RESPONSE OF SELECTED TREE SPECIES TO CULLING OF INTRODUCED AUSTRALIANBRUSHTAIL POSSUMS TRICHOSURUS-VULPECULA AT WAIPOUA FOREST, NORTHLAND,NEW-ZEALAND

Population trends of colonising brushtail possums Trichosurus vulpecul a were monitored between 1981 and 1993 at Waipoua Forest in Northland, New Zealand. Canopy defoliation and possum-related damage to eight pl ant indicator species were assessed annually for four years after an a erial poison operation in 1990, to determine whether the operation had reduced possum numbers and halted possum-related damage to the vegeta tion. Trap catch estimates of possum density increased by over 70% dur ing the 1980s. The poison operation killed 87% of possums, and ongoing leg-hold trapping (1991-1993) maintained population levels at 7-9% tr ap catch. Reduction of the possum population did not produce a signifi cant positive short-term vegetation response, but did halt the continu ing downward trend in vegetation condition evident in nearby forested areas where possum densities were not reduced. Other measures of possu m impact (browsed foliage, stem damage) improved significantly within a year of the poison operation, and may prove more sensitive short-ter m indicators of reduced possum-related damage. The apparent ability of residual possum populations to retard vegetation recovery has importa nt implications for forest health strategies based on intermittent poi son operations. While a large initial population reduction is required to halt vegetation decline, an ability to maintain residual possum po pulations at very low levels appears necessary for significant vegetat ion recovery. Targetting the increased resources now available for red ucing possum populations to fewer areas of higher conservation value m ay well return greater ecological dividends than attempts to sustain l ong-term reduction of possum populations over large areas of more modi fied indigenous forest. (C) 1997 Elsevier Science Ltd.