University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln

Great Plains Research: A Journal of Natural and Social Sciences Great Plains Studies, Center for

Fall 2006

River Size and Fish Assemblages in Southwestern South Dakota

Christopher Hoagstrom South Dakota State University

Steven S. Wall South Dakota State University

Jeremy P. Duehr South Dakota State University

Charles R. Berry, Jr. South Dakota State University

Follow this and additional works at: https://digitalcommons.unl.edu/greatplainsresearch

Part of the Other International and Area Studies Commons

Hoagstrom, Christopher; Wall, Steven S.; Duehr, Jeremy P.; and Berry, Jr., Charles R., "River Size and Fish Assemblages in Southwestern South Dakota" (2006). Great Plains Research: A Journal of Natural and Social Sciences. 845. https://digitalcommons.unl.edu/greatplainsresearch/845

This Article is brought to you for free and open access by the Great Plains Studies, Center for at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Great Plains Research: A Journal of Natural and Social Sciences by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. Great Plains Research 16 (Fall 2006):117-26 © Copyright by the Center for Great Plains Studies, University of Nebraska-Lincoln

RIVER SIZE AND FISH ASSEMBLAGES IN SOUTHWESTERN SOUTH DAKOTA

Christopher W. Hoagstrom,l Steven S. Wall,2 Jeremy P. Duehr'

Department of Wildlife and Fisheries Sciences South Dakota State University, Box 2140B Brookings, SD 57007 pecospupfish@ hotmail. com

and

Charles R. Berry, Jr.

South Dakota Cooperative Fish and Wildlife Research Unit U.S. Geological Survey South Dakota State University, Box 2140B Brookings, SD 57007

ABSTRACT-We studied relations between river size, fish species diversity, and fish species composition along four major rivers in the Great Plains of southwestern South Dakota to assess patterns of species diversity and composition. We expected diversity to increase with river size and fish composition to change via species addition downstream. Previous surveys of 52 sampling stations provided fish assemblage data, and we used the Geographic Information System (GIS) to determine watershed area by station. Watershed area did not predict species richness or species diversity (Fisher's a), so species richness of 12 ± 3.5 SD species and Fisher's a of2.3 ± 0.87 SD characterized species diversity in the study area. Cluster analysis of faunal similarity (S!i>rensen's Index) among the 52 sampling stations identified two geographically distinct faunal divisions, so species composition was variable within the study area, but changed via species replacements among faunas rather than species addi­ tions downstream. Nonnative species were a minor component of all faunas. Uniform species diversity may be a recent phenomenon caused by impacts of Missouri River dams on native large-river fishes and the unsuitability of rivers in the Great Plains for nonnative species. Variation in faunal composition may also be recent because it was affected by dams.

Key Words: Bad River, Belle Fourche River, Cheyenne River, longitudinal succession, White River

INTRODUCTION

Regional studies that compare stream fish assemblage of the importance of geomorphic context for such studies characteristics are important for understanding the effects (Montgomery 1999; Walters et al. 2003). For example, of environmental factors and determining the impor­ local geomorphic features increasingly influence stream tance of regional versus local conditions (Matthews and fish assemblages in regions with complex geology, but are Robison 1998; Marsh-Matthews and Matthews 2000). less important where geology is uniform (Montgomery Recognition that stream habitat and disturbance regimes 1999; Walters et al. 2003). Stream fish assemblages in affect stream fish assemblages has increased awareness regions with uniform geology are typically influenced by longitudinal changes in stream habitat and disturbance regimes. Downstream increases in stream size correspond 1 Current address: Department of Zoology, Weber State Uni­ versity, 2505 University Circle, Ogden, UT 84408 to higher fish species diversity as a result of the higher habitat diversity and less severe disturbance regimes 2 Current address: James River Water Development District, Box 849, Huron, SD 57350 (Lotrich 1973; Schlosser 1982). Longitudinal succession

3 Current address: Westwood Professional Services, 7699 Ana­ of stream fish assemblages in regions with uniform geol­ gram Drive, Eden Prairie, MN 55344 ogy usually consists of species additions from upstream

117 118 Great Plains Research Vol. 16 No.2, 2006 to downstream, whereas species replacements (faunal to form the Lower Cheyenne River. The Belle Fourche, breaks) normally correspond to geomorphic boundaries Cheyenne, and White rivers are perennial throughout (Rahel and Hubert 1991). South Dakota because they are fed by extensive water­ Historical and biogeographical contexts are also sheds that extend upstream into adjacent states, whereas important for understanding ecological patterns (Rick­ the entire Bad River drainage lies within South Dakota lefs 1987). Stream fish assemblage patterns may change and much of the mainstem is intermittent during dry dramatically following environmental alterations (e.g., periods. Cross and Moss 1987; Hoagstrom 2003). For example, Rivers of southwestern South Dakota have deeply dams modify assemblages by truncating fish dispersal up­ dissected valleys (Thornbury 1965) and a long history of stream (Luttrell et al. 1999) and interrupting longitudinal human modification for irrigated agriculture (Riley et al. environmental gradients downstream (Bonner and Wilde 1955; Caldwell 1983; Sando 1991) and livestock (Culler 2000). The introduction of nonnative fishes accelerates 1961). Large dams are present on the Belle Fourche and fish assemblage change, but the effect of nonnatives varies Upper Cheyenne rivers (Sando et al. 2001), but the White from enhanced assemblages with persistent natives and River is undammed in South Dakota and the Bad River established nonnatives (Gido and Brown 1999) to reduced is entirely undammed. However, dams on the Missouri assemblages dominated by nonnatives (Lemly 1985). River impound the mouths of the Cheyenne, Bad, and This study is a regional analysis of fish assemblages White rivers and isolate them from each other (Fig. 1). from 52 sampling stations distributed along four major rivers of the Great Plains in southwestern South Da­ River Size kota, where the underlying geology is relatively uniform (Thornbury 1965). The first objective is to document the We used an analysis of U.S. Geological Survey gage regional relation of river size and disturbance regimes to data from water years 1995 through 1999 to describe flow fish assemblage structure. Previous studies in the Great regimes of the region. These data were available for 14 Plains suggest that species diversity will increase with gaging stations on the rivers we studied and corresponded river size and that species composition will change via with the sampling period. We used simple linear regres­ species additions downstream (Rahel and Hubert 1991; sion (Sokal and Rohlf 1995) to summarize relations of Barfoot and White 1999). Rahel and Hubert (1991) pro­ mean discharge and disturbance regime to watershed area posed that reduced disturbance downstream accounted with the expectation that watershed area would predict for increased species diversity. The second objective is both (Horwitz 1978). We used the R-B Index (Baker et al. to compare species diversity trends with species com­ 2004) to quantify disturbance regimes. It is a measure of position trends. Studies outside the Great Plains suggest discharge flashiness, derived using the formula: local species diversity is determined by regional factors, whereas local environmental factors determine species composition (Matthews and Robison 1998; Marsh-Mat­ thews and Matthews 2000; Walters et al. 2003). The third where q is mean daily discharge. Rivers with higher dis­ objective is to assess the influence of nonnative species charge flashiness (higher R-B Index values) have relative­ on fish assemblage patterns. Previous studies in the Great ly low base-flow coupled with a relatively high frequency Plains suggest nonnatives increase species diversity, es­ of short-lived, high-discharge events (flash floods). pecially in association with reservoirs and dense human populations (Gido et al. 2004; Falke and Gido 2006). Fish Assemblages

METHODS Previous investigators (Doorenbos 1998; Hampton 1998; Fryda 2001; Milewski 2001) captured fishes along Study Area each river at evenly-spaced sampling stations (Fig. 1). Each sampling station was 36 mean wetted widths in We studied the mainstem Belle Fourche, Cheyenne, length. All studies had the common goal of collecting all Bad, and White rivers, South Dakota (Fig. 1). The Belle species present. Researchers used 4.7 or 8.0 mm mesh bag Fourche River is the north fork of the Cheyenne River, seines that were 5 to 9 m long to capture fishes. They made which runs north of the Black Hills, whereas the Upper three seine passes through all habitats at each station. Cheyenne River runs south of the Black Hills. They join Milewski (2001) used block seines to improve estimates of River Size and Fish Assemblages in Southwestern South Dakota 119

+ Assemblage I * Assemblage II A Assemblage III • Assemblage IV

NEBRASKA o 25 50 100 150 200 EH3::HECE===C===E==::::31 Kilometers

Figure 1. Map of southwestern South Dakota showing the four major rivers. Sampling stations are designated by symbols that cor­ respond to fish assemblage types. U.S. Geological Survey gaging stations are designated by letters. Gaging stations designated by "F" were combined for analysis. species richness in the Bad River and conducted surveys where S = number of species and N = number of indi­ once per station. Doorenbos (1998), Hampton (1998), and viduals. We also calculated nonnative species richness Fryda (2001) used trap nets to improve species richness and dominance (percentage of all fish caught) for each estimates in the Belle Fourche, Cheyenne, and White sampling station. We estimated watershed area for each rivers and conducted surveys twice per sampling station, sampling station with a 30 m digital elevation model from once each in consecutive years. We combined their repli­ the National Elevation Database using the Arc Hydro cated collections for our analyses. tools version 1.0 in ArcGISTM 9.1 (ESRI® Inc.) to measure We estimated fish species diversity of each sampling river size. We used simple linear regression to determine station by calculating species richness (number of species if watershed area predicted species richness, Fisher's a, collected) and Fisher's a (Fisher et al. 1943 ; Magurran nonnative species dominance, or nonnative species rich­ 1988). Species richness is a common diversity measure, ness. but it ignores differences in species dominance (MacAr­ We conducted a hierarchical, agglomerative, polythet­ thur and MacArthur 1961) and is affected by sample size ic cluster analysis using Sprensen (Bray-Curtis) distances (Preston 1962). Fisher's a, on the other hand, represents and flexible clustering with ~ = -0.25 linkage (Legendre species of average abundance (neither highly abundant and Legendre 1998) to document variation in stream fish species nor rare species) and is unaffected by sample size assemblage composition. We scaled the cluster dendo­ (Kempton and Taylor 1974; Magurran 1988). Fisher's a is gram with Wishart's objective function that measures derived using the formula: information loss for each step in a hierarchical cluster (Wishart 1969; McCune and Grace 2002), determined the a = N(1-x) I x number of clusters in order to maximize the amount of in­ formation conserved and provide a reasonable number of where x is from iterative solution of: interpretable species groups (sensu Godinho et al. 1998; Newall and Magnuson 1999), and plotted the distribution SIN = (l-x)lx[-ln(1-x)] of stream fish assemblages on a map. We determined the 120 Great Plains Research Vol. 16 No.2, 2006 dominance of each fish species by assemblage and used simple linear regression to determine whether watershed 50 • area predicted assemblage type. :;- 40 -5 OJ 30 • RESULTS f • :0 20 • c '" • • ~ 10 River Size •• .+ • • Watershed area predicted mean discharge (N = 15, r = 10.000 20.000 30.0no 40.000 50.000 6D.noo 0.8, F = 48.4, P < 0.01, mean = 16 ± 16.8 SD), but not dis­ Watershed area (km') charge flashiness (N = 15, r = 0.1, F = 1.5, P = 0.24, mean =

0.3 ± 0.15 SD). In other words, increasing river size did not 0.6 correspond to more stable disturbance regimes (Fig. 2). 0.5 • 0.4 Fish Communities >< • ]" 0.3 , • III + ~ • • • • A total of 52 stations were sampled (Fig. 1). All fish 0.2 • • • sampling occurred in June and August 1996 through 1999, 0.1 except for three stations in the Upper Cheyenne River, upstream from Angostura Dam, which were sampled in 10,000 20,000 30,000 40,000 50,000 60,000 August 2003. A total of 33,787 fish of 36 species were Watershed area (krn2) collected (Table 1), including 11 nonnative species that Figure 2. Mean discharge (top) and environmental harshness composed 1.2% of all fishes collected. Watershed area of (R-B Index values, bottom) plotted versus watershed area for 2 U.S. Geological Survey mean daily discharge data (water years sampling stations ranged from 521 to nearly 6,300 km , 1995 through 1999). Gage locations are shown on Fig. 1. but there was no correlation between river size and fish assemblage metrics or nonnative species prevalence (Fig. 3). That is, watershed area did not predict species richness dominated assemblage II faunas (Table 1). Assemblage I (N = 52, r2 = 0.0, F = 1.5, P = 0.22, mean = 12 ± 3.5 SD), included 16 more species than assemblage II. Ten nonna­ Fisher's ex (N = 52, r2 = 0.1, F = 3.9, P = 0.05, mean = 2.3 tive species were present in Assemblage I, representing ± 0.87 SD), nonnative species dominance (N = 52, r2 = 2% of the individuals collected, whereas two norinatives 0.0, F = 0.1, P = 0.80, mean = 2 ± 2.7 SD), or nonnative were present in Assemblage II faunas, representing less species richness (N = 52, r2 = 0.1, F = 3.0, P = 0.09, mean than 1% of all individuals collected. = 1 ± 1.0 SD). Faunal division II occupied the White River, Lower In contrast, cluster analysis indicated the presence of Cheyenne River, lower Bad River, and portions of the multiple fish assemblages. Based on the cluster dendogram, Belle Fourche and Upper Cheyenne rivers (Fig. 1). Flat­ we recognized two faunal divisions and four assemblages head chub (Platygobio gracilis) and channel catfish (lc­ (Fig. 4). Assemblages were subgroups of divisions. Faunal talurus punctatus) were co-dominant in assemblage III division I (assemblages I and II) was distributed upstream and IV, but plains minnow (Hybognathus placitus) was of faunal division II (assemblages III and IV), except for also co-dominant in assemblage III (Table 1). Assemblage the presence of division II faunas upstream of the Belle IV included seven species absent from assemblage III, Fourche Dam (Fig 1). Within each faunal division, faunal whereas assemblage III included 3 species absent from assemblages were geographically intermixed. As a result, assemblage IV. Five nonnative species were present in as­ watershed area was a poor predictor of assemblage type semblage III, representing 1% of the individuals collected, (N = 52, r2 = 0.1, F = 6.6, P = 0.01). Faunal division I oc­ and six nonnative species were present in assemblage IV, cupied the Belle Fourche River below Belle Fourche Dam, representing more than 3% of all individuals collected. the Upper Cheyenne River adjacent to Angostura Dam and Reservoir, and the majority of the Bad River (Fig. DISCUSSION 1). Red shiner (CyprineZZa lutrensis lutrensis) and plains sand shiner (Notropis stramineus missuriensis) were co­ There was no relation between river size and stream dominant in assemblage I faunas, but red shiner alone fish species diversity even though the range of river sizes River Size and Fish Assemblages in Southwestern South Dakota 121

TABLE 1 FISH SPECIES DOMINANCE (PERCENT ABUNDANCE) BY ASSEMBLAGE TYPE BASED ON A CLUSTER ANALYSIS OF FAUNAL SIMILARITY (FIG.4).

Assemblageb Fish speciesa I II III IV Goldeye Hiodon alosoides 1 * * 1 Red shiner Cyprinella lutrensis lutrensis 25 71 7 4 Common carp C)prinus carpio 1 * 1 3 Western silvery minnow Hybognathus argyritis 3 - 9 3 Plains minnow Hybognathus placitus 7 6 23 8 Sturgeon chub Macrhybopsis gelida - - 1 3 Golden shiner Notemigonus crysoleucas * - - - Emerald shiner Notropis atherinoides * * 1 * Spottail shiner Notropis hudsonius * - * * Plains sand shiner Notropis stramineus missuriensis 22 12 5 4 Fathead minnow Pimephales promelas 6 1 3 * Flathead chub Platygobio gracilis 4 2 22 40 Longnose dace Rhinichthys cataractae cataractae * - * 2 Creek chub Semotilus atromaculatus * - - * Northern river carpsucker Carpiodes carpio carpio 2 1 4 1 White sucker Catostomus commersonii 2 1 * 3 Mountain sucker Catostomus platyrhynchus - - - * Shorthead redhorse Moxostoma macrolepidotum 3 * 1 3 Black bullhead Ameiurus melas 1 * * * Yellow bullhead Ameiurus natalis * - * - Channel catfish Ictalurus punctatus 9 6 23 23 Stonecat Noturus fiavus * - * 2 Northern pike Esox lucius * - * - Northern plains killifish Fundulus kansae 6 - - * Plains topminnow Fundulus sciadicus * - - - White bass Morone chrysops - - * * Green sunfish Lepomis cyanellus 3 * * * Orangespotted sunfish Lepomis humilis 3 * * - Bluegill Lepomis macrochirus macrochirus 1 - - - Smallmouth bass Micropterus dolomieu * - - * Largemouth bass Micropterus salmoides salmoides * * - - Black crappie Pomoxis nigromaculatus * - - * Yellow perch Percafiavescens * - - * Sauger Sander canadensis - - * * Walleye Sander vitreus * - - * Freshwater drum Aplodinotus grunniens * - * * a Fishes are listed in taxonomic order

b < 1.0% = * 122 Great Plains Research Vol. 16 No.2, 2006

" ~ :: , 4 'g I·· · ~ '1-··· 4- I ...... o 10 -j .... . , .. ] I :. .. E I. .. " .. . i 5 I . '". . i O~' ----~------~--~ o~------o 10.000 20.000 30.000 40,000 50.000 60.000 70,000 o 10.000 20.000 30,000 40.000 50.000 60,000 70.000 (km") 2 Watershed area Watershed area (km )

5,,------. 10

00 : I Il> : CIl I· 'u 4·~ Il> 8 Il> ' 'u ~ Il> Il> ~ I Il> ..5 3]' ;;. 6 j .~ § I c I o ~ ~ i 0 '0 2 i' c C 41 ~ i Il> .D ; ~ ...... • ... Il> 2 j § l'~ 0... : z i I· • .: i . .. o " .. o .,! '.,.'".", . ,. II o 10.000 20,000 30.000 40.000 50,000 60,000 70.000 0 10,000 20,000 30.000 40.000 50.000 60.000 70.000 Watershed area (km") Watershed area (km")

Figure 3. Fish species richness, diversity (Fisher's a), nonnative species richness, and nonnative species dominance (percent non­ native species) plotted versus watershed area. we studied was large compared to other studies that have shovelnose sturgeon (Scaphirhynchus platorynchus), documented downstream diversity increases (e.g., Hor­ paddlefish (Polyodon spathula), American eel (Anguilla witz 1978; Rahel and Hubert 1991). This supports Marsh­ rostrata), sicklefin chub (Macrhybopsis meeki), silver­ Matthews and Matthews (2000), who concluded that fish band shiner (Notropis shumardi), blue sucker (Cycleptus species diversity in streams is determined by landscape­ elongatus), blue catfish (lctalurus !urcatus), and flathead scale factors rather than local habitat features such as river catfish (Pylodictis olivaris) were historically found in size and its correlates, though the specific factors that one or more of the tributaries we studied (Hoagstrom cause this relation remain unclear. Previous researchers and Berry 2006), but were absent from the present study. suggested that harsh disturbance regimes of streams in It seems unlikely that disturbance caused their absence the Great Plains precluded longitudinal increases in spe­ because native Missouri River fishes are tolerant of harsh cies diversity (Morris 1960; Summerfelt 1967; Bramblett disturbance regimes (Hesse and Mestl 1993; Pegg et al. and Fausch 1991). Our findings support this suggestion 2003; Dieterman and Galat 2004). However, disturbance because disturbance regimes were similar throughout our may limit the distribution of nonnative fishes that are study area. established in reservoirs of arid regions by making inflow­ Nevertheless, there is an alternative hypothesis. Dams ing rivers unsuitable (Cross 1985; Gido et al. 2004; Falke and reservoirs may reduce species richness upstream by and Gido 2006), which could explain why nonnatives flooding suitable habitat and blocking dispersal routes composed a minor portion of the fish assemblages we (Luttrell et al. 1999; Wilde and Ostrand 1999; Herbert studied. Low nonnative species abundance may also result and Gelwick 2003). This could explain the lack of down­ from low human population density (Gido et al. 2004), stream species additions in the rivers we studied because but we observed the highest abundance of nonnatives in dams on the Missouri River have led to the decline of assemblage IV faunas, which were located in regions of large-river fishes (Hesse et al. 1993; Ruelle et al. 1993). low human density such as along the White River and Species such as pallid sturgeon (Scaphirhynchus albus), Lower Cheyenne River. Thus, we hypothesize that there River Size and Fish Assemblages in Southwestern South Dakota 123

Information Remaining (%) lQO 75 5p 9 BROI UCROI UCR02 UCR03------~ BR02 Assemblage I BR03 BR04 BFR03 BR05 BR08 BRIO BFR04 BFR06 Division I BFR07 ------' BFR05------~------~ UCR04 ------' BR06 BR07 BRll BR09 BRI5 BRl2 Assemblage II BRB BRl6 BRl4 BRI7 BRI8 BRI9 WR04 ------~ Assemblage III BFR09 LCR04 UCR06----..... UCR07 ~---~ WR03 UCR08----~~------~ LCR02 ------' BR20 Division II WR06 WR07 BFR08 WRII WR02 WRIO Assemblage IV WR08 WR09 WROI BFROI BFR02 UCR05 LCROI LCR03------~~------' WR05

Figure 4. Cluster dendogram summarizing fish faunal similarity among sampling stations of the Belle Fourche River (BFR), Upper Cheyenne River (UCR), Lower Cheyenne River (LCR), Bad River (BR), and White River (WR). Sampling stations are numbered in upstream to downstream order. 124 Great Plains Research Vol. 16 No.2, 2006 was a historical longitudinal increase in species diversity river species with disturbance regimes that preclude the along rivers of southwestern South Dakota, but Missouri spread of nonnative species. Hence, longitudinal increases River dams disrupted this pattern by eliminating large­ in species diversity documented in undammed river sys­ river fishes. Subsequently, harsh disturbance regimes tems of the Great Plains (Rahel and Hubert 1991; Barfoot that are characteristic of the region maintained disrupted and White 1999) may better represent historical patterns longitudinal patterns by limiting the spread of nonnative along Great Plains rivers. species. Variation in faunal composition among sampling sta­ ACKNOWLEDGMENTS tions supports Marsh-Matthews and Matthews (2000) by suggesting that local environmental factors affected fish South Dakota Game Fish and Parks supported this distributions. Faunal succession along the Bad River was work with Federal Aid in Sport Fish Restoration funds presumably related to longitudinal succession of stream (Project nos. F-21-R and F-15-R). We are particularly fish faunas from the headwaters to the mouth. However, grateful to all landowners who allowed access to sample faunal turnover was not longitudinal in the Belle Four­ on private lands. R.B. Jacobson, S.K. Sando, D.R. DeVr­ che River. Faunal division II, dominated by the riverine ies, and anonymous reviewers provided helpful sugges- species flathead chub and channel catfish, was present . tions for early versions of this manuscript. upstream of faunal division I assemblages, which were dominated by fishes typical of small to medium streams, REFERENCES the red shiner and sand shiner. This distribution was apparently caused by surface-water diversions via the Baker, D.B., R.P. Richards, T.T. Loftus, and lW. Kramer. Belle Fourche Dam, which deplete river flows and make 2004. A new flashiness index: Characteristics and the river downstream less riverine and more streamlike, applications to midwestern rivers and streams. Jour­ similar to findings elsewhere in the Great Plains (Cross nal of the American Water Resources Association and Moss 1987; Bonner and Wilde 2000; Eberle et al. 40:503-22. 2002). In any case, changing faunal composition was Barfoot, C.A., and R.G. White. 1999. Fish assemblages not a result of species additions downstream, which was and habitat relationships in a small northern Great expected and previously documented in the Great Plains, Plains stream. Prairie Naturalist 31:87-107. but was the·result of species replacements among faunas. Bonner, T.H., and G.R. Wilde. 2000. Changes in the This discrepancy is explained by the influence of dams, Canadian River fish assemblage associated with which have created geomorphic and hydrologic diversity reservoir construction. Journal of Freshwater Ecol­ that is uncharacteristic of pristine Great Plains rivers. ogy 15:189-98. Given this, we suggest that faunal division II assem­ Bramblett, R.G., and K.D. Fausch. 1991. Variable fish blages characterized the pre-dam faunas of the Belle communities and the Index of Biotic Integrity in Fourche River in South Dakota, which is supported by the a western Great Plains river. Transactions of the presence of such faunas throughout the undammed White American Fisheries Society 120:752-69. River in South Dakota. It is likely that the Upper Chey­ Caldwell, I. 1983. Bad River (Wakpa Sica): Ripples, enne River in South Dakota was historically populated by Rages, and Residents. Pierre, SD: State Publishing faunal division II as well. Fish assemblage changes down­ Company. stream and upstream of Angostura Dam are consistent Cross, F.B., and R.E. Moss. 1987. Historic changes in fish with patterns observed in other Great Plains rivers where communities and aquatic habitats in plains streams environmental conditions are periodically harsh and dams of Kansas. In Community and Evolutionary Ecol­ have disrupted fish dispersal (Wilde and Ostrand 1999; ogy of North American Stream Fishes, ed. W.l Bonner and Wilde 2000). Matthews and D.C. Heins, 155-65. University of Overall, this study supports the findings of Marsh­ Oklahoma Press, Norman, OK. Matthews and Matthews (2000) that species composition Cross, IN. 1985. Distribution offish in the Virgin River, a varies due to local environmental factors, whereas species tributary of the lower Colorado River. Environmen­ diversity varies due to large-scale factors. However, this tal Biology of Fishes 12:13-21. pattern may be a recent phenomenon in southwestern Culler, R.C. 1961. Hydrology ofStock- Water Reservoirs in South Dakota, caused by the combination of human im­ Upper Cheyenne River Basin. U.S. Geological Sur­ pacts that interrupt dispersal routes and eliminate large- vey Water-Supply Paper 1531-A, Washington, DC. River Size and Fish Assemblages in Southwestern South Dakota 125

Dieterman, D.l, and D.L. Galat. 2004. Large-scale fac­ Hoagstrom, C.W. 2003. Historical and recent fish fauna tors associated with sicklefin chub distribution in the of the Lower Pecos River. In Aquatic Fauna of the Missouri and lower Yellowstone rivers. Transactions Northern Chihuahuan Desert, ed. G.P. Garrett and of the American Fisheries Society 133:577-87. N.L. Allan, 91-110. Special Publications of the Mu­ Doorenbos, R.D. 1998. Fishes and habitat of the Belle seum of Texas Tech University 46. Fourche River, South Dakota. Master's thesis, South Hoagstrom, C.W., and C.R. Berry, Jr. 2006. Island bioge­ Dakota State University, Brookings, SD. ography of native fish faunas among Great Plains Eberle, M.E., E.G. Hargett, T.L. Wenke, and N.E. Man­ drainage basins: Basin scale features influence com­ drak. 2002. Changes in fish assemblages, Solomon position. American Fisheries Society Symposium River Basin, Kansas: habitat alterations, extirpa­ 48:221-64. tions, and introductions. Transactions of the Kansas Horwitz, R.l 1978. Temporal variability patterns and the Academy of Science 105: 178-92. distributional patterns of stream fishes. Ecological Falke, lA., and K.B. Gido. 2006. Spatial effects of res­ Monographs 48:307-2l. ervoirs on fish assemblages in Great Plains streams Kempton, R.A., and L.R. Taylor. 1974. Log-series and log­ in Kansas, USA. River Research and Applications normal parameters as diversity discriminants for the 22:55-68. Lepidoptera. Journal of Animal Ecology 43:381-99. Fisher, R.A., A.S. Corbet, and C.B. Williams. 1943. The Legendre, P., and L. Legendre. 1998. Numerical Ecology, relation between the number of species and the num­ 2nd English ed. Elsevier Science B.V., Amsterdam. ber of individuals in a random sample of an animal Lemly, A.D. 1985. Suppression of native fish populations population. Journal of Animal Ecology 12:42-58. by green sunfish in first-order streams of Piedmont Fryda, D.D. 200l. A survey of the fishes and habitat of the North Carolina. Transactions of the American Fish­ White River, South Dakota. Master's thesis, South eries Society 114:705-12. Dakota State University, Brookings. Lotrich, V.A. 1973. Growth, production, and community Gido, K.B., and lH. Brown. 1999. Invasion of North composition of fishes inhabiting a first-, second-, and American drainages by alien fish species. Freshwa­ third-order stream of eastern Kentucky. Ecological ter Biology 42:387-99. Monographs 43:377-97. Gido, K.B., IF. Schaefer, and l Pigg. 2004. Patterns of Luttrell, G.R., A.A. Echelle, W.L. Fisher, and D.l Eisen­ fish invasions in the Great Plains of North America. hour. 1999. Declining status of two species of the Biological Conservation 118:12l-3l. Macrhybopsis aestivalis complex (Teleostei: Cy­ Godinho, F.N., M.T. Ferreira, and M.l. Portugal e Castro. prinidae) in the Arkansas River basin and related 1998. Fish assemblage composition in relation to effects of reservoirs as barriers to dispersal. Copeia environmental gradients in Portuguese reservoirs. 1999:981-89. Aquatic Living Resources 11 :325-34. MacArthur, R.H., and lW. MacArthur. 1961. On bird spe­ Hampton, D.R. 1998. A survey of the fishes and habitat of cies diversity. Ecology 42:594-98. the Cheyenne River in South Dakota. Master's the­ Magurran, A.E. 1988. Ecological Diversity and Its Mea­ sis, South Dakota State University, Brookings, SD. surement. Princeton University Press, Princeton, Nl Herbert, M.E., and F.P. Gelwick. 2003. Spatial variation of Marsh-Matthews, E., and W.l Matthews. 2000. Geo­ headwater fish assemblages explained by hydrologic graphical, terrestrial and aquatic factors: Which variability and upstream effects of impoundment. most influence the structure of stream fish assem­ Copeia 2003:273-84. blages in the midwestern United States? Ecology of Hesse, L.W., and G.E. Mestl. 1993. An alternative hydro­ Freshwater Fish 9:9-2l. graph for the Missouri River based on the precontrol Matthews, W.l, and H.W. Robison. 1998. Influence of condition. North American Journal of Fisheries drainage connectivity, drainage area and regional Management 13:360-66. species richness on fishes of the interior highlands in Hesse, L.W., G.E. Mestl, and lW. Robinson. 1993. Status of Arkansas. American Midland Naturalist 139:1-19. selected fishes in the Missouri River in Nebraska with McCune, B., and lB. Grace. 2002. Analysis of Ecologi­ recommendations for their recovery. In Proceedings of cal Communities. MjM Software Design, Glenden the Symposium on Restoration Planning for the Rivers Beach, OR. of the Mississippi River Ecosystem, 327-40. U.S. Na­ Milewski, c.L. 200l. Local and systemic controls on fish tional Biological Survey, Biological Report 19. and fish habitat in South Dakota rivers and streams: 126 Great Plains Research Vol. 16 No.2, 2006

Implications for management. PhD diss., South Da­ Sando, S.K. 1991. Estimation and Characterization of the kota State University, Brookings. Natural Streamflow of the White River Near the Ne­ Montgomery, D.R. 1999. Process domains and the braska-South Dakota State Line. U.S. Geological Sur­ river continuum. Journal of the American Water vey Water-Resources Investigations Report 91-4096. Resources Association 35:397-410. Sando, S.K., lE. Williamson, K.K. Dickerson, and E.A. Morris, L.A. 1960. The distribution of fish in the Platte Wesolowski. 200l. Irrigation Drainage Studies River, Nebraska. Master's thesis, University of Mis­ of the Angostura Reclamation Unit and the Belle souri, Columbia, MO. Fourche Reclamation Project, Western South Da­ Newall, P.R., and II Magnuson. 1999. The importance of kota: Results of 1994 Sampling and Comparisons ecoregion versus drainage area on fish distributions with 1988 Data. U.S. Geological Survey Water-Re­ in the St. Croix River and its Wisconsin tributaries. sources Investigations Report 01-4103. Environmental Biology of Fishes 55:245-54. Schlosser, 1.l 1982. Fish community structure and func­ Pegg, M.A., e.L. Pierce, and A. Roy. 2003. Hydrological tion along two habitat gradients in a headwater alteration along the Missouri River Basin: A time stream. Ecological Monographs 52:395-414. series approach. Aquatic Sciences 65:63-72. Sokal, R.R., and Fl Rohlf. 1995. Biometry, the Principles Preston, FW. 1962. The canonical distribution of com­ and Practice of Statistics in Biological Research, monness and rarity: Part 1. Ecology 43:185-215. 3rd ed. W.H. Freeman, New York. Rahel, Fl, and W.A. Hubert. 1991. Fish assemblage and Summerfelt, R.C. 1967. Fishes of the Smoky Hill River, habitat gradients in a Rocky Mountain-Great Plains Kansas. Transactions of the Kansas Academy of stream: Biotic zonation and additive patterns of Science 70:102-39. community change. Transactions of the American Thornbury, W.D. 1965. Regional Geomorphology of the Fisheries Society 120:319-32. United States. John Wiley and Sons, New York. Ricklefs, R.E. 1987. Community diversity: Relative roles of Walters, D.M., D.S. Leigh, M.e. Freeman, B.J. Free­ local and regional processes. Science 235: 167-7l. man, and e.M. Pringle. 2003. Geomorphology and Riley, M.P., W.F Kumlien, and D. Tucker. 1955.50 Years' fish assemblages in a Piedmont river basin, U.S.A. Experience on the Belle Fourche Irrigation Project. Freshwater Biology 48:1950-70. South Dakota State College, Agricultural Experi­ Wilde, G.R., and K.G. Ostrand. 1999. Changes in the fish ment Station, Bulletin 450. assemblage of an intermittent prairie stream up­ Ruelle, R., R. Koth, and e. Stone. 1993. Contaminants, stream from a Texas impoundment. Texas Journal fish, and hydrology of the Missouri River and west­ of Science 51:203-10. ern tributaries, South Dakota. In Proceedings of the Wishart, D. 1969. An algorithm for hierarchical classifica­ Symposium on Restoration Planning for the Rivers tions. Biometrics 25:165-70. of the Mississippi River Ecosystem, 449-80. U.S. National Biological Survey, Biological Report 19. Manuscript received for review, April 2006; accepted for publication, August 2006.