THE EFFECTS OF A MODIFIED FLOW REGIME ON FISH ASSEMBLAGES IN A THIRD-ORDER PIEDMONT STREAM IN LEE COUNTY, AUBURN, ALABAMA Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. __________________________________ James Edmund Gleason Certificate of Approval: E. Cliff Webber, Co-Chair David R. Bayne, Co-Chair Research Fellow Professor Fisheries and Allied Aquacultures Fisheries and Allied Aquacultures Jack W. Feminella Joseph F. Pittman Professor Interim Dean Biological Sciences Graduate School THE EFFECTS OF A MODIFIED FLOW REGIME ON FISH ASSEMBLAGES IN A THIRD-ORDER PIEDMONT STREAM IN LEE COUNTY, AUBURN, ALABAMA James Edmund Gleason A Thesis Submitted to the Graduate Facility of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama December 17, 2007 iii THE EFFECTS OF A MODIFIED FLOW REGIME ON FISH ASSEMBLAGES IN A THIRD-ORDER PIEDMONT STREAM IN LEE COUNTY, AUBURN, ALABAMA James Edmund Gleason Permission is granted to Auburn University to make copies of this thesis at its discretion, upon the request of individuals or institutions and at their expense. The author reserves all publication rights. _______________________ Signature of Author _______________________ Date of Graduation iv THESIS ABSTRACT THE EFFECTS OF A MODIFIED FLOW REGIME ON FISH ASSEMBLAGES IN A THIRD ORDER PIEDMONT STREAM IN LEE COUNTY, AUBURN, ALABAMA James Edmund Gleason Master of Science, December 17, 2007 (B.S., University of Central Florida, 2003) 134 Typed Pages Co-Directed by E. Cliff Webber and David R. Bayne I examined the effects of a modified flow regime on fish assemblage condition over a 2-year period at two sites in Chewacla Creek, a third-order Piedmont stream (Lee County, AL). My goal was to assess the biotic condition of fish assemblages in the creek impacted by dewatered conditions that occurred prior to 2003 as a result of drought, no water releases from Lake Ogletree, and quarry mining in the watershed. The two Chewacla sites were compared with a relatively undisturbed reference site, Cane Creek (Cane-1), another Piedmont stream in Lee County. Water quality and fishes were sampled once per year at all three sites during low-flow conditions in the fall of 2004 and 2005. In addition, a physical habitat assessment at each site was conducted in the fall of 2006. The principal method used in this study for comparison of fish assemblages across v sites was the Index of Biotic Integrity (IBI). Differences across sites in water quality, physical habitat quality and fish community health were compared. Results showed few differences across sites in water quality and physical habitat once continuous flows were reestablished in the study reach in December of 2003. IBI ratings for Cane-1were good to excellent; thus, it compared well with other Piedmont reference streams in terms of fish assemblage condition. In contrast, IBI ratings for the two Chewacla Creek sites (CH-U and CH-D) rated poor to fair, indicating that fish assemblages at both sites were degraded and did not meet fish assemblage conditions present at Cane-1 in either 2004 or 2005. At CH-U, IBI ratings were poor both years suggesting no recovery of the fish assemblage from altered flow conditions that occurred periodically between 1999 and 2001. At CH-D, IBI ratings changed from poor to marginally fair in 2005 suggesting slight recovery of the fish assemblage from periodic dewatering of the stream channel that occurred at the site. Differences in fish assemblage condition between Cane-1, CH-U, and CH-D were mostly attributable to differences in flow regime. Although water quality and physical habitat influenced the fish assemblages at CH-U and CH-D, a highly modified flow regime resulting from drought in previous years, mining in the watershed, and the presence of Lake Ogletree appeared to be most influential in structuring the fish assemblages at both sites during this study. vi ACKNOWLEDGEMENTS I would like to thank Dr. Cliff Webber, Dr. David Bayne, Wendy Seesock and Eric Reutebuch for their outstanding guidance and support during the past several years. Also, I wish to thank Dr. Jack Feminella for serving on my graduate committee and providing advice throughout the study. All have generously shared their time and knowledge and have contributed greatly to my education and training during my tenure at Auburn University. Special thanks to fellow students Jason Dickey, Robbie McKinney, David Saalfield, Daniel Smith, Andy Myrick, Jonathan Jackson and Nathan Brown for their participation in field and laboratory work for this study ? I am fortunate to have worked with such an outstanding and dedicated group of individuals. Finally, I express my deepest gratitude to my parents Edmund and Bebe Jo, my brother Andrew, and my sister JoBeth for their love, encouragement and support throughout many years of study. vii Style manual or journal used: Transactions of the American Fisheries Society Computer software used: Microsoft Word, Excel 2002 for Windows, JMP Version 4.04 Statistical Software (SAS Institute, Inc.) viii TABLE OF CONTENTS LIST OF TABLES .................................................................................................... ix LIST OF FIGURES ................................................................................................... xi I. INTRODUCTION ......................................................................................... 1 II. LITERATURE REVIEW .............................................................................. 6 III. MATERIALS AND METHODS................................................................... 29 Description of study sites............................................................................... 30 Assessment of physical habitat ...................................................................... 32 Measurement of discharge ............................................................................ 33 Water chemistry sampling ............................................................................. 33 Fish sampling ?????............................................................................ 34 Index of Biotic Integrity (IBI) ????.. .................................................... 38 Index of Well-Being (Iwb)???????????????????. 48 Data Analysis????????????????????????.. 48 IV. RESULTS AND DISCUSSION .................................................................... 50 Characterization of habitat quality in Chewacla and Cane creeks ................ 50 Characterization of flow regime in Chewacla Creek..................................... 53 Characterization of water chemistry in Chewacla and Cane creeks.............. 55 Fish data collected across sites????????????????? .. 58 IBI and Iwb .................................................................................................... 58 Analysis of fish assemblage structure across sites......................................... 70 Analysis of fish assemblage function across sites??????????. 75 Summary?????????????????????????? 80 LITERATURE CITED???????????????????? .............. 105 ix LIST OF TABLES Table 1 Metrics developed by Karr (1981) to assess fish communities in streams located in the midwestern United States............................... 82 Table 2 Candidate IBI metrics evaluated for use in Chewacla and Cane creeks, Lee County, Alabama ............................................................ 83 Table 3 Physical characteristics measured at Chewacla and Cane creek sites, Lee County, Alabama ............................................................... 85 Table 4 Metrics used to assess physical habitat quality at Chewacla and Cane creek sites, Lee County, Alabama ............................................ 87 Table 5 Scoring criteria used to visually estimate physical habitat quality at sites in the Chewacla and Cane creek watersheds during the fall 2006 ............................................................................................. 88 Table 6 Assessment used to evaluate physical habitat quality at all sites in Chewacla and Cane creeks during the fall 2006............................ 89 Table 7 Index of biotic integrity (IBI) metrics and scoring criteria developed for Chewacla and Cane creek sites, Lee County, Alabama???????????????????????. 91 Table 8 Iwb scoring criteria and integrity classes for wadeable Piedmont streams ............................................................................................... 92 Table 9 Physical habitat assessment scores for sites in the Chewacla and Cane creek watersheds, fall 2006 ...................................................... 93 Table 10 Physicochemical data measured at sites in the Chewacla and Cane creek watersheds during the fall of 2004 and 2005 ........................... 95 Table 11 Total abundance, percent relative abundance and total number of fishes collected by family at sites in the Chewacla and Cane creek watersheds, 2004 and 2005................................................................ 96 x Table 12 Fish collection data for all sites sampled in Chewacla and Cane Creeks, 2004 and 2005....................................................................... 97 Table 13 IBI metrics, raw metric scores and total IBI scores for sites in Chewacla and Cane creeks, 2004 and 2005................................... 99 Table 14 IBI and Iwb scores with associated integrity classes for all sites in Chewacla and Cane creeks, 2004 and 2005?????????? 100 Table 15 Core IBI metrics and F-test results between raw metric values across sites ......................................................................................... 101 Table 16 Descriptive measures of community structure for sites in Chewacla and Cane creeks, 2004 and 2005????????....... 102 xi LIST OF FIGURES Figure 1 Map of Chewacla Creek and its tributaries showing locations of sampling stations, Lake Ogletree, and Martin Marietta quarry for physical habitat assessment, discharge measurement, water quality sampling, and fish sampling .................................................................... 84 Figure 2 Map of Cane Creek showing sampling locality for physical habitat assessment, water quality sampling, and fish sampling........................... 86 Figure 3 Map showing locations of stream gages for measuring flow operated by AU Department of Fisheries ............................................................... 90 Figure 4 Hydrographs showing mean daily discharge for the upstream and downstream Chewacla Creek sites from August 2004 to April 2006...... 94 Figure 5 Feeding guild composition ? percent total number of fishes collected by feeding guild at sites in Chewacla and Cane creeks, 2004 and 2005 ...... 103 Figure 6 Tolerance class composition ? percent total number of fishes collected by tolerance class at sites in Chewacla and Cane creeks, 2004 and 2005.......................................................................................................... 104 1 I. INTRODUCTION Most lotic waters in the world have been impacted by human activities, resulting in modification of their physical, chemical, and biological condition (Allan 1995). Humans have altered natural flow regimes by building dams, channelizing rivers and diverting water for flood control, navigation, irrigation, industrial and municipal water supply, and hydropower. Furthermore, major changes in the surrounding landscape, such as urbanization, agricultural practices, and deforestation have altered natural flow regimes resulting in degraded ecological integrity of lotic ecosystems (Mulholland and Lenat 1992). Mining is a major anthropogenic factor affecting streams and rivers and its impact on streams and riparian areas in the United States and other regions of the world has been severe (Allan 1995). Within Alabama, mining mineral deposits for crushed stone, sand and gravel in or adjacent to streams is a common practice. For example, in Lee County quarry mining has been practiced intermittently since the early 1900s. Mining activities can have a number of detrimental impacts on streams including degrading water quality from increased sediment load (Kondolf 1994; Waters 1995), increasing stream bed bank erosion (Bull and Scott 1974; Sandecki 1989; Kondolf 1994; Pond 2004), and altering natural flow regimes and instream habitat through flow impediments; each of these activities can reduce overall biological diversity and production (Benke 1990; Lyttle 1993; Waters 1995; Brown et al. 1998). 2 Chewacla Creek, a third-order Piedmont stream in East Central Alabama, has been the subject of much interest in recent years, in particular the 1.5 mile reach between Lake Ogletree and Chewacla State Park in Lee County. This reach of Chewacla Creek is a regulated stream because of the presence of Lake Ogletree. Portions of the reach downstream of Lake Ogletree, the primary potable water supply for the city of Auburn, have been directly impacted by operation of a crushed rock quarry adjacent to the creek channel. Mining operations at the quarry resulted in sinkholes in reaches of the stream channel near the quarry (United States Fish and Wildlife Service 2003). In addition, periodic drought conditions since 1999 resulted in low lake levels and no water flow over the spillway into the downstream channel (USFWS 2003). Thus, the only water input from Lake Ogletree to the downstream channel during these dry periods was seepage below the base of the dam; however, this input of water was usually insufficient to maintain constant flows in the creek between the lake and the quarry. The combination of sinkholes and no outflow of water from the lake has resulted in dewatered reaches adjacent to the quarry during drought (Rex Griffin, Water and Sewer Director, personal communication). Furthermore, on several occasions between 1999 and 2001 the creekbed was so dry that only small, isolated pools were present (Cliff Webber, AU Fisheries, personal communication). In 2003 legal action by local citizens resulted in a Chewacla Creek Safe Harbor Agreement between the citizens, the City of Auburn Water Board, Martin Marietta Materials, Inc., the Alabama Department of Conservation and Natural Resources (ADCNR), and the U.S. Fish and Wildlife Service. The agreement provided for establishment of minimum flows of 2.0 million gallons per day (MGD) in the stream just 3 below the dam on Lake Ogtletree. Three requirements were included in the Chewacla Creek Safe Harbor Agreement (2003) to ensure that a flow of 2.0 MGD was maintained in the impacted reach. First, the City of Auburn Water Board (now the Water Resources Management Department) was required to siphon water from the lake at a depth of 1.5 m (5 ft) below the water surface to maintain a minimum flow of 2.0 MGD at the base of Lake Ogletree dam. Second, Martin Marietta was required to monitor and repair sinkholes in the channel near the quarry so that the minimum flow of 2.0 MGD released from the lake was maintained downstream. Third, Martin Marietta could be asked by the Water Board to pump water from the quarry pit to the lake to replenish water siphoned from the lake to maintain the minimum 2.0 MGD flow at the base of Lake Ogletree dam. Pumping water from the quarry pits to the lake rather than directly into Chewacla Creek would allow the settling of any sediment present in the quarry water and a moderation of pH levels. To implement these measures, Martin Marietta and the City of Auburn Water Board were required to monitor flow conditions continuously and conduct annual bioassessments in the impacted reach downstream of Lake Ogletree. Concern for the survival of 3 threatened species of mussels listed in 1993 by the U.S. Fish and Wildlife Service: the fine-lined pocketbook (Lampsilis altilis), southern clubshell (Pleurobema decisum) and ovate clubshell (Pleurobema perovatum) prompted a study of the stream and assessment of the mussel assemblage and of fish their host species, the largemouth bass (Micropterus salmoides). Females of these species release glochidia which are free-living parasitic larvae that attach to the gills of host fish (Wetzel 2001). Glochidia remain on the gills for several weeks where they metamorphose into juvenile mussels and then drop off the host and settle within the substrate (Wetzel 2001). 4 The purpose of the Safe Harbor Agreement (2003) was to provide constant flows in the previously dewatered reach of Chewacla Creek in hopes that the fine-lined pocketbook mussel would repopulate all sections of the reach. In addition, migration of largemouth basses into previously dewatered sections would increase the likelihood that southern and ovate clubshell mussels would repopulate the impacted reach. Since 2003 cooperative agreements have been in place between AU Fisheries and the City of Auburn Water Resources Department to monitor water quality and stream flow and conduct annual bioassessments of the macroinvertebrate and mussel assemblages in the impacted reach. The Safe Harbor Agreement (2003) secured minimum flow and subsidence control requirements in sections of Chewacla Creek to enhance and expand available instream habitat for the three threatened mussel species, but it did not require assessment of the fish assemblage in the reach downstream from Lake Ogletree and adjacent to the quarry. This lack of monitoring of the fish assemblage left a critical gap in the assessment of the aquatic biota in Chewacla Creek as it recovered from the dewatering of the stream channel that occurred between 1999 and 2001 (Webber and Blevins 2000; 2001). In 2004, funding was obtained by AU Fisheries for a survey of the fish assemblages at two sites in Chewacla Creek through a United States Environmental Protection Agency (US EPA) grant for stream restoration. This survey included sampling of the fish fauna at 2 sites in Chewacla Creek, one just downstream of Lake Ogletree and the other further downstream adjacent to the quarry, to examine the impact of an altered flow regime on the resident fish assemblages. In addition, the fish fauna was sampled concurrently in a reference stream during the same 2-year period in Cane Creek, an unregulated third-order Piedmont stream. 5 This 2-year study allowed me to examine the extent of recovery of the fish assemblages at two sites in Chewacla Creek from an altered flow regime that occurred between 1999 and 2001 after consistent flows were reestablished in the reach downstream of Lake Ogletree. The main objective of this study was to compare fish assemblage condition in a reach recovering from altered flows with one exhibiting a natural flow regime. Specific questions were related to comparisons between Chewacla Creek and reference sites in terms of (1) differences in abundance, species richness and diversity, (2) altered trophic structure and function with an increase in tolerant species, and (3) an increase in dominance of certain species with lower evenness in Chewacla Creek relative to the reference site. 6 II. LITERATURE REVIEW Stream velocity, depth, substrate, and instream cover are considered the most important physical habitat characteristics affecting distribution and abundance of fish species (Angermeier 1989). Stream alterations from human activity often disturb physical habitats, changing fish assemblage structure upstream, downstream, and within disturbed areas (Nunnally 1978; Schlosser 1990; Aadland 1993). Managing the integrity of stream water resources requires the characterization and careful analysis of physical, chemical, and biological components. Thus, resource managers often must quantify these changes in stream ecosystems, and distinguish between changes occurring naturally and those caused by anthropogenic impacts. Frequent, long-term, and continuing anthropogenic impacts are inevitable in Piedmont streams in the southeastern United States because of continued landscape alteration associated with human population growth (Mulholland and Lenat 1992). However, stronger regulatory efforts by state and federal agencies utilizing sound biological principles can potentially offset the effects of physical habitat degradation on instream biological communities. Habitat degradation often results in changes in resident fish assemblages (Karr 1981; Schlosser 1990); thus, careful analysis of fish assemblage characteristics relative to the levels of disturbance can provide benchmarks guiding future habitat improvement programs for Piedmont streams. 7 Mining Impacts on Aquatic Biota One of the most highly visible forms of land use change in Alabama is surface mining, in particular crushed rock mining (Alabama Department of Industrial Relations 2003). Crushed rock (aggregate) is one of the most important and highly demanded resources in the United States, having uses in nearly all commercial, industrial, and residential construction (Langer and Glanzman 1993). Aggregates are produced from sand and mineral deposits or from bedrock sources (Poulin et al. 2004). Behind fuel mineral mining, aggregate (non-fuel) mining is the second largest mining industry in the United States (Bull and Scott, 1974; Waters 1995). The amount of aggregate produced annually in the United States increased from a modest 58 million tons mined in 1980 to 2.3 billion tons in 1996 (Bolen 1997; Tepordei 1997). In Alabama, mining for aggregate is currently practiced in 21 counties, including Lee County (United States Geological Survey 2004). About 49,100 metric tons of crushed rock was produced in the state in 2004 with an economic value of about $290 million dolars (USGS 2004). One of the most common methods of aggregate extraction in the United States is open pit (quarry) mining in the flood plains and terraces of small and medium-sized streams (Roell 1999). The impact of surface mining operations such as quarry mining on the surrounding landscape, stream hydrology and geomorphology, and aquatic communities can be profound (Starnes and Gasper 1996). Surface mining, particularly in the flood plains of streams, can cause temporary elimination of surface vegetation and alteration of surface topography (Starnes and Gasper 1996). Furthermore, mining can alter subsurface geologic structure and subsurface hydrologic regimes. The deep quarry pits that form often result in subsidence features (cracks, fissures, and sinkholes) in 8 stream channels that can dewater reaches and divert flows (Starnes and Gasper 1996). Degraded water quality in aquatic systems impacted by aggregate mining usually occurs as a result of wastewater discharges into streams that increase concentrations of suspended solids as well as sedimentation in pools and riffles (Gammon 1970). Typically, heavy sedimentation and degraded stream channels are common in areas in which rock quarries occur (Roell 1999). Furthermore, streams located near rock quarries often exhibit altered flow regimes and loss of critical habitat (Roell 1999). Such negative effects can result in loss of sensitive fish species and ecological integrity. Changes may occur in the structure and function of fish assemblages through mining-related effects on growth, behavior, migration, and reproduction (Gammon 1970; Ahle and Jobsis 1996; Bell and Payne 1993). In particular, these effects on stream fishes may include reduced feeding efficiency and, thus, reduced growth, respiratory impairment, reduced tolerance to diseases and toxicants, and increased physiological stress (Waters 1995). Additionally, inorganic sediment and alterations in flow regime as a result of mining activities can reduce the reproductive success of certain warmwater fishes, in particular those requiring clean stony sites for spawning (Berkman and Rabeni 1987). Streams near heavily mined areas often show lower benthic macroinvertebrate diversity, abundance, and productivity, which, in turn, reduce food availability for foraging fishes (Waters 1995). The impact of aggregate mining on freshwater fish assemblages and other instream biotic communities in Alabama is not well understood because < 50% of surface mining activity is inspected in the state while mining is in progress (Alabama Department of Industrial Relations 2003). Relationship between Flow Regime and Stream Fish Assemblage 9 The effects of altered flow regimes on streams can include changes in physical characteristics such as sediment transport, channel depth and width, current velocity and substrate composition. These physical changes, in turn, can alter both the structure and function of fish assemblages (Ward and Stanford 1983; Bain et al. 1988; Ligon et al. 1995; Poff et al. 1997; Marchetti and Moyle 2001). Drought and other unpredictable environmental disturbances commonly alter many aspects of stream systems, including water chemistry, stream size, water temperature, streambed structure and substrate, and flow (Taylor et al. 1996; Medeiros and Maltchik 2001; Matthew and Matthews 2003). Such environmental variation can dramatically alter the living conditions and aquatic habitats within the water, affecting much of the aquatic fauna inhabiting streams (Keaton et al. 2005). Several studies have examined how flow variability influences fish community structure and function in streams. The effects of low flows on the distribution of stream fishes have been examined following drought (Larimore et al. 1959; Deacon 1961). Larimore et al. (1959) found that variable flows in Illinois streams as a result of severe drought (created suboptimal habitat conditions) that led to decreased survival of fishes during early life stages or physiological stresses on adults. However, they found that most fish assemblages rapidly reestablished pre-drought populations, noting most species returned to previously dewatered sites. Only 3 species remained absent from sites 3 years after normal flows had returned: the longear sunfish Lepomis megalotis, blackstripe topminnow Fundulus notatus and hornyhead chub Nocomis biguttatus (Larimore et al. 1959). Adult fish reentering a previously dewatered stream reproduced rapidly, resulting in populations dominated by young-of-year individuals. 10 Deacon (1961) documented the effects of drought on the fishes of the Marais des Cygnes and Neosho rivers, Kansas, noting reductions in some species during the drought and expansions of others in post-drought years. He found that the channel catfish Ictalurus punctatus, Neosho madtom Noturus placidus and two darter species, the fantail darter Etheostoma flabellare and the orangethroat darter Etheostoma spectabile rapidly declined in abundance during drought years and did not fully recover to pre-drought numbers. In contrast, Deacon found that that spotted bass Micropterus punctulatus and several Lepomis species, including the green sunfish Lepomis cyanellus increased in numbers during drought years in the two rivers. These habitat generalists quickly repopulated reaches that had been previously dewatered once normal flows had resumed. Species that occupied restricted habitats, such as the Neosho madtom, fantail darter and orangethroat darter, were slow to increase in numbers post-drought. As a result, these species disappeared or became restricted to a single tributary in the stream system (Deacon 1961). Gorman and Karr (1978) studied the impact of flow modification (as channelization) on fish assemblages in Indiana and Panama streams. There, streams with a more natural flow regime supported fish assemblages with high species diversity and greater seasonal stability than streams with altered flow regimes. The authors concluded that most fishes were habitat specialists and that their presence or absence in a particular stream was dependent on the degree of habitat heterogeneity and flow stability in a stream (Gorman and Karr 1978). As a test of the resiliency of fish assemblages, Meffe and Sheldon (1990) analyzed fish assemblage structure at 37 sites in South Carolina streams before and one 11 year after defaunation resulting from drought. The authors noted rapid post-defaunation recovery with little change in richness, density, and biomass between the pre- and post- defaunation period. They concluded that assemblages were not randomly structured units, but instead were highly correlated with local habitat structure including depth, width, current velocity, substrate, and type and availability of cover (Meffe and Sheldon 1990). Poff and Allan (1995) studied the relationship of flow variability on the organization of stream fish assemblages using data from 34 sites in Minnesota and Wisconsin for each of 106 species to test the premise that hydrological factors (water depth, current velocity, food availability, and thermal regime) were significant determinants of assemblage structure. Findings indicated that unstable sites with highly altered flow regimes supported resource generalist species classified as resource generalists where sites with more natural flow regimes were characterized by a higher proportion of specialist species. The authors concluded that hydrological variability induced by climate change could modify stream fish assemblages in the region. Spence et al. (1999) assessed the health and condition of fish assemblages in several Oklahoma streams during a drought year in 1998. During the summer drought, the authors noted that highly variable flows from local drought resulted in poor habitat quality (dewatered riffles and runs and increased isolated pools). The authors used the Index of Biotic Integrity (IBI) to examine the condition of fish assemblages in streams impacted by the drought. The authors observed high correlation between IBI scores and flow-dependent parameters including pool variability and the presence of rocky runs and riffles. They noted that drought-impacted streams characterized by low flows, high pool 12 variability, and little to no riffle or run habitat had much lower IBI scores than reference streams relatively unimpacted by the summer drought. Marchetti and Moyle (2001) studied the effects of flow regime on fishes in Putah Creek, a highly regulated stream in central California, as a result of water diversion over the last 40 years. Because of the combination of extreme dry and extreme wet years that were encountered over the 5-year study, the authors were able to test the hypothesis that a more natural flow regime favors native fishes and suppresses exotic fishes because climatic conditions abruptly changed stream flow from conditions characteristic of a dammed stream to conditions similar to a natural flow regime. The authors noted higher numbers of native fishes during wet years with large peak flows in winter and sustained flows during the summer. In contrast, during extremely dry years, exotic species were favored when winter flows were lower and summer flows were intermittent. The authors concluded that higher flows during wet years flushed many exotic species from the stream while simultaneously creating conditions that favored reproduction by native species (Marchetti and Moyle 2001). This study concluded that restoration of natural flow regimes is important for reversing the decline of native fish populations in streams of the western United States. This study illustrated the need for examining the recovery of fish assemblages following restoration of natural flow regimes in artificially dewatered streams, such as in the Chewacla Creek stream reach that comprised my study. Keaton et al. (2005) studied the effects of drought on fish assemblages in selected streams in the Piedmont ecoregion of South Carolina. Fishes were sampled during drought conditions in 2000 and then again during post-drought conditions in 2003. Between the two time periods the authors observed many more juveniles and young of 13 the year (YOY) in the 2003 (vs. 2000) collections, which suggested that most species exhibited greater reproductive success following the drought. The authors also noted higher variability in common measures of assemblage structure, including abundance, richness, and diversity in collections from 2000 than in 2003. At all sites fish assemblage structure was highly correlated with instream habitat structure, with sites that had greater discharge and deeper water exhibiting more habitat heterogeneity and greater stability in fish community structure (Keaton et al. 2005). Gelwick et al. (2001) found that instream habitat structure was a good predictor of fish assemblage condition, so habitat instability associated with variation in stream flow will adversely impact the fish assemblage residing within it. For example, low flows during drought can limit habitat resources and fish mobility and kill fishes directly, either by desiccation or predation (Lohr and Fausch 1997). Furthermore, reproduction and juvenile recruitment can be negatively impacted by environmental stress associated with drought (Freeman et al. 1998; Schlosser et al 2000; Keaton et al. 2005). Altered natural flow regimes and their impact on fish assemblages in streams resulting from impoundment, channelization, urbanization, water diversion, agricultural practices, clearcutting for timber, and construction have been extensively studied. Mining, another highly visible anthropogenic activity, has degraded stream quality in the US and many parts of the world. Research has focused mostly on water quality impacts from mining activities on streams and resident fish communities, such as sedimentation, erosion, altered water chemistry (Gammon 1970; Knight and Newton 1977; Harkins et al. 1980) and decreased stream water pH (Leonard and Orth 1986; Robb 1994; Ahle and Jobsis 1996; Henry et al. 1999; Sams and Beer 2000; Soucek 2001). However, very few 14 studies have been conducted on how mining activities disrupt natural flow regimes in streams and the degree to which this impact aquatic biota (Starnes and Gasper 1996). Thus, my study of how fish assemblages have been impacted by altered stream flows in Chewacla Creek between Lake Ogletree and the quarry was unique because it examined several factors that have contributed to the modified flow regime in the reach: (1) the influence of Lake Ogletree, (2) drought conditions occurring between 1999 and 2001, and (3) sinkhole activity in the stream channel near the quarry as a result of mining. Biomonitoring The Water Pollution Control Act was implemented in 1948 to protect aquatic systems in the United States (Dauwalter et al. 2003). Now known as the Clean Water Act after 1972 amendments, this statute mandates state and federal agencies to monitor US surface water conditions and restore and maintain the biological integrity of the nation?s water resources (Karr 1991; Shaner 1999). Historically, monitoring of US surface waters focused primarily on point-source effluents and setting criteria levels for chemical contaminants (Bowen 1996; Shaner 1999). For many years, it was believed that monitoring point-source discharges was sufficient to protect water resources and, thus, provided water regulatory agencies with a means to satisfy the directives of the Clean Water Act. Thus, assessment of water resources in the US historically emphasized non- biological variables, specifically physical and chemical water quality, because it was presumed that improvements in the physical and chemical water quality would restore biological integrity (Karr 1993; Shaner 1999). Implementation of effluent regulatory programs for point-source discharges helped to maintain adequate water quality, but this approach was insufficient to protect a 15 variety of aquatic resources, particularly fishes, from non-point source impacts (Karr 1991). In addition, habitat alteration, flow regime modification, and changes in the energy base for stream biota are detrimental impacts not detected by point-source effluent monitoring programs. A need existed for a more comprehensive approach to monitoring water quality in response to the continued decline in biological integrity of aquatic resources. This prompted some regulatory agencies to integrate a biological approach, or biomonitoring, into their water quality monitoring programs (Karr 1991; Shaner 1999). Karr (1999) defined biomonitoring as the evaluation of the health of a biological system to assess degradation to aquatic ecosystems from a multitude of human impacts. Biomonitoring was based on the direct observation of aquatic communities, for which traditional physical and chemical water quality monitoring programs generally have proven to be unreliable alternatives (Shaner 1999, 2001). To address the directives of the Clean Water Act, biological assessment methods have incorporated a variety of taxonomic groups including algae, benthic macroinvertebrates, and fishes, all of which reflect water quality through the structure, composition, and functional relationships of their communities (Karr 1981; Fausch et al. 1984; Angermeier and Karr 1986; Saylor and Scott 1987; Hughes 1990; Shields et al. 1995; Yoder and Rankin 1998; Roth et al. 1998; Barbour et al. 1999). Fishes are especially useful in assessing environmental impacts resulting from anthropogenic activities because they occupy multiple trophic levels in aquatic communities, are relatively long-lived, are sensitive to a variety of stressors, and are easily sampled (Karr et al. 1986; Helms et al. 2005). Furthermore, fishes are an important, visible component 16 of any aquatic ecosystem and, as such, the condition of the fish assemblage can be easily communicated to the public. Biomonitoring Tools Index of Biotic Integrity (IBI). Monitoring and managing the integrity of water resources can be a cumbersome and time-consuming process, often lacking a suitable framework for the analysis and interpretation of biological data (Weber 1981). A practical method of assessing stream quality using fish assemblages is the index of biotic integrity, first introduced by Karr (IBI, 1981). The IBI is widely used to rapidly, efficiently, and directly assess the quality of water resources based on the condition of the fish community (O?Neil et al. 2006). The IBI originally developed by Karr (1981) consisted of 12 metrics that assessed 3 attributes of the fish assemblage: species richness and composition, trophic composition and dynamics, and fish abundance and condition. Each of the 12 metrics was scored by comparing its value to expected values determined from regional reference sites (Karr et al. 1984). A reference site was considered a stream of similar size in the same ecoregion that remained relatively undisturbed by human impacts (Shaner 1999). An ecoregion is a large area of land and water that contains a geographically distinct assemblage of natural communities and is defined primarily by similar landforms, climate and vegetation (Omernick 1987). The metrics were scored based on whether they approximated (5), deviated somewhat (3), or deviated strongly (1) from the values measured from the reference site (Karr et al. 1986; Shaner 1999, 2001), and metric scores were summed to generate a total IBI score. Based on the total IBI score, a stream was assigned into an integrity class 17 (excellent, good, fair, poor, very poor, and no fish), which allowed for easy interpretation of the condition of the instream fish assemblages (Karr et al. 1984; Shaner 1999). Table 1 lists the original metrics suggested by Karr et al. (1984) for streams in the midwestern US. The IBI offers a broad range of application when added to any water quality monitoring program. It can be used to evaluate current conditions at a site and identify local sources of disturbance, detect temporal trends at a site sampled repeatedly, and to compare fish assemblages across multiple sites (Karr et al. 1986; O?Neil et al. 2006). In addition to evaluating fish assemblages, IBIs have been developed for other aquatic organisms including macroinvertebrates (Kerans and Karr 1994; Resh et al. 1995; Barbour et al. 1999; Maxted et al. 2000; Klemm et al. 2003) and periphyton (Barbour et al. 1999; Hill et al. 2000). Past studies have shown the IBI to be an effective tool in identifying many types of environmental degradation in aquatic systems, including: (1) wastewater treatment effluents (Karr et al. 1985; Hughes and Gammon 1987), (2) flow modification (Bowen et al., 1996), (3) urbanization and riparian zone disturbance (Steedman 1988; Schleiger 2000, Helms et al. 2005), (4) agricultural and construction practices (Karr et al. 1987; Crumby et al. 1990; Rabeni and Smale 1995; Frenzel and Swanson 1996), sedimentation impacts (Walters et al. 2001), and (6) acid mine drainage (Leonard and Orth 1986; Ahle and Jobsis 1996). Since Karr (1981) originally proposed the IBI, improvements in IBI procedures have been made. Guidelines for identifying reference sites (Hughes 1995), improvements in metric selection (Scott and Hall 1997, Hughes et al. 1998; McCormick et al. 2001; O?Neil et al. 2001, 2006), detection of metric redundancy (Bowen et al. 1996; 18 Angermeier et al. 2000; Dauwalter et al. 2003; Brambett et al. 2005) and modification of the original scoring criteria and development of new scoring methods (Fausch et al. 1984; Minns et al. 1994; Hughes et al. 1998; Smogor and Angermeier 1999; Dauwalter and Jackson 2004; Bramblett et al. 2005) all have enhanced the performance of the IBI. The IBI was originally developed for streams in the midwestern United States (Fausch et al. 1984; Karr et al. 1984), but has been shown to be highly adaptable to different geographic regions and aquatic ecosystems (Fore et al. 1994). For example, an IBI has been developed in the United States for lake ecosystems (Minns et al. 1994), modified rivers (Bowen et al. 1996), tailwaters (Scott 1999), coldwater streams (Leonard and Orth 1986; Lyons et al. 1995; Mundahl and Simon 1999), prairie streams (Lydy et al. 2000; Shearer and Berry 2002), reservoirs (Scott 1992; McDonough and Hickman 1999), and wetlands (Danielson 1998; Wilcox et al. 2002; Teels et al. 2004). Outside of the US, the IBI has been successfully utilized for biological assessment of aquatic systems worldwide (Steedman 1988; Oberdoff and Hughes 1992; Lyons et al. 1995; Hugueny et al. 1996; Hughes and Oberdoff 1999; Karr and Chu 1999). Miller et al. (1988) found that the IBI was applicable to a broad range of stream sizes with varying levels of habitat and water quality degradation. The authors also noted that the IBI was more responsive to disturbance after modification of the original Karr (1981) metrics to account for regional differences. When developing an IBI for a particular ecoregion, it must be adjusted to local ecological conditions (Paller et al. 1996). This typically requires selection and (or) modification of the original Karr (1981) metrics to accurately discriminate between disturbed and undisturbed sites in the region in which the IBI will be used. However, the 19 basic index structure developed by Karr (1981) was maintained to maximize metric sensitivity to a range of impacts (Paller et al. 1996). Metrics were included within each of the basic categories in keeping with Karr?s original index: species richness and composition, trophic composition, abundance, condition and local indicator species (Paller et al. 1996). Karr?s original metrics (1981) were developed for midwestern streams, many modifications have been necessary in streams in different US regions (Miller et al. 1988). Five of Karr?s original metrics have required frequent modification in different published versions of the IBI, including the number of darter species, the number of native sunfish species, the proportion of individuals as insectivorous cyprinids, the proportion of individuals as green sunfish and the proportion of individuals as hybrids (Hughes 1990). Where darter species were depauperate or rare, the numbers of benthic insectivore cyprinids and sculpin species have been used as substitutes in streams in New England and on the West Coast, respectively (Hughes 1990). In the Southeast, madtoms and sculpins have been added to this metric (Hughes 1990; Paller et al. 1996; Schleiger 2000; Walters et al. 2001; Compton et al. 2003). Schleiger (2000), in particular, noted the need to include benthic species in the IBI because those species are often disproportionately affected by degradation of aquatic systems. Thus, Schleiger (2000) modified the number of darter species metric to include the number of madtom and sculpin species in lower Piedmont and upper Coastal Plain streams in Georgia. The number of sunfish species metric has been replaced by the number of minnow species, or water-column species in the Northeast and on the West Coast, respectively (Hughes 1990). This modification was necessary because sunfish species 20 were introduced or depauperate in both regions (Hughes 1990). In the Southeast, the number of sunfish species metric required modification to include native sunfish species only, and excluded non-native species and hybrids such as the redbreast sunfish Lepomis auritus because of their increase in both the Tennessee and Alabama drainages (Georgia Department of Natural Resources 2005). Onorato et al. (1998) suggested that the original number of sunfish species metric be substituted with the number of centrarchid species because of the increase in the species richness of the centrarchid family in the southeastern United States. Hlass et al. (1998) and O?Neil et al. (2006) substituted the number of native minnow species metric for the original number of sunfish species because minnows exhibit a wider range of pollution tolerances, habitat preferences, and trophic and reproductive guilds and, as such, have greater value as an ecological indicator of water quality than sunfish. O?Neil and Shepard (1998) suggested that the original number of sunfish species should be substituted or eliminated entirely in IBI studies in the southeastern United States because the metric did not reliably distinguish between moderately and severely impacted streams. They found that Lepomis species often dominate in moderately degraded streams in Alabama and as such the original number of sunfish species metric showed low sensitivity to disturbance (O?Neil and Shepard 1998). The proportion of insectivorous cyprinids metric was originally suggested by Karr et al. (1984) because these species are a dominant trophic group in streams in the Midwest. However that metric has been replaced by ?all insectivores?, ?specialized invertebrate feeders?, or ?all invertivores? in other geographical regions. Ahle and Jobsis (1996), Hughes and Oberdoff (1999), O?Neil and Shepard (2000), and O?Neil et al. (2006) substituted the proportion of insectivorous cyprinids metric with the proportion as 21 invertivores metric because it was more inclusive and ecologically accurate than percent insectivorous cyprinids for assessing the degree to which the invertebrate community was degraded in streams. Compton et al. (2003) expanded the original percent insectivorous cyprinids metric to include all insectivorous individuals because of the high number of darter and madtom species classified trophically as insectivores. The original proportion of individuals as green sunfish metric has been replaced frequently, particularly in southeastern streams (Hughes 1990). The most commonly substituted metric for percent green sunfish has been the proportion of tolerant species, which was viewed as more inclusive than green sunfish alone (GA DNR 2005, O?Neil et al. 2006). Crumby et al. (1990), Shields et al. (1995), Ahle and Jobsis (1996), Shaner (1999), and Compton et al. (2003) replaced the percent green sunfish metric with a percent tolerant species metric because of the scarcity of green sunfish in streams. Several studies conducted by the Geological Survey of Alabama (GSA) have substituted percent green sunfish with percent Lepomis species as several of these Lepomis spp. have become dominant in disturbed habitats in Alabama streams (Shepard et al. 1997; O?Neil and Shepard 2000). IBI studies conducted by Schleiger (2000) and Bowen et al. (1996) eliminated the percent green sunfish metric entirely because these fishes composed < 1% of the total fish fauna collected in sampled streams. The Tennessee Valley Authority (TVA) has replaced the percent green sunfish metric with a percent tolerant species metric in its IBI assessments in the Tennessee River basin (Saylor and Scott 1987; Saylor et al. 1988; Wade and Stalcup 1987; Saylor 1989; Saylor and Ahlstedt 1990; Scott 1992). The original proportion of individuals as hybrids metric has been problematic in other regions outside the Midwest and has been regularly substituted or eliminated 22 entirely in IBI studies (Hughes 1990). Researchers have found this metric to be of limited use because of difficulty in identification of hybrid individuals and the lack of a consistent relationship between hybridization and environmental degradation (Shields et al. 1995; Schleiger 2000). An exotic species metric or reproductive guild metric has been substituted regularly for the percent hybrids metric in southeastern streams as recommended by the Ohio EPA (1987). Schleiger (2000) and O?Neil et al. (2006) substituted the percent hybrids metric with a reproductive guild metric because (1) few hybrids were found in the study streams, and (2) reproductive guild metrics were extremely sensitive to even moderate stream degradation and, as such, were deemed more reliable metrics than the percent hybrids metric. A list of metrics previously used in published IBI studies is presented in Table 2. These metrics were considered as potential metrics for this study from which a group of twelve metrics were selected based on evaluation of fish assemblages representative of southeastern Piedmont streams. IBI Studies of Streams in the southeastern United States A number of regional IBIs have been successfully developed for warmwater aquatic systems in the southeastern United States. In Alabama, the IBI has been used by the TVA in the Tennessee River basin since 1986 (Saylor and Scott 1987; Saylor et al. 1988; Wade and Stalcup 1987; Saylor 1989; Saylor and Ahlstedt 1990; Scott 1992). The Bowen et al. (1996) developed an IBI for the Tallapoosa River system to evaluate the sensitivity of the IBI to fish community changes resulting from streamflow alteration. The authors determined that the IBI successfully discriminated between unimpacted sites and sites undergoing severe flow fluctuations as a result of dam impoundment. The 23 authors recommended that the IBI be used in long-term monitoring programs in regulated southeastern river systems. The IBI has been applied by Onorato et al. (1998) and Davenport et al. (2005) to fish assemblages in the upper Cahaba River system. To assess the ecological health of the upper Cahaba, Onorato et al. (1998) conducted biomonitoring at 13 sites between 1995 and 1997. IBI scores were found to be lower at sites located downstream, and within the metro Birmingham area, which were most heavily influenced by urbanization. The authors concluded that these changes were related to point- and non-point sources of pollution from extensive urban development. Davenport et al. (2005) surveyed fishes at 15 sites on the upper Cahaba and used a modified version of the Shepard et al. (1997) IBI for the Cahaba River. The authors obtained similar IBI results to those of Onorato et al. (1998) as IBI scores were lowest in samples from the middle reaches of the river located in the most heavily urbanized areas in Birmingham. Sediment accumulation was found as the principal source of environmental degradation in those reaches where IBI scores were lowest (Shepard et al. 1997). An extensive watershed study by the Geological Survey of Alabama (GSA) of the Tallapoosa and Coosa river systems incorporating the IBI was conducted between 2003 and 2005 (O?Neil et al. 2006). IBI metrics and scoring criteria were calibrated to conditions in the Tallapoosa and Coosa River systems and then applied to a watershed study in Terrapin Creek (Coosa River drainage) to demonstrate the newly calibrated IBI and evaluate biological conditions in the stream. The authors showed that the IBI fits well into the framework of biomonitoring because of its integrative nature, ease of calculation, and broad regional applicability. However, they concluded that a better 24 understanding of the relationship between fish assemblages and classification factors such as drainages, ecoregions, and physiography needed to be developed to guide future IBI assessments. Finally, they concluded once IBI regions have been delineated, they should be thoroughly sampled across the gradient of human disturbance and stream sizes, and IBI metrics should be calibrated to account for both natural faunal variation and type and intensity within each IBI region. Regional IBIs also have been developed for other southeastern states. Crumby et al. (1990) applied the IBI to fish collected from Roaring River in north-central Tennessee to determine IBI sensitivity to known disturbances. They found the IBI to be useful in assessing habitat and water quality disturbances in Roaring River and that IBI scores were consistently lower at sites heavily impacted by sedimentation. An IBI study conducted by Shields et al. (1995) was used to quantify environmental quality in 27 watersheds bordering the Mississippi River alluvial plain in northwestern Mississippi, and to test the IBI for relating fish community characteristics to habitat degradation. The authors concluded that the IBI was effective for assessing biotic integrity over a large area with widespread degradation of physical habitats. The authors cautioned that future IBI studies must take into consideration the need for suitable reference data for setting metric scoring criteria, variability in IBI scores at different sites over time, and variability in IBI scores at sites with small sample sizes. Ahle and Jobsis (1996) developed an IBI to assess the impact of effluents from gold mine operations in nine Piedmont streams of South Carolina. It was determined that the IBI successfully differentiated between streams receiving mining effluents from those affected by only minor anthropogenic impacts. The authors concluded that the IBI was 25 an excellent tool for monitoring the fish assemblages of South Carolina Piedmont streams. Paller et al. (1996) developed an IBI to assess the health of fish assemblages in streams in the Sand Hills ecoregion on the upper coastal plain of South Carolina. Metrics were selected for the IBI based on their ability to effectively discriminate disturbed and undisturbed sites in the Sand Hills ecoregion. One unique aspect of this study was the development of a procedure based on species-area curves to remove the potentially confounding effects of unequal stream size and sampling effort from selected species richness metrics. A modified IBI was developed that was minimally affected by stream size and sampling effort and accurately discriminated between disturbed and undisturbed sites in South Carolina coastal plain streams (Paller et al. 1996). Schleiger (2000) conducted an intensive IBI study to detect land-use effects on fish communities in > 500 Piedmont streams in west-central Georgia. It was determined that the IBI effectively identified streams in low-density urban areas adversely impacted by nonpoint and point-source runoff. Schleiger (2000) concluded that the IBI was useful in differentiating stream quality between sites and that it was a valuable tool in long-term biological monitoring of Piedmont streams. Paller (2002) developed an IBI for assessing fish assemblages sampled over a 2- to 10-year period in 17 undisturbed and disturbed streams in the Savannah River Site (SRS), a Department of Energy reservation located on the upper coastal plain of South Carolina. An IBI was used to compare temporal fish assemblage variability between the disturbed and undisturbed sites. IBI scores were significantly lower in disturbed (vs. undisturbed) streams and showed greater variation over time in disturbed streams. Paller 26 (2002) also found that assemblages exhibited greater temporal variability in fish community structure in disturbed streams. In contrast, assemblages were more stable and persistent in undisturbed sites. Regulatory agencies in other southeastern states that have developed an IBI for inclusion in water resources biomonitoring include the Alabama Department of Environmental Management (ADEM), the Kentucky Department for Environmental Protection, the North Carolina Department of Environment, Health, and Natural Resources, and the South Carolina Department of Health and Environmental Control (Shaner 1999). In addition, the Fisheries section of the Georgia Department of Natural Resources has developed a standardized IBI for wadeable streams in the Piedmont ecoregion of Georgia (GA DNR 2005). Although the IBI has been applied extensively to studies relating many different types of anthropogenic disturbances to the health of fish assemblages in aquatic ecosystems, few studies used the IBI to assess how mining in a stream watershed impacts stream fish assemblages (Leonard and Orth 1986; Ahle and Jobsis 1996). Although Spence et al (1999) used an IBI to monitor fish assemblage health in streams with variable flow regimes during a drought year, no studies were found that used the IBI to examine changes in the resident fish community in a stream following an extended period in which the flow regime was altered. Thus, the IBI is an underutilized tool for assessing how an altered flow regime, resulting from anthropogenic activities in the watershed such as mining, can affect stream fish assemblages. As an effective biomonitoring tool, IBIs can provide biological data that can be used to identify perturbations in the watershed and 27 measure effects based on changes to fish assemblages (Karr 1981; Moyle 1994; Stewart and Loar 1994). Index of Well-Being (Iwb). Another approach to assessing fish assemblage data is the Index of Well-Being (Iwb). It was originally developed by Gammon (1976) to evaluate large riverine fish assemblages (Yoder and Smith 1999). It incorporates measures of abundance, biomass, and the Shannon diversity index (H?) into a single index value for assessing fish assemblage health (Gammon 1976; Ohio EPA 1987; Yoder and Smith 1999; Schleiger 2000). The Iwb was based on 4 measures of species diversity, abundance, and biomass, and was an attempt to produce an integrated evaluation of these fish assemblage attributes (Yoder and Smith 1999). Individual performances of abundance, biomass, and the Shannon index as indicators of fish assemblage quality have historically proven disappointing; however, when combined in the Iwb, these individual attributes were better predictors of fish assemblage condition (Yoder and Smith 1999). However, the Iwb does not incorporate the more sensitive aspects of community function such as trophic and reproductive guilds. A high Iwb value is indicative of a healthy fish assemblage characterized by high diversity and abundance of fishes. The Ohio EPA (1987) developed a modified Iwb that was different from the original index developed by Gammon (1976). Any species designated as highly tolerant, exotic, and hybrid were eliminated from the numbers and biomass components of the Iwb. This eliminated the positive bias produced by increased abundance of tolerant, hybrid, and introduced species at degraded sites, but retained their influence on the Shannon diversity indices (Yoder and Smith 1999; GA DNR 2005). 28 Relationship between the IBI and the Iwb The Iwb is used as a complementary evaluation to the IBI by the Ohio EPA and GA DNR for monitoring environmental quality in streams and rivers. It reflects overall production in terms of fish abundance and biomass and usually precedes the IBI in indicating the initial stages of fish assemblage recovery following a disturbance (Yoder and Smith 1999). The Iwb is very responsive to severe disturbances that result in reduced abundance, biomass, and species richness but does not incorporate the more sensitive attributes of the fish community such as the functional guilds of the IBI (Ohio EPA 1987). The IBI incorporates attributes of the fish assemblage such as the trophic and reproductive guilds, so it tends to respond more slowly than the Iwb in terms of exhibiting fish assemblage recovery (Yoder and Smith 1999). Both indices, when used together, can provide a more accurate assessment of fish assemblage response to anthropogenic disturbance (Ohio EPA 1987; Yoder and Smith 1999; Schleiger 2000; GA DNR 2005). Use of the Iwb requires the collection of biomass data, which is not required in current versions of the IBI (Yoder and Smith 1999; Schleiger 2000). Because both indices have their strengths and weaknesses, the Ohio EPA and GA DNR recommended both be used in bioassessments of stream fish assemblages. 29 III. MATERIALS AND METHODS Physiography and Geology For this study, three third-order stream sites located in the Tallapoosa River Drainage, East Central Alabama, USA, were sampled in fall 2004 and 2005. All three streams were in the Southern Outer Piedmont subecoregion above the Fall Line in Lee County. The Southern Outer Piedmont is a Level 4 subecoregion of the Piedmont ecoregion, portions of which are in both East Central Alabama and West Central Georgia (Omernick 1987; Griffith et al. 2001). The Piedmont ecoregion comprises a transitional area between the mountainous Appalachians to the northwest and the relatively flat coastal plain to the southeast (Omernick 1987; Griffith et al. 2001). The topography is characterized by rolling hills, broad and narrow ridges, and deeply narrow valleys (Mulholland and Lenat 1992). Piedmont soils are fine-textured, acidic, and nutrient-poor (Mulholland and Lenat 1992; Boyd 2000; Webber 2002). Major forest cover types in the Piedmont ecoregion are loblolly-short leaf pine and oak-pine (Turner 1987; Walser and Bart 1999). Piedmont streams have low to moderate gradients with cobble, gravel and sandy substrates, although many streams now contain large quantities of sand and silt as a result of sediment deposition from human activities (Webber 2006). Stream elevations range from about 50 to 600 m above sea level (Mulholland and Lenat 1992). Almost all Piedmont streams have been dramatically affected by historical land use (Mulholland and Lenat 1992). In the past 50 years landscape alterations resulting 30 from agriculture, silviculture, and watershed urbanization in heavily populated areas have altered stream hydrology, water quality, and physical habitat; these alterations have adversely affected biological communities (Mullholland and Lenat 1992; Helms et al. 2005). East Central Alabama and the areas surrounding the cities of Opelika and Auburn are notable examples of this trend of increased urbanization (Webber 2002). Description of Study Sites Chewacla Creek in the study area is a third-order reach in Lee County, Alabama. The stream basin encompasses a total drainage area of about 383 km 2 (148 mi 2 ). As part of the Tallapoosa River drainage, Chewacla Creek?s headwaters originate south of Opelika. The stream flows through pasture, commercial, and private developments before emptying into Lake Ogletree, a 1.2 km 2 (300 acres) lake that is the primary drinking water supply for the city of Auburn (Figure 1). Flow into Chewacla Creek downstream of Lake Ogletree dam is regulated by water levels in the lake. When the lake is at full pool, water flows over the spillway into the channel downstream. When the lake is low during dry periods of the year, little or no flow goes over the spillway. During the latter periods only seepage below the dam and local springs supply water to the downstream channel. Downstream from Lake Ogletree, Chewacla Creek flows through several geological formations including the Manchester Schist, the Chewacla Marble, and the Hollis Quartzite (USFWS 2003). Extensive mining of the Chewacla Marble first began in the early 1900s (Prouty 1916). The Chewacla Marble is a narrow strip of highly crystalline dolomite that originates ~ 8 km (5 miles) southeast of Opelika (Prouty 1916). The belt in which the dolomite occurs is ~ 0.5 km (a third of a mile) wide and extends ~ 4 km (2.5 miles) in a northeast and southwest direction (Prouty 1916). The primary aggregate mined in the Chewacla Marble has been limestone. The chief quarry located along Chewacla Creek, now owned and operated by the Martin Marietta Company, was first opened in 1961 for the purpose of supplying building materials to the Alabama Department of Transportation (DOT) to construct Interstate 85 (Scott Weinhold, Martin Marietta Company, personal communication). The Martin Marietta Company purchased the quarry in 1995 from Drevo Industries and has operated the stone-crushing facilities at the quarry since 1995 (Scott Weinhold, Martin Marietta Company, personal communication). The major tributaries into Chewacla Creek downstream of Lake Ogletree are Moores Mill Creek and, further downstream, Town and Parkerson Mill creeks. Moores Mill Creek was an ADEM 303 (d)-listed stream, impaired because of pollution from excessive sediment (Webber 2006). Moores Mill Creek flows into Chewacla Lake in Chewacla State park. Downstream from Chewacla Lake, Moores Mill Creek joins Chewacla Creek to form a fourth-order reach that flows south to Opintlocco Creek in Macon County, Alabama, where they join to form Uphapee Creek. Major tributaries into Chewacla Creek upstream of Lake Ogletree include Nash and Robinson creeks. Two sampling sites were chosen in the reach of Chewacla Creek between Lake Ogletree and the quarry for the collection of fish samples and analysis of hydrological and chemical variables. Both sampling stations were on the Chewacla Creek main stem. The reach has in the past exhibited a highly altered flow regime during dry periods as a 31 result of no water flowing over the lake spillway and sinkholes in the channel in the vicinity of the quarry further downstream. The upstream site (CH-U) was ~ 300 m downstream from the Lake Ogletree dam. The downstream site (CH-D) near the quarry was ~ 100 m downstream of the ?pretty hole?, a deep pool providing refuge for fishes during periods when this reach was completely dewatered. Location and stream characteristics for CH-U and CH-D are presented in Table 3. The reference stream for this study was a third-order reach of Cane Creek (Figure 2). The Cane Creek site (Cane-1) was in a relatively undisturbed area of the Ropes Creek watershed. Ropes Creek is a tributary to Sougahatchee Creek, also in the Piedmont ecoregion in Lee County, Alabama. The sampling site was located just upstream of the bridge on Lee County Road 217. Collection of biological and chemical data was conducted concurrently with CH-U and CH-D in 2004 and 2005; however, no hydrological data were collected in Cane-1 both years. Location and stream characteristics for Cane-1 are also presented in Table 3. Physical Habitat Assessment A detailed habitat assessment was conducted at each site according to methods described in the US EPA Rapid Bioassessment Protocols (Barbour et al. 1999). Habitat in each reach was scored on the basis of instream variables including substrate and available cover, channel morphology, bank features, and streamside (riparian) vegetation. A total of 10 metrics developed by Barbour et al. (1999) were used in the habitat assessment for each study site (Table 4). Each metric was scored in the field based on observation and best professional judgment. Scoring criteria used to visually estimate 32 33 physical habitat quality at sites in Chewacla and Cane creeks are presented in Table 5. The assessment used to evaluate physical habitat quality at all sites is presented in Table 6. Characteristics of Flow Regime in Chewacla Creek As part of the Chewacla Creek Safe Harbor Agreement (USFWS 2003), continuous monitoring of stream flow was conducted at 4 stations downstream of Lake Ogletree (Figure 3). Water level was measured at each station using Solinst Model 3001 Levellogger gages (pressure transducers) (Webber 2006). These data were available to characterize the flow regime in the study reaches downstream of Lake Ogletree. Discharge was measured in the field at least monthly using the velocity-area method as described in Rantz et al. (1982), Wahl et al. (1995), and Gore (1996). A Marsh- McBirney Flo-Mate Model 2000 portable flowmeter was used to measure the velocity. A Microsoft Excel program was developed to calculate discharge from the velocities recorded. Flow measurements taken in the field were used to calculate stage/discharge regressions for each station (Webber 2006). Using the regression model, discharge (MGD) was calculated from stage level readings for each station. Hydrographs showing mean daily discharge (MGD) were prepared from these data for both CH-U and CH-D. Water Chemistry Analysis Field and laboratory measurements of physical and chemical variables were conducted annually at Cane-1, CH-U, and CH-D during the study. Dissolved oxygen and water temperature were measured in situ using a Yellow Springs Instrument (YSI) Model 34 51B Dissolved Oxygen meter. Specific conductance was measured in situ using a YSI Model 30 Conductivity meter. Water samples were collected in 2-L Nalgene bottles at each site for analysis in an AU Fisheries laboratory. Nine water quality variables were measured: turbidity, total alkalinity, total suspended solids (TSS), total hardness, total nitrogen (TN), total phosphorous (T-PO 4 ), orthophosphate (O-PO 4 ), total ammonium (NH 3 -N) and nitrate nitrogen (NO 3 -N). All water quality analyses conducted in the laboratory followed methods recommended by the American Public Health Association (APHA 1998). Turbidity was measured using the nephelometric method with a Model 18900 Ratio Turbidimeter. Total alkalinity was measured by titrating a water sample with a standard 0.1 N hydrochloric acid (HCl) solution to a pH of 4.5. Total hardness was measured by titrating a water sample using an EDTA titrant and Erichrome Black T dye as a color indicator. Total nitrogen (TN) and total phosphorous (TP) were measured using the persulfate digestion method. Orthophosphate (O-PO 4 ) was measured using an ascorbic acid test. Total ammonium (N-NH 3 ) was measured using the automated phenate method. Nitrate nitrogen (N-NO 3 ) was measured using the cadmium reduction method. Fish Sampling Methodology Fishes were collected annually over a 2-year period at both impacted and the reference sites using Rapid Bioasssessment Protocols (RBP) recommended by the US EPA (Barbour et al. 1999). Sampling locations were selected to include riffle, run, and pool habitat. Sampling all major habitat types was important to obtain a representative sample of fishes illustrating assemblage attributes (such as abundance, richness, and trophic composition as these attributes can change along a habitat gradient (Schlosser 35 1982). Electroshocking was the primary method of fish collection. Equal sampling effort was used at each site to ensure the capture of the most representative sample possible at each site and so that direct comparisons of richness would be possible without species-area bias. Block nets using 30-foot seines were set at the upstream and downstream ends of each reach to minimize fish escaping during sampling. Two separate electroshocking runs were conducted in each 100 meter reach. Pulsed DC current was used to shock fishes at each site. The first run utilized a small 10-foot jonboat with a portable generator and booms with attached electrodes hanging from the end of each boom. The boom electrodes were the anodes, and the boat hull was the cathode. Starting at the most downstream end of the reach, a three-person crew was utilized with one person pushing the jonboat upstream and two other persons positioned on each side of the boat to net stunned fish. A maximum current of 4 to 5 amps was applied by the generator which sufficiently stunned fish for netting. The second run involved the use of a Coffelt Mark- 10 portable backpack electroshocker powered by a Honda EX350 generator. The backpack shocker consisted of a separate anode connected to a pole held by the operator and a trailing rattail cathode. Starting at the most downstream end, a two-person crew was utilized with one person carrying the shocker and a second person netting stunned fish. A maximum current of 4 to 5 amps was applied to stunned fish. The use of the backpack electroshocker was important because it allowed access to potential fish habitat difficult for the jonboat to access such as rootwads, undercut banks, and shallow riffle areas. Both runs utilized a bank-to-bank sweeping technique to ensure maximum 36 coverage of all available fish habitat (Barbour et al. 1999). Captured fishes were transported in buckets to an aerated livewell located on the stream bank. One minor logistical problem encountered during sampling was the natural low baseline conductivity in Cane Creek, which is typically ~ 50 ?S (Webber and Blevins 2001; Webber 2002, 2006). Salt was used to increase the conductivity of the water during electroshocking. Fifty pound bags of salt were positioned in a riffle area at the most upstream end of the reach and then opened to allow the salt to disperse downstream and dissolve (Vincent 1960 and 1971; Hindman et al. 1990). This procedure was not necessary at the impacted sites because conductivity was typically greater than 100 ?S at each site (Webber 2006), which was sufficient for generating the 4 to 5 amps needed without the addition of salt (Webber 2006). As a result, at no time during sampling was conductivity considered a limiting factor in electrofishing success at any of the sites. Fishes were identified on the stream bank using appropriate reference materials (Mettee et al. 1996; Boschung and Mayden 2004). Individuals were weighed, measured for total length, inspected for lesions, tumors, and other abnormalities, and then returned alive to the stream. Individuals that could not be identified in the field were preserved in 10% formalin and returned to the laboratory at Auburn University for identification. All individuals captured were identified to species Designation of Guilds and Tolerance Classes Fish species were assigned to one of five feeding guilds (FG) based on published descriptions of their feeding ecology: herbivores (H), insectivores (INS), omnivores (O), top carnivores (TC), and invertivores (INV) (Mettee et al. 1996; Boschung and Mayden 37 2004; O?Neil et al. 2006). Herbivores consume plant material, either living or dead. Insectivores consume insects or insect nymphs almost exclusively. Omnivores consume both plant and animal matter. Invertivores are more general feeders consuming insects, crustaceans, mollusks, and other invertebrates. Top carnivores are opportunistic predators that consume a wide range of terrestrial and aquatic animal species; many consume fish as their major food source (O?Neil et al. 2006). Fish species were assigned to one of four tolerance classes (TC): pioneer (P), intolerant (INT), intermediate (MI), and tolerant (TOL) (Ohio EPA 1987; Mettee et al. 1996; Goldstein and Simon 1999; Smogor and Angermeier 1999; Schleiger 2000; Tabit and Johnson 2002; Boschung and Mayden 2004; O?Neil et al. 2006). Schleiger (2000) defined pioneer species as those that will rapidly invade a previously disturbed habitat, and their presence in a site indicates temporal or unstable habitats. Examples of pioneer species include Campostoma, Nocomis, and Lepomis species (Schleiger 2000; Tabit and Johnson 2002). Intolerant species are classified as those whose ranges or abundances have decreased, presumably because of anthropogenic disturbance, and tolerant species as those known to be least detrimentally affected by anthropogenic disturbances to stream watersheds; many of these species historical ranges or local abundances have increased with increases in anthropogenic disturbance (Ohio EPA 1987; Smogor and Angermeier 1999). Ohio EPA (1987) defined intermediate species as those that are commonly observed and strongly associated with healthy fish assemblages, but are occasionally observed in slight to moderately degraded streams. Assignment of tolerance classes can be highly subjective because a number of fish species are sensitive to some types of 38 disturbance and tolerant of others (Leonard and Orth 1986; Ebert and Filipek 1988; Hlass et al. 1998). Fish species were assigned to one of two breeding guilds based on published descriptions of spawning behavior: lithophilic spawners and non-lithophilic spawners (Mettee et al. 1996; Boschung and Mayden 2004; O?Neil et al. 2006). Lithophilic spawners obligately use clean mineral substrates (i.e., sand, gravel, cobble) in which to deposit their eggs. Non-lithophilic spawners deposit their eggs on the remaining types of substrate or spawning surfaces such as aquatic vegetation, coarse organic matter, and other features not related to lithophilic substrates (O?Neil et al. 2006). Characterization of Fish Assemblages Using an Index of Biotic Integrity (IBI) Selection and modification of IBI metrics. Fish data collected during this study were analyzed using the Index of Biotic Integrity (IBI) (Karr, 1981). Previous IBI studies conducted in the southeastern United States were thoroughly researched to assist in the selection of IBI metrics for this study. Metrics were grouped into three categories as originally suggested by Karr et al. (1986): (1) species richness and composition, (2) trophic and reproductive guilds and (3) fish abundance and condition. Modifications of Karr?s (1981) original metrics were made to reflect the fish fauna typically found in southeastern Piedmont streams. Each metric was scored and individual scores were summed to calculate a total IBI score for each site. A separate IBI score was calculated for each site for each year sampled (2004 and 2005). Thus, a total of six IBI scores were calculated for the three study sites. Each IBI score was then equated to a particular 39 stream condition as originally developed by Karr (1981): Excellent, Good, Fair, Poor, and Very Poor. Defining the reference condition. Many IBI studies used the concept of the reference condition to establish a network of reference streams that exhibited the most natural conditions against which biotic integrity can be measured (Compton et al. 2003). The reference condition refers to the range of measurable elements (i.e., chemistry, habitat and biology) that are found in natural environments. Thus, stream location, physical and chemical attributes of the water and surrounding watershed, and the resident biological communities are criteria often used for selection of regional reference sites and the establishment of the reference condition (Simon and Lyons 1995; ADEM 1999; Compton et al. 2003). Natural, undisturbed land cover is almost nonexistent in the southeastern Piedmont, so regional reference sites should not be considered pristine or undisturbed. Rather, regional reference sites should be defined as representing the best possible conditions given anthropogenic impacts that have occurred historically (O?Neil et al. 2006). Typically, multiple reference sites must be sampled to obtain a reasonable estimate of the reference condition for a specific region (Shaner 1999). Since only one stream (Cane-1) was selected and sampled as a reference site, it was necessary to determine if it sufficiently met the reference condition established for Piedmont streams in other IBI studies. Analysis of fish data from museum collection records (Auburn University Museum of Natural History, unpublished data; University of Alabama Ichthyological Collection, unpublished data; Tulane University Museum of Natural 40 History, unpublished data, University of Georgia Museum of Natural History, unpublished data), previous ichthyological studies (Gilbert 1969; Dahlberg and Scott 1971; Webber 1971; Pierson et al. 1986), and ADEM reference streams provided reference data to compare with data from Cane-1. The ADEM reference sites were Cornhouse Creek (Randolph County Road 821), Hatchet Creek (Tyler Ford), Hurricane Creek (upstream of Randolph County Road 26), Emuckfaw Creek (3 miles west of Daviston, AL on unnamed Tallapoosa County road) and Weogufka Creek (County Road 41 near Stewartville, AL) (D. Miller, unpublished data) Description of Core Metrics. The method of Karr et al. (1984) was used in applying the IBI to data sets from Chewacla and Cane creek sites with modifications intended to make the IBI more sensitive to regional conditions found in Piedmont streams. Twelve core metrics were chosen from the pool of 33 (Table 2) candidate metrics used commonly in previous IBI studies. These metrics were selected because they were most representative of fish assemblages found in Piedmont streams. The twelve core metrics selected for this study appear in Table 7. Metrics 1-4 evaluated species richness and composition at each site. A description of each metric in this category is given below. 1. Total number of native fish species. This metric was a count of all the native fish species in the sample. The number of native fish species supported by similar streams in a given region generally decreases with environmental degradation (Karr et al. 1986). Exotic and hybrid species were excluded in the calculation of this metric as 41 recommended by the Ohio EPA (1987) because their presence does not provide an accurate assessment of long-term biotic integrity. 2. Total number of native cyprinid species. Fish species belonging to the family Cyprinidae are highly ubiquitous in streams in Alabama with a broad range of tolerances, habitat preferences, and trophic and reproductive guilds (Shepard et al. 1995; Mettee et al. 1996; O?Neil and Shepard 2000; Boschung and Mayden 2004). Many cyprinid species are specialized feeders, and high richness and abundance in a sample is indicative of a diverse aquatic macroinvertebrate community and a healthy trophic structure of a stream fish assemblage (Niemela et al. 1999). Cyprinids are found in a variety of habitats over a diverse array of substrates, thus providing a measure of the quality of instream cover and bottom substrates (O?Neil and Shepard 1998; GA DNR 2005). Many cyprinid species are sensitive to environmental disturbance and thus have value as ecological indicators (Fausch and Schrader 1987; Hlass et al. 1998). 3. Total number of native round-bodied sucker species. Sucker diversity is high in Alabama streams (23 species) (Boschung and Mayden 2004). Most sucker species are sensitive to environmental degradation (O?Neil and Shepard 1998; Boschung and Mayden 2004). Schleiger (2000) found sucker species in Georgia Piedmont streams to be sensitive to habitat alteration, siltation, and changes in water quality. Sucker species are found in all types of habitats and typically have long life spans and, thus, can provide long-term assessments of past and current environmental conditions (O?Neil and Shepard 1998). 42 4. Number of intolerant species. This metric included species intolerant of several types of environmental disturbance including siltation, low dissolved oxygen levels, flow alterations or chemical contamination (O?Neil and Shepard 2000). Intolerant species are typically the first to disappear after some form of disturbance to a stream (Ohio EPA 1987). Examples include minnow (Hybopsis and Cyprinella), darter (Etheostoma and Percina), madtom (Noturus), sunfish (Ambloplites and Micropterus) and sculpin (Cottus) species (Tabit and Johnson 2002; Boschung and Mayden 2004; O?Neil et al. 2006). Metrics 5-10 measure species composition and trophic dynamics at each site. These metrics assess the quality of the energy base and the flow of energy through the stream community; they also provided a measure of the availability of suitable spawning habitat in each stream (Karr et al. 1986). A description of each metric in this category is given below. 5. Proportion of individuals as tolerant species. This metric has been commonly used as a replacement for Karr?s (1984) original metric of proportion as green sunfish (Ohio EPA 1987; Hughes and Oberdorff 1999). In degraded streams, more pollution tolerant species typically dominate at the expense of less pollution tolerant species (Karr et al. 1986). Thus, diversity of the fish assemblage can decrease without a loss in species richness because of an increase in relative abundance of one or more pollution tolerant- species (GA DNR 2005). Examples include certain minnow (Campostoma, Luxilus and Nocomis), sucker (Minytrema), catfish (Ameiurus), sunfish (Lepomis), and darter 43 (Percina) species (Mettee et al. 1996; Boschung and Mayden 2004; GA DNR 2005; O?Neil et al. 2006). 6. Proportion of individuals as top carnivores. Viable and healthy populations of top carnivore species typically indicate a healthy, trophically diverse community (Karr et al. 1986; Hlass et al. 1998). Hughes and Oberdorff (1999) noted that top carnivores can be affected by long-term environmental impacts because they are typically long-lived. Only species that feed primarily on fish and larger invertebrates as adults were included (O?Neil and Shepard 2000; O?Neil et al. 2006). Examples include Cottus, Lepomis, and Micropterus species (O?Neil et al. 2006). 7. Proportion of individuals as Lepomis species. This metric represented the proportion of individuals in a sample that were Lepomis species. Non-native species and Lepomis hybrids were included in this metric. Lepomis species are highly responsive to increased disturbance with relative proportions increasing dramatically at disturbed sites (GA DNR 2005). An assemblage dominated by Lepomis species may be indicative of a reduction in macroinvertebrate diversity and a lack of suitable spawning habitat for broadcast spawners (GA DNR 2005). Lepomis species were dominant in moderately disturbed Alabama streams, sometimes exceeding 50% of the sample (O?Neil and Shepard (1998). Also, the proportion of Lepomis species differed significantly between disturbed and undisturbed sample sites in South Carolina coastal plain streams (Paller et al. 1996). 8. Proportion of individuals as benthic invertivores. This metric represents the proportion of darters, madtoms, and sculpins collected in a sample. Because of their 44 specificity for feeding and reproducing in benthic habitats, benthic invertivores species assess the availability of suitable benthic habitat at a site (Ohio EPA 1987). Many IBI assessments strictly used darter species as a measure of environmental degradation in streams; however, because darters, madtoms, and sculpins are naturally scarce in streams, are benthic species, and exhibit similar feeding and spawning requirements these three groups were combined into a single metric (Paller et al. 1996; Schleiger 2000). 9. Proportion of individuals as non-lithophilic spawners. Non-lithophilic spawners included those species that use a variety of spawning substrates including snags, aquatic vegetation, and coarse organic matter, or spawn in open water (O?Neil et al. 2006). Non-lithophilic spawners include Cyprinella, Ameirus, Noturus, Etheostoma, and Micropterus species (Boschung and Mayden 2004; O?Neil et al. 2006). O?Neil et al. (2006) found that the proportion of non-lithophilic spawners was more strongly related to disturbance and discriminated more effectively between least and most-disturbed sites in their Terrapin Creek watershed study than did simple lithophils. Non-lithophilic spawners showed a broader response to disturbance than simple lithophils by incorporating not only the effects of substrate siltation but also more complex disturbances such as changes in flow regime and loss of large woody debris that can adversely affect instream fish habitat (O?Neil et al. 2006). 10. Proportion of individuals as omnivores and herbivores. This metric evaluated the shift in trophic composition of the fish assemblage in degraded streams (Karr et al. 1986). The abundance of omnivores often increases at degraded sites as specialist feeders emigrate or are eliminated because of limited food resources (Hlass et al. 1998). 45 Examples of generalists include Nocomis and Ameiurus species (Smogor and Angermeier 1999; O?Neil et al. 2006). This metric also includes those species that feed primarily as herbivores, such as Campostoma species, whose increase abundance in a sample is associated with elevated nutrient levels (O?Neil and Shepard 1998). Metrics 11 and 12 evaluated the abundance and condition of the fish assemblage. A description of each metric in this category is given below. 11. Number of individuals per 100 meters. This metric represented the total number of fish collected over a designated 100 meter stream reach (Ohio EPA 1987; Barbour et al. 1999). Abundance was one of the original Karr metrics (1986) and has been widely employed in IBI applications (Hughes and Oberdorff 1999). Fish abundance is generally assumed to decrease with habitat disturbance (Ohio EPA 1987); however, in some disturbed streams it can lead to higher proportions of tolerant species. Therefore, the scoring criteria were adjusted to avoid rewarding a possible degraded site with a very high catch rate as a result of increased numbers of tolerant individuals (Ohio EPA 1987; Simon 1991). Sites with catch rates <100 individuals or > 400 individuals received a score of 0. Sites with catch rates within an optimum range of 200 to 300 individuals received a score of 100. Sites where catch rates deviated moderately, either above or below the optimum range, received a score of 50. 12. Proportion of individuals with external anomalies plus hybrids. The original metric proposed by Karr et al. (1986) represented the proportion of individuals with deformities, eroded fins, lesions, and tumors (DELTs). The occurrence of unhealthy individuals in a fish assemblage often reflects the health and condition of the assemblage 46 as whole (O?Neil et al. 2006). However, occurrence of large numbers of diseased individuals is rare except in the most highly degraded streams (Karr et al. 1986). Similarly, hybridization between species is usually indicative of highly disturbed streams (Karr et al. 1986). However, occurrence of hybrid individuals is rare except in the most degraded streams (O?Neil et al. 2006). Most streams exhibit only low to moderate numbers of diseased and hybridized fishes, so I combined the 2 into a single metric to better distinguish between least and most disturbed sites. Development of Expected Values for IBI Metrics. Expected values are defined as those metric values that would be expected for fish assemblages typically observed in relatively undisturbed Piedmont streams (O?Neil et al. 2006). Fish data from the reference site and from the 5 ADEM Piedmont reference streams as well as IBIs developed for Piedmont streams (Bowen et al. 1996; Schleiger 2000; O?Neil et al. 2006) were used to establish expected values (EV) (Table 7). Often, richness can vary with upstream watershed area (drainage size) with larger streams typically supporting a higher number of species (Fausch et al. 1984; Karr et al. 1996; Osborne et al. 1992; Bowen et al. 1996). Therefore, expected values (EV) were adjusted to account for variation in drainage size in Piedmont streams using the method outlined by O?Neil et al. (2006) (Table 7). Relative abundance is often not influenced by the size of the watershed (GA DNR 2005), so relative abundance metrics were not adjusted to account for differences in watershed area (Table 7). Development of Scoring Criteria for IBI. The original method of scoring metrics using a trisection of the scoring range has been well-documented (Karr et al. 1986; Ohio 47 EPA 1987; Fore et al. 1996; Barbour et al. 1999; O?Neil et al. 2006). However, Minns et al. (1994) and Hughes et al. (1998) standardized metrics as percentages of an expected metric value. A similar metric scoring system was developed for this study. Raw metric scores were standardized to a scale of 0-100 by dividing each raw metric score by an expected value for that metric as determined from reference data (Cane Creek and ADEM data) and different studies in Piedmont streams using the IBI (Bowen et al. 1996; Schleiger 2000; O?Neil et al. 2006) and multiplying by 100 (Hughes et al. 1998, Barbour et al. 1999; Compton et al. 2003; Bramblett et al. 2005). A continuous scale was used because it was shown to be more responsive to environmental impacts than categorical scoring (5, 3, 1) used by Karr et al. (1984) (Hughes et al. 1998; McCormick et al. 2001; Compton et al. 2003). Metrics from sites that performed well and scored >100 were set at 100. Metrics from sites that performed poorly and had negative values were set at zero. The equations for calculating standardized metric scores were: Standardized metric score = (raw metric score/expected metric value) x 100 for positive-response metrics and Standardized metric score = 100 ? (raw metric score/expected metric value) x 100 for negative-response metrics. Positive-response metrics are defined as metrics whose scores are expected to decrease with degradation, and negative-response metrics are those whose scores are expected to increase with degradation (Compton et al. 2003). The final IBI score, on a 0-100 point scale, was the average of the standardized metric scores. Streams were placed into one of six integrity classes based on IBI scores (adapted from Minns et al. 1994 and Dauwalter et al. 2003). The integrity classes for evaluating stream 48 condition were: 0 (no fish), > 0 - < 20 (very poor), 20 - < 40 (poor), 40 - < 60 (fair), 60 - < 80 (good) and > 80 (excellent). Characterization of Fish Assemblages Using an Index of Well-Being (Iwb) The Index of Well-Being (Iwb) was calculated for each site for both 2004 and 2005. The Iwb was expressed as: Iwb = 0.5 ln (N / 100 m) + 0.5 ln (B / 100 m) + H' (no.) + H' (wt.) where N / 100 m = number of individuals per 100 m B / 100 m = total biomass for all species per 100 m H' (no.) = Shannon diversity index based on numbers of fish H' (wt.) = Shannon diversity index based on biomass of fish In the calculation of the Iwb, any highly tolerant, hybrid, and introduced species were excluded from the relative abundance components of the Iwb, but retained in the calculations for H? as recommended by the Ohio EPA (1987) and GA DNR (2005). Each Iwb score was then equated to a particular stream condition: Excellent, Good, Fair, Poor, and Very Poor. Scoring criteria for the Iwb used were the same as that used by the GA DNR (2005) and Ohio EPA (1987) in their stream bioassessments (Table 8). Data Analysis Index of Biotic Integrity. For statistical purposes, the distribution of the 12 metric scores for each site were categorized as k independent groups where k = 3. A Kruskall- 49 Wallis (K-W) test was then used to compare the groups of metric scores across sites to determine if differences in median metric scores existed. To determine where group differences occurred, a Steel-Dwass (S-D) nonparametric post-test was used. The S-D test is the nonparametric equivalent of the Tukey-Kramer HSD test and is used to assess pairwise differences between groups (Zar 1996). Raw metric scores for each individual metric were evaluated using a, an F-test to determine if median metric values across sites were statistically equivalent or distinct (US EPA 2006). Differences for all statistical tests were considered significant at p ? 0.05. Measures of Assemblage Structure and Function. Descriptive measures related to community structure included total abundance (N), species richness (S), species diversity (H?), and evenness (E). These measures were compiled for each site and sampling period. Total abundance (N) was the total number of individuals present in a sample. Species richness (S) was the total number of species present in a sample. To measure the diversity of the fish assemblages at each site, the Shannon diversity equation (H?) was selected. This index weighs species exactly by their relative proportions in a sample, without favoring rare or common species (Peet 1974; Jost 2006). This equation is given as: H? = - ? (n i /N) log e (n i / N) (Shannon and Weaver 1949). Wilhm and Dorris (1968) interpreted Shannon diversity values as follows: (1) little stress / relatively unimpacted (? 3.0), (2) moderate stress / moderately impacted (1.0 ? 2.9) and (3) stressful / impact is evident (< 1.0). 50 Evenness, the distribution of individuals among species in a sample, was calculated using the following equation: Evenness (E) = e H? /S (Sheldon 1969) Evenness values range from 0 to 1 and approach 0 as a single species becomes more dominant in the community (Brower and Zar 1984; Ludwig and Reynolds 1988). Although evenness indices can be used to assess the condition of the fish assemblage, comparisons across sites and years must be interpreted with caution because most evenness indices are highly sensitive to the number of species in a sample (Alatalo 1981; Zar 1984; Ludwig and Reynolds 1988). A chi-squared (? 2 ) test for association (involving r x c contingency tables) was used to examine across-site differences in trophic structure using feeding guilds and tolerance class for each sampling period (Sokal and Rohlf 1987; Tabit and Johnson 2002; Keaton et al. 2005). Trophic structure of the resident fish assemblages at each site was determined based on the percentage of the number of captured individuals represented by each feeding guild, respectively relative to the total number of individuals in the sample. The relative abundance of fishes in each tolerance class was calculated as the number of captured individuals in each tolerance class divided by the total number of individuals in the sample. For the chi-square tests, differences were considered significant at p ? 0.05. To test for normality of fish abundance data, histograms showing the distribution of abundance data were plotted. If the data was non-normal, a Wilcoxon signed-rank test was used to examine between-year differences in trophic structure (feeding guilds) and relative abundance in each tolerance class (Tabit and Johnson 2002; Keaton et al. 2005). 51 The Wilcoxon test is useful for between-year comparisons of fish community data when the same localities are sampled over time (fixed-station sampling) resulting in correlated, non-independent samples (McCall 1986; Maceina et al. 1994; Sokal and Rohlf 1987). For the Wilcoxon test, differences were considered significant at p ? 0.05. All statistical analyses were conducted with JMP Version 4.04 Statistical Software (SAS Institute, Inc.) (Sall et al. 2001). 52 IV. RESULTS AND DISCUSSION Physical Habitat Assessment Scores of the visual assessment of physical habitat at each site are presented in Table 9. In terms of habitat quality, the upstream and downstream impacted sites (CH-U and CH-D) were comparable to the reference site (Cane-1), each scoring 191 out of a possible 200 points. Cane-1 scored 195 out of a possible 200 points. The main differences between Cane-1 and CH-U and CH-D were related to the amount of sediment present in the stream channel and the frequency of large, well-developed pools. Because of local clearcutting for timber on the Cane Creek watershed, more sediment deposition was evident at Cane-1, particularly in the deeper pool areas, relative to CH-U and CH-D. Because of the presence of Lake Ogletree, there was little sediment deposition at either CH-U or CH-D. This was because lake water flowing over the spillway or through the siphon (drawing water at about 5 feet below the surface) was typically low in sediments as heavier sediment particles entering the lake with stream water were settling to the lake bottom. In addition, there were a series of deep pools at Cane-1 that were noticeably absent at both CH-U and CH-D. At CH-U and CH-D, pools were present in each reach but were smaller and shallower than those at Cane-1. In addition, there were only portions of large, deep pools located at the lower ends of both CH-U and CH-D. However, large pools were located upstream and downstream of both sites. At CH-U, 53 there was a long pool upstream of a small ford that crosses the creek at the upper end of the site as well as several large pools located just downstream of the site. At CH-D, there was a very large pool located upstream (known locally as the ?pretty hole?) as well as a long series of pools located just downstream of the site. Although deep pools were mostly absent within each site, the presence of local pools both upstream and downstream of the sites likely served as refuge and cover for fishes, particularly during low flows. Schlosser (1982) found that stable stream reaches were characterized by large, deep pools containing older age class fishes and larger pool species, particularly catostomids and centrarchids. In comparing physical habitat between the two impacted sites minor differences were evident. In terms of available fish habitat, there were no undercut bank areas present at CH-D as was observed at CH-U. In addition, there were no large, deep pools present at CH-D needed by older age class fishes and larger pool species (Schlosser 1982). Emergent vegetation including American water willow Justicia Americana and river weed Podostemon ceratophyllum was evident at both sites. Instream cover in the form of woody debris was more prevalent at CH-U than at CH-D. Both sites were observed to have nearly identical substrate characteristics. Riffles were composed of gravel, medium-sized cobble, and small boulders at both sites. There was much more exposed bedrock in the run areas at CH-D than at CH-U, although medium-sized cobble and small boulders were present similar to that observed in the runs at CH-U. Although habitat scores were the same for both sites, the lack of undercut bank habitat at CH-D was the most noticeable difference between the two sites. Flow Regime between Lake Ogletree and Martin Marietta Quarry 54 The hydrographs indicated very similar flow conditions in the channel at both sites (Figure 4). At CH-U, discharge approached 2.0 MGD only during August and September 2004, May 2005 and again in March 2006. CH-D exhibited a similar flow pattern with lowest flows in August and September 2004 and May 2005. Both sites exhibited highest discharge values during the spring 2005 and 2006 as a result of several large spring rain events. During the study period from August 2004 to April 2006, water levels never dropped below 2.0 MGD at either site. Thus, there was consistent water flow in the stream channel such that fish migration was not impeded at any time during the study. However, both hydrographs, in particular for CH-U, exhibited modified flow conditions common in streams downstream of an impoundment (Poff et al. 1997). These changes in flow have the potential for detrimental effects on stream ecosystems (Poff et al. 1997; Richter and Thomas 2007). Between August 2004 and April 2006, there were a high number of low flow pulses but few high flow pulses (Figure 4). In addition to low flow pulses, periodic high flow pulses are necessary to maintain natural seasonal variability in stream flows to support natural biota (Poff et al. 1997; Richter and Thomas 2007). During rainfall events, water levels in Lake Ogletree were regulating downstream flows into the stream. This dampening of flow amplitude by the lake resulted in reduced high flow pulses and higher base flow variability, as seen on the hydrographs for CH-U and CH-D (Figure 4). Although measures were taken to stabilize base flows downstream of the lake (USFWS 2003); periodic high flows necessary for maintaining natural flow variability were lacking in the study reach (Richter and Thomas 2007). Unfortunately, because no flow data was collected at Cane-1, comparisons in mean daily discharge 55 (MGD) between flow-regulated sites (CH-U, CH-D) and an unregulated site (Cane-1) could not be made. Water Chemistry Water chemistry data for sampling dates in 2004 and 2005 are presented in Table 10. Water chemistry in Chewacla Creek was influenced by both point and nonpoint sources (Webber and Blevins 2001; Webber 2002, 2006). However, downstream of Lake Ogletree water chemistry was influenced primarily by the dynamics of water flow through the lake and by discharge from an Opelika limestone quarry into a tributary flowing into Lake Ogletree (Webber 2002, 2006). No differences in water chemistry were found at Cane-1, CH-U, and CH-D that indicated potential problems for the local fish fauna. Because of the presence of Lake Ogletree, lake water flowing over the spillway or through the siphon was typically low in nutrients and sediments. This was probably because plankton communities in the lake utilized much of N and P washed into the lake from the upstream watershed. In addition, heavier sediments entering with stream water settled to the bottom of the lake so that soil particles leaving the lake were greatly reduced in quantity (Webber and Blevins 2001; Webber 2002; Webber 2006). Water temperatures ranged from a low of 12.5 o C at Cane-1to a high of 15.5 o C at CH-U just downstream of Lake Ogletree (Table 10). The flow of warmer surface water (epilimnion) from the lake over the spillway into the stream in addition to periodic siphoning of warmer, epilimnetic water from the lake into the stream probably caused the higher temperatures at CH-U. Dissolved oxygen (DO) ranged from 7.50 mg/L to 10.8 mg/L across sites (Table 8). DO was highest at Cane-1 in both 2004 and 2005, with 56 lowest DO measurements recorded at CH-U both years. The variation in DO levels across sites was related to water temperature, which explains why Cane-1 had the highest DO and lowest temperature measurements (Horne and Goldman 1994). The highest water temperature and lowest DO measurements occurred at CH-U suggesting that Lake Ogletree greatly influenced hydrological and chemical properties at the site. All sites in 2004 and 2005 had sufficient oxygen at the time of sampling to meet the needs of the resident fish communities. The pH of Piedmont streams is typically slightly acidic based on the nature of soils (Boyd 2000). The values for pH ranged from 6.74 to 7.12 across sites (Table 10). In 2004 and 2005, Cane-1 was slightly acidic while CH-U and CH-D were slightly alkaline. The pH at all sites during the time of sampling was within the acceptable range of 6.5 ? 9.0 for supporting healthy fish communities in 2004 and 2005 (Boyd 2000). N and P levels were low at all 3 sites in 2004 and 2005 (Table 10). Nitrate (NO 3 - N), ammonium (NH 3 -N), orthophosphate (O-PO 4 ) and total phosphorous (T-PO 4 ) concentrations at the two Chewacla Creek sites were similar to those measured at the Cane Creek site in both 2004 and 2005. NO 3 -N levels ranged from 0.012 mg/L to 0.09 mg/L across sites in 2004 and 2005. NH 3 -N levels ranged from 0.006 mg/L to 0.095 mg/L across sites in 2004 and 2005. T-PO 4 levels ranged from 0.006 mg/L to 0.02 mg/L across sites in 2004 and 2005. Total phosphorous concentrations at all sites were within the acceptable range of 0.01 to 0.05 mg/L for natural waters at all sites in 2004 and 2005 (Wetzel 2001). O-PO 4 levels were a fraction of the T-PO 4 levels at all sites Turbidity was variable across sites with a high reading observed at Cane-1 in 2004 (Table 10). However, turbidity levels at CH-U and CH-D were low both years. In 57 2005, turbidity measured in Cane-1 was close to the low turbidity measurements observed at CH-U and CH-D. The high turbidity but low TSS measured in 2004 at Cane- 1 was difficult to explain, suggesting that a rain event prior to sampling had occurred on the watershed upstream of the sampling site (Webber 2006). The low turbidity and TSS levels at CH-U and CH-D both years reflected the influence of Lake Ogletree. Heavier solids settled in the lake and were not discharged downstream. Alkalinity, conductivity, and hardness were higher at CH-U and CH-D than at Cane-1 both years (Table 10). In fact, total alkalinity and hardness levels at CH-U and CH-D were twice as high as those at Cane-1 both years. Conductivity levels were more than 20 times higher at CH-U and CH-D than at Cane-1 in 2004 and almost twice as high in 2005. The elevated alkalinity, conductivity and hardness measurements may have resulted from discharge from an Opelika limestone quarry into a tributary of Chewacla Creek above Lake Ogletree (Webber and Blevins 2001; Webber 2002, 2006). Water from limestone quarries is relatively high in alkalinity, often exceeding 100 mg/L CaCO 3 (Webber 2006). However, these measurements were not high enough to have deleterious effects on the resident fish assemblages at either site. The alkalinity and hardness measurements observed at Cane-1 were typical of relatively undisturbed Piedmont streams (Boyd 2000). The conductivity measurement at Cane-1 was unusually low (5.0 ?S) in 2004, suggesting that a rain event occurred prior to sampling in the watershed above the sampling site. Webber (2002, 2006) reported that the natural baseline conductivity at Cane-1 is ~ 50 ?S. The conductivity measurement at Cane-1 (63.0 ?S) in 2005, although slightly elevated, was closer to conductivity levels typically observed at the site 58 Fish Data A total of 1,647 fish representing 7 families and 35 species were collected over the 2-year study from the three sites (Table 11). Cyprinidae (carps and minnows), Centrarchidae (sunfish), Catostomidae (suckers), and Percidae (perches) were the dominant families in all samples, composing 97.3% of the total number of fish. The remaining 2.7% belonged to Ictaluridae (freshwater catfishes), Fundulidae (topminnows and killifishes), and Cottidae (sculpins). Of the total fish captured, 45.8% were centrarchids, 30.2% were cyprinids, 15.1% were catostomids, and 6.2% were percids. A comprehensive list of fish species collected at all sites with associated tolerance classes, feeding guilds, and breeding guilds is presented in Table 12. Nine species were common to all sites, including the largescale stoneroller Campostoma oligolepis, blacktail shiner Cyprinella venusta, Alabama hogsucker Hypentelium etowanum, speckled madtom Noturus leptacanthus, green sunfish, bluegill sunfish Lepomis macrochirus, longear sunfish, spotted bass, and speckled darter Etheostoma stigmaeum (Table 12). Six species, including the weed shiner Notropis texanus, spotted sucker Minytrema melanops, channel catfish, shadow bass Ambloplites ariommus, redear sunfish Lepomis microlophus, and largemouth bass were found only at CH-U. The banded sculpin Cottus carolinae was found only at CH-D. Six species were unique to Cane-1, including the Tallapoosa shiner Cyprinella gibbsi, lined chub Hybopsis lineapunctata, striped shiner Luxilus crysocephalus, bluehead chub Nocomis leptocephalus, rough shiner Notropis baileyi, and silverjaw minnow Notropis buccatus. Index of Biotic Integrity (IBI) and Index of Well-Being (Iwb) 59 Assessment of Cane Creek as a Reference Site. Cane Creek (Cane-1) was chosen as a reference site for this study, based on (1) a high proportion of its watershed in forest (>70%) and correspondingly low proportions in urban/suburban (<1%) and agriculture (< 6%) areas (Webber 2002), (2) a diverse macroinvertebrate assemblage (Bayne et al. 2004), and (3) historically good water quality and physical condition (Webber and Blevins 2001, Webber 2002, 2006). However, because it was the only reference site used for this study, it was necessary to compare fish data collected at the site with data collected from other established Piedmont reference sites. For comparison with Cane-1, ADEM data from designated Piedmont reference sites including Hatchet Creek (HATC- 3), Weogufka Creek (WGFC-1), Hurricane Creek (HCR-1), Emuckfaw Creek (EMKT- 14) and Cornhouse Creek (CRHR-9) were used (D. Miller, ADEM, unpublished data). Comparing data sets between Cane-1 and the 5 ADEM reference site, it appeared that Cane-1compared favorably with the ADEM sites in terms of fish species richness and diversity. Cane-1 exhibited high cyprinid richness and diversity similar to data from the ADEM sites. Furthermore, samples from Cane-1 contained high numbers of centrarchid species with relatively low abundance per species, similar to the ADEM samples, indicating an equitable distribution of individuals among species and low dominance, typical of relatively undisturbed streams (GA DNR 2005). Schlosser (1982) noted that balanced, healthy assemblages in streams typically exhibit high cyprinid and centrarchid diversity. One area where the Cane-1 did not compare as favorably with the ADEM sites was the total number of darter species collected. At Cane-1 only 3 darter species, the 60 speckled darter, the Tallapoosa darter and the blackside darter were collected whereas at the ADEM sites, typically ? 5darter species were found, including the blackbanded darter, greenbreast darter Etheostoma jordani, speckled darter, rock darter Etheostoma rupestre, lipstick darter Etheostoma chuckwachattee, Tallapoosa darter, muscadine darter Percina sp., and the bronze darter Percina palmaris. Despite fewer darter species present, the data from Cane-1 compared reasonably well with ADEM reference data such that the site provided reasonable reference data for measuring environmental degradation at CH-U and CH-D. Metric Selection for IBI. Reference data from Cane Creek and ADEM sites in addition to previous applications of the IBI was used to select the 12 core metrics for this study. Percent tolerant species, % Lepomis, and % benthic invertivores were included as metrics because they were commonly used in other IBI studies and added trophic composition information to the IBI. Commonly used madtom, darter, and sculpin metrics were combined into a single metric, % benthic invertivores, for 2 reasons. First, the scarcity of these fishes in samples at all sites necessitated their combination into a single metric. Second, as benthic fishes they often respond to environmental disturbances similarly, allowing them to be grouped together (GA DNR 2005). Percent DELTs and % hybrids, 2 metrics used frequently in IBI studies, were combined into a single metric because of their scarcity in samples at all sites. Two commonly used metrics in IBI studies, % insectivorous cyprinids and number of native sunfish species were not included because they correlate with the number of native cyprinid and % Lepomis metrics, respectively (O?Neil and Shepard 61 1998; ADEM 1999; GA DNR 2005),. It was important to select metrics that were not redundant and presented unique information to the IBI index (Dauwalter et al. 2003; Bramblett et al. 2005). Finally, % non-lithophilic spawners was chosen as a trophic composition metric over the more commonly used % lithophilic spawners metrics. O?Neil et al. (2006) found that the % non-lithophilic spawners metric better discriminated most from least disturbed sites in streams in both small and large watersheds in the Tallapoosa and Coosa river drainages than did % lithophilic spawners. CH-U and CH-D were not heavily impacted by sedimentation based on low TSS and turbidity levels at the 2 sites both years (Table 10); this metric was not very useful in discriminating across sites. Rather, the % non-lithophilic spawners metric was more suitable for my IBI because fishes in this breeding guild exhibit a more diverse array of spawning behaviors as well a greater variety of spawning substrates than lithophilic spawners (Boschung and Mayden 2004). Thus, the % non-lithophilic spawners metric can detect more broad disturbances across sites such as changes in flow regime that reduce habitat diversity resulting in loss of instream fish cover (O?Neil et al. 2006). Comparison of IBI and Iwb Scores Across Sites. IBI metrics, raw metric scores and IBI scores for Cane-1, CH-U, and CH-D for 2004 and 2005 are presented in Table 13. In addition, IBI scores with associated integrity classes for each site are presented in Table 14. IBI scores ranged from 29.5 to 83.3 across sites in 2004 and from 35.6 to 69.3 across sites in 2005 (Tables 13 and 14). Iwb scores ranged from 6.0 to 13.5 across sites in 2004 and from 6.4 to 12.1 across sites in 2005 (Table 14). 62 At Cane-1, IBI scores were 83.3 and 69.3 in 2004 and 2005, respectively, placing the site in the ?Excellent? integrity class in 2004 and the ?Good? integrity class in 2005. Moreover, Iwb scores were 13.5 and 12.1, respectively placing the site in the ?Excellent? class both years. The fish assemblage at Cane-1 was characterized by high cyprinid diversity, relatively low abundance of tolerant fishes, presence of intolerant species and high habitat diversity. Shields et al. (1995) found sites with high biotic integrity were associated with greater habitat diversity, particularly at sites with deep pools and abundant woody debris. The high sunfish diversity in Cane-1 was apparently indicative of the presence of deep pools with abundant instream cover. Also, the presence of several intolerant species, lower numbers of tolerant fishes, and no fishes with DELTs or hybrids in Cane-1 suggested minimal site impacts in comparison with CH-U and CH-D. The tolerant species collected in Cane-1 both years (metric score = 40.8 and 0, respectively) included the largescale stoneroller, bluehead chub, striped shiner, longear sunfish, redbreast sunfish, bluegill sunfish and green sunfish (Table 12). However, because these species were observed in relatively low numbers in 2004, their presence most likely represented natural abundance that occurs in relatively undisturbed Piedmont streams rather than a response to some perturbation. In 2005, IBI and Iwb scores at Cane-1 were lower than in 2004. The decrease in scores of both indices was related to increases in relative proportions of tolerant species, % Lepomis species (metric score = 63.3 and 0, respectively), and % omnivores and herbivores (metric score = 25.5 and 29.5, respectively) and decreases in % non- lithophilic spawners (metric score = 83.0 and 60.1, respectively) and number of individuals per 100 m (metric score = 100 and 50.0, respectively). These increases in 63 relative proportions of tolerant fishes, omnivores and herbivores, and Lepomis species can indicate a disturbance to a stream; furthermore, degradation of stream habitats often results in lower abundances of fishes in streams (Kovalak 1981; Karr et al. 1986; Schleiger 2000). Schleiger (2000) found that tolerant fishes increased in Georgia Piedmont and Coastal Plain streams with degraded water quality. However, water quality in Cane-1 was characteristic of relatively undisturbed Piedmont streams both years (Boyd 2000) (Table 10). Degraded water quality and physical habitat in a stream often results in a decrease in diversity of the macroinvertebrate food base for fishes (GA DNR 2005). However, Webber (2006) reported high EPT (Ephemeroptera, Plecoptera, Tricoptera) richness (23 taxa) collected at Cane-1 in 2005, suggesting that a diverse macroinvertebrate assemblage was present at the site during this study. Furthermore, there were increases in relative proportions of intolerant fishes and benthic invertivores at Cane-1 in 2005; high proportions of intolerant and benthic species are often characteristic of healthy fish assemblages in streams (Karr et al. 1986; Ahle and Jobsis 1996). Based on water quality, macroinvertebrate, and fish data from the site, it appeared that the lower IBI and Iwb scores were probably related to natural year-to-year variation in fish assemblage structure and function at Cane-1. Although logging did occur in the Cane Creek watershed during this study, the effects on the fish assemblage were not measured. Thus, the impact of logging on the fish assemblage at Cane-1 was unknown. The high IBI and Iwb scores observed at Cane-1, particularly in 2004, typically occur at ?high-quality? sites and illustrated ecosystem stability based on other studies (Karr et al. 1987; Ahle and Jobsis 1996). However, IBI and individuals metric scores were lower in 2005, suggesting either high natural variation in fish assemblage structure 64 and function, or the presence of some disturbance, possibly related to logging in the watershed. The fish assemblage in Cane Creek was characteristic of an unmodified stream system having a natural flow regime with a relatively high number of intolerant species, relatively low abundance of tolerant species, high cyprinid richness, and low numbers of lentic species such as Lepomis spp., spotted bass, and largemouth bass. In addition, the redeye bass was present at Cane-1. The redeye bass is considered a relatively intolerant species (Schleiger 2000; Boschung and Mayden 2004), and is typically found in higher numbers than either the spotted bass or the largemouth bass in free-flowing streams (Boschung and Mayden 2004). In contrast, the redeye bass was absent in stream reaches at CH-U and CH-D, both of which have exhibited historically modified flow regimes. At CH-U, IBI scores were 37.0 and 35.6 in 2004 and 2005, respectively, placing the site in the ?Poor? integrity class both years. Moreover, Iwb scores were 9.4 and 7.8, respectively, placing the site in the ?Good? class in 2004 and the ?Fair? class in 2005. Thus, IBI and Iwb scores showed that CH-U had low biotic integrity when compared with Cane-1 (Tables 13 and 14), even though the physical habitat score at CH-U was similar to that at Cane-1 (Table 9). In comparison with Cane-1, higher abundances of tolerant Lepomis species, in particular, bluegill and longear sunfish (metric score = 0 for 2004 and 2005), coupled with no intolerant fishes (metric score = 0 for 2004 and 2005) contributed most to the low IBI scores at CH-D (Table 13). High abundances of Lepomis and Micropterus species were present at CH-U compared to Cane-1 and CH-D, probably because of the presence of large pools both upstream and downstream of CH-U. These local pools possibly served as sources of fish colonists to the site, particularly for lentic 65 species like the bluegill and longear sunfish and the largemouth bass. In addition, these pools may have offered refuge for these species during the periodic dewatering of the channel that occurred prior to 2003 such that local extinction was minimized. Once consistent flows returned to the reach, fishes in these pools might have been able to repopulate the site quickly. Other studies have found similar occurrences in which the proximity of pools to stream sites highlighted the role of pools as sources of immigration and refuge for fishes during severe disturbance events such as drought (Meador and Matthews 1992; Hoyt et al. 2001; Medeiros and Maltchick 2001). In contrast, there were few flow-dependent minnow and darter species collected at CH-U except for the blacktail shiner, and two tolerant species, the blackbanded darter and the largescale stoneroller. Because there was no pre-drought fish data available, it was possible that minnows and darters were naturally scarce at CH-U before dewatering of the study reach occurred. However, the relative absence of these fishes was still surprising considering there was gravel/cobble riffle habitat present at CH-U during the study. This suggested that during the periodic dewatering that occurred prior to 2003, minnow and darter species that perhaps were present at the site disappeared, either desiccating in the dry riffles or becoming trapped in isolated pools and consumed by larger predatory fishes or terrestrial predators. Furthermore, once consistent flows were present in 2003, flow-dependent species appeared to colonize more slowly at CH-U than lentic species such as Lepomis spp., spotted and largemouth basses, and the largescale stoneroller. At CH-D, IBI scores were 29.5 and 43.5 in 2004 and 2005, respectively, placing the site in the ?Poor? integrity class in 2005 and the ?Fair? integrity class in 2005. 66 Moreover, Iwb scores were 6.0 and 6.4, respectively, placing the site in the ?Poor? class both years. Thus, IBI and Iwb scores indicated that CH-D exhibited low biotic integrity compared with Cane-1 (Tables 13 and 14). Higher IBI scores in 2005 indicated some improvement in the condition of the fish assemblage at CH-D, compared to Cane-1. Higher sucker richness (metric score = 66.7), an increase in the relative proportion of benthic invertivores (metric score = 100) and decreases in relative proportions of omnivores and herbivores (metric score = 76.4) and diseased fishes/hybrids (metric score = 100) contributed most to the improved IBI scores in 2005. There were several possible reasons for the low IBI and Iwb scores at CH-D. First, the complete absence both years of intolerant species contributed to lower IBI scores at the site. In terms of available habitat, CH-D exhibited no deep pools and little instream cover, especially undercut banks. Uniform shallow habitat, as was observed at CH-D, often results in simple fish assemblages usually dominated by 1 or 2 opportunistic species (Schlosser 1987; Ahle and Jobsis 1996). This pattern was evident at CH-D, as 2 tolerant Lepomis species, the bluegill and longear sunfish dominated the assemblage. Like CH-U, there were few minnows and darters present at CH-D either year, with the exception of the largescale stoneroller and blackbanded darter. Riffle habitat at CH-D was less abundant that at CH-U, so the low richness of minnow and darter species might have represented natural scarcity at the site rather than slow recovery from dewatered conditions. Finally, the dewatering of the creek channel at CH-D caused by periodic drought that occurred from 1999 to 2001 and sinkhole development may have affected fish biotic integrity resulting in low IBI scores, compared with Cane-1. An unstable flow 67 environment can disrupt fish assemblage balance resulting in dominance by a single family or trophic group (Gorman and Karr 1987; Ahle and Jobsis 1996). Unlike Cane-1, CH-U and CH-D had lower abundances of intolerant species, higher abundances of tolerant species in particular the bluegill sunfish, longear sunfish, blackbanded darter, and largescale stoneroller, and lower numbers of flow-dependent minnows and darters. In addition, there were higher numbers of lentic species such as the yellow bullhead, channel catfish, largemouth bass, spotted sucker, black redhorse, redbreasted sunfish, redspotted sunfish, largescale stoneroller, and blacktail shiner, which are more common in impounded streams (Mettee et al. 1996; Boschung and Mayden 2004). The relative absence of flow-dependent fishes such as darters and minnows is characteristic of streams with highly altered flow regimes impacted by drought, impoundment or other disturbance (Mammoliti 2002). Samples collected at all 3 sites indicated that fish assemblage structure at CH-U and CH-D was dramatically different from that at Cane-1, a stream exhibiting a natural flow regime. The fish assemblages at both CH-U and CH-D have been slow to recover to that present in a healthy, unmodified Piedmont stream such as Cane Creek, even though consistent flows have been present in the reach between Lake Ogletree and the quarry since 2003. This suggested that Lake Ogletree has influenced fish assemblage structure and function at CH-U and CH-D and that both sites support fish assemblages characteristic of regulated streams (Mammoliti 2002). Statistical Assessment of Metrics. In 2004, the K-W test showed that group medians were significantly different among the 3 sites (K-W, ? 2 = 13.43, df =2, p = 0.0012). The S-D test showed that significant pairwise differences existed between the 68 Cane-1and CH-U (p ? 0.01) and Cane-1 and CH-D (p ? 0.01) but not between CH-U and CH-D (p > 0.05) in 2004. In 2005, group medians were not significantly different among the 3 sites (? 2 = 3.54, df = 2, p = 0.170). The K-W and S-D tests showed that in 2004, the groups of metric scores among the 3 sites were different. Based on the K-W and S-D tests for 2005, it appeared that the groups of metric scores among the 3 sites were more similar. This occurred because of a decline in the number of cyprinids as well as an increase in the number of tolerant and Lepomis species at Cane-1, suggesting that the fish assemblage was being impacted, perhaps from increased sediment input to the stream as a result of logging in the upper watershed. Not surprisingly, K-W and S-D tests showed that fish assemblages at CH-U and CH-D were similar in terms of trophic structure. The main differences in the fish assemblages between CH-U and CH-D were in terms of total abundance and species richness. Samples from CH-U contained higher species richness and overall abundance of fishes than at CH-D suggesting that proximity to Lake Ogletree, and differences in physical habitat contributed to differences in the fish assemblages between the sites. F-test results for raw metric scores showed that 5 of the 12 metrics effectively discriminated among the 3 study sites (Table 15), including 3 species richness and composition metrics (number of native fish species, number of native cyprinid species, number of intolerant species) and 2 trophic composition metrics (% top carnivores, % benthic invertivores). Two additional metrics (number of native round-bodied sucker species, % non-lithophilic spawners) also gave good discrimination among the 3 sites; however, the differences in group means were close but not statistically significant (F 2,3 = 69 8.96, p = 0.054 and F 2,3 = 0.081, p = 0.081). Two trophic composition metrics and both fish abundance and condition metrics discriminated poorly among the 3 sites (p > 0.05). Based on F-test results, it was evident that species richness and composition metrics (number of native fish species, number of native cyprinid species, number of intolerant species) contributed the most to the high variation in IBI scores among the 3 sites. At CH-U and CH-D, the numbers of cyprinid species were low compared to Cane-1. Also, the relative abundance of Lepomis species, particularly, bluegill and longear sunfish, was much higher compared to the reference site. These above differences were likely related to the modification of the natural flow regime in the study reach in Chewacla Creek compared with the natural flow regime in Cane Creek. Based on the data from this study between 1999 and 2001, the fish assemblages at CH-U and CH-D had very few riffle-dwelling cyprinids and high abundances of Lepomis species occupying the small, isolated pools that were present at both sites during the dewatering of the creek channel. Larimore et al. (1959) and Deacon (1961) also found similar distributions of stream fishes with low numbers of cyprinid species and higher numbers of more lentic fishes including Lepomis species in study streams following drought and in regulated streams. As consistent flows have been reestablished in my study reach since 2003, habitat generalists such as bluegill, longear and other Lepomis species capable of movement from permanent pools quickly repopulated both sites. Species that occupied restricted habitats, such as cyprinids, have been slow to increase in numbers at both CH-U and CH-D now that consistent flows are present in the creek channel between Lake Ogletree and the quarry. 70 The IBI and Iwb are useful tools for measuring biotic integrity in streams because they summarize a large amount of data into a single index value (Hughes 1990). The Iwb may have overestimated biotic integrity at CH-U, which was characterized by high abundance and species richness, but also high dominance, low evenness, and low diversity, because of the Iwb?s high dependence on abundance and biomass data (Table 16). In contrast, IBI scores rated biotic integrity lower than Iwb scores at CH-U both years, suggesting that the IBI better reflected the high dominance and overall low trophic diversity at the site. At CH-D, low Iwb scores both years reflected the low abundance and biomass at the site; however, the IBI rated biotic integrity higher than the Iwb in 2005 suggesting moderate improvement in the fish assemblage at the site. The IBI may have rated biotic integrity at CH-D higher than the Iwb because it was more sensitive to the trophic composition of the fish assemblage. The IBI offered one major advantage over the Iwb in measuring biotic integrity at each site. Although the IBI measured both abundance and species richness, it also incorporated functional aspects of the fish assemblage including breeding guilds, feeding guilds, and tolerance classes, which the Iwb did not. As a result, the IBI was perhaps more sensitive to fish assemblage response to disturbance than the Iwb and probably offered a truer reflection of fish assemblage condition at each site. However, the Iwb incorporated biomass, an important component of the fish assemblage. Because these indices incorporated different aspects of the fish assemblages, their combined use appeared to accurately assess fish assemblage condition across sites and years. Analysis of Fish Assemblage Structure Across Sites and Years 71 Total abundance (N) varied considerably across sites and years, ranging from 114 to 537 captured individuals (Table 16). Species richness (S) varied across sites and years, ranging from 13 to 21 captured species in samples (Table 16). At Cane-1, total abundance decreased considerably from 327 to 114 captured individuals between 2004 and 2005 whereas species richness was the same both years (19 species). It was unclear exactly why there was such a large decline in total numbers of fishes collected at Cane-1 in 2005. Sampling methodology was the same both years, and there was no difference in collecting efficiency. The data from Cane-1 suggested that the large between-year difference in abundance represented natural variation in the fish assemblage rather than a response to disturbance. At the upstream site (CH-U), total abundance and species richness decreased from 537 to 332 captured individuals, and 21 to 17 species, respectively, at CH-U. The high abundance at CH-U, particularly in 2004, resulted from high numbers of fishes commonly found in regulated streams downstream of an impoundment including the largescale stoneroller, blacktail shiner, spotted sucker, black redhorse, bluegill sunfish, and longear sunfish (Table 12). The presence of these species in relatively high numbers suggested that Lake Ogletree was influencing the fish assemblage at CH-U. In 2005, the decline in total abundance and species richness at CH-U most likely represented natural variation in the fish assemblage rather than a response to disturbance. At the downstream site (CH-D), total abundance decreased from 186 to 151 captured individuals, whereas richness increased slightly from 12 to 13 species in 2005. The low abundance and richness at CH-D perhaps was indicative of an unhealthy fish 72 assemblage and reflected the site?s low habitat diversity, particularly the lack of undercut bank and large, deep pools. Furthermore, abundance and richness changed little from 2004 to 2005, suggesting that overall recovery of the fish assemblage from the dewatering of the channel has been slow, even though consistent flows were present in the study reach both years. Shannon diversity (H?) values were relatively low across sites and years and there was little difference in species diversity across sites (Table 16). At Cane-1, H? increased slightly from 2.35 to 2.54 between 2004 and 2005. This increase occurred because of the decline in total abundance in 2005 but richness remained the same as in 2004. Relatively low H? values for Cane-1 was surprising given the high IBI and Iwb scores both years of the study. There were no known impacts that could account for the relatively low H? values unless continued logging in the Cane Creek watershed impacted the fish assemblage, which seemed unlikely. Furthermore, because of the stable flow regime present at Cane-1, there was high habitat heterogeneity and abundant fish habitat in the form of glide riffles, deep pools, undercut bank areas and instream woody debris. In addition, Webber (2006) found a highly diverse macroinvertebrate assemblage at the site, suggesting that Cane-1was relatively undisturbed even with logging occurring in the watershed. There was a decline in H? at CH-U between 2004 and 2005, from 2.38 to 1.98, respectively, indicating a decline in fish assemblage at the site. This decrease was most likely related to the decline in both abundance and richness in 2005. At CH-D, there was no change in H? from 2004 to 2005 indicating slow recovery of the fish assemblage from 73 altered flow conditions that occurred periodically from 1999 to 2001. However, because of the lack of pre-drought fish data for CH-U, there was also the possibility that the low H? values were indicative of a naturally sparse fish assemblage at the site. Based on the interpretation of Shannon index (H?) values by Wihlm and Dorris (1968), fish assemblages at all 3 sites were moderately stressed or impacted. This contradicted the IBI and Iwb results, which showed that both CH-U and CH-D exhibited highly stressed fish assemblages compared with Cane-1. However, the Shannon diversity index did not incorporate functional attributes of the fish community such as trophic structure and tolerance class and, perhaps, was a less reliable indicator of fish assemblage condition across sites. There was relatively little difference in evenness (E) values among sites (Table 16). Surprisingly, E values were relatively low across sites, particularly at Cane-1 in 2004. Pielou (1966) noted that evenness values < 0.6 are typical for stressed biotic communities. This indicated that the fish assemblage at Cane-1 was moderately stressed in 2004 but less stressed in 2005 which contradicted IBI and Iwb scores at the site. In 2004, E values at CH-D indicated that the fish assemblage was only slightly stressed; however, IBI and Iwb scores for the site were indicative of a highly stressed fish assemblage. Also, E values were relatively high at CH-D which was contradictory to the very low H? values (1.97 and 2.05, respectively). It was unclear exactly why CH-D exhibited low diversity but relatively high evenness both years. In 2005, evenness was highest at Cane-1 because abundance was lower but richness was the same as in 2004, suggesting a more equitable distribution of species present at the site in 2005. 74 There was no clear explanation for the relatively low H? and E values both years and the low abundance in 2005 at Cane-1. Despite these low values, IBI and Iwb scores clearly showed that Cane-1 had the healthiest and most balanced fish assemblage of the three sites both years. There had been no alterations to the natural flow regime, so fish assemblage structure at Cane-1 had been largely determined by instream physical habitat. The greater habitat heterogeneity and diverse macroinvertebrate assemblage at Cane-1, compared with CH-U and CH-D, likely explained the healthier and more balanced fish assemblage present at Cane-1. Studies by Schlosser (1982, 1987, 1990), Capone and Kushlan (1991) and Hoyt et al. (2001) all reported that healthy fish assemblages were often found in stream reaches with greater diversity of habitat including deep pools, swift-flowing gravel/cobble riffles, and available instream cover in the form of woody debris, such as that present at Cane-1. Fish assemblage structure at CH-U appeared to have been largely structured by the presence of Lake Ogletree, and appeared typical of fish assemblages located downstream of an impoundment (Mammoliti 2002). In terms of overall abundance and richness, the fish assemblage at CH-U was dominated by species found in regulated streams, in particular the blacktail shiner, spotted sucker, bluegill sunfish, longear sunfish, black redhorse and largescale stoneroller. With the exception of large, deep pools, habitat diversity at CH-U was comparable to that in Cane Creek. However, during the periodic drought that occurred from 1999 to 2001, there was no water flowing throughout most of the upstream reach and the channel was almost completely dry with the exception of small, isolated pools (Cliff Webber, AU Fisheries, personal communication). It was very likely that during dewatered conditions in the channel, 75 species that favor lentic waters survived in the large pools both upstream and downstream of the study reach and then repopulated the site once flows were reestablished. In contrast, minnows and darters, fishes that require swift-flowing water, disappeared and have not repopulated the site as quickly despite the presence of gravel/cobble riffle habitat. CH-D was in a section of Chewacla Creek where numerous sinkholes were present, and, thus, during dry periods this reach was often completely dewatered. Also, habitat diversity at the site was low, lacking deep pools and undercut banks. Once consistent flows were reestablished in 2003, fish assemblage data for CH-D suggested that the assemblage had not recovered to that expected for an undisturbed stream, such as Cane Creek. Schlosser (1982) noted that streams with low habitat diversity, particularly those lacking deep pools often have fish assemblages with low abundance and richness. Other studies related richness and diversity to availability of pool habitat (Evans and Noble 1979; Meffe and Sheldon 1988; Ebert et al. 1991; Shields and Hoover 1991; Shields et al. 1994). Deep pools are important because they serve as refuge for many fish species during drought and as sources of dispersal migration when normal flows return to a stream (Medeiros and Maltchik 2001). Analysis of Fish Assemblage Function Across Sites Feeding Guilds. There were significant differences in feeding guild composition among the 3 sites in 2004 (X 2 = 266.99, df = 6, p ? 0.001) and 2005 (X 2 = 186.10, df =8, p ? 0.001) (Figure 6). However, between-year differences were not significant at any site (Wilcoxon test, p < 0.001, Figure 6). At Cane-1, invertivores and insectivores were the 76 dominant feeding guilds both years whereas invertivores were the dominant feeding guild both years at CH-U and CH-D. Cane-1 was dominated by insectivorous cyprinids in 2004 with invertivores, omnivores, herbivores and top carnivores in relatively even proportions (Figure 6). Insectivorous cyprinids are trophic specialists who feed at all levels of the water column, so high cyprinid richness often indicates a diverse aquatic macroinvertebrate assemblage is present in a stream, and, thus, a healthy fish assemblage (Niemela et al. 1999; GA DNR 2005). Webber (2006) reported high aquatic macroinvertebrate diversity at Cane-1 with total EPT (Ephemeroptera, Plecoptera, Tricoptera) taxa of 23 species in 2006 whereas EPT richness was 13, 15, and 16 taxa, respectively, in sites located in the Chewacla Creek study reach. This might explain the higher trophic complexity of the fish assemblage at Cane-1, as fishes there were exploiting a more diverse macroinvertebrate food base, compared to CH-U and CH-D. In addition, top carnivores were found in higher proportions at Cane-1 than at either CH-U or CH-D. An abundance of top carnivores (> 5%) can be indicative a healthy and trophically diverse fish community (Karr et al. 1986; GA DNR 2005). However, an overabundance of top carnivores (> 20%) can be indicative of stream degradation (GA DNR 2005). The relative proportion of top carnivores was not > 15% at any site in either 2004 or 2005. Their presence in higher proportions at Cane-1 possibly reflected the more abundant pool habitat at the site, compared to CH-U and CH-D. Schleiger (2000) reported higher numbers of top carnivores in lower Piedmont streams with abundant pool habitat. 77 In 2005, there was a shift in feeding-guild dominance from insectivores to invertivores at Cane-1. It was unclear what may have contributed to this trophic shift at Cane-1 in 2005. Often, shifts in trophic structure from highly specialized feeders such as insectivores to more generalist feeders are caused by a decrease in the diversity of the food base in a stream (Niemela et al. 1999; GA DNR 2005). However, it appeared that a relatively diverse macroinvertebrate assemblage was present at Cane-1, as a high number of EPT taxa were observed in 2005 (Webber 2006). Although logging did occur in the Cane Creek watershed during the study, water quality, macroinvertebrate and fish data from Cane-1 suggested minimal impacts to the stream. Thus, it appears that the trophic shift that occurred at Cane-1 in 2005 represented natural year-to-year variation rather than a response of the fish assemblage to disturbance. At CH-U and CH-D, trophic structure was characterized by dominance by a single feeding guild, invertivores. Insectivores, omnivores, and herbivores were observed in much lower proportions relative to invertivores at both sites. Typically, a high proportion of invertivores in a sample is indicative of a healthy fish community (Niemela and Feist 2002). However, the high abundance of invertivores in samples collected at CH-U and CH-D both years was the result of the dominance of two Lepomis species, bluegill and longear sunfish, which are habitat generalists (Boschung and Mayden 2004). An aquatic assemblage dominated by Lepomis species can be indicative of poor diversity of the macroinvertebrate food base (O?Neil and Shepard 1998; GA DNR 2005), as Webber (2006) reported for sites in Chewacla Creek between Lake Ogletree and the quarry. As food resources become less reliable in degraded stream habitats, habitat generalists often become the dominant members of the fish assemblage 78 because they are capable of exploiting a less diverse food base (Karr et al. 1986; GA DNR 2005). Thus, the high abundance of invertivores relative to other feeding guilds at CH-U and CH-D suggested degraded fish assemblages were present both years, compared to the fish assemblage at Cane-1. Lepomis species are highly resilient to stream disturbance and can thrive in degraded stream environments (Detenbeck et al. 1992). Disturbed stream sites often exhibit simple community structure with one or several species in a single family dominating the fish assemblage at the expense of species in other families (Schlosser 1987; Ahle and Jobsis 1996). Bluegill and longear sunfish were probably more abundant relative to other species at CH-U and CH-D because there was no predation pressure on them from piscivorous Micropterus species such as the spotted and largemouth bass. There were higher numbers of piscivorous predators such as the spotted and largemouth bass at CH-U because of the close proximity of Lake Ogletree; however, their numbers relative to Lepomis sunfish were so small they likely had little influence on populations of the smaller-sized Lepomis sunfish. The absence of large piscivorous fishes at the downstream site was likely related to lack of suitable habitat, particularly deep pools, and undercut banks, and woody debris to serve as cover for these ambush predators (Boschung and Mayden 2004). Tolerance Classes. The relative abundance of tolerance classes was statistically different across sites in 2004 (X 2 = 279.32, df = 6, p ? 0.001) and 2005 (X 2 = 117.08, df = 6, p ? 0.001) (Figure 7). However, between-year differences did not differ significantly at any site (Wilcoxon test, p < 0.05, Figure 7). During both years of this study samples 79 were dominated by sensitive and intolerant species. Species designated as intolerant observed at Cane-1 included the lined chub, Tallapoosa shiner, redeye bass and the blackside and Tallapoosa darters. Six tolerant species including the striped shiner, largescale stoneroller, green sunfish, longear sunfish, bluegill sunfish and yellow bullhead were observed at Cane-1 both years. However, numbers of these species most likely represented natural abundance in the fish assemblage rather than a response to some disturbance in the watershed. The higher numbers of intolerant and moderately intolerant fishes and lower numbers of tolerant fishes, compared to CH-U and CH-D, suggested a healthier fish assemblage at Cane-1 both years. At CH-U, tolerant species were numerically dominant both years, and moderately intolerant species were the second most abundant tolerance class. The relative abundance of intolerant species at CH-U was low relative to other tolerance classes both years. The high proportions of tolerant species such as the largescale stoneroller, spotted sucker, yellow bullhead, bluegill sunfish, longear sunfish and blackbanded darter suggested that these species quickly repopulated the site once normal flows were reestablished in the creek channel. In contrast, even with normal flows, moderately intolerant and intolerant minnow and darter species appear to be colonizing more slowly at CH-U. There was no change in numbers of tolerant and pioneer fishes, or in number of moderately intolerant and intolerant fishes at CH-U from 2004 to 2005. This result suggested little to no recovery of the fish assemblage from dewatered conditions at the site, as IBI and Iwb scores both years indicate. Despite the presence of consistent flows 80 in the channel since 2003, it appeared that Lake Ogletree influenced fish assemblage structure and function at CH-U more than at CH-D. At CH-D, tolerant species including the largescale stoneroller, bluegill sunfish, longear sunfish and blackbanded darter were numerically dominant both years. Pioneer species such as the largescale stoneroller, and blackbanded darter were the second most abundant tolerance class in 2004, with moderately intolerant species the second most abundant tolerance class in 2005. There were no intolerant species collected at CH-D in either 2004 or 2005. Tolerant and pioneer species were present in fewer numbers in 2004, suggesting that fish assemblage condition had improved slightly during the two- year study. Some recovery of the fish assemblage from dewatered conditions had taken place, but perhaps not to the level present at Cane-1. This assertion was supported by the IBI and Iwb scores for CH-U, showing improvement in fish assemblage condition in 2005. Summary. Results from my study showed that fish assemblages at two sites in Chewacla Creek between Lake Ogletree and the quarry (CH-U and CH-D) were highly stressed in comparison with the fish assemblage at the reference site in Cane Creek (Cane-1). Several natural and anthropogenic factors have greatly influenced fish assemblage structure and function at CH-U and CH-D, including: (1) drought that occurred periodically between 1999 and 2001, (2) sinkhole development in the channel as a result of quarry mining in the watershed, and (3) the presence of Lake Ogletree. Both CH-U and CH-D were dominated by tolerant species commonly found in disturbed streams, in particular two Lepomis species, the bluegill and longear sunfish. These 81 species quickly colonized both sites and became dominant once consistent flows were reestablished in 2003. In contrast, flow-sensitive minnow and darter species were slow to colonize both sites once consistent flows were reestablished. The fish assemblage at Cane-1 was characteristic of a relatively undisturbed Piedmont stream with a stable flow regime, good instream habitat diversity and a relatively undisturbed watershed. In terms of fish assemblage condition, the site compared well with other Piedmont reference streams and thus served as a high quality site against which fish assemblages at more disturbed sites could be measured. In comparison with Cane-1, both CH-U and CH-D have exhibited slow recovery from dewatered conditions in the study reach. Despite the presence of consistent flows in the stream channel since 2003, fish assemblages at both sites were similar to those found in highly modified or degraded streams. Higher IBI and Iwb scores at CH-D in 2005 suggested that the fish assemblage had recovered slightly from dewatered conditions, but not to that present at Cane-1. There was no recovery from dewatered conditions at CH- U, based on similar IBI and Iwb scores in both years. Although fish assemblage structure at CH-U was characterized by high abundance and species richness, a large proportion of the total abundance was tolerant species such as longear and bluegill sunfish and other lentic fishes commonly found in regulated streams. Despite the presence of consistent flows at CH-U, the fish assemblage continues to be influenced by the presence of Lake Ogletree. 82 Table 1. Index of biotic integrity (IBI) metrics originally developed by Karr et al. (1984) to assess stream fish assemblages in the midwestern United States. Category Metric Species richness and Total number of fish species composition Number and identity of darter species Number and identity of sunfish species Number and identity of sucker species Number and identity of intolerant species Proportion of individuals as green sunfish Trophic composition Proportion of individuals as omnivores Proportion of individuals as insectivorous cyprinids Proportion of individuals as piscivores Fish abundance and Number of individuals in sample condition Proportion of individuals as hybrids Proportion of individuals with disease, tumors, fin damage, and skeletal anomalies 83 Table 2. Candidate IBI metrics evaluated for use in Chewacla and Cane creeks, Lee County, Alabama. *DELTS includes fishes with diseases, erosions, lesions, and tumors Barbour et al. (1999). Category Metric Species richness and composition Number of native fish species Number of darter species Number of native sunfish species Number of native round-bodied sucker species Number of intolerant species Number of native centrarchid species Number of native minnow species Number of native benthic insectivore species Number of benthic invertivores Number of Lepomis species Number of benthic fluvial specialists Number of native families Trophic composition Proportion of individuals as green sunfish Proportion of individuals as omnivores Proportion of individuals as insectivorous cyprinids Proportion of individuals as piscivores Proportion of individuals as green sunfish + yellow bullheads Proportion of individuals as stonerollers Proportion of individuals as invertivores Proportion of individuals as benthic invertivores Proportion of individuals as omnivores + herbivores Proportion of individuals as lithophilic spawners Proportion of individuals as non-lithophilic spawners Proportion of individuals as Lepomis species Proportion of individuals as bluegill sunfish Proportion of individuals as pioneer species Proportion of individuals as benthic fluvial specialists Ratio of generalist feeders to specialist feeders Fish abundance and condition Number of individuals per 100 meters Proportion of individuals as hybrids Proportion of individuals with DELTs * Proportion of individuals with DELTs + hybrids Proportion of individuals with DELTS + hybrids + exotics Biomass (g) per 100 meters Figure 1. Map showing sampling stations on Chewacla Creek for 2004 and 2005. The locations of the Martin Marietta quarry and Lake Ogletree are included. Sampling stations (CH-U and CH-D) are shown in bold. 84 85 Table 3. Physical characteristics measured in Chewacla (CH-U, CH-D) and Cane (Cane-1) creek sites, Lee County, Alabama. Station County Location Watershed Mean Mean Mean Mean Area Elevation Stream Width Depth Velocity (km 2 ) (m) (m) (m) (m/s) CH-U Lee N 32.54702 90.7 167.1 8.7 0.25 0.17 W 85.45155 CH-D Lee N 32.54846 116.5 138.1 7 0.33 0.14 W 85.46910 Cane-1 Lee N 32.63252 17 172.1 10.9 0.34 0.13 W 85.68861 Figure 2. Map showing sampling station in Cane Creek for 2004 and 2005. Sampling station (Cane-1) is shown in bold. 86 87 Table 4. Metrics used to assess physical habitat quality at Chewacla and Cane creek sites, Lee County, Alabama during the fall 2006. Category Metric Epifaunal Substrate/Available Cover Epifaunal substrate/Available cover Embeddedness Velocity/Depth Regime Sediment Deposition Channel Morphology Channel flow status Channel Alteration Frequency of Riffles (or Bends) Riparian and Bank Structure Bank stability Bank Vegetative Protection Riparian Vegetative Zone Width Table 5. Scoring criteria used to visually estimate physical habitat quality at sites in the Chewacla and Cane Creek watersheds during the fall 2006. A total score is tabulated and compared with the reference stream (Barbour et al. 1999). Category Excellent Good Fair Poor Epifaunal Substrate/Available Cover Substrate and cover 20-16 15-11 10-6 5-0 Embeddedness 20-16 15-11 10-6 5- Velocity/Depth regime 20-16 15-11 10-6 5-0 Sediment deposition 20-16 15-11 10-6 5-0 Channel Morphology Channel flow status 20-16 15-11 10-6 5-0 Channel alteration 20-16 15-11 10-6 5-0 Frequency of riffles (or bends) 20-16 15-11 10-6 5-0 Riparian and Bank Structure Bank stability 10-9 8-6 5-3 2-0 Vegetative protection 10-9 8-6 5-3 2-0 Riparian zone width 10-9 8-6 5-3 2-0 88 89 Table 6. Assessment used to evaluate physical habitat quality at all sites in Chewacla and Cane creeks during the fall 2006 (Barbour et al. 2006). Assessment Category Percent Comparability Comparable to Reference Site ? 90% Supporting 1 75-88% Partially Supporting 60-73% Non-Supporting ? 58% 1 Potential to support an acceptable assemblage of fishes. Figure 3. Stations for measuring flow in Chewacla Creek between Lake Ogletree and Martin Marietta quarry. Stream gage stations 1-4 are operated and maintained by the AU Department of Fisheries. Station 5 is a USGS stream gage downstream from the confluence of Moores Mill and Chewacla creeks. Station 6 is a bioassessment site, not a stream gage for measuring flow. 90 91 Table 7. IBI metrics and scoring criteria developed for Chewacla (CH-U, CH-D) and Cane (Cane-1) creek sites, Lee County, AL. Watershed Expected Size Value Score Metric (km 2 ) (EV) 1 Calculation Species richness and composition Number of native fish species #NAT < 26 17 (#NAT/EV)x100 26-130 21 > 130 25 Number of native cyprinid species #CYP < 26 5 (#CYP/EV)x100 26-130 7 > 130 8 Number of native round-bodied sucker species #RBS < 26 2 (#RBS/EV)x100 26-130 3 > 130 Number of intolerant species #INT < 26 2 (#INT/EV)x100 26-130 4 > 130 6 Trophic composition % tolerant %TOL all sizes 0.10 100-(%TOL/EV)x100 % top carnivores %TC all sizes 0.04 (%TC/EV)x100 % Lepomis %LEP all sizes 0.30 100-(%LEP/EV)x100 % benthic invertivores %BI all sizes 0.20 (%BI/EV)x100 % non-lithophilic spawners %NLS all sizes 0.35 (%NLS/EV)x100 % omnivores and herbivores %OH all sizes 0.20 100-(%OH/EV)x100 Fish abundance and condition Number of fishes per 100 m #Sample all sizes 200-300 100 % DELTs + hybrids %DH all sizes 0.005 100-(%DH/EV)x100 1 EV = metric value expected for relatively healthy Piedmont streams. Expected values were determined using fish data from Cane Creek, ADEM reference data and previous applications of the IBI. 92 Table 8. Index of Well-Being (Iwb) scoring criteria and integrity classes for wadeable Piedmont streams. Adapted from GA DNR (2005) and Ohio EPA (1987). Iwb Watershed Size Integrity Score (km 2 ) class ? 8.1 < 40 Excellent ? 9.6 ? 40 8.1 - ? 7.3 < 40 Good 9.6 - ? 8.6 ? 40 7.3 - ? 5.7 < 40 Fair 8.6 - ? 6.6 ? 40 5.7 - ? 4.9 < 40 Poor 6.6 - ? 5.6 ? 40 < 4.9 < 40 Very poor < 5.6 ? 40 93 Table 9. Physical habitat assessment metrics, individual metric and total scores, percent comparison, and final habitat assessment for sites in the Chewacla (CH-U, CH-D) and Cane (Cane-1) creek watersheds, September 2006. Category CH-U CH-D Cane-1 Epifaunal Substrate/Available Cover Substrate and cover 20 18 20 Embeddedness 20 Velocity/Depth regime 15 15 18 Sediment deposition 20 20 17 Channel Morphology Channel flow status 18 20 20 Channel alteration 20 20 20 Frequency of riffles (or bends) 18 20 20 Riparian and Bank Structure Bank stability 20 20 20 Vegetative protection 20 20 20 Riparian zone width 20 18 20 Total score 191 191 195 Percent comparison with reference stream 97.9 97.9 Habitat assessment C 1 C 1 N = non-supporting; C = comparable to reference site; S = supporting. Figure 4. Hydrographs showing mean daily discharge for the upstream and downstream Chewacla Creek sites from August 2004 to April 2006. 94 95 Table 10. Physicochemical data measured at sites in Chewacla (CH-U, CH- D) and Cane (Cane-1) creeks during the fall of 2004 and 2005. 2004 2005 Variable CH-U CH-D Cane-1 CH-U CH-D Cane-1 Dissolved Oxygen (mg/L) 8.2 10.0 10.8 7.50 8.20 9.80 Temperature ( o C) 15.5 14.0 12.5 18.5 18.0 15.8 pH 7.05 7.1 6.74 7.01 7.12 6.71 Conductivity (?S) 128.0 137.8 5.0 106.6 126.9 65.3 Turbidity (NTU) 4.40 1.39 62.10 3.60 1.03 3.00 NO 3 -N (mg/L) 1 0.063 0.09 0.058 0.051 0.079 0.012 NH 3 -N (mg/L) 1 0.027 0.006 0.095 0.04 0.02 0.03 O-PO 4 (mg/L) 2 0.004 0.002 0.000 0.004 0.005 0.004 T-PO 4 (mg/L) 2 0.02 0.006 0.006 0.02 0.01 0.02 Total alkalinity (mg/L) 52.50 62.50 26.25 46.25 56.25 27.50 Total hardness (mg/L) 45.65 61.91 19.49 51.05 55.14 29.61 TSS (mg/L) 2 4.34 0.52 2.96 3.84 1.09 2.38 96 Table 11. Total abundance, percent relative abundance and total number of fishes by family collected at sites in the Chewacla (CH-U, CH-D) and Cane (Cane-1) creek watersheds, 2004 and 2005. Family CH-U CH-D Cane-1 % Cyprinidae 166 69 262 30.2 Catostomidae 151 24 74 15.1 Ictaluridae 18 4 7 1.8 Cottidae 0 5 0 0.3 Fundulidae 6 5 0 0.67 Centrarchidae 491 174 89 45.8 Percidae 37 56 9 6.2 Total 869 337 441 100.0 97 Table 12. Comprehensive list of fish species including species abundance and total abundance for sites in Chewacla and Cane creeks during the two-year study period, 2004 and 2005. TC = tolerance class, FG = feeding guild, P = pioneer, INT = intolerant, TOL = tolerant, MI = intermediate. CH-U, CH-D = upstream and downstream Chewacla Creek sites, Cane- 1= reference site in Cane Creek. Species names follow Boschung and Mayden (2004). Family Common Name TC FG CH- U CH- D Cane- 1 Cyprinidae Campostoma oligolepis Largescale stoneroller * P HER 1 106 59 14 Cyprinella gibbsi Tallapoosa shiner INT INS 0 0 33 Cyprinella venusta Blacktail shiner MI INS 53 10 57 Hybopsis lineapunctata Lined chub INT INS 0 0 10 Luxilus crysocephalus Striped shiner TOL INS 0 0 27 Nocomis leptocephalus Bluehead chub P O 0 0 14 Notropis baileyi Rough shiner MI INS 0 0 103 Notropis buccatus Silverjaw minnow MI INS 0 0 4 Notropis texanus Weed shiner MI INS 7 0 0 Catostomidae Hypentelium etowanum Alabama hogsucker MI O 20 23 33 Minytrema melanops Spotted sucker TOL O 29 0 0 Moxostoma duquesnei Black redhorse MI INV 94 0 41 Moxostoma poecilurum Blacktail redhorse MI INV 8 1 0 Ictaluridae Ameiurus natalis Yellow bullhead TOL O 13 0 2 Ictalurus punctatus Channel catfish MI O 4 0 0 Noturus funebris Black madtom MI INV 0 0 4 Noturus leptacanthus Speckled madtom MI INS 1 4 1 Fundulidae Fundulus olivaceous Blackspotted topminnow MI INS 6 5 0 Cottidae Cottus carolinae Banded sculpin MI TC 0 5 0 Centrarchidae Ambloplites ariommus Shadow bass INT TC 1 0 0 Lepomis auritus Redbreast sunfish MI INV 43 3 15 Lepomis cyanellus Green sunfish TOL/P TC 2 2 10 Lepomis gulosus Warmout MI TC 0 0 4 Lepomis macrochirus Bluegill sunfish TOL INV 251 64 5 Lepomis megalotis Longear sunfish TOL INV 124 87 36 Lepomis hybrid TOL IN 1 0 0 Lepomis microlophus Redear sunfish MI INV 4 0 0 Lepomis miniatus Redspotted sunfish MI INV 47 16 0 Micropterus coosae Redeye bass INT TC 0 0 16 Micropterus salmoides Largemouth bass MI TC 6 0 0 Micropterus punctulatus Spotted bass MI TC 12 2 3 Table 12. ? continued. 98 Table 12. ? continued. Percidae Etheostoma stigmaeum Speckled darter MI INS 17 13 5 Etheostoma tallapoosae Tallapoosa darter INT INS 0 0 1 Percina maculata Blackside darter INT INS 0 0 3 Percina nigrofasciata Blackbanded darter TOL/P INS 20 43 0 Total Count 2 869 337 441 1 species abundance, 2 total abundance by site 99 Table 13. IBI metrics, raw metric scores, and total IBI scores for sites in Chewacla and Cane creeks, 2004 and 2005. CH-U, CH-D = upstream and downstream sites in Chewacla Creek, Cane-1 = reference site in Cane Creek. CH-U CH-D Cane-1 Metric 2004 2005 2004 2005 2004 2005 Species richness and composition Number of fish species 95.2 76.2 57.1 51.9 100 100 Number of cyprinid species 42.9 28.6 28.6 28.6 100 100 Number of round-bodied sucker 100 100 33.3 66.7 100 100 species Number of intolerant species 25.0 0 0 0 100 100 Trophic composition % tolerant 0 0 0 0 40.8 0 % top carnivores 60.3 60.2 46.0 19.4 100 100 % Lepomis 0 0 0 0 63.3 0 % benthic invertivores 27.0 13.6 86.0 100 87.2 92.1 % non-lithophilic spawners 39.5 16.3 9.2 29 83.0 60.1 % omnivores and herbivores 54.0 82.0 43.6 76.4 25.5 29.5 Fish abundance and condition Number of fishes per 100 m 0 50.0 50.0 50.0 100 50.0 % DELTs + hybrids 0 0 0 100 100 100 ? metric scores 443.9 426.9 353.8 522.0 999.8 831.7 IBI Score 1 37.0 35.6 29.5 43.5 83.3 69.3 1 IBI Score = ?metric scores / 12 100 Table 14. IBI and Iwb scores with associated integrity classes for all sites in Chewacla (CH-U, CH-D) and Cane (Cane-1) creeks, 2004 and 2005. 2004 2005 Site Score Integrity Score Integrity Class Class IBI CH-U 37.0 Poor 35.6 Poor CH-D 29.5 Poor 43.5 Poor Cane-1 83.2 Excellent 66.9 Good Iwb CH-U 9.4 Good 7.8 Fair CH-D 6.0 Poor 6.4 Poor Cane-1 13.5 Excellent 12.1 Excellent 101 Table 15. Core IBI metrics, and F-test results between raw metric values across sites. Metric F-value p-value Species richness and composition No. native species 5.25 0.102 No. native cyprinid species 39.38 0.006 No. native round-bodied sucker species 15.56 0.024 No. intolerant species 1.83 0.324 Trophic composition % tolerant 5.10 0.106 % top carnivores 1.61 0.363 % Lepomis 18.28 0.019 % benthic invertivores 32.56 0.009 % non-lithophilic spawners 7.34 0.067 % generalist feeders 16.51 0.022 Fish abundance and condition No. of individuals in sample per 100 m 2.73 0.211 Percent DELTs + hybrid species 2.09 0.270 102 Table 16. Descriptive measures of fish assemblage structure at sites in the Chewacla (CH-U, CH-D) and Cane (Cane-1) creek watersheds, 2004 and 2005. 2004 2005 CH-U CH-D Cane-1 CH-U CH-D Cane-1 Abundance (N) 521 186 327 332 148 114 Richness (S) 21 12 19 17 13 19 Diversity (H?) 2.38 1.97 2.35 1.98 2.05 2.54 Evenness (E) 0.514 0.600 0.553 0.429 0.600 0.674 Figure 6. Feeding guild composition ? percent total number of fishes collected by feeding guild at sites in Chewacla and Cane creeks, 2004 and 2005. 103 Figure 7. Tolerance class composition ? percent total number of fishes collected by tolerance class in Chewacla and Cane creeks, 2004 and 2005. 104 105 LITERATURE CITED Aadland, L.P. 1993. Stream habitat types: their fish assemblages and relationship to flow. North American Journal of Fisheries Management. 13: 790-806. Alabama Department of Environmental Management (ADEM). 1999. Monitoring of watersheds associated with Alabama state parks utilizing physical, chemical, and biological assessments. Environmental Indicators Section. Field Operations Division. Montgomery, Alabama. Alabama Department of Industrial Relations. 2003. Surface mining of non-fuel minerals. FY 2002 Annual Report. Montgomery, Alabama. Alatalo, R.V. 1981. Problems in the measurement of evenness in ecology. Oikos 37: 199- 204. Ahle, R.C. and G. J. Jobsis. 1996. Using an index of biotic integrity to assess gold mine impacts on streams. Proc. Annu. Conf. Southeast. Assoc. Fish and Wildl. Agencies 50: 38-50. Allan, R.J. 1995. Impact of mining activities on the terrestrial and aquatic environment with emphasis on mitigation and remedial measures. Heavy Metals: problems and solutions. Springer-Verlag. New York. pp. 119-140. Angermeier, P.L. and J.R. Karr. 1986. Applying an index of biotic integrity based on stream fish communities: considerations in sampling and interpretation. North American Journal of Fisheries Management 6: 418-429. Angermeier, P.L. 1989. Spatio-temporal variation in habitat selection by fishes in small Illinois streams, in W.J. Matthews and D.C Heins (Eds.), Community and Evolutionary Ecology of North American Stream Fishes, Univ. of Oklahoma Press, pp. 52-60. Angermeier, P.L., R.A. Smogor, and J.L. Stauffer. 2000. Regional frameworks, and candidate metrics for assessing biotic integrity in Mid-Atlantic Highland streams. Trans. Am. Fish. Soc. 129: 962-981. 106 APHA. 1998. Standard methods for the examination of water and wastewater. American Public Health Association, American Water Works Association and Water Pollution Control Federation. 20 th ed. Washington, D.C. Bain, M.B., J.T. Finn, and H.E. Booke. 1988. Streamflow regulation and fish community structure. Ecology 69: 382-392. Barbour, M.T., J. Gerritsen, B.D. Snyder, and J.B. Stribling. 1999. Rapid Bioassessment Protocols for Use in Streams and Rivers: Periphyton, Benthic Macroinvertebrates, and Fish. 2 nd Ed. EPA 841-B-99-02. U.S. Environmental Protection Agency. Office of Water. Washington, D.C. Bayne, D.R., E.M. Reutebuch, W.C. Seesock, E.C. Webber. 2004. Nutrient and sediment loading in Sougahatchee Creek and the impacts on aquatic biota. Report submitted to West Point Stevens and the Cities of Auburn and Opelika, Alabama. Department of Fisheries and Allied Aquacultures. Alabama Agriculture Experiment Station. Auburn University. Bell, D.E. and F.E. Payne. 1993. Mining. Pages 197-207 in C.F. Bryan and D.A. Rutherford, eds. Impacts on warmwater streams: guidelines for evaluation. South. Div. Am. Fish. Soc. Little Rock, Arkansas. Benke, A.C. 1990. A perspective on America?s vanishing streams. Journal of the North American Benthological Society 9: 77-88. Berkman, H.E. and C.F. Rabeni. 1987. Effects of siltation on stream fish communities. Environ. Biol. Fish 18: 285-294. Bolen, W.P. 1997. Construction sand and gravel: U.S. Geological Survey. Minerals Yearbook 1: 703-714. Boschung, H.T. and R.L. Mayden. 2004. Fishes of Alabama. Smithsonian Books. Washington, D.C. Bowen, Z.H., M.C. Freeman, and D.L. Watson. 1996. Index of biotic integrity applied to a flow-regulated river system. Proc. Annu. Conf. Assoc. Fish and Wildl. Agencies 50: 26-37. Boyd, C.E. 2000. Water quality: an introduction. Kluwer Academic Publishers. Boston, MA. Bramblett, R.G., T.R. Johnson, A.V. Zale, and D.G. Heggem. 2005. Development and evaluation of a fish assemblage index of biotic integrity for northwestern Great Plains streams. Transactions of the American Fisheries Society 134: 624- 640. 107 Brower, J.E. and J.H. Zar. 1984. Field and laboratory methods for general ecology. 2 nd ed. Wm. C. Brown Publishing. Dubuque, Iowa. Brown, A.V., M.M. Lyttle and K.B Brown. 1998. Impacts of gravel mining on gravel bed streams. American Fisheries Society 127: 979-994. Bull, W.B. and K.M. Scott. 1974. Impact of mining gravel from urban stream beds in the southwestern United States Geology. 2: 171-174. Capone, T.A. and J.A. Kushlan. 1991. Fish community structure in dry-season stream pools. Ecology 72: 983-992. Compton, M.C., G.J. Pond, J.F. Brumley. 2003. Development and application of the Kentucky index of biotic integrity. Division of Water. Water Quality Branch. Kentucky Department of Environmental Protection. Frankfort, Kentucky. Crumby, W.D., M.A. Webb, and F.J. Bulow. 1990. Changes in biotic integrity of a river in north-central Tennessee. Transactions of the American Fisheries Society 119: 885-893. Dahlberg, M.L. and D.C. Scott. 1971. The freshwater fishes of Georgia. Bull. Georgia Acad. Sci 29: 1-64. Danielson, T.J. 1998. Wetland bioassessment fact sheets. U.S. EPA843-f-98-001. Environmental Protection Agency. Office of Wetlands, Oceans, and Watersheds. Wetlands Division. Washington, D.C., USA. Dauwalter D.C., E.J. Pert, and W.E. Keith. 2003. An index of biotic integrity for fish assemblages in Ozark Highland streams of Arkansas. Southeastern Naturalist 2: 447-468. Dauwalter, D.C. and J.R. Jackson. 2004. A provisional fish index of biotic integrity for assessing Ouachita Mountains streams in Arkansas, USA. Environmental Monitoring and Assessment 91: 27-57. Davenport, L.J., W.M. Howell, K.J. Morse, K. Yancie, and J.L. Wood. 2005. Fishes and macroinvertebrates of the upper Cahaba River: a three-year study. Journal of the Alabama Academy of Science 76: 1-18. Deacon. J.E. 1961. Fish populations, following a drought, in the Neosho and Marais des Cygnes rivers of Kansas. Museum of Natural History. Volume 13. University of Kansas, Lawrence. 108 Detenbeck, N.E., P.W. DeVore, G.J. Niemi, and A. Lima. 1992. Recovery of temperate-stream fish communities from disturbance: a review of case studies and synthesis of theory. Environmental Management 16: 33-53. Ebert, D.J. and S.P. Filipek. 1988. Response of fish communities to habitat alteration in a small Ozark stream. Proceedings of the Arkansas Academy of Science 42: 28-32. Ebert, D.J., T.A. Nelson and J.L. Kershner. 1991. A soil-based assessment of stream fish habitats in coastal plain streams. Warmwater Fisheries Symposium I. General technical Report RM-207. USDA Forest Service. Washington, D.C. pp. 217-224. Evans, J.W. and R.L. Noble. 1979. The longitudinal distribution in an east Texas stream. Am. Midl. Nat. 101: 333-343. Fausch, K.D., J.R. Karr, and P.R. Yant. 1984. Regional application of an index of biotic integrity based on stream fish communities. Transactions of the American Fisheries Society 113: 39-55. Fausch, K.D. and L.H. Schrader. 1987. Use of the index of biotic integrity to evaluate the effects of habitat, flow, and water quality on fish communities in three Colorado front range streams. Final report to the Kodak-Colorado Division and the Cities of Fort Collins, Loveland, Greeley, Longmont, and Windsor, Colorado. Department of Fishery and Wildlife Biology. Colorado State University. Fort Collins, Colorado. Fore, L.S., J.R. Karr, and L.L. Conquest. 1994. Statistical properties of an index of biological integrity used to evaluate water-resources. Canadian Journal of Fisheries and Aquatic Sciences 51: 1077-1087. Fore, L.S., J.R. Karr, and R.W. Wisseman. 1996. Assessing invertebrate responses to human activities: evaluating alternative approaches. J. N. Amer. Benthol. Soc. 15: 212-231. Freeman, M.C., M. Crawford, J.C. Barrett, D.E. Facey, M.G. Flood, J. Hill, D.J. Stouder, and G.D. Grossman. 1988. Fish assemblage stability in a southern Appalachian stream. Canadian Journal of Fisheries and Aquatic Sciences 45: 1949-1958. Frenzel, S.A. and R.B. Swanson. 1996. Relations of fish community composition to environmental variables in streams of central Nebraska, USA. Environmental Management 20: 689-705. 109 Gammon, J.R. 1970. The effect of inorganic sediment on stream biota. Water Pollution Control Research Series 1805ODWC12/70. United States Environmental Protection Agency. Water Quality Office, Washington, D.C. Gammon, J.R. 1976. The fish populations of the middle 340 km of the Wabash River. Purdue Univ. Water Res. Cen. Tech. Rep. 86. Gelwick, F.P., S. Akin, D.A. Arrington and K.O. Winemiller. 2001. Fish assemblage structure in relation to environmental variation in a Texas gulf coastal wetland. Estuaries 24: 285-296. Georgia Department of Natural Resources (GA DNR). 2005. Part I: Standard operating procedures for conducting biomonitoring on fish communities in wadeable streams in Georgia. Fisheries Management Section. Wildlife Resources Division. Gilbert, R.J. 1969. The distribution of fishes in the central Chattahoochee River drainage. Masters Thesis. Auburn University. Auburn, Alabama. Goldstein, R.M. and T.P. Simon. 1999. Toward a united definition of guild structure for feeding ecology of North American Freshwater fishes. Pages 123-202 in T.P. Simon, ed. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. Boca Raton, Florida. Gore, J.A. 1996. Discharge measurements and streamflow analysis. Pages 53-74 in F.R. Hauer and G.A. Lamberti, eds. Methods in stream ecology. Academic Press. San Diego, California. Gorman, O.T. and J.R. Karr. 1978. Habitat structure and stream fish communities. Ecology 59: 507-515. Griffith, G.E., J.M. Omernick, J.A. Comstock, S. Lawrance, G. Martin, A. Geddard, V.J. Hulchjer, and T. Foster. 2001. Ecoregions of Alabama and Georgia, (color poster with map, descriptive text, summary tables, and photographs): Reston, Virginia. U.S. Geological Survey. (1:1,700,000-scale). Harkins, J.R. and others. 1980. Hydrologic assessment, Eastern Coal Province, Area 23, Alabama. United States Geological Survey. Open-file Report 80-683. 76 pp. Hlass, L.J., W.L. Fisher, and D.J. Turton. 1998. Use of the index of biotic integrity to assess water quality in forested streams of the Ouachita Mountains Ecoregion, Arkansas. Journal of Freshwater Ecology 13(2): 181-192. 110 Helms, B.S., J.W. Feminella, and S. Pan. 2005. Detection of biotic responses to urbanization using fish assemblages from small streams of western Georgia, USA, Urban Ecosystems 8: 39-57. Henry, T.B., E.R. Irwin, Grizzle, J.M., Wildhaber, M.L., and Brumbaugh, J.G. 1999. Acute toxicity of an acid mine drainage mixing zone to juvenile bluegill and largemouth bass. Trans. Am. Fish. Soc. 128: 919-928. Hill, B.H., A.T. Herlily, P.R. Kaufmann, R.J. Stevenson, F.H. McCormick, and C.B. Johnson. 2000. Use of periphyton assemblage data as an index of biotic integrity. J.N. Amer. Benthol. Soc. 19: 50-67. Hindman, J., G. McMichael, J. Olson, and S. Leider. 1990. Yakima River species interactions studies. Annual Report, Project No. 198910500, 85 electronic pages, (BPA Report DOE/BP-01483-1). Hlass, L.J., W.L. Fisher, and D.J. Turton. 1998. Use of the index of biotic integrity to assess water quality in forested streams of the Ouachita Mountain ecoregion, Arkansas. Journal of Freshwater Ecology 13: 181-192. Horne, A.J. and C.R. Goldman. 1994. Limnology 2 nd ed. McGraw-Hill, Inc. New York, New York. Hoyt, R.D. 1 , D. McElroy, and D. Stiles 2 . 2001. Spatio-temporal assessment of stream fish communities adjacent to an agricultural best-management practices operation. 1 Department of Biology, Center for Biodiversity Studies, Western Kentucky University, Bowling Green, Kentucky. 2 Department of Agriculture, Western Kentucky University, Bowling Green, Kentucky. Hughes, R.M. and J.R. Gammon. 1987. Longitudinal changes in fish assemblages and water quality in Willamette River, Oregon. Transactions of the American fisheries Society 116: 196-209. Hughes, R.M. 1990. The IBI: a quantitative, easily communicated assessment of the health and complexity of entire fish communities. Pages 26-28 in M.B. Bain, editor. Ecology and assessment of warmwater streams: workshop synopsis. U.S. Fish Wildl. Serv., Biological Report 90. 44 pp. Hughes, R.M. 1995. Defining acceptable biological status by comparing with reference conditions. Pages 31-48 in W.S. Davis and T.P. Simon, eds. Biological assessment and criteria: tools for water resource planning and decision making. CRC Press, Inc. Boca Raton, Florida. 111 Hughes, R.M., P.R. Kaufmann, A.T. Herlihy, T.M. Kincaid, L. Reynolds, and D.P. Larsen. 1998. A process for developing and evaluating indices of fish assemblage integrity. Canadian Journal of Fisheries and Aquatic Sciences 55: 1618-1631. Hughes, R.M. and T. Oberdorff. 1999. Applications of IBI concepts and metrics to waters outside of the United States and Canada. Pages 79-93 in T.P. Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. Hugueny, B.S., S. Camara, B. Samoura, and M. Magassouba. 1996. Applying an index of biotic integrity based on fish assemblages in a West African river. Hydrobiologia: 331: 71-78. Jost, L. 2006. Entropy and diversity. Oikos 113: 363-375. Karr, J.R. 1981. Assessment of biotic integrity using fish communities. Fisheries 6: 21-27. Karr, J.R. 1991. Biological integrity: a long-neglected aspect of water resource management. Ecol. Appl. 1: 66-84. Karr, J.R. 1993. Defining and assessing ecological integrity: beyond water quality. Environmental Toxicology and Chemistry 12: 1521-1531. Karr, J.R. 1999. Defining and measuring river health. Freshwater Biology 41: 221- 234. Karr, J.R. P.R. Yant, K.D. Fausch and I.J. Schlosser. 1984. Evaluation of an index of biotic integrity: temporal variability and regional application in the Midwest. EPA-600/D-84- 053. Environmental Research Laboratory. United States Environmental Protection Agency. Corvallis, Oregon. Karr, J.R., R.C. Heidenger, and E.F. Helmer. 1985. Effects of chlorine and ammonia from wastewater treatment facilities on biotic integrity. J. Water Pollut. Control 57: 912-915. Karr, J.R., K.D. Fausch, P.L., Angermeier, P.L.Yant, and I.J. Schlosser. 1986. Assessing biological integrity in running waters: a method and its rationale. Illinois Natural History Survey. Special Publication 5. Karr, J.R., P.R. Yant, K.D. Fausch, and I.J. Schlosser. 1987. Spatial and temporal variability of the index of biotic integrity in three Midwestern streams. Trans. Am. Fish. Soc. 116: 1-11. 112 Karr, J.R. and E.W. Chu. 1999. Restoring life in running waters: better biological monitoring. Island Press. Washington, D.C. 206 pp. Keaton, M., D. Haney, and C.B Andersen. 2005. Impact of drought upon fish assemblage structure in two South Carolina Piedmont streams. Hydrobiologia 545: 209-223. Kerans, B.L. and J.R. Karr. 1994. A benthic index of biotic integrity (B-IBI) for rivers of the Tennessee Valley. Ecological Applications 4: 768-785. Klemm, D.J., K.A. Blocksom, F.A. Fulk, A.T. Herlily, R.M. Hughes, P.R. Kaufmann, D.V. Peck, W.T. Thoeny, M.B. Griffith and W.S. Davis. 2003. Development and evaluation of a macroinvertebrate biotic integrity index (MBII) for regionally assessing mid-Atlantic highlands streams. Environmental Management 31: 656- 669. Knight, A.L., and J.G. Newton. 1977. Water and related problems in coal mine areas of Alabama. United States Geological Survey. Water Resources Inv. 76-130. 51 pp. Kondalf, G.M. 1994. Geomorphic and environmental effects of instream gravel mining. Landscape Urban Planning 28: 225-243. Kovalak, W. 1981. Assessment and prediction of impacts of effluents on communities of benthic stream invertebrates. Pages 255-263 in J.M Bates and C.I Weber, eds. Ecological assessments of effluent impacts on communities of indigenous aquatic organisms. American Society of Testing and Materials, STP 730, Philadelphia, Pennsylvania. Langer, W.H. and Glanzer, V.M. 1995. Natural aggregate ? building America?s future; public issues in earth science. U.S. Geological Survey Circular 1110. U.S. Department of Interior, Denver, Colorado. Larimore, R.W., W.F. Childers, and C. Heckrotta. 1959. Destruction and re- establishment of stream fish and invertebrates affected by drought. Transactions of the American Fisheries Society 88: 261-285. Leonard, P.M. and D.J. Orth. 1986. Application and testing of an index of biotic integrity in small, coolwater streams. Transactions of the American Fisheries Society 115: 401-414. Ligon, F.K., W.E. Dietrich, and W.J. Trush. 1995. Downstream ecological effects of dams: a geomorphic perspective. Bioscience 45: 183-192. 113 Lohr, S.C., and K.D. Fausch. 1997. Multiscale analysis of natural variability in stream fish assemblages of a western Great Plains watershed. Copeia 1997: 706-724. Ludwig, J.A. and J.F. Reynolds. 1988. Statistical ecology: a primer on methods and computing. John Wiley and Sons, Inc. New York. Lyons, J. S., P.A. Navarro-Perez, E. Cochran, C. Santana, and M.Guzman-Arroyo. 1995. Index of biotic integrity based on fish assemblages for the conservation of streams and rivers in west-central Mexico. Conservation Biology 9: 569-584. Lydy, M.J. C.J. Crawford, and J.W. Frey. 2000. A comparison of selected diversity, similarity and biotic indices for detecting changes in benthic-invertebrate community structure and stream quality. Archives of Environmental Contamination and Toxicology 39: 469-479. Lyttle, M.M. 1993. Impacts of gravel mining on fish communities in three Ozark streams. Masters Thesis. University of Arkansas. Fayetteville, Arkansas. Maceina, M.J., P.W. Bettoli, and D.R. Devries. 1994. Use of a split-plot analysis for repeated-measures fishery data. Fisheries 19: 14-20. Mammoliti, C.S. 2002. The effects of small watershed impoundments on native stream fishes: a focus on the Topeka shiner and horneyhead chub. Trans. of Kansas Academy of Science 105: 219-231. Marchetti, M.P and P.B. Moyle. 2001. Effects of flow regime and habitat structure on fish assemblages in a regulated California stream. Ecological Applications 11: 530-539. Matthews, W.J. and E. Marsh-Matthews. 2003. Effects of drought on fish across axes of space, time, and ecological complexity. Freshwater Biology 48: 1232-1253. Maxted, J.R., M.T. Barbour, M.T. Gerritsen, J. Poretti, V. Primrose, N. Silvia, A. Penrose, D. Renfrow, and R. Renfrow. 2000. Assessment framework for Mid- Atlantic Coastal Plain streams using benthic macroinvertebrates. J. N. A. Benthol. Soc. 19: 128-144. McCall, R.B. 1986. Fundamental statistics for behavioral sciences 4 th ed. Harcourt Brace Jovanovich, Publishers. San Diego, California. McCormick, F.H., R.M. Hughes, P.R. Kaufmann, and A.T. Herlihy. 2001. Development of an index of biotic integrity for the Mid-Atlantic Highlands region. Transactions of the American Fisheries Society 130: 857-877. 114 McDonough, T.A. and G.D. Hickman. 1999. Reservoir fish assemblage index development: a tool for assessing ecological health in Tennessee Valley Authority Impoundments. Pages 523-562 in T.P Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. Meador, M.R. and W.J. Matthews. 1992. Spatial and temporal patterns in fish assemblage structure in an intermittent Texas stream. Am. Midl. Nat. 127: 106- 114. Medeiros, E.S. and L. Maltchik. 2001. Fish assemblage stability in an intermittently flowing stream from the Brazilian semiarid region. Austral Ecology 26: 156-164. Meffe, G.K. and A.L. Sheldon. 1988. The influence of habitat structure on fish assemblage composition in southeastern blackwater streams. Am. Midl. Nat. 120: 225-240. Meffe, G.K. and A.L. Sheldon. 1990. Post-defaunation recovery of fish assemblages of southeastern blackwater streams. Ecology 71: 657-667. Mettee, M.F., P.E. O?Neil, and J.M. Pierson. 1996. Fishes of Alabama and the Mobile Basin. Oxmoor House, Inc. Birmingham, Alabama. Miller, D.L., P.M. Leonard, and R.M. Hughes. 1988. Regional application of an index of biotic integrity for use in water resource management. Fisheries 13: 12-20. Minns, C.K., V.W. Cairns, R.G. Randall, and J.E. Moore. 1994. An index of biotic integrity (IBI) for fish assemblages in the littoral zone of Great Lakes? area of concern. Canadian Journal of Fisheries and Aquatic Sciences 51: 1804-1822. Moyle, P.B. 1994. Biodiversity, biomonitoring, and the structure of stream fish communities. Pages 171-186 in S.L. Loeb and A. Spacie, eds. Biological monitoring of aquatic systems. Lewis Publ. Boca Raton, Florida. Mulholland, P.J. and D.R. Lenat. 1992. Streams of the southeastern Piedmont, Atlantic drainage. Pages 193-232 in C.T. Hackney, S.M Adams, and W.A. Martin, eds. Biodiversity of the southeastern United States-aquatic communities. John Wiley and Sons, Inc. New York. Mundahl, N.D. and T.P. Simon. 1999. Development and application of an index of biotic integrity for coldwater streams of the upper Midwestern United States. Pages 383-415 in T.P Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. 115 Niemela, S.N., E. Pearson, T.P. Simon, R.M. Goldstein, and P.A. Bailey. 1999. Development of an index of biotic integrity for the species-depauperate Lake Agassiz Plain ecoregion, North Dakota and Minnesota. Pages 339-366 in T.P Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. Niemela, S.N. and M.D. Feist. 2002. Index of biotic integrity (IBI) guidance for coolwater rivers and stream of the Upper Mississippi River basin. Minnesota Pollution Control Agency. St. Paul, Mn. 56 pp. Nunnally, N.R. 1978. Stream renovation: an alternative to channelization. Environmental Management 2: 403-411. Oberdoff, T. and R.M. Hughes. 1992. Modification of an index of biotic integrity based on fish assemblages to characterize rivers of the Seine Basin, France. Hydrobiologia 228: 117-130. Ohio Environmental Protection Agency. 1987. Biological criteria for the protection of aquatic life. Volumes I, II, and III. State of Ohio Environmental Protection Agency. Division of Water Quality Planning and Assessment. Columbus, Ohio. 376 pp. Omernick, J.M. 1987. Ecoregions of the conterminous United States. Annals of the Association of American Geographers 77: 118-125. O?Neil, P.E. and R.V. Chandler. 2005. Water quality and biological monitoring in the Choccolocco Creek watershed, Alabama, 1996-2001. Alabama Geological Survey Bulletin 177, 72 pp. O?Neil, P.E. and T.E. Shepard. 1998. Standard operating procedure manual for sampling freshwater fish communities and application of the index of biotic integrity for assessing biological condition of flowing, wadeable streams in Alabama. Report prepared in partial fulfillment of Contract No. AGY7042 with the Alabama Department of Environmental Management. O?Neil, P.E. and T.E. Shepard. 2000. Application of the index of biotic integrity for assessing biological condition in the Black Warrior River system, Alabama. Alabama Geological Survey Bulletin 169, 71 pp. O?Neil, P.E. and T.E. Shepard. 2004. Hatchet Creek regional reference watershed study. Alabama Geological Survey Open-File Report 0509. 71 pp. 116 O?Neil, P.E, T.E. Shepard, and M.R. Cook. 2006. Habitat and biological assessment of the Terrapin Creek watershed and development of the index of biotic integrity for the Coosa and Tallapoosa River systems. Open-File Report 0601. Water Investigations Program. Geological Survey of Alabama. Tuscaloosa, Alabama. Onorato, D., K.R. Marion, and R.A. Angus. 1998. Longitudinal variations in the icthyo-faunal assemblages of the upper Cahaba River: possible effects of urbanization in a watershed. Journal of Freshwater Ecology 13(2): 139-154. Osborne, L.L., S.L. Kohler, P.B. Bayley, D.M. Day, W. A. Bertrand, M.J. Wiley, and R. Sauer. 1992. Influence of stream location in a drainage network on the index of biotic integrity. Transactions of the American Fisheries Society 121: 635-643. Paller, M.H., M.J.M. Reichert, and J.M. Dean. 1996. Use of fish communities to assess environmental impacts in South Carolina Coastal Plain streams. Transactions of the American Fisheries Society 125: 633-644. Paller, M.H. 2002. Temporal variability in fish assemblages from disturbed and undisturbed streams. Journal of Aquatic Ecosystems Stress and Recovery 9: 149- 158. Peet, R.K. 1974. The measurement of species diversity. Annual Review of Ecology and Systematics 5: 496-498. Pielou, E.C. 1966. Ecological diversity. Wiley and Sons, Inc. New York. pp. 165. Pierson, M.J., R.S. Krotzer, and E.J. Tyberghein. 1986. Fishes from the upper Tallapoosa River and three of its major tributaries in Alabama. Journal of the Alabama Academy of Science. 57: 1- 18. Poff, L.N. and J.D. Allan. 1995. Functional organization of stream fish assemblages in relation to hydrological variability. Ecology 76: 606-627. Poff, L.N., J.D. Allan, M.B. Bain, J.R. Karr, K.L. Prestegaard, B.D. Richter, R.E. Sparks, and J.C. Stromberg. The natural flow regime: a paradigm for river conservation and restoration. Bioscience 74: 769-784. Pond, G.J. 2004. Effects of surface mining and residential land use on headwater stream biotic integrity in the eastern Kentucky coalfield region. Kentucky Department for Environmental Protection, Division of Water, Frankfurt, Kentucky. Poulin, R.C., K. Pakalnis, and K. Sinding. 1994. Aggregate resources: production and environmental constraints. Environmental Geology 23: 221-227. 117 Prouty, W.F. 1916. Preliminary report on the crystalline and other marbles of Alabama. Alabama Geological Survey Bulletin 18. pp. 78-80. Tuscaloosa, Alabama. Rabeni, C.F. and M.A. Smale. 1995. Effects of siltation on stream fishes and the potential mitigating role of the buffering riparian zone. Hydrobiologia 303: 211- 219. Rantz, S.E., and others. 1982. Measurement and computation of streamflow: Volume 1. Measurement of stage and discharge. U.S. Geological Survey Water-Supply Paper 2175. Resh, V.H., R.H. Norris, and M.T. Barbour. 1995. Design and implementation of rapid assessment approaches for water-resource monitoring using benthic macroinvertebrates. Australian Journal of Ecology 20: 108-121. Richter, B.D., and G.A. Thomas. Restoring environmental flows by modifying dam operations. Ecology and Society 12(1): 12. [online] URL: http://www.ecologyandsociety.org./vol12/iss1/art12/. Robb, G.A. 1994. The consequences of coal mine closure. The Geographical Journal 106: 33-40. Roell, M.J. 1999. Sand and gravel mining in Missouri stream systems: aquatic resource effects and management alternatives. Executive Summary. Missouri Department of Conservation. Columbia, Missouri. Roth, N., M. Southerland, J. Chaillou, R. Klauda, S. Stranko, S. Weisberg, L. Hall, Jr., and R. Morgan II. 1998. Maryland biological stream survey: development of a fish index of biotic integrity. Environmental Monitoring and Assessment 51: 89- 106. Sall, J., A. Lehman and L. Creighton. 2001. JMP start statistics ? A guide to statistics and data analysis using JMP and JMPIn software 2 nd ed. SAS Institute, Inc. Duxbury Press/Thomson Learning. Sams, J.I. and K.M. Beer. 2000. Effects of coal-mine drainage on stream water quality in the Allegheny and Monongehela river basins-sulfate transport and trends. Water Resources Investigations Report 99-4208. National Water Quality assessment Program. U.S. Geological Survey. Sandecki, M. 1989. Aggregate mining in river systems. California Geologist 42: 88- 94. 118 Saylor, C.F. and E.M. Scott. 1987. Application of the index of biotic integrity to existing TVA data. Division of Services and Field Operations. Office of Natural Resources and Economic Development. Tennessee Valley Authority. Norris, Tennessee. 25 pp. Saylor, C.F., G.D. Hickman, and M.P. Taylor. 1988. Application of index of biotic integrity (IBI) to fixed station water quality monitoring sites-1987. Aquatic Biology and Fisheries. Tennessee Valley Authority. Norris, Tennessee. 72 pp. Saylor, C.F. 1989. Sand Mountain-Guntersville Reservoir biological monitoring 1988 index of biotic integrity. Report TVA/WR/WQ?89/4 prepared by Aquatic Biology Department for Water Quality Department. Chattanooga, Tennessee. 29 pp. Saylor, C.F. and S.A. Ahlstedt. 1990. Application of index of biotic integrity (IBI) to fixed station water quality monitoring sites. Aquatic Biology Department. Tennessee Valley Authority. Norris, Tennessee. 94 pp. Schleiger, S.L. 2000. Use of an index of biotic integrity to detect effects of land use on stream fish communities in west-central Georgia. Transactions of the American Fisheries Society 129: 1118-1133. Schlosser, I.J. 1982. Fish community structure and function along two habitat gradients in a headwater stream. Ecological Monographs 52: 393-414. Schlosser, I.J. 1987. A conceptual framework for fish communities in small warmwater streams. Pages 17-24 in W.J. Matthews and D.C. Heins, eds. Community and evolutionary ecology of North American stream fishes. Univ. Okla. Press, Norman, Oklahoma. Schlosser, I.J. 1990. Environmental variation, life history attributes, and community structure in stream fishes: implications for environmental management and assessment. Environmental Management 14: 621-628. Schlosser, I.J., J.D. Johnson, W.L. Knotek, and M. Lapinska. 2000. Climate variability and size-structured interactions among juvenile fish along a lake- stream gradient. Ecology 4: 1046-1057. Scott, E.M. 1992. TVA reservoir monitoring program fish community results-1991. Aquatic Biology Group. Water Resources. Tennessee Valley Authority. Norris, Tennessee. Scott, M.C. and L.W. Hall. 1997. Fish assemblages as indicators of environmental degradation in Maryland coastal streams. Trans. Amer. Fish. Soc. 126: 349-360. 119 Scott, E.M. 1999. Tailwater fish index (TFI) development for Tennessee River tailwaters. Pages 507-540 in T.P. Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. Shaner, B.L. 1999. Development of a standardized index of biotic integrity for the Piedmont region of Georgia. Proceedings of the 1999 Georgia Water Resources Conference, held March 26-27, at the University of Georgia. K.J. Hatcher, editor. Institute of Ecology. University of Georgia. Athens, Georgia. pp. 287-290. Shaner, B.L. 2001. An index of biotic integrity for wadeable streams in the Appalachicola and Atlantic slope basin drainages in the Piedmont ecoregion. Proceedings of the 2001 Georgia Water Resources Conference, held March 30-31, at the University of Georgia. K.J. Hatcher, editor. Institute of Ecology. University of Georgia. Athens, Georgia. pp. 287-290. Shannon, C.E. and W. Weaver. 1949. The mathematical theory of communication. University Illinois Press, Urbana, IL. Sheldon, A.L. 1969. Equitability indices: dependence on the species count. Ecology 50: 466-467. Shearer, J.S. and C.R. Berry, Jr. 2002. Index of biotic integrity utility for the fishery of the James River of the Dakotas. Journal of Freshwater Ecology 17: 575-588. Shepard, T.E., S.W. McGregor, M.F. Mettee, and P.E. O?Neil. 1995. Status of the bluestripe shiner (Cyprinella callitaenia) in Alabama, 1994-1995. Alabama Geological Survey Open File Report. 78 pp. Shepard, T.E., P.E. O?Neil, S.W. McGregor, M.F. Mettee, and S.C. Harris. 1997. Biomonitoring and water-quality studies in the upper Cahaba River drainage of Alabama, 1989-1994. Alabama Geological Survey Bulletin 165. 255 pp. Shields, F.D., Jr. and J.J. Hoover. 1991. Effects of channel restabilization on habitat diversity, Twenty Mile Creek, Mississippi. Regulated rivers: Res. Mgmt 6: 163- 181. Shields, F.D., Jr., C.M. Cooper, and S.S. Knight. 1994. Effects of channel incision on base flow stream habitats and fishes. Envir. Mgmt 18: 43-57. Shields, F.D., S.S. Knight, and C.M. Cooper. 1995. Use of the index of biotic integrity to assess physical habitat degradation in warmwater streams. Hydrobiologia 312: 191-208. 120 Simon, T.P. 1991. Development of index of biotic integrity expectations for the ecoregions of Indiana. I. Central Corn Belt Plain. U.S. Environmental Protection Agency. Region V. Environmental Sciences Division, Monitoring and Quality Assurance Branch: Ambient Monitoring Section. EPA 905/9-91/025. Chicago, Illinois. Simon, T.P. and J. Lyons. 1995. Application of the index of biotic integrity to evaluate water resource integrity in freshwater ecosystems. Pages 245-262 in W.S. Davis and T.P. Simon, eds. Biological assessment and criteria: tools for water resource planning and decision making. Lewis Publishers. Boca Raton, Florida. Smogor, R.A. and P.L. Angermeier. 1999. Relations between fish metrics and measures of anthropogenic disturbance in three IBI regions in Virginia. Pages 585-610 in T.P Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. Sokal, R.R. and F.K. Rohlf. 1987. Introduction to biostatistics. 2 nd ed. W.H. Freeman. New York. Soucek, D.J. 2001. Integrative bioassessment of acid mine drainage impacts on the Upper Powell River watershed, Southwestern Virginia. Ph.D. Dissertation. Virginia Polytechnic Institute and State University. Blacksburg, Virginia. Spence, A.M., M.L. Smith, and R.W. Nairn. 1999. Ecological assessment of several central Oklahoma streams through evaluation of fish communities and habitat in a drought year. Proc. Okla. Acad. Sci. 79: 61-72. Starnes, L.B. and D.C. Gasper. 1996. Effects of surface mining on aquatic resources in North America. Fisheries 21: 24-26. Steedman, R.J. 1988. Modification and assessment of an index of biotic integrity to quantify stream quality in southern Ontario. Canadian Journal of Fisheries and Aquatic Sciences 45: 492-501. Stewart, A.J. and J.M Loar. 1994. Spatial and temporal variation in biological monitoring data. Pages 91-124 in W.J. Matthews and D.C. Heins eds. Biological monitoring of aquatic systems. Lewis Publ. Boca Raton, Florida. Tabit, C.R. and G.M. Johnson. 2002. Influence of urbanization on the distribution of fishes in a southeastern upper Piedmont drainage. Southeastern Naturalist: 253- 268. 121 Taylor, C.M., M.R. Winston, and W.J. Matthews. 1996. Temporal variation in tributary and main stem fish assemblages in a Great Plains stream system. Copeia 1996: 280-289. Teels, B.M., L.E. Mazanti, and C.A. Rewa. 2004. Using an IBI to assess effectiveness of mitigation measures to replace loss of a wetland-stream ecosystem. Wetlands 24: 375-384. Tepordei, V.V. 1997. U.S. Geological Survey Minerals Yearbook 1: 783-809. Tomljanovich, D.A. and C.F. Saylor. 1989. Results of fisheries investigations in Douglas tailwater, August 1987 ? September 1988. Norris, Tenn: Tennessee Valley Authority, Water Resources, Aquatic Biology Department. Turner, M.G. 1987. Land use changes and net primary production in the Georgia, USA, landscape: 1935-1982. Environmental Management 11: 237-247. United States Environmental Protection Agency (USEPA). 2006. Draft wadeable streams assessment: a collaborative survey of the nation?s streams. EPA 841-B06- 002. Office of Water. Washington, D.C. United States Fish and Wildlife Service (USFWS). 2003. Chewacla Creek Safe Harbor Agreement. Ecological Services Field Office. Daphne, Alabama. United States Geological Survey (USGS). 2004. U.S. Geological Survey Minerals Yearbook, Vol. II Area Reports: Domestic. Geologic Division, Reston, Virginia. Vincent, R. 1960. Some influences of domestication upon three stocks of brook trout (Salvelinus fontinalis Mitchill). Trans. Amer. Fish. Soc. 89: 39-52. Vincent, R. 1971. River electrofishing and fish population estimates. Prog. Fish Culturist 33: 163-182. Wade, D.C. and S.B. Stalcup. 1987. Assessment of the sport fishery potential for the Bear Creek floatway: biological integrity of representative sites, 1986. Division of Air and Water Resources. Office of Natural resources and Economic Development. Tennessee Valley Authority. Norris, Tennessee. 35 pp. Wahl, K.L., W.O. Thomas, Jr., and R.M. Hirsch. 1995. Stream-gaging program of the U.S. Geological Survey. U.S. Geological Survey Circular 1123. Reston, Virginia. Walser, C.A. and H.L Bart, Jr. 1999. Influence of agriculture on in-stream fish habitat and fish community structure in Piedmont watersheds of the Chattahoochee River system. Ecology of Freshwater Fish 8: 237-246. 122 Walters, D.M., M.J. Freeman, D.S. Leigh, B.J. Freeman, M.J. Paul, and C.M. Pringle. 2001. Bed texture and turbidity as indicators of fish biotic integrity in the Etowah River system. Proceeding of the 2001 Georgia Water Resources Conference. Athens, Georgia. Ward, J.V. and J.A. Stanford. 1983. The serial discontinuity concept of lotic ecosystems. Pages 29-42 in T.D. Fontaine and S.M. Bartell, editors. Dynamics of lotic ecosystems. Ann Arbor Science, Ann Arbor, Michigan, USA. Waters, T.F. 1995. Sediment in streams ? sources, biological effects, and control. American Fisheries Society Monograph 7. Weber, C.I. 1981. Evaluation of the effects of effluents on aquatic life in receiving waters-an overview. American Society for Testing and Materials Special Technical Publication 730: 3-13. Webber, E.C. 1971. Changes in limnological characteristics of two small streams from the headwaters to mouth. PhD Dissertation. Auburn University. Auburn, Alabama. Webber, E.C. 2002. Bioassessment of stream quality in Chewacla Creek and its tributaries. Final report submitted to the City of Auburn and the Auburn Water Works Board. Department of Fisheries and Allied Aquacultures. Auburn University. Webber, E.C. 2006. Bioassessment of stream quality in Chewacla Creek and its tributaries. Final report submitted to the City of Auburn and the Water Works Board. Department of Fisheries and Allied Aquacultures. Auburn University. 1 Webber, E.C. and 2 W.T. Blevins. 2000. Bioassessment of stream quality in Chewacla Creek and its tributaries. Final report submitted to the City of Auburn and the Auburn Water Works Board. 1 Department of Fisheries and Allied Aquacultures, Auburn University and 2 Suncrest Laboratories, L.L.C. 1 Webber, E.C. and 2 W.T. Blevins. 2001. Bioassessment of stream quality in Chewacla Creek and its tributaries. Final report submitted to the City of Auburn and the Auburn Water Works Board. 1 Department of Fisheries and Allied Aquacultures, Auburn University and 2 Suncrest Laboratories, L.L.C. Wetzel, R.G. 2001. Limnology: lake and river ecosystems. 3 rd ed. Academic Press. San Diego, California. Wilhm, J.L. and T.C. Dorris. 1968. Biological parameters for water quality criteria. Bioscience 18: 477-481. 123 Wilcox, D.A., J.E. Meeker, P.L. Hudson, B.J. Armitage, M.G. Black, D.G. Uzarski. 2002. Hydrologic variability and the application of index of biotic integrity metrics to wetlands: a Great Lakes evaluation. Wetlands 22: 588-615. Yoder, C.O. and E.T. Rankin. 1998. The role of biological indicators in a state water quality management process. Environmental Monitoring and Assessment. 51: 61- 88. Yoder, C.O. and M.A. Smith. 1999. Using fish assemblages in a state biological assessment and criteria program: essential concepts and considerations. Pages 17- 56 in T.P Simon, editor. Assessing the sustainability and biological integrity of water resources using fish communities. CRC Press, Inc. New York, New York. Zar, J.H. 1984. Biostatistical analysis. Prentice Hall, Inc. Englewood Cliffs, New Jersey. Zar, J.H. 1996. Biostatistical analysis. 3 rd ed. Prentice Hall, Inc. Englewood Cliffs, New Jersey.