2006 SOUTHEASTERN NATURALIST 5(1):69–84 Distribution and Population Biology of the Black Warrior Waterdog, Necturus alabamensis MICHELLE C. DURFLINGER MORENO1,2, CRAIG GUYER1,*, AND MARK A. BAILEY3 Abstract - Streams were sampled throughout the Upper Black Warrior River basin to (1) determine the current distribution of Necturus alabamensis (Black Warrior Waterdog) and (2) identify its habitat requirements. Individuals were observed at 14 of 112 localities within 60 streams. Discriminant function analysis of stream characteristics revealed that waterdog populations were associated with the presence of ephemeropteran larvae on a local scale and with water depths of 1–4 m, large leaf packs, and low percentage of fine sediments on a regional scale. Where present, N. alabamensis occurs at relatively low population densities, and populations monitored in our survey exhibited a 1:1 sex ratio with no apparent age-class structure or sexual size dimorphism. We conclude that this species is rare on the basis of its restricted geographic range, low abundance, and unpredictable occurrence in suitable habitat; as such, N. alabamensis should be considered for federal protection. Introduction In his revision of the genus Necturus, Viosca (1937) described a species now known as Necturus alabamensis (Black Warrior Waterdog; Fig. 1). This permanently aquatic salamander is restricted to scattered locations within the Upper Black Warrior River drainage of northwestern Alabama (Ashton and Peavy 1986, Bart et al. 1997). Most specimens of N. alabamensis have been collected within the Bankhead National Forest in Lawrence, Winston, and Walker counties, where the species often occurs in accumulated leaf litter in stream beds from early November through early March (Neill 1963). However, consistent captures are confined largely to Sipsey Fork. Therefore, Black Warrior Waterdogs appear to be rare, not only because of their restricted geographic range, but also because they may occur only at low densities in a small number of streams within the range. 1Department of Biological Sciences, Auburn University, Auburn, AL 36849. 2Current address - US Fish and Wildlife Service, 2730 Loker Avenue West, Carlsbad, CA 92008. 3Conservation Southeast, Inc., 7746 Boggan Level Road, Andalusia, AL 36420. *Corresponding author - cguyer@acesag.auburn.edu. Figure 1. Adult Necturus alabamensis (Black Warrior Waterdog). Photograph © by Greg Sievert. 70 Southeastern Naturalist Vol. 5, No. 1 Despite their apparent rarity, Black Warrior Waterdogs currently are not protected by state and/or federal regulations, largely because the taxonomic status of N. alabamensis was confused until recently (Bart et al. 1997). Moreover, the current distribution, habitat requirements, and key life history characteristics of this species are not well known, making identification of optimal stream habitat for conservation purposes difficult. In this study, we delimit the current distribution of N. alabamensis based on stream surveys throughout the presumed geographic range and describe habitat characteristics associated with sites possessing this species. We also characterize demographic parameters of the species at Sipsey Fork, the best known population. Methods Regional samples We surveyed 112 sites throughout the Upper Black Warrior River basin. Ninety-nine sites were visited during the winters of 1991 and 1992. Of these, 22 sites were revisited from November 1994 through March 1995, and all original 99 sites together with 13 new ones were surveyed in 1997. Sites were examined for the presence of waterdogs by removing sections of submerged leaf packs with a large dip net (6.35-mm mesh); sample contents were placed on the stream bank and sifted for Necturus. This process was repeated until the entire leaf pack had been processed or until 10 parcels of a large leaf pack had been sampled. In 1997, habitat characteristics were measured at each of the 112 localities. We used stream crossings to access each site, and the samples were collected along a 100-m transect oriented along the mid-line of each stream. Substrate composition, stream depth, water temperature, and leaf pack length (longest horizontal axis), width (greatest width perpendicular to length), and depth were recorded at 10-m intervals. Substrate composition was determined by estimating the linear distance along each 10-m interval that crossed rock, sand, and sediments (fine organic and inorganic matter). These data were converted into a percentage of each substrate type along the entire transect. We then sampled leaf packs with dip nets, as described previously, recording the presence of plethodontid salamanders, fish, crayfish, mollusks, and aquatic insect larvae, as well as Necturus. Because several streams contained more than one sampling site, we pooled within-stream samples, calculating mean values for the variables described above. This pooling of data yielded 60 streams for use in further analyses (Appendix 1). Local samples We intensively sampled Sipsey Fork, a third-order stream in the Black Warrior River drainage near Alabama Highway 33 (T9S R8W S33 and 34, Bailey and Guyer 1998). The stream originates in the Sipsey Wilderness, flows through the Bankhead National Forest, and drains into Lewis Smith Lake in Winston County. We selected a 100-m stream section starting from 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 71 the cement entry point for canoes and oriented upstream from that locality. The total area for the site was approximately 834 m2. We conducted a mark-recapture study of Necturus at Sipsey Fork during January–May of 1999 through 2001, using the dip-net sampling technique, as described previously. Samples were taken once each month during 1999 and 2000 and three times per week during 2001. During 1999 and 2000, 15 minnow traps (32 cm x 62 cm) were placed on leaf packs or along submerged logs and rocks throughout the study site. In 2001, a 3 x 11 trapping grid was established with traps placed at 10-m intervals along 3 parallel transect lines, one in the center of the stream and one along each bank. Traps were baited with wet cat food placed inside perforated plastic 35-mm film containers, and bait was replaced weekly. We recorded total length (TL; nearest 1.0 mm) and wet mass (nearest 0.01 g) for all captured individuals. Large individuals (> 100 mm TL) were identified as to sex (following Shoop 1965) and were given a unique mark by clipping toes (no more than one toe per foot). Subadult and larval specimens were not marked. All waterdogs were returned to their site of capture. During 2001, the trapping grid was used to create sampling localities for data describing the abiotic and biotic environment at the Sipsey Fork site. The abiotic environment was characterized by recording substrate composition, stream depth, water temperature, leaf-pack width, and leaf-pack depth at each grid locality. Substrate composition was sampled within a 0.5 m2 metal frame used to estimate the percentage of area covered by rock, sand, and fine sediments. To characterize the biotic environment, sections of leaf packs and underlying substrate were removed with a dip net, as described previously, for each 10-m interval along the three transects (stream center and the two stream edges). The presences of plethodontid salamanders, fish, crayfish, mollusks, and insect larvae were recorded. All insect larvae were preserved in ethanol (70%) and subsequently identified to order. Statistical analyses To describe the current distribution of Necturus alabamensis, we determined the map coordinates for the mouth of each stream examined. These were georeferenced and processed as two categories in an Arc/Info Geographical Information System database: sites known to contain waterdogs, and sites not known to contain waterdogs. We used discriminant function analysis (SPSS 1999) to determine whether habitat characteristics at local and/or regional scales could be used to predict sites that contained N. alabamensis versus those that did not. For the local scale, 14 abiotic and biotic variables were used to create the discriminant function (Table 1). To determine the separate contribution of each habitat variable to the overall model, we used the stepwise procedure (SPSS 1999). For variables that were significantly associated with Necturus, we used box plots to determine the direction (positive or negative) of each association. Finally, we used G-tests to determine whether the significant predictor variables were associated with each other. 72 Southeastern Naturalist Vol. 5, No. 1 For regional scale analysis, 23 environmental variables (Table 1) were partitioned into three sets: (1) abiotic features (5 variables); (2) biotic features (13 variables); and (3) features of leaf packs (5 variables). Because the original data were heavily skewed toward sites without waterdogs (45 of 60 sites), we randomly selected 10 sites without waterdogs for analysis. To decrease the probability that results were affected by specific stream combinations, we repeated this random selection process 20 times to create 20 trials. For each trial, we used stepwise discriminant analysis separately on each of the three categories listed above to examine the contribution of each variable to an overall model. We then performed a discriminant analysis (separately for each trial) to determine the significance of the overall model. This series of analyses was then repeated, but restricted to the 14 variables analyzed at the local scale (Table 1). To summarize overall trends from this series of analyses, we used the combined probabilities test (Sokal and Rohlf 1969) to determine the overall probability that a particular variable contributed to discriminating sites with waterdogs from those without. This procedure was repeated to evaluate whether all variables contributed to a significant Table 1. Summary of variables used in the discriminant function analyses to determine characteristics of sites with and without Necturus alabamensis. Columns indicate variables used (X) for local analysis of Sipsey Fork site and regional analyses of abiotic, biotic, and leaf pack features throughout the Upper Black Warrior River basin. A fifth discriminant function analysis was performed at a regional scale for the variables indicated under local analysis. Measurements of distance are in m; taxon variables represent presence or absence of that taxon at a particular site. Local Regional analysis Variable analysis Abiotic Biotic Leaf packs Stream width X Stream depth X X % sand X X % rock X X % sediment X X Desmognathus X Pseudotriton X Unknown salamander X Crayfish X X Fish X X Snail X X Odonata X X Plecoptera X X Trichoptera X X Diptera X Ephemeroptera X X Corbicula X Mussel X # of leaf packs X Leaf-pack length X Leaf-pack width X X Leaf-pack depth X X Distance to bank X 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 73 overall model. Finally, we used G-tests to determine whether the significant predictor variables were associated with each other. The capture data were used to characterize the waterdog population at Sipsey Fork. We used histograms to depict patterns of TL and mass within the population. Kolomogorov-Smirnov and Shapiro-Wilk normality tests were used to compare the observed distributions to a normal distribution. We used G-tests to compare the observed sex ratio to an expected 1:1 ratio and independent-samples t-tests to compare the average TL and mass of males and females. Results Of the 60 streams and 112 localities sampled within the Upper Black Warrior River basin, Necturus alabamensis were found at 14 localities (Fig. 2). Figure 2. Distribution of sites known to contain Necturus alabamensis (solid circles) and sites not known to contain N. alabamensis (open circles). Solid line indicates the Fall Line. Inset shows location of Upper Black Warrior River basin (shaded) within Alabama. 74 Southeastern Naturalist Vol. 5, No. 1 Thus, Black Warrior Waterdogs were found at only 12% of localities and 24% of streams sampled. The total number of visits and the number of N. alabamensis captured varied among streams (Table 2). The average number of visits to sites containing N. alabamensis was 7 (SD = 15.7; N = 14), and the average number of times that at least one N. alabamensis was observed at these sites during a visit was 4 (SD = 8.7; N = 14). Thus, the chance of observing N. alabamensis at sites known to contain them was 0.57. On a local scale, stepwise discriminant function analysis revealed that the presence of ephemeropterans was significantly associated with the presence of waterdogs (X2 = 7.7, df = 1, p = 0.001; Fig. 3A). However, an Table 2. Summary of visitations to sites containing Necturus alabamensis. Columns indicate the number of times a site was visited, the number of visits during which N. alabamensis was observed, and the total number of N. alabamensis captured during all visits. Number of Number of visits Number of Site visits with waterdogs waterdogs Blackburn Fork 5 2 2 Blackwater Creek 3 2 2 Browns Creek 1 1 1 Brushy Creek 5 3 3 Capsey Creek 1 1 1 Carroll Creek 1 1 1 Little Blackwater 4 2 2 Lost Creek 4 2 2 Lye Branch 1 1 1 Mulberry Fork 2 1 1 North River 3 1 1 Sipsey Fork 61 34 59 Slab Creek 1 1 1 Yellow Creek 2 1 1 Table 3. Sites misclassified by discriminant function analysis as containing Necturus alabamensis. Township, range, and section as in Appendix 1. Location County Big Sandy Creek Tuscaloosa Binton Creek Tuscaloosa Bluewater Creek Fayette Cabin Creek Cullman Calvert Prong Blount Cane Creek Fayette Hurricane Creek Cullman Inman Creek Winston Lewis Smith Lake Winston Little Sandy Creek Tuscaloosa Mill Creek Jefferson Pan Creek Winston Rock Creek Winston Rush Creek Cullman Sandy Creek Winston Splunge Creek Winston Valley Creek Jefferson 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 75 analysis considering all 14 environmental variables simultaneously was not significant (X2 = 7.8, df = 7, p = 0.35). Figure 3. Association of A) Ephemeroptera with Necturus alabamensis at the Sipsey Fork study site and of B) stream depth, C) sediments, D) leaf-pack length, E) Ephemeroptera, and F) Trichoptera with N. alabamensis throughout the Upper Black Warrior River basin. Error bars indicate 95% confidence interval. Figure 4. Larval N. alabamensis. Photograph © by Greg Sievert. 76 Southeastern Naturalist Vol. 5, No. 1 For abiotic and biotic stream characteristics on a regional scale, stepwise analysis followed by a combined probabilities test revealed that stream depth and percent of stream substrate covered by fine sediments were negatively associated with Necturus presence (Figs. 3B and C), whereas ephemeropteran presence and leaf-pack length were positively associated with Necturus presence (Figs. 3D and E). All variables were independent of one another. In each Figure 5. Frequency distributions of A) total length and B) mass for larval (shaded bars), adult female (white bars), and adult male (dark bars) Necturus alabamensis at Sipsey Fork during 1999–2001. 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 77 of the three groups of variables, an analysis considering all variables simultaneously was not significant (X2 = 5.5, df = 40, p = 0.40 for abiotic stream features; X2 =14.1, df = 40, p = 0.44 for biotic stream features; X2 = 3.9, df = 40, p = 0.70 for leaf-pack features). When stepwise analysis was restricted to the 14 variables examined for local habitats, the presences of trichopterans (X2 = 58.6, df = 40, p = 0.05; Fig. 3F) and ephemeropterans (X2 = 113.7, df = 40, p = 0.001) were found to be positively associated with the presence of Necturus. Furthermore, presences of trichopterans and ephemeropterans were significantly and positively associated with each other (G = 17.01, df = 1, p > 0.001). An analysis considering all variables simultaneously also was significant (X2 = 82.1, df = 40, p = 0.001). The model correctly classified 28 of 45 sites (62%) without waterdogs and 10 of 14 sites (71%) with them. The 17 streams currently lacking specimen vouchers for N. alabamensis but classified by discriminant function analysis as appropriate habitat were distributed in six counties throughout the presumed range of the Black Warrior Waterdog (Table 3). Fifty-two waterdogs were captured during 173,160 trap hours at Sipsey Fork. Of these, 35 were adults (> 100 mm TL; 17 males and 18 females), 5 were subadults (< 100 mm TL; sex not determinable), and 12 were larvae (Fig. 4). Only two individuals were recaptured, one in 2000 (marked in 1999), and one in 2001 (marked in 2000). Adult TL ranged from 145 to 210 mm and body mass ranged from 8.1 to 40.9 g. The frequency distributions for length and mass did not differ from a normal distribution (D = 0.12, p = 0.20; W = 0.97, p = 0.54 for length; D = 0.11, p = 0.20; W = 0.96, p = 0.34 for mass; Fig. 5). TL and body mass did not differ significantly between adult males and females (t = 0.90, p = 0.38 for males; t = 0.62, p = 0.54 for females). Numbers of adult males and females did not differ significantly from an expected 1:1 sex ratio (G = 0.874; p = 0.50). Discussion Although the Upper Black Warrior River basin is a network of more than 150 first-, second-, and third-order streams, our survey of 112 localities within 60 streams confirmed the presence of Necturus alabamensis at only 14 localities (representing 14 streams). These localities span five counties, indicating that this species is still widely distributed within its presumed historical range (Ashton and Peavy 1986). However, our data indicate that this species is present in 23% of streams and 12% of localities within this range. Even for sites harboring known populations, waterdogs are captured only 57% of the time that these localities are sampled. With the exception of Sipsey Fork, our observations were generally limited to one or two captures. Over the period of intense sampling at Sipsey Fork, waterdogs were captured at 5 of 30 trapping stations. These five stations were located within deep pools containing leaf packs inhabited by aquatic insects, but also containing sunken logs and large flat rocks. These habitat features are similar to 78 Southeastern Naturalist Vol. 5, No. 1 descriptions by Fedak (1971) for Necturus punctatus (Gibbes), Ashton (1985) for N. lewisi (Brimley), and Shoop (1965) for N. beyeri Viosca. Habitat analyses on a regional scale show similarities to those on a local scale. Discriminant function analyses of abiotic and biotic stream characteristics indicate that stream depth, percent of stream substrate covered by fine sediments, leaf-pack length, and presence of ephemeropterans (and perhaps trichopterans) are significantly associated with the presence of N. alabamensis. Waterdog presence is negatively associated with stream depth on a regional scale, but positively associated with stream depth at a local scale. This seemingly incongruous pair of results stems from the fact that variation in water depth at the local scale is much less than that recorded for the regional scale. Therefore, our data suggest that N. alabamensis occupies a narrow range of water depths from 1–4 m. In deeper, more central portions of streams, water flow is faster, possibly preventing development of leaf beds and decreasing food availability. In shallower waters, temperature may be elevated and water flow minimized, resulting in decreased levels of dissolved oxygen. Ashton (1985) considered high level of dissolved oxygen to be an important factor for habitat selection by N. lewisi. Based on their similar morphologies and potentially similar physiology, we speculate that this factor is also important for N. alabamensis. The percentage of stream substrate covered by fine sediments was also negatively associated with the presence of N. alabamensis. This association suggests that waterdogs are vulnerable to increased levels of sedimentation linked to certain types of land use (e.g., strip mining, intensive forestry, road construction); such alterations in sedimentation rates are common within the waterdog’s historic range (Hartfield 1990). Heavy sedimentation may reduce food availability as well as adversely affect reproduction by smothering nests and eggs (Ashton 1985). Dodd et al. (1988) documented that sedimentation in the Upper Black Warrior River system led to decreased abundance for many aquatic insects. These fine sediment particles may also coat the gills of N. alabamensis, potentially reducing the efficiency of oxygen uptake. The feeding habits of N. alabamensis are unknown, but discriminant function analysis indicates that the presence of ephemeropterans and perhaps trichopterans are significantly associated with the presence of Black Warrior Waterdogs. Braswell and Ashton (1985) found that ephemeropterans are important food items of N. lewisi, and the association of Black Warrior Waterdogs with these aquatic insect taxa may indicate trophic relationships. Alternatively, the association between N. alabamensis and these aquatic insect larvae may indicate similar habitat requirements. Both ephemeropteran and trichopteran larvae inhabit clear, cold, free-flowing waters (Edmunds and Waltz 1995, Wiens 1996) as do Black Warrior Waterdogs. 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 79 We observed a normal distribution of body length and mass in N. alabamensis, as well as a lack of sexual dimorphism in body size. Therefore, development time and age structure of N. alabamensis cannot be inferred from these size data. The ranges of TL and body mass observed in our study suggest that the Sipsey Fork population consists largely of sexually mature individuals (46%). Studies of N. beyeri and N. maculosus (Rafinesque) indicate that the average development time to sexual maturity is 5–6 years (Bishop 1941, Petranka 1998). Average size for adult N. alabamensis (170– 190 mm) is similar to that of N. maculosus (175–200 mm; Pope 1944) and N. beyeri (160–203 mm; Bishop 1943). The percentage of the population comprising sexually mature individuals at Sipsey Fork is higher than that observed in N. lewisi (34%) or N. punctatus (14%) (Fedak 1971). Similarly, our observation of a 1:1 sex ratio of N. alabamensis at Sipsey Fork mirrors that described for N. maculosus (Gibbons and Nelson 1968) and N. punctatus (Meffe and Sheldon 1987). Our recapture data were too limited to calculate population size. However, we captured Black Warrior Waterdogs at a rate of 3 individuals per 10,000 trap hours, whereas Braswell and Ashton (1985) captured N. lewisi at a rate of 6 individuals per 10,000 hours. Baited minnow traps were used as the method of capture in both studies, so comparisons should be reasonable. Both species appear to have extremely low abundances or are extremely difficult to sample. A species is defined as being rare by having a small geographic range, restricted habitat use within that range, and small populations at those locations where it is known to exist (Rabinowitz 1981). Following these criteria, the restriction of Necturus alabamensis to the Upper Black Warrior River basin, its patchy distribution within that range, and the small current population sizes at those sites identify this species as rare relative to all three components of rarity. These features document that N. alabamensis warrants consideration for protection and that long-term monitoring of this species and its key environmental variables will be required to retain it in the landscape. Low levels of sediments and a periodic input of organic matter in the form of dead leaves are optimal stream conditions for N. alabamensis. However, increased levels of silviculture and mining in this region have the potential to alter water chemistry, increase sediment loads, and decrease organic input and leaf pack development within streams of the Upper Black Warrior River basin (Dyer 1982, Knight and Newton 1977). The fact that N. alabamensis appears locally abundant only at one site makes this species particularly vulnerable to stochastic catastrophic events associated with regional human activities. However, of the 45 sites without waterdogs, our regional discriminant function model misclassified 17 sites, which suggests that localities with suitable habitat features are available for translocation projects or habitat management to encourage development of a broader distribution of populations of similar density to that at Sipsey Fork. 80 Southeastern Naturalist Vol. 5, No. 1 Acknowledgments We are grateful to Chris Adams, Erik Corredor, Matt Green, Marisa Lee-Sasser, Keith Patton, Stacy Pugh-Towe, and Maggie Smith for assistance with field work. Emmett Blankenship assisted with marking of animals. The staff at Camp McDowell, especially Mark Johnston, provided us many courtesies that aided our ability to perform research in the remote areas of the Bankhead National Forest. The text was improved by suggestions from Jack Feminella, George Folkerts, and two anonymous reviewers. However, all remaining faults are entirely our own. This research was performed under Auburn University IACUC protocol numbers 9912-R-0835 and 0403-R-2416. Literature Cited Ashton, Jr., R.E. 1985. Field and laboratory observations on microhabitat selection, movements, and home range of Necturus lewisi (Brimley). Brimleyana 10:83–106. Ashton, R.E., and J. Peavy. 1986. Black Warrior Waterdog. Pp. 63–64, In R.H. Mount (Ed.). Vertebrate Animals of Alabama in Need of Special Attention. Alabama Agricultural Experiment Station, Auburn University, Auburn, AL. 124 pp. Bailey, K.A., and C. Guyer. 1998. Demography and population status of the Flattened Musk Turtle, Sternotherus depressus, in the Black Warrior River basin of Alabama. Chelonian Conservation and Biology 3:77–83. Bart, Jr., H.L., M.A. Bailey, R.E. Ashton, Jr., and P.E. Moler. 1997. Taxonomic and nomenclatural status of the upper Black Warrior River Waterdog. Journal of Herpetology 31:192–201. Bishop, S.C. 1941. The Salamanders of New York. New York State Museum Bulletin 324:1–365. Bishop, S.C. 1943. Handbook of Salamanders. Comstock Publishing Co., Inc. Ithaca, NY. 555 pp. Braswell, A.L., and R.E. Ashton, Jr. 1985. Distribution, ecology, and feeding habits of Necturus lewisi (Brimley). Brimleyana 10:13–25. Dodd, C.K., K.M. Enge, and J.N. Stuart. 1988. Aspects of the biology of the Flattened Musk Turtle Sternotherus depressus in northern Alabama USA. Bulletin of the Florida State Museum Biological Sciences 34:1–64. Dyer, K.L. 1982. Stream water quality in the coal region of Alabama and Georgia. US Department of Agriculture, Forest Service, Northeastern Forest Experiment Station, Broomall, PA. General Technical Report NE-73. 109 pp. Edmunds, Jr., G.F., and R.D. Waltz. 1995. Ephemeroptera. Pp. 126–163, In R.W. Merritt and K.W. Cummins (Eds.). An Introduction to the Aquatic Insects of North America. Kendall-Hunt, Dubuque, IA. 862 pp. Fedak, M.A. 1971. A comparative study of the life histories of Necturus lewisi Brimley and Necturus punctatus Gibbes (Caudata: Proteidae) in North Carolina. Unpublished M.Sc. Thesis. Duke University, Durham, NC. 103 pp. Gibbons, J.W., and S. Nelson, Jr. 1968. Observations of the mudpuppy, Necturus maculosus, in a Michigan lake. American Midland Naturalist 80:562–564. Hartfield, P. 1990. Status survey for mussels in the tributaries of the Black Warrior River, Alabama. US Fish and Wildlife Service, Jackson, MS. 8 pp. 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 81 Knight, A.L., and J.G. Newton. 1977. Water and related problems in coal-mine areas of Alabama. US Geological Survey, Water-Resources Investigations 76– 130:1–51. Meffe, G.K., and A.L. Sheldon. 1987. Habitat use by Dwarf Waterdogs (Necturus punctatus) in South Carolina streams, with life history notes. Herpetologica 43:490–496. Neill, W.T. 1963. Notes on the Alabama Waterdog, Necturus alabamensis Viosca. Herpetologica 19:166–174. Petranka, J.W. 1998. Salamanders of the United States and Canada. Smithsonian Institution Press. Washington DC. 587 pp. Pope, C.H. 1944. Amphibians and Reptiles of the Chicago Area. Chicago Natural History Museum Press, Chicago, IL. 275 pp. Rabinowitz, D. 1981. Seven forms of rarity. Pp. 205–217, In H. Synge (Ed.). The Biological Aspects of Rare Plant Conservation. John Wiley, Chichester, UK. 558 pp. Shoop, C.R. 1965. Aspects of reproduction in Louisiana Necturus populations. American Midland Naturalist 74:357–367. Sokal, R.R., and F.J. Rohlf. 1969. Biometry. W.H. Freeman and Company, San Francisco, CA. 776 pp. SPSS, INC. 1999. SPSS Base 10.0 User’s Guide. SPSS, Inc., Chicago, IL. 537 pp. Viosca, Jr., P. 1937. A tentative revision of the genus Necturus with descriptions of three new species from the southern Gulf drainage area. Copeia 1937:120–138. Wiens, J.A. 1996. Wildlife in patchy environments: Metapopulations, mosaics, and management. Pp. 53–84, In D.R. McCullough (Ed.). Metapopulations and Wildlife Conservation. Island Press, Washington, DC. 429 pp. 82 Southeastern Naturalist Vol. 5, No. 1 Appendix 1. Sites sampled within the Black Warrior River basin of Alabama for use in habitat analyses and analysis of current distribution of Necturus alabamensis. Site number refers to locations pooled for habitat analyses at the regional scale (NS = sites not sampled for analysis at regional scale but used for Necturus distribution). Site letters refer to specific-localities sampled for Necturus and habitat data. TRS = township, range, and section. Site # Stream TRS County Necturus 1 Barbee Creek t18s r11w sec 20 Tuscaloosa No 2 Beckworth Creek t9s r1e sec 29 Cullman No 3a Big Mud Creek tl0s r3e sec 27 Marshall No 3b Big Mud Creek tl0s r3e sec 29 Blount No 4a Big Sandy Creek t24n r6e sec 14 Tuscaloosa No 4b Big Sandy Creek t24n r6e sec 20 Tuscaloosa No 5 Big Yellow Creek t17s r8w sec 17 Tuscaloosa No 6 Binton Creek t19s r11w sec l Tuscaloosa No 7a Blackburn Fork t13s r1e sec 30 Blount No 7b Blackburn Fork t13s r1w sec 13 Blount Yes 7c Blackburn Fork t14s r1e sec 5 Blount No 8a Blackwater Creek t13s r7w sec 15 Walker Yes 8b Blackwater Creek t13s r8w sec 13/24 Walker No 8c Blackwater Creek t13s r8w sec 16 Walker No 8d Blackwater Creek t12s r8w sec 32 Walker No 8e Blackwater Creek t14s r6w sec 23 Walker No 8f Blackwater Creek t12s r9w sec 17 Winston No 8g Blackwater Creek t11s r10w sec34 Winston No 8h Blackwater Creek t12s r9w sec 24 Walker No 8i Blackwater Creek t13s r8w sec 14 Walker No 9a Blevens Creek t9s r5w sec 17 Cullman No 9b Blevens Creek t9s r5w sec 29 Cullman No 9c Blevens Creek t10s r6w sec 11 Winston No 10 Blue Creek t18s r9w sec 15 Tuscaloosa No 11a Blue Springs Creek t11s r1w sec 22 Blount No 11b Blue Springs Creek t11s r1w sec 20 Blount No 12 Blue Water Creek t16s r9w sec13 Fayette No 13 Browns Creek t12s r9w sec 10/11 Winston Yes 14 Brushy Creek t9s r7w sec 23 Winston Yes 15a Buck Creek t24n r3e sec 22 Tuscaloosa No 15b Buck Creek t13s r8w sec 8 Walker No 16 Cabin Creek t9s r1e sec 17 Cullman No 17 Calvert Prong Creek t13s r1e sec6 Blount No 18 Cane Creek t15s r10w sec 32 Fayette No 19 Capsey Creek t9s r6w sec 18 Winston Yes 20 Carroll Creek t20s r10w sec 22 Tuscaloosa Yes 21a Clear Creek t10s r9w sec 8 Winston No 21b Clear Creek t11s r8w sec 19 Winston No 21c Clear Creek t16s r11w sec 2 Fayette No 21d Clear Creek t10s r3e sec 25 Marshall No 21e Clear Creek t10s r3e sec 34 Marshall No 22 Cripple Creek t18s r10w sec 6 Tuscaloosa No 2006 M.C. Durflinger, C. Guyer, and M.A. Bailey 83 Site # Stream TRS County Necturus 23a Crooked Creek t15s r3w sec 17 Jefferson No 23b Crooked Creek t10s r4w sec 6 Cullman No 23c Crooked Creek t11s r5w sec 3 Cullman No 23d Crooked Creek t9s r4w sec 32 Cullman No 23e Crooked Creek t20s r2w sec 2 Tuscaloosa No 24 Davis Creek t20s r2w sec 2 Tuscaloosa No 25 Dorsey Creek t13s r4w sec 20 Cullman No 26a Duck River t9s r1w sec 32 Cullman No 26b Duck River t10s r2w sec 25 Cullman No 27 Frost Creek t15s r9w sec 28 Walker No 28 Grant Creek t24n r4e sec 5 Tuscaloosa No 29a Gurley Creek t14s r2w sec 21 Jefferson No 29b Gurley Creek t14s r2w sec 23 Jefferson No 30 Guthrie Creek t15 r8w sec13 Walker Yes 31a Hurricane Creek t9s r1e sec 16 Cullman No 31b Hurricane Creek t9s r1e sec 21 Cullman No 32 Inman Creek t9s r7w sec 36 Winston No 33 Little Blackwater Creek t14s r6w sec 23 Walker Yes 34 Little Sandy Creek t22s r5e sec 26 Tuscaloosa No 35a Locust Fork t11s r3e sec 6 Blount No 35b Locust Fork t12s r3e sec 15 Blount No 35c Locust Fork t12s r1 w sec 29 Etowah No 35d Locust Fork t11s r3e sec 25 Etowah No 35e Locust Fork t11s r3e sec 14 Etowah No 35f Locust Fork t12s r3e sec 11 Etowah No 36a Lost Creek t14s r8w sec 7 Walker No 36b Lost Creek t14s r8w sec 27 Walker No 36c Lost Creek t14s r9w sec 2 Walker No 36d Lost Creek t13s r9w sec 29 Walker No 37 Lye Branch t24s r7e sec 5 Tuscaloosa Yes 38 Marriott Creek t12s r3w sec 33 Cullman No 39 Mill Creek t13s r9w sec 21 Walker No 40a Mud Creek t18s r6w sec 19 Jefferson No 40b Mud Creek t19s r6w sec 10 Cullman No 41 Mulberry Fork t11s r2w sec 13 Cullman Yes 42a North River t18s r10wsec 16 Tuscaloosa Yes 42b North River t15s r10w sec 32 Fayette No 42c North River t16s r10w sec 7 Fayette No 43 Pan Creek t10s r1e sec 7 Winston No 44a Rock Creek t9s r6w sec 23/26 Winston No 44b Rock Creek t9s r6w sec 23 Winston No 44c Rock Creek t11s r4w sec4 Cullman No 45 Rush Creek t9s r7w sec 10 Winston No 46a Ryan Creek tl0s r3w sec 31 Cullman No 46b Ryan Creek t11s r5w sec 3 Cullman No 46c Ryan Creek tl0s r3w sec 30 Cullman No 47a Sandy Creek tl0s r8w sec 11 Winston No 47b Sandy Creek tl0s r8w sec 10 Winston No 47c Sandy Creek tl0s r8w sec 5 Winston No 84 Southeastern Naturalist Vol. 5, No. 1 Site # Stream TRS County Necturus 47d Sandy Creek tl0s r8w sec 11 Winston No 48 Shoal Creek t18s r7w sec 22/27 Jefferson No 49 Short Creek t17s r5 w sec 18 Jefferson No 50a Sipsey Fork t9s r8w sec 8 Winston Yes 50b Sipsey Fork t9s r8w sec 34 Winston No 51a Slab Creek t9s r3e sec 33 Marshall Yes 51b Slab Creek tl0s r2e sec 12 Marshall No 52a Lewis Smith Lake t12s r7w sec 24 Winston No 52b Lewis Smith Lake t12s r7w sec 13 Winston No 53a Splunge Creek tl1s r9w sec 32 Winston No 53b Splunge Creek tl1s r10w sec24 Winston No 54 Sullivan Creek t13s r4w sec 33 Cullman No 55 Turkey Creek t15s r3w sec 13 Jefferson No 56a Valley Creek t18s r6w sec 23 Jefferson No 56b Valley Creek t18s r6w sec 10 Jefferson No 57 Village Creek t16s r5w sec 22 Jefferson No 58a Wolf Creek t14s r9w sec 34 Walker No 58b Wolf Creek t15s r9w sec 15 Walker No 59 Wynnville Creek t11s r2e sec 3 Blount No 60a Yellow Creek t19s r9w sec 22 Tuscaloosa Yes 60b Yellow Creek t20s r9w sec 2 Tuscaloosa No NS Black Warrior River t11s r3e sec 6 Tuscaloosa No NS Borden Creek t8s r8w sec 28 Lawrence No NS Indian Creek t17s r8w sec 27 Walker No NS Little Warrior River t13s r1w sec 13 Blount No NS Little Warrior River t13s r1w sec 30 Blount No NS Murphy Creek t13s r3w sec 11 Blount No NS Thacker Creek t13s r3w sec 12 Cullman No