Linking Measures of Colony and Individual Honey Bee Health to Survival among Apiaries Exposed to Varying Agricultural Land Use

Abstract

We previously characterized and quantified the influence of land use on survival and productivity of colonies positioned in six apiaries and found that colonies in apiaries surrounded by more land in uncultivated forage experienced greater annual survival, and generally more honey production. Here, detailed metrics of honey bee health were assessed over three years in colonies positioned in the same six apiaries. The colonies were located in North Dakota during the summer months and were transported to California for almond pollination every winter. Our aim was to identify relationships among measures of colony and individual bee health that impacted and predicted overwintering survival of colonies. We tested the hypothesis that colonies in apiaries surrounded by more favorable land use conditions would experience improved health. We modeled colony and individual bee health indices at a critical time point (autumn, prior to overwintering) and related them to eventual spring survival for California almond pollination. Colony measures that predicted overwintering apiary survival included the amount of pollen collected, brood production, and Varroa destructor mite levels. At the individual bee level, expression of vitellogenin, defensin1, and lysozyme2 were important markers of overwinter survival. This study is a novel first step toward identifying pertinent physiological responses in honey bees that result from their positioning near varying landscape features in intensive agricultural environments.

Fig 1. Pupating brood population.

The mean area (summed number of combs) containing pupae ± s.e. by site and year, n = 24 colonies per site). Letters denote sites with significantly different total areas of pupating brood within each year.

Fig 2. Pollen collection per day in each site and year, 2010–2012.

Traps were open for 24 hours and fresh weight (g) of incoming pollen from three colonies per site were averaged on each date (± s.e.). Letters denote significant differences among sites within each year.

Fig 3. Status of colonies prior to mortality 2010–2013.

Each apiary started out with a total of 24 colonies in North Dakota each May. Final colony losses were determined in March (almond pollination) of the following year. Status of the queen was assessed every 6 weeks in each colony.

Fig 4. Levels of Varroa destructor in September shown as percent mite infestation per 100 adult bees.

N = 24 colonies per site, letters denote significant differences among sites within each year.

Fig 5. September viral expression by site and year, 2010–2012.

Adult bees were taken from the brood chamber from each of the 24 assessed colonies at each site. Expression levels of viruses were determined relative to the reference gene, RPS5. The lowest levels of viral expression are depicted in black, while the highest levels are in white.

Fig 6. Measures of statistically modeled nutritional status and immune gene expression in individual 7-day old bees collected from colonies in September each year.

Measures of nutrition included gene expression levels of vitellogenin (a) and weight of abdominal lipids (b). Measures of immune functioning included gene expression levels of lysozyme2 (c), and defenisin1 (d). (See S5 and S6 Figs for other related measures). Vitellogenin, defensin1, and lysozyme2 expression were determined from the same bees (n = 47 per site), while abdominal lipids were quantified from separate bees (n = 180 per site). Box and whisker plots depict the median (black line), upper and lower 25% quartiles (the box), maximum and minimum (whiskers) not including outliers, and outliers (open circles) that were greater or less than 1.5 times the upper or lower quartile, respectively.