It’s about time: temporal fragmentation and metrics for habitat change

Habitat fragmentation is commonly analyzed in space, but continuity can also fail in time. Temporal habitat fragmentation describes this disruption of continuity and can be defined as the division of a continuous period of habitat availability into multiple shorter or irregular intervals. Although classic theories such as patch dynamics and metapopulation address habitat change through time, the ecological role of temporal configuration and structure is worth further exploration. My objectives are fourfold: First, I make the case that a broader, explicit set of metrics can help quantify temporal fragmentation and unpack its complexity in greater detail. Second, I define and formalize a suite of temporal metrics by direct analogy to well-established spatial metrics. Third, I demonstrate the behavior of these metrics through a simulated example of daily habitat availability, showing how they separate temporal loss from fragmentation per se. Finally, I discuss how these temporal metrics complement and extend established spatial analyses, and outline future directions for their application. I define metrics for amount, arrangement, short-term persistence, and sustained windows of suitability. These include total habitat time, number of periods, temporal isolation (mean and maximum interior gaps), duration variability, temporal edge density, temporal autocorrelation, temporal aggregation, and a core time index based on minimum run length. I demonstrate their behavior with five simulated daily habitat presence-absence scenarios representing continuous, loss only, fragmentation only, combined loss and fragmentation, and irregular multi-interval seasons. The metrics distinguish temporal loss from fragmentation per se. Equal totals in the amount of time that habitat is available can mask differences in temporal arrangement, revealed by higher isolation and edge density. Shortened seasons can retain substantial usable time when remaining intervals are long, as captured by the core time index, whereas highly split seasons with short bursts reduce persistence and yield a core time index near zero. Treating time explicitly with simple, interpretable metrics complements spatial analyses and improves diagnosis of habitat change. With the emergence of continuous spatiotemporal datasets that can generate accurate representations of the Earth for any specified date (e.g., AlphaEarth), these metrics can be computed at the pixel level to produce spatially explicit surfaces of temporal loss and fragmentation. The simulated dataset, code, and metric definitions provide a practical framework to advance temporal ecology.