Spatially explicit power analyses to inform occupancy‐based multi‐species wildlife monitoring programmes

Current and accurate information on wildlife populations is integral to successful biodiversity management and conservation globally. Nevertheless, many monitoring programmes fail in their attempts to accurately monitor populations of interest due to interlinked issues including insufficient sample sizes, inappropriate duration, lack of reproducibility and lack of clearly stated objectives. These common pitfalls could be avoided through the elicitation of explicit monitoring objectives and the a priori use of simulations to inform minimum sampling design requirements to meet said objectives. Here, we provide a blueprint for using spatially explicit power analyses to inform the design and implementation of multi‐species monitoring programmes on landscape‐scales. As a demonstration, we used spatially explicit simulations to devise a suitable sampling regime to meet clearly specified monitoring objectives in New York State, USA: to use annual occupancy‐based monitoring to be able to detect 25% and 50% changes in abundance of populations over 5‐ and 10‐year periods for all species of management interest. We focused our simulation efforts on three challenging focal species that differ notably in their morphology, life histories, space use, detection probability, habitat suitability and population sizes/trajectories, and thus provide extremes in the challenges presented when it comes to sampling appropriately to detect changes in abundance. Our simulations demonstrate context‐dependent trade‐offs in sampling designs (i.e. number of sites [ J ] and number of occasions [ K ]) and identify necessary minimum detection probabilities that must be attained to achieve statistical power to detect changes of varying magnitudes in populations of varying sizes in the three focal species. The simulations highlight that monitoring population increases is likely beyond the reach of occupancy‐based monitoring programmes for wide‐ranging or locally abundant species. Synthesis and applications : We combine the results from the single‐species simulations to produce a multi‐species sampling design that meets the specified decline‐based objectives for all three species. While the case study is centred on developing a multi‐species sampling regime for New York State, it provides a reproducible step‐by‐step framework using established methods for wildlife managers and other practitioners to inform their own context‐ and objective‐specific multi‐species occupancy‐based monitoring programmes.