We've used my long term attack data for Nile crocs in S Africa and Swaziland to investigate possible links with rainfall, water levels and temperature. Initial findings suggest that water levels and rainfall are not significant, and only minimum daily temperature shows a (strong) correlation.