Merging Crop Pollination with Remote Sensing to Better Understand Spatial Variability in Australian Tree Orchards

Abstract

<p>Ensuring the sustainability of crop production, whilst simultaneously taking actions to mitigate the environmental impacts of agriculture, is a current global priority. Given around 75% of global food crop yields benefit from pollination services provided by diverse wild and managed insect taxa, management strategies that support diverse communities of pollinator taxa are valuable to ensure ongoing pollination service provisioning and agricultural production. In addition to pollination, realised crop yields are also influenced by other biotic and abiotic factors which vary across different spatial and temporal scales. This thesis addresses three important aspects of crop pollination, namely the need to merge disparate research fields, the degree to which pollinator taxa service multiple crops and regions and how pollination interacts with crop tree physiological factors such as tree vigour.</p> <p>First, I reviewed the literature to evaluate the knowledge gaps concerning pollinator effectiveness and the utility of using remote sensing in crop pollination research. I conducted surveys and pollen deposition trials to identify pollinators in avocado, mango and macadamia crops in three geographically distinct growing regions in Australia across three years. Using single visit deposition rates, bipartite networks and spatial analyses I also investigated pollinator service provisioning and the land use types that influence pollinator communities in these crop and regions. Using hand pollination trials over two years I investigated the impact of supplemental cross pollination on the yield of avocado trees.</p> <p>My first review identified important research directions to account for pollination processes occurring at a community level including: plant-pollinator interactions, heterospecific pollen transfer and variation in pollination outcomes. My second review identified the areas in which remote sensing technologies can facilitate our understanding of interactions between pollinators, pollination services, environmental and plant physiological factors which affect final harvest measures. </p> <p>Using multi-crop, multi-year and multi-region crop-pollinator networks I demonstrated that shared wild pollinator taxa visit multiple crops across several regions. In particular, honey bees (<i>A. mellifera</i>) and two families of wild visitors, Syrphidae and Calliphoridae, are present across all regions and crops. Further, regional comparisons for both avocado and mango crops identified additional shared families that were locally abundant such as Coccinellidae and native Apidae. </p> <p>I found that the effect of additional cross pollination on trees of different vigour varied between individual orchard blocks and across years. General patterns relating to the impact of interaction between tree vigour and pollination on yield were discernible in this study, with lower and medium vigour trees responding more positively to supplemental pollination than high vigour trees. High variability in results and differences in effect response across orchard blocks highlight the need to investigate further factors at a tree and block scale, in future analyses.</p> <p>My research indicates that there is significant potential to identify shared pollinators that provide services across multiple crops. Pollination management strategies that are regionally specific and that include bee and non-bee taxa and co-flowering crop species are needed to ensure ongoing effective and resilient pollination services are delivered to crop systems. The merging of different research fields, such as remote sensing, pollinator ecology and precision agriculture offers exciting new approaches to facilitate our understanding of these complex crop-pollinator interactions.</p>