Using citizen‐collected wildlife sightings to predict traffic strike hot spots for threatened species: a case study on the southern cassowary

Abstract

Summary Assessing the causal factors underpinning the distribution and abundance of wildlife road‐induced mortality can be challenging. This is particularly ubiquitous for rare or elusive species, because traffic strikes occur infrequently for these populations and information about localized abundance, distribution and movements are generally lacking. Here, we assessed whether citizen‐collected sightings data may serve as a low cost and efficient means of gathering long‐term animal roadside presence and road‐crossing information, which could then be used to assess the causative factors and direct mitigation actions aimed at reducing wildlife traffic strike frequency. We explored this principle using two decades of traffic strike records and citizen‐collected sightings of the southern cassowary Casuarius casuarius johnsonii. Roads have bisected the cassowaries’ rain forest habitat and despite considerable investment into mitigation strategies for this species, road‐induced mortality is considered one of the primary threatening processes affecting the population. Using a Bayesian approach and controlling for spatial autocorrelation with conditional autoregressive models, we demonstrate that traffic strikes are primarily a density‐dependent process in the southern cassowary. That is, traffic strike clusters occurred along stretches of road where cassowaries were most frequently sighted. There were, however, road stretches where traffic strike frequency was greater than predicted by the number of roadside sightings, illustrating when and where density‐independent processes increased the mortality potential for a road‐crossing cassowary. Synthesis and applications. This is the first time that citizen‐collected sightings data have been used to systematically inform upon the abundance and distribution of wildlife traffic strike. The technique not only predicts where incidents are likely to occur but also helps us to understand the factors responsible for strike clustering. While not a replacement for systematic surveys, we highlight citizen‐collected sightings data as a low‐cost option when assessing contributing factors to vehicle‐induced mortality. Accounting for density‐dependent and density‐independent processes will ensure the most effective allocation of resources when implementing wildlife traffic strike mitigation.