Abstract
Foliage loss and possum browsing were scored separately on 1475 southern rata trees and seedlings in three neighbouring study areas in central Westland which had different histories of possum occupation. As the time of possum occupation increased, foliage loss from individual trees and the numbers of trees showing such losses increased sharply in most stands. Dead and dying trees (>75% foliage loss) were scattered in a salt-and-pepper pattern among healthier trees. Basal area of dead and dying trees constituted from 1.5% (5 years' possum colonisation) to 32% (20 years' possum occupation) of the total. Foliage loss occurred over all altitudes and sites sampled, but in one area foliage loss on sites with exposed physiography or with seral vegetation was significantly greater than on other sites. In all areas foliage loss increased with increasing tree diameter and was least from small rata trees in even-sized stands. Foliage loss was also greater in the canopy than in the sub-canopy. Possum browsing patterns were similar to foliage loss patterns. Possum browsing was heaviest in the canopy of large trees but, in the two areas with high numbers of possums, seedlings and small trees were also browsed. Both intensity and occurrence of possum browsing were significantly correlated with foliage loss on individual trees, and dying trees were the most heavily browsed. It is concluded that the stress imposed by possum browsing is a principal cause for the observed dieback of southern rata.