Table 1.
Compounds identified using GC-MS from mandibular gland extracts of honey bee queens that were either raised from first-instar worker larvae (i.e., "high-quality" queens) or third-instar worker larvae (i.e., "low-quality" queens) in 2010 and 2012.
The relative amount (in %) of each compound is given for each treatment group (mean ± s.e.m.). Kovats index values were calculated for the N-methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA) derivatives obtained from the GC retention times. Differences in relative amounts of compounds between high-quality queens and low-quality queens were analyzed with two-tailed non-parametric Wilcoxon tests because of unequal variances. All tests were performed separately for queens raised in 2010 and queens raised in 2012 (see Methods for details).
Fig 1.
Queen grafting age significantly altered the chemical profile of mandibular gland contents.
The chemical composition of mandibular gland extracts of honey bee queens raised from either first instar worker larvae (i.e., “high-quality” queens) or third instar worker larvae (i.e., “low-quality” queens) were analyzed using gas chromatography and mass spectrometry. Principal component analysis of the mandibular extracts from both queen types was done based on the relative proportion of each compound. The first principal component (PC1) explained 46.7% of the variation in mandibular gland composition. The second principal component (PC2) explained an additional 32.6% of the variation. There was a significant difference in the two-dimensional composite measure between high-quality queens (black squares) and low-quality queens (open circles). Solid and dashed ellipses signify 50% confidence intervals for PC1 and PC2 for high-quality queens and low-quality queens, respectively.
Fig 2.
Levels of worker attraction to honey bee queens varied during retinue response bioassays depending on the age at which a worker larvae are chosen to be raised as queen (i.e., "grafting age," see Methods for details).
Worker retinue size was significantly higher when observation colonies were headed by queens raised from first instar worker larvae (i.e., "high-quality" queens) compared to those headed by queens raised from third instar worker larvae (i.e., "low-quality" queens). The total number of instantaneous sampling points for each queen type (n) is denoted within each bar. Retinue size across both treatments was compared with a matched-pair t-test (* P < 0.0001).