Negative feedback enables fast and flexible collective decision-making in ants

Abstract

Positive feedback plays a major role in the emergence of many collective animal behaviours. In many ants pheromone trails recruit and direct nestmate foragers to food sources. The strong positive feedback caused by trail pheromones allows fast collective responses but can compromise flexibility. Previous laboratory experiments have shown that when the environment changes, colonies are often unable to reallocate their foragers to a more rewarding food source. Here we show both experimentally, using colonies of Lasius niger, and with an agent-based simulation model, that negative feedback caused by crowding at feeding sites allows ant colonies to maintain foraging flexibility even with strong recruitment to food sources. In a constant environment, negative feedback prevents the frequently found bias towards one feeder (symmetry breaking) and leads to equal distribution of foragers. In a changing environment, negative feedback allows a colony to quickly reallocate the majority of its foragers to a superior food patch that becomes available when foraging at an inferior patch is already well underway. The model confirms these experimental findings and shows that the ability of colonies to switch to a superior food source does not require the decay of trail pheromones. Our results help to resolve inconsistencies between collective foraging patterns seen in laboratory studies and observations in the wild, and show that the simultaneous action of negative and positive feedback is important for efficient foraging in mass-recruiting insect colonies.

Figure 1. Proportions of ants visiting two identical 1 molar sucrose feeders each with 1, 3, 9 or 27 feeding holes (Experiment 1).

The blue line represents the feeder that had more ants after 5 minutes, the red line the other feeder. The dashed black line indicates an equal distribution of ants at both feeders. Data represent the mean of 6 test colonies, 1 trial per colony for each number of holes. The shaded areas (light blue and pink) represent the standard errors (SE) of the mean.

Figure 2. The number and behaviour of ants on both branches (Experiment 1).

(A) Mean average difference in the proportions of ants feeding at the two feeders during the whole of the 120 minute trial. Bars show the mean and standard error for the 6 test colonies, one trial each per treatment. The letters above the bars indicate statistically significant differences (linear mixed-effect models: P<0.05; see results for details). (B) Mean number of ants per hole during the whole of the trial. (C) The mean number of successful, i.e. full ants leaving a feeder (averaged for the two feeders) counted over 2 minutes. (D) The mean ratio of empty to full ants returning to the nest for the 6 colonies, measured every 15 min. (E) The mean number of ants laying trail pheromone on either branch. Ants were counted during 2 min every 15 min. (F) The mean proportion of empty (blue line) and full (red line) ants leaving a feeder under high crowding conditions (both feeders had 1 hole) and walking towards the other feeder instead of back to the nest. As can be seen the proportion of empty ants walking towards the other feeder was considerable higher than the proportion of full ants. The shaded areas represent the SE of the mean.

Figure 3. The mean proportion of ants at the two feeders, in which the second feeder (red line) had three times as many feeding holes but was made available 15 minutes after ants starting collecting syrup at the first feeder (blue line) (Experiment 2).

As can be seen, the lines cross for the 1 versus 3 hole situation, but not for 9 versus 27. The shaded areas represent the SE of the mean.

Figure 4. Proportions of agents visiting two identical food patches each with space for 8, 24, 72 or 216 foraging agents (Model 1).

The blue line represents the patch that had more agents after 600 time steps, the red line the other. The dashed black line indicates an equal distribution of agents at both feeders. Data averaged from 30 simulations in each situation. The standard deviation (StDev) is shown in light blue and pink. The StDev used instead of the SE because the SE is too small to be seen by eye.

Figure 5. Proportions of agents foraging at the two food patches, in which the second patch (red line) allowed three times as many agents to feed simultaneously but was made available 900 times steps after agents started foraging at the first food patch (blue line) (Model 2).

Data averaged from 30 simulations in each situation. The StDev is shown in light blue and pink.

Figure 6. The number of agents on both branches and the time colonies needed to switch.

(A) Switch point of agents (solid lines, *) and pheromone trail strength (dashed lines, **). The solid lines show the proportion of ants foraging at the first (blue) and second patch (red). The second patch was available after a 900 time step (15 minutes) delay, but allowed 3 times more agents to collect food simultaneously (8 vs. 24 agents). The dashed lines show the relative amounts of pheromone on the branches leading to the first (blue) and second patch (red). The pheromone switch happened some time after the switch in the number of foraging ants (average of 30 simulations). (B) Relationship between the pheromone decay rate and the time until more agents foraged at the second food patch. Note that the switch happens even with a pheromone decay rate of zero. A decay rate of 2.0 corresponds to a pheromone decay below the perception threshold of the agents in less than 10 minutes.