Landscape factors influencing honey bee colony behavior in Southern California commercial apiaries

Authors: Meikle, William; Weiss, Milagra; Beren, Eli

Submitted to: Nature Scientific Reports
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/27/2020
Publication Date: 3/19/2020
Citation: Meikle, W.G., Weiss, M., Beren, E.D. 2020. Landscape factors influencing honey bee colony behavior in Southern California commercial apiaries. Nature Scientific Reports. 10:5013. https://doi.org/10.1038/s41598-020-61716-6.
DOI: https://doi.org/10.1038/s41598-020-61716-6

Interpretive Summary: Commercial beekeepers place honey bee hives in many different kinds of landscapes, from land under intensive agriculture to land that has little or no agriculture or other forms of development. These landscapes differ in many ways in terms of agrochemical exposure, food availability and the density of hives in the area. In this study we monitored the size and internal temperature of hives placed in three very different landscapes in southern California: heavy agriculture (Imperial Valley), a peri-urban landscape of citrus and avocado orchards near Escondido, and a higher altitude landscape with few orchards or agriculture of any kind. The same hives were monitored from August until after almond and blueberry pollination in March, so they went through one or two additional landscape changes. We monitored pesticide residues in the honey, bee bread (pollen) and wax. Pesticide residues were higher in the hives from the Imperial Valley, but the hive health appeared to be higher in those hives. We attributed this result to the greater abundance of food in the agricultural landscape. We found that 1) visually assessing the adult bee population is fast and easy but did not reveal many landscape differences; 2) taking photographs of frames with larval bees (brood) was able to detect more differences than the visual assessment but analyzing the photos took a lot of time; 3) internal hive temperature was easy to collect, did not disturb bees and found many differences among hive groups but it is not easy to interpret what the data mean; and 4) weight data from hive scales could be obtained with little hive disturbance and it revealed many differences with clear meaning, but installing hive scales is difficult.

Technical Abstract: In this study we monitored, for each of two years, internal hive temperature of 60 colonies as they moved from agricultural, peri-urban and mountain landscapes in southern California to a common site for blueberry pollination and then to almond pollination. At the commercial sites we also monitored hive weight continuously for approximately 6 weeks in February and March, and we conducted visual assessments of adult bee masses as well as image analysis of brood frame photographs. Varroa densities, and pesticide residue analyses of wax, honey and bee bread were conducted. Significant differences were observed among all response variables, although not necessarily at the same time. Adult bee masses and brood surface areas were higher in the mountain sites than in the agricultural sites in August; by October adult bee mass estimates were not different while brood levels were significantly higher in colonies from the agricultural sites than those from mountain sites. Starting in September, hives from agricultural sites had higher and more constant temperatures than hives from mountain sites, although by February those original landscape-based differences were no longer significant. In February and March of each year approximately half the hives were sent to commercial almond pollination and the remaining hives kept in a commercial blueberry farm. Colonies in almonds gained on average, across both years, 287 g per d while those in blueberry lost on average 68 g per d. Within-day weight data showed different foraging activity patterns between almonds and blueberries. Mite levels were not different among the initial landscape sites. Hives originating from agricultural sites had a greater diversity of pesticides (insecticides, miticides, fungicides, herbicides and synergists) and higher levels of chlorpyrifos than those from the mountain sites, but colony health compared favorably with colonies from non-agricultural landscapes, possibly due to food availability in commercial sites. The costs and benefits of the different kinds of colony-level data were considered, and temperature monitoring was found to have the lowest costs in terms of equipment cost, installation and colony disturbance but information yield is still comparatively low.