Plenty of flowers in the field: the roles of resource availability and habitat type in the conservation of flower-visiting insects

Abstract

The conservation of flower-visiting insects is a necessary and worthwhile endeavor given their critical role in the pollination of wildflowers and crops. A lack of flowers is often cited as one of the reasons for declines of important flower-visiting taxa such as bees, so conservation and restoration efforts often revolve around the planting of native wildflowers. Given the wide variety of native wildflowers available to choose from and the expense of sourcing local seed, selecting the most preferred (i.e., the most widely visited) flowers is of the utmost importance. Flower abundance obscures preference because neutral processes, which are thought to be an important determinant of flower-visitor interactions, suggest that an abundant flower is more likely by chance to receive a visit than a rare flower. Therefore, to control for flower abundance and remove the effect of neutral processes on the data, many researchers use preference metrics to select the most preferred flowers from flower visitation data. There are a wide variety of metrics to use, and when I compared five of them using the same flower visitation dataset, I found that they produced wildly different results. The contrasting results were likely due to the varying extent to which each metric controlled for flower abundance and how strongly each metric responded to undersampling. When I experimentally controlled for flower abundance using potted plants of the five most and five least preferred plants as calculated by each metric, I found that the metrics that least controlled for flower abundance best predicted flower visitation. Furthermore, I conducted the potted plant trials in two adjacent natural regions (Grassland and Parkland) and found that insect-flower preferences changed between them, likely due to the differences in the insect communities in each natural region. However, flower abundance in the original Grassland dataset predicted flower visitation better than any metric in both regions. Thus, abundant flowers were preferred by flower-visiting insects, but not because they were abundant, and the effect of neutral processes on flower-visitor data may be overestimated. Understanding how flower-visiting insect communities and flower-visitor interactions vary among already established flower communities and habitat types will help guide flower-visiting insect conservation efforts. In agricultural systems, mass-flowering crop bloom causes the amount of floral resources on the landscape changes dramatically throughout the flight period of many insect species, which can change how they interact with other flowers and how they move about agricultural fields. When I compared flower-visiting insect species composition and flower-visitor interactions within and in field borders adjacent to canola fields in central Alberta, I found that field border type (herbaceous or treed) ultimately drove flower-visiting insect and flower-visitor interaction diversity in agricultural systems. In fact, mass-flowering crop bloom generally had a negative effect on flower-visitor interactions as interactions decreased significantly during canola bloom compared to before or after. However, that result was not consistent in 2022 as it depended on border type. I also found that hoverflies, and particularly one species, Toxomerus marginatus, are integral insects in agricultural systems in central Alberta because they were common crop visitors, especially compared to wild bees which rarely visited canola. When I measured hoverfly movement into and out of canola crops using bi-directional Malaise traps set in the borders of canola fields, I found that field border type and not mass-flowering crop bloom affected hoverfly movement. Finally, from my Malaise trap sampling, I found new Alberta records for a hoverfly species (Platycheirus varipes), along with multiple rare species that were previously either of conservation concern, rarely found (< 15 records across their entire range), or only found via iNaturalist records. Overall, my thesis demonstrates that habitat type (through the comparison of natural regions and field border types respectively) matters more than the quantity of floral resources in shaping insect-flower visitation and insect communities. Flower-visiting insect conservation should focus on diversifying resources, even on a relatively small scale, to conserve the broadest number of species possible. Finally, I also demonstrated the prevalence and likely importance to ecosystem services of hoverflies in agricultural areas in central Alberta.