Exploring timescales of predictability in species distributions

Accurate forecasts of how animals respond to climate-driven environmental change are needed prepare for future redistributions, however, it is unclear which temporal scales variability give rise predictability species distributions. We examined the that best predicted spatial abundance a marine predator, swordfish Xiphias gladius, in California Current. To understand provide biological predictability, we decomposed physical variables into three components: monthly climatology (long-term average), low frequency component representing interannual variability, and high (sub-annual) captures ephemeral features. then assessed each component's contribution predictive skill spatially-explicit catch. The was primary source catch, reflecting distribution associated with seasonal movements this region. Importantly, found (capturing variability) provided significant predicting anomalous cannot. addition added only minor improvement predictability. By examining models' ability predict anomalies, assess models way consistent goal – deviations distributions from their average historical locations. critical importance climate describing catch matches target timescales forecasts, suggesting potential skillful ecological across short (seasonal) long (climate) timescales. Understanding sources prediction responses gives confidence our accurately abundance, know likely less predictable, under change. This important as continues cause an unprecedented redistribution life on Earth.