Extinction debt of a common shrub in a fragmented landscape

Environmental stochasticity and low demographic rates may cause delayed extinctions of habitat-specialist species that were initially retained within remnant patches after habitat loss and fragmentation. Detecting such extinction debts opens opportunities to counteract future biodiversity loss, yet knowing the underlying causes of population declines is a basic need for targeting specific guidelines for conservation and restoration (e.g. habitat quantity, quality or connectivity). 2. Here, we examine the extinction debt in the common Mediterranean shrub Myrtus communis (myrtle) occurring in woodland patches of a highly fragmented region that has lost nearly half of the remnant woodland cover during the last 50 years (1956?2002). 3. We sampled myrtle occurrence in 304 woodland patches and modelled its probability of occurrence in relation to patch size, patch disturbance and woodland cover in the surrounding landscape. In order to test for extinction debt evidence, we tested whether myrtle occurrence is better predicted by past (1956) than by present (2002) woodland cover. 4. We found that the probability of myrtle occurrence is associated with present patch features (size and disturbance) that are linked to causes of individual mortality and/or recruitment collapse. However, it was associated with past ? rather than present ? woodland cover in the surrounding landscape, proving a still unpaid extinction debt. Specifically, myrtle occurrence is very unlikely in small and highly disturbed patches located in long-term deforested landscapes. 5. Synthesis and applications. Individual longevity of the Mediterranean shrub Myrtus communis (myrtle) can delay local extinctions, but paying the debt is a matter of time, especially considering the elevated susceptibility of small and disturbed populations to environmental stochasticity. However, these populations still offer management opportunities, and reducing disturbances and improving habitat quality, especially within small woodland patches, should be a top priority for conservation and restoration.Fil: González Varo, Juan P.. Consejo Superior de Investigaciones Cientificas. Estacion Biologica de Doñana; EspañaFil: Albaladejo, Rafael. Universidad de Sevilla; EspañaFil: Aizen, Marcelo Adrian. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Patagonia Norte. Instituto de Investigación en Biodiversidad y Medioambiente; Argentina. Universidad Nacional del Comahue; ArgentinaFil: Arroyo, Juan. Universidad de Sevilla; EspañaFil: Aparicio, Abelardo. Universidad de Sevilla; Españ