On the identification of mortality hotspots in linear infrastructures

摘要: Abstract One of the main tasks when dealing with impacts infrastructures on wildlife is to identify hotspots high mortality so one can devise and implement mitigation measures. A common strategy divide an infrastructure into several segments determine number collisions in a segment above given threshold, reflecting desired significance level that obtained assuming probability distribution for collisions, which often Poisson distribution. The problem this approach, applied each individually, identifying false (Type I error) potentially high. way solve recognize it requires multiple testing corrections or Bayesian approach. Here, we apply three different methods required identification hotspots: (i) familywise error rate correction, (ii) discovery rate, (iii) hierarchical procedure. We illustrate application these data two bird species collected road Brazil. proposed provide practitioners procedures are reliable simple use real situations and, addition, reflect practitioner’s concerns towards positive missing true hotspots. Although may argue overly cautionary approach (reducing type be beneficial from biological conservation perspective, lead waste resources probably worse, raise doubts about methodology adopted credibility those suggesting it.