Fig. 3

Evolutionary trajectories and alternative outcomes of evolution. Four replicate simulations illustrate that the co-evolution of individual and social learning can lead to markedly different outcomes. Evolutionary trajectories are coloured as indicated in the colour spectrum to the right: from dark blue in the initial generations to dark red towards the end of the simulation. The four trajectories all start with 20 learning episodes and a proportion of social learning of 0.5 (marked by a dark blue cross). After 20 K generations (dark red part of the trajectories), learning was lost in replicate (A), pure individual learning had evolved in replicate (B), and a learning strategy combining about 20% of social learning with 80% of individual learning had evolved in replicate (C). Replicate (D) illustrates that social learning, after having been lost initially, can later be regained, in this example resulting in a combination of 25% social learning and 75% individual learning. The legend to the right provides guidelines on how to interpret the simulation outcomes: ‘no learning’ if the evolved number of learning episodes is very low; ‘pure individual learning (IL)’ if the evolved proportion of social learning is very low; ‘pure social learning (SL)’ if the proportion of social learning is close to 1.0; and ‘combined individual and social learning (IL/SL)’ if the proportion of both types of learning is intermediate. The simulations were run for the same scenario as in Fig. 8, which will be discussed in detail later.