Honey bees increase their foraging performance and frequency of pollen trips through experience

Abstract

Honey bee foragers must supply their colony with a balance of pollen and nectar to sustain optimal colony development. Inter-individual behavioural variability among foragers is observed in terms of activity levels and nectar vs. pollen collection, however the causes of such variation are still open questions. Here we explored the relationship between foraging activity and foraging performance in honey bees (Apis mellifera) by using an automated behaviour monitoring system to record mass on departing the hive, trip duration, presence of pollen on the hind legs and mass upon return to the hive, during the lifelong foraging career of individual bees. In our colonies, only a subset of foragers collected pollen, and no bee exclusively foraged for pollen. A minority of very active bees (19% of the foragers) performed 50% of the colony's total foraging trips, contributing to both pollen and nectar collection. Foraging performance (amount and rate of food collection) depended on bees' individual experience (amount of foraging trips completed). We argue that this reveals an important vulnerability for these social bees since environmental stressors that alter the activity and reduce the lifespan of foragers may prevent bees ever achieving maximal performance, thereby seriously compromising the effectiveness of the colony foraging force.

Figure 1

Colony entrance with sensors. Bees entered and departed the hive using two different paths. On each path, the bees were individually recognised (RFID), weighed (balance) and filmed (webcam). The entrance and exit tubes were 1 cm diameter transparent plastic tubes. (1) automatic gates, (2) infrared emitter/receiver, (3) RFID antennae, (4) balance, (5) Plastic bristles (forcing the passage of a bee from one direction only), (6) landing platform (open on the outside), (7) webcam.

Figure 2

Changes of foraging performance with experience. (A) Probability for a bee to collect pollen on a given trip according to experience. Estimated curves for each bee according to a binomial GLMM are shown in colours (N = 154 bees, linear predictor Chi_1,2.215 = 97.07, P < 0.001; quadratic predictor: Chi_1,2.215 = 34.72, P < 0.001). (B) Total number of trips performed per bee across successive foraging days. Boxplot: the line shows the median; boxes and the whiskers represent interquartile ranges; dots represent outliers (data greater than third quartile + 1.5* (interquartile range), or less than first quartile − 1.5* (interquartile range)). Blue lines indicate the best fitted linear model obtained from a segmented regression analysis estimating break point at day number 9. (C) Foraging performance (weight difference between end and start of a non-pollen trip) per bee across successive trips. Bees performed better as they gained foraging experience. Only the non-pollen trips were analysed here (N = 277 bees). LMM: F_(1,925) = 6.45, P = 0.016.

Figure 3

Lorenz curves of relative individual contributions to the colony foraging activity. For each colony, bees were ranked by the lifetime number of trips they performed in ascending order. The fraction of each bee’s contribution to the total number of the colony trips was cumulatively plotted in the Y axis. Black dotted lines represent the distribution predicted by an evenly distributed contribution of each bee. Grey dotted horizontal lines indicate the threshold of a contribution to 50% of the total activity. Vertical green dotted lines represent the fraction of foragers, for each colony, for which this threshold was reached. In colony 1, light green, (N = 296 foragers in total): 17.29% of the total of bees performed 50% of the total number of trips. In colony 2, dark green, (N = 270 foragers in total): 20.45% of the total of bees performed 50% of the total number of trips. See Fig. S3 for a similar analysis showing the proportion of foragers that contribute to the number of trips per bee per day.