Game theory is a mathematical framework used in economics to understand how  a contestant’s strategy in a competition over resources is determined by rational decisions regarding the strategy of opponents. Over the last century since game theory was first developed in mathematical language, it has been adapted for evolutionary and animal behavior contexts, and tested across a number of animal species with different types of competitive behaviors. The consensus has historically been that weaker animals tend to lose and that the timing of their decision to give up relies on how motivated they are to attain the resources they are fighting for. This highlights how winning and losing relies on both the perceived value of the resource and the skills of the contestants.

However, simplifying the model to” losers quit” neglects the fact that animals may adapt their fighting behavior and strategically gain advantages against the odds. For example, some animals have more aggressive personalities, independently of their opponent, while others  may use aggression strategically and exaggerate their strength to ward off opponents that are stronger than them. In order to better understand these competitive animal behaviors, we studied male Siamese fighting fish and found that they display both of these unexpected strategic behaviors: small males more readily engage relatively bigger opponents and flair their fins to overstate their size.

It is likely that the behavior of these fish relies on a combination of motivated aggression and calculated strategy.

Aggression towards bigger opponents relies on the fish’s motivation to defend its territory from greater threat. Therefore, fish show less aggressive behavior when their territory is negatively affected by acoustic noise, such as that imposed on aquatic environments by light human traffic. However, male Siamese fighting fish become more aggressive once they have built bubble-nests. Males build these nests to signal their preparedness for mating and to deposit fertilized eggs into them after mating. This increases their young’s survival, is attractive to the females and therefore highly beneficial to the males. To defend territory, especially territory with bubble-nests, males use elaborate displays to ward off intruders by displaying their size and aggressiveness, including lateral displays where they present their extended fins. Interestingly, the display of fins is used to exaggerate their strength by overstating their size, but this tactic can be recognized by fish during a fight. As a result, contestants progressively disregard the size-information from fin displays as the fight develops. This shows that male Siamese fighting fish can adjust their assessment of their opponent’s abilities and take them into account when fighting. Therefore, it is likely that the behavior of these fish relies on a combination of motivated aggression and calculated strategy.

In our recent study published in BMC Zoology we decided to test this hypothesis by characterizing the modulators of fighting behavior in Siamese fighting fish in detail. First, we measured each individual’s aggression level against its mirror image on two separate instances to spot any baseline personality differences. We used this information to see whether the level of aggressiveness could predict the fish’s behavior during fights against other fish. Then, we examined each animal’s behavior during fights and how they were affected by their opponent’s size and behavior when compared to their own. In order to test the fish’s motivation to fight, we interrupted the fight at 5 and again at 10 minutes into the interaction and measured how fast fish re-engaged opponents.

Using a composite score of consistent aggressiveness measured from the repeated mirror fights, we found that individual personalities only affected early stages of the live fights but not decisions to attack opponents or retreat. As expected, smaller animals engaged bigger opponents faster and more aggressively, but could become submissive when ongoing attacks increased the risk of injury. In addition, fish used new information from ongoing behavior as well as their energy levels to decide whether to continue fighting, and whether to escalate or retreat, respectively. Finally, the motivation to defend resources, drives behaviors meant to achieve this in the least costly manner, like signalling their size and strength or submitting to attack. On the contrary, behaviors with potentially dangerous consequences- such as engaging opponents and attacking them – depended on informed decisions like the relative size or behavior of the opponent and the territory’s value.

Decisions rely on personality differences, resource value and are adjustable to changes in own and opponent fighting performance during the contest.

These strategies show that male Siamese fighting fish can make calculated decisions. However, contrary to most models used in economics, where strategic decisions are considered dependent largely on fixed assessments of opponents, in Siamese fighting fish we find that decisions rely on personality differences, resource value and are adjustable to changes in own and opponent fighting performance during the contest. The evaluation of resources follows long-standing expectations by evolutionary applications of game-theory, while the composite effects of personality and the use of flexible assessments are a developing field. We hope that similar approaches can help further develop accurate models of contest theory and that their future application in other help elucidate both archetypal mechanisms and phenotypic variation across the evolutionary scale.