Comparing sampling effort and errors in abundance estimates between short and protracted nesting seasons for sea turtles

Sea turtles have iteroparous reproduction, migrating periodically from foraging habitat to nesting grounds where they generally lay several clutches at regular intervals throughout a nesting season. The total length of the nesting season depends ultimately on environmental conditions that are conducive to the production of viable hatchlings, and varies from 3 to 4 months to year-round nesting. As with many migratory marine species, the ease of monitoring marine turtles on their nesting beaches opposed to on their foraging grounds, has resulted in a focus of research on breeding females for population studies and provides a useful albeit limited population index. To explore the precision of monitoring regimes to sample nesting turtle populations, we developed theoretical models for 3 1/2-month and 9-month nesting populations. We used individual-based models for tagged animals and parametric and non-parametric models to estimate annual nest abundance for track count data. These simulation models show that seasonality substantially influenced both the length and temporal position of optimal sampling regimes, showing a five to sevenfold greater effort in monitoring required for longer nesting seasons in order to encounter between 83 and 90% of the annual population. The implications for trend detection and inter-annual and inter-species variations are discussed.