• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Response of Fish Assemblages to Winter in Two Adjacent Warmwater Streams

Posted on: Wednesday, 27 July 2005, 03:00 CDT

ABSTRACT.-

Fish assemblages of small temperate streams are strongly influenced by harsh winter conditions, which are tolerated by year- round resident species and avoided by seasonal migrants. We tested effects of season and longitudinal position on total biomass, species abundances and body condition offish in two neighboring streams (Plum Run, Brintons Run) in southeast Pennsylvania, USA. In each stream we sampled four pools in a downstream reach in close proximity to a larger (6th order) stream and four pools in a more isolated upstream reach during fall, winter and spring. Total biomass depended strongly on pool volume, with which other habitat variables (area, maximum depth, extent of undercut bank) were also correlated. Downstream pools of Plum Run (mean volume 18.1 m^sup 3^, link magnitude 5) had 21 fish species, including 6 species of Centrarchidae that largely vacated the pools in winter. The assemblage in upstream pools of Plum Run and in both upstream and downstream pools of Brintons Run (mean volumes 5.6 m^sup 3^, 5.7 m^sup 3^ and 4.4 m^sup 3^, respectively; all link magnitude 2), consisted of 7 to 11 species. These pools were dominated by Cyprinidae, generally lacked centrarchids and showed little seasonal variation. Three year-round resident cyprinids [Semotilus atromaculatus (Creek chub), Rhinichthys atratulus (Blacknose dace) and Clinostomus funduloides (Rosyside dace)] and the winter emigrant Catastomus commersonii (White sucker) were the most abundant species in the two streams. Body condition of all four species declined substantially in winter; effects of longitudinal position (upstream vs. downstream) on body condition were less pronounced and varied among species. Abundances of the three resident cyprinids were all negatively correlated with the abundances of large, potentially predatory, centrarchids. Occupation of downstream pools in the larger stream by the centrarchids during fall and spring, but emigration from the study area during winter, further modified the effects of winter on the fish assemblage.

INTRODUCTION

Winter has been widely considered a time of both physiological stress and of changing species abundances for fish in small temperate streams (e.g., Schlosser, 1987). Little detailed information, however, is available describing fish assemblages in winter because of difficulties in sampling (Jackson et al., 2001). Overwintering fish often stop feeding and may experience declines in lipid content and body weight, sometimes leading to death (Oliver et al., 1979; Cunjak and Power, 1986; Wicker and Johnson, 1987; Shuter et al, 1989; Smale and Rabeni, 1995; Cargnelli and Gross, 1997). The presence of toxic substances, parasites and other external stressors may accentuate the effects of low temperature (Lemly, 1996).

Physiological stress and associated mortality during winter are thought to be more severe in upstream waters, owing primarily to lower or more variable temperatures than occur downstream (Schlosser, 1990), often resulting in populations dominated by young individuals (Schlosser, 1987). Fish that are year-round residents in small streams, thus, may be more physiologically tolerant of winter temperatures than fish in larger streams (Matthews and Styron, 1981).

Small streams may also be occupied during non-winter months by seasonal migrants that move to larger streams during winter. As discussed by Smithson andjohnston (1999), Skalski and Gilliam (2000) and Rodriguez (2002), movement tendency is known to vary considerably not only between species, but also among individuals within populations, determined in part by body size and condition (Sabo and Orth, 2002). Migration may depend strongly on the proximity of a suitable winter refugium. As a result, seasonal migrants are often more prevalent in fish assemblages close to larger streams than at the periphery of stream networks (Osborne and Wiley, 1992).

Piscivory, like physiological stress and seasonal migration, may reduce the local abundance of smaller fish both through direct consumption and predator avoidance (e.g., Jackson et al., 2001). Piscivores are typically thought to be more abundant in larger streams, which provide more stable year-round conditions suitable for the growth of larger fish (Schlosser, 1987), but may also colonize small streams during non-winter months.

This study describes the response of pool-dwelling fish assemblages to winter in two adjacent warmwater streams. In each stream we sampled pools in a downstream reach near the confluence with a larger stream and an upstream reach more distant from the confluence. The pools were visited during fall, winter and spring. We tested the effects of season and longitudinal position on: (a) biomass, (b) species abundances and (c) body condition in the two streams. We further use these results to address the following questions: (1) Does body condition of fish in upstream pools decline more in winter than in downstream pools?, (2) Does seasonal migration affect the fish assemblage in downstream pools with greater access to a winter refugium more than in upstream pools?, (3) Are piscivore abundances seasonally greater in downstream pools and are they negatively associated with the abundances of smaller resident species? and (4) Are these seasonal patterns in biomass, species abundances and body condition replicated in two adjacent streams with different habitat features?

MATERIALS AND METHODS

STUDY SITES

Research was conducted on two small tributaries of Brandywine Creek, a 6th order stream at the points of confluence, in southeastern Pennsylvania (Fig. 1). Plum Run has a main stem length of 5.1 km and drainage area of 9.4 km^sup 2^. Land use within the watershed consists primarily of single-family homes (48%) and agriculture (26%), a golf course (4%) and university property (2%). The riparian corridor consists of mixed tree cover, fields and mowed grass. Brintons Run has a main stem length of 3.3 km and drainage area of 3.9 km^sup 2^. Land use within the watershed is dominated by agriculture (47%), deciduous forest (30%) and single family homes (21%). Unlike Plum Run, Brintons Run is bordered by forest throughout its length. Additional watershed information is available in Butler (2001).

Upstream and downstream reaches in both Plum Run and Brintons Run were each 200 m long, with multiple pools and intervening riffles. Downstream reaches were within 300 m of Brandywine Creek and in its floodplain. The confluences of both streams with Brandywine Creek were shallow and partly occluded by sand bars, but discharge at the mouth of Plum Run at base flow (0.11 m^sup 3^/s) exceeded that of Brintons Run (0.03 m^sup 3^/s). Upstream reaches were more isolated from Brandywine Creek by distance (720 m and 440 m for Plum Run and Brintons Run, respectively), shallow riffles and steep gradient. Four pools at each reach in both streams were chosen for study. All 16 pools were ≥ 30 cm deep and were bounded by well-defined riffles ≤ 5 cm deep during baseflow.

FIG. 1.-Map of study area, showing the watersheds and sampled reaches of Plum Run and Brintons Run, tributaries of Brandywine Creek, Chester County, Pennsylvania

FISH COLLECTION

Fish were sampled at each pool in fall (F) during 4-10 October 1997, winter (W) during 2-10 January 1998 and spring (S) during 5- 13 May 1998 (n = 48 sampling events). Fish were collected during three complete passes from the downstream to upstream end of each pool, using a Coffelt model BP-1C backpack electrofisher set at 50- 70 W direct current. Block nets (1.7 mm mesh) were placed at the beginning and at the end of each pool to prevent escape of fish during sampling. After each pass, fish were identified to species, weighed (0.1 g) and measured for total length (1 mm). Following sampling, live fish were returned to the stream.

Total capture efficiency in each pool, based on removal- depletion from the three passes (Van Deventer and Platts, 1983), averaged 88.6% (SE = 1.7%) in Plum Run and 94.6% (SE = 1.3%) in Brintons Run. Neither season nor pool volume (see below) significantly affected these estimates (Butler, 2001). Mortality caused by sampling was <1% during fall and winter (mortality was not determined in spring), due in large part to the low levels of direct current applied during capture (Butler, 2001). Thus, we are confident that a large proportion offish actually present in the pools was sampled, with little habitat or seasonal bias and that handling of the fish during capture did not affect the assemblages during subsequent visits.

HABITAT MEASUREMENTS

In order to describe pool size we set up parallel transects, 1 m apart, across the stream from the beginning to the end of each pool and measured depth every 50 cm along each transect. Pool volume and surface area then were calculated using Surfer 6.0 software (Golden Software, Inc.). No major floods modified stream habitat features during the study period. Pool volume and area were considered constants through the three seasonal visits based on visual inspection and the maintenance of nearly uniform water height at a staff gage in the center of each pool. The extent of undercut bank, measured during mapping, was likewise treatedas a constant for each pool.

Percent canopy cover was calculated using an American Forestry spherical densiometer (Lemmon, 1956), placed at the center of each pool. Position within each stream network was described using link magnitude, the number of 1st-order source streams upstream of each pool (Shreve, 1966). Water temperature was measured at mid-depth in each pool during all three seasons. Additional measures of pool morphology and water chemistry are available in Butler (2001).

STATISTICAL ANALYSES

All data were analyzed using Statistica (Statsoft, Inc.). We used a split-plot repeated measures analysis of covariance (ANCOVA) to describe the combined effects of season, pool size and reach (upstream vs. downstream) on total fish biomass in each pool. Pools were considered as subjects, reach as a between-subjects factor, season as a repeated measures factor within subjects and log^sub 10^(pool volume) as a covariate. Streams were considered separately. The Student-Newman-Keuls (SNK) test was performed to detect significant pairwise differences in biomass per pool between seasons.

Species populations comprising at least 50 individuals and present in both streams were classified as "winter emigrants" if abundances in winter were significantly less than the abundances in fall and spring based on a chi-square test (significance set at P < 0.001); "winter immigrants" had significantly greater abundances in the study area during winter (P < 0.001). Populations with at least 50 individuals and that persisted in approximately equal numbers in all three seasons were considered "residents."

Canonical correspondence analysis (CCA) was used to relate species abundances to key environmental variables and to assess the relative importance of those variables in structuring the fish assemblage. Effects of seasonal change were evaluated by introducing season as a nominal variable and percent canopy cover (arcsin- square root transformed) as a continuous variable. Log^sub 10^(pool volume) was used to summarize effects of differing pool size. Position of each reach within the stream network was represented using link magnitude and by including the four reaches as nominal variables. Significance of these environmental variables in explaining unique additional variation in the fish assemblages among sites was evaluated using a stepwise forward procedure with 999 Monte Carlo permutations (ter Braak and Smilauer, 1998). Log- transformed species abundances (number of individuals) were used to describe the fish assemblage, with the exclusion of four species represented by only one individual among all pools.

Second, in order to test for environmental effects on condition, we used 3-way ANCOVA, with season, stream and reach as fixed factors and fish length as a covariate. Analysis for C. funduloides did not include downstream pools in Plum Run because of low sample sizes. The SNK test was used to compare condition among seasons.

In evaluating the impact of predation on the fish assemblage, we considered Rock bass (Ambloplites rupestris) >75 mm, other bass (Micropterus spp.) >100 mm and all sunfish (Lepomis spp.) >75 mm total length to be potential piscivores (Schlosser, 1987). Two cyprinids (Semotilus atromaculatus and S. corporalis) were considered potential piscivores at total lengths >81 mm (Copes, 1978; Reed, 1971). Abundances of these taxa were related to pool volume and to the abundances of the same four numerically dominant species (Rhinichthys atratulus, Clinostomus funduloides, Semotilus atromaculatus, Catostomus commersonii) using Spearman rank correlations, with all pool sampling visits (n = 48) included in the analysis. Because pools were sampled in all three seasons, observations lacked complete independence and significance was set conservatively at P = 0.01.

TABLE 1.-Measurements summarizing differences in pool morphology, season and reach (mean SE) of pools in Plum Run and Brintons Run

RESULTS

ENVIRONMENTAL FEATURES OF POOLS

Mean pool volume, pool surface area, maximum depth and the area of undercut bank were all greater at the downstream reach on Plum Run than in the other three reaches (Table 1). Log^sub 10^ (pool volume) was strongly correlated with area, maximum depth and the areal extent of undercut bank (r = 0.89, 0.61 and 0.83, respectively) and is used as a surrogate for pool morphology in subsequent analyses. Pool substrate in all four reaches was dominated by fine sediments (sand, silt), with coarse gravel and organic mud in small areas of unusually high or low current velocity, respectively. Two variables, water temperature and percent canopy, were strongly influenced by season. Canopy cover varied substantially among pools within reaches, but was lower in winter than in either fall or spring. Temperature varied little between streams or longitudinal position (a portion of the variation within seasons was attributable to sampling reaches on different days). Season was used as a discrete variable to represent changing temperature and canopy cover in subsequent analyses. Link magnitude, used as a further indicator of longitudinal position in both streams, was greater for the downstream reach of Plum Run than for the other three reaches.

EFFECTS OF SEASON AND REACH ON FISH BIOMASS

Seasonal changes in biomass were evident, but there was no consistent pattern among reaches (Fig. 2a). In Plum Run, biomass declined during winter in downstream pools (P < 0.01), but not upstream. Biomass in Brintons Run also varied significantly with season (P < 0.01), with greatest biomass during spring, but not between upstream and downstream reaches. Biomass increased with increasing pool volume in both streams (Fig. 2b).

SPECIES ABUNDANCES

Plum Run contained 22 species, of which 20 were found downstream and 11 upstream (Table 2). Centrarchids, especially Lepomis spp., were common in the downstream pools during spring and fall, but were rarely found upstream; centrarchids were nearly absent during winter and their absence largely accounted for the winter decline in total biomass. Species of Cyprinidae were found both upstream and downstream, with Semotilus atromaculatus and Clinostomus funduloides typically more abundant upstream. By contrast, Brintons Run had 12 species, including 11 downstream and 7 upstream. Centrarchids were rare and upstream and downstream assemblages were similar. Based on chi-square evaluation of their seasonal occurrences, two species of centrarchids and C. commersoni were classified as winter emigrants (although many individuals of C. commersoni were found in the study area during all three seasons) and the taxon "other minnow species" was identified as a winter immigrant, based largely on increased abundances in downstream pools of Plum Run during winter. Three species of cyprinids (S. atromaculatus, Rhinichthys atratulus, C. funduloides) were considered permanent residents.

Impacts of environmental variables on species composition in the two streams, based on CCA, are summarized in Table 3. Together the environmental variables explained 63% of variation in the species data. Reach position was the most important determinant of assemblage structure, as indicated both by link magnitude and the downstream reach in Plum Run, which each explained 44% of total variation in the species data if considered alone. Pool volume accounted for 37% of total variation in the fish assemblages among pools when considered alone.

Ordination diagrams based on the CCA results provided a convenient means of summarizing the relationship of individual species abundances to particular environmental variables and of visually identifying species associations (Fig. 3). The continuous environmental variables are shown as vectors, with larger values toward the head of the arrows and the inferred habitat "optimum" for each species is indicated by its position relative to the habitat vectors (ter Braak, 1986). For example, optima for the sunfishes Lepomis auritus, L. cyanellus, L. macrochirus and L. gibbosus and basses Micropterus salmoides and Ambloplites rupestns occurred near the heads of the vectors for pool volume and area in the lower righthand quadrant of Figure 3, reflecting their greater use of larger pools. Smaller pools in both streams were typically occupied by the the year-round resident species Semotilus atromaculatus, Rhinichthys atratulus and Clinostomus funduloides and by smaller individuals of Catostomus commersonii, shown on the left side of the ordination diagram. What appears clear is that pool size and link magnitude (with long vectors oriented along the first canonical axis) differentiated the downstream fish assemblage of Plum Run rather than simply its proximity to Brandywine Creek.

FIGS. 2A-B.-(a) Seasonal variation in total biomass in upstream (U) vs. downstream (D) reaches in Plum Run (P) and Brintons Run (B): mean SE; (b) effect of pool volume on biomass: Δ = fall, [black circle] = winter, [white square] = spring

BODY CONDITION

Season significantly affected condition (corrected for body weight) in all four species examined (Table 4). Lowest condition consistently occurred during winter. Condition factors of individuals of all four species were similar in Plum Run vs. Brintons Run. We also found significant, but less consistent, effects of reach (upstream vs. downstream) and of interactions between reach and either season or stream for some species. Mean condition factors for the four species are compared by stream, reach and season in Figures 4a-d.

EFFECTS OF POTENTIAL PISCIVORES

The abundances of larger centrarchids and cyprinids considered to be potential piscivores were greatest in downstream pools of Plum Run during fall and spring (Table 2) and were positively correlated with pool volume (Table 5). Large sunfish were the most frequently observed of the piscivore taxa and were negatively related to abundances of the three resident species Semotilus atrom\aculatus (all individuals), Khinichthys atratulus and Clinostomus funduloides. Abundances of large Semotilus, by contrast, were not negatively associated with coexisting R. atratulus or C. funduloides.

DISCUSSION

Migrants (particularly centrarchids) gained seasonal access to downstream pools but not upstream pools in Plum Run and their emigration during winter greatly changed total biomass and species abundances of the assemblage as predicted. The pattern, however, was not replicated in Brintons Run, with downstream pools that were either too small, lacked critical habitat (e.g., cover, undercut banks) or were inaccessible to these seasonal migrants. second, several smaller cyprinids were negatively associated with the centrarchids, but coexisted with potentially predatory cyprinids Semotilus spp. Third, winter caused strongly declines in body condition of the resident fish, but fish in upstream pools did not consistently experience greater weight loss than fish in downstream pools as had been predicted. We evaluate the plausibility of these three assertions in the sections that follow.

SEASONAL MOVEMENT OF CENTRARCHIDS

Documentation of seasonally based movement of centrarchids has been infrequent. Berra and Gunning (1972), in their study of longear sunfish (Lepomis megalotis) in three headwater Louisiana streams, observed relatively little movement offish within well-defined home ranges during the warmer months, but a substantial downstream exodus prior to winter. Larger individuals were particularly likely to migrate.

Just how large pools would have to be to sustain year-round centrarchid populations is not known. Taylor (1997), who studied the fish assemblage of a 3rd order stream in Oklahoma during early summer, noted that longear sunfish (Lepomis megalotis) were typically present only in pools >10 m^sup 3^; minimum size of occupied pools with no surface-water connection to the stream was much larger (ca. 60 m^sup 3^). Green sunfish (L. cyanellus) displayed a similar pattern. Johnston and Smithson (2000) have further shown that movement of L. megalotis and L. cyanellus from small pools (5-7 m3) was greater than from larger pools. Pools in our study were consistently <10 m^sup 3^ except downstream in Plum Run.

In effect, seasonal migration by the centrarchids in Plum Run may be an infrequent phenomenon associated with particular populations of normally more sedentary species, occasioned by a combination of pools sizes that were adequate for seasonal occupancy but apparently not attractive for year-round persistence and ready access to a larger winter refugium.

TABLE 2.-Total abundances of species and (at bottom) of potential predators, by reach (U, D) and season (F, W, S) in Plum Run and Brintons Run. CCA codes indicate species shown in Figures 3a-b. Common species are identified as residents (R), winter emigrants (WE) or winter immigrants (WI) based on chi-square analysis (see text)

TABLE 2.-Total abundances of species and (at bottom) of potential predators, by reach (U, D) and season (F, W, S) in Plum Run and Brintons Run. CCA codes indicate species shown in Figures 3a-b. Common species are identified as residents (R), winter emigrants (WE) or winter immigrants (WI) based on chi-square analysis (see text)

EFFECTS OF PISCIVORES ON SMALLER FISH

Large centrarchids, dominant in fall and spring in downstream pools of Plum Run, have been shown repeatedly to cause either substantial mortality or predator avoidance behavior in smaller fish (soejackson et al, 2001 for a review). Semotilus atromaculatus, often characterized as an inhabitant of more isolated upstream pools (Lotrich, 1973; Felley and Hill, 1983; Taylor, 1997), was rarely found in pools with centrarchids in our study. By contrast, larger individuals of S. atromaculatus and S. corporalis coexisted with Rhinichthys atratulus, Clinostomus funduloides and Catostomus commersonii'm the smaller pools of Plum Run and Brintons Run. S. atromaculatus has been regarded as a relatively ineffective piscivore compared to bass (Schlosser, 1988). Schlosser and Ebel (1989), who experimentally manipulated densities of 5. atromaculatus in pools of a headwater stream in North Dakota, observed minimal impacts on population densities or avoidance behavior of three species of coexisting minnows. Fraser and Emmons (1984), who documented the presence of R. atratulus in die diet of adult creek chub, noted that the dace showed little predator avoidance during daylight hours but greater avoidance at dusk when the chubs were more actively feeding. Large individuals of 5. corporalis share with S. atromaculatus a very generalized diet. Fish comprised 9.6% and 6.8% of food volume in Reed's (1971) study of age 1+individuals of S. corporalis in the Winooski River, Vermont and Yellow Creek, Pennsylvania, respectively. Our view that the seasonal distribution of centrarchids, but not of S. atromaculatus or S. corparalis influenced the distributions of smaller fish is thus consistent with previous work.

TABLE 3.-CCA of Plum Run and Brintons Run, showing the significance of individual environmental variables in uniquely explaining variation in the fish assemblages among pools, λ = eigenvalue; F(p) indicate significance when tested alone

FIG. 3.-Effects of pool size (VOL = Iog10pool volume), season (CANOPY = percent canopy cover; centroids for sampling date are FALL, WINTER and SPRING) and reach (LINK = link magnitude; centroids are PU = upstream Plum Run, PD = downstream Plum Run, BU = upstream Brintons Run, BD = downstream Brintons Run) on pool fish assemblages. Codes for fish species are in Table 4. The first canonical axis accounted for 43.5% of unconstrained variation in the species data and 73.9% of the species-environment relation; cumulative variation explained by both axes was 49.3% of unconstrained variation and 83.6% of the species-environment relation

TABLE 4.-Effects of season, stream, reach and body length on condition (K) of four species based on 3-way ANCOVA. Shown are F statistics for all main effects and two-way interactions (** = P < 0.01, *** = P < 0.001). In addition, pairwise differences based on SNK comparisons are summarized for season (F = fall, W = winter, S = spring). Standardized slopes (β) are shown for effects of the covariate body length

FIGS. 4A-D.-Effects of season on body condition in (a) -R. atratulus, (b) S. atromaculatus, (c) C. funduloides and (d) C. commersonii in pools in Plum Run upstream (PU), Plum Run downstream (PD), Brintons Run upstream (BU) and Brintons Run downstream (BD): mean SE

BODY CONDITION

Body condition of fish has been measured using a variety of metrics (e.g., Bolger and Connelly, 1989; Vila-Gispert and Moreno- Amich, 2001). Simultaneous evaluation of both spatial and temporal effects on condition, however, has not been previously attempted to our knowledge.

The largest effect on condition was clearly seasonal for all four species measured. For example, Rhinichthys atmtulus experienced an average decline of 18-31% between fall and winter depending on reach. Cunjak and Power (1986), in their study of R. atratulus in a southern Ontario stream, similarly noted substantial declines in condition (measured as in equation 1, but with fork length rather than total length) to minimum values between January and March. Juveniles lost proportionally more weight (K declined by 20% between fall and winter) than adults (K declined by 6%). Despite the very substantial energy losses and increased mortality, downstream migration may be a less attractive option for these small fish because of increased predation pressure downstream.

TABLE 5.-Spearman correlation coefficients of 4 potential piscivore taxa with pool volume and the abundances of 4 numerically abundant species of smaller fish (* = P < 0.01, ** = P < 0.001)

SUMMARY

Our study of two adjacent streams within the same drainage supports several paradigms of long standing in stream fish ecology, perhaps most succinctly articulated in Schlosser's ( 1987) conceptual framework. Upstream pools in both streams supported similar, species-poor assemblages dominated by small cyprinids and with few piscivores. The assemblage differed little among seasons and most species were presumed to be year-round residents. Substantial winter declines in body condition were observed among the four species measured.

Downstream pools in the two streams, by contrast, supported different fish assemblages. In one stream (Plum Run) larger pools and likely greater access to and from a winter refugium permitted larger fish (especially Centrarchidae) to occupy the reach during spring and fall and emigrate in winter. Their presence in turn produced marked seasonal changes in total biomass and the abundances of smaller species.

Acknowledgments.-We thank T. Andrews, M. Carfioli, A. Faulds, L. Fords, M. Johnson and A. Thompson for their many hours of help in fieldwork. M. Boyer provided assistance with fish identifications, A.Johnson helped with the preparation of bathymetie maps for the pools sampled and D. Nieman provided support with our statistical analyses. We thank D. Nieman, M. Boyer and R. Horwitz for helpful suggestions on an earlier draft of the manuscript.

LITERATURE CITED

BERRA, T. M. AND G. E. GUNNING. 1972. Seasonal movement and home range of the longear sunfish, Lepomis megalotis (Rafinesque) in Louisiana. Am. Midi. Nat., 88:368-375.

BOLGER, T. AND P. L. CONNOLLY. 1989. The selection of suitable indices for the measurement and analysis offish condition./ KsA Biol, 34:171-182.

BUTLER, L. H. 2001. The effects of seasonality and location on fish communities in two adjacent warmwater streams in southeastern Pennsylvania. M.S. Thesis, West Chester University, West Chester, Pennsylvania.

CARGNELLI, L. M. AND M. R. GROSS. 1997. Fish energetics: larger individuals emerge from winter in better condition. Trans. Am. Fish. Soc., 126:153-156.

CARLANDER, K. D. 1977. Handbook offreshwater fishery biology, Vol. 2. Iowa State University Press, Ames.

COPES, F. C. 1978. Ecology of the creek chub. Reports on the Fauna and Flora of Wisconsin, 1978:1-21.

CUNJAK, R. A. AND G. POWER. 1986. Winter biology of the blacknose dace, Rhinichthys atratulus, in a southern Ontario stream. Environ. Biol. Fishes, 17:53-60.

FELLEY, J. D. AND L. G. HILL. 1983. Multivariate assessment of environmental preferences of cyprinid fishes of the Illinois River, Oklahoma. Am. Mid. Nat., 109:209-221.

FRASER, D. F. AND E. E. EMMONS. 1984. Behavioral response of blacknose dace (Rhinichthys atratulus) to varying densities of predatory creek chub (Semotilus atromaculatus). Can. J. Fish. Aquat. Sd., 41:364-370.

JACKSON, D. A., P. R. PERES-NATO AND J. D. OLDEN. 2001. What controls who is where in freshwater fish communities-the roles of biotic, abiotic and spatial factors. Can. J. Fish Aquat. Sd, 58:157- 170.

JOHNSTON, C. E. AND E. B. SMITHSON. 2000. Movement patterns of stream-dwelling sunfishes: effects of pool size./. FreshwaterEcol, 15(4):565-566.

LEMLY, A. D. 1996. Wastewater discharges may be most hazardous to fish during winter. Envir. Pollution, 93:169-174.

LEMMON, P. E. 1956. A spherical densiometer for estimating forest overstory density. Forest Science, 2:314-320.

LOTRICH, V. A. 1973. Growth, production, and community composition of fishes inhabiting a first-, second-, and third-order stream of eastern Kentucky. Ecol. Monogr., 43:377-397.

MATTHEWS, W. J. AND J. T. STYRON. 1981. Tolerance of headwater vs. mainstream fishes for abrupt physicochemical changes. Am. Midi. Nat., 105:149-158.

OLFVTR, J. D., G. F. HOLETON AND K. E. CHUA. 1979. Overwinter mortality of fingerling smallmouth bass in relation to size, relative energy stores, and environmental temperature. Trans. Am. Fish. Soc., 108:130-136.

OSBORNE, L. L. AND M. J. WILEY. 1992. Influence of tributary spatial position on the structure of warmwater fish communities. Can. J. Fish. Aquat. Sd., 49:671-680.

PETERSON, J. T. AND C. F. RABENI. 2001. The relation offish assemblages to channel units in an Ozark stream. Trans. Am. Fish. Soc., 130:911-926.

REED, R. J. 1971. Biology of the fallfish, Semotilus corparalis (Pisces, Cyprinidae). Trans. Am. Fish. Soc., 1971: 717-725.

RODRIGUEZ, M. A. 2002. Restricted movement in stream fish: the paradigm is incomplete, not lost. Ecology, 83:1-13.

SABO, M. J. AND D. J. ORTH. 2002. Effects of hunger on foraging behavior of age-0 smallmouth bass in a heterogeneous environment. Envir. Biol. Fishes, 63:217-222.

SCHLOSSER, I. J. 1987. A conceptual framework for fish communities in small warmwater streams, p. 17-24. In: W. J. Matthews and D. C. Heins (eds.). Community and evolutionary ecology of North American stream fishes. University of Oklahoma Press, Norman.

_____. 1988. Predation rates and the behavioral response of adult chub and smallmouth bass predators. Copeia, 1988:691-697.

_____. 1990. Environmental variation, life history attributes, and community structure in stream fishes: implications for environmental management and assessment. /. Envir. Man., 14:621- 628.

_____ AND K. K, EBEL. 1989. Effects of flow regime and cyprinid predation on a headwater stream. Ecol. Mon., 59:41-47.

SHREVE, R. L. 1966. Statistical law of stream numbers./ Geol., 74:17-37.

SHUTER, B. J., P. E. IHSSEN, D. L. WALESAND E. J. SNUCINS. 1989. The effects of temperature, pH and water hardness on winter starvation of young-of-the-year smallmouth bass, Micropterus dolomieui Lacepede. /. Fish Biol., 35:765-780.

SKALSKI, G. T. AND J. F. GILLIAM. 2000. Modeling diffusive spread in a heterogeneous population: a movement study with stream fish. Ecology, 81:1685-1700.

SMALE, M. A. AND C. F. RABENI. 1995. Hypoxia and hypothermia tolerances of headwater stream fishes. Trans. Am. Fish. Soc., 124:698-710.

SMITHSON, E. B. AND C. E. JOHNSTON. 1999. Movement patterns of stream fishes in a Ouachita Highlands stream: an examination of the restricted movement paradigm. Trans. Am. Fish. Soc., 128:847-853.

TAYLOR, C. M. 1997. Fish species richness and incidence patterns in isolated and connected stream pools: effects of pool volume and spatial position. Oecologia, 110:560-566.

TER BRAAK, C. J. F. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology, 65:1167-1179.

_____ AND P. SMILAUER. 1998. CANOCO Reference Manual and User's Guide to CANOCO for Windows (Version 4). Centre for Biometry, Wageningen, Netherlands. 351 p.

VAN DEVENTER, J. S. AND W. S. PLATTS. 1983. Sampling and estimating fish populations from streams. Trans. N. Am. Wildlife and Natural Resources Conf., 48:349-354.

ViLA-GiSPERT, A. AND R. MoRENO-AMiCH. 2001. Testing alternative methods of analysing differences in fish condition, with an example using the mosquitofish Gambusia holbrooki. Archiv. Hydrobiol, 153: 145-154.

WICKER, A. M. AND W. E.JOHNSON. 1987. Relationships among fat content, condition factor, and first-year survival of Florida largemouth bass. Trans. Am. Fish. Soc., 116:264-271.

SUBMITTED 30 OCTOBER 2003 ACCEPTED 18 OCTOBER 2004

LANCE H. BUTLER1

Philadelphia Water Department, Office of Watersheds, 1101 Market St., 4th floor, Philadelphia, Pennsylvania 19107

AND

G. WINFIELD FAIRCHILD

Department of Biology, West Chester University, West Chester, Pennsylvania 19383

1Corresponding author.

Copyright American Midland Naturalist Jul 2005

Source: American Midland Naturalist, The

More News in this Category