Freshwater decapod (Aegla longirostri) uses a mixed assessment strategy to resolve contests

How animals decide to withdraw from a contest has puzzled researchers for years. Currently, four models try to explain how this decision is made: war of attrition (WOA); cumulative assessment (CAM); opponent-only assessment (OOA); and sequential mutual assessment (SAM). Although their predictions differ, they must be simultaneously tested to infer which model best describes contests. Herein, we identified the traits related to the resource-holding potential (RHP) in the decapod Aegla longirostri, and used these traits to test the predictions of each model. We identified which morphological/performance traits affect contest outcome, related these traits to the contest duration of male dyads, and tested the differences in aggressive acts (claw grabs). We tested the models using pairs of random and RHP-matched opponents. Additionally, we performed contests where RHP-matched opponents could communicate before the contest ('previewed'), and contests where the focal animal would only communicate with one individual and fight another unseen individual afterwards ('unseen'). In comparing these groups we tested whether information was being exchanged. The best predictor of contest outcome included a combination of cephalothorax length and claw height, and claw grabs increased with opponent similarity. Contest duration increased with loser's cephalothorax length and decreased with winner's cephalothorax length in random pairs, and winners spent more time in claw grab than losers. These findings refute WOA and OOA. In RHP-matched pairs, no relation was found and contests with previewed opponents were shorter than contests with unseen opponents, both results suggest SAM. However, the time spent in claw grab did not differ between previewed and unseen opponents. We argue that SAM is cognitively complex, and mutual assessment without comparison of RHP could be a better explanation. Furthermore, claw grab is important in contest resolution. Thus, the costs inflicted may suggest a mixed assessment strategy for A. longirostri's contests.