NorEs AND FrEln Reponrs r09

Table 2.Carapace length (rnm) by scr"rte annr"rli number of western United States and Canada. Srnithsonian Institution Press, Wash- pond turtles (Clemntls tnerntorotct) at Hayfork Creek. Califomia ington. 578 pp. from 1968-13. Means given with standard deviation (SD;. FnnnEr-r-, R.F.. AND GnnHena. T.E. 1991. Ecological notes on the tr"rftle Carapace Length (rnm) Clenunt's irtsailltta tn nofihwestem New Jersey. J. Herpetol. 25:l-9. Annuli Number n Mean + SD Ran_ge Ge.LeRAmr. D.A., AND BRoorcs, R.J. 1981 .Addition of annual growth lines

in adr"rlt snappin-e tuft les. C h e I y d ru s e rp e nti na . J .Herpetol. 2 | :3 59-363 . 0 1 32.73 + I .'18 30.5 - 33.5 I 10 5 1.93 + 4.86 12.5 - 60 GEnveNo. D.J.. AND BuRv. R.B. 1998. Age determination in tr"rftles: 2 30 66.55 + 6.,13 55 - 8r evidence of annual deposition of scute rings. Chelonian Conserva- 3 58 78.1/+ + 6.00 66-92 tion and Biology 3:123-132. 4 58 87.51 + 6.25 t3 - t02 5 6t 94.7 5 + 5.20 83 - r03.5 HenorNc" J.H.. AND Br-ooven, T.J. 1979. The wood tuftle, Clenunys + 6 58 102.47 7.01 76-n1 irtsculpto . . . a natural history. HERP - Bull. New York Herpetol. l 61 107.20 + 15.82 93 t26 - Soc. 15:9-26. 8 60 I t6.73 + 8.37 95 - t26 Males JENr.rrNcs. M.R.. nNo Hnves. M.P. 1994. Amphibian and Reptile Species 9 t9 Izt .21+ 6.08 llr - 13r.5 of Special Concern in California. California Depaftment of Fish and l0 t2 t29.04 + 8.58 l2l - I'15 Females Game. Final Repoft. Contract No. 8023, Rancho Cordova,255 pp. 9 36 t22.31 + 7.88 I 10.5 - r39 LovrcH, J.E.. ERNsr, C.H.. AND McBnEEN, J.F. 1990. Growth. matu- + l0 20 t29.15 9.30 I 13.5 - I'10.5 rity. and sexual dimorphism in the wood tuftle, Clenunt's irtsculptct.

Can. J. Zool. 68:672-671 . but turtle no. 353 from Hayfork Creek only increased carapace Ross. D.A.. BREwsrER. K.N.. ANoERSoN, R.K.. RATNER, N., and length I mm in2 years while still adding 2 annuli (Table l). BRewsrER. C.M. 1991. Aspects of the ecology of wood tuftles,

If annuli are deposited each year, body size should C I e nunt's i r t s c ulpta. inWi sconsin. Can. Field Nanrrali st I 05 : 36 3 -361 . progressively and consistently increase with age (i.e., larger SrEserNs, R.C. 1985. A Field Guride to Western Reptiles and Arnphib- turtles have more annuli than smaller turtles). To test this ians. Houghton Mifflin Cornpany, Boston, 336 pp. relationship, we compared annuli and measured carapace SroRER. T.I. 1930. Notes on the range and life-history of the Pacific fresh-water tuftle. Clemnn's mennorata. Univ. Calif. Publ .Zool. length of 490 C. mannorata from Hayfork Creek, Trinity 32:429-441. Co., California. The size of turtles increased steadily with age based on counting annuli (Table 2). This separate data Receivecl: l8 November 1997 set corroborates that size corresponds to age and annuli in C. Reviev'ecl: 3 May 1998 marmorata up to at least l0 years of age. Revisecl ancl Acceptecl: l0 Jr"rne 1998 Annual deposition of scute rings has been shown for the congeners Clemmys guttata (Ernst, 1975) and C. insculptct 0e- I I I (Harding and Bloomer, 1979; Lovich et al., 1990, Ernst et o ,ee' ,, .'Jliilli::l?:1::1,,'J,l,l,;.i,l,l;rr al., 1994). The maximum number of annuli deposited by congeners is 14-18 for C. guttctra (Ernst, 1975), 13 for C. Bait Preferences of muhlenbergii (Ernst, 1917), and 15-20 for C. inscuplta Southeastern United States Coastal Plain (Harding and Bloomer, 1979; Farrell 199 I and Graham, ; Riverine Turtles: Fish or Fowl? Ross et al., l99l). Most of the C. mannorata we examined seemed to stop depositing countable scute annuli after 12 to B. 14years, but one turtle had l6 countable rings, the maximum JonN JpNSENI number of scute annuli that we have seen for this species. I Depctrtnrcnt o.f lr'lcttural Re source s, These are the first data indicating that scute rings are depos- lrl on garne -Endctn ge recl Wildlife P ro gram, ited essentially annually and that size corresponds to age in I I6 Rurn Creek Drive, Forst'tlt, Georgia 31029 USA juvenile C. marmorata. I F a.r : 9 I 2 -99 3 - 3 050 : E-mai I : j olut j en s en @ ntail. clrtr. state. gct. us ]

Acknowledgments. - We thank J.E. Lovich, P.C.H. Few studies have been conducted comparing the Pritchard, A.G.J. Rhodin, and G.R. Zugfor critically evalu- effic acy of different types of baits for trapping aquatic ating an earlier draft of this paper. We were helped by a turtles. Lagler (1943) considered fresh fish and fowl variety of field assistants but especially L.R. Gangle, G.W. entrails best, but provided no supportive data. Ernst Bury, and R.N. Germano. ( 1965) compared the attractiveness of six different baits (including fresh fish and fowl entrails) to Chrysemys Literature Cited picta, Cltelyclra serpentina, and Sternotherus odoratus. He concluded that, although fish- and fowl-baited traps Bnoorcs, R.J., Knawcnur, M.A., SrEvENs, C., AND KoeeR, N. 1997. Testing captured equal numbers of C. serpentina and S. oclorotLts,, the precision and accuracy of age estimation using lines in scutes of Chelydra serpentirm and Chn,senns picta. J. Herpetol. 3l:521-529. fowl was far superior to fish for attracting C. picta.

EnNSt, C.H. 197 5. Growth of the spotted turtl e, Clenu??.I.r guttata. J . However, since he ran each of the seven bait trials Herpetol. 9:3 13-318. (including one unbaited control) consecutively, with a ERNsr, C.H. l9l7 . Biological notes on the bog turtle. Clenmiys week of no trapping between each trial, biases related to muh I e nb e r gi i . Herpetologica 33 :24 I -246. seasonality, weather, trap habituation, and trap shyness ERNST, C.H., LovrcH, J.E.. AND Bnnnoun, R.W. 1994. Turtles of the may have influenced his results. Further, I am aware of l l0 CHEloNtaN CoNSERVATToN AND BroLoGv, Volume 3, Nuntber I - 1998

no studies comparing bait preferences of riverine turtles Thirty-one traps were baited with chicken and 34 with of the southeastern United States Coastal Plain. fish. Eight to l4 traps were placed in each stretch of stream In conjunction with a status survey of the alligator depending on habitat and trap-site availability. Traps were snapping turtle, Macroclemys temminckii, in Georgia, and set mid-afternoon, left overnight to accommodate the noc- upon learning of the availability of fresh chicken entrails in turnal activity period of M. temminckii (Collins, 1993), and quantity from a local processing plant, I developed a study checked and removed early the next morning. Duration of to compare the effectiveness of chicken entrails vs. fresh fish trapping varied from l2-18 hrs among sites, but by less than as trap bait for this species and other sympatric turtles. Fresh 2 hrs between traps at the same site. Trapping only took place fish, which I find more difficult to acquire in sufficient on days and nights with clear or mostly clear skies and never amounts, does successfully attract M. temmirtckii (pers. during turbid water conditions. Water temperature among obs.; P. Moler, pers. comm.; Pritchard, 1989). Though fish sites varied by no more than 3"C (26-29"C). All captured was likely used by most former commercial trappers of M. turtles were identified to species and released. temminckii, at least some used chicken successfully and Results Six species of turtles M. temminckii, nearly exclusively (P. Moler, pers. comm.); commercial Chelvdra serpentina, Pseuclemys concinna,- Trachemys harvest is now illegal in most states. scriptct, Apalone fero.r, and Apalone spinifera were Methocls The study was conducted during late captured; only M. tentminckii and T. scripta,however, were summer 1997 in six stretches of four Gulf Coastal Plain caught in sufficient numbers for statistical evaluation of bait streams within Georgia, USA: , Chattahoochee preferences (Table l). Macroclentys temminckii showed an River, Spring Creek (Decatur County), and apparent preference for fresh fish (Xt=9.96,df = l, p <0.05). (including Suwannoochee Creek). Single opening hoop- Conversely, T. scripta preferred fresh chicken entrails (1r = traps were employed, consisting of four 4-ft diameter fiber- 9.63, df = 1,, p < 0.05). Neither bait successfully attracted glass hoops supporting 4-tn mesh netting. Bait, either fresh more turtles when all six species were combined (X'= 1.54, df chicken entrails or fresh cut fish (including entrails and - 1, p < 0.05). The large mesh netting precluded the capture of flesh), was stuffed approximately half-full into 1 liter plastic any juvenile turtles, other than Macroclemlts and Chelydra, of bottles, which were then topped off with water and frozen. which no juveniles were caught, and therefore ontogenetic The lid of each bottle was fitted with a clip for attaching the differences in bait preference could not be analyzed. bait to a line within the posterior region of the trap. This Discttssion It is not surprising that both bait types positioned the bait in the center of the cylinder, thus prevent- successfully attracted M. temminckii, given the species' ing turtles from reaching it without entering the trap. Before catholic feeding habits (Sloan et al, 1996). However, the final setting of each trap, the bait bottle was punctured species' apparent preference for fish bait beckons explana- several times using a knife. The closed end of each trap was tion. One hypothesis may be that M. temminckii has evolved tied to a tree, root, or snag on the stream bank with at least with fish being a major dietary component, whereas birds are some portion above the water, while the open funnel end was a minor component at best. Indeed, M. temminckii is well oriented downstream and anchored with a weighted line. All noted for capturing live fish using its unique tongue lure traps were nearly parallel with the water flow and positioned (Drummond and Gordon, 1979). Therefore, M. temminckii so that the hoop at the opening of the trap was resting upright may have evolved a greater sensitivity to fish odor. An on the stream bottom. This positioning presumably facili- alternative may relate to the density differences between the tates cumbersome, bottom-walkin g M. temminckii (Ashton two baits. After removing traps, I noticed that unconsumed and Ashton, 1985) in entering the traps. All traps were placed chicken entrails and their oils floated on the water's surface just upstream from log jams, undercut banks, or other struc- after discarding, whereas all portions of cut fish quickly

tures assumed to be favored aquatic microhabitats of M. sank. As a primarily benthic species , M. temminckii is likely temminckii. While setting traps, I endeavored to alternate bait to smell water-carried odors trapped in the upper surface of types as I moved upstream or downstream. Colored flagging slow-flowing streams only when it surfaces to breathe. on nearby snags or limbs was used to indicate bait type. Though turbulence may permit some of the chicken entrails'

Table 1. Number of turtles captured in traps using two different baits. C = chicken entrails; F = cut fish; TN = trap-nights. No. of individuals captured M. tentminckii C. serpentina T. scripta P. concinnct A. ferox A. spinfera Stream CFCFCFCF CF CF Suwannee River (C=8TN,F=6TN) Alapaha River T2 (C=5TN,F=6TN) Spring Creek (C=8TN,F= 12TN) 52 24 (C=l0TN,F=l0TN)

TOTALS l3 67 39 Norgs AND FIEI-p Rgponrs lll odor to reach deeper water, the possible lack of a continuous bellied turtle, Chn'sern\:s scr'pta scriprc (Schoepf0. Bull. Florida odor trail may hinder the ability of M. temminckii to locate St. Mus. Biol. Sci. 20: l-45. CRcLe, F.R., AND CunNEv, A.H. 1950. Turtle populations in Louisi- the source. Other predominantly benthic turtles, such as ana. Amer. Midl. Nat. 43:383-389. Sternotherus minor and Chel,y,*clro serpentina, may be simi- CoLLnrs, J.T. 1993. Amphibians and Reptiles in Kansas. University larly affected. Though the sample size was too low for Press of Kansas, Lawrence, Kansas, 397 pp. statistical analysis, this point is reinforced by the capture of DnuvvroND, H., AND GonooN, E.R. 1979. Luring in the neonate four of the five C. serpentina in traps baited with fish. alligator snapping turtle (Macroclenws temminckii): description Unfortunately, the relatively large mesh size used precluded and experimental analysis. Z. Tierpsychol. 50: 136-152. capture of Sternotherus. ERNsr, C.H. 1965. Bait preferences of some freshwater turtles. Trachemys scripta, also considered an opportunistic Joumal of the Ohio Herpetological Society 5:53. omnivore (Ernst et al., 1994), was decidedly more attracted ERNsr, C.H., LovrcH, J.E., AND Bnnsoun, R.W. 1994. Turtles of the by chicken than fish. Basking occupies a great deal of the United States and Canada. Washington: Smithsonian Institution Press, 578 pp. daily activity cycle of T. scripta (Auth, l9l5), which there- FRezeR, N.B., GrenoNs, J.W., AND OwENS, T.J. 1990. Turtle rapping: fore involves more frequent exposure to the upper, rather preliminary tests of conventional wisdom. Copeia 1990:1 150- 1152. than lower, water It may be possible that T. column. scripta LnclEn, K.F. 1943. Methods of collecting freshwater turtles. Copeia is more likely than M. temminckii to smell near-surface 1943:21-25. odors than those submerged below. PnrrcHnnn, P.C.H. 1989. The Alligator Snapping Turlle: Biology and Cagle and Chaney (1950) stated, and later Frazer et al. Conservation. Milwaukee: Milwaukee Public Museum, 104 pp. ( 1990) demonstrated that the mere presence of an individual SLoRN, K.N., Burn-nreNN, K.A., nNo Lovtcn, J.E. 1996. Stomach contents turtle in a trap may attract other conspecifics. This attraction of commercially harvested adult alligator snapping turtles, M acroclenu s may have biased the capture data for i". scripta, but not likely tenuninckii. Chelonian Conservation and Biology 2:96-99. for M. temminckii. Two particular traps, baited with chicken, Receivecl: I I January 1998 captured 19 and 14 T. scripta, respectively. Seven T. scripta Review'ed:3 May 1998 per trap was the next highest concentration. If the turtles Revised cmcl Accepted:20 May 1998 from those two traps were eliminated from the analysis (due to possible influence from the presence of other turtles), no tlorticttt Cortsenatiort utttl Biologr, 1998. -3( I ): I I l- I I 2 significant bait preference would have been revealed for Z. O 1998 by Chelonian Research Foundation scripta. Two traps containing four M. temminckii each represented the only multiple captures of this species. Since Sexual Dimorphism of Neonate Eastern one of these traps was baited with fish and the other with Spiny Softshells, Apalone spinrfura spinifera chicken, it is unlikely that bias associated with attractiveness to other conspecifics influenced the analysis. The interest in capturing aquatic turtles for survey and Tpnnv E. GnAHAMI a,Nn CvNrHrA B. CoBBI inventory, rather than simply for subsistence or commercial- tzatron' has made studies such as this important. Increasing I Deportment of Biolog,,-, Worcester State College, the odds of capturing a larger percentage of a particular Worcester, Mcrssachusetts 0 I 602 U SA species' population increases the accuracy of distributional I E-mail : t graham@ worc.nras s. edu ] and demographic surveys, but care must be taken that new and improved capture techniques are not utilized to the Sexual dimorphism in the carapacial pattern of adult detriment of turtle populations. eastern spiny softshells (Apalone spintfera spinrfera) is well known (Webb , 1962; Ernst et al. , 1994). Males have clear Acknowledgments.- I thank Lee Andrews, Paul Moler, paravertebral spots with dark bordering rings and females and Jim Ozier for trapping assistance. Personnel with Geor- lack clear spots. The development of obvious sexual pattern gia DNR's Fisheries Section provided fish. Chris Poplaski differences in headstarted juvenile softshells has been re- helped prepare fish for bait. A special thanks is extended to ported by Graham (1991) at a carapace length (CL) of as Cagle's Inc. for supplying chicken entrails. Georgia DNR little as 52 mm. law enforcement officers provided information on boat On 14 August l99l a clutch of 19 softshell eggs was launch sites and stream conditions. Kathy Gault graciously discovered on a sandbar in the Lamoille River, Chittenden assisted me with the statistical analysis. The manuscript was County, Vermont, USA, and removed to the laboratory for much improved by Dale Jackson, who provided editorial incubation. Seventeen eggs hatched on 9 September and the suggestions. neonates ranged in CL from 39.0 to 41 .7 mm (x - 40.2). We carefully examined each hatchling under a 7X dissecting Literature Cited microscope. Two patterns in the appearance of the anterior paravertebral spots were noted. In the first pattern, the spots AsHroN, R.E., JR., RNo AsuroN, P.S. 1991. Handbook of reptiles and amphibians of Florida, Part II,Lizards, turtles, and crocodilians. were bordered by a distinct dark ring, as in adult males (Fig. Winward Press, Miami, l9l pp. lA), and in the second pattern, the spots had either an AutH, D.L. 1975. Behavioral ecology of basking in the yellow- indistinct or absent ring as in adult females (Fig. 1B). These