Abstract:
|
Dynamic occupancy models allow for a Markov dependence between the occupancy status of sites across years. They also provide a flexible framework for modeling relationships between covariates and species occupancy, and covariates and colonization/extinction over time. Although dynamic occupancy models have the ability to detect changes in occupancy of imperiled (or invasive) species, guidelines for assessing power to detect such trends are sparse. Additionally, when sample sizes are small or species are difficult to detect, estimators may not converge. Omnibus programs such as the North American Bat Monitoring Program (NABat) are designed to detect changes in occupancy for more than just one species. Therefore, tools for investigating tradeoffs in different trend parameters, numbers of sites, numbers of visits, spatial subsets of interest, and species-specific characteristics are essential. We developed flexible and accessible tools in the form of an R-package for NABat (NABatpow) that can also be used more broadly to address similar questions for other taxa. We present these tools and discuss challenges regarding survey design for large-scale monitoring programs in general.
|