They say that all's fair in love and war, and that certainly seems to be the case of Atlantic mollies (Poecilia mexicana). These freshwater fish are small and unassuming, but in their quest to find the best mates, they rely on and Machiavellian misdirection.
The males always prefer larger females but not if they are being watched. Under the voyeuristic gaze of a rival male, Atlantic mollies will either feign disinterest or direct their attentions toward a smaller, less attractive female.
Deception is par for the course in the animal kingdom. Plovers will try to lure predators away from nests by feigning broken wings and ravens will fool their rivals over the burial locations of food, as just two examples. But this may be the first instance of one male tricking another about which mate they fancy the most.
You have chosen... wisely
Martin Plath at the University of Potsdam discovered these surprisingly sophisticated strategies by giving male Atlantic mollies a choice between two females, who they shared a communal tank with. Without anyone watching, they always directed their first advances toward the larger female, nipping at her to get chemical cues, and thrusting at her for more obvious reasons. It seems that Atlantic mollies like 'em big. They also preferred a female of their own species to a similarly sized female from a related species - the Amazon molly (see footnote for more on this bizarre species).
That all changed when Plath added a second male into the mix. This rival was enclosed within a transparent cylinder - he couldn't be part of the action but his presence was nonetheless noticed by the other males. For a start, their apparent interest in mating dropped considerably and they displayed fewer nipping and thrusting actions.
But their preferences also shifted toward the smaller females or the Amazon mollies that they had previously shunned. Almost all of the males showed this turnaround. You might think that perhaps they were just cowed by the presence of a competitor into going for their second-best choice, but if that were the case, you would also expect them to change their minds only when being watched by a bigger fish. Instead, Plath found that the males switched their behaviour regardless of the size of their watchers.
He suggests that they are doing it to actively deceive the onlookers, directing attention away from their mate-of-choice and diluting their competition. Mollies compete aggressively for mates, both through direct physical confrontation and through indirect "sperm competition", where one male will mate after a rival does in order to flush his competing sperm out of the female. The male's deception provides him with relief from this conflict.
The tactic may be particularly effective in animals like mollies that are known to copy the mate choices of competitors. That in itself is a sound strategy - it allows individuals to make decisions based on information that others have attained - and the deception may be a counter-move.
Now, Plath wants to see how common this behaviour is in natural situations. If mollies could always deceive their rivals, then you would expect fewer and fewer fish to copy the behaviour of their peers, for individuals that didn't fall for the con would have free access to the best mates.
A tangent on the Amazon molly
It's worth sidetracking slightly to talk about the Amazon molly and their truly bizarre mating system. The females - and they are all females - need an injection of sperm for her eggs to develop, but unlike most other fish, she has no call for the DNA it contains. Her eggs are clones and already contain a full complement of genes; the sperm does nothing more than trigger the transition from egg to embryo.
To get this stimulation, the female Amazon mollies mate with males from several related species including the Atlantic molly. From the male's point of view, she is a sexual parasite; in her body, his sperm is wasted and contributes nothing to the next generation. This is why he, as expected, prefers a female of his own species.
Reference: Current Biology10.1016/j.cub.2008.06.067