Fig 1.
Experimental design and timeline.
Age-marked bees were painted from newly emerging bees between Weeks 1 and 2, then collected at 9 or 16 days old during the aggression assays (above line). Assays and samples collected per week are described below the weekly headings. Pictograms: Colony box and lid = resource (honey, pollen, brood, population) assessment; bee stinging hand = aggression assay; Bee with pollen on corbicula = sampled bees (broodnest and forager); bee with paint mark = bees collected in Groups 1 or 2 (if below the timeline) or that emerging bees were paint marked (above the timeline), referred to as “Age-marked bees”. Image created in BioRender.
Fig 2.
Pollen-deprived colonies had higher defensive assay scores than control colonies.
The figure illustrates the distribution of factors scores across our sampling timeline and between control (black) and pollen deprived (light blue) colonies. Across all observations (N = 75) x-axis denotes discrete time points, and y-axis provides the defensive score. The y-axis extends from lowest to highest score obtained in the data set and tick marks are made at each quartile. For each group, segments extend from minima to maxima with gaps bounding the interquartile range of the data, open circles identifying the mean, and horizontal bars identifying the median. Trending and significant timepoints are indicated with the p-value above the week. Pollen deprived colonies showed a trend of more aggressive behavior (p = 0.086) than control colonies in Week 4. Pollen deprived colonies were significantly more aggressive than control colonies (p < 0.001) in Week 5.
Fig 3.
Behavior-associated gene expression.
The figure illustrates associations between colony defensive behavior and target genes previously associated with the trait. Within panel A, light blue denotes Group 1 individuals and darker blue denotes Group 2 individuals. “Control” refers to colonies that had continual access to all pollen foragers returned with while “Treatment” refers to colonies that had activated pollen traps and were pollen deprived. In (A) measures of gene expression are highlighted for all targets and indicated effects of treatment, time and rearing environment for age-marked bees (top) and foragers (bottom). Significant differences between groups (p < 0.05) are indicated with different letters above the box plots. Cyp6g1/2, GB53860, and inos are all genes historically associated with significant changes in expression in individual bees with high aggression. Expression of Cyp6g1/2 in heads of 16d old age-marked bees was significantly reduced by Treatment in Group 1 bees, while expression in Group 2 bees was low overall (N = 136). Profiles of GB53860 (N = 133) and inos (N = 126) showed expression in age-marked bees was not significantly influenced by Group or pollen deprivation treatment. Among foragers, Cyp6g1/2 expression did not differ due to time or treatment (N = 133), though GB53860 expression increased generally from week 1 to week 5 (N = 146). Expression of inos in foragers increased across time in bees from pollen control colonies, while those from pollen deprivation colonies maintained similar expression across weeks (N = 149). Illustrated in panel (B) is the relationship between GB53860 and Colony Defensiveness Score in foragers. In the panel figure the x-axis corresponds to our Colony Defensiveness Score and they y-axis to the relative expression level for GB53860, with each point indicating an individual bee. The dashed line highlights the observed correlation while the colors delineate time point with green for Week 1 samples and purple for Week 5 samples. Across our panel of gene targets, GB53860 was the only one whose expression level significantly correlated with our measure of defense.
Fig 4.
Physiological and metabolic assessment of groups.
The figure illustrates an analysis of multiple physiological and metabolic parameters including gene expression for honey bees sampled in this study. Panel (A) highlights differences observed in Hypopharyngeal gland (HPG) size of brood nest bees with overall lower size measured in the Treatment (pollen deprived) group in Week 5. In panel (B) differences in lipid content of age-marked bees is demonstrated, with two independent effects observed between rearing groups and age (days old). Between groups, there was significantly greater lipid content in bees that experienced pollen deprivation throughout their entire developmental cycle (Group 2; dark blue) compared to Group 1 (light blue). Across ages, we saw an overall decrease of lipid content independent of group membership. Panel (C) illustrates results from gene expression analyses in age-marked bees where Vg expression showed a Group effect independent of PD or PC treatment, while Ilp1 showed significant differences only in the pollen deprived treatment Group 2. In panel C, light blue denotes Group 1 and dark blue denotes Group 2.