Drivers of Non-Random Nest-Site Selection in an Oviparous Vertebrate

Selection of a suitable nest site is critical for the reproductive success of oviparous animals. Habitat choices can take place across multiple special and temporal scales. Species that exhibit non-random patterns of nest-site fidelity or nesting phenology may be particularly at risk of declines due to anthropogenic causes. We performed a 7-year mark-recapture study (2008 - 2014) to evaluate the nesting behavior of a population of western pond turtles, Emys marmorata, at a semi-permanent vernal lake in Lake County, CA. We marked 95 adult female turtles and mapped 226 nest locations. Mark/recapture data indicated that there are approximately 108 reproductively active females at this site. Turtles made nesting excursions in the late afternoon and early evening during June and early July of each year, and nested between 1 and 332 m from the lake, usually within 15 m of tree canopy. Individual females nested closer to their own previous nests than the mean inter-nest distance for the population as a whole, and an interaction between carapace length and high temperature was significantly correlated with increasing inter-nest distances. The date of observed first activity in each year was predicted by mean temperature in the preceding month. Initial excursions were positively correlated with increasing maximum temperatures, as well as a significant interaction between maximum temperature and barometric pressure. On average, turtles made their first nesting excursion of the year within 4 days of the previous year's initial excursion date, and day of season out of the water in previous year significantly predicted the day of initial excursion in the current year. Larger turtles came out significantly earlier in the season than smaller ones. These data have potential implications for land management for E. marmorata as well as other oviparous reptilian species.