Postbreeding Movements of the Dark Gopher Frog, Rana sevosa Goin and Netting: Implications for Conservation and Management

-Conservation plans for amphibians often focus on activities at the breeding site, but for species that use temstrial habitats for much of the year, an understanding of nonbreeding habitat use is also essential. We used radio telemetry to study the postbreeding movements of individuals of the only known population of dark gopher frogs, Rana sevosa, during two breeding seasons (1994 and 1996). Movements away from the pond were relatively short (< 300 m) and usually occurred within a two-day period after frogs initially exited the breeding pond. However, dispersal distances for some individuals may have been constrained by a recent clearcut on adjacent private property. Final recorded locations for all individuals were underground retreats associated with stump holes, root mounds of fallen trees, or mammal burrows in surrounding upland areas. When implementing a conservation plan for Rana sevosa and other amphibians with similar habitat utilization patterns, we recommend that a temstrial buffer zone of protection include the aquatic breeding site and adjacent nonbreeding season habitat. When the habitat is fragmented, the buffer zone should include additional habitat to lessen edge effects and provide connectivity between critical habitats. For our study site, we recommend a 1000-m buffer zone around the primary breeding site and each of two other potential breeding ponds. For amphibians that breed in temporary ponds, the hibernation sites, breeding sites, and foraging areas may be temporally and spatially separated, and individuals must migrate to and from these sites in seasonal cycles (Semlitsch, 1981; Sinsch, 1990). Because individuals are generally concentrated only during the breeding season, many studies of amphibian biology take place in and adjacent to breeding sites and not in the nonbreeding habitats. However, designing a comprehensive management plan for any amphibian that uses terrestrial habitats for mu& of the year requires an understanding of habitat use during both breeding and nonbreeding seasons (Dodd, 1996; Dodd and Cade, 1998). Postbreeding movements of amphibians into adjacent terrestrial habitats are poorly understood, and distances that most species normally disperse are unknown (Dodd, 1996; Dodd and Cade, 1998). Although mark-recapture studies have provided valuable data on dispersal distances, recent advances in radio-telemetric methods make this a more suitable method for studying postbreeding movement pattems and habitat selection. Although these data are critical " in understanding habitat use by frogs, few telemetry data are yet published. One study of ranid frogs (Rana clamitans) using terrestrial habitats showed a maximal dispersal distance of 560 m, although there was considerable interindividual variation (Lamoureux and Madison, 1999). Gopher frogs are rare and poorly studied frogs whose geographic range once extended throughout the southeastern coastal plain from North Carolina to Louisiana. Although once common to abundant in coastal Mississippi and Louisiana (Allen, 1932; Dundee and Rossman, 1989), breeding populations of gopher frogs west of Alabama have been severely reduced in numbers. They are thought to be extirpated in Louisiana and are now known to occu; only at a single location in the De Soto National Forest in Harrison County, Mississippi. A recent study by Young and Crother (2001) indicated that this population is genetically distinct from other populations of gopher frogs and that it should be recognized as Rana smsa Goin and Netting. Thus, &ere 1s a v~tal need for ~nformat~on 0% Author Present both breeding and postbreedlng actlvltles ment of Zoology, University of Oklahoma, Norman, Oklahoma 73019, USA, E-mail richter@ou edu Because gopher frogs are secretwe and d~ffiPresent Address Applied Ecology Rewarch to locate of the breeding Group, Un~vers~ty of Canberra, AustralIan Capit,,l knowledge of their ecology 1s generally llmlted Territory 2601, Australia to stud~es of reproductwe ecology at breed~ng MOVEMENTS AND CONSEIiVATlON OF GOPIiER FliOGS 31 7 s~ tes ( e g , Ra~ley, 1991; Sernl~tsch ct a1 , 1995, Young, 1997, I'al~s, 1998, Rtcliter, 1998) During the nonbreed~ng season, gopher frogs have been reported to take shelter in the burrows of gopher torto~scs, Gop\lcrlri po/t/p/rc7~iilli (Franz, 1986) D~staiices moved from the breed~ng s ~ t c after reproduct~on are unknown, except for two ~nd~viduals in Flor~da that were' tound 1 6 and 2 0 l<m trom a brecd~ng s ~ t e (Carr, 1940, Franz et al , 1988) The absence of quant~f~ed data address~iig postbreed~ng movement patterns makes ~t d~ff~cul t to destgn and assess an appropr~ate conscrvat~on plan for t h ~ s peclcs In tills study, we used r a d ~ o telemetry to determ~ne postbrced~ng movement patterns of R ic.ia~io at ~ t s Mississippi breed~ng s ~ t e and niake spec~ i~c management recommcndat~ons Shr~iy S11c -All work was performed at Glen's Pond and its surroundtngs, located In tlie De Soto National Forest In Harrison County, In southern Miss~sslpp~. Glen's Pond IS an upland, winter-f~ll~ng, ephemeral pond w ~ t h an open canopy located In a primarily longleaf pine (PInus pal~istris) ecosystein. Although most of the surrounding hab~tat IS part of the De Soto Nat~onal Forest, the land approxtmately 200 m north of Glen's Pond was managed by International Paper Company (IP) as a prne plantat1011 unt~l 1999, when ~t was acquired by a pr~vate company for res~deiittal development Rolllo P/cv11~try -Frogs used for r a d ~ o telemetry were captured d u r ~ n g postbreeding Inlgratlons by hand or by driit fence w ~ t h 25 liter p ~ t f d l traps (G~bbons and Sttnl~tsch, 1981) All frogs wcre measured (snout-vent length, SVL) to the nearest m~ll~ineter, weightd to tlie nearest 0 5 g w ~ t h a I'esola '"pr~ng balance, glveli an ~nd~v~c lua l toe c l ~ p fo l lon~~~ig the scheme of Donncdly (1989), trtted w ~ t h transm~ttcrs, and released at the s ~ t e of capture w~tliin 24 h Males were d ~ s t ~ n g u ~ s h e d from tc~nales by tlicir thumbs, wli~cli enlarge d u r ~ n g the breed~ng season, and by the patred lateral vocal s'tcs rrdnsni~tters werc attached to trogs by ustng a s111a1l p~ece of polyetliylene m~croc~itheter tub~ i i g and a bdrb from a large flyline cyelet to ni,il<c a harness (B;trtel t and I'etcrson, 2000) The tub~ng was thre'tded through a prefabr~c'itcd hole In the transm~tter, and tlie free ends of tlie tub~ng werc connected wtth the barb Tlie liarness was posit~oned on tlie wa~s t of the frog by sl~dlng ~t over the cxtcnded h~ncf legs Tlie i ~ t of tlic harness was sii~ig over thc th~glis arid sltghtly loowr aronrid tlic wa~s t We used external t1~1nsrntttcrs (IHolohtl Systems lnc , Canada) w ~ t h a battery life of c~pprox~~n~i te ly 70 days ancl a wc~ght ot 1 44 g ILlrnesscs weighed less than 0 001 g, and the percent of body weight tor the harness plus transm~tter for all ~ndivtduals was 3-5'%, of the total mass of the frog T h ~ s is well below the general rule of 1O1% as tlie maximum we~ght ratio of traiism~tter packages to body Illass (R~cliards et al., 1994) To determ~ne potcnt~al ~ iegat~ve effects on gopher trogs, thc harness design n7as tested on southern leopard frogs, Roll(? s f ~ l ~ ~ ~ i z o ~ ~ ~ ~ ~ l ~ o i o , 111 the laboratory for 60 clays No sk111 abras~ons or other problc~ns were observed, aiid frogs continued to eat norinally The steel barb used to hold thc tub~ng t(3gethc.r was suscept~ble to ~iict~st co~id~tions and would, over time, dcltcrrorate and allow the harness to be lost W~tli few exceptions, frogs were relocated d a ~ ly For each s~gliting, we recorded date, trme, gcncral habitat, '2nd any beliavioral observattons Care was taken to avo~d d ~ s t u r b ~ n g the frogs Fach relocat~on site was marked ~71th plast~c flagg~ng, and tlie distance to the lait sight~ng was meLisured w ~ t h a measuring wheel or li~pchain Directtons of these pos~tions relat ~ v e to one another ancl to tlie center of the pond werc obtained by compass; the coordinates of the final locations were determ~ned wtth a Global Posltion~ng Systein u n ~ t (Tr~mbleNav~gat~on NavBeacon XL@, 1-m accuracy) Migration d ~ s tances were measured from the center of the pond to determ~nc the area used by the population after hrecdtng Stnfl~llci-Statistical tests were performed uslng SYSTAT 7 0 (SPSS, Inc ) Means are follonved by i 1 SE Alpha was 0 05 Gcr~erill M(ri)cvirc~lf IJoftrr 175 -Ei,urteen frogs ( t i~ne males and five females) were equ~pped w ~ t h r a d ~ o transni~tttlrs Two trogs subsequently lost tlie~r transm~tters prlor to movement from the pond, result~ng In a total ot 12 rad~o-telenietrrcd gopher frogs (seven males and five temales) Although the study spanned two separate breed~ng seasons (1994 and 1996), n o 1nd1viducil frog was tracked both years (Table 1 ) Frogs wcre tollowed for 21-88 clays (mean = 52 days) from 5 February to 25 May 1994 or froni 29 February to 6 June 1996 (Table 1 ) All frogs mo\ed relat~vely short d~stances (< 300 m ) troin the pond and cliclngcd location infrequently (Table 1, F I ~ 1) All ~ ~ i l t ~ a l movements occurred < 24 h Collow~ng rele'ise Mean d~stance moved froni [lie center of the pond wds 170 0 i 23 43 tn (range. 49-299 m) Tlie known (= m~ntmum) number of moLcments (changes In loc'it~oii) recorclcd per frog rangccl froni 1-5 (mean = 2 3 t 0 43, Table I ) , and most movements werc assoc~ated w ~ t h I atiifall cvcnts (65'XI ~n 1994, lOO'%, In 1996, Ivg 2) During m~gratron, five ~ n d ~ v ~ d ~ i a l s used clumps of grass for refuge, onc was fo~ind bur~cd approx~m'ately 15 cni tinI I 1 Iota1 s~iovement dlst'rncc, <ind I-rie~r\ulement data lor all frog5 inotiltorccl lirrougli radio telemetry In the 1904 and 1996 brcedrng sca\on\ I Iistance M,iss SVI. N . 1 irom pitn(3 l'r<icking Ilates I.'ro): SCX ( ) : I (iilnl) rnolrc\ ccnlcr (ni) period (d) moni torcd 11 Mar-I5 Api 91 1'3 Mar-5 Apr 91 1'3 Ma1 -3 May '13 1'3 Mar-26 Apr ' 14 I? Mar-? May 91 1'3 Mar-25 May 94 5 tkb-'3 May 94 I A p r l May 96 1'3 Apr-20 May 96 18 ~M'tr-10 Mav 96