Fig 1.
Flight cage, artificial flowers and visual landmarks.
(a) The flight cage in the experimental field (Egham, Surrey, UK). (b) Artificial flower consisting of a blue horizontal landing platform, an electric syringe pump, a webcam and their supports (see details in Fig 2a). (c) A laptop protected by a golf umbrella with a unique two-colour pattern acting as a three-dimensional landmark for the bees (photographs by Mathieu Lihoreau).
Fig 2.
(a) Schematic of an artificial flower and the video tracking system (drawing by Pierre Vedel). As the electric pump depresses the syringe plunger, sucrose solution is pushed through a plastic tube and accumulates at a constant rate (3.3 μL/min) in a feeding cup (capacity = 40 μL) accessed by the bee through a hole in the middle of the horizontal landing platform. A webcam connected to a laptop computer running motion detection software is mounted directly above the landing platform. The webcam was fitted with a neutral density filter (Neutral Density = 0.6) placed on a truncated cardboard cone, to reduce the amount of light entering the lens. Bee movements in the camera field of view trigger recording of a video clip (minimum duration = 5 s), from which the bee’s tag number, its arrival and departure times, and any interactions with another forager can be identified (e.g. S1 and S3 Videos). (b) Spatial arrangement of the nest box, artificial flowers and laptops within the flight cage. Coordinates (x, y) of the nest box (black square), the pre-training flower (open circle) and the test flowers (black circles numbered 1–10) are in metres. Distance between the nest box and the nearest test flower (number 1) was 12.5 m. Distance between any two nearest neighbour test flowers within the array was 9 m (e.g. flowers 1 and 4). Distance between a flower and the nearest laptop was 5.20 m. Each laptop was protected by a golf umbrella, with a unique two-colour pattern, to provide additional three-dimensional landmarks for the bees. The black cross indicates the position of the experimenter. The black arrow (bottom left) indicates north. Photos of the flight cage, the artificial flowers and the visual landmarks are shown in Fig 1.
Fig 3.
Maps showing the cumulative movement patterns of (a) resident bees during the one-forager phase, (b) resident bees during the two-forager phase, and (c) newcomer bees during the two-forager phase. Visitation networks of bees from the same pair are presented in vertical columns. For each panel, we have represented the nest box (black square), the flowers that were visited at least once (circles), and the flowers that were never visited (grey crosses), by the focal bee. The diameter of each circle is proportional to the cumulative frequency of visits to that flower relative to the total number of visits to all flowers. The four flowers visited most frequently by the focal bee are shown in red. Arrows indicate the frequency and direction in which the bee moved between pairs of locations. Arrow thickness is proportional to the cumulative frequency of transitions between locations relative to the total number of transitions observed. Single-headed arrows indicate the bee was significantly more likely to move in one direction during that transition (binomial test with probability 0.5 to move in either direction, P < 0.05). Two-headed arrows indicate symmetrical transition direction. Labels (Bee 1–14, Pair 1–7) refer to the same individuals throughout the study.
Fig 4.
Average data (mean ± s.e.) for the seven resident (black dots) and the seven newcomer (white dots) bees during all consecutive foraging bouts of the two experimental phases. (a) Mean number of different flowers visited per foraging bout; (b) number of revisits to the same flowers per foraging bout; (c) travel speed per foraging bout; (d) net energy intake rate per foraging bout. Grey arrows show the moment when newcomers were introduced in the array of flowers (bout 26). Additional foraging bouts by resident bees during the two-forager phase (> bout 50) are not shown.
Fig 5.
Cumulative data (mean ± s.e.) for the seven resident (black bars) and the seven newcomer (white bars) bees showing: (a) total number of following events (one bee landed on a flower less than 5 s after the other bee departed) (b) total number of joining events (one bee landed on an already occupied flower); (c) total number of pushing events (one bee pushed the other bee with its head or legs); (d) total number of evictions (one bee moved away, or fell off, the landing platform immediately after being pushed). For each pairwise comparison, the results of a GLMM are shown (Poisson GLMM, random effect: pair identity, fixed effect: bee status (resident or newcomer) on variable of interest). Asterisks represent significant differences (P < 0.05).
Fig 6.
Consequences of direct encounters on subsequent flower visits.
Average (mean ± s.e.) number of visits to a flower during a foraging bout in relation to the outcome of encounters on that flower in the previous bout for the seven resident and the seven newcomer bees: either the focal bee evicted the other forager or the focal bee was evicted by the other forager. Data shown are for resident (black bars) and newcomer (white bars) bees. Upper case letters represent significant differences (P < 0.05; Poisson GLMM, random effects: flower and pair identities, fixed effect: outcome of eviction from a flower in the previous bout on the number of visits to that same flower in the current bout).