|dc.description.abstract||This research investigated the ecology of Ixerba brexioides (tāwari) with regard to pollination, breeding strategy, seed biology, and forest composition. Research was focused around three main questions:
1. How is tāwari pollinated with regard to vector and pollen source? How is it adapted for this?
2. How is seed dispersal of tāwari achieved and under what conditions is germination most successful?
3. What are the dominant community associations of tāwari what how are they constrained by environmental variables?
The first research question was addressed using video surveillance, nectar analysis, and artificial pollination experiments in a small tract of tāwari forest at Tūī Ridge Park in the Mamaku Range, North Island, New Zealand. Analysis of 125 hours of video footage showed that tāwari is predominantly insect pollinated with occasional bird visitation. The most frequent flower visitors were flies and nocturnal moths. Tāwari nectar volume peaked at midday, and declined toward the late afternoon, before increasing again at dusk. This pattern of nectar secretion followed closely the activity patterns of flies during the day and moths at night. Nectar sugar concentration was 11% on average (range 3% to 20%) which is considered low, but is suited to moths, bats, birds, and bees. Exclusion experiments demonstrated that tāwari is capable of producing viable seed under cross-fertilisation, self-fertilisation, and agamospermy. Breeding system indices demonstrated that tāwari is medium pollen limited (PLI = 0.31), self-compatible (SCI = 0.93), and autonomously selfing (ASI = 0.65).
Question two was addressed by video surveillance of tāwari seed capsules, morphological observation, and by a series of germination experiments. Video failure meant that with limited footage (mostly at night) no dispersal activity was captured on film. However, using available literature and observations of seed morphology birds were considered the most probable effective disperser. Germination trials included standard conditions, shade, seeds left in fruit, seeds deposited on soil surface, and seeds buried at 5 cm soil depth. Germination was most effective in the buried treatment (average 85% germination) and the soil surface treatment (average 75% germination).
Cluster analysis and NMS (Non-Metric Multi-Dimensional Scaling) ordination of 641 plots of tāwari forest from the NVS (National Vegetation Survey) database was used to assess the community associations of tāwari. Based on plot species assemblage, the analysis showed four main forest types that are largely separate in geographical space: Northland, Coromandel, Kaimai, and Urewera. Each type had significant indicator species that were constrained in range by latitudinal limits. Recent literature suggests that environmental variables that most influence tāwari forest distribution are annual temperature and rainfall in conjunction with solar radiation. Tāwari forest occupies only areas which are cool and moist, and seedlings are biased toward high light conditions. In the central North Island tāwari had the highest probability of occurrence in sites with a mean annual temperature of 11° to 13° Celsius, mean annual rainfall between 2000 and 2250 mm annually, and mean solar radiation of 143 to 145 MJ m2 per day. Parent material also plays an important part, especially the depth of Taupō Pumice deposits as tāwari forest is more prone to occur in areas of mature soil where depth of volcanic deposits is not excessive.
Information on tāwari from the present thesis combined with other published and unpublished sources is presented in the form of a New Zealand Biological Flora Series journal contribution.
Recommendations for further research on tāwari and New Zealand reproductive biology in general are given.||