|dc.description.abstract||New Zealand’s apiculture industry is the fastest growing in the world, expanding in agricultural landscapes, as well as in native ecosystems. While this has obvious benefits for economy and industry, the impacts on sustainability of native ecosystems are less easy to discern. Honey bees (Apis mellifera) have a suite of potential impacts, both positive and negative, on native plants and flower visitors in native ecosystems. This research aimed to investigate the impact of managed introductions of Apis mellifera in native forest dominated by Weinmannia racemosa and Ixerba brexioides, two native forest trees used extensively for monofloral honey production in New Zealand. Research focused on three key areas: 1) timing and availability of floral nectar resources; 2) impacts on plants, particularly I. brexioides and W. racemosa; and 3) impacts on invertebrate flower visitor communities.
Availability of floral nectar resources from I. brexioides and W. racemosa was assessed using a combination of nectar collection and phenology data. Pollination potential of honey bees was assessed using video surveillance and effects on seed set of I. brexioides and W. racemosa were observed using exclusion experiments. Community-level effects on invertebrate flower visitors were assessed using collection of flower visitors and assessment of community data using multivariate and other statistical approaches.
Timing and availability of floral nectar resource showed extreme variation between annual cycles. Nectar sugar production was lowest during a hot, dry summer compared with a cooler, wetter summer, in terms of both sugar production per flower and flower production per tree. At a landscape scale, this can have serious flow-on effects for foraging nectar-feeders, and hence for seed set of flowering plants.
Video surveillance showed that suitability of different flower visitors for pollination of I. brexioides and W. racemosa differed. For W. racemosa all groups of flower visitors contacted reproductive structures, allowing for successful pollination. However for I. brexioides, pollination potential was greatest for birds, beetles and native bees and least for spiders, wasps, ants and honey bees. Seed set for I. brexioides was highest at pest-proof fenced sites and lowest at high hive density sites, whereas W. racemosa seed set was highest at high hive density sites, and lowest at low density sites. Weinmannia racemosa had lowest levels of pollen limitation at sites of high hive density. The combination of W. racemosa responses at the high hive density site suggest that small-flowered species, such as W. racemosa, have the potential to benefit from increased pollination success in areas where honey bees are frequent visitors.
Community analysis showed differences in flower visitor communities between sites with high and low honey bee hive density. Honey bees were the key species contributing to differences between high and low hive density sites. Diversity of insect flower visitors was higher at low hive density sites. Network analysis highlighted structural differences in networks between high and low hive density sites in terms of connectance, nestedness, and species-level indices. High and low hive density sites had a similar number of species that were native and non-native, but high hive density sites had more frequent interactions with non-natives, and 45 % of those interactions were from honey bees.
Pilot studies investigating methods for studying plant-pollinator interactions highlighted a need to tailor methods of pollen isolation to fit the research question, i.e. whether research is focused on pollination interactions or diet-related questions. Comparison of methods for understanding plant-pollinator interactions concluded that identification of pollen by microscopic means identified a greater breadth of plant-pollinator interactions than that identified by field observations alone. However, DNA-based methods of pollen identification have the potential for even greater specificity, cost-effectiveness, and answering a range of questions not possible with traditional methods.
Analysis and findings from this research support a case indicating that honey bees can affect seed set of native plants, and communities of invertebrate flower visitors in a number of ways. Prevention of permanent changes to flower visitor communities should be prioritised by preserving large areas of intact native forest where low levels of fragmentation create refuges for native flower visitors. Estimates of annual sugar production for I. brexioides and W. racemosa call attention to the need to build greater flexibility into legislated stocking rates in native forest, to minimise competitive effects on native flower visitors during low production years. Developing these measures will build sustainability into New Zealand models of apicultural practise, ensuring longevity of honey operations and protection of native ecosystems.||