Avian community composition and its variation across kauri forests in the Waitākere Ranges infected with Phytophthora agathidicida

Kauri (Agathis australis) is an environmentally and culturally important tree species that is threatened by the disease known as kauri dieback, caused by the pathogen Phytophthora agathidicida (PA). Given the known effect kauri has on controlling the forest species around it, there is still little known on the effect that kauri dieback is having on the larger ecosystem, and whether bird communities are responding to the disease presence. Acoustic recordings were used to assess the prevalence of bird species over four seasons in six plots across the Waitākere Ranges. Each plot was given an infection score from a combination of PA presence and the canopy health of 10 sample kauri trees. Tree community composition was established from stem counts and total basal area from all the tree species in each plot. Multidimensional scaling analyses were done to assess the similarities between plots based on the tree community composition and separately for the bird community composition. The axes from the basal area MDS as well as the infection score were used as predictors in a linear mixed effect model testing the response of each bird species. While controlling for seasonal differences as a random effect. Twenty bird species were recorded across the six plots, with only 10 appearing above at least once per day when averaged. The composition of species changed significantly at each plot across the seasons, with all species being more abundant in summer and spring. When bird prevalence was plotted in summer/spring seasons against infection levels, Pīwakawaka (fantail), Miromiro (tomtit) , and Tūī showed a suggestive trend towards higher prevalence at lower infection plots, with Tauhou responding in the opposite direction. However, when run through the linear mixed effect model, only three species had any significant response, and this was in response to the first axis of the multidimensional scaling plot assessing tree community composition that appeared to correlate with kauri dominance. Blackbird responded positively, while Pīwakawaka and Ruru responded negatively. There was no statistical response to infection presence in any species. A mantel test between the bird community and tree community dissimilarity index also found no statistical correlation between the two (rM = -0.175, p-value 0.73). These findings suggest that while bird communities do differ between plots, it is unclear what is the cause of the difference.