SALAMANDRA 46(1) 37–47 20 FebruaryEcology 2010 of RhacodactylusISSN 0036–3375 auriculatus Ecological observations on the Gargoyle Gecko, Rhacodactylus auriculatus (Bavay, 1869), in southern New Caledonia Joshua Snyder1+2, Leslie Snyder2 & Aaron M. Bauer1 1) Department of Biology, Villanova University, 800 Lancaster Avenue, Villanova, Pennsylvania 19085, USA 2) 883 South 630 West, Payson, Utah 84651, USA Corresponding author: Aaron M. Bauer, e-mail: [email protected] Manuscript received: 1 September 2009 Abstract. Ecological niches of squamate reptiles are delineated by diet, microhabitat use, and activity period. In the cur- rent study field data were gathered to characterize the ecology of Rhacodactylus auriculatus with regard to these three axes. Rhacodactylus auriculatus regularly consume a wide taxonomic and ecological variety of arthropods, lizard prey including geckos and skinks, and various plant materials including floral parts and sap. Based upon the variety of dietary constituents and the regularity with which they are utilized, R. auriculatus may have the most atypical of all gekkonid diets. Rhacodacty- lus auriculatus partition their microhabitat with conspecifics based on perch height, presumably to avoid aggressive interac- tions. During winter in southern New Caledonia, R. auriculatus were most active from one to four hours after sunset. Key words. Diplodactylidae, Rhacodactylus auriculatus, New Caledonia, ecology, diet, activity, microhabitat. Introduction is also known to utilize mangroves and strand vegetation (Bauer & Vindum 990, Bauer & Sadlier 2000). Within The gargoyle gecko, Rhacodactylus auriculatus, is a com- these areas, R. auriculatus is most frequently encountered monly kept and bred diplodactylid gecko. It is the small- on shrubs, saplings, and trees of small diameter. It is often est of six recognized species in the genus, reaching only found within a few meters of these and is regularly active about 25 mm SVL (Sameit 986, Bauer & Sadlier 2000). on the ground, at least seasonally, in association with for- Rhacodactylus auriculatus occurs in both humid forest and aging and mating (Bauer & Sadlier 2000). maquis (a shrub-dominated habitat underlain by ultrama- In order to establish a more complete picture of the au- fic substrates) across the southern third of the New Cal- tecology of Rhacodactylus auriculatus, a field study was un- edonian mainland (Grande Terre) from sea level to nearly dertaken to quantify aspects of microhabitat use and activ- 000 m, but appears to be absent from the Isle of Pines and ity and to assess diet during the cool period of the austral its satellite islets (Bauer & Sadlier 994a, 2000). It has re- winter. cently been recorded from numerous isolated areas in the north of the island (Whitaker et al. 2004), and the tax- onomic status of these populations is under investigation Materials and methods (Bauer et al., unpublished). Study site and specimen documentation Although there is a large body of captive data about the species (e.g., Mertens 964, Rösler 988, Henkel 99, Rhacodactylus auriculatus were studied in Le Parc Provin- Pether 999, Tytle 2000, de Vosjoli et al. 2003, Bach cial de la Rivière Bleue (22°06’S, 66°39’E) in the southern 2004, Henkel & Schmidt 2007), field-based observa- part of the Province Sud on the Grande Terre of New Cale- tions are more limited, and have chiefly been based on op- donia (Figures –2). The park is a protected area with limit- portunistic data gathering (e.g., Roux 93, Meier 979, ed access. It was selected because of its location in the cent- Bauer & Vindum 990, Bauer & Sadlier 994b, 200, de er of the known contiguous distribution of R. auriculatus Vosjoli & Fast 995, Seipp & Henkel 2000). The infor- (Bauer & Sadlier 2000). The site contained both humid mation available, however, suggests that R. auriculatus has forest and maquis habitats (Figure 2). Other geckos in the a particularly diverse diet that is highly atypical for geckos, study area included Bavayia septuiclavis, chiefly in maquis incorporating vertebrate prey, mollusks, and plant mate- and R. ciliatus, R. sarasinorum, Eurydactylodes symmetri- rial, as well as a broad range of arthropods (Bauer & Rus- cus, and Bavayia geitaina in forest. The study took place sell 990, Bauer & Sadlier 994b, Snyder et al. 2008). from 22 June 2004 to 4 August 2004 during the cool New The preferred habitats of this species are shrubby maquis Caledonian winter season. Data were recorded on a total vegetation, primary and secondary forest edges, and the of 30 nights within this interval as heavy rains occasion- ecotonal region between forest and maquis, although it ally precluded locating geckos by eye-shining. Transects © 2010 Deutsche Gesellschaft für Herpetologie und Terrarienkunde e.V. (DGHT), Rheinbach, Germany All articles available online at http://www.salamandra-journal.com 37 Joshua Snyder et al. were followed along unpaved roads within the Parc Rivière altimeter, and measurements of vegetation were made with Bleu, typically from 8:00 to 22:30 h (although on warm- a tape measure (perch height) and digital caliper (perch er nights, when activity remained high, searching contin- diameter). Temperature was recorded each night at the ued until 0:00 h). Searches were not randomized but were beginning of each search period. Ambient light was cal- concentrated in areas found to be most productive. culated using data from the U.S. Naval Observatory site Geckos were located primarily by eye-shining and cap- (http://aa.usno.navy.mil/data/docs/RS_OneDay.html), tured by hand. For each specimen SVL (mm), total body which provides information about percent reflected visible mass (g), tail condition (original versus regenerated), sex, moonlight and moonrise/set time. and age class (juvenile [< 88 mm SVL] versus adult) were Single factor ANOVAs and regression analyses were recorded. A permanent black marker was used to write a performed in order to investigate the relationship between unique number on the ventral surface of the body so that perch height and diameter and gecko mass, SVL, and sex, individuals could be recognized when recaptured. Such and between temperature, ambient light, and precipitation markings are lost at ecdysis, but this occurs less frequently and average hourly capture rate. at cooler temperatures and lower metabolic rates (Alex- ander & Brooks 999) and none of the recaptured geckos shed during the period of the study. Nonetheless, to ensure Diet unambiguous identification of individuals digital photo- graphs were taken to record color pattern, which is highly Diet was determined through stomach flushing. Although variable in this species (Meier 979, Henkel 986, Bauer this method does not retrieve items that have passed below 990, Wirth & Peukert 2009; Figures 3A–B) and can be the pyloric sphincter, it has been found to be highly reli- used to identify individuals. able at recovering stomach contents and is non-destruc- tive (Legler 977, Legler & Sullivan 979). The mouth was held open by a plastic cuff and a human infant feeding Habitat and activity tube with sub-terminal holes to decrease pressure placed on the pyloric sphincter was connected to a 0 cc syringe For each specimen captured the following data were re- and passed down the throat to the center of the stomach corded: time of collection, latitude and longitude, eleva- (James 990). Approximately 0 cc of water (adjusted for tion, weather condition, phase of moon, vegetation type body size) was injected into the stomach. This was repeated (maquis or forest), perch height (m), and perch diameter 3–5 times unless stomach contents were regurgitated soon- (mm). Geographic coordinates were recorded with a Mag- er, in which case one additional flushing took place to en- ellan GPS, elevations were determined using a barometric sure the removal of all stomach contents. Stomach contents Figure 1. Map of New Caledonia showing the position of the study site (Parc Rivière Bleue). 38 Ecology of Rhacodactylus auriculatus Figure 2. Habitat of Rhacodactylus auriculatus. (A) View of the Rivière Bleue valley from an adjacent hilltop. (B) View of the Rivière Bleue from the Pont Germain showing humid forest trees flanking the banks. (C) Humid forest vegetation near a roadside in the Parc Rivière Bleue. Such roads served as transects for locating Rhacodactylus auriculatus. (D) Maquis vegetation, characterized by low canopy and generally small-diameter trees and shrubs along a roadside transect in Parc Rivière Bleue. were stored in 70% alcohol for subsequent identification. Results Specimens were released at the site of capture. Stomach Microhabitat flushing resulted in no fatalities or obvious damage to the geckos. Specimens weighing less than 2 g and those that One hundred and two captures, representing 88 unique had been stomach-flushed within the previous seven days specimens, were made (Figure 4). Six specimens were re- were not examined for dietary items. captured once and four were recaptured twice. Thirteen Dietary niche breadth for Rhacodactylus auriculatus Rhacodactylus auriculatus were found in humid forest and was calculated by number and volume of items in each 89 were found in maquis. All individuals found in the hu- food category using the Shannon index of diversity (Pian- mid forest were located along the roadside or at the ecotone ka 966, Peet 974, Pielou 977, Castanzo & Bauer 993, with maquis, where Bavayia septuiclavis was also common. 997, Vitt 995). Prey items were identified to the lowest The mean (+ S.D.) perch diameter for R. auriculatus re- taxonomic possible and larval coleopterans and lepidop- corded in humid forest was 40.0 ± 46.9 mm and the mean terans were treated as prey categories separate from adults. perch height was .6 ± 0.8 m. The mean perch diameter Calculations were based on the combination of results of and height for specimens recorded in the maquis was 2.0 the present study and previously published quantitative ± 23.2 mm and .8 ± . m, respectively. dietary data (Bauer & Devaney 987; Bauer & Sadlier Differences between mean perch height usage for males, 994b).