Figure 1.
Summary of model dynamics and anti-predator behavior.
A) Model dynamics. Solid boxes indicate model aspects that are implemented, namely the initial environment and pseudo-genes defining sensing and decision making of foragers. Dashed boxes indicate emergent processes on longer timescales: As foragers interact with local contexts in the environment, they generate behavioral patterns which in turn determine reproduction and survival and cause resource depletion. Reproduction and survival affect population growth and competition. If variation in pseudo-genes (due to mutations) results in differential reproduction and survival, natural selection emerges. B) Anti-predator behavior. Foragers start in a normal foraging state (top dashed box) and can decide to scan for a predator with probability pV. Only if a predator is detected, or a neighbor flees, does the forager flee, at which point it reaches a safe state (bottom dashed box). Here it continues to scan until it no longer detects the predator. After that, the forager continues scanning for another tP minutes, before it returns to a normal foraging state.
Table 1.
Evolvable parameters of foragers.
Figure 2.
A: An overview of evolutionary simulations reveals that larger groups evolve when predation risk dP is greater (blue dashed). Compared to simulations without evolution of grouping (grey dotted), vigilance evolves to smaller values (orange solid) if grouping evolves. Shown are means and standard deviation. B: The time course of mean group size (blue) and mean vigilance (orange) in a case-study for dP = 5 exhibits three phases: (I) a solitary phase; (II) a small group phase; and (III) a large group phase where group size and vigilance co-evolve: when group size increases vigilance declines, and vice versa.
Figure 3.
Lineages and behavior in phase II and III.
A: Lineages between year 40 (top left) and 600 (bottom right), with the transition from phase II to III (crossing thick vertical line, year 320). At the origin of grouping (circle at a), there are two co-existing lineages (red and orange) which differ in their repulsion zone (radius of orange circle F80 and L80, where the blue segment shows the repulsion turning angle aR, and the large arrow indicates the forager’s heading). This difference in repulsion zone leads to the leader-follower movement shown in B, which is needed to overcome the lack of coordination when two leaders (or followers) form groups (shown in C). Later, leaders and followers are differentiated in the repulsion turning angle (F180 and L180). Red lineages b, c, and d are independently evolved leader lineages which co-exist with followers. During the transition from phase II to III, followers become independent from leaders and bigger groups evolve (shown in D). From this point on, non-vigilant lineages (pV = 0, blue) repeatedly evolve and go extinct (small arrows e to i). B–D: Visualization of movement of leader-follower (B) and two followers (C) at year 70, and movement of group foragers at year 400 in phase III (D).
Figure 4.
Co-existence of vigilant and non-vigilant foragers.
A: Invasion simulation showing a selected non-vigilant genotype (orange) invading a population of a selected vigilant genotype (blue), after which both types co-exist. The non-vigilant genotype has a stronger grouping tendency and this enables it to coexist with the vigilant genotype. Co-existence ends when non-vigilant foragers are prevented from fleeing when vigilant foragers flee (after dashed line). Parameters for vigilant and non-vigilant foragers were taken from year 330 in the simulation of Fig. 2B. B: Average group size in vigilant (blue) and non-vigilant foragers (orange), using the same genotypes as in A, decline as population size declines. We used simulations in which population size and the ratio of vigilant and non-vigilant foragers was fixed. A greater proportion of non-vigilant foragers (dashed lines) leads to larger groups sizes since non-vigilant foragers have stronger grouping tendencies.
Figure 5.
Eco-evolutionary interactions of vigilance and group size in phase III.
Relatively small groups with high vigilance evolve after phase II, but this generates a niche for less vigilant foragers. Reduced vigilance and larger groups therefore evolve, which reinforces the social niche of less vigilant foragers. This can generate a within-group selection cycle (solid arrows). However, non-vigilant foragers in large groups forage inefficiently, leading to a reduced global competition for resources. As a result the group-level selection of ‘small group’ foragers with high vigilance becomes possible, and these can outcompete large groups with less vigilance. The population can therefore return to the situation with small groups and high vigilance (dashed arrows), where the niche for less vigilant foragers will re-emerge. Evolutionary changes are in blue and there ecological consequences are in orange.