Effects of sampling methods on bee pollinators observed in Cucurbita pepo

This study was performed to compare catch abundance of three commonly used bee sampling methods within pumpkin and squash (Cucurbita pepo) fields. The three methods of sampling included visual sampling, vacuum sampling, and bowl trapping. Cucurbita pepo require bee pollinators to move pollen from male flower to female flower for fruit to set. In Cucurbita pepo the three major expected bee types that should be observed are squash bees Xenoglossa pruinosa Say (formerly Eucera (Peponapis) pruinosa) [Hymenoptera: Apidae], bumble bees (Bombus spp.) [Hymenoptera: Apidae], and honey bees (Apis mellifera L. [Hymenoptera: Apidae]). Knowing what bees are present in a grower's field could help them to make field management decisions and potentially reduce input costs, thus we chose methods where collected specimens could be submitted to experts for identification. We used visual observations as our control as we could confirm these bees were contacting flower reproductive segments, and this method is widely adopted in literature. Through this study, a total of 2,502 bees were observed, of which 82% were squash bees, 3.4% were honey bees and 2.4% were bumble bees. Bowl sampling differed significantly from visual observations for all three major bee types, having significantly fewer catches. Vacuum sampling recovered no bumble bees. Vacuum sampling was not significantly different from visual observation for honey bees but caught fewer squash bees relative to visual sampling. This study also looked at other collected morphotaxa of bees, particularly in bowl trapping, however bees beyond the three taxa presented above had limited presence in visual observations, and as such were not considered to play a major role in the pollination of these plants. This study indicates that visual observation of flowers provides a more reliable estimate of active pollinators in C. pepo fields than the two other bee sampling methods, vacuum sampling and bowl trapping, and they should likely not be deployed by growers to gain a representative sample of active pollinators in the field.