PARENT SESSION
Contributed Oral Session 17: Reptiles and Amphibians: Habitat; and Communities
Monday, August 8, 8:00 AM - 11:30 AM, Meeting Room 524 B, Level 5, Palais des congrès de Montréal
Larviposition strategies of the fire salamander: A dynamic state variable model and empirical evidence.
Segev, Ori*,1, Blaustein, Leon2, Mangel, Marc3, Wolf, Nicolas4, 1 University of Haifa, Haifa, Israel2 University of Haifa, Haifa, Israel3 University of California, Santa cruz, California4 University of California, Santa cruz, California
ABSTRACT- In Israel, the terrestrial fire salamander, Salamandra salamandra infraimmaculata, migrates to a breeding site once winter rains begin with a full clutch of developed larvae. Since rains are infrequent at the beginning of the rainy season, breeding early comes with an increased risk of desiccating and breeding later comes with an increased risk of being out-competed and cannabilized by larger conspecific larvae. We used a combination of modelling, experiments and field surveys, to assess how Salamandra spread their young both spatially and temporally. A state-dependent model based on stochastic dynamic programming considers physiological state (offspring complement), hydrologic environment (including rainfall probabilities, size and permanence of ponds, probabilities of encountering different types of ponds), larval growth and survival. According to the model females are predicted to spread their young spatially (across pools) and temporally, additionally females are predicted to adjust the number of young deposited according to pool size. To test the model predictions, we performed outdoor enclosure experiments where gravid Salamandra females could freely choose where to deposit in four pools within the enclosure. Salamandra larviposition was consistent with the model predictions. Most females (86%) spread their young over more than one pool. Salamandra also spread their offspring temporally within the pen. Salamandra females, when given the choice of two shallow (5 cm) and two deeper (13 cm) pools, deposited significantly more larvae in the deeper pools. Finally, a three-year field survey of Salamandra breeding sites indicates that individual females adopt different temporal larviposition strategies. Gravid females carrying a full larval load were found at the breeding sites throughout the entire rainy season, from September to March.
Key words: larviposition strategies, dynamic state variable model, Salamandra salamandra infraimmaculata