M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 112 Vol. 12, Special Issue 4 Changes in Fish Community Structure and Effects on Blackside Dace (Chrosomus cumberlandensis) Populations Following Beaver Colonization in Davis Branch, Cumberland Gap National Historical Park, Bell County, Kentucky Michael C. Compton1, Michael A. Floyd2,*, and Douglas E. Stephens3 Abstract - Davis Branch in Cumberland Gap National Historical Park was considered important for the recovery of the Federally threatened Chrosomus cumberlandensis (Blackside Dace). Monitoring of the Davis Branch population began in 1990. Castor canadensis (North American Beaver) colonization was reported in 1994, and a series of dams have persisted in the upper section of the stream. The pervasive effects of Beavers on Blackside Dace populations were unknown. Our goals were: 1) to characterize the fish community structure and relative abundance of Blackside Dace within the assemblage, 2) to determine the direct and indirect effects of beaver dams on the fish community, and Blackside Dace population, 3) to monitor the age structure of the Blackside Dace population, and 4) to recommend management actions that will enhance the Blackside Dace population within Davis Branch. Fishes were collected from eight stations annually from 1990–2010, except for 2003–2005. Stations were delineated within downstream and upstream sections based on Beaver presence, and within three periods, prior (1990–1993), post (1994–2002), and recent (2006–2010). Fish community structure indicated distinct downstream and upstream assemblages prior to Beaver colonization, with the relative abundance of Blackside Dace approximately 6% and 18%, respectively. The fish assemblage and relative abundance of Blackside Dace shifted over time, with the relative abundance of Lepomis gulosus (Warmouth), Lepomis auritus (Redbreast Sunfish), and Chrosomus erythrogaster (Southern Redbelly Dace) increasing, and Blackside Dace decreasing; only 25 Blackside Dace were encountered from 2006–2010. Comparison of Blackside Dace age classes between the 3 periods indicated the 0+ age class decreased over time and 1+ and 2+ age classes increased initially but eventually declined within both the downstream and upstream sections. It is speculated that the continued presence of beaver dams increased stream temperatures downstream and upstream, altered the flow regime, and changed the available habitat, thus creating conditions more suitable for centrarchid species and Southern Redbelly Dace, which probably preyed upon and out-competed Blackside Dace, respectively. It is recommended that the National Park Service and its partners 1) develop a Beaver management program, 2) return Davis Branch to its free-flowing, pre-Beaver condition, 3) significantly decrease nonnative species and decrease native centrarchid species to abundances prior to Beaver colonization, and 4) establish a Davis Branch Blackside Dace propagation and reintroduction program. 1Kentucky State Nature Preserves Commission, 801 Schenkel Lane, Frankfort, KY 40601. 2US Fish and Wildlife Service, 330 West Broadway, Suite 265, Frankfort, KY 40601. 3PO Box 243, Whitley City, KY 42653. *Corresponding author - mike_floyd@fws.gov. Ecology and Conservation of the Threatened Blackside Dace, Chrosomus cumberlandensis 2013 Southeastern Naturalist 12(Special Issue 4):112–128 113 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 Introduction The federally threatened Chrosomus cumberlandensis (Starnes and Starnes) (Blackside Dace) was first reported from Davis Branch, Cumberland Gap National Historical Park (CUGA), Bell County, KY, by Starnes and Starnes (1978). Subsequent surveys by Starnes (1981) and O’Bara (1990), and the species’ recovery plan (USFWS 1988), ranked the Blackside Dace population in Davis Branch as excellent, and O’Bara (1990) considered it one of 9 populations most important to the species’ continued existence. Davis Branch continues to be one of only two Blackside Dace streams in public ownership that are located outside of the Daniel Boone National Forest, Winchester, KY. Due to the importance of Davis Branch toward Blackside Dace recovery, CUGA initiated a monitoring program for the stream that included annual surveys at several mainstem locations. In 1994, a large Castor canadensis Kuhl (North American Beaver) dam and pond were discovered at approximately 2.9 km upstream of the confluence with Little Yellow Creek. Beaver activity increased over subsequent years, resulting in the construction of at least 15 dams and the inundation of a continuous 1-km reach in the upper section of the Davis Branch mainstem (Fig. 1). Within this area, Davis Branch was transformed from a free-flowing, shaded, gravel/cobble stream to a series of ponds and canal-like reaches with decreased canopy cover and silty substrates. As habitat conditions changed, the status of the Blackside Dace population was a serious concern and continued monitoring was conducted. Beavers cause profound changes to stream habitats through dam building and manipulation of riparian trees (Collen and Gibson 2001, Muller-Schwarze and Sun 2003). These habitat alterations can influence fish communities in a number of ways (Collen and Gibson 2001). Positive effects include increased production for some species (Gard 1961, Hanson and Campbell 1963, Neff 1957); creation of pool habitats preferred by large-bodied species such as trout, centrarchids, and esocids (Gard 1961, Neff 1957, Pullen 1971, Rupp 1955); creation of cover associated with lodge debris and food caches (France 1997, Gibson 1969); reduced bed and bank erosion below dams and more stable downstream flows (Apple et al. 1984, Parker 1986); trapping of sediment in streams with high sediment loads (Grasse 1951, Macdonald et al. 1995); and creation of refugia during low-flow periods (Hanson and Campbell 1963). Beavers can negatively affect fish communities through creation of migration barriers (Gard 1961, Knudsen 1962, Munther 1983, Schlosser 1995); increased water temperatures and siltation of spawning substrates (Cook 1940, Knudsen 1962, Rohde and Arndt 1991); and creation of habitats for avian, mammalian, and piscine predators (Dubuc et al. 1990, Grover and Baldassarre 1995, Collen and Gibson 2001). In this paper, we evaluate Davis Branch fish community data collected from 1990–2010. We attempt to show the effect of Beaver colonization on the fish community structure and the Blackside Dace population by comparing the data within three periods—an initial period with no Beaver activity (prior: 1990–1993), an immediate period following Beaver colonization (post: 1994–2002), and a period M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 114 Vol. 12, Special Issue 4 after Beaver establishment (recent: 2006–2010). Project goals included: 1) to characterize the fish community structure and relative abundance of Blackside Dace within the assemblage, 2) to determine the direct and indirect effects of Figure 1. Generalized map of Davis Branch stations within Cumberland Gap National Historical Park (gray shading). Hashed area indicates inundated portion of Davis Branch affected by Castor canadensis (Beaver) dams. Stations 1–3 were classified as downstream and stations 4–8 were classified as upstream. 115 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 Beaver colonization on the fish community structure and Blackside Dace population, 3) to monitor the age structure of the Blackside Dace population, and 4) to make management recommendations that benefit Blackside Dace persistence in Davis Branch. Study Area Davis Branch is located in the Central Appalachians (Level III) and Cumberland Mountain Thrust Block (Level IV) ecoregions (Woods et al. 2002) of southeastern Kentucky. Its catchment (3.7 km2) lies almost entirely (>90%) within the boundary of CUGA, a 97-km2 national park that straddles the ridgeline of Cumberland Mountain in Kentucky, Tennessee, and Virginia (Fig. 1). Elevations within the Davis Branch watershed range from approximately 347 m to 633 m above mean sea level. Davis Branch is roughly 4.5 km long and flows southward to its confluence with Little Yellow Creek, just east of Middlesboro, KY. Prior to the establishment of CUGA, several farms and home sites were located along the stream, but land use within the watershed now consists primarily of second- and third-growth forest. The lower 0.1 km of Davis Branch is bordered on the northwest by a residential neighborhood within the city limits of Middlesboro. Davis Branch is characterized as a cool, upland stream with gravel and cobble substrates and moderate gradient (D.E. Stephens, unpubl. data). Riffle/ pool ratios range from 60/40 in the headwaters to 30/70 near the mouth (D.E. Stephens, unpubl. data). Riparian vegetation is dominated by Platanus occidentalis L. (Sycamore), Tsuga canadensis (L.) Carriere (Eastern Hemlock), Magnolia macrophylla Michx. (Bigleaf Magnolia), Rhododendron maximum L. (Rhododendron), Carpinus caroliniana Walter (American Hornbeam), Fagus grandifolia Ehrh. (American Beech), and Liriodendron tulipifera L. (Tulip Poplar). Beavers have significantly modified approximately 1.2 km of the Davis Branch mainstem, resulting in several ponds, wetlands, and canal-like reaches. These areas are characterized by reduced canopy cover, reduced flow, and silty substrates. Methods Field surveys Fish surveys in Davis Branch were conducted at 8 stations during 1990–2010 (Figs. 1, 2). Each station encompassed a stream reach approximately 200 m in length. Because of stream conditions and other study logistics, not all stations were surveyed annually, and no surveys were conducted during 2003–2005. Surveys were completed in early spring or fall to avoid disruption of Blackside Dace reproduction. Surveys were conducted in an upstream direction, and fishes were captured using a battery-powered, backpack electrofisher (Smith-Root, Vancouver, WA). All available habitats (e.g., riffles, pools, runs, undercut banks with root mats, and woody debris jams) were searched, and captured fishes were identified in the field, enumerated, recorded, and released. Electrofishing time M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 116 Vol. 12, Special Issue 4 (seconds) and total length (mm) of Blackside Dace were recorded during each sampling effort. Data analyses Fish community data were classified temporally and spatially based on the initial presence and location of beaver dams. Temporally, data were grouped into three periods: prior to Beaver colonization (prior: 1990–1993), during early pond establishment (post: 1994–2002), and established pond development (recent: 2006–2010). Spatially, stations 1–3 and stations 4–8 were grouped as downstream and upstream stations, respectively, based on beaver dam location. These groupings mitigated the absence of data from missed sampling years (2003–2005) and absence of annual sampling from some stations. We used nonmetric multidimensional scaling (NMS) to identify any potential differences and gradients in fish community structure across the temporal and spatial scales. The NMS was conducted using PC-ORD v. 6.0 software (McCune and Mefford 2011) and followed the general procedure outlined in McCune and Grace (2002). We used the Bray-Curtis similarity coefficient (Bray and Curtis, 1957, McCune and Grace, 2002) based on a power transformation (p = 0.5) of abundances. The data were computed with 250 iterations, 50 real runs, and 50 randomized runs. Figure 2. Photographs of Davis Branch. (A) Station 1, upstream view (April 2005); (B) Station 6, downstream view (April 2005); (C) Station 7, upstream view—beaver dam and pond (November 2011); (D) Station 8, upstream view (April 2005). 117 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 We used the nonparametric multi-response permutation procedure (MRPP, PCORD v. 6.0; McCune and Mefford 2011) to determine if fish community structure differed significantly among the spatial and temporal categories. We tested the null hypotheses of 1) no difference in fish community structure between downstream and upstream stations prior to Beaver colonization, 2) no differences between temporal periods within the downstream fish communities, and 3) no differences between temporal periods within the upstream fish communities. We calculated the relative abundance of each species encountered, for each spatial group across each temporal period, to determine if certain species were dominant or if a species made a noticeable shift within the community. A catchper- unit-effort (CPUE) was then calculated for selected species (i.e., Blackside Dace) and compared across the temporal periods. The CPUE was the log (x + 1) transformed result of the total number of individuals encountered divided by the number of 15-second electrofishing units of the total electrofishing time (seconds). We then used the Mann-Whitney U-tests to test the hypothesis of no difference in CPUE for a species across temporal periods (P < 0.05). Lastly, a frequency histogram of all Blackside Dace TL (mm) measurements was made to determine the appropriate cut-off for age classes. The 0+ age class was <40 mm TL, 1+ was 40–60 mm TL, and 2+ was >60 mm TL. These classifications were similar to previous descriptions by Starnes and Starnes (1981:368) and Etnier and Starnes (1993:244–245). A CPUE, as described above, was made for each age class across the temporal periods. The Mann-Whitney U-test was used to test the hypothesis of no difference in Blackside Dace age class CPUE across temporal periods (P < 0.05). Results Nearly 20,000 individuals, representing 5 families and 23 species, were encountered during 76 surveys completed during 1990–2010 (Table 1). Semotilus atromaculatus (Creek Chub), Lepomis auritu (Redbreast Sunfish), and Catostomus commersonii (White Sucker) were the most common species, and they represented 65% of the total abundance. Lampetra aepyptera (Least Brook Lamprey), Campostoma anomalum (Central Stoneroller), Luxilus chrysocephalus (Striped Shiner), Pimephales notatus (Bluntnose Minnow), Hypentelium nigricans (Northern Hog Sucker), Ambloplites rupestris (Rock Bass), and Etheostoma caeruleum (Rainbow Darter) were present primarily in the downstream section. Only Cyprinella spiloptera (Spotfin Shiner) was present (1 individual) exclusively in the upstream section. In general, Blackside Dace and Southern Redbelly Dace were more abundant in the upstream section than in the downstream section across each temporal period (Table 1). Blackside Dace were detected during 52 of the 76 surveys, and 1508 individuals were encountered. NMS produced a 2-dimension solution with a final stress of 16.1. The first and second axes accounted for 34% and 54% of the variation, respectively. The fish community structure among the different categories was distinct, and temporal and spatial gradients were evident along the first and second axes, respectively (Fig. 3). The separation of the fish community structure in ordinational space M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 118 Vol. 12, Special Issue 4 Table 1. Species abundance of Davis Branch fishes downstream and upstream of beaver dams and across temporal periods. Numbers in parentheses indicate the total number of surveys, total richness, and total abundance. Downstream Upstream Species Prior Post Recent Prior Post Recent Petromyzontidae Lampetra aepyptera (Abbott) (Least Brook Lamprey) 24 41 12 Cyprinidae Campostoma anomalum (Rafinesque) (Central Stoneroller) 62 342 87 5 4 1 Chrosomus cumberlandensis (Starnes & Starnes) (Blackside Dace) 97 490 2 333 561 25 Chrosomus erythrogaster Rafinesque (Southern Redbelly Dace) 78 1609 283 292 1665 420 Cyprinella spiloptera (Cope) (Spotfin Shiner) 1 Luxilus chrysocephalus Rafinesque (Striped Shiner) 166 323 241 3 4 Notemigonus crysoleucas (Mitchill) (Golden Shiner) 1 Notropis buccatus (Cope) (Silverjaw Minnow) 2 Pimephales notatus (Rafinesque) (Bluntnose Minnow) 6 79 20 Rhinichthys obtusus Agassiz (Western Blacknose Dace) 141 331 162 10 18 3 Semotilus atromaculatus (Mitchill) (Creek Chub) 779 2525 322 1000 1152 425 Catostomidae Catostomus commersonii (Lacepede) (White Sucker) 83 614 80 109 918 85 Hypentelium nigricans (Lesueur) (Northern Hog Sucker) 18 47 15 3 Moxostoma erythrurum (Rafinesque) (Golden Redhorse) 2 Centrarchidae Ambloplites rupestris (Rafinesque) (Rock Bass) 14 69 20 Lepomis auritus (Linnaeus) (Redbreast Sunfish) 71 1273 288 3 399 239 Lepomis gulosus (Cuvier) (Warmouth) 2 302 1 30 Lepomis macrochirus Rafinesque (Bluegill) 4 35 30 2 26 Micropterus punctulatus (Rafinesque) (Spotted Bass) 7 1 Micropterus salmoides (Lacepede) (Largemouth Bass) 3 1 Percidae Etheostoma caeruleum Storer (Rainbow Darter) 78 181 73 Etheostoma kennicotti (Putnam) (Stripetail Darter) 68 107 57 61 49 50 Etheostoma sagitta (Jordan & Swain) (Arrow Darter) 5 32 1 18 50 2 Surveys (76): 10 27 8 8 16 7 Richness (23): 16 21 20 10 13 12 Abundance (19,773): 1694 8113 1999 1834 4823 1310 119 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 was significant across all spatial and temporal categories based on the MRPP (Table 2). Spatially, distinct fish assemblages were observed for each temporal Table 2. Results of MRPP pairwise comparison of disturbance categories. Category Comparison T A P-value Prior: Downstream vs. upstream -8.594 0.226 <0.001 Post: Downstream vs. upstream -23.139 0.155 <0.001 Recent: Downstream vs. upstream -3.615 0.100 0.005 Downstream: prior vs. post -7.310 0.067 <0.001 Downstream: post vs. recent -8.847 0.088 <0.001 Downstream: prior vs. recent -6.577 0.132 <0.001 Upstream: prior vs. post -10.965 0.123 <0.001 Upstream: post vs. recent -4.915 0.507 <0.001 Upstream: prior vs. recent -5.703 0.230 <0.001 Figure 3. Non-metric multidimensional scaling scatterplot of Davis Branch stations. Stations were grouped into disturbance categories based on their spatial and temporal relation to beaver dams. Spatially, stations were assigned as upstream (circles) and downstream (squares) of beaver dams. Temporally, stations were assigned as prior-beaver dam construction (white), post-beaver dam construction (gray), and recent (black). MRPP results indicate that the distances in ordination space between each disturbance group was significantly different. Axis 1 indicates the temporal shift in fish community structure following beaver colonization. Axis 2 indicates the longitudinal spatial shift in fish community structure in Davis Branch from upstream to downstream. M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 120 Vol. 12, Special Issue 4 period. Prior to Beaver colonization, the fish community had distinct upstream and downstream faunas (A = 0.226, P < 0.001) and maintained that separation during the post-colonization period (A = 0.155, P < 0.001) and recent colonization period (A = 0.100, P = 0.005). Temporally, the fish assemblages were distinct within both the downstream and upstream sections, and shifted over time. Downstream, the fish community prior and post (A = 0.067, P < 0.001), prior and recent (A = 0.088, P < 0.001), and post and recent periods were separate (A = 0.132, P < 0.001). Upstream, the fish community prior and post (A = 0.123, P < 0.001), prior and recent (A = 0.507, P < 0.001), and post and recent periods were separate (A = 0.230, P < 0.001). Increased species diversity and changes in minnow and sunfish abundances were the primary factors in the observed succession in community structure (Fig. 4). Prior to Beaver colonization, 10 species were present in the upstream section, and the community was comprised predominantly of small-bodied fishes, such as Blackside Dace, Southern Redbelly Dace, Creek Chubs, and Stripetail Darters. The downstream assemblage was more diverse with 16 species, but it was still comprised predominantly of small-bodied fishes, such as Central Stoneroller, Striped Shiner, and Blacknose Dace. In addition, largerbodied species, such as sunfish, were more diverse and abundant. Following pond establishment, Northern Hog Sucker was the only species extirpated, and 4 new species immigrated into the upstream section even though the habitat changed from lotic to lentic. Downstream of the ponds, the habitat remained lotic (Fig. 2A, B), but the hydrology was altered. No species were extirpated, and 5 new species, mostly centrarchids (i.e., Lepomis gulosus [Warmouth] and Micropterus spp.), immigrated into the downstream section following pond establishment. Creek Chubs decreased across the temporal periods within both sections, but other species were mostly static (i.e., Central Stoneroller downstream, and Blacknose Dace upstream) or fluctuated among temporal periods (i.e., Striped Shiner downstream, and White Sucker upstream). Sunfish diversity and abundance increased within both sections following beaver pond establishment. Immediately following Beaver colonization (post), Blackside Dace CPUE values were not significantly different from those prior to colonization for the downstream and upstream sections (Fig. 5). However, the population during the recent period had decreased substantially and was significantly different from earlier periods (prior and post). Southern Redbelly Dace CPUE values increased significantly following the initial Beaver colonization but decreased in the following period and ultimately were not significantly different than those prior to Beaver colonization, for either the downstream and upstream sections. Changes in CPUE values for Centrarchidae in the downstream section were significant and increased during each subsequent period (Fig. 5). Warmouth, Micropterus salmoides (Largemouth Bass), and Micropterus punctulatus (Spotted Bass) were only encountered after Beaver colonization. Warmouth abundance increased greatly during the recent period. In the upstream section, Centrarchidae CPUE values were significantly different following Beaver colonization. War121 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 Figure 4. Stacked bars of the Davis Branch fish taxa across temporal periods for the upstream (A) and downstream (B) sections. In general, cyprinid abundance decreased and centrarchid abundance increased over time. BSD = Blackside Dace, SRBD = Southern Redbelly Dace. M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 122 Vol. 12, Special Issue 4 mouth and Lepomis macrochirus (Bluegill) were only encountered after Beaver colonization, and they increased in abundance over time. For both sections, the abundance of the nonnative Redbreast Sunfish increased greatly following Beaver colonization, and it was one of the most common species encountered. The Blackside Dace age class structure between the downstream and upstream sections indicated the CPUE values for each age class were greater in the upstream section, and that separate patterns were observed over time. Recent CPUE values for each age class were near zero (Fig. 6). For the downstream section, the Mann-Whitney U-test indicated that the median CPUE values for the 0+, 1+, and 2+ age classes were not significantly greater immediately following Beaver colonization (U = 118.0, P < 0.485; U = 182.5, P < 0.103; U = 176.0, P < 0.159, Figure 5. Box plots of downstream (A) and upstream (B) Blackside Dace, downstream (C) and upstream (D) Southern Redbelly Dace, and downstream (E) and upstream (F) Family Centrarchidae CPUE values across temporal periods. Significant difference of CPUE (Mann-Whitney U-test: P < 0.005) across temporal periods is indicated by a distinct letter among the box plots. 123 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 respectively), but were significantly lower after prolonged colonization (U = 27.0, P < 0.039; U = 17.0, P < 0.011 U = 22.0, P < 0.030, respectively). In the upstream section, the 0+ and 1+ Blackside Dace age class decreased immediately following colonization but were not significantly different than prior to coloniza- Figure 6. Median CPUE values for Blackside Dace age classes upstream and downstream across temporal periods. Blackside Dace were able to persist immediately after Beaver colonization but recruitment ceased at some point between 1994–2002, which contributed to the collapse of the population. M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 124 Vol. 12, Special Issue 4 tion (U = 41.0, P < 0.156; U = 49.0, P < 0.358, respectively). The 2+ age class CPUE was significantly greater immediately following colonization (U = 100.0, P < 0.027). However, after prolonged (greater than 10 years) colonization all age classes were significantly lower than the previous period (U = 22.0, P < 0.041; U = 4.5, P < 0.001 U = 19.0, P < 0.031, respectively). Discussion Distinct upstream and downstream fish assemblages were identified within Davis Branch prior to Beaver colonization. The dynamic spatial and temporal activities of Beavers, pond age, pond placement within the watershed, valley shape, and the larger-scale stream network can influence the successional shift in community structure and diversity following pond establishment (Johnston and Naiman 1987, Schlosser and Kallemeyn 2000, Snodgrass and Meffe 1998). In addition, dam integrity and precipitation can influence dispersal rates and the connectivity of downstream and upstream fishes (Schlosser 1995, Schlosser and Kallemeyn 2000, Snodgrass and Meffe 1998). Intact dams and periods of reduced precipitation decrease upstream colonization from downstream sources, reduce hydrologic dynamics, and can result in decreased dissolved oxygen levels, thus selecting for more tolerant species (Schlosser and Kallemeyn 2000). In our study, the upstream and downstream sections maintained distinct assemblages following pond establishment, but exhibited similar successional community-structure shifts. In general, centrarchid diversity and abundance increased within the upstream and downstream sections. Redbreast Sunfish and Southern Redbelly Dace became the predominant species, while Blackside Dace and Creek Chub abundances decreased. Most other species were static or did not exhibit a noticeable pattern. The Blackside Dace population initially showed little change after pond establishment within the upstream and downstream sections. However, after prolonged Beaver colonization (greater than 10 years), the population eventually collapsed in the upstream section, with only one individual encountered since 2007. The extreme upstream section contained a small reach of lotic habitat likely essential for spawning (Fig. 2D), but a series of droughts during the past decade reduced that habitat and forced the remaining individuals into the inundated stream portion (Fig. 2C). Recruitment was never substantial in the downstream section (Fig. 6), suggesting suitable spawning habitat was only present within the upstream section prior to Beaver colonization. We speculate that Blackside Dace were heavily preyed upon as centrarchid diversity and abundance increased within the lentic habitat, and the cumulative factors of habitat alteration, drought, decreased downstream connectivity, and reduced spawning habitat most likely resulted in the collapse of the Blackside Dace population within the upstream section. The downstream Blackside Dace population collapsed after prolonged Beaver colonization, but it did not show a measurable difference immediately following pond establishment. Prior to Beaver colonization, the Blackside Dace population within the downstream section consisted mostly of adult individuals (Fig. 6), 125 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 which suggested minimal recruitment and that colonization would have to come either from upstream, from Little Yellow Creek, or a combination of the two. The shift in community structure was similar within the downstream and upstream sections, and predation from centrarchid species most likely played a role in the population dynamics, but the abiotic factors that contributed to the population collapse within the downstream section were different from those in the upstream section. Droughts and the establishment of beaver dams over the last decade have diminished the hydrologic fluxes downstream, thus reducing habitat complexity and likely concentrating species in pools, which increased predation success. Dam presence also limited the colonization from upstream to downstream sites, which increased the importance of Little Yellow Creek as a source for Blackside Dace into Davis Branch. However, the downstream reaches of Little Yellow Creek and most tributaries of Yellow Creek have been degraded by a variety of anthropogenic activities; therefore, water quality and habitat conditions have not supported a robust population of Blackside Dace. Other factors not measured in our study but which may have contributed to the changes in community structure and Blackside Dace populations after Beaver colonization are alterations in the nutrient cycle and temperature regime. It has been documented that the abundance of standing material increases, and decomposition and transportation rates of material are altered following pond establishment (Hodkinson 1975, Francis et al. 1985, McDowell and Naiman 1986, Naiman et al. 1991, Smith et al. 1991). Aquatic plant, invertebrate, and fish communities are also altered by changes in water quality (Clifford et al. 1993, Feldman 1995, Smith et al. 1991). Although only a few species demonstrated a significant change in abundance following Beaver colonization, several species showed consistent increases in abundance over time (i.e., Central Stoneroller, Striped Shiner, Bluntnose Minnow, and White Sucker). These species are associated with nutrient and temperature increases in eastern Kentucky’s mountain headwater streams (Compton et al. 2003). As Beaver colonization persists, canopy cover within the watershed decreases, resulting in increased temperatures and primary production, as well as increased coarse particulate organic material within the stream. Management implications The presence and influence of beaver colonization within Blackside Dace streams is not a unique phenomenon, and the two species have coexisted for centuries. In our study, we demonstrated that prolonged Beaver colonization was detrimental to the local persistence of Blackside Dace. However, in the largerscale context of a drainage basin (i.e., Cumberland), the two species can persist through source/sink dynamics (Pulliam 1988), where Beaver activity influences the dispersal of fishes in and out of ponds, based on their habitat tolerances (Schlosser 1995). Meffe and Sheldon (1988) and Snodgrass and Meffe (1998) suggested that Beaver colonization influenced the evolution of stream fishes, where the distribution of impounded, unimpounded, and previously impounded M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 126 Vol. 12, Special Issue 4 streams created a mosaic of habitats, and species’ habitat preferences allowed for resource partitioning and persistence. Unfortunately, Beaver and Blackside Dace distributions across eastern Kentucky no longer resemble pre-settlement distributions due to landscape alteration and stream degradation. Streams that have high water quality, like Davis Branch, serve as refuges for both species. Historically, the Blackside Dace population within Davis Branch was recognized as an important population in the eastern extent of the species’ range (O’Bara 1990); therefore, its restoration within Davis Branch would be beneficial for the dace’s overall recovery. Below, we recommend several management actions to benefit the Federally threatened Blackside Dace within the Cumberland Gap National Park, with the goal of species restoration and maintenance within Davis Branch: 1. Removal and management of Beavers within Davis Branch. 2. Removal of existing beaver dams and restoration of the stream channel and flow patterns present prior to Beaver colonization. 3. Restoration of the riparian zone. 4. Development of a systematic program to significantly reduce the nonnative Redbreast Sunfish population and to decrease native centrarchid species abundances. 5. Establish a Blackside Dace propagation and reintroduction program for Davis Branch. Acknowledgments We would like to thank Mark Woods (Park Superintendent), Jenny Beeler (Biologist), Megan Blanchard (GIS Specialist), Amy Wilson (Biologist) and other CUGA staff for their invaluable assistance with many aspects of this project, but especially for the creation of the map used in this paper. Field assistance was provided by Carrie Allison, Jonathan Baxter, Dirk Bradley, Stephanie Brandt, Sue Bruenderman, Jeff Duncan, Danny Parks, Tim Pinion, Matt Thomas, and John Williams. Virgil Lee Andrews, Jr. (Field Supervisor, Kentucky Ecological Services Field Office) provided financial support in the form of equipment and travel expenses. The Kentucky Aquatic Resource Fund provided financial support for publication costs. Disclaimer The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the US Fish and Wildlife Service. Literature Cited Apple, L.L., B.H. Smith, J.D. Dunder, and B.W. Baker. 1984. The use of Beavers for riparian/aquatic habitat restoration of cold-desert gulley-cut stream systems in southeastern Wyoming. Pp. 123–130, In Proceedings, American Fisheries Society/Wildlife Society Joint Chapter Meeting, February 8–10, Logan, UT. Bray, J.R., and J.T. Curtis. 1957. An ordination of the upland forest communities in southern Wisconsin. Ecological Monographs 27:325–349. Clifford, H.E, G.M. Wiley, and R.J. Casey. 1993. Macroinvertebrates of a Beaver-altered 127 M.C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist Vol. 12, Special Issue 4 boreal stream of Alberta, Canada, with special reference to fauna on the dam. Canadian Journal of Zoology 71:1439–1447. Collen, P., and R.J. Gibson 2001. The general ecology of beavers (Castor spp.), as related to their influence on stream ecosystems and riparian habitats, and the subsequent effects on fish: A review. Reviews in Fish Biology and Fisheries 10:439–461. Compton, M.C., G.J. Pond, and J.F. Brumley. 2003. Development and application of the Kentucky Index of Biotic Integrity (KIBI). Kentucky Department for Environmental Protection, Division of Water, Frankfort, KY. Cook, D.B. 1940. Beaver-trout relations. Journal of Mammalogy 21:397–401. Dubuc, L.J., W.B. Krohn, and R.B. Owen. 1990. Predicting occurrence of River Otters by habitat on Mount Desert Island, Maine. Journal Wildlife of Management 54:594–599. Etnier, D.A., and W.C. Starnes. 1993. The Fishes of Tennessee. The University of Tennessee Press, Knoxville, TN. 689 pp. Feldmann, A.L. 1995. The effects of Beaver (Castor canadensis) impoundment on plant diversity and community composition in the coastal plain of South Carolina. M.Sc. Thesis, University of Georgia, Athens, GA. France, R.L. 1997. The importance of Beaver lodges in structuring littoral communities in boreal headwater lakes. Canadian Journal of Zoology 75:1009–1013. Francis, M.M., R.J. Naiman, and J.M. Melillo. 1985. Nitrogen fixation in subarctic streams influenced by Beaver (Castor canadensis). Hydrobiologia 121:193–203. Gard, R. 1961. Effects of Beaver on trout in Sagehen Creek, California. Journal Wildlife Mananagement 25:221–242. Gibson, R.J. 1969. East Blue Lake fisheries investigation, 1967–1968. Manitoba Department of Mines and Natural Resources, Fisheries Branch, Report Number 69-7. 85 pp. Grasse, J.E. 1951. Beaver ecology and management in the Rockies. Journal of Forestry 49:3–6. Grover, A.M., and G.A. Baldassarre. 1995. Bird species richness within beaver ponds in south-central New York. Wetlands 15:108–118. Hanson, W.D., and R.S. Campbell. 1963. The effects of pool size and Beaver activity on distribution and abundance of warm-water fishes in a north Missouri stream. American Midland Naturalist 69:136–149. Hodkinson, I.D. 1975. Energy flow and organic matter de-composition in an abandoned beaver pond ecosystem. Oecologia 21:131–139. Johnston, C.A., and R.J. Naiman. 1987. Boundary dynamics at the aquatic–terrestrial interface: The influence of Beaver and geomorphology. Landscape Ecology 1:47–57. Knudsen, G.J. 1962. Relationship of Beaver to forests, trout, and wildlife in Wisconsin. Technical Bull. No. 25, Wisconsin Conservation Department, Madison, WI. 50 pp. McCune, B., and M.J. Mefford. 2011. PC-ORD. Multivariate analysis of ecological data, version 6.0. MjM Software. Gleneden Beach, OR. McCune, B., and J.B. Grace. 2002. Analysis of ecological communities. MjM Software Design, http://www.mjmsoft.com/. Macdonald, D.W., F.H. Tattersall, E.D. Brown, and D. Balharry. 1995. Reintroducing the European Beaver to Britain: Nostalgic meddling or restoring biodiversity. Mammal Review 25:161–200. McDowell, D.M., and R.J. Naiman. 1986. Structure and function of a benthic invertebrate stream community as influenced by Beaver (Castor canadensis). Oecologia (Berlin) 68:481–489. Muller-Schwarze, D., and L. Sun. 2003. The Beaver: Natural History of a Wetlands EngiM. C. Compton, M.A. Floyd, and D.E. Stephens 2013 Southeastern Naturalist 128 Vol. 12, Special Issue 4 neer. Comstock Publishing Associates, Cornell University Press, Ithaca, NY. 189 pp. Munther, G.L. 1983. Integration of Beaver into forest management. Pp. 73–80, In Proceedings of the 18th Annual Meeting of the American Fisheries Society, Laramie, WY. Naiman, R.J., T. Manning, and C.A. Johnston. 1991. Beaver population fluctuations and tropospheric methane emission in boreal wetlands. Biogeochemistry 12:1–15. Neff, D.J. 1957. Ecological effects of Beaver habitat abandonment in the Colorado Rockies. Journal of Wildlife Management 21:80–84. O’Bara, C.J. 1990. Distribution and ecology of the Blackside Dace, Phoxinus cumberlandensis (Osteichthyes: Cyprinidae). Brimleyana 16:9–15. Parker, M. 1986. Beaver, water quality and riparian systems. Pp. 88–94, In Proceedings: Wyoming Water and Streamside Zone Conference. Wyoming Water Research Center, University of Wyoming, Laramie, WY. Pullen, T.M. 1971. Some effects of Beaver (Castor canadensis) and beaver pond management on the ecology and utilization of fish populations along warm-water streams in Georgia and South Carolina. Ph.D. Dissertation. University of Georgia, Athens, GA. 81 pp. Pulliam, R.H. 1988. Sources, sinks, and population regulation. American Naturalist 132:652–661. Rohde, F.C., and R.G. Arndt. 1991. Distribution and status of the Sandhills Chub, Semotilus lumbee, and the Pinewoods Darter, Etheostoma mariae. Journal of the Elisha Mitchell Society 107:71–70. Rupp, R.S. 1955. Beaver-trout relationship in the headwaters of Sunkhaze Stream, Maine. Transactions of the American Fisheries Society 84:5–85. Schlosser, I.J. 1995. Dispersal, boundary processes, and trophic-level interactions in streams adjacent to beaver ponds. Ecology 76:908–925. Schlosser, I.J., and L.W. Kallemeyn. 2000. Spatial variation in fish assemblages across a Beaver-influenced successional landscape. Ecology 81:1371–1382. Smith, M.E., C.T. Driscoll, B.J. Wyskowski, C.M. Brooks, and C.C. Cosentini. 1991. Modification of stream eco-system structure and function by Beaver in the Adirondack Mountains, New York. Canadian Journal of Zoology 69:55–61. Snodgrass, J.W., and G.K. Meffe. 1998. Influence of Beavers on stream fish assemblages: Effects of pond age and watershed position. Ecology 79:928–942. Starnes, W.C. 1981. Listing package for the Blackside Dace, Phoxinus cumberlandensis. US Fish and Wildlife Service, Asheville, NC. 23 pp. Starnes, W.C., and L.B. Starnes. 1978. Status report on a new and threatened species of Phoxinus from the upper Cumberland drainage. Proceedings of the Southeastern Fishes Council 2:1–3. US Fish and Wildlife Service. 1988. Blackside Dace recovery plan. Atlanta, GA. 23 pp. Woods, A.J., J.M. Omernik, W.H. Martin, G.J. Pond, W.M. Andrews, S.M. Call, J.A. Comstock, and D.D. Taylor. 2002. Ecoregions of Kentucky (color poster with map, descriptive text, summary tables, and photographs, map scale 1:1,000,000): US Geological Survey, Reston, VA.