In order to understand the mechanisms underlying pollinator-dependent plant reproduction in cultivated or remnant landscapes, we need an in-depth knowledge of fine-scale interactions between insects and flowering plants. The advent of robust, effective and high-resolution molecular techniques, such as DNA/RNA sequencing, have been pivotal in facilitating the plight of pollination ecologists to track pollen movement between flowers and insects from specific plant species or cultivars. This thesis aims to progress this knowledge by investigating pollen carried by honeybees in apple orchards to (i) investigate the use of transcriptome analyses as a novel molecular metric and to evaluate its utility in measuring pollinator effectiveness" (ii) examine the gene expression response to honeybee flower visits" (iii) generate molecular markers for different apple cultivars, and (iv) examine the microbiome communities related to pollination using metagenomics approaches.

First, we developed a field experiment in a model pollinator-dependent crop (apple, Malus domestica Borkh.) and used high throughput RNA sequencing (RNA-seq) to quantify pollen flow by measuring changes in gene expression between different pollination treatments. By combining the genotyping data, the differential expression analysis, and traditional fruit set field experiments, we evaluated the pollinator effectiveness of honey bee visits under orchard conditions. This is the first time that pollen-stigma-style mRNA expression analysis has been conducted after a pollinator visit (honeybee) to a plant (in vivo apple flowers), providing evidence that mRNA sequencing can be used to address complex questions related to stigma–pollen interactions over time in pollination ecology (Chapters 2 and 3). Second, we used the pollen loads carried by honeybees foraging in an apple orchard to boost the potential to genotype crop cultivars. In this experiment, we used the Oxford Nanopore (Minion) sequencing to recover long DNA reads from pollen. Using the pollen DNA long reads information, we developed cultivar specific low-cost molecular PCR markers that can be used to test pollinator effectiveness at broad scales (Chapter 4).

Finally, by combining the RNA/DNA sequencing information from pollinated stigmas, apple pollen and pollen carried by honey bees, we conducted a metagenome analysis to identify pathogen and probiotic microorganisms transported in pollen during the pollination process. The results demonstrate that fungi, archaea, bacteria, and viruses can be detected in pollen. More importantly, pollen arriving on stigmas from different cultivars and plant-insect interactions, increases the chance of pathogen transport. Importantly, the results indicate that insect pollinators can be used as indicators of pollinator and orchard health via environmental DNA/RNA assessment. These results demonstrate additional applications of sequencing information to elucidate new relationships in plant-pollinator network dynamics (Chapter 5).

Insects are essential for the reproduction of pollinator-dependent crops and contribute to the pollination of 87% of wild plants and 75% of the world’s food crops. Understanding pollen flow dynamics between plants and pollinators is thus essential to manage and conserve wild plants and ensure yields are maximized in food crops. However, the determination of pollen transfer in the field is complex and laborious. We developed a field experiment in a pollinator-dependent crop and used high throughput RNA sequencing (RNA-seq) to quantify pollen flow by measuring changes in gene expression between pollination treatments across different apple (Malus domestica Borkh.) cultivars. We tested three potential molecular indicators of successful pollination and validated these results with field data by observing single and multiple visits by honey bees (Apis mellifera) to apple flowers and measured fruit set in a commercial apple orchard. The first indicator of successful outcrossing was revealed via differential gene expression in the cross-pollination treatments after 6 h. The second indicator of successful outcrossing was revealed by the expression of specific genes related to pollen tube formation and defense response at three different time intervals in the stigma and the style following cross-pollination (i.e. after 6, 24, and 48 h). Finally, genotyping variants specific to donor pollen could be detected in cross-pollination treatments, providing a third indicator of successful outcrossing. Field data indicated that one or five flower visits by honey bees were insufficient and at least 10 honey bee flower visits were required to achieve a 25% probability of fruit set under orchard conditions. By combining the genotyping data, the differential expression analysis, and the traditional fruit set field experiments, it was possible to evaluate the pollination effectiveness of honey bee visits under orchards conditions. This is the first time that pollen-stigma-style mRNA expression analysis has been conducted after a pollinator visit (honey bee) to a plant (in vivo apple flowers). This study provides evidence that mRNA sequencing can be used to address complex questions related to stigma–pollen interactions over time in pollination ecology.