Social behaviour of Sphaerodactylus vincenti vincenti SALAMANDRA 44 1 15-22 Rheinbach, 20 February 2008 ISSN 0036-3375 Social behaviour of the dwarf gecko Sphaerodactylus vincenti vincenti on St. Vincent, Lesser Antilles Martha A. Marcum, Michael A. Powell, Alexander J. Muensch, Helen F. Arnold & Robert Powell Abstract. In June 2006, we staged male/male (m/m), male/female (m/f), and female/female (f/f) (seven each) interactions between pairs of Sphaerodactylus vincenti vincenti (Sphaerodactylidae) from St. Vin- cent, West Indies, and evaluated 23 types of behaviours. Behavioural repertoires diff ered signifi cantly be- tween types of interactions. Six individual behaviours occurred signifi cantly more frequently in at least one type of interaction: crouch, gular-infl ation, and antiparallel-positioning more in m/m than in m/f, tongue-fl ick and rooting more in f/f than either m/m or m/f, and head-turn more in f/f than in m/f en- counters. Female/female interactions consistently included the greatest number of behaviours that could be construed as being aggressive, although attack, the most egregious agonistic behaviour documented, occurred at almost equal frequencies in both types of consexual interactions. Th e most consistently ob- served pattern (found in all three combinations of interactions) began with stare, followed by signaling from a distance, approach or attack, contact, and fi nally withdrawal. Some display/approach sequences were performed with a “limp”. In intersexual encounters, males would initiate an approach from behind or alongside a female, moving toward her tail, and lick-sniffi ng her there. Th e female would then allow the male to mount or, more frequently, violently tail-wave and withdraw. Key words. Squamata, Sphaerodactylidae, Sphaerodactylus vincenti vincenti, St. Vincent, social behav- iour. Introduction Sphaerodactylus vincenti vincenti (Fig. ) is the only species of dwarf gecko on St. Vin- Dwarf geckos in the genus Sphaerodactylus cent and is an island endemic. Other subspe- are small, voiceless lizards that may be active cies of S. vincenti occur on the Martinique, by day, night, or during crepuscular periods St. Lucia, and Dominica island banks (Stein- (Leuck et al. 988, Steinberg et al. 2007). berg et al. 2008). Sphaerodactylus v. vincen- Th e systematic literature on West Indian ti is a relatively large sphaerodactyl, with a species of Sphaerodactylus is abundant, but maximum snout-vent length (SVL) of 40 that concerned with behaviour of Caribbean mm, and is found in leaf litter, bromeliads, (Leuck et al. 988, Regalado 997, 2003) and under rocks and rotten logs (Schwartz and specifi cally Lesser Antillean (Wissmann & Henderson 99), particularly in mesic et al. 2005) dwarf geckos lags far behind, with areas (Steinberg et al. 2007). only fi ve of the over 80 species known to oc- cur in the region (Powell et al. 996) hav- ing received more than cursory examination. Materials and methods According to the few detailed studies, visual signals are a primary means of sex recogni- In June 2006, we collected four groups of tion, although geckos in general appear to 6-0 geckos in a densely shaded area with rely heavily on olfactory cues (e.g., Dial & thick (> 5 cm) mango (Mangifera sp.) leaf lit- Schwenk 996). ter in the Botanical Garden in Kingstown, © 2008 Deutsche Gesellschaft für Herpetologie und Terrarienkunde e.V. (DGHT) http://www.salamandra-journal.com 15 Martha A. Marcum et al. St. Vincent. We observed and videotaped 2 with a paper towel substrate, and misted con- intraspecifi c interactions: seven male/male, tainers twice a day for the duration of captivi- seven female/female, and seven male/female ty. Containers were kept outside on a covered encounters. All animals were released at the porch at ambient temperatures and humidity, site of capture aft er a maximum of three days and were washed thoroughly with soap and in captivity. warm water for each new inhabitant to elimi- We sexed geckos and measured SVL to the nate the possibility of chemical distractions. nearest 0. mm. Individuals were housed in Aft er the settling period, we placed two 6 x 9 x 0-cm plastic cages and circular plas- selected geckos into a washed and dried 2 tic containers with screened lids. In order to x 4 x 3-cm plastic cage with a fresh, white calm and acclimate geckos prior to each in- paper substrate cut to fi t the bottom of the teraction, we isolated them by placing paper cage. Each individual participated in a maxi- between neighbouring cages for a period of mum of three encounters. No animals were at least 20 minutes. We provided each gecko used more than once per day and no pairs A B Fig. 1. Adult male (top) and female Sphaerodactylus vincenti vincenti from the Botanical Garden, Kingstown, St. Vincent, showing some of the colour and pattern variation observed in these geckos. Despite the diff erences evident in these two individuals, S. v. vincenti is not sexually dimorphic, with variation in dorsal and ventral ground colours, colour of the iris, and distinctiveness of pattern elements found in both males and females. Sexes of all lizards used in this study were determined by the presence (males) or absence of enlarged escutcheons. Th ese lizards reach estimated population densities to 5625/ha in moist, shaded leaf-litter (Steinberg et al. 2007). 16 Social behaviour of Sphaerodactylus vincenti vincenti were repeated. Every interaction was taped other animal), on wall (individual climbed on a video-recorder that simultaneously re- the wall of the cage), and rooting (individual corded our audio observations, which were touched nose to the substrate). supplemented with notes by at least two ob- We noted other behaviours observed by servers. We later reviewed the video record- D.S. Steinberg (pers. comm.) that included: ings to obtain a more detailed analysis and gular-pumping (a series of throat infl ations more descriptive data set. As in Wissmann et and defl ations), gular-infl ation (individual al. (2005), we began recording each individ- maintains an engorged throat sac for a peri- ual’s behavioural repertoire, which included od of time), and antiparallel-positioning (in- notations of behavioural sequences, from the dividuals are parallel to one another but fac- time that one lizard initiated a behaviour or ing opposite directions). We modifi ed the lat- assumed a position directed toward the oth- ter to include variations of head positioning er. We ceased recording aft er 5 minutes of at the neck or side of the other animal. We inactivity following any interactions or af- believe that redefi ning this behaviour is ap- ter 20 minutes of total inactivity had passed propriate because each head placement was since the introduction. accompanied by licking and/or sniffi ng and Behaviours were characterised as in Re- therefore seemed to serve the same purposes galado (997, 2003) and Wissmann et al. of attaining information and sexual identifi - (2005) and included: approach (individual cation (Regalado 2003). moved toward the other animal at a normal We combined two behaviours that were pace), withdrawal (individual moved away identifi ed separately by Wissmann et al. from the other animal), stiff (legs extended (2005), as it was diffi cult to distinguish be- posteriorly, sometimes with body tilted to- tween lick (individual tongue fl icks at the ward the opponent, almost always accompa- other animal) and sniff (individual touches nied by arched back and head pointed down), the other animal with its snout without lick- crouch (individual rested its legs alongside ing), labeling the combined behaviour lick- the body while keeping its trunk close to the sniff . We divided raised-tail (Wissmann et al. substrate and head held above the substrate), 2005) into three behaviours distinguished by head-turn (stationary individual assumes duration and extent: raised-tail A (elevated a 45-90° angle between head and long axis tail curved upward), raised-tail B (tail elevat- of the body, oft en alternating the direction ed at the base and center making an arc with of fl exion to the opposite side of the body), the tip pointed down), raised-tail C (tail is tongue-fl ick (rapid extension and retraction fl at on the substrate with only the tip curved of tongue), head-bob (up and down move- upward). ment of the head, noted from the beginning We observed and added one behaviour to the end of a series), and stand (individu- (attack: rapid charge followed by a bite) that al lift ed body partially away from substrate was not noted in any of the previous studies. while stationary). We analysed our data using Statview 5.0 Additional behaviours drawn from Wiss- (SAS Inst., Cary, North Carolina); for all mann et al. (2005) included: stare (individu- tests, α = 0.05. al looked intently at the other animal), walk (individual moved around without appearing to interact with the other animal), tail-wave Results (tail waved or wagged quickly and rhythmi- cally from side to side, noted from beginning Repertoires of behaviours (Table ) diff ered to the end of a series), touch (part of one in- signifi cantly between types of interaction dividual’s body touched the other animal), (contingency test, df = 23, χ2 = 44.82, P < mount (individual climbed on top of the 0.000), but when we compared individual 17 Martha A. Marcum et al. types of behaviours, the only signifi cant dif- Tab. 1. Frequencies of behaviours exhibited by ference was in the occurrence of antiparallel- Sphaerodactylus vincenti vincenti during male/ positioning (ANOVA, df = 2, F = 3.93, P = male (MM), male/female (MF), and female/female (FF) interactions. See text for distinctions between 0.04). Pairwise comparisons (Fisher’s PLSD) “raise-tail A”, “raise-tail B”, and “raise-tail C”.