Southeastern Naturalist 117 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 Sexual Dimorphism, Movement Patterns, and Diets of Sternotherus carinatus (Razorback Musk Turtle) Brendan T. Kavanagh1 and Matthew A. Kwiatkowski2,* Abstract - Sternotherus carinatus (Razorback Musk Turtle) is an aquatic species for which natural history information is lacking. We explored natural history differences between males and females using analyses of sexual dimorphism, movement patterns, and diets in 2 East Texas populations. We found male-biased sexual dimorphism in both head and body size, a pattern partially consistent with analyses of preserved museum specimens. Using radiotelemetry, we found no differences between males and females in home ranges, maximum distance moved, and total distance moved. However, females exhibited significantly higher mean daily speeds. Analysis of the proportional use of food resources indicated a high degree of overlap between males and females and a low degree of overlap between populations. The selective forces influencing the observed patterns of sexual dimorphism in this species remain unclear but our data seem to rule out partitioning of food resources. Introduction Selection pressures often vary between males and females, resulting in differences in natural history traits, including mating behavior, foraging, habitat use, and coloration (Cox et. al 2007, Darwin 1871, Lindeman 2003). Males and females can also differ in morphology, and many studies on sexual size-dimorphism (SSD) have focused on reptiles because sexual dimorphism is common and life histories are variable within this clade (reviewed in Cox et al. 2007). Chelonians are no exception, and exhibit a broad range in SSD from female-biased SSD in Emydidae (Gibbons and Lovich 1990, Lindeman 2000) to male-biased SSD in Testudinidae (Auffenberg and Weaver 1969, Moskovits 1988). Explanations for SSD in reptiles have focused on 3 selective pressures (Berry and Shine 1980; reviewed in Cox et al. 2007): (1) selection which favors large females due to associated increases in fecundity, (2) sexual selection for large males due to male–male combat and differential mating success, and (3) natural selection causing sexual dimorphism that reduces competition between the sexes. Across chelonid taxa, female-biased SSD is more frequently observed than male-biased SSD (Cox et al. 2007, reviewed in Gibbons and Lovich 1990). Female- biased SSD in turtles appears to result from fecundity benefits, where larger females produce larger clutches or larger eggs (Congdon and Gibbons 1985, 1987; Congdon and Tinkle 1982; Congdon et al. 1983, 1987; Gibbons et al. 1982; St. Clair 1998). Eggs with greater mass result in larger hatchlings, which may have higher survival rates (Brooks et al. 1991, Janzen et al. 2000, although see Congdon 1Center for Watershed Sciences, University of California, Davis, CA 95616. 2Department of Biology, Stephen F. Austin State University, PO Box 13003, Nacogdoches, TX 75962. *Corresponding author - kwiatkowm@sfasu.edu. Manuscript Editor: Jerry Cook Proceedings of the 6th Big Thicket Science Conference: Watersheds and Waterflow 2016 Southeastern Naturalist 15(Special Issue 9):117–133 Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 118 Vol. 15, Special Issue 9 et al. 1999, Kolbe and Janzen 2001). Male-biased SSD has also been documented in some turtle species (reviewed in Gibbons and Lovich 1990) and may result when larger males are superior in agonistic encounters during mating events or territorial disputes (Auffenberg 1977, Berry and Shine 1980, Cagle 1950, Lardie 1983, Lovich et al. 1998). Aggressive interactions among multiple males courting a female have been documented for a number of turtle species in natural populations (Barzilay 1980, Burge 1977, Hammer 1969, Jackson 1969, Mann et al. 2006, Rovero et al. 1999). Although male-biased SSD resulting from male territorial interactions has been documented in many reptiles (e.g., Cox et al. 2003, Shine 1994, Stamps 1983), true territoriality (as defined by Noble 1939 or Pitelka 1959), has not been conclusively confirmed in chelonians. Copulation can be an aggressive activity that resembles combat; hence, although some researchers consider the phenomenon improbable (Gibbons and Lovich 1990), greater male size may improve the ability of male turtles to forcibly inseminate non-receptive females (Berry and Shine 1980, Tanaka and Sato 1983),. Head size may also play an important role during aggressive interactions. Because bite-force performance increases with increasing head size (Herrel et al. 2005), male turtles with larger heads may be more successful at warding off competitors during mating events or territorial disputes. In addition, male turtles often bite the head and neck of females during courtship and a larger, stronger head may enable a male to better control the female during copulation. Larger males may also have an advantage if males make terrestrial or aquatic movements in search of receptive females. Males of several turtle species are known to travel greater distances than females during the mating season (Gibbons 1986, Morreale et al. 1984), and there is evidence that the longest movements are made by the largest males (Gibbons and Lovich 1990). Larger male turtles may also be more capable of mating with the largest females due simply to physical proportions (Gibbons and Lovich 1990). Males will presumably mate with any receptive female, but a male turtle, if presented with a choice, should mate with the larger female because of the benefits gained by mating with a female that has a higher probability of laying a larger clutch (Gibbons et al. 1982). Competition between sexes for limited resources has been attributed as the cause of SSD in some reptiles (Cox et al. 2007). Resource partitioning and SSD are especially prominent in gape-limited organisms such as snakes, where the larger sex is able to feed on larger or more-diverse prey (Shine 1989). Resource partitioning and SSD have also been documented in some turtle species in which the larger sex generally feeds on larger prey or a greater diversity of prey (Chen and Lue 1999, Lindeman 2003, Tucker et al. 1995, Wilhelm and Plummer 2012). Problems with this hypothesis arise in determining the causal relationship between SSD and dietary partitioning because few studies have actually shown that intersexual dietary partitioning actually reduces competition (Cox et al. 2007). Although there have been many studies on SSD in turtles, most have focused on the families Emydidae (e.g., Berry and Shine 1980; Forsman and Shine 1995; Gibbons and Lovich 1990; Lindeman 2000, 2003; Rovero et al. 1999; Tucker et al. Southeastern Naturalist 119 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 1995) or Testudinidae (Auffenberg and Weaver 1969, Lagarde et al. 2001, McRae et al. 1981, Moskovits 1988, Willemsen and Hailey 2003). Fewer studies have focused on SSD within the family Kinosternidae (Iverson 2002). This family includes Sternotherus carinatus Gray (Razorback Musk Turtle), which is a little-studied aquatic species distributed throughout the western portion of the Gulf Coastal Plain (Ernst et al. 1994, Mahmoud 1969, Tinkle 1958). This turtle is a small, bottom-dwelling species found in various habitats including ponds, lakes, rivers, slow-moving streams, and other permanent bodies of water (Mahmoud 1969, Tinkle 1958). Evidence suggests that the Razorback Musk Turtle exhibits male-biased sexual dimorphism in body traits (Atkinson 2013, Mahmoud 1967), although head dimorphism has yet to be quantified. Understanding the natural history of an organism is essential to development of sound conservation practices in increasingly human-dominated landscapes. Our goal was to provide insight into the factors affecting sexual differences by quantifying natural history characteristics of Razorback Musk Turtle, including sexual dimorphism of the head and body, and sexual differences in movement patterns and diet. Field-site Description We selected 2 sites for this study. Bernaldo Creek is a 2nd-order stream, which flows through the Stephen F. Austin Experimental Forest (SFAEF), southern Nacogdoches County (31°29'N, 94°47'W) in East Texas. The creek is relatively undisturbed and is not directly used by any municipality or industry (McCulloch 1981). The SFAEF is comprised of 728 ha of bottomland hardwood forest and 310 ha of upland pinewoods with ~670 ha of the bottomland hardwood forest within the Angelina River floodplain. Intense timber harvesting has not occurred since the 1920s, and most canopy trees are greater than 70 y old. Most of the entire extent of Bernaldo Creek has an abundance of woody debris in the form of individual logs, log-jams, roots, downed branches, and a beaver dam within the confines of the channel. Additionally, the portion found within the SFAEF is within the Angelina River floodplain and is covered by a dense canopy of hardwoods whose leaves form a thick detritus layer on the stream bottom. Due to an abundance of leaf litter, Bernaldo Creek supports a diverse and abundant benthic macroinvertebrate community (McCulloch 1981). La Nana Creek is a 3rd–4th-order stream that passes through the town of Nacogdoches, TX. La Nana Creek is impacted by many anthropogenic disturbances that affect stream-water quality including municipal-sewage effluent, runoff from foodpacking houses, industrial-wastewater effluent, plywood-mill and asphalt-plant runoff, urban runoff, and runoff from agricultural lands (Ahle 1991). A 1500-m portion of the study area in La Nana Creek has been channelized, resulting in very few bends and meanders, somewhat unstable banks, and a lack of riparian vegetation. During flood events, large volumes of water pass through the stream very quickly and with great depth, often flushing any woody debris downstream and scouring the banks, which are mowed regularly along Stephen F. Austin State University’s property. A secondary effect of channelization is the homogenization of aquatic habitat, Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 120 Vol. 15, Special Issue 9 and the channelized portion of La Nana Creek is essentially 1500 m of relatively deep water with a gravel/sand composite as the dominant substrate. Methods Trapping We used aquatic hoop-traps and crab-pots baited with chicken or fish scraps to trap turtles in Bernaldo Creek from March until August of 2007 and 2008 and in La Nana Creek from May until June 2008. We also caught Razorback Musk Turtles opportunistically by hand and with dipnets while checking traps, radiotracking, and performing visual searches. We used 16 hoop-traps adn 2 crab-pots in Bernaldo Creek and 6 hoop-traps in La Nana Creek. We placed traps 50 m apart and periodically moved them in increments of 50 m up- and downstream to sample a greater portion of the sites. We checked the traps each day or every other day, depending on catch rate and weather conditions. Razorback Musk Turtle morphometrics We brought captured Razorback Musk Turtles into the lab for radiotransmitter attachment and morphometric measurements. We measured curved carapace length (CCL), plastron length (PL), carapace width (CW), and plastron width (PW) to the nearest 1.0 mm using a flexible tape measure, and determined mass to the nearest 0.5 g with a balance scale. We anesthetized each turtle by placing it in a lidded gallon jar with a cotton ball soaked in Isoflurane (Halocarbon Products Corporation, River Edge, NJ). We used dial calipers to measure to the nearest 0.05 mm maximum head width (HW), maximum head depth (HD), and maximum head length (HL) of anesthetized turtles. We kept turtles in shallow water, monitored them until they fully recovered from the anesthesia, and transferred them to 37.85-L (10-gallon) aquariums for 24 h until release at the capture location. We included 39 preserved Razorback Musk Turtle museum specimens (Texas Cooperative Wildlife Collection, Texas A&M University, College Station, TX) in our study. We sexed and measured to the nearest 0.05 mm using dial calipers CCL, HW, HD, and HL (carapace width was not measured in museum specimens). We analyzed data from museum and field-caught specimens separately. We calculated sexual dimorphism indices following the methods of Lovich and Gibbons (1992), as modified by Forsman and Shine (1995), using CCL, HW, HD, and HL. The sexual dimorphism index is calculated by dividing body measurements of the larger sex by measurements of the smaller sex and assigning a positive value if females are larger, a negative value if males are larger, and adding one if males are larger or subtracting one if females are larger, thus giving a value symmetrical around zero. Normal quantile plots indicated that all data distributions were approximately normal; therefore, we did not transform our data. We employed Pearson’s correlation coefficients to assess the correlation of HL, HD, and HW with increasing CCL of turtles. We examined factors potentially influencing differences in head size between the sexes using 7 models created from all possible combinations of independent Southeastern Naturalist 121 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 variables for which there was existing data. For each model, the dependent variable, which we refer to as the head matrix, was a 3-column matrix composed of the HW, HD, and HL. Independent variables used in the models included CCL, sex, population, sex × CCL, and CCL × population. We used Akaike’s information criterion (AIC) to rank models. We examined relationships of morphometric variables using Pearson correlation matrices, and tested factors influencing sexual dimorphism in head size with multivariate analysis of variance (MANOVA). Statistical tests were carried out using R v.2.7.1 (www.r-project.org). In order to further verify that differences in body size between males and females were not influencing head-size comparisons, we divided HW, HL, and HD by each turtle’s CCL, thus providing head size-variables controlled for by body size. We employed Mann-Whitney tests to compare each body-size-controlled head measurement between the sexes. All statistical tests were conducted with an α = 0.05. Radiotelemetry We attached 10- and 8-g radio transmitters (Model RI-2B, Holohil Systems Ltd., Carp, ON, Canada) or custom transmitters (Phillip Blackburn, Nacogdoches, TX) to the lateral-posterior portion of Razorback Musk Turtle carapaces using quick-drying, waterproof epoxy (Devcon 60-sec epoxy). All transmitters weighed less than 5% of the turtle’s body mass. We used a portable telemetry receiver (Model R-1000, Communication Specialists, Inc., Orange, CA) and hand-held directional antennae to locate turtles within 1 m using homing methods. We typically radiotracked turtles 2–3 times per week during the spring, summer, and early fall (April–September) and once per week during the winter months (October–March). Upon locating each turtle, we recorded universal transverse mercator (UTM; WGS 84) coordinates as close to the turtle as possible (≤1 m) using a handheld GPS (Model eTrex Legend C, Garmin International Inc., Olathe, KS). We digitized each location into a GIS for placement on a digital aerial photograph. We calculated linear aquatic home-range (LAHR), mean daily speed (MDS), maximum distance moved (MDIS), and total distance moved (TDIS) using the animal movement extension (Hooge and Eichenlaub 2000) in ArcView v3.2 (Environmental Systems Research Institute, Redlands, CA). We carried out all statistical analyses in SYSTAT (version 12.0, 2007, Systat Software Inc, San Jose, CA). We used non-parametric Mann-Whitney U-tests (Zar 1999) for univariate analyses comparing male and female movement variables. To assess whether variables associated with spatial ecology could identify subgroups (sex and population), we averaged LAHR, MDS, MDIS, and TDIS values for each of the 23 turtles monitored throughout the course of the study and performed a PCA with varimax rotation (Manly 1986). All variables were standardized (normal deviates) before the PCA. Subsequent to the PCA, we used the same variables and undertook a discriminant function analysis to identify any individuals misclassified by sex or site. Diet analyses We filtered the water from the aquariums in which turtles were held following anesthetization to collect and preserve in 10% formalin turtles’ fecal deposits. Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 122 Vol. 15, Special Issue 9 We later sorted fecal samples under a dissecting microscope and calculated the total volume of each sample and the volume of each food category for each sample to the nearest 0.1 ml by displacement in water. We assessed dietary overlap between the sexes using the simplified Morisita index of niche overlap (Horn 1966, Krebs 1989): n CH = 2Σpijpik n i = 1 Σp ij 2 + Σpik 2 i = 1 where n is the number of diet categories, pij and pik are the proportions of the diet composed of diet category i for turtle species j and k. The index gives a value of dietary overlap ranging from 0 (no overlap) to 1 (complete overlap). Results Razorback Musk Turtle sexual size-dimorphism Trapping and opportunistic captures from both study sites yielded a total of 28 adult (16 male, 12 female) Razorback Musk Turtles for use in the sexual dimorphism analyses. Average male CCL, CW, HW, HL, and HD were greater than those of females (Table 1). Sexual dimorphism indices (SDI) for all morphometric measurements suggested male-biased sexual dimorphism (Table 1). Pearson’s correlation coefficients indicated that HW, HL, HD, and CCL were highly correlated (Table 2); HW, HL, and HD increased proportionately with increasing CCL. Head size increased with carapace length, but males consistently had larger heads relative to their carapace lengths (Fig. 1). Results from AIC-model rankings are summarized in Table 3. The top-ranked model only included CCL and sex (head matrix ~ ccl + sex, ΔAIC = 0.00); the 2ndranked model included those terms as well as the CCL × sex interaction (ΔAIC = 0.55). Results from MANOVA using the full model (full model: head matrix ~ ccl + sex + pop + pop × sex + ccl × sex; Table 4) indicated that CCL and sex were significant terms in the model for predicting head size (Sex: F = 12.838, P < 0.001; CCL: F = 46.775, P < 0.0001). Univariate comparisons of field-caught specimens provided further evidence that, when controlling for CCL, the HW, HD, and HL of Table 1. Summary of Razorback Musk Turtle morphometrics from live and museum specimens. Curved carapace length (CCL), carapace width (CW), head width (HW), head length (HL), head depth (HD) in mm (± 1 SE), and sexual dimorphism indices (SDI) are reported. Sex n CCL (± SE) CW (± SE) HW (± SE) HL (± SE) HD (± SE) Wild-caught Male 16 161.19 (3.61) 143.19 (2.98) 34.08 (0.95) 53.33 (1.46) 27.27 (0.77) Female 12 154.33 (2.09) 138.42 (2.92) 29.16 (0.57) 47.72 (1.21) 23.28 (0.59) SDI -0.044 -0.034 -0.168 -0.117 -0.170 Museum Male 23 149.96 (3.41) 32.15 (0.88) 48.60 (1.37) 29.01 (1.06) Female 16 135.18 (3.10) 26.60 (0.76) 40.19 (2.73) 24.07 (0.88) SDI -0.109 -0.209 -0.209 -0.205 Southeastern Naturalist 123 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 male turtles were significantly greater than those of females (HW/CCL: U = 4.0, P < 0.001; HL/CCL: U = 40.0, P < 0.001; HD/CCL: U = 19.0, P < 0.001). Our analysis of the 39 (23 male, 16 female) preserved museum specimens partially corroborated the results of our analyses of wild-caught Razorback Musk Turtles. Average male CCL, CW, HW, HL, and HD were larger than those of females (Table 1). Pearson correlation coefficients indicated that HW, HL, HD, and CCL were correlated, but the correlations between body measurements from museum specimens were less strong than the correlations between body measurements of wild-caught turtles (Table 2). The full model (head matrix ~ ccl + sex + pop + Table 2. Pearson correlation matrix from wild-caught (above the diagonal) and museum specimens (below the diagonal) of Razorback Musk Turtle. Pearson correlation coefficients are reported for head width, head depth, head length, and curved carapace length. Head width Head length Head depth Curved carapace length Head width - 0.926 0.939 0.851 Head length 0.689 - 0.910 0.839 Head depth 0.793 0.575 - 0.819 Curved carapace length 0.863 0.702 0.728 - Figure 1. Curved carapace length versus head size (head width, head length, and head depth) for male and female Razorback Musk Turtle. Open symbols represent males and closed symbols are females for head width (triangles), head length (squares), and head depth (circles). Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 124 Vol. 15, Special Issue 9 pop × sex + ccl × sex, df = 13, ΔAIC = 0.00) was the top ranked model (Table 3). Results from MANOVA using the full model indicated that CCL and sex are significant terms in the model predicting head size (Sex: F = 11.30, P < 0.01, CCL: F = 169.46, P < 0.001; Table 4). MANOVA on the full model also yielded significant values for the population (pop) and CCL × pop terms (Table 4) due to the greater number of populations (n = 16) from which the museum specimens were collected. Univariate comparisons of museum specimens partially corroborated the analyses on live-caught turtles, but there were some discrepancies. Male turtles had significantly wider heads than females, but HD and HL were not significantly different when controlling for body size (HW/CCL: U = 77.0, P = 0.002; HL/CCL: U = 147.0, P = 0.291; HD/CCL: U = 130.0, P = 0.123). Table 4. Summary of MANOVA for wild-caught and museum specimens of Razorback Musk Turtle. The dependent variable, head matrix, is a 3-column matrix composed of the head width, head depth, and head length measurements from each turtle. Wild-caught Museum df F P df F P ccl 1 46.775 less than 0.0001 1 169.46 less than 0.001 sex 1 12.838 less than 0.0010 1 11.300 less than 0.01 pop 1 0.475 0.7035 15 4.383 less than 0.001 pop×ccl 1 0.413 0.7455 7 2.498 less than 0.01 ccl×sex 1 1.809 0.1779 1 2.183 0.1475 Residuals 22 13 Table 3. AIC rankings from wild-caught and museum specimens of Razorback Musk Turtle. Models tested, degrees of freedom, and AIC values for each model are reported. The dependent variable, head matrix, is a 3-column matrix composed of the head width, head depth, and head length measurements from each turtle. Model df AIC value ΔAIC Wild-caught Head matrix ~ ccl + sex 25 184.76 0.00 Head matrix ~ ccl + sex + ccl×sex 24 185.31 0.55 Head matrix ~ ccl + sex + pop 24 189.25 4.49 Head matrix ~ ccl + sex + pop + pop×ccl + ccl×sex 22 193.03 8.27 Head matrix ~ ccl + sex + pop + pop×ccl 23 193.75 8.99 Head matrix ~ ccl 26 206.05 21.29 Head matrix ~ sex 26 226.44 41.68 Head matrix ~ intercept 27 238.10 53.34 Museum Head matrix ~ ccl + sex + pop + pop×ccl + ccl×sex 13 329.94 0.00 Head matrix ~ ccl + sex + pop + pop×ccl 14 342.16 12.22 Head matrix ~ ccl + sex + pop 21 410.31 80.37 Head matrix ~ ccl + sex 36 448.08 118.14 Head matrix ~ ccl + sex + ccl×sex 35 448.20 118.26 Head matrix ~ ccl 37 451.90 121.96 Head matrix ~ sex 37 492.26 162.32 Head matrix ~ intercept 38 503.39 173.45 Southeastern Naturalist 125 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 Home ranges and movement patterns of Razorback Musk Turtles We captured and radiotracked 24 turtles (5 Bernaldo Creek, 18 La Nana Creek; we excluded from our analyses 1 juvenile with less than 10 locations). Trapping and opportunistic captures yielded 3 males and 2 females from Bernaldo Creek and 9 males and 9 females from La Nana Creek. We located the 23 turtles a total of 725 times (min = 6, max = 80, mean = 31 per turtle). Male and female LAHR, TDIS, MDIS, and MDS did not differ between Bernaldo and La Nana Creeks (Table 5). We pooled spatial data between study sites and calculated home ranges from the total amount of data available for each turtle. Female Razorback Musk Turtles had significantly higher MDS than male turtles (U = 103.0, P = 0.023). LAHR, TDIS, and MDIS did not differ between the sexes (Table 6). The principle component analysis of the combined variables associated with home range (LAHR) and movement behavior (MDIS, TDIS, MDS) indicated that the first 2 components accounted for 74% of the variation within the overall sample (PC 1: 49%; PC 2: 25%; loadings provided upon request). However, there was no obvious separation by population or sex corroborating the univariate comparisons described above (Fig. 2). Based on the multivariate model, the discriminant function analysis was insignificant with respect to classification by population or sex Table 6. Summary of home-range and movement-pattern analyses calculated from total data sets for each turtle. Means (±1 SE) for male and female linear aquatic home range (LAHR, m), maximum distance moved (MDIS, m), total distance moved (TDIS, m), mean daily speed (MDS, m/d), Mann- Whitney U-statisic, and P-value are reported. Male (n = 12) Female (n = 11) U P LAHR (m) 277.166 (99.66) 362.73 (91.84) 88.0 0.176 MDIS (m) 236.699 (85.47) 293.9645 (70.22) 85.0 0.242 TDIS (m) 1148.513 (414.41) 1522.408 (415.50) 88.0 0.176 MDS (m/day) 3.89 (1.42) 15.67 (9.55) 103.0 0.023 Table 5. Summary of home-range and movement-pattern comparisons between Bernaldo and La Nana creeks. Means (±1 SE) for male and female linear aquatic home ranges (LAHR), maximum distances moved (MDIS), total distance moved (TDIS), mean daily speed (MDS), Mann-Whitney U-statisic, and P-value are reported. Bernaldo La Nana U/P Female LAHR (m) 334.00 (144.00) 369.12 (110.82) 11.0/0.727 MDIS (m) 234.23 (150.63) 307.24 (82.35) 9.0/1.00 TDIS (m) 875.71 (40.17) 1666.12 (499.52) 11.0/0.727 MDS (m/day) 3.48 (0.21) 18.38 (11.59) 15.5/0.145 Male LAHR (m) 109.67 (18.66) 333.00 (128.92) 15.0/0.864 MDIS (m) 82.96 (17.64) 287.95 (109.78) 17.0/0.600 TDIS (m) 466.23 (260.49) 1375.94 (532.29) 17.0/0.600 MDS (m/day) 3.39 (1.52) 4.06 (1.87) 16.0/0.727 Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 126 Vol. 15, Special Issue 9 (Population: Wilk’s λ = 0.923, F = 0.829, P = 0.451; Sex: Wilk’s λ = 0.955, F = 0.472, P = 0.630). Home ranges of the 3 male turtles from Bernaldo Creek did not overlap with other male home ranges, but 2 of the 3 male home ranges did overlap with the home range of 1 or more females and the 2 female home ranges also overlapped. Home ranges of 7 of the 9 males radio-tracked in La Nana Creek overlapped with at least 1 other male home range and all male turtles had home ranges that overlapped with at least 1 other female home range. All female home ranges overlapped with at least 1 other female home range. Razorback Musk Turtle food habits Volumes of fecal material by sex and population are reported in Table 7. Although the dominant food item differed between study sites, turtles from both La Nana Creek and Bernaldo Creeks fed on the same food items; thus, we pooled dietary data between sites to provide a broader picture of how diets differed between the sexes in the 2 populations. When we pooled the dietary data from both sites, we found that male and female turtles fed on a similar variety of organisms in similar proportions; food items included freshwater mussels, crustaceans (crayfish), arthropods, and plant material (Table 7). Analysis of the proportional use of food resources between males and females yielded a simplified Morisita index value of 0.764, indicating a high degree of food-resource use overlap. Crustacean (crayfish) material comprised 55% of the total volume of fecal material collected from Bernaldo Creek turtles (n = 4) and mollusk (freshwater mussel) material comprised 73% of the total volume of fecal material collected from La Nana Creek turtles (n = 19) (Table 7). Analysis of the proportional use of food resources between Bernaldo Creek and La Nana Creek turtles yielded a simplified Morisita index value of 0.163, indicating a low degree of foodresource use overlap. Figure 2. Principle component analysis (PCA) scores by (a) sex and (b) population for 23 Razorback Musk Turtles. Variables used in the PCA were linear aquatic home range, maximum distance moved, total distance moved, and mean daily speed. Southeastern Naturalist 127 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 Discussion We found that Razorback Musk Turtles exhibited male-biased sexual sizedimorphism, and that males have disproportionately larger heads than females relative to their curved carapace lengths. This conclusion was corroborated by the outcome of our analyses of both wild-caught and preserved turtle specimens. However, population and population x ccl were also significant factors in the analyses of museum specimens, though this result is likely due to the relatively high number of populations from which we collected data and the relatively low number of individuals from each population. Although sexual dimorphism may result from food-resource partitioning, male and female turtles had 76% overlap in their food-resource use. Hence, dietary data collected for this study do not support the resource-partitioning hypothesis as an explanation for the observed patterns of sexual dimorphism (our sample size was low for Bernaldo Creek, so conclusions should be considered carefully). Alternatively, male-biased sexual dimorphism often results from sexual selection. We did not observe combat between male turtles during the course of this study, and our spatial data did not conclusively confirm territoriality for this species. Mating-system spatial structure can be difficult to quantify in turtles (Pearse and Avise 2001), although dominance hierarchies have been documented for a few species (Galbraith 1991; Galbraith et al. 1987, 1993; Kauffman 1992; McRae et al. 1981). Generally, males of territorial species move more than females because they are faced with the challenge of defending a territory from their rival conspecific males (reviewed in Gehring and Swihart 2004). Male turtles of most aquatic species move more than their female conspecifics during the active season because Table 7. Summary of fecal-material analysis with mean volume and proportional volume. Mean volume (ml) ±1 SE of females (populations combined), males (populations combined), Bernaldo Creek turtles (males and females combined), and La Nana Creek (males and females combined). Total volume (ml) and proportional food-resource use are reported for male and female data combined from the 2 study sites and also for Bernaldo turtles and La Nana turtles with data for each sex combined. Males Females Bernaldo La Nana Mean volume Mollusk 1.12 (0.41) 4.51 (1.58) 0.05 (0.03) 3.31 (0.98) Crustacean 0.42 (0.27) 0.14 (0.09) 1.03 (0.73) 0.13 (0.08) Arthropod 0.21 (0.04) 0.11 (0.03) 0.13 (0.03) 0.17 (0.04) Plant 0.66 (0.14) 0.63 (0.28) 0.28 (0.18) 0.72 (0.17) Other 0.46 (0.16) 0.2 (0.10) 0.45 (0.29) 0.31 (0.11) Total volume 2.79 (0.38) 5.5 (1.56) 1.85 (0.60) 4.56 (0.94) Proportional volume Mollusk) 0.400 0.820 0.027 0.725 Crustacean 0.149 0.025 0.554 0.028 Arthropod 0.075 0.020 0.068 0.037 Plant 0.236 0.114 0.149 0.158 Other 0.164 0.036 0.243 0.068 Total volume 33.5 60.5 7.4 86.6 Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 128 Vol. 15, Special Issue 9 their reproductive strategy is to mate with as many females as possible, and the probability of encountering a female turtle increases as they move greater distances (Gibbons et al. 1990). Male turtles defending territories could forego extensive journeys throughout the aquatic landscape and would, thus, exhibit smaller home ranges or reduced movement patterns compared to females who would still be required to make relatively large movements during the nesting season. Male turtles establish territories to either defend females from other rival males or defend resources that attract females (Emlen and Oring 1977). La Nana Creek exhibits a clumped distribution of resources important to Razorback Musk Turtles (i.e., structure and substrate type; B.T. Kavanagh, unpubl. data), and clumped resource distributions favor the formation of female groups (Travis et al. 1995). Although male Razorback Musk Turtles in our study seemed to move less than females, notably during the active seasons, and males were particularly abundant in sections of La Nana Creek that offered structural refugia and sand/gravel substrate, this pattern does not provide conclusive evidence of territoriality. Seven of the 9 male home-ranges in La Nana Creek overlapped, suggesting that turtles may not be territorial, but we did not determine temporal overlap, and doing so would require further investigation. It is unknown whether patterns of home-range overlap in Bernaldo and La Nana creeks are the result of differences in habitat characteristics and resource distributions between the 2 sites, or small sample sizes. Male home-ranges in Bernaldo Creek did not overlap, but it is difficult to interpret this result because we radiotracked only 3 male turtles. It is generally accepted that mating-system structure can be influenced by ecological variables, especially population density (Davies 1991, Maher and Lott 2000, Travis et al. 1995). When population densities are extremely high or low, the cost of maintaining a territory usually outweighs the benefit gained (Emlen and Oring 1977), therefore males in low-density populations exhibit no site defense, males in moderately dense populations exhibit territoriality, and males in high density populations usually resort to dominance hierarchies, leks, or scramble competition (Maher and Lott 1995). The data from this study suggests that Razorback Musk Turtles have higher population densities in La Nana Creek compared to Bernaldo Creek. Both male and female turtles from Bernaldo Creek were smaller than those found in La Nana Creek. There are 3 possible scenarios that may explain the observed patterns. First, the observation that Razorback Musk Turtles from Bernaldo Creek were smaller than those from La Nana Creek could be a result of small sample sizes collected from Bernaldo Creek. Second, that Razorback Musk Turtles were smaller in Bernaldo Creek may be an indication of this creek’s suboptimal condition related to the ecological needs of this species. Though Bernaldo Creek has adequate structure and overstory canopy cover, the mud substrate may make feeding more difficult. Compared to La Nana Creek, Bernaldo Creek turtles had lower mean volumes of fecal contents, suggesting that turtles from Bernaldo Creek consumed less prey. La Nana Creek turtles fed predominantly on freshwater mollusks, which were seemingly abundant in this creek; crushed shells of these animals made up a high proportion of the total volumes from stomach contents of Southeastern Naturalist 129 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 turtles collected from there. Turtles from Bernaldo Creek fed predominantly on crayfish, which may be more difficult for Razorback Musk Turtles to capture, and may contain less nutrients than freshwater mussels. Further data on diet effects on Razorback Musk Turtle growth patterns would be informative. Razorback Musk Turtles from Bernaldo Creek may be smaller because arthropods offer less energy per unit effort than freshwater mussels, and thus, La Nana Creek turtles may have exhibited better feeding efficiency and accelerated growth patterns. Third, the difference in CCL between Bernaldo Creek and La Nana Creek male Razorback Musk Turtles was greater than that of females. Density-dependent effects associated with mating structures may have influenced sexual dimorphism in Bernaldo and La Nana creeks. Data from this study suggest the population density of Bernaldo Creek is lower than that of La Nana Creek. Male turtles from Bernaldo Creek may garner few reproductive benefits by defending females or resources because the probability of interacting with other males is low and physical confrontations less frequent, thus there may not be any selection favoring large male size. Conversely, Razorback Musk Turtle populations from La Nana Creek exist at higher densities than Bernaldo Creek, and competition between males may be more frequent, explaining the more-pronounced sexual dimorphism observed in this creek. Acknowledgments We thank Dan Saenz, Warren Conway, and Brent Burt for technical assistance, and Taylor Hall, Taylor Cotten, Richard Adams, Erica Lopez, Erin Fucik, and Scott Wahlberg for their assistance in the field. Literature Cited Ahle, R.C. 1991. A study of three creeks in Nacogdoches County, Texas using an index of biological diversity. M.Sc. Thesis. Stephen F. Austin State University, Nacogdoches, TX. Atkinson, C.A. 2013. Razor-backed Musk Turtle (Sternotherus carinatus) diet across a gradient of invasion. Herpetological Conservation and Biology 8:561–570. Auffenberg, W. 1977. Display behavior in tortoises. American Zoologist 17:241–250. Auffenberg, W., and W.G. Weaver Jr. 1969. Gopherus berlandieri in southeastern Texas. Bulletin of Florida State Museum of Biological Science 13:141–203. Barzilay, S. 1980. Aggressive behavior in the Wood Turtle, Clemmys insculpta. Journal of Herpetology 14:89–91. Berry, J.F., and R. Shine. 1980. Sexual size-dimorphism and sexual selection in turtles (Order Testudines). Oecologia (Berlin) 44:185–191. Brooks, R.J., M.L. Bobyn, D.A. Galbraith, J.A. Layfield and E.G. Nancekivell. 1991. Maternal and environmental influences on growth and survival of embryonic and hatchling Snapping Turtles (Cheldra serpintina). Canadian Journal of Zoology 69:2667–2676. Burge, B. 1977. Daily and seasonal behavior and areas utilized by the Desert Tortoise (Gopherus agassizii) in Southern Nevada. Proceedings of the Desert Tortoise Council Symposium 1977:59–94. Cagle, F.R. 1950. The life history of the Slider Turtle, Pseudemys scripta troostii (Holbrook). Ecological Monographs 20:31–54. Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 130 Vol. 15, Special Issue 9 Chen, Tien-His, and Kuang-Yang Lue. 1999. Food habits of the Chinese Striped-necked Turtle, Ocadia sinensis, in the Keelung River, Northern Taiwan. Journal of Herpetology 33:463–471. Congdon, J.D., and J.W. Gibbons. 1985. Egg components and reproductive characteristics of turtles: Relationships to body size. Herpetologica 41:194–205. Congdon, J.D., and J.W. Gibbons. 1987. Morphological constraints on egg size: A challenge to optimal egg-size theory? Proceedings of the National Academy of Sciences of the United States of America 84:4145–4147. Congdon, J.D., and D.W. Tinkle. 1982. Reproductive energetics of the Painted Turtle (Chrysemys picta). Herpetologica 38:228–237. Congdon, J.D., J.W. Gibbons, and J.L. Greene. 1983. Parental investment in the Chicken Turtle (Deirochelys reticularia). Ecology 64:419–425. Congdon, J.D., G.L. Breitenbach, R.C. Van Loben Sels, and D.W. Tinkle. 1987. Reproductive and nesting ecology of Snapping Turtles (Chelydra serpintina) in southeastern Michigan. Herpetologica 43:39–54. Congdon, J.D., R.D. Nagle, A.E. Dunham, C.W. Beck, O.M. Kinney and S.R. Yeomans. 1999. The relationship of body size to survivorship in hatchling Snapping Turtles (Chelydra serpintina): An evaluation of the “bigger is better” hypothesis. Oecologia 121:224–235. Cox, R.M., S.L. Skelly, and H.B. John-Adler. 2003. A comparative test of adaptive hypotheses for sexual size-dimorphism in lizards. Evolution 57:1653–1669. Cox, R.M., M.A. Butler, and H.B. John-Adler. 2007. The evolution of sexual size-dimorphism in reptiles. Pp. 38–49, In D.S. Fairbairn, W. Blankenhorn, and T. Szekely (Eds.). Sex, Size, and Gender Roles: Evolutionary Studies of Sexual Size Dimorphism. Oxford University Press, Oxford, UK. 280 pp. Darwin, C.R. 1871. The Descent of Man, and Selection in Relation to Sex. J. Murray, London, UK. Davies, N.B. 1991. Mating systems. Pp. 263–294, In J.R. Krebs and N.B. Davies (Eds.). Behavioural Ecology: An Evolutionary Approach. 3rd Edition. Wiley-Blackwell, Oxford, UK. 432 pp. Emlen S.T, and L.W. Oring. 1977. Ecology, sexual selection, and the evolution of mating systems. Science 197:215–223. Ernst, C.H., J.E. Lovich, and R.W. Barbour. 1994. Turtles of the United States and Canada. Smithsonian Institution Press, Washington, DC. 578 pp. Forsman, A., and R. Shine. 1995. Sexual size-dimorphism in relation to frequency of reproduction in turtles (Testudines: Emydidae). Copeia 3:727–729. Galbraith, D.A. 1991. Studies of mating systems in Wood Turtles (Clemmys insculpta) and Snapping Turtles (Chelydra serpintina) using DNA fingerprinting. Ph.D. Dissertation. Queen’s University, Kingston, ON, Canada. Galbraith, D.A., M.W. Chandler, and R.J. Brooks. 1987. The fine structure of home ranges of male Chelydra serpintina: Are Snapping Turtles territorial? Canadian Journal of Zoology 66:2623–2629. Galbraith, D.A., B.N. White, R.J. Brooks, and P.T. Boag. 1993. Multiple paternity in clutches of Snapping Turtle (Chelydra serpintina) detected using DNA fingerprints. Canadian Journal of Zoology 71:318–324. Gehring, T.M., and R.K. Swihart. 2004. Home range and movements of Long-tailed Weasels in a landscape fragmented by agriculture. Journal of Mammalogy. 85:79–86. Southeastern Naturalist 131 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 Gibbons, J.W. 1986. Movement patterns among turtle populations: Applicability to management of the Desert Tortoise. Herpetologica 42:104–113. Gibbons, J.W., and J.E. Lovich. 1990. Sexual dimorphism in turtles with emphasis on the Slider Turtle (Trachemys scripta). Herpetological Monographs 4:1–29. Gibbons, J.W., J.L. Greene, and K.K. Patterson. 1982. Variation in reproductive characteristics of aquatic turtles. Copeia 1982:776–784. Gibbons, J.W., J.L. Greene, and J.D. Congdon. 1990. Temporal and spatial-movement patterns of Sliders and other turtles. Pp. 223–232, In J.W. Gibbons (Ed.). Life history and Ecology of the Slider Turtle. Washington: Smithsonian Institution Press, Washington, DC. 367 pp. Hammer, D.A. 1969. Parameters of a Marsh Snapping Turtle population, La Creek Refuge, South Dakota. Journal of Wildlife Management 33:995–1005. Hooge, P.N., and B. Eichenlaub. 2000. Animal movement extension to Arcview. ver. 2.0. Alaska Science Center, Biological Science Office, US Geological Survey. Available online at http://gcmd.nasa.gov/records/USGS_animal_mvmt.html. Horn, H.S. 1966. Measurement of “overlap” in comparative ecological studies. American Naturalist 100:419–424. Herrel, A., E. De Grauw, and J.C. O’Reilly. 2005. Ontogenetic scaling of bite force in lizards and turtles. Physiological and Biochemical Zoology 79:31–41. Iverson, J.B. 2002. Reproduction in female Razorback Musk Turtles (Sternotherus carinatus). Southwestern Naturalist 47(2):215–224. Jackson, C.G. 1969. Agonistic behavior in Sternotherus minor minor Agassiz. Herpetologica 25:53–54. Janzen, F.J., J.K. Tucker, and G.L Paukstis. 2000. Experimental analysis of an early lifehistory stage: Selection on size of hatchling turtles. Ecology 81:2290–2304. Kauffman, J.H. 1992. The social behavior of Wood Turtles, Clemmys insculpta, in central Pennsylvania. Herpetological Monographs 6:1–25. Kolbe, J.J., and F.J. Janzen. 2001. The influence of propagule size and maternal nest-site selection on survival and behavior of neonate turtles. Functional Ecology 15(6):772–781. Krebs, C. 1989. Ecological Methodology. Harper and Row, New York, NY. 654 pp. Lagarde, F., X. Bonnet, B. Henen, J. Corbin, K. Nagy, and G. Naulleau. 2001. Sexual sizedimorphism in Steppe Tortoises (Testudo horsfieldi): Growth, maturity, and individual maturity. Canadian Journal of Zoology 79:1433–1441. Lardie, R.L. 1983. Aggressive interactions among melanistic males of the Red-eared Slider, Pseudemys scripta elegans (Wied). Bulletin Oklahoma Herpetological Society 8:105–117. Lindeman, P.V. 2000. Evolution of the relative width of the head and alveolar surfaces in map turtles (Testudines: Emydidae: Graptemys). Biological Journal of the Linnean Society 69:549–576. Lindeman, P.V. 2003. Sexual difference in habitat use in Texas Map Turtles (Emydidae: Graptemys versa) and its relationship to size dimorphism and diet. Canadian Journal of Zoology 81:1185–1191. Lovich, J.E., and J.W. Gibbons. 1992. A review of techniques for quantification of sexual size-dimorphism. Growth, Development, and Aging 56:269–281. Lovich, J.E., C.H. Ernst, R.T. Zappalorti, and D.W. Herman. 1998. Geographic variation in growth and sexual size dimorphism of Bog Turtles (Clemmys muhlenbergii). American Midland Naturalist 139:69–78. Southeastern Naturalist B.T. Kavanagh and M.A. Kwiatkowski 2016 132 Vol. 15, Special Issue 9 Maher, C.R., and D.F. Lott. 2000. A review of ecological determinants of territoriality within vertebrate species. American Midland Naturalist 143:1–29. Mahmoud, I.Y. 1967. Courtship behavior and sexual maturity in four species of Kinosternid turtles. Copeia 1967:314–319. Mahmoud, I.Y. 1969. Comparative ecology of the Kinosternid turtles of Oklahoma. The Southwestern Naturalist 14:31–66. Manly, B.F.J. 1986. Multivariate Statistical Methods: A Primer. Chapman and Hall Press, London, UK. 208 pp. Mann, G.K.H., M.J. O’Riain, and M.D. Hofmeyer. 2006. Shaping up to fight: Sexual selection influences body shape and size in the Fighting Tortoise (Chersina angulata). Journal of Zoology 269(3):373–379. McCulloch, D.L. 1981. The benthic macroinvertebrate communities of Alazan Creek and Bernaldo Bayou in Nacogdoches County, TX. M.Sc. Thesis. Stephen F. Austin State University, Nacogdoches, TX. McRae, W.A., J. Landers, and G.D. Cleveland. 1981. Sexual dimorphism in the Gopher Tortoise (Gopherus polyphemus). Herpetologica 37:46–52. Morreale, S.J., J.W. Gibbons, and J.D. Congdon. 1984. Significance of activity and movement in the Yellow-bellied Slider Turtle (Pseudemys scripta). Canadian Journal of Zoology 62:1038–1042. Moskovits, D.K. 1988. Sexual dimorphism and population estimates of the two Amazonian tortoises (Geochelone carbonaria and G. denticulate) in northwestern Brazil. Herpetologica 44:209–217. Noble, G.K. 1939. The role of dominance in the social life of birds. The Auk 56:263–273. Pearse, D.E., and J.C. Avise. 2001. Turtle mating systems: Behavior, sperm storage, and genetic paternity. The American Genetic Society 92:206–211. Pitelka, F.A. 1959. Numbers, breeding schedule, and territoriality in Pectoral Sandpipers of northern Alaska. The Condor 61:233–264. Rovero, F., M. Lebboroni, and G. Chelazzi. 1999. Aggressive interactions and mating in wild populations of the European Pond Turtle, Emys orbicularis. Journal of Herpetology 33:258–263. Shine, R. 1989. Ecological causes for the evolution of sexual dimorphism: A review of the evidence. The Quarterly Review of Biology 64:419–461. Shine, R. 1994. Sexual size-dimorphism in snakes revisited. Copeia 1994:326–346. Stamps, J.A. 1983 Sexual selection, sexual dimorphism, and territoriality. Pp. 169–204, In R.B Huey, E.R. Pianka, and T.W. Schoener (Eds.). Lizard Ecology: Studies of a Model Organism. Harvard University Press, Cambridge, MA. 501 pp. St. Clair, R. 1998. Patterns of growth and sexual size-dimorphism in two species of box turtles with environmental sex determination. Oecologia 115:501–507. Tanaka, S., and F. Sato. 1983. Brief observation of the mating behavior of the box turtle, Cuora flavomarginata flavomarginata, in nature. Biological Magazine Okinawa 21:75–76. Tinkle, D.W. 1958. The systematics and ecology of the Sternotherus carinatus complex (Testudinata: Chelydridae). Tulane Studies of Zoology 6:1–56. Travis S.E., C.N. Slobodchikoff, and P. Keim. 1995. Ecological and demographic effects on intraspecific variation in the social system of Prairie Dogs. Ecology 76:1794–1803. Tucker, A.D., N.N. Fitzsimmons, and J.W. Gibbons. 1995. Resource partitioning by the estuarine turtle Malaclemys terrapin: Trophic, spatial, and temporal foraging constraints. Herpetologica 51:167–181. Southeastern Naturalist 133 B.T. Kavanagh and M.A. Kwiatkowski 2016 Vol. 15, Special Issue 9 Wilhelm, C.E., and M.V. Plummer. 2012. Diet of radiotracked Musk Turtles, Sternotherus odoratus, in a small urban stream. Herpetological Conservation and Biology 7:258–264. Willemsen, R.E., and A. Hailey. 2003. Sexual dimorphism of body size and shell shape in European tortoises. Journal of Zoology 260:353–365. Zar, J.H. 1999. Biostatistical Analysis. 4th Edition. Prentice-Hall, Upper Saddle River, NJ. 663 pp.