REPRODUCTION OF THE SMALLEST TORTOISE, THE NAMAQUALAND SPECKLED PADLOPER, HOMOPUS SIGNATUS SIGNATUS Author(s): Victor J. T. Loehr, Brian T. Henen, Margaretha D. Hofmeyr Source: Herpetologica, 60(4):444-454. 2004. Published By: The Herpetologists' League DOI: http://dx.doi.org/10.1655/03-59 URL: http://www.bioone.org/doi/full/10.1655/03-59 BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/ terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder. BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. Herpetologica, 60(4), 2004, 438–444 Ó 2004 by The Herpetologists’ League, Inc. HABITAT SELECTION BY SPHAERODACTYLUS NICHOLSI (SQUAMATA: GEKKONIDAE) IN CABO ROJO, PUERTO RICO 1 2,3 RICARDO LO´ PEZ-ORTIZ AND ALLEN R. LEWIS 1Departamento de Recursos Naturales y Ambientales de Puerto Rico, Negociado de Pesca y Vida Silvestre, P.O. Box 491 Boquero´n, Puerto Rico 00622 2Department of Biology, University of Puerto Rico, Mayagu¨ez, Puerto Rico 00681-9012 ABSTRACT: We studied habitat selection by the litter-dwelling gecko Sphaerodactylus nicholsi in the Cabo Rojo National Wildlife Refuge in southwestern Puerto Rico to determine how this small diurnal species exists in an apparently xeric habitat. We divided the 240 ha refuge into 956 quadrats each measuring 50 m 3 50 m and selected 60 quadrats at random to search for geckos. The 60 quadrats yielded 33 discrete locations with sphaerodactylids that we called occupied sites. Occupied sites were located beneath the dense canopies of broad-leaved evergreen trees and were completely shaded. The average area of an occupied site was 103 m2. We paired each occupied site with a nearby random site and compared relative sphaerodactylid density and habitat variables to determine features important in habitat selection. Occupied sites were in dense shade of evergreen trees and contained trees that produced large seeds or fruits; random sites commonly were exposed to sunlight, either directly or filtered through a thin deciduous canopy. Occupied sites had significantly lower daytime substrate temperatures, more leaf litter, less grass, less bare ground, and an order of magnitude more sphaerodactylids than the random sites. The four random sites with sphaerodactylids also had dense shade. Despite its small size, diurnal habit, and apparent vulnerability to dehydration, S. nicholsi thrives through selection of the scattered patches of relatively mesic environment within a xeric landscape. Key words: Gekkonidae; habitat selection; metapopulation; Puerto Rico; Sphaerodactylus nicholsi THE GEKKONID genus Sphaerodactylus Puerto Rico. It ranges in snout–vent length includes the world’s smallest amniotes (Hed- (SVL) from 8.0 mm in hatchlings to 23.0 mm in ges and Thomas, 2001; Thomas, 1965). Species large gravid females (Lo´pez-Ortiz and Lewis, of this genus often inhabit dry forests on 2002). In the Gua´nica State Forest this species Caribbean islands where their small body size, has been reported to be most abundant in deep relatively large surface/volume ratio, and di- leaf litter in humid ravines (Genet et al., 2001). urnal activity appear to make them vulnerable The closest relative to S. nicholsi (Hass, 1991), to thermal stress and dehydration. At least one S. townsendi (Townsend’s Dwarf Sphaero), of the smallest species, S. parthenopion (Virgin was reported to inhabit leaf litter, decompos- Islands Dwarf Sphaero, after Hedges, 2001), ing logs, and palm trash in eastern Puerto does not have the specific physiological adap- Rico (Gaa-Ojeda, 1983). Similar habitats are tations for water conservation possessed by reported for sphaerodactylids from other desert lizards and shows water loss rates similar islands. Sphaerodactylus parthenopion was to lizards from mesic habitats (Hedges and captured hidden in moist microhabitats on Thomas, 2001; MacClean, 1985). It has been the leeward side of Virgin Gorda (MacLean, argued that this species and other small 1985; Thomas, 1965), S. becki (Navassa sphaerodactylids survive through selection of Sphaero) was found during the day in areas relatively mesic microhabitats within larger of deep shade beneath leaf litter or protected xeric settings (Hedges and Thomas, 2001; by other objects (Powell, 1999), and S. ariasae Powell, 1999) along with adjustment of the (Jaragua Sphaero) of Isla Beata was found reproductive cycle so that abundance of under leaf litter (Hedges and Thomas, 2001). delicate hatchlings peaks during the wetter These patterns are consistent with selection part of the year (Lo´pez-Ortiz and Lewis, 2002). of mesic microenvironments by the geckos Sphaerodactylus nicholsi (Nichol’s Crescent in order to limit exposure to physiologically Sphaero) is a small ground-dwelling gecko challenging conditions of low humidity and inhabiting dry coastal forests in southwestern high temperature. The hot, dry climate of southwestern Puerto Rico was exemplified at the refuge weather 3 CORRESPONDENCE: e-mail, [email protected] station in the year of this study by monthly 438 December 2004] HERPETOLOGICA 439 annual rainfall and temperature with 95% An occupied site was a location within one of confidence intervals of 47.8 6 23.5 mm and the quadrats where at least one individual of S. 26.9 6 0.6 C (J. Schwagerl, personal commu- nicholsi was found. As geckos were not found nication). These conditions combined with the beyond the densely shaded leaf litter zone high surface to volume ratio of S. nicholsi meant beneath particular trees or clusters of trees, choice of microhabitat by this litter-dwelling all occupied sites were patches of litter. We gecko should be critical to its survival. We measured the length and width of each looked for evidence of microhabitat selection to occupied site defining the length as the longest determine how this small diurnal species exists axis of the patch of shaded litter and the width in an apparently xeric habitat. We compared as the longest axis perpendicular to the first. microclimatic and other ecological variables at Area was calculated as the ellipse defined by locations used by the geckos with values of the these axes. All occupied sites were surrounded same variables at nearby randomly selected by a boundary area devoid of geckos that ex- locations. The conditions at the random loca- tended at least 5 m beyond the litter patch tions provided a null hypothesis against which occupied by the geckos. The boundary area to test our prediction that the geckos occupied was not part of the occupied site. If two mesic patches in a relatively xeric setting. occupied sites were less than 5 m apart, they were collapsed into a single site. An index of relative abundance (RA) was MATERIALS AND METHODS established for each occupied site (Brower and The study was conducted at the Cabo Rojo Zar, 1984). The RA was defined as the number National Wildlife Refuge located in the Sub- of S. nicholsi seen per minute of searching tropical Dry Forest Life Zone (Ewel and with a constant level of effort. The observer Whitmore, 1973) in southwestern Puerto Rico would search with the same speed and in- from June 1997, through May 1998. The study tensity of search over each occupied site and site was the 240 ha core area of the refuge that divide the final count of geckos by the number houses the administrative facilities and backs of minutes to obtain the RA. As the observer up to coastal mangroves and salt flats. The moved systematically through the occupied vegetation of the refuge was a mixture of sites the geckos could be counted as they secondary woodland with both deciduous and emerged from the leaf litter, moved over the evergreen broad-leaved trees, grassland, and surface, and then disappeared back into the brush (Startek, 1997). litter. The same person (RLO) carried out all To assure that the area searched for geckos RA measurements. Counts were conducted included a representative sample of habitat on 0930–1100 h and 1430–1700 h. the refuge, we divided the refuge into a grid of As we wished to determine whether occu- 956 quadrats (Krebs, 1989). Each quadrat pied sites were more mesic than average sites measured 50 m 3 50 m (2500 m2). We used on the refuge, each occupied site was paired random numbers to select 60 quadrats to be with a random site located 33 m away along searched for geckos. a random bearing. The random sites were The observer walked a zigzag pattern a representative sample of habitat on the through each of the 60 quadrats searching the refuge near occupied sites. All random sites surface of the ground for S. nicholsi. When were close enough to occupied sites that we a gecko was found, the location was flagged for believed geckos would not be absent simply subsequent delimitation of the occupied site because of distance from a source population. and estimation of relative population density. It A random site did not have to fall within the quickly became apparent that geckos were quadrats used to search for occupied sites. associated with leaf litter beneath shady trees. The random sites were circles 5 m in di- Not all trees supported geckos, but no geckos ameter. If geckos were found in a random site, were found outside patches of shaded litter the area that they utilized was delimited and during the searches of the quadrats.