Southeastern Naturalist E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 252 2017 SOUTHEASTERN NATURALIST 16(2):252–260 A Novel Bait for Capturing Eastern Musk Turtles Eric C. Munscher1,2,*, Andrew D. Walde2,3, Angela M. Walde3, Brian P. Butterfield4, and Nicole Salvatico5 Abstract – Freshwater-turtle ecologists rely on effective baits to capture target species. We conducted a bait-preference study for Sternotherus odoratus (Eastern Musk Turtle) in 2 phases during the 2014, 2015, and 2016 field seasons at Comal Springs, New Braunfels, TX. During Phase 1, we compared the effectiveness of buffalo chicken to other malodourous or previously used successful baits for capturing Eastern Musk Turtles. During Phase 2, we compared catch rates when using buffalo chicken, raw chicken, or fried chicken as bait. We also tested for differences in the effectiveness of buffalo chicken as a bait choice between sexes and among seasons. During Phase 1, we captured 231 of 276 Eastern Musk Turtles (84%) in traps baited with buffalo chicken. During Phase 2, we captured 46 of 72 Eastern Musk Turtles (63.9%) in traps baited with buffalo chicken. We captured significantly more turtles when we used Buffalo chicken in comparison to all other baits. Our results indicate that buffalo chicken is an effective novel bait that could be used to increase capture rates of Eastern Musk Turtles. Additional studies are needed in other portions of the Eastern Musk Turtle’s range to determine if the attraction to buffalo chicken as a trap bait is a local or widespread phenomenon. Introduction Freshwater-turtle ecologists must have reliable methods for capturing their target species. A wide variety of traps and baits have been used to collect freshwater turtles, including hoop nets (Legler 1960), deep-water crawfish nets (Glorioso and Niemiller 2006), funnel traps (Iverson 1979), and basking traps (Ream and Ream 1966). Most of these traps are set in shallow water over a prolonged period with bait placed in containers that prevent actual consumption (Mali et al. 2014a). Several factors can influence trap success, including species targeted, placement within habitat, trapping duration, bait age, and bait preference (Brown et al. 2011, Mali et al. 2014a, Voorhees et al. 1991). Bait choice has been shown to greatly affect capture rates, and numerous baitpreference studies have been conducted over the past 50 years (reviewed in Mali et al. 2014b). One of the earliest studies compared 6 different baits used to capture a turtle assemblage in Pennsylvania, and those researchers found that different turtle species preferred different baits (Ernst 1965). Specifically, canned sardines in oil 1SWCA Environmental Consultants, 200 Bursca Drive, Suite 206, Bridgeville, PA 15017. 2Turtle Survival Alliance–North American Freshwater Turtle Research Group, PO Box 12074, Fort Worth, TX 76109. 3Walde Research and Environmental Consulting, 8000 San Gregorio Road, Atascadero, CA 93422. 4Freed–Hardeman University, 158 East Main Street, Henderson, TN 38340. 5University of Central Florida, Department of Biology, 4110 Libra Drive, Orlando, FL 32826. *Corresponding author - emunscher@swca.com. Manuscript Editor: Will Selman Southeastern Naturalist 253 E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 were preferred by Chrysemys picta Schneider (Painted Turtle) and to a lesser degree by Sternotherus odoratus Latreille (Eastern Musk Turtle), whereas Chelydra serpentina L. (Eastern Snapping Turtle) were attracted to fresh fish and chicken parts (Ernst 1965). Canned sardines became the standard for turtle trapping and are still used today by many researchers (e.g., Failey et al. 2007, Gamble 2006, Voorhees et al. 1991). Another study conducted in Kansas tested the effectiveness of 17 different baits in attracting semi-aquatic turtles (Voorhees et al. 1991). The authors captured 9 turtle species and identified fresh mussels, canned creamed corn, canned sardines, fresh shrimp, and fresh scallops as the most effective baits (Voorhees et al. 1991). Macrochelys temminckii Troost (Alligator Snapping Turtle) preferred traps baited with fresh-cut fish (Jensen 1998). Mali et al. (2014a) found that dried cat and dog food worked well for Red-eared Sliders and to a lesser degree for Eastern Musk Turtles. Little information is available on seasonal aspects of bait choice; however, Thomas et al. (2008) inferred that bait effectiveness might vary seasonally, sexually, and ontogenetically, and suggested that these variables were worth exploring. It is clear that different species can have different bait preferences. Therefore, it is important for turtle ecologists to determine effective capture methods, including bait effectiveness, for their target species. While conducting population surveys for freshwater turtles at Comal Springs, New Braunfels, TX, the first author observed a young boy capture 24 Eastern Musk Turtles in 20 min on a hand-line baited with a buffalo chicken wing. Using this information, we designed a study to specifically ask: (1) Is there a difference in catch rate between traps baited with buffalo chicken and highly odoriferous and commonly used alternative baits? (2) Are there bait-preference differences between the sexes? (3) Is there a seasonal difference in capture rates for traps baited with buffalo chicken? and (4) Is there a difference in capture rate among traps baited with buf falo, fried, and raw chicken? Field-site Description Comal Springs arises from the Edwards Aquifer, and 13 springs feed Landa Lake, the headwaters of the Comal River in New Braunfels (Comal County, TX). At the southern end of the lake (where it drains into the Comal River), there is an elevated spillway that creates an obstacle to animal movement in the waterway. Following the spillway, the water-flow becomes the Comal River, which flows for 4.8 km before its confluence with the Guadalupe River (Fig 1). Water temperatures from Comal Springs are relatively constant and range from 23.1 °C to 23.9 °C (Brune 1981). The spring has an average flow of ~9000 L per second or ~726,799,000 L per day (Brune 1981). The system is highly sensitive to fluctuations in climate, with peak flows recorded following tropical storms, and no flow during drought periods (Herbert and Frank 1974, Kelton 1973). The surface area of Landa Lake is ~10.1 ha (8.4 ha excluding islands). The deepest parts of the lake are ~4.5 m deep. The lake is heavily vegetated with Zostera marnia L. (Eelgrass) and Ludwigia repens J.R. Forst. (Red Ludwigia). The lake is surrounded by a largely developed shoreline that includes private lawns, a city golf course, and a highly manicured city park. Southeastern Naturalist E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 254 Landa Park receives heavy recreational use by an estimated 1 million people every year (Wade Tomlinson, Parks and Recreation, New Braunfels, TX; pers. comm.). The park accommodates multiple recreational activities including fishing, boating, golfing, picnics, and train rides. Methods Trap placement and bait We conducted this bait comparison study in 2 phases. Phase 1 compared 5 highly odoriferous and commonly used baits against buffalo chicken. We conducted 2–3- day trapping sessions in February/March, April, July, and November 2014, and another 3-day session in February 2015. We studied bait preference by Eastern Musk Turtles using 15 collapsible crab traps (81.3 cm × 50.8 cm × 30.5 cm; Promar, Gardena, CA). In this first phase of the study, we set traps in groups of 3, with 2 baited traps—1 with buffalo chicken and the other with a comparison bait. An unbaited control trap was also placed amongst the baited traps. We placed traps in the same formation and location during every sampling session with buffalo chicken on the left, control in the middle, and comparison bait on the right. Traps were placed ~2 m apart from one another. We chose 2 sites within the lake (Heidelberg and Canoe Island) because of the known presence of Eastern Musk Turtles (E. Munscher, Figure 1. Study site map showing Landa Lake (Comal Springs), the Comal River and the Guadalupe River. Southeastern Naturalist 255 E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 pers. observ.). We deployed 3 sets of traps (9 total) at the Heidelberg Lodge and 2 sets of traps (6 total) at Canoe Island. All traps were fully submerged in ~1.2 m of water. We tested the buffalo chicken wing bait (H-E-B brand; H-E-B, San Antonio, TX) during each sampling session. For each sampling session, we selected a comparison bait that was malodorous in nature and or had previously been used with success in other studies (Ernst 1965, Mali et al. 2014a, Voorhees et al. 1991). The 1st sampling session (28 February–2 March 2014) compared buffalo chicken vs. stink bait (Magic Bait Ole Whiskers 10 oz. Beef Blood Catfish Bait; Academy Sports, Katy, TX), the 2nd sampling session (18–20 April 2014) compared buffalo chicken vs. wet canned cat food (Iams Purrfect Delights Winner; Winner with Chicken Dinner Flaked in Sauce, Mason, OH; Mali et al. 2014a), the 3rd sampling session (18–20 July 2014) compared buffalo chicken vs. potted meat (Armour, Miami, FL), the 4th sampling session (7–8 November 2014) compared buffalo chicken vs. Hill Country Fare 4-Flavor Morsels dry cat food (H-E-B brand; Mali et al. 2014a), and the 5th sampling session (13–15 February 2015) compared buffalo chicken vs. canned sardines in oil (Bumble Bee, San Diego, CA; Ernst 1965, Failey et al. 2007, Gamble 2006, Voorhees et al. 1991). Each trap was set for 3 h around dusk and checked twice during the sample period. After the late check, we pulled the traps and baited and reset them the following evening. Fifteen traps were active for a total of 630 trap hours during the 5 sampling sessions. We conducted phase 2 during the periods 25–26 September 2015 and 19–20 February 2016 to determine if there was a difference among chicken types. Specifically, we searched for differences among raw, fried, and buffalo chicken. All chicken types were purchased from the H-E-B grocery store in New Braunfels, TX. We employed the same methodology as described for phase 1 in this second phase of the study. A total of 15 traps were set during each sampling session as described above, but no unbaited control traps were used. Buffalo chicken was set to the left, raw chicken in the middle, and fried chicken on the right. Each trap was set for 3 h around dusk and checked twice. After the late check, we pulled the traps and baited and reset them the following evening. Traps were set for a total of 90 trap hours. Data analysis All turtles captured were weighed and measured, sexed, and marked with both an external hard mark and passive integrated transponder (PIT) tag (Biomark, Inc., Boise, Idaho, USA). We experienced a low recapture rate of 24 of 348 (6.9%) total captures across all sampling trips. Turtles recaptured during the same sampling trips were not counted in our analysis. We used chi-square tests to make all comparisons. We specifically compared baited traps and non-baited traps, buffalo chicken and each other type of bait, buffalo chicken and each other type of bait by sex, buffalo chicken and all phase 1 baits pooled by sex (male, female, juvenile), capture rates among and between trapping sessions, and buffalo chicken among and between fried chicken and raw chicken. When we made multiple comparisons, we adjusted the alpha to control for family-wise error rates using the Holm method (Norman and Streiner 2000). We conducted our analyses in JMP 10.0 (SAS Institute, Inc., Cary, NC) with an alpha set at 0.05. Southeastern Naturalist E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 256 Results During phase 1, we captured 276 Eastern Musk Turtles, of which 231 (84%) were caught with buffalo chicken bait (Table 1). Significantly more captures were made in baited traps than in unbaited controls, and buffalo chicken was more effective at attracting Eastern Musk Turtles than any other tested bait (Table 2). No significant differences were found between the sexes with respect to bait type (Table 3). A chi-square test indicated that capture rates among males, females, and Table 2. Summary of chi-square tests of bait preferences of Eastern Musk Turtles (Sternotherus odoratus) during phase 1 trials at Comal Springs, TX. df χ2 P Bait vs. no bait 1 209.02 less than 0.0001 Buffalo chicken vs. stink bait 1 23.27 less than 0.0001 Buffalo chicken vs. cat food 1 32.40 less than 0.0001 Buffalo chicken vs. potted meat 1 23.56 less than 0.0001 Buffalo chicken vs. dry cat food 1 17.69 less than 0.0001 Buffalo chicken vs. sardines 1 43.38 less than 0.0001 Table 1. Capture totals of Eastern Musk Turtles (Sternotherus odoratus) at Comal Springs, TX, comparing buffalo chicken and other malodorous baits, and other chicken types (BC = buffalo chicken; SB = stink bait; CC = canned cat food; PM = potted meat; DC = dry cat food; SA = sardines; FC = fried chicken; RC = raw chicken; and NB = No bait). Bait Sampling period Sex BC SB CC PM DC SA FC RC NB February/March M 30 5 - - - - - - 1 F 8 1 - - - - - - 0 JV 0 0 - - - - - - 0 April M 22 - 0 - - - - - 0 F 16 - 2 - - - - - 0 JV 0 - 0 - - - - - 0 July M 18 - - 1 - - - - 0 F 11 - - 1 - - - - 0 JV 0 - - 0 - - - - 0 November M 38 - - - 17 - - - 0 F 30 - - - 10 - - - 0 JV 2 - - - 0 - - - 0 February M 37 - - - - 7 - - 0 F 13 - - - - 0 - - 0 JV 6 - - - - 0 - - 0 September 2015 M 15 - - - - - 8 3 - F 12 - - - - - 5 2 - JV 0 - - - - - 0 0 - February 2016 M 11 - - - - - 3 1 - F 8 - - - - - 0 4 - JV 0 - - - - - 0 0 - Totals 277 6 2 2 27 7 16 10 1 Southeastern Naturalist 257 E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 juveniles did not differ between buffalo chicken and all phase 1 baits pooled (χ2 = 3.00, df = 2, P = 0.2232). A chi-square test revealed that capture rates differed significantly among trapping sessions (χ2 = 23.01, df = 4, P < 0.0001). Significant differences were found between November (n = 97) and February/March (n = 45) (χ2 = 19.49, df = 1, P < 0.0001), November (n = 97) and April (n = 40) (χ 2 = 24.45, df = 1, P < 0.0001), November (n = 97) and July (n = 31) (χ 2 = 35.73, df = 1, P < 0.0001), November (n = 97) and February 2015 (n = 63) (χ 2 = 7.28, df = 1, P = 0.007), and February 2015 (n = 63) and July (n = 31) (χ 2 = 11.11, df = 1, P = 0.0009). During phase 2, we captured 72 Eastern Musk Turtle captures, of which 63.9% were caught with buffalo chicken. A total of 45 captures were made during September 2015, and 27 of these were captured using buffalo chicken, 13 using fried chicken, and 5 using raw chicken. A total of 27 captures were made during February 2016 with 19 captures using buffalo chicken, 3 using fried chicken, and 5 using raw chicken. We pooled data from both September 2015 and February 2016 trapping sessions to increase our sample size. We found a significant difference in the number of captures among traps baited with raw chicken (n = 10), fried chicken (n = 16), and buffalo chicken (n = 46) (χ2 = 29.37, df = 2, P < 0.0001). Significantly more turtles were captured with buffalo chicken compared to fried chicken (χ2 = 15.14, df = 1, P = 0.0001). Likewise, significantly more turtle captures were made with buffalo chicken than raw chicken (χ2 = 25.08, df = 1, P < 0.0001); no difference was found between fried chicken and raw chicken (χ 2 = 1.40, df = 1, P = 0.237). Discussion It is not surprising that Eastern Musk Turtles were attracted to all baits used due to the species’ broad dietary preferences (Ford and Moll 2004, Iverson and Meshaka 2006), the previous successful use of these baits in other studies (Ernst 1965, Mali et al. 2014a, Voorhees et al. 1991), and their malodourous nature. However, the apparent preference of Eastern Musk Turtles for selecting buffalo chicken warrants explanation. One potential explanation is that these highly carnivorous turtles are attracted to the grease emitted by the cooked chicken meat. Regular fried chicken was the second most-effective bait in phase 2, and seems to support this assumption. We have noticed that as soon as these traps were deployed, the buffalo chicken created a grease sheen on the water’s surface. Eastern Musk Turtles are primarily a benthic species that has a highly developed olfactory system (Ernst 1986) and they may be able to smell the odors within the water column. In contrast, Jensen (1998) Table 3. Summary of chi-square tests of bait preferences between sexes of Eastern Musk Turtles (Sternotherus odoratus) during phase 1 trials at Comal Springs, TX. df χ2 P Sex*bait (buffalo chicken vs. stink bait) 1 0.06 0.8045 Sex*bait (buffalo chicken vs. cat food) 1 2.57 0.1087 Sex*bait (buffalo chicken vs. potted meat) 1 0.12 0.7347 Sex*bait (buffalo chicken vs. dry cat food) 1 0.40 0.5284 Sex*bait (buffalo chicken vs. sardines) 1 3.91 0.0481 Southeastern Naturalist E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 258 hypothesized that predominately benthic turtle species may have difficulty locating baits that generate a grease trail at the water’s surface; the lack of a continuous odor trail at lower levels in the water column might hinder the ability of these species to locate the bait. An alternate hypothesis is that Eastern Musk Turtles are attracted to some other component of buffalo chicken, perhaps the sauce. We have successfully used raw and fried chicken in the past at other sites and on other species; however, we have had much greater success with buffalo chicken. During phase 2, we demonstrated that buffalo chicken is much more successful than plain fried chicken. It could be argued that a novel bait may offer a unique smell and taste (e.g., buffalo sauce) and become highly attractive. However, we compared buffalo chicken directly with other novel baits, and buffalo chicken was always the preferred bait. Therefore, we reject the notion that turtles demonstrate a preference for a specific bait because it is novel. Clearly, there is some component of buffalo chicken that is consistently attractive to Eastern Musk Turtles because we are still capturing high numbers of turtles in this system after multiple years of sampling. The most successful trapping session in our study was in November 2014. Whereas trapping success was similar during the previous 3 samples, we made significantly more captures in November. An earlier study from a nearby system in Texas had peak captures in July and January and few captures in November (Swannack and Rose 2003). It is unclear why these results would be so different, and we are not aware of any biologically meaningful explanation for this result. Clearly, further studies on seasonal bait preferences and seasonal-activity patterns of spring-dwelling Eastern Musk Turtles are needed. We have demonstrated herein that buffalo chicken is a very effective bait for capturing Eastern Musk Turtles when compared to several traditionally used baits and novel baits, including other chicken types. We made more than 80% of Eastern Musk Turtle captures with buffalo chicken during phase 1 bait trials. When we directly compared buffalo chicken with other chicken types in phase 2, buffalo chicken was the bait used to capture more than 63% of all turtles sampled. We suspect that our findings have implications for researchers studying Eastern Musk Turtles in other systems. Future studies should evaluate this bait throughout the Eastern Musk Turtle’s range as well as on other species such as Sternotherus minor Agassiz (Loggerhead Musk Turtle), Sternotherus carinatus Gray (Razor-backed Musk Turtle), and the critically endangered and federally listed Sternotherus depressus Tinkle and Webb (Flattened Musk Turtle). Freshwater-turtle ecologists may want to try this novel bait in investigations of other kinosternids and members of other turtle families as it may increase capture rates. Acknowledgments We thank the following people for supporting this research: past and current staff of the City of New Braunfels, Texas Parks and Wildlife Department, Pennsylvania State University, Freed-Hardman University, University of North Florida, and Peninsula College. We also thank volunteers M. Farris, S. Ross, C. Collins, J. Gray, J. Munscher, B. Bower, Southeastern Naturalist 259 E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 B. Hauge, R. Losey, and D. Rogers, and the many other students and biologists who make up the Turtle Survival Alliance-North American Freshwater Turtle Research Group (NAFTRG). We would like to thank Allan Brown for providing the detailed site map. We are appreciative of the San Antonio Zoo and Busch Gardens Sea World for their generous monetary and personnel support for this project. We would also like to thank the Heidelberg Lodge for their constant hospitality over the years of this project. Additional thanks go to SWCA Environmental Consultants for their constant support. Capture and handling protocols were approved by the Texas Parks and Wildlife Department, the City of New Braunfels, and conform to the ASIH/SSAR animal use guidelines (ASIH/SSAR, 2001). The study was conducted under permit # SPR-0212-019 from the Texas Parks and Wildlife Department. Literature Cited Brown, D.J., I. Mali, and M.R.J. Forstner. 2011. No difference in short-term temporal distribution of trapping effort on hoop-net capture efficiency for freshwater turtles. Southeastern Naturalist 10:245–250. Brune, G. 1981. Springs of Texas. Vol. 1. Branch-Smith, Inc., Fort Worth, TX. 566 pp. Ernst, C.H. 1965. Bait preferences of some freshwater turtles. Journal of the Ohio Herpetological Society 5:53. Ernst, C.H. 1986. Ecology of the turtle, Sternotherus odoratus, in southeastern Pennsylvania. Journal of Herpetology 20:341–352. Failey, E.L., J.C. McCoy, S.J. Price, and M.E. Dorcas. 2007. Ecology of turtles inhabiting golf-course and farm ponds in the western piedmont of North Carolina. Journal of North Carolina Academy of Science 123:221–232. Ford, K.D., and D. Moll. 2004. Sexual and seasonal variation in foraging patterns in the Stinkpot, Sternotherus odoratus, in southwestern Missouri. Journal of Herpetology 38:296–301. Gamble, T. 2006. The relative efficiency of basking and hoop traps for Painted Turtles (Chrysemys picta). Herpetological Review 37:308–312. Glorioso, B.M., and M.L. Niemiller. 2006. Using deep-water crawfish nets to capture aquatic turtles. Herpetological Review 37:185–187. Herbert, P.J., and N.L. Frank. 1974. Atlantic hurricane season of 1973. Monthly Weather Review 102:280–282. Iverson, J.B. 1979. Another inexpensive turtle trap. Herpetological Review 10:55. Iverson, J.B., and W.E. Meshaka Jr. 2006. Sternotherus odoratus: Common Musk Turtle or Stinkpot. Pp. 207–223, In P.A. Meylan (Ed.). Biology and Conservation of Florida Turtles. Chelonian Research Monographs 3. Chelonian Research Foundation, Lunenberg, MA. 376 pp. Jensen, J.B. 1998. Bait preferences of southeastern United States coastal-plain riverine turtles: Fish or fowl? Chelonian Conservation and Biology 3:109 –111. Kelton, E. 1973. The Time it Never Rained. Tom Doherty Associates, LLC, New York, NY. 416 pp. Legler, J.M. 1960. Natural history of the Ornate Box Turtle, Terrapene ornata ornata Agassiz. University of Kansas Publications, Museum of Natural Histo ry 11:527–669. Mali, I., D. Haynes, and M.R.J. Forstner. 2014a. Effects of bait types, bait age, and trap hours on capture success of freshwater turtles. Southeastern Naturalist 13:619–625. Mali, I., D.J. Brown, J.R. Ferrato, and M.R.J. Forstner. 2014b. Sampling freshwaterturtle populations using hoop nets: Testing potential biases. Wildlife Society Bulletin 38:580–585. Southeastern Naturalist E.C. Munscher, A.D. Walde, A.M. Walde, B.P. Butterfield, and N. Salvatico 2017 Vol. 16, No. 2 260 Norman, G.R., and D.L. Streiner. 2000. Biostatistics: The Bare Essentials, 2nd Edition. B.C. Decker Inc., Hamilton, ON, Canada. 324 pp. Ream, C., and R. Ream. 1966. The influence of sampling methods on the estimation of population structure in Painted Turtles. American Midland Naturalist 75:325–338. Swannack, T.M., and F.L. Rose. 2003. Seasonal and ontogenetic changes in the sex ratio of a population of Stinkpots (Kinosternidae: Sternotherus odoratus). The Southwestern Naturalist 48:543–549. Thomas, R.B., I.M. Nall, and W.J. House. 2008. Relative efficacy of three different baits for trapping pond-dwelling turtles in east-central Kansas. Herpetological Review 39:186–188. Voorhees, W., J. Schnell, and D. Edds. 1991. Bait preferences of semi-aquatic turtles in southeast Kansas. Kansas Herpetological Society Newsletter 85:13–15.