Hotspots, Population Turnover and Long-Term Data: Ecological Insights From a Short-Lived Frog

Population turnover (local extinction and recolonization) is a fundamentally important ecological process but one that is not often (or easily) studied. Even more poorly known are the local and landscape factors that influence population turnover. We leverage a long-term monitoring dataset on a short-lived frog (Blanchard's cricket frog, Acris blanchardi) to provide empirical estimates of local extinction and colonization frequency in an occupancy modeling framework that explicitly addresses detection probability. We surveyed 102 aquatic habitats in northwestern Ohio (USA) using 1491 listening surveys from 2004-2008 and 2017-2022. Estimated colonization and extinction rates were very similar over the entire period 2004-2022 (0.1099 (SE = 0.0149) and 0.1014 (SE = 0.0301), respectively). However, colonization and extinction rate estimates were substantially higher in the 2017-2022 interval compared to the 2004-2008 interval. The two most important variables influencing colonization and extinction probabilities were the distance to the nearest "hotspot" (a site that was continuously or nearly continuously occupied throughout our study) and the proximity to a large, interstate highway. The strong importance of "hotspot" proximity suggests that resilient core populations reduce extinction rates and boost colonization rates in nearby populations and generally play an outsized role in overall regional population dynamics. While many other habitat, landscape structure and connectivity variables were not consistently informative, landscape connectivity metrics generally were more powerful than habitat or landscape structure variables. These results highlight the importance of long-term ecological data in understanding the role of episodic processes operating in natural populations.