnow traps were set to help detect scent-oriented fish
such as catfish and madtoms.
The following data were recorded for all fish and crayfish captured
during the movement study: (1) species, (2) total length (or carapace
length), (3) sex and spawning condition (if apparent), (4) capture
location and (5) direction of movement (upcurrent or downcurrent). Fish
and crayfish weights were calculated using length versus wet weight
power equations given by Carlander (1977) and those developed in this
study using the Statistical Analysis System (SAS) non-linear regression
(NLIN) procedure with a general power equation:
WT = BqL^I
where: WT = wet weight in grams,
L = length in centimeters, and
Bq and Bi = constants (Goodnight 1979).
Initially, various set times (1-8 days) were used to determine the
most efficient sampling duration for each collection device. One-day
set periods were found to be most effective and were used for the
remainder of the study. Special effort was made to sample at the onset
of and immediately after major rainfall events to determine what effect
rapidly increasing and decreasing water levels have on fish movements
into and out of the floodplain.
12
FOOD HABITS STUDY
Invertebrate Standing Crop
Benthic and drift samples were taken when fish were collected for
gut analysis so that invertebrates available to fish on the inundated
floodplain could be determined. Some samples were also taken in the
stream channel for comparison. Benthic invertebrates (excluding
crayfish) were more extensively sampled by Sniffen (1981) and his
findings are most likely more representative of the overall benthic
invertebrate production in Creeping Swamp than was possible to determine
from invertebrate sampling in this study. However, because conditions
fluctuate from year-to-year, invertebrate sampling was required for a
concurrent comparison to fish stomach contents. Three collection
methods were used: (1) smal 1 diameter cores, (2) m2 1 ift nets, and (3)
drift nets.
CORES. Three to six benthic samples (8.4 cm depth) were taken with
a small diameter (5.5 cm) coring tube at randomly selected locations in
the collection area. The 3-6 cores for each area were then combined for
processing as a composite sample. The sample was preserved in the field
with 70% ethanol. Rose bengal was added to the ethanol later to
facilitate separation of invertebrates from detritus. Small diameter
core sampling is designed for accurate enumeration of all types of
benthic invertebrates which usually occur in clumped or contiguous
distributions in aquatic environments (Elliott 1973). For the relatively
small or less mobile invertebrates such as ol i gochaetes, flatworms,
chironomids, micro-crustaceans, nematodes, diptera larvae and bivalves.
13
it is an excellent method. These taxa constitute over 98% of the
invertebrates on the floodplain (Sniffen 1981).
LIFT NETS. For larger and more mobile invertebrates such as
crayfish and adult beetles, core sampling is not a very efficient
technique. These relatively large invertebrates were found in the gut
analyses of swamp fish species in Eastern North Carolina (reported by
Huish and Pardue 1978) and were therefore suspected to be an important
component of the diet of larger (>10.0 cm) fish in Creeping Swamp. To
sample these more mobile invertebrates, I randomly placed three 1.0 m'-
lift nets of 0.2 cm mesh on one of the f 1 oodpl ai n sampl ing areas. The
lift nets were placed beneath the leaf litter and left undisturbed for
over one month. Harvested samples were preserved with 70% ethanol
containing rose bengal.
DRIFT NETS. Drift nets (363-micron mesh, entrance apertures 50 cm
in diameter) were used during daylight hours and at night to determine
when and what species of invertebrates were available to fish in the
floodplain and stream drift. Nets were set for periods ranging from 3
to 24 hours in length. Duration of set and current velocity at the
mouth of the net were recorded for each set. Drift samples were
preserved with 70% ethanol containing rose bengal. The volume of water
sampled by each drift set was calculated using net aperture size and
current flow. Results were expressed as number of invertebrates per
100 cubic meters of water sampled.
Core and lift net samples were washed through three U.S. Standard
sieves (No. 60, No. 30, No. 10) in the laboratory. The material on each
sieve was placed in a white enamel pan for examination by eye, and the
invertebrates were removed with forceps. All invertebrate samples were
14
labeled and stored in 70% ethanol. Organisms were identified under a
dissecting scope and their lengths were recorded. Length versus dry
weight regression equations developed by Sniffen (1981) for the
dominant invertebrate taxa in Creeping Swamp were used to calculate dry
weights (Table 2). General equations (non-species-specific) for aquatic
and terrestrial insects were used when specific length-weight
relationships could not be found in the literature. All organisms not
identified under the dissecting scope were mounted on slides for
examination under a compound microscope. Benthic and drift invertebrate
organisms were identified to the lowest taxonomic level of which
invertebrates examined in fish stomachs could be determined.
Fish Gut Analysis
To determine food habits of fish occupying the inundated
floodplain, fish were collected in the latter half of the wet season for
gut analysis. Fish collections were also made in the stream channel for
comparison.
Collection of fish for gut analysis normally requires instantaneous
capture techniques. Available techniques included el ectroshocking,
spotlighting, rotenone and lift nets. Electroshocking was performed
with a Smith-Root backpack Type VIII-A electroshocker operated at 60
cycles, 450 volts direct current. Spotlighting involved the use of
headlamps to locate and immobilize fish at night, thus enabling capture
by hand or with dip nets. Unfortunately, the use of el ectroshocki ng,
spotlighting and lift nets did not yield enough fish from the floodplain
for statistical ly valid gut analyses. Non-instantaneous techniques
(e.g. traps with long set periods) allow trapped fish to digest
15
Table 2. Invertebrate Length versus Dry Weight Relationships. Derived from
non-linear regression equation DWT = BqL^Iwhere DWT = dry weight
in mg, L = length in mm and Bq and Bi are constants (Sniffen 1981).
Taxon Bo Bl
Turbellaria* 0.0720 1.220
Nematoda DWT = 0.005
01 igochaeta** 0.1243 1.659
Arachnida
Acariña 0.0397 2.761
Crustacea
Cladocera Length >1 DWT = 0.02; Length <1 DWT = 0.005
Ostracoda Length >1 DWT = 0.02; Length <1 DWT = 0.005
Copepoda Length >1 DWT = 0.02; Length <1 DWT = 0.005
Isopoda 0.0094 2.681
Amphipoda 0.0058 2.798
Decapoda
Astacidae+ 0.0251 3.568
Insecta
Plecoptera++ 0.0194 2.460
Odonata 0.0099 2.760
Ephemeroptera 0.0067 2.880
Hemiptera 0.0308 2.400
Trichoptera 0.0024 3.110
Diptera 0.0046 2.430
Ceratopogonidae 0.0005 2.950
Chironomidae 0.0054 2.320
Culicidae 0.0046 2.430
Simulidae 0.0052 2.760
Coleoptera++ 0.0194 2.460
Gastropoda DWT = ((Length/2)3)0.2933
Pelecypoda 0.1028 2.614
Encysted invertebrate DWT = ((Length/2)3)0.2933
Terrestrial invertebrate# 0.0305 2.620
* Body area (L x W) rather then length
**Body width rather than length
Developed from data collected during this study using carapace lengths in cm
^*General equation for aquatic insects (Smock 1980)
General equation for terrestrial insects (Rogers 1976)
16
previously ingested food and provide larger piscivorous fish the
opportunity to feed upon smaller fish and crayfish inside the trap. Two-
way fixed weir traps were used, utilizing short set times (1-4 hrs), to
supplement fish caught instantaneously. Leaf litter placed in the
bottom of the weir traps provided shelter from predation for small fish
and crayfish. Small one-way movable weir traps and minnow traps, with
small entrance apertures that exclude larger predators, were also used
with short set periods.
From preliminary sampling and observations in the study area, most
fish in the swamp appeared to be nocturnal. Intermittent shade and
sunlight reflection off the water surface of the inundated floodplain
hindered locating fish immobilized by el ectroshocki ng during daylight
hours. Intensive sampling of fish for gut analysis was conducted
primarily at night when surface reflections were absent and shocked fish
were easier to see. Headlamps and lanterns provided adequate light for
setting and harvesting traps and made electroshocking more productive.
In the field, the peritoneal cavities of all fish >5 cm total
1 ength col 1 ec ted for gut anal y si s were i njected wi th 10% formal i n to
arrest digestion. Each fish was then labeled, placed in a plastic ziploc
bag containing 10% formalin, and kept on ice until processed in the
laboratory.
Fish stomachs were removed in the lab, labeled and stored in 70%
ethanol. Percent stomach fullness was later estimated using a zero to
four scale where 0 = empty, 1 = l%-25% ful 1 , 2 = 25%-50% ful 1 , 3 = 50%-
75% full and 4 = 75%-100% full. The species, total length, sex and
spawning condition (if apparent) of each fish were then recorded and the
body discarded. Under a dissecting scope, stomach contents were
17
identified to the lowest recognizable taxonomic category and lengths
recorded. Percent total lengths of broken organisms were estimated so
that dry weights could be calculated. All organisms not identified
under the dissecting scope were mounted on slides for examination with a
compound microscope.
PHYSICAL AND CHEMICAL MEASUREMENTS
Water level in the stream was monitored continuously by a battery-
powered Stevens recorder (Type F, Model 68). Water temperatures were
measured with a glass, mercury-filled thermometer on each sampling trip.
Dissolved oxygen, conductivity, water temperature, water depth and
current vel oci ty were measured periodical ly at al 1 trap sites and at
each location where fish were collected by electroshocking and
spotlighting. Dissolved oxygen and conductivity were measured with a
YSI Model 57 Oxygen Meter. Current velocities were determined by timing
movement of entrained leaf fragments over a 50-100 cm distance.
SPAWNING/NURSERY AREA OBSERVATIONS
Sampling and observations were conducted primarily during the
latter half of the wet season to detect (1) adult fish in spawning
condition, (2) eggs, (3) larvae and (4) juveniles. Methods used were:
(1) for ripe adults—all fish collected during the movement study
were examined for spawning condition before
release. Fish collected for gut analysis
were examined during dissection;
18
(2) for eggs fish stomach contents and drift net, lift net
and core samples used for invertebrate collec-
tions were examined;
(3) for larvae fish stomach contents, drift and lift net
samples and visual field observations;
(4) for juveniles fish stomach contents, small weir traps,
minnow traps, lift nets, casual observa-
tion and rotenone sampling (discussed
below).
ROTENONE STUDY
A 75 meter long section of the primary stream channel was sampled
with rotenone on May 13, 1980. Objectives for the sampling effort were:
(1) To determine what fish were present in the stream late in the
wet season, when the floodplain was no longer inundated and
low water levels prevented effective use of other collection
techniques. Rotenone sampling provided a basis for comparing
effectiveness of other collection techniques during the latter
part of the wet season;
(2) To collect fish for use in developing length-versus-wet weight
power equations for those species not found in the literature;
(3) To detect the presence of juvenile fish too small to be
caught effectively with the other collection techniques. This
would provide evidence of the swamp's function as a spawning
and/or nursery area.
Sampling was conducted in the stream channel the first time the
floodplain dried following the end of the wet season. Water levels had
19
been falling for 14 days. One-to-two ppm of diluted 5% emulsified
rotenone was applied for one hour to the stream section (75 m length,
5.1 m average width, 0.5 m average depth and 0.04 ha total area). The
stream was blocked off with 0.64 cm mesh netting and the outflow
detoxified with potassium permanganate. All affected fish were
dipnetted and fixed in 10% formalin. Collections of dead fish were made
for two days. All fish were identified to species and total lengths and
wet weights were measured in the laboratory.
RESULTS
All fish data collected 1n this study was species specific,
however, the results are presented as a composite of the entire fish
community for fish movements, food habits and spawning/nursery area
function. Results for fish occurrence and comparison of collection
techniques are presented by species.
Three thousand six hundred and sixty-five fish (24 species) were
caught using all fish collection techniques, excluding rotenone (Table
3). Two thousand two hundred and fifty-one crayfish (2 species) were
also collected. Crayfish (Procambarus acutus a., Fal 1 icambarus uhleri)
represented 38.0% of the total catch by number and 19.0% by weight.
The most common fish species (92% of the total fish catch by number
and 86% by weight) were pirate perch (Aphredoderus sayanus), redfin
pickerel (Esox americanus a.), flier (Centrarchus macropterus), mud
sunfish (Acantharcus pomotis), eastern mudminnow (Umbra pygmaea),
American eel (Angui1 1 a rostrata) and bowfin (Amia cal va) (Table 3).
Pirate perch and redfin pickerel were the dominant fish species,
accounting for 50.9% and 20.2%, respectively, of the total number of
fish caught (Table 3). Bowfin were the largest fish captured (23.3 cm
average length, 191.2 gm average weight) and ranked third behind pirate
perch and redfin pickerel in percent total weight (9.1%), although
representing only 1.3% of the total number (Table 3). The majority
(67.3%) of species collected averaged less than 10 cm in length and 15
grams wet weight (Table 3).
The majority of the total crayfish catch by number (69.9%) and by
weight (64.7%) were Procambarus acutus a. while Fal1icambarus uhleri, a
21
Table 3. Fish and Crayfish Catch in Creeping Swamp from November
1979 Through May 1980 (excluding rotenone sampling).
N is the_total number caught, %TN is the percent total
number, TL is the average total lengthy WT is the weight,
%TWT is the percent total weight and WT is the average
weight.
Scientific Name Common Name N %TN TL (cm) WT (gm) %TWT WT (gm)
Aphredoderus sayanus pirate perch 1866 50.9 8.9 24493.0 24.2 13.1
Esox americanus a. redfin pickerel 742 20.2 15.1 34264.0 33.9 46.2
Centrarchus macropterus fl ier 253 6.9 9.2 5622.3 5.6 22.2
Acantharcus pomotis mud sunfish 214 5.8 10.5 7539.3 7.5 35.2
Umbra pygmaea eastern mudminnow 135 3.7 7.6 889.0 0.9 6.6
Anguilla rostrata American eel 106 2.9 27.1 4931.3 4.9 46.5
Enneacanthus obesus banded sunfish 75 2.0 5.6 1346.4 1.3 18.0
Erimyzon oblongus creek chubsucker 50 1.4 17.8 6275.9 6.2 125.5
Ami a calva bowfin 48 1.3 23.3 9175.9 9.1 191.2
Chologaster cornuta swampfish 29 0.8 5.9 40.9 <0.1 1.4
Enneacanthus gloriosus bluespotted sunfish 27 0.7 8.2 2093.3 2.1 77.5
Elassoma zonatum pygmy sunfish 22 0.6 3.9 10.8 <0.1 0.5
Notropis sp. shiner 21 0.6 6.2 85.1 0.1 4.1
Lepomis microlophus redear sunfish 19 0.5 9.8 520.8 0.5 27.4
Ictalurus nebulosus brown bullhead 14 0.4 14.1 2589.4 2.6 185.0
Lepomis gibbosus pumpkinseed 13 0.4 9.0 262.7 0.3 20.2
Notemigonus crysoleucas golden shiner 13 0.4 10.6 110.7 0.1 8.5
Lepomis macrochi rus bluegill 6 0.2 10.3 155.3 0.2 25.9
Lepomis gulosus warmouth 5 0.1 15.5 439.4 0.4 87.9
Le pomi's auritus long-eared sunfish 2 0.1 13.3 105.0 0.1 52.5
Et neostoma fusiforme eastern swamp darter 2 0.1 5.3 3.2 <0.1 1.6
Pomoxis nigromaculatus black crappie 1 <0.1 17.0 60.0 0.1 60.0
Lepomis cyanellus green sunfish 1 <0.1 12.0 28.5 <0.1 28.5
Gambusia affinis mosquito fish 1 <0.1 3.0 0.3 <0.1 0.3
Total 3665 101042.5
Procambarus acutus a. 1573 69.9 3.2* 15308.0 64.7 9.7
Fa 11icambarus uhleri 400 17.8 4.0* 8113.4 34.3 20.3
Inmature crayfish 278 12.3 1.6* 249.4 1.0 0.9
Total 2251 23670.8
Fi sh 3665 62.0 101042.5 81.0
Crayfi sh 2251 38.0 23670.8 19.0
Total 5916 124712.3
?Carapace length
22
slightly larger crayfish with heavier carapace, represented 17.8% of
crayfish caught. Immature crayfish, too small for identification,
accounted for the remaining 12.3% of the total crayfish catch.
MOVEMENT STUDY
Fish and Crayfish Catch
Two-way fixed weir traps and fyke nets accounted for 95% of the
total fish catch (Table 4), sampled all fish species captured during the
study except the mosquito fish and yielded 11.6 and 5.8 fish per set
day, respectively (Table 5). Pirate perch and redfin pickerel were the
species caught most frequently by the two collection devices (Table 4).
The weir traps and fyke nets also accounted for 81.3% of total crayfish
catch, yielding 5.7 crayfish per set day (Table 5).
Minnow traps, baited minnow traps and small, one-way movable weir
traps accounted for only 2.6% of the total fish catch and only 9 of the
24 fish species captured during this study (Table 4). Their combined
fish catch per set day was only 0.2 (Table 5). Pirate perch and redfin
pickerel were the fish species most frequently caught using these three
small collection devices. Unexpectedly, scent-oriented fish (e.g.
catfish) were not caught in baited minnow traps. The collection devices
did account for 18.8% of the total crayfish catch, averaging 1.0
crayfish caught per set day.
Sufficient data was collected from the two-way fixed weir traps and
fyke nets to differentiate floodplain and stream catch. However,
differences between floodplain and stream occurrence (i.e. percent of
total number) of each fish species were small (Table 6). Twenty-one of
Table 4. Fish and Crayfish Catch by Collection Device.
WR is two-way fixed weir traps, FN is fyke nets,
MT is minnow traps, SW is small, one-way movable
weir traps and BMT is baited minnow traps.
Species WR FN MT SW BMT
Pirate perch 1728 76 28 11 2
Redfin pickerel 565 97 12 26
FI ier 216 33
Mud sunfish 153 54 1
Eastern mudminnow 118 2 2 6
American eel 96 8 1
Banded sunfish 68 5 2
Creek chubsucker 48 1
Bowfin 18 28
Swampfish 29
Bluespotted sunfish 26 1
Pygmy sunfish 21 1
Shiner 20
Redear sunfish 19
Brown bullhead 10 3
Pumpkinseed 13
Golden shiner 13
Bluegill 6
Warmouth 5
Long-eared sunfish 2
Eastern swamp darter 2
Black crappie 1
Green sunfish 1
Mosquito fish 1
Total 3178 307 46 44 4
CTN* 86.7 8.4 1.3 1.2 0.1
Procambarus acutus a. 1028 199 160 71 115
Fallicambarus uhleri 337 55 5 1 2
Immature crayfTsTi 208 2 50 17 1
Total 1573 256 215 89 118
%JH** 69.9 11.4 9.6 4.0 5.2
* Total fish catch by all collection techniques (excluding
rotenone) = 3665
**Total crayfish catch by all collection techniques = 2251
24
Table 5. Number of Sets, Total Number of Set Days, Fish and Crayfish
Catch per Set Day (CPSD) by Collection Device. WR is two-
way fixed weir traps, FN is fyke nets, MT is minnow traps,
SW is small, one-way movable weir traps and BMT is baited
minnow traps.
Collection Total No. of CPSD
Device No. of Sets Set Days Fish Crayfish
WR 158 275.1 11.6 5.7
FN 23 52.0 5.9 4.9
MT 151 237.4 0.2 0.9
SW 91 102.6 0.4 0.9
BMT 54 86.1 < 0.1 1.4
25
Table 6. Comparison of Floodplain and Stream Fish Catches by Species Using
Fixed Weir Traps and Fyke Nets^ N is the total number caught, %TN
is the percent total number, TL is the average tojUl length, WT is
the weight, %TWT is the percent total weight and WT is the average
weight.
Floodplain Stream
Species N UN TL ( cm) WT (gm) TTWT WT (gm) N %TN TL (cm) WT (gm) TTWT WT (gm)
Pirate perch 1218 53.5 8.8 15679.3 27.8 12.9 586 48.5 9.1 8187.7 19.3 14.0
Redfin pickerel 424 18.6 15.0 18188.5 32.3 42.9 238 19.7 17.9 15384.6 36.2 64.6
FI i er 176 7.7 9.1 3722.5 6.6 21.2 73 6.0 9.5 1861.4 4.4 25.5
Mud sunfish 117 5.1 10.5 4131.5 7.3 35.3 90 7.4 10.5 3209.9 7.6 35.7
Eastern mudminnow 73 3.2 8.2 531.2 0.9 7.3 47 3.9 8.5 345.4 0.8 7.3
American eel 62 2.7 27.4 3026.8 5.4 48.8 42 3.5 26.8 1842.5 4.3 43.9
Banded sunfish 52 2.3 5.1 584.5 1.0 11.2 21 1.7 7.1 755.4 1.8 36.0
Creek chubsucker 26 1.1 17.1 2756.1 4.9 106.0 23 1.9 18.7 3385.8 8.0 147.2
Bowfin 14 0.6 23.5 2693.3 4.8 192.4 32 2.6 23.3 6145.1 14.5 192.0
Swampfish 21 0.9 5.8 29.0 0.1 1.4 8 0.7 5.9 12.1 <0.1 1.5
Bluespotted sunfish 17 0.7 8.9 1681.5 3.0 98.9 9 0.7 7.1 380.6 0.9 42.3
Pygmy sunfish 19 0.8 4.0 9.3 <0.1 0.5 2 0.2 3.9 0.9 <0.1 0.5
Shiner 11 0.5 6.0 39.4 0.1 3.6 9 0.7 6.2 36.4 0.1 4.0
Redear sunfish 15 0.7 10.0 423.4 0.8 28.2 4 0.3 8.9 97.5 0.2 24.4
Brown bullhead 11 0.5 14.2 2249.2 4.0 204.5 2 0.2 15.3 308.2 0.7 154.1
Pumpkinseed 6 0.3 9.8 139.1 0.2 23.2 7 0.6 8.4 123.6 0.3 17.7
Golden shiner 3 0.1 8.7 11.8 0.1 3.9 10 0.8 11.2 98.9 0.2 9.9
Bluegill 4 0.2 10.0 97.6 0.2 24.4 2 0.2 10.8 57.7 0.1 28.9
Warmouth 3 0.1 16.0 272.8 0.5 90.9 2 0.2 14.8 166.6 0.4 83.3
Long-eared sunfish 1 <0.1 11.5 32.6 0.1 32.6 1 0.1 15.0 72.4 0.2 72.4
Eastern swamp darter 1 <0.1 5.5 1.9 <0.1 1.9 1 0.1 5.0 1.4 <0.1 1.4
Black crappie 1 <0.1 17.0 60.0 0.1 60.0
Green sunfish 1 <0.1 12.0 28.5 0.1 28.5
Total 2276 56389.8 1209 42474.1
26
the 24 fish species sampled were caught both on the inundated floodplain
and in the stream. General ly, the fish caught on the fl oodpl ain were
slightly smaller than stream-caught fish (Table 6). Fish catch per set
day was higher on the floodplain than in the stream, while crayfish
catch exhibited little difference between the two areas (Table 7).
Diurnal and Nocturnal Habits:
Preliminary sampling and observation in the study area suggested
that fish in Creeping Swamp are primarily nocturnal. Further
investigation included three day-night comparison sets (data not
presented) using floodplain fixed weir traps during April, 1980. This
small number of comparison sets failed to verify exclusive nocturnal
habits.
Seasonal Occurrence and Movement:
The dominant fish species quickly moved into the upstream aquatic
system upon autumnal flooding and were present there throughout
inundation (Table 8). Crayfish were also abundant in the swamp
throughout the study. The number of fish species at the study site
steadily increased as the wet season progressed, until rapidly declining
water levels in late spring forced evacuation to deeper downstream
locations. Early in the wet season, average fish catch per set day was
greater in the stream than in floodplain areas reflecting initial
recruitment upstream and into the study area (Figure 4). Once fish
populations were established and floodplain inundation maintained, fish
movements on the floodplain began to exceed that in the stream and were
greatest during April and May (Figure 4).
27
Table 7. Comparison of Floodplain and Stream Catches of
Fish and Crayfish Using Fixed Weir Traps and Fyke
Nets. N is the total number caught, %TN is the
percent total number and CPSD is catch per set day.
Set Days N %TN CPSD
FLOODPLAIN 166.7
Fish 2276 71.7 13.7
Crayfish 900 28.3 5.4
Total 3176
STREAM 160.4
Fish 1209 56.5 7.5
Crayfish 929 43.5 5.8
Total 2138
28
Table 8. Monthly Catches of Fish and Crayfish in Creeping Swamp During
the 1979-80 Wet Season.
Species NOV DEC JAN FEB MAR APR MAY
Pirate perch 56 132 989 159 204 269 57
Redfin pickerel 88 42 107 35 40 271 159
FI i er 28 16 56 23 39 81 10
Mud sunfish 50 5 67 6 24 49 13
Eastern mudminnow 2 2 66 3 13 40 4
American eel 6 12 17 2 22 40 7
Banded sunfish 2 4 17 9 10 28 5
Creek chubsucker 1 6 3 14 21 5
Bowfin 24 7 2 4 3 6 2
Swampfish 1 9 2 9 7 1
Bluespotted sunfish 2 2 16 7
Pygmy sunfish 17 5
Shiner 2 5 2 3 6 3
Redear sunfish 10 8 1
Brown bullhead 2 1 1 1 8 1
Pumpkinseed 12 1
Golden shiner 1 1 2 2 7
Bluegill 2 4
Warmouth 4 1
Long-eared sunfish 2
Eastern swamp darter 1 1
Black crappie 1
Green sunfish 1
Mosquito fish 1
Total Fish 258 225 1359 258 404 877 284
No. of Species 9 12 13 16 17 22 17
Procambarus acutis a. 187 94 397 200 193 316 186
Fallicambarus uhleri 46 56 199 27 32 39 1
Immature crayfish 1 10 73 11 13 125 45
Total Crayfish 234 160 669 238 238 480 232
Figure 4. Average Combined Fixed Weir and Fyke Wet Fish Catch per Set Day in the Floodplai
and Stream by Month.
30
Environmental Effects on Fish Occurrence and Movement
Water Level :
Eight major spates occurred during this study, maintaining
continuous inundation of varying portions of the floodplain from
November through mid-May (Figure 5). During an average spate, water
levels rose 35 cm over four days, followed by a decline of 32 cm over 11
days (Table 9). The stream channel was considered full at a water level
of approximately 40 cm (rel ati ve to the stream bottom), whi 1 e minimum
water levels for effective use of floodplain weir traps were about 50-55
cm. In this study, average water temperatures were usually somewhat
higher during the falling water level periods than when the water was
rising (Table 9).
During the first half of the wet season, fish movements indicated
an overall recruitment upstream and into floodplain areas (Table 10).
Average fish catch during the last three months indicated increased rate
of movement and an increase in occurrence of new species. In April
overall movements were primarily downstream and out of floodplain areas
both on rising and falling water levels (Table 10).
Water Temperature;
Water temperatures in Creeping Swamp gradually decreased during the
first half of the wet season from 15oc in November to freezing in
February and early March, and then increased rapidly as air temperatures
rose and water level declined (Figure 6). Average fish catches
indicated that minimal movements occurred when water temperatures were
less than 60C. Moderate movements occurred from 60C-130C and movements
T , , , , J
no V
100
90
80
70
60
50
40
30
20
Stream Flow Ceases
10
O
ÏÏÔV ' DEC JAN FEB ‘ MAR 7VPÏÏ ' MAY
1979 1980
gur 5. Hydrograph of Water Level Fluctuations at the Creeping Swamp Study Site During
the 1979-80 Wet Season. Roman numerals indicate major spates.
32
Table 9. Characteristics of Major Spates for the 1979-80 Wet Season in Creeping
Swamp During Rising and Falling Water Levels. Days is days of spate
duration, WLR is water level range, WLC is water level change and WTM
is average water temperature.
Rising Water Level Falling Water Level
WLR WLC WTM WLR WLC wTm
Spate* Days (cm) (cm) (OC) Days (cm) (cm) (oC)
I 4 36.5-88.0 +51.5 13.0 12 88.0-58.0 -30.0 11.4
II 3 58.0-91.0 +33.0 15.8 7 91.0-68.0 -23.0 8.0
III 9 51.0-91.5 +40.5 5.9 10 91.5-75.0 -16.5 9.3
lY 4 70.0-87.0 +17.0 0.8 13 87.0-65.5 -21.5 8.4
V 4 65.0-108.0 +43.0 1.3 5 108.0-87.0 -21.0 10.4
VI 3 76.0-102.0 +26.0 14.0 13 102.0-56.5 -45.5 17.0
VII 3 56.5-86.0 +29.5 16.0 12 86.0-51.0 -35.0 15.2
VIII 2 51.0-93.0 +42.0 17.3 18 93.0-28.0 -65.0 19.0
Average 4.0 +35.3 11.3 -32.2
*Refer to Figure 5
33
Table 10. Average Fish Catches on Rising and Falling Water Levels During Major
Spates. -?FP is movements into the floodplain, FP-»is movements out of
the floodplain, ts is upstream movements and •iS is downstream movements.
Rising Water Level Fal1ing Water Level
Avg. Catch per Set Day Avg. Catch per Set Day
Harvest Harvest
Spate* Days -=>FP FP-^ fS iS Days ~>FP FP^ ts
I 1 „ 3.5 7 7.4 8.5 13.7
II 2 13.0 -- 6.0 28.0 2 1.3 — 5.0
III 4 11.5 4.9 6.4 6.0 2 22.3 5.7 5.5 5.1
IV 1 0.5 0.0 0.0 0.5 5 24.3 21.8 30.6 12.8
V 1 0.4 0.7 0.3 0.1 1 29.0 8.5 — --
VI 0 — — — -- 4 9.6 12.1 6.1 4.2
VII 0 — — — 4 9.7 23.3 8.0 6.0
VIII 1 41.0 68.0 9.5 29.0 2 12.5 26.5 7.5 12.5
*Refer to Figure 5
Figure 6. Thermograph of Water Temperature Fluctuations at the Creeping Swamp Study Site
During the 1979-80 Wet Season.
35
sharply increased after water temperatures rose above 140C (Figure 7).
Dissolved Oxygen:
Dissolved oxygen levels in Creeping Swamp were generally greater
than 8 mg/1 at the weir col 1 ection sites. At times other than 1 ate in
the wet season, there were only small differences in oxygen level
between floodplain and stream areas (Table 11). Low D.O. levels (1.6-
3.3 mg/1) in April were found only in floodplain edge areas where there
was little or no flow during periods of low water levels and high water
temperatures (Table 11). Juvenile redfin pickerel (< 5 cm in length)
were the only fish commonly observed in these shallow fringe areas.
Dissolved oxygen levels in Creeping Swamp normally were high (> 8
mg/1) until water temperatures exceeded 14oc. Although D.O. steadily
declined with higher water temperatures, fish movements appeared to be
more inhibited by low water temperatures than by decreasing D.O. levels
(Table 12).
FOOD HABITS STUDY
Invertebrate Standing Crop
Floodplain Taxa:
Twenty invertebrate taxa were collected on the inundated floodplain
with drift nets, lift nets and core samplers (Table 13). The dominant
floodplain invertebrates (76.5% of the total number and 76.3% of the
total weight) were copepods, amphipods, isopods, chironomids,
oligochaetes and crayfish (Astacidae). Copepods and amphipods were the
most abundant invertebrates sampled, representing 47.0% and 18.2%,
Figure 7. Average Number of Fish Caught in Fixed Heir Traps and Fyke Nets per Collection
Device per Set Day versus Water Temperature (N=150 sets).
CO
fTi
37
Table 11. Floodplain-Stream Comparison of Average Dissolved Oxygen
Levels from February 1980 through April 1980. D.O. is
average dissolved oxygen, range is dissolved oxygen range
and N is the total number of D.O. measurements.
Floodplain Stream
Month D.O. (mg/1) Range N D.O. (mg/1) Range N
FEB 10.2 4.0-12.9 46 10.3 6.5-12.9 10
MAR 10.0 5.7-12.5 85 9.6 7.9-10.2 6
APR 5.4 1.6- 7.6 68 7.0 6.2- 7.5 4
38
Table 12. Water Temperature-Dissolved Oxygen Relationship
to Fish Catch. WTM is water temperature, D.O. is
average dissolved oxygen, range is dissolved
oxygen range and CPSD is average fixed weir fish
catch per set day.
WTM (oC) Cmr. (mg/l) Range CPSD
0-6 11.8 10.0-12.8 3.6
7-13 9.9 7.9-11.6 22.7
14-18 6.7 3.3- 8.0 51.9
39
Table 13. Invertebrates Collected on the Inundated Creeping Swamp Floodplain Using
Drift Nets, Lift Nets and Core Samplers. N is the total number of
invertebrates collected, %JH is the percent total number, TL is the
average tot^ length, WT is the dry weight, %TWT is the percent total
weight and WT is the average dry weight.
Taxon N %TN TT (mm) WT (mg) %TWT FT (mg)
Turbellaria 49 1.2 3.5 15.1 0.9 0.3
Nematoda 21 0.5 0.1 <0.1 <0.1
01igochaeta 188 4.7 7.9 52.4 2.9 0.3
Arachnida
Acariña 46 1.1 1.0 2.8 0.2 0.1
Crustacea
Cladocera 29 0.7 1.6 0.6 <0.1 <0.1
Copepoda 1887 47.0 0.9 25.2 1.4 <0.1
Isopoda 150 3.7 5.2 169.0 9.5 1.1
Amphipoda 731 18.2 5.1 648.8 36.4 0.9
Decapoda
Astacidae 12 0.3 12.1* 595.2 33.4 49.6
Insecta
Odonata 3 0.1 8.8 64.9 3.6 21.6
Ephemeroptera 137 3.4 3.2 49.3 2.8 0.4
Hemiptera 10 0.2 3.0 5.4 0.3 0.5
Trichoptera 5 0.1 7.7 11.3 0.6 2.3
Diptera
Ceratopogonidae 25 0.6 4.8 1.8 0.1 0.1
Chironomidae 254 6.3 4.0 39.3 2.2 0.2
Culicidae 41 1.0 4.1 6.3 0.4 0.2
Simulidae 28 0.7 4.2 11.1 0.6 0.4
Coleóptera 62 1.5 3.0 36.8 2.1 0.6
Gastropoda 37 0.9 3.0 41.4 2.3 1.1
Pelecypoda 17 0.4 1.3 5.5 0.3 0.3
Encysted invertebrate 1 <0.1 1.0 0.4 <0.1 0.4
Terrestrial invertebrate 3 0.1 2.2 1.5 0.1 0.5
Invertebrate exuviae 203 5.1
Unidentifiable material 72 1.8
Total 4011 1784.2
*Carapace length
40
respectively, of the total floodplain catch. Crayfish, however, were not
adequately sampled by the drift nets or core samples. Crayfish probably
make up a significantly greater percentage of the floodplain
invertebrates than results show, as indicated by the large number caught
by fish collection techniques. Crayfish ranked second behind amphipods
in percent total weight (33.4%) of the floodplain invertebrates sampled,
although they were only 0.3% of the total number.
Stream Taxa:
Sixteen invertebrate taxa were collected in the stream with drift
nets and core samplers (Table 14). The dominant stream invertebrates
(81.9% of the total number and 67.5% of the total weight) were copepods,
ol igochaetes, chironomids and amphipods. Copepods and oligochaetes were
the most abundant invertebrates sampled, comprising 44.8% and 15.1%,
respectively, of the total stream catch. Again, crayfish were not
adequately sampled with the invertebrate collection devices used and, in
fact, probably make up a significant percentage of the stream
invertebrates.
Comparison of Collection Techniques:
CORES. A total of thirty core samples yielded only 9 invertebrate
taxa, although large numbers of each taxa were collected by this method
(Table 15). The dominant invertebrates in core samples were
ol igochaetes, chironomids and amphipods. 01 igochaetes were the most
abundant invertebrate sampled, averaging 5188.5 individuals per square
41
Table 14. Invertebrates Collected in the Creeping Swamp Stream Using Drift Nets and
Core Samplers. N is the total number of invertebrates collected, %TN is
the percent total number, TL is the average tftal length, WT is the dry
weight, %TWT is the percent total weight and WT is the average dry weight.
Taxon N %JH TL (mm) WT (mg) %TWT WT (mg)
Turbellaria 2 0.3 2.9 0.5 0.3 0.3
Nematoda 1 0.1 2.0 <0.1 <0.1 <0.1
Oligochaeta 105 15.1 14.7 86.2 45.8 0.8
Crustacea
Ostracoda 3 0.4 0.5 <0.1 <0.1 <0.1
Copepoda 312 44.8 0.9 4.5 2.4 <0.1
Isopoda 2 0.3 6.3 2.6 1.4 1.3
Amphipoda 69 9.9 2.1 23.3 12.4 0.3
Insecta
Ephemeroptera 34 4.9 2.9 18.0 9.5 0.5
Trichoptera 5 0.7 10.6 18.9 10.1 3.8
Diptera 1 0.1 5.2 0.3 0.1 0.3
Ceratopogonidae 3 0.4 8.8 1.0 0.5 0.3
Chironomidae 84 12.1 4.4 12.9 6.9 0.2
Culicidae 2 0.3 2.8 0.1 0.1 0.1
Simulidae 9 1.3 3.9 2.2 1.2 0.2
Coleóptera 15 2.2 1.9 1.6 0.8 0.1
Pelecypoda 4 0.6 3.9 15.5 8.2 3.9
Terrestrial invertebrate 1 0.1 2.5 0.7 0.3 0.7
Invertebrate exuviae 44 6.3
Total 696 188.3
42
Table 15. Average Number of Invertebrates Sampled by Collection Device. Sampling
was concurrent with fish collections for gut analyses. Blanks represent
invertebrate taxa not collected.
Drift Nets Lift Nets Core Samples
Taxon Avg. No./lOO m3 Avg. No./m2 Avg. No./m2
Turbellaria 1.7 4.7 104.8
01 igochaeta 0.6 11.3 5188.5
Arachnida
Acariña 2.4 4.3
Crustacea
Cladocera 1.9 0.3
Ostracoda 1.0
Copepoda 30.5 6.7 366.9
Isopoda 0.1 48.7 209.6
Amphipoda 12.0 207.3 642.0
Decapoda
Astacidae 0.6 2.7
Insecta
Odonata 1.0 0.3
Ephemeroptera 6.7 0.7
Hemiptera 0.5
Trichoptera 1.5 0.7
Diptera
Ceratopogonidae 2.0 3.7 209.6
Chironomidae 11.0 26.3 668.3
Culicidae 0.5 1.7
Simulidae 5.5 0.3
Coleóptera 3.1 8.0 104.8
Gastropoda 0.1 12.0
Pelecypoda 0.1 4.3 174.7
Encysted invertebrate 52.4
Terrestrial invertebrate 1.0 0.7
43
meter of substrate (0.084 m3), while chironomids and amphipods averaged
668.3 and 642.0 individuals per square meter, respectively (Table 15).
LIFT NETS. Three lift net samples taken in one floodplain area
collected 18 invertebrate taxa (Table 15). The dominant invertebrates
captured with lift nets were forms normally inhabiting the leaf litter
layer including amphipods, isopods and chironomids (Table 15).
Amphipods were the most abundant invertebrates captured, averaging 207.3
individuals per square meter of inundated floodplain. The invertebrates
collected with lift nets were considerably larger than those taken in
drift nets and core samples.
DRIFT. Twenty invertebrate taxa were collected by 11 drift net
samples during the time and in the same areas that fish were collected
for gut analysis. The dominant invertebrates in the drift were
copepods, amphipods and chironomids. Copepods were the most abundant
invertebrate in the drift, averaging 80.5 individuals per 100 cubic
meters of water. One day-night drift net comparison revealed few
differences between invertebrates collected in drift nets during
daylight and those collected at night (Table 16). Mayfly larvae
(Ephemeroptera), however, were more prevalent in the night sample than
in the day sample while chironomids, Turbellaria (flatworms). Acariña
(water mites), Cladocera and Coleóptera occurred more frequently during
the day (Table 16).
44
Table 16. Day-Night Comparison of Drift Net Collections. Results are from two 10-hour
sets in April 1980. N is the total number of invertebrates collected, %TN
is the percent total number, WT is the dry weight and %TWT is the percent
total weight.
Day Night
Taxon N %TN WT (mg) %TWT N 7oTN WT (mg) %TWT
Turbellaria 11 3.0 1.8 14.4 1 0.3 0.4 1.3
01igochaeta 1 0.3 0.1 0.8 1 0.3 0.1 0.3
Arachnida
Acariña 13 3.6 0.7 5.6 2 0.6 0.2 0.6
Crustacea
Cladocera 7 1.9 0.2 1.6
Copepoda 263 72.5 2.8 22.4 249 73.9 2.3 7.2
Isopoda 1 0.3 0.2 1.6
Amphipoda 12 3.3 0.4 3.2 11 3.3 0.6 1.9
Insecta
Ephemeroptera 13 3.6 4.6 36.8 60 17.8 27.7 87.1
Hemiptera 1 0.3 0.2 1.6
Diptera
Ceratopogonidae 1 0.3 0.1 0.8
Chironomidae 15 4.1 0.8 6.4 8 2.4 0.5 1.6
Coleóptera 3 0.8 0.4 3.2
Gastropoda 1 0.3 0.1 0.8
Pelecypoda 1 0.3 0.1 0.8
Invertebrate exuviae 20 5.5 4 1.2
Unid, material 1 0.3
Total 363 12.5 337 31.8
45
Fish Gut Analysis
Collection Techniques:
Five fish collection trips were made from February through April,
1980 to obtain fish for gut analysis. One hundred and forty fish (14
species) were col 1 ected on the fl oodpl ain and sixty fish (10 species)
were taken from the stream (Table 17).
Over half (109) of the floodplain caught fish and all 60 of the
stream-caught fish were captured by two-way fixed weir traps utilizing
short set periods (1-4 hours). Since fixed weir traps were located near
the confluence of primary floodplain drainage sloughs with the stream,
fish caught moving upcurrent into floodplain fixed weir traps were
considered, for gut analyses, to be stream-caught. Exact feeding
locations, however, could not be assured for these fish.
El ectroshocking accounted for 59 of the fish collected for gut
analysis. All were captured on the floodplain even though
electroshocking was also performed in the stream channel.
Electroshocking was effective when water temperatures were above 6oC and
the floodplain was only 10-40% inundated. Small, one-way movable weir
traps and minnow traps accounted for only 5 of the 200 total fish
collected.
Stomach Fullness:
Stomach fullness of all floodplain-caught fish averaged 2.6 on the
0-4 scale, representing, by interpolation, about 55% fullness, whereas,
stream fish fullness averaged only 27% (Table 18). Twenty of the 200
46
Table 17. Floodplain-Stream Comparison of Fish Collected for Gut Analysis.
Results of 5 collections from February through April, 1980. N is
the total number of fish collected, TL is the average total length
and range is the length range.
Floodplai n Stream
Species N TL (cm) Range N TT (cm) Range
Pirate perch 36 7.9 5.2-11.9 32 8.7 5.8-11.7
Redfin pickerel 58 8.5 1.6-18.5
FI ier 12 7.4 6.0- 8.8 12 9.2 7.6-16.1
Mud sunfish 6 10.1 9.0-11.5 4 11.4 10.0-13.8
Eastern mudminnow 11 2.5 1.7- 7.8 1 6.1
American eel 3 25.0 23.5-27.2 3 27.6 25.1-29.0
Banded sunfish 5 4.4 3.2- 5.6
Creek chubsucker 1 19.5
Bowfin 2 22.5 21.0-24.0 1 27.0
Pygmy sunfish 2 3.9 3.8- 4.0
Shiner 1 8.8 2 7.0
Brown bullhead 1 10.2 3 8.4 5.7-10.5
Golden shiner 1 9.2
B1uegill 1 11.0
Eastern swamp darter 1 5.0
Green sunfish 1 12.0
Total 140 60
47
Table 18. Floodplain-Stream Comparison of Average Estimated Stomach
Fullness of Fish Collected for Gut Analysis. Stomach
fullness was estimated using a zero to four scale where
0 = empty, 1 = \%-lS% full, 2 = 25%-50% full, 3 = 50%-75%
full and 4 = 75X-100% full.
Floodplain Stream
Species Avg. Fullness (N) Avg. Fullness (N)
Pirate perch 2.5 (36) 1.3 (32)
Redfin pickerel 2.7 (58)
FI ier 2.8 (12) 2.0 (12)
Mud sunfish 2.7 ( 6) 2.8 ( 4)
Eastern mudminnow 2.5 (11) 1.0 ( 1)
American eel 2.0 ( 3) 2.3 ( 3)
Banded sunfish 2.6 ( 5)
Creek chubsucker 0.0 ( 1)
Bowfin 2.0 ( 2) 2.0 ( 1)
Pygmy sunfish 2.0 ( 2)
Shiner 1.0 ( 1) 0.5 ( 2)
Brown bullhead 3.0 ( 1) 2.7 ( 3)
Golden shiner 0.0 ( 1)
Bluegill 3.0 ( 1)
Eastern swamp darter 1.0 ( 1)
Green sunfish 4.0 ( 1)
Total 2.6 (140) 1.6 (60)
48
total fish collected for gut analyses had empty stomachs (7 floodplain-
caught and 13 stream-caught).
Stomach Contents:
Nineteen invertebrate taxa and one fish (eastern mudminnow) were
found in the stomachs of 140 floodplain-caught fish (Table 19). The
most frequently occurring food items in these fish were isopods,
copepods, amphipods and chironomids (69.7% of the total number).
Crayfish (Astacidae) ranked highest in percent total weight of food
items (61.4%), although representing only 0.9% of the total number
(Table 19). The one fish and most of the crayfish consumed were found in
the stomachs of the larger piscivorous fish species (e.g. redfin
pickerel, American eel, bowfin and brown bullhead).
Fourteen invertebrate taxa were found in the stomachs of 60 stream-
caught fish (Table 20). The most frequently occurring food items in
these fish were simulids (black fly larvae), chironomids, copepods,
isopods and amphipods (59.4% of the total number). Again, crayfish
ranked highest in percent total weight of food items (72.5%), although
representing only 1.2% of the total number (Table 20).
SPAWNING/NURSERY AREA OBSERVATIONS
Fish Spawning Condition
Five hundred and fifty-seven fish caught on the floodplain and two
hundred and twenty-one fish caught in the stream from December 29 to
April 29 were found to be in spawning condition (Table 21). Overal 1 ,
there were 493 males and 285 females in spawning condition, representing
49
Table 19. Stomach Contents of 140 Floodplain-Caught Fish Collected for Gut Analysis.
N is the_total number of items in fish stomachs, %TN is the percent total
number, TL is the average_total length, WT is the dry weight, %TWT is the
percent total weight and WT is the average dry weight.
Taxon H %TN TL (mm) WT (mg) %TWT WT (mg)
Turbellaria 1 <0.1 6.3 0.7 <0.1 0.7
01igochaeta 6 0.2 6.2 3.1 0.1 0.5
Arachnida
Acariña 8 0.2 1.2 0.9 <0.1 0.1
Crustacea
Ostracoda 113 4.0 0.4 0.8 <0.1 <0.1
Copepoda 581 18.0 0.9 5.3 0.1 <0.1
Isopoda 796 24.7 6.5 1424.2 27.0 1.8
Amphi poda 515 16.0 5.0 459.0 8.7 0.9
Decapoda
Astacidae 29 0.9 15.2* 3242.2 61.4 111.8
Insecta
Plecoptera 1 <0.1 4.0 0.6 <0.1 0.6
Odonata 1 <0.1 19.5 36.2 0.7 36.2
Ephemeroptera 57 1.8 5.9 34.6 0.7 0.6
Trichoptera 18 0.6 5.9 12.9 0.2 0.7
Diptera 3 0.1 8.5 2.2 <0.1 0.7
Ceratopogonidae 20 0.6 5.3 0.9 <0.1 <0.1
Chironomidae 354 11.0 4.2 33.6 0.6 0.1
Culicidae 1 <0.1 1.5 0.1 <0.1 0.1
Simulidae 30 0.9 4.7 10.3 0.2 0.3
Coleóptera 15 0.5 3.6 8.9 0.2 0.6
Pelecypoda 37 1.1 0.9 3.8 0.1 0.1
Encysted invertebrate 1 <0.1 1.0 0.4 <0.1 0.4
Invertebrate exuviae 1 <0.1
Parasitic worm 38 1.2
Eastern mudminnow 1 <0.1 16.0
Vegetative material 9 0.3
Unidentifiable material 586 18.2
Total 3223 5280.5
*Carapace length
**Length/weight equation for eastern mudminnow was not appropriate for length <20.0 mm
50
Table 20. Stomach Contents of 60 Stream-Caught Fish Collected for Gut Analysis.
N is the total_number of items in fish stomachs, %TN is the percent
total number, TL is the average total_Length, WT is the dry weight,
%TWT is the percent total weight and WT is the average dry weight.
Taxon N %TH TL (mm) WT (mg) %TWT WT (mg)
Turbellaria 1 0.1 4.0 0.3 <0.1 0.3
01igochaeta 3 0.3 7.1 1.1 0.1 0.4
Crustacea
Ostracoda 2 0.2 0.7 0.1 <0.1 <0.1
Copepoda 109 9.5 0.9 1.0 0.1 <0.1
Isopoda 64 5.6 7.1 141.4 7.8 2.2
Amphipoda 64 5.6 5.9 73.0 4.0 1.1
Decapoda
Astacidae 14 1.2 14.5* 1323.0 72.5 94.5
Insecta
Odonata 2 0.2 15.5 54.7 3.0 27.4
Ephemeroptera 18 1.6 6.4 22.4 1.2 1.2
Trichoptera 5 0.4 9.7 34.8 1.9 7.0
Diptera
Ceratopogonidae 17 1.5 4.0 0.5 <0.1 <0.1
Chironomidae 129 11.2 4.2 10.9 0.6 0.1
Simulidae 317 21.S 5.2 160.2 8.8 0.5
Pelecypoda 4 0.3 1.1 0.5 <0.1 0.1
Parasitic worm 44 3.8
Vegetative material 5 0.4
Unidentifiable material 354 30.7
Total 1152 1823.9
*Carapace length
51
Table 21. Number of Male (M) and Female (F) Fish for Each Species in Spawning Condition
by Month.
DEC JAN FEB MAR APR
Species M F M F M F M F M F Total
Pirate perch 269 71 69 50 96 81 19 29 684
Redfin pickerel 1 1 5 3 10
FI ier 6 4 1 7 6 24
Mud sunfish 2 15 4 21
Eastern mudminnow 1 1
Banded sunfish 1 1
Creek chubsucker 2 1 4 1 3
Bowfin 1 1
Swampfish 1 5 1 9 3 19
Bluespotted sunfish 1 1
Pygmy sunfish 3 3
Brown bullhead 2 2
Golden shiner 1 1
Warmouth 1 1
Green sunfish 1 1
Total 0 1 269 77 76 68 96 92 52 47 778
% Total Fish Catch* 0.4 25.5 55.8 46.5 11.3
Floodplain Total 1 276 71 140 69 557
Stream Total 0 70 73 48 30 221
No. of Species 1 3 8 4 13 15
*Refer to Table 8 for monthly totals
52
a total of 15 species. Percentages of fish in spawning condition were
greatest in February and March (55.8% and 46.5%, respectively), and
decreased to only 11.3% in April (Table 21).
Young~of-the-Year
Fish eggs and larvae were not collected by any of the collection
techniques used in this study. However, juvenile fish were frequently
collected and observed, particularly late in the wet season. Twenty-two
fish under 5 cm in length (representing less than one-year class for
most species) were trapped on the floodplain, while several clumped
distributions of juveniles (25-100 individuals, 1-2 cm in length) were
observed on the floodplain and at the stream edge. Juveniles also
accounted for over half of the fish collected in the Rotenone Study
(discussed below).
ROTENONE STUDY
Six hundred and ninety-eight fish (14 species) were collected using
rotenone in a 75 m section of the Creeping Swamp stream on May 13, 1980
(Table 22). Pirate perch, redfin pickerel and eastern mudminnow were
the most numerous species sampled (81.7% of the total catch), while
bowfin were the largest fish captured (25.9 cm average length, 224.8 gm
average weight). Overall, average fish size (Table 22) was considerably
less when compared to the same species caught in the fixed weir traps
(Table 6) except for flier, mud sunfish and bowfin. Juveniles (< 5 cm
in length) accounted for 361 fish collected, representing over half of
the total catch. Juvenile species collected most frequently were pirate
perch, redfin pickerel, eastern mudminnow and swampfish (Table 22).
53
Table 22. Fish Catch from Rotenone Sampling of a Section of the
Creeping Swamp Stream on May 13, 1980. N is the total number
of fish caught, %TN is the percent total number, TL is the
average total length, WT is the weight, %TWT is the percent
total weight and WT is the average weight.
Species N %jn TL (cm) WT (gm) %TWT WT (gm)
Pirate perch 225 32.2 2.6 421.2 4.9 1.9
Redfin pickerel 209 29.9 12.6 4730.6 54.7 22.6
Flier 37 5.3 8.7 615.5 7.1 16.6
Mud sunfish 8 1.1 11.4 319.7 3.7 40.0
Eastern mudminnow 137 19.6 2.7 34.0 0.4 0.2
American eel 3 0.4 20.6 83.4 1.0 27.8
Banded sunfish 1 0.1 5.0 2.5 <0.1 2.5
Creek chubsucker 11 1.2 13.1 394.3 4.6 35.8
Bowfin 8 1.1 25.9 1798.1 20.8 224.8
Swampfish 35 5.0 1.7 1.9 <0.1 0.1
Shiner 8 1.1 5.8 21.3 0.2 2.7
Brown bullhead 8 1.1 10.9 189.3 2.2 23.7
Golden shiner 7 1.0 7.2 29.5 0.3 4.2
B1uegil1 1 0.1 4.4 2.7 <0.1 2.7
Total 698 8644.0
54
Fish length versus wet weight relationships were developed for
those species of which at least 8 individuals were collected during the
rotenone sampling (Table 23A). Ten other non-linear regression
equations were found in the literature (Table 23B).
55
Table 23, Fish Length versus Weight Relationships. (A.) derived from non-linear regression
equation WT = BqLBi where WT = wet weight in grams, L = total length in cm,
Bq and Bi are constants and r = asymptotic correlation of Bq and Bi- (B.) for
Log WT = Log Bq + B^ Log L.
A. From This Study;
Length Range
Species (cm) r Bo one S.E.''”'' ^1 - rone S.E. N
Pirate perch 6.3-10.2 -.999 0.0258 0.0078 2.8203 0.1389 41
Redfin pickerel 4.2-20.5 -.998 0.0100 0.0012 3.0000 0.0426 209
FI ier 4.5-10.5 -.999 0.0293 0.0148 2.9123 0.2273 37
Mud sunfish 10.0-12.5 -.999 0.0451 0.0516 2.8000 0.4628 8
Eastern mudminnow 2.1- 3.5 -.995 0.0071 0.0008 3.2697 0.1015 134
American eel* 18.5-45.3 -.999 0.0041 0.0033 2.8028 0.2180 28
Creek chubsucker 10.5-16.0 -.999 0.0055 0.0024 3.4000 0.1637 11
Bowfin 23.5-28.5 -.999 0.0510 0.0762 2.6000 0.4554 8
Swampfish 1.2- 2.1 -.987 0.0133 0.0023 2.6159 0.2988 35
Pygmy sunfish** -.987 0.0133 0.0023 2.6159 0.2988
Shiner 5.2- 7.0 -.998 0.0210 0.0183 2,8000 0.4780 8
Brown bullhead 8.8-14.0 -.999 0.0199 0.0037 3.1821 0.1332 8
Eastern swamp darter* -.995 0.0071 0.0008 3.2697 0.1015
Mosquito fisn* -.995 0.0071 0.0008 3.2697 0.1015
B. From Literature:
Species Source Log Bq Bl
Banded sunfish Carlander (1977) -5.5000 3.7300
Bluespotted II IIsunfish -5.5000 3.7300
Redear sunfish II II -4.6220 2.9600
Pumpkinseed II II -5.2130 3.2620
Golden shiner II (1969) -5.9750 3.3440
Bluegil1 II (1977) -4.8870 3.0700
Warmouth II It -4.8410 3.0800
Long-eared sunfish 11 II -4.6900 3.0100
Black II IIcrappie -4.8460 2.9700
Green sunfish II II -5.2700 3.2345
* Length and weight data obtained from Dr. Margie Gallagher of the Institute for
Coastal and Marine Resources at East Carolina University
**Same as swampfish; not present in literature and not enough data to develop
equation
Same as eastern mudminnow; not present in literature and not enough data to
develop equation
++One Standard Error
DISCUSSION
This study established that fish which migrate upstream in Creeping
Swamp during the wet season do utilize the adjacent inundated
floodplain. Fish movements between the stream and floodplain, and
within each area, were influenced by several environmental factors.
Floodplain inundation greatly increased the aquatic habitat available to
swamp-stream fish, providing them with more cover, a substantial
invertebrate food supply and a relatively predator-free spawning and
nursery area.
Species Composition of Fish Occupying the Floodplain
Pirate perch and redfin pickerel were the dominant fish species in
the inundated floodplain (Table 6). Other frequently occurring species
i ncluded f 1 ier, mud sunfish, eastern mudminnow, American eel, banded
sunfish, creek chubsucker, bowfin and swampfish. Huish and Pardue
(1978) identified these same species as numerically important fish in
their study of three wooded coastal swamp streams in northeastern North
Carolina. Pirate perch and redfin pickerel accounted for 63.4% of the
total number of fish they collected from Duke Swamp Creek. This
compares very well to the 71.1% total number captured in Creeping Swamp
for the same two species. Pirate perch and redfin pickerel are usually
considered minor species in most habitats, but their abundance in small
coastal plain swamps of eastern North Carolina indicates that these
shallow, seasonal systems are prime habitat for the pirate perch and
redfin pickerel.
57
The fish connnunity in Creeping Swamp is composed predominantly of
small-sized species. All year classes of these smaller species (e.g.
pirate perch, eastern mudminnow, banded sunfish, swampfish, blue-spotted
sunfish and pygmy sunfish) were found. Most younger year classes of
redfin pickerel, flier, mud sunfish, creek chubsuckers, shiners and
bowfin were commonly collected but larger specimens were seldom trapped.
Whether this was due to trap avoidance because of small entrance
aperture size or absence of the 1 arger fish in the shall ow waters of
Creeping Swamp is not known. Electroshocking and rotenone sampling did
not produce any older classes of these fish. Also, only juvenile eels
(less than 30 cm TL) were found in this study.
Differences in the relative occurrence of each fish species between
the floodplain and stream were small (Table 6), indicating fish moved
freely in the inundated system, apparently not differentiating between
the two areas. Fish caught on the floodplain were slightly smaller than
fish collected in the stream channel. This difference most likely
occurred because shallow floodplain water depths excluded larger-sized
fish, restricting them to the deeper stream channel. The number of fish
caught per set day on the floodplain was nearly twice the catch in the
stream (Table 7), indicating that most fish were moving to-and-from the
inundated floodplain and exhibiting relatively fewer movements up-and-
down the stream channel. 0 veral 1, the f i sh communi ty appeared to use
the floodplain more than the stream.
Diurnal and Nocturnal Habits:
During a diel study in April, weir-caught fish did not exhibit
exclusive nocturnal habits as shown in preliminary collections.
58
However, many of the dominant species in Creeping Swamp have been
reported as nocturnal (Holder 1970; Strommen and Rasch 1975). Indeed,
Holder found that 90% of the catch in his study of fish movements from
the Okefenokee Swamp occurred during the night.
Diurnal patterns of fish activity in Creeping Swamp were not
ascertained in this study. Diel periods of fish activity probably are
affected by a variety of factors such as diel activity of invertebrates,
diel fish predation, seasonal spawning migration urges and water level
fluctuations. Determining diel activity patterns was beyond the scope
of this study. The use of 24-hour set periods for weir traps
effectively "black boxed" the problem.
Seasonal Occurrence and Movement:
During the wet season, normally occurring from mid-December through
April a greatly expanded, relatively predator-free aquatic environment
with an abundant invertebrate food supply is available for fish
utilization. The swamp fish community is well-adapted to the seasonal
nature of smal 1 stream swamps in the coastal pi ain of North Carol ina.
Comparison of Fish Collection Techniques
Two-way fixed weir traps were found to be the most effective fish
collection device for both the inundated floodplain and the stream
(Tables 4 and 5). Although similar to those used by Huish and Pardue
(1978), the weir traps used in this study had smaller mesh size (0.64 cm
versus 1.28 cm for Huish and Pardue) and more shallow construction. This
allowed for the capture of smaller sizes of fish at lower water levels.
Even with the smaller mesh size, juvenile fish (e.g. pirate perch.
59
eastern mudminnow and swampfish) were not effectively sampled due to
within-trap predation and their ability to swim through the 0.64 cm
mesh. Fyke nets were also effective collection devices once installed.
However, the abundance of underwater snags and obstructions made
installation and maintenance difficult.
Twenty-four hours was the most efficient sampl ing period for the
fixed weir systems used in this study. Shorter set periods might
provide biased data due to differential diel movements of fish. Short
set periods al so may not al low time for fish to readjust after being
disturbed during the resetting of traps. For set periods greater than 24
hours, the accumulation of fish inside the traps leads to an increase in
the rate of within-trap predation. Activity within the trap may
discourage smaller fish from entering. Eventually the rate of trap
avoidance and within-trap predation would equal the rate of entrapment,
and apparent catch rates would decrease. Longer set periods also
allowed the accumulation of floating debris and clogging of entrance
apertures.
Minnow traps, baited minnow traps and small, one-way movable weir
traps were not very effective collection devices in Creeping Swamp
(Tables 4 and 5). Small entrance apertures (2-3 cm) excluded most adult
fish. Because extensive weir fencing was not used with these small traps
to concentrate fish into a smaller area, fewer fish were captured.
El ectroshocking was an effective technique for collecting fish on
the inundated floodplain, if performed at night, when water temperatures
were above 6oc and when the floodplain was only 10-40% inundated.
El ectroshocking efficiency was limited by two factors: (1) fish density
and (2) the ability of the user to see the shocked fish. When the
60
floodplain was inundated more than 40%, fish dispersal resulted in lower
densities and obviously lower shocking catch. Shocked fish were easier
to see at night with a headlamp or spotlight than during the daylight.
At temperatures below 6oc, shocked fish slowly rolled and settled to the
bottom where they were nearly impossible to see.
Environmental Effects on Fish Occurrence and Movement
Changes in environmental conditions during the wet season
significantly affected fish occurrence and movement in Creeping Swamp.
Water level, water temperature and dissolved oxygen were the most
important factors affecting fish activity. General water quality, as
shown in Table 1, during the wet season in Creepi ng Swamp was high and
should not have affected the fish community.
Water Level :
Water level is the most important factor affecting fish occurrence
and movement in fluctuating aquatic environments. Fish communities that
utilize seasonal aquatic systems such as Creeping Swamp have to be very
mobile. Small changes in water level can greatly affect the amount of
floodplain inundation. At the study site in Creeping Swamp, an increase
in water level from 50 cm to 75 cm more than doubled the aquatic area of
the swamp (Sniffen 1981). A further increase from 75 to 100 cm again
more than doubled the aquatic area. Thus during the wet season of 1979-
80, spates caused the aquatic area of Creeping Swamp to increase 2 to 5
times the amount available at a water level of 50 cm. Therefore, fish
occupying the floodplain must be very sensitive to water level
fluctuations.
61
During the first half of the wet season, overa! 1 movements were
upstream and into floodplain areas (Table 10). During the last three
months of the study, average water levels progressively declined. During
the last spate of the wet season, fish moved downstream and out of
f 1 oodplain areas both on rising and fal 1ing water levels (Table 10).
These movements indicate that many of the fish occupying floodplain
areas and upstream reaches of Creeping Swamp late in the wet season
leave these areas just prior to the time of the normally-occurring dry
season and move to more permanent downstream locations even though water
levels may remain high. Similar behavior was observed by Wei comme
(1969) in a study of the fishes of a small tropical swamp stream. He
observed young fish migrating downstream during the time of a normal dry
season, even though the stream remained unusually high and the
surrounding swamps full of water.
Water Temperature;
Water temperature was a significant factor affecting fish movements
in Creeping Swamp. Fluctuations in water temperature which occurred in
this study (Figure 6) typically cause changes in the rate of metabolism
of most temperate fish species (Lagl er et 1977). These changes of
metabolic rate are particularly evident as water temperatures approach
freezing. Water temperatures less than 6oc severely reduced fish
movements (Figure 7), while fish activity increased as temperatures rose
from 6oc to 20oc. Although water temperature fluctuations appeared to
have a direct influence on fish movements, other coincidental factors
(e.g. spring spawning migrations, declining water levels) may have been
equally important. Holder (1970) suggested that high water temperature
62
is frequently inversely related to fish movement. This probably is a
result from sluggish behavior in response to high temperatures (>20° C)
and resul tant 1 owered D.O. (<4 mg/1 ). Low water temperature i s a more
limiting factor affecting fish movements in Creeping Swamp since
temperatures above 20oc and adverse D.O. conditions were not commonly
found during the wet season.
Dissolved Oxygen:
Dissolved oxygen was not usually a limiting factor upon fish
movements through the fixed weir traps in Creeping Swamp during the wet
season. As mentioned above, adverse D.O. conditions rarely occurred at
the fixed weir sites and when they did occur usually did not persist for
long periods of time. Catch versus D.O. comparisons (Table 12) showed
that fish movements actually increased as D.O. decreased. However, the
D.O. levels at warmer temperatures (14-180 C) averaged 6.7 mg/1, a level
which should not limit fish activities. Increased movements more likely
occurred, because of warmer water temperatures, spring spawning
migrations of minor species and late-season migrations of swamp species
out of the floodplain and downstream.
Low (<4 mg/1) D.O. levels did occur during the wet season in the
shallow fringe areas of the floodplain where current velocities were 0-4
cm/sec and when temperatures were relatively high, i.e. >14o c. Most
fish appeared to avoid these low D.O. areas. The only fish seen or
captured under these conditions were young redfin pickerel, which are
reported to be tolerant to low D.O. levels (Huish and Pardue 1978).
During the normal dry season (May-November) when water temperatures
are usually over 15° C, low oxygen conditions are common in waters
63
remaining in coastal plain swamp streams (Kuenzler £t 1977). It is
during the dry season that tolerance to low oxygen conditions increases
the survivability of typical swamp fish species.
Food Habits of Fish Occupying the Floodplain
Invertebrate Food Supply:
Large numbers of aquatic invertebrates were available to fish on
the inundated floodplain during the wet season. In a study of benthic
invertebrate production in Creeping Swamp, Sniffen (1981) found that
invertebrate biomass levels peaked in January/February at 2-6 gm dry
weight per m2. The dominant invertebrate taxa included copepods,
amphipods, isopods, chironomids, oligochaetes and crayfish. Similar
findings in eastern North Carolina floodplain swamps were reported by
Sniffen (1981) and Chapin (1975) for Creeping Swamp and Huish and Pardue
(1978) for slightly larger swamp streams. This study showed differences
in the dominant invertebrates between the floodplain and stream to be
small (Tables 13 and 14), although the floodplain appeared to support
more amphipods and isopods, while oligochaetes seemed to be more common
in the stream. Nevertheless, the greater aquatic area of the inundated
floodplain and its abundant invertebrate food supply was easily utilized
by fish as an extension of the stream.
Crayfish are an important component of the food web in Creeping
Swamp. The two species of crayfish (Procambarus acutus £.,
Fal1icambarus uhleri) collected during this study were not adequately
sampled using smaller, more conventional invertebrate sampling devices.
However, weir traps designed to catch fish yielded large numbers of
64
crayfish (38% of the total catch by number, 19% by weight). Holder
(1971) similarly found large numbers of crayfish on the Suwannee River
floodplain that represented over one-third of the existing animal
biomass. While crayfish were also not adequately sampled by Sniffen
(1981) or Chapin (1975) in their invertebrate studies in Creeping Swamp,
Sniffen suggested that crayfish biomass probably represents an amount
equal to the total of all other invertebrates in the swamp combined
(about 2-4 gm dry weight per m2). Crayfish represent an important
potential food source for fish populations, as well as racoons and other
swamp inhabitants using this seasonal aquatic system.
Except for not adequately sampling crayfish, all of the
invertebrate collection techniques used in this study were effective.
Each type of collection device sampled specific invertebrate habitats
where fish are known to feed. Copepods were found to be the dominant
invertebrates in the drift, while the most commmon benthic organisms
included ol igochaetes, amphipods, isopods and chironomids. Sniffen
(1981) stated that the floodplain soil/litter layer was the dominant
habitat available to invertebrates in Creeping Swamp. Snags and other
underwater obstructi ons can al so be important locations of fish food
organisms (Benke et al_. 1979). However, investigation of these types of
invertebrate habitats was outside the scope of this study. Fish may
have fed more selectively on invertebrates occupying snag habitats
(aufwuchs).
The inundated floodplain represents over 90% of the total aquatic
area in Creeping Swamp during the wet season on an average-area-flooded-
per-day basis and 80% of the total swamp invertebrate production
(Sniffen 1981). Channelization, which eliminates floodplain inundation.
65
would therefore destroy 90% of the available fish habitat and 80% of the
invertebrate food supply. These predictions agree very well with the
earlier fish population studies of Tarpl ee et (1971), who found
about 75% less game fish biomass in channelized streams than in natural
swamp streams in eastern North Carolina.
Food Habits:
Fish, occupying both floodplain and stream, fed almost exclusively
on invertebrates, especially the dominant crustaceans (e.g. amphipods,
isopods, copepods, crayfish) and chironomids. Although several of the
fish collected for gut analysis are known piscivores (e.g. redfin
pickerel), only one fish was found in all of the stomachs examined.
This rather unusual behavior of non-piscivorous activity most likely
reflects the large amounts of invertebrates and cover that is available
to fish when the floodplain is inundated, effectively eliminating the
need and likelihood of fish predation by other fish. While usually
accounting for only l%-2% of the total number of invertebrates consumed,
crayfish represented a large portion of the food consumed by weight and
were most often preyed upon by the normal ly pi sci vorous fi sh species.
Apparently, crayfish and other relatively large invertebrates replaced
fish in the diet of these larger piscivorous fish during the wet season.
Core samples indicated that oligochaetes constituted a large portion
of the benthic invertebrates that occurred in the swamp (Table 15), but
these rapidly-digested soft-bodied organisms (e.g. oligochaetes,
flatworms) were seldom found in fish stomachs. This apparent problem
represents a classical weakness in stomach analysis techniques used to
determine fish food habits. Even though few soft-bodied organisms were
66
seldom found in fish stomachs, there was no conclusive evidence
indicating fish were ignoring these organisms. Indeed, fish probably
fed more heavily upon these abundantly available soft-bodied
invertebrates than gut analysis results indicated. The results from gut
analysis, however, compare fairly well with drift and lift net results.
Average fish stomach fullness estimates indicated that fish fed more
heavily on the floodplain than in the stream (Table 18). The greater
amount of aquatic habitat and invertebrate food supply of the inundated
floodplain may explain this difference. Overall, the fish community
appeared to be feeding on the available invertebrates (except for soft-
bodied invertebrates, e.g. flatworms and oligochaetes) and fed more
actively on the floodplain.
Spawning/Nursery Area Observations
The inundated floodplain appears to serve as a spawning and nursery
area for most of the typical swamp fish species that move into Creeping
Swamp during the wet season. Large numbers of fish in spawning
condition occupied both the floodplain and the stream, appearing most
frequently from February through April. Fish eggs and larvae were not
col 1 ected in cores, 1 i ft or dri ft nets. In the extensi ve area of the
inundated floodplain, fish eggs and 1 arvae woul d be widely-dispersed
and, consequently, difficult to collect using any standard techniques.
However, juvenile fish were prevalent late in the wet season and
indicate that the swamp-stream fish species use Creeping Swamp as a
spawning and nursery area.
67
Importance of Creeping Swamp to Downstream Systems
Ecological Importance;
Sniffen (1981) estimated that small swamp stream systems like
Creeping Swamp represented the majority of the aquatic area of Coastal
Plain river systems on an inundated-per-day basis. The typical swamp
fish species found in this study may be the primary link in the food web
between abundant invertebrate populations of the extensive inundated
floodplains and larger piscivorous fish inhabiting the permanent streams
and rivers located downstream. It was not within the scope of this
study to determine how many typical swamp fish migrate downstream at the
end of the wet season. Tarplee et £]_. (1971) did establish that natural
swamp streams contained 75X more fish biomass than channelized swamp
streams. From this study it is obvious that the higher fish biomass of
natural swamp streams is a direct result of the larger amount of aquatic
habitat available during the wet season. What remains to be
investigated is the degree to which economical ly important fish (e.g.
basses and catfish) feed upon typical swamp fish (e.g. pirate perch,
redfin pickerel, eastern mudminnows, etc.) during their downstream
migrations to permanent waters at the beginning of the dry season. Fish
trapped in stream pools during the dry season appear to be consumed by
semi-aquatic (e.g. water snakes, turtles) or terrestrial (e.g.
kingfishers, herons) predators.
The role of crayfish in the food web of coastal plain stream
swamp/river systems is probably more important than indicated from the
results of this study. The larger (>3 cm carapace length) crayfish were
generally too large to be eaten by the typical small fish species found
68
in this study. From examination of feces, it was evident that racoons
were feeding on the larger crayfish. The extent to which crayfish
migrated out of the system in late spring or survived in situ in burrows
is also unknown. From visual observations during night el ectroshocking
operations, it was apparent that crayfish standing stock biomass greatly
exceeded that of the fish conmunity on the inundated floodplain.
Economic Importance:
The only commercial fish species commonly found in Creeping Swamp
was the American eel. All eel s caught were juveni 1 es in young year
classes (i.e. less than 35 cm TL). The inundated floodplain was a
fertile foraging area for this widely distributed commercial species.
If typical swamp fish are an important seasonal component of the diet of
economically important piscivorous fish in permanent downstream waters,
then the main economic importance of Creeping Swamp is derived from its
ecological role in the coastal plain riverine food web.
SUMMARY
1. The seasonal inundation of Creeping Swamp from December to April
creates a shallow, slow-flowing reservoir some 7-10 km long and
varying from 10 to 800 meters wide depending on changes in water
level of only 50 cm.
2. The fish community of the Creeping Swamp system was dominated by
pirate perch and redfin pickerel. Other frequently occurring
species included flier, mud sunfish, eastern mudminnow, American
eel, banded sunfish, creek chubsucker, bowfin and swampfish.
3. Two-way fixed weir traps were the most effective fish and crayfish
collection devices for both the inundated floodplain and the stream.
Twenty-four hour set duration times were the most efficient and
practical sampling periods.
4. Seasonality of the aquatic system significantly affected fish
occurrence and movement. Important environmental factors were:
A. Water level — the dominant fish species quickly moved into the
aquatic system upon initial fl ooding and were present for the
entire inundation season until declining water levels in late
spring forced evacuation to more permanent waters downstream.
B. Water temperature — minimal fish movements occurred when water
temperatures were less than 6° C.
C. Dissol ved oxygen — general 1 y, D.O. was not a 1 imiting factor
upon fish occurrence or movement during the wet season.
5. Abundant standing crops of aquatic invertebrates were available for
fish foraging early in the wet season.
70
6. Crayfish were present in large numbers in Creeping Swamp during the
wet season and represented an important food source for larger
10.T(>10 cm) fish in the swamp.7. Fish fed almost exclusively on invertebrates, especially thedhoeminant crustaceans (e.g. amphipods, isopods, copepods, crayfish)and chironomids. Crayfish and other relatively large invertebratesappeared to have replaced fish in the diet of the larger piscivorousfish species (e.g. redfin pickerel).8. Oligochaetes constituted a large portion of the benthicinvertebrates in the swamp, but these rapidly-digested soft-bodiedorganisms were seldom found in fish stomachs.9. Fifty-one percent of fish collected in February and March were foundto be in spawning condition.inundated floodplain/stream system serves as a spawning and
nursery area for several swamp-stream fish species during the wet
season.
LITERATURE CITED
Benke, A.C., D.M. Gillespie and F.K. Parrish. 1979. Biological basis
for assessing impacts of channel modification; Invertebrate
production, drift and fish feeding in a southeastern blackwater
river. ERC 06-79, Environmental Resources Center, Georgia
Institute of Technology, Atlanta, 187 p.
Carl ander, K.D. 1969. Handbook of freshwater fishery biology. Vol. 1.
Iowa State Univ. Press. Ames, 752 p.
Carlander, K.D. 1977. Handbook of freshwater fishery biology. Vol. 2.
Iowa State Univ. Press. Ames, 431 p.
Chapin, J.W. 1975. A study of benthic macroinvertebrates in
channelized and unchannelized coastal plain stream-swamps. M.A.
Thesis. East Carolina University.
Crossman, E.J. 1962. The redfin pickerel, Esox a. americanus in North
Carolina. Copeia 1:114-123.
Elliott, J.M. 1973. Some methods for the statistical analysis of
samples of benthic invertebrates. Freshw. Biol. Assoc., Scientific
Publ. No. 25., 148 p.
Forbes, S.A. and R.E. Richardson. 1920. The fishes of Illinois. State
of Illinois, Natural History Survey, 357 p.
Goodnight, J.H. 1979. NLIN Procedure, p. 317-329. Jji: SAS User's
Guide, 1979 Edition. SAS Institute, Raleigh, North Carolina.
Guillory, V. 1979. Utilization of an inundated floodplain by
Mississippi River fishes. Florida Sci. 42(4):222-228.
Holder, D.R. 1970. A study of fish movements from the Okefenokee Swamp
into the Suwannee River. Proc. 24th Ann. Conf. S.E. Assn. Game
Fish Comm. 24:591-608.
Holder, D.P. 1971. Benthos studies in warmwater streams. Ann.
Progress Report, Project F-21-2. Statewide Fisheries Investig.,
Ga. Game and Fish Comm., Dept, of Natural Resources, Atlanta.
Huish, M.T. and G.B. Pardue. 1978. Ecological studies of one
channelized and two unchannelized wooded coastal swamp streams in
North Carolina. U.S. Fish and Wildlife Service, Office of
Biological Services, Harpers Ferry, West Virginia. FWS/OBS-78/85,
72 p.
Kuenzler, E.J., P.J. Mul hoi land, L.A. Ruley and R.P. Sniffen. 1977.
Water quality in North Carolina Coastal Plain streams and effects
of channelization. Water Resources Research Institute. Report No.
127. University of North Carolina, Chapel Hill, 160 p.
72
Lagler, K.F., J.E. Bardach, R.R. Miller, and D.R.M. Passino. 1977.
Ichthyology. John Wiley and Sons, Inc., Second Edition, 506 p.
Mulholland, P.J. 1979. Organic carbon cycling in a swamp-stream
ecosystem and export by streams in eastern North Carol ina. Ph.D.
Dissertation, University of North Carolina, Chapel Hill, 152 p.
Rogers, L.E., W.T. Hinds and R.L. Buschbom. 1976. A general weight
versus length relationship for insects. Ann. Entomol. Soc. Am.
69(2):387-389.
Smock, L.A. 1980. Relationship between body size and biomass of
aquatic insects. Freshw. Biol. 10:375-383.
Sniffen, R.P. 1981. Benthic invertebrate production during seasonal
inundation of a floodplain swamp. Ph.D. Dissertation, University
of North Carolina, Chapel Hill, 176 p.
Strommen, C.A. and E.M. Rasch. 1975. Activity patterns, feeding and
behavior of the pirate perch, Aphredoderus sayanus. Copeia 3:572-
574.
Tarplee, W.H., Jr., D.E. Louder and A.J. Weber. 1971. Evaluation of
the effects of channelization on fish populations in North
Carolina's coastal plain streams. N.C. Wildlife Resources
Commission, Raleigh, 22 p.
Welcomme, R.L. 1969. The biology and ecology of the fishes of a small
tropical stream. J. Zool. Lond. 158, 485-529.
Welcomme, R.L. 1979. Fisheries ecology of floodplain rivers. Longman
Inc. 317 p.
Winner, M.D., Jr. and C.E. Simmons. 1977. Hydrology of the Creeping
Swamp watershed, N.C., with reference to potential effects of
stream channelization. U.S. Geological Survey Water Resources
Investigations 77-26, 54 p.
Yarbro, L.A. 1979. Phosphorus cycling in the Creeping Swamp floodplain
ecosystem and exports from the Creeping Swamp watershed. Ph.D.
Dissertation, University of North Carolina, Chapel Hill, 231 p.