Red deer adult sex ratio: comparing three approaches using camera trap data

Abstract European forests face severe challenges due to climate change, highlighting the importance of tree species diversity for regeneration. The management of browsing species such as red deer ( Cervus elaphus ) is therefore crucial. An important metric for managing population size is the adult sex ratio (ASR), for example, derived from camera trap data. In our study, we address 2 issues specific to camera trap methods separately: avoiding the loss of information contained in the proportion of unknown‐sex sightings and accounting for sex‐specific detection probabilities. Our hypothesis was that these advanced statistical methods reveal a potential bias in the estimation of red deer ASR in a naive approach directly from raw camera trap data. We conducted research in an alpine area of Austria from January 2020 to May 2022. We chose weeks 9 to 17 in spring to obtain pre‐hunting and pre‐birth data. We included unknown‐sex sightings (8.9% of all sightings) with predicted sex (precision of 75%) in the ASR estimation after conducting a model selection process with the goal of the highest precision in prediction. These predicted ASRs showed a non‐significant slight shift towards males, indicating a more balanced ASR than a naive ASR. To include sex‐specific detection probabilities, we ran a Royle Nichols occupancy model for both sexes and calculated ASR from estimated latent abundances. Contrary to our expectations, there were no significant differences between the occupancy‐based ASRs and the naive or predicted ASRs. This suggests that, under the conditions of our study, both the naive approach and more advanced statistical methods yield comparable results in terms of ASR. In the future, this could enable wildlife managers to make informed decisions with minimal effort and cost using tools that do not require complex analysis. If desired, additional ecological information can be derived from advanced statistical methods, such as occupancy modeling, allowing more nuanced interpretations of spatiotemporal behavior.