Fig. 4

Competition experiments show that selective advantage of a regulatory strategy depends on presence of adhesion. Setup of competition experiment: two groups (with differing regulatory networks or adhesion values) are made up of 16 cells each, and are placed adjacent to each other. The simulation is then run for multiple seasons (without mutations), until one of the groups has gone extinct. In the snapshots, cells in one group are coloured green and in the other purple (lighter shade when dividing). A-D Snapshots show the first season of the competition simulation. Graph plots group size against median distance of cells to the peak for the first season (\(0\%\)=maximum distance, \(100\%\)=at the peak). Light-to-dark colour gradient of the line indicates time in the season, grey lines connect equal time points between the two groups. Replicates can be found in [Additional file 17]. A Competition between adhering (\(\gamma =6\), green) and non-adhering (\(\gamma =-4\), purple) cells, either both with a division-late strategy (left) or division-early strategy (right). B Competition between ancestral, division-early strategy (evolved with adhesion before the switch to non-adhering conditions; green in snapshots) and evolved, division-late strategy (purple in snapshots); both non-adhering. C Two seasons of the competition experiment between two adhering groups, one with an ancestral, division-late strategy (evolved without adhesion before the switch to a simulation in which adhesion could evolve; green in snapshots) and one with a strategy evolved after 500 seasons with adhesion (having become more division-early; purple in snapshots). D Group size vs. median distance from peak over the course of one season, for the competition experiment in C) E Time dynamics of the competition in C. Top plot shows group sizes, bottom plot shows distance to the peak. Shading indicates \(25^{\text {th}}\) and \(75^{\text {th}}\) percentile of population