Résultat de la recherche
Collaborative hunting in the yellow saddle goatfish (Parupeneus cyclostomus)
Cooperation is of great interdisciplinary interest because we need to reconcile its occurrence with evolutionary theory and its emphasis on self-regarding individuals. Group hunting in various vertebrates has attracted much and continued attention from researchers because it provides the opportunity to study the evolution and stability of cooperation, the potential links between cooperation and cognitive abilities. Until now, all studies on coordinated hunting have been field observations and are hence correlative. Therefore, many conclusions are rather preliminary, like the repeatedly observed positive correlation between group size and hunting success as evidence for cooperation. In my PhD-thesis I conducted an unprecedented experimental study using yellow saddle goatfish (Parupeneus cyclostomus) as a study species. Yellow saddle goatfish are the first fish species described to be collaborative hunters. Individuals play different roles during a hunt (‘chasers’ and ‘blockers’), encircling prey hiding in coral crevices and trying to pry it out with means of inserting their barbels. I designed an experimental set-up in which the yellow saddle goatfish were confronted with a) mock prey that was pulled towards a shelter with multiple entrances; b) lively mobile prey hiding in an artificial coral reef. Cameras which were installed above and below allowed me to conduct detailed behavioural analyses i) to find out which decision rules underlie collaborative hunting in the yellow saddle goatfish; ii) to test the relationship between group size and hunting success; iii) to find out which payoff matrix conforms best to the hunting scenario when yellow saddle goatfish try to pry out the prey hidden in the shelter. The findings of my first chapter demonstrate that collaborative hunting in yellow saddle goatfish is based on simple, distance-based, self-serving decision rules. The individual that first detected the moving mock-prey always initiated a direct pursuit. Similarly, goatfish that were second to react in our experiments directly pursued the prey in almost all trials when they were in close proximity to the initiator. However, when lagging behind the initiator, the follower opted for a longer, less direct path to the prey. In the second chapter I showed that overall hunting efficiency (probability and speed of catching prey) is a function of group size. Larger groups of yellow saddle goatfish performed better and generally caught prey faster than smaller groups did. Groups of all sizes (2-4 individuals) were significantly more successful and faster than singleton hunters. Furthermore, I demonstrated that efficiency as singleton hunters did not predict hunting success when in a group. However, with my experimental set-up I could not address the question of how net calorie intake is affected by group size, as singleton success rate was already much higher than success rates observed in nature. The findings of my third chapter, where I investigated on how yellow saddle goatfish behave in order to pry out the hidden prey from the coral rock, demonstrate that actions were mainly maintained in order to obtain immediate benefits. Only the first barbel insertion decreased immediate benefits to the actor and created a public good resulting in payoff matrices similar to Snowdrift (2 players) and Volunteer’s Dilemma (N-players) games. However, further insertion effort did not lower capture probability, which even increased for the individuals that touched the prey first. Hence, besides being the first to insert, barbel-insertion can be considered as self-serving. Barbel insertion effort decreased from the singleton to the group hunting level and remained constant from 2-player to N-player situations, a finding which stands in contrast to theoretical predictions. These would expect a decline in the described situations, however under the assumption that the entire game is a Snowdrift/Volunteer’s Dilemma game. Interestingly, I could not find a correlation between insertion effort from singleton to group hunts, meaning that individuals seem to adjust their behaviour independently from whether they hunt alone or with others. In conclusion, the results of my PhD-thesis show that yellow saddle goatfish predominantly hunt self-servingly by following strategies which underlie simple decision rules. Overall, I propose that seemingly complex cooperation / collaboration can be based on simple rules. The challenge for researchers studying large-brained species is hence to test whether larger brains lead to more complex decision rules or whether collaborative hunting is generally a rather simple story.