Do lifetime contest costs affect the evolution of assessment strategies? A meta-analysis

In animal contests, the decision to flee or to remain in a contest is crucial to the individual's fitness because a wrong decision can dramatically change the costs individuals accrue over their lifetime. Empirical evaluations of the rules that individuals adopt to remain on a contest often provide support for two major possibilities: self-assessment strategies, in which each individual remain in the contest until reaching its own cost threshold, and a mutual-assessment strategy, in which individuals assess the rival's capacity to sustain or to impose costs during the contest and decide to flee when they consider that the rival is stronger. However, it is unclear what drives the evolution of different assessment strategies among species. A factor that may be key to understanding such variation is contest cost. If contests provide a fast and/or large cost accrual, mutual assessment strategies may be favoured because they avoid contests in which individuals always reach their own cost threshold. On the other hand, in species in which cost accrual during the contest is slow and/or small, self-assessment may prevail. Here, we performed a meta-analysis using information on the way individuals engage in contests in different species. Our goal is to test the hypothesis that contests involving high-cost accrual, such as contests in species in which rivals use weapons, can favour the evolution of mutual assessment strategies, while contests with lower costs, such as contests without physical contact, should favour the evolution of self-assessment strategies. We obtained 80 effect sizes spanning 36 species. Contrary to our hypothesis, we found that species with high-cost contests consistently adopted self-assessment strategies, while species with low-cost contests did not present a consistent assessment strategy. We suggest that high-cost contests may evolve in species in which individuals experiencing defeat or opting to flee from a contest have a significant decrease in their reproductive success. Consequently, individuals would be compelled to engage in contests regardless of their opponent. In low-cost contests, however, we suggest that any potential selective pressure for a specific assessment strategy is relaxed, which could explain the diversity of assessment strategies observed in this category.