Patchiness in distribution of rodents and mustelids in New Zealand forests

Context. Relative density indices assuming uniform distribution of the target species are often the only cost-effective method for monitoring a population over the long term and at landscape scale, and the only source of valuable historical data. Yet, theoretical models emphasise the dangers of ignoring spatial heterogeneity, especially in short-term field data. Aims. To test whether Brown's index of patchiness (BIP) can offer a simple means of checking rodent and mustelid survey data for violations of the assumption of uniform distribution. Methods. We use BIP to interrogate long-term legacy data collected by index trapping of mice (Mus musculus), rats (Rattus rattus and R. norvegicus) and stoats (Mustela erminea) in New Zealand forests. Key results. We found evidence of moderately patchy distributions that were independent of abundance in all three species. In two South Island beech (Nothofagus) forest valleys, 19% (6 of 31) of mouse samples and 8% (3 of 36) of stoat samples were significantly patchy, correlated with a seedfall event; in mixed forest at Pureora in the North Island, significant patchiness in distribution of ship rats was recorded in 19% (16 of 84) of Fenn trap samples and 5% (2 of 42) of rodent trap samples. Conclusions. Moderate patchiness is common. The consequences for any given study depend on the purpose of the work, but may be more important for practical management than for population modeling.