The long-term survival of species with temperature-dependant sex determination requires a sufficient range of incubation temperatures to ensure that both males and females are produced. The primary sex ratio of sea turtles is determined by the temperature experienced by eggs during the middle third of incubation. Here, we investigated the variability in the production of male and female offspring by loggerhead sea turtles Caretta caretta at six nesting beaches in the temperate breeding area of Zakynthos, Greece. Hatchling sex ratios were estimated using incubation durations and sand temperatures for 2007-2009, while the empirical relationship between air and sand temperature was used to infer historical (1875-2010) and future (2011-2100) hatchling sex ratios. First, all six beaches produced males; 55% of production was across five beaches, while 45% was focused on one beach (primarily in July). Second, male production varied across the season in different years; there was an initial peak in June, with production (rising, declining or plateauing) later in the season being regulated by August air temperatures. Interestingly, the annual male production rate estimated from the 3-year dataset (23%) was half that estimated from the 135-year reconstruction (50%), with the latter showing broad interannual variation. Finally, modelled predictions of future sex ratios ranged from a conservative 7.6% decline in male production by 2100 versus no production by 2038. This study provides a baseline from which predicted trends in hatchling sex ratio, in parallel to regular field assessments, could be used to provide guidelines for the selection of appropriate nest protection management protocols, to maintain required sex ratios that safeguard the future of this population. bs_bs_banner Animal Conservation. Print ISSN 1367-9430 508 Animal Conservation 15 (2012) 508-518

The effect of a summer wildfire on the abundance and community composition of soil macroinvertebrates of a phryganic ecosystem was studied. The numerical responses of macroinvertebrates to fire were variable, most notably the reduction of the saprophagous fauna. The magnitude of the changes in abundance due to fire was comparable or even lower than the variations imposed by the seasonality of the mediterranean climate. fire did not seem to affect strongly the community composition of the soil inhabiting fauna, but caused considerable changes in the composition of the litter-dwellers. Furthermore, seasonal changes in community composition were less pronounced in the burnt site compared to those in the control. The observed effects of fire should be attributed to the destruction of the litter layer and the above ground vegetation cover, rather than to the immediate death of the animals, since most of them were absent from the upper soil layers during summer, that is when fire occured.