Clarifying habitat niche width using broad-scale, hierarchical occupancy models: a case study with a recovering mesocarnivore

• Published in: Journal of zoology (1987) 2016-11, Vol.300 (3), p.177-185
• Main Authors: Moll, R. J., Kilshaw, K., Montgomery, R. A., Abade, L., Campbell, R. D., Harrington, L. A., Millspaugh, J. J., Birks, J. D. S., Macdonald, D. W.
• Format: Article
• Language: English
• Subjects: Bayesian analysis; camera traps; Carnivores; habitat niche; Habitats; hierarchical occupancy models; Martes martes; specialist-generalist; trap happiness
• ISSN: 0952-8369; 1469-7998
• DOI: 10.1111/jzo.12369

A species’ habitat niche width informs its position on the generalist–specialist continuum, which is central to life‐history theory and crucial to conservation planning. However, assessments of niche width are often based on local‐scale studies or qualitative descriptions rather than broad, quantitative assessments conducted in heterogeneous landscapes. Here, we show how broad‐scale, hierarchical occupancy models can clarify a species’ niche width and degree of habitat specialism by evaluating the woodland‐specialist classification of the European pine marten Martes martes. We deployed 526 camera‐trap stations at 27 sites throughout a vast extent (~50 000 km2) in Scotland and modeled pine marten occupancy as a function of habitat characteristics using a hierarchical Bayesian analysis. Our model was flexible to trap happiness due to baiting at camera traps and accounted for spatial autocorrelation among and imperfect detection at camera‐trap stations. We detected a positive association between pine marten occupancy probability and wooded habitats. However, pine marten occupancy probability was also high in numerous non‐wooded habitats, including agricultural land, heather and heather grassland, semi‐natural grassland and areas near anthropogenic structures. Our study is the first to record high pine marten occupancy in open habitats at broad spatial scales and thereby corroborates recent smaller scale indications that pine martens are more of a habitat generalist than previously thought. Our results guide ongoing conservation efforts by identifying that pine martens are not strict woodland specialists, but rather inhabit a mosaic of habitat types in the landscape. More broadly, our case study exemplifies how coupling hierarchical occupancy models with large‐scale experimental designs can clarify a species’ niche width and associated position on the generalist–specialist continuum. Evaluations of wide‐ranging species’ habitat niche width require broad‐scale, quantitatively rigorous approaches, but these are often lacking, especially for species of lesser conservation concern. We present a case study of how a broad‐scale, hierarchical occupancy model, based on data collected across Scotland from 526 camera traps, can be used to update the habitat niche width of a recovering mesocarnivore. Given the dual ubiquity of camera trapping and hierarchical modeling, our approach could be applied in a variety of settings to accurately position species on the generalist