Towards precision apiculture: Traditional and technological insect monitoring methods in strawberry and raspberry crop polytunnels tell different pollination stories.

Over one third of crops are animal pollinated, with insects being the largest group. In some crops, including strawberries, fruit yield, weight, quality, aesthetics and shelf life increase with insect pollination. Many crops are protected from extreme weather in polytunnels, but the impacts of polytunnels on insects are poorly understood. Polytunnels could reduce pollination services, especially if insects have access issues. Here we examine the distribution and activity of honeybees and non-honeybee wild insects on a commercial fruit farm. We evaluated whether insect distributions are impacted by flower type (strawberry; raspberry; weed), or distance from polytunnel edges. We compared passive pan-trapping and active quadrat observations to establish their suitability for monitoring insect distribution and behaviour on a farm. To understand the relative value of honeybees compared to other insects for strawberry pollination, the primary crop at the site, we enhanced our observations with video data analysed using insect tracking software to document the time spent by insects on flowers. The results show honeybees strongly prefer raspberry and weed flowers over strawberry flowers and that location within the polytunnel impacts insect distributions. Consistent with recent studies, we also show that pan-traps are ineffective to sample honeybee numbers. While the pan-traps and quadrat observations tend to suggest that investment in managed honeybees for strawberry pollination might be ineffective due to consistent low numbers within the crop, the camera data provides contrary evidence. Although honeybees were relatively scarce among strawberry crops, camera data shows they spent more time visiting flowers than other insects. Our results demonstrate that a commercial fruit farm is a complex ecosystem influencing pollinator diversity and abundance through a range of factors. We show that monitoring methods may differ in their valuation of relative contributions of insects to crop pollination.