Cost-effectiveness of two monitoring strategies for the great crested newt (Triturus cristatus)

Designing cost-effective monitoring protocols is a fundamental prerequisite for amphibian conservation. Here, we report a comparison of flashlight survey and trapping (with and without light sticks as trap baits) in order to determine flashlight detectability and trap detectability of great crested newts (Triturus cristatus). Twelve ponds were surveyed in Switzerland where T. cristatus had been known to occur. We measured covariates affecting both flashlight detectability and trap detectability. Newt flashlight detectability using 20 min long flashlight surveys was on average +/- SE = 39% +/- 10%). Flashlight detectability was mostly influenced by surface and submerged vegetation density, as well as by water temperature. Newt trap detectability during one night using six funnel traps per pond was on average +/- SE = 41% +/- 10%. Trap detectability was mainly affected by trap position in the pond, with traps lying on the pond floor being more likely to attract newts. The use of light sticks did not enhance the trap detectability. Estimates of flashlight detectability and trap detectability were used to define how many times the sites have to be visited to be 95% certain of not missing T. cristatus in ponds where they are present. In both cases multiple visits (7 flashlight surveys or 6 trapping sessions) have to be performed. Flashlight surveys are the most easily applied and most cost-effective method to use in large scale programs.