Seasonal and habitat-related changes in population density of North Island Saddlebacks (Philesturnus rufusater) on a small island: using distance sampling to determine variation

Abstract

We applied line-transect distance sampling to look at small-scale seasonal and habitat-related changes in abundance of North Island Saddlebacks (Philesturnus rufusater) on a small protected offshore island. We also assessed the recovery of Saddleback density in a mature forest patch on the island after removal of a small number of Saddlebacks for translocation (harvesting). From June 2003 to June 2004, we conducted 34 km of line-transect surveys on 15 transects in five forest patches (two mature forest patches, two replanted forest patches and the harvested mature forest patch). Different seasonal trends were identified for the two habitat types. Densities varied in replanted forest from a low of 3.18 ± 2.9 ha−1 (95% confidence intervals (CI)) in December-January (austral summer) to a high of 18.9 ± 2.5 ha−1 (95% CI) in April-May (austral autumn). In remnant mature forests, density varied less and ranged from a low of 5.23 ± 3.74 ha−1 (95% CI) in June—July (austral winter) to a peak of 14.2 ± 2.55 ha−1 (95% CI) in April—May. Differences in the densities of Saddlebacks in harvested and mature forest patches did not vary significantly. Juvenile density was higher than adult density in replanted patches, and we suggest that the decline in density in spring was caused by juvenile mortality. Although obtaining such detailed density estimates required considerable effort (34 km), this study has provided a current estimate of North Island Saddleback numbers on Tiritiri Matangi Island and has monitored a translocation source population. Finally, this analysis has highlighted testable hypotheses regarding juvenile dispersal and mortality, and seasonal use of different habitat types.