Estimation of breeding probability can make monitoring data more revealing: a case study of amphibians.
Sam S CruickshankAriel BergaminiBenedikt R SchmidtPublished in: Ecological applications : a publication of the Ecological Society of America (2021)
Monitoring programs serve to detect trends in the distribution and abundance of species. To do so, monitoring programs often use static state variables. Dynamic state variables that describe population dynamics might be more valuable because they allow for a mechanistic understanding of the processes that lead to population trends. We fit multistate occupancy models to data from a country-wide multispecies amphibian occupancy monitoring program and estimated occupancy and breeding probabilities. If breeding probabilities are determinants of occupancy dynamics, then they may serve in monitoring programs as state variables that describe dynamic processes. The results showed that breeding probabilities were low and that a large proportion of the populations had to be considered to be non-breeding populations (i.e., populations where adults are present but no breeding occurs). For some species, the majority of populations were non-breeding populations. We found that non-breeding populations have lower persistence probabilities than populations where breeding occurs. Breeding probabilities may thus explain trends in occupancy but they might also explain other ecological phenomena, such as the success of invasive species, which had high breeding probabilities. Signs of breeding, i.e., the presence of eggs and larvae, were often hard to detect. Importantly, non-breeding populations also had low detection probabilities, perhaps because they had lower abundances. We suggest that monitoring programs should invest more in the detection of life history stages indicative of breeding, and also into the detection of non-breeding populations. We conclude that breeding probability should be used as a state variable in monitoring programs because it can lead to deeper insights into the processes driving occupancy dynamics.