Wagner, N., J. Pfrommer & M. Veith

Comparison of different methods to estimate abundances of larval fire salamanders (Salamandra salamandra) in first-order creeks. pp. 265-274 plus Supplementary documents.

Abstract. The European urodelan diversity is threatened by the recent range expansion of the chytrid fungus Batrachochytrium salamandrivorans. The fire salamander can be classified as especially sensitive as infection of individuals with this parasitic fungus usually leads to chytridiomycosis and rapid mortality. Hence, to observe effects at the population level, it is crucial to monitor population sizes and trends. The abundance of adult fire salamanders is far more difficult to estimate compared to larval stages. Therefore, population monitoring of fire salamanders focusses on the aquatic larvae. We examined abundances of larval fire salamanders in two different first-order creeks. Four different methods were compared: two simple count methods, and two abundance estimates using removal sampling and capture-mark-recapture (CMR), respectively. The results of the count methods strongly correlated with the abundance estimates. Furthermore, the results of the removal sampling correlated with the estimates using CMR data. The estimates of the CMR study should get closer to real larval abundances in the creeks compared to the estimates of the removal sampling approach. Count methods but also removal sampling analysis underestimate real larval abundances. Due to (i) the strong correlations of the results, (ii) less time and cost effort of removal sampling compared to CMR, we suggest the removal sampling method, keeping in mind that real larval population is larger.

Key words. Amphibia, Caudata, Capture-mark-recapture (CMR), program MARK, program POPAN, removal sampling, Batrachochytrium salamandrivorans.