Cheats never prosper – social policing maintains honesty in cooperative fish

What pre­vents anim­als from lying to each oth­er? Can cheats ever prosper? A study pub­lished in Evol­u­tion Let­ters has provided new insight into the evol­u­tion of hon­est sig­nals in cooper­at­ive Prin­cess of Bur­undi cich­lid fish, demon­strat­ing that the reli­ab­il­ity of com­mu­nic­a­tion between indi­vidu­als is main­tained by social poli­cing. We asked Hugo Gante, a research­er at the Uni­ver­sity of Basel and co-author of the paper, to tell us more about the feisty fish he stud­ies, and why these new find­ings are so interesting.

EL: Your new study invest­ig­ates how the Prin­cess of Bur­undi cich­lid fish uses the strik­ing col­our pat­terns on its face to com­mu­nic­ate its intent to fight with oth­er fish of the same spe­cies. Con­spicu­ous, col­our­ful pat­terns like these are more typ­ic­ally asso­ci­ated with mat­ing dis­plays and attract­ive­ness. Why did you think these facial pat­terns might be of par­tic­u­lar import­ance to how the fish sig­nal their aggressiveness?

HG: Stud­ies on con­spicu­ous col­our pat­terns that dif­fer between males and females, a con­di­tion known as sexu­al dichro­mat­ism, clearly dom­in­ate the lit­er­at­ure on col­our pat­tern evol­u­tion. In the case of sexu­al dif­fer­ences in col­or­a­tion, the col­our­ful sex is fre­quently the male, and research has shown that the need for males to dis­play and attract females has lead to the evol­u­tion of extra­vag­ant col­or­a­tion. Because females invest a sub­stan­tial amount of resources in egg pro­duc­tion, they want to make sure the male is of good enough qual­ity, and we now think females use male col­our­a­tion as a reli­able indic­at­or of their over­all con­di­tion. In oth­er words, males are com­mu­nic­at­ing their intrins­ic qual­ity to females using col­our signals.

Con­versely, in the par­tic­u­lar spe­cies we have stud­ied, males and females look abso­lutely alike, a con­di­tion that is known as sexu­al mono­chro­mat­ism. The reas­ons for the evol­u­tion of a col­our pat­tern that is the same in females and males must be dif­fer­ent from those when the sexes are dif­fer­ent. In par­tic­u­lar, it sug­gests that both sexes (not just one) are com­mu­nic­at­ing some inform­a­tion, and that inform­a­tion must be sim­il­ar in con­tent. We know from their bio­logy that both male and female Prin­cess of Bur­undi cich­lids are very ter­rit­ori­al and both invest in rais­ing their young. In fact, they live in fam­ily groups that togeth­er defend their ter­rit­or­ies in the rocky shores of Lake Tan­ga­nyika against intruders. There­fore, com­mu­nic­a­tion dur­ing aggress­ive inter­ac­tions was an obvi­ous con­text for us to test a func­tion of this facial col­our pat­tern, and it turned out to be the case.

EL: You found that a fish’s face pat­tern changes depend­ing on wheth­er they win or lose a fight. How does this change hap­pen and what does it tell us about how these fish sig­nal their inten­tions to those around them?

HG: This facial mask is com­posed of sev­er­al more or less dis­crete ele­ments – blue, yel­low, black and white patches. We decided to indi­vidu­ally char­ac­ter­ise each one of them, and we found that their phys­ic­al prop­er­ties and arrange­ment make this pat­tern very con­spicu­ous to oth­er fish. It is like the col­our­ful masks used by humans in the Mex­ic­an free wrest­ling Lucha Libre.

In the case of the fish, besides the high con­spicu­ous­ness of the facial pat­tern, we found that the intens­ity of the black hori­zont­al stripe depends on wheth­er a fish switches to a dom­in­ant or sub­or­din­ate beha­viour, while all the oth­er col­our patches remain rel­at­ively stable. It is import­ant that such col­our change hap­pens fast, if you want to sig­nal to your oppon­ent that you’re no longer a threat before things get to a point of no return. It’s like wav­ing the white flag. This fast change hap­pens in a few seconds and is medi­ated by physiolo­gic­al col­our changes with­in the black pig­ment cells, called melan­o­phores. Essen­tially the black pig­ment, melan­in, is trans­por­ted with­in the melan­o­phore to cre­ate a dark or pale aspect, depend­ing on wheth­er it is dis­persed or aggreg­ated, respect­ively. In the heat of the fight you can lit­er­ally see the hori­zont­al facial stripe change its intens­ity before your eyes, such that the black stripe fades and becomes light grey. In the same way the hori­zont­al stripe pales, this state can be reversed and the stripe can become dark­er again in a mat­ter of a few seconds. It is a very dynam­ic pro­cess. The fact that intens­ity of the stripe is so labile already hints at how it func­tions as a sig­nal of motiv­a­tion or aggress­ive state.

EL: Does the ‘baseline’ (e.g. pre-fight) intens­ity of the facial stripe vary across dif­fer­ent fish, and is this linked to any oth­er phys­ic­al characteristics?

HG: Very small indi­vidu­als do not have a facial mask. It even­tu­ally devel­ops when the fish reach about 2–3 cm in length and once they have it, they keep it. There is always some level of indi­vidu­al vari­ation asso­ci­ated with any trait, but the baseline intens­ity of the facial stripe tends to be dark­er, rather than paler. When a fight starts they can sig­nal moment­ar­ily to an oppon­ent until peace is re-estab­lished. This said, we don’t know yet how aggres­sion and dom­in­ance is trans­lated into col­our change. That is one of our next ques­tions, but it is pos­sible that hor­mones medi­ate this inform­a­tion transfer.

EL: You also con­duc­ted an exper­i­ment where you arti­fi­cially altered the intens­ity of the facial stripe on some fish. How did you do this and what more did it reveal about the func­tion of the stripe as a sig­nal to oth­er fish?

HG: In this study we wanted to more gen­er­ally address how com­mu­nic­a­tion can be hon­est. Why would anim­als provide hon­est inform­a­tion about their qual­ity, con­di­tion, and inten­tions, if they could sig­nal unre­li­ably and profit from it? This ques­tion has puzzled research­ers for dec­ades and still does to some extent. The­or­eti­cians seem to have reached an agree­ment that costs are respons­ible for main­tain­ing hon­est com­mu­nic­a­tion, but the exact source and type of costs still remain elu­sive. While empir­i­cists have tried to gath­er exper­i­ment­al evid­ence to sup­port dif­fer­ent altern­at­ives, it is also unclear wheth­er costs have been really measured.

We tried to address this issue by provid­ing a frame­work to guide empir­i­cists study­ing col­our sig­nals. First we char­ac­ter­ise the col­our pat­tern, then determ­ine its func­tion, and finally identi­fy the source of the costs of hon­esty. To under­stand what kind of costs could be guar­an­tee­ing com­mu­nic­a­tion reli­ab­il­ity, it is essen­tial to manip­u­late the sig­nal of indi­vidu­als that could gain from provid­ing inac­cur­ate inform­a­tion. So essen­tially, we force indi­vidu­als to lie by send­ing an inac­cur­ate sig­nal, and ask wheth­er receiv­ers of that sig­nal can detect it. We expect that if reli­ab­il­ity is main­tained through viab­il­ity and fecund­ity, as would likely be the case for males dis­play­ing their col­ours to attract females, this hap­pens too far into the future for liar detec­tion to evolve. Costs of hon­esty must there­fore have a physiolo­gic­al basis. Con­versely, liar detec­tion is pre­dicted to evolve for sig­nals of labile inform­a­tion, such as motiv­a­tion, in which case social poli­cing can act instant­an­eously and impose costs on liars.

We manip­u­lated the hori­zont­al stripe and intro­duced the fish into a new tank with a mir­ror, where they saw their reflec­tion as if it is the own­er of a ter­rit­ory. We either arti­fi­cially darkened the stripe by apply­ing water­proof liquid eye­liner, or we arti­fi­cially paled the stripe with a powder nor­mally used to stop wounds in fish. In this way, we cre­ated bluffers that sig­nalled high aggres­sion and fight­ing motiv­a­tion, or Tro­jans (in ref­er­ence to the Tro­jan Horse) that would pre­tend they’re not a threat and use that as a sub­ter­fuge to strike without notice.

What we found is that both types of manip­u­lated fish received more aggres­sion, and faster, than con­trols, indic­at­ing that Prin­cess cich­lids can identi­fy liars and pun­ish them. Our find­ings sup­port the hypo­thes­is that social selec­tion through receiv­er retali­ation main­tains reli­able sig­nalling of a labile state. This is import­ant because it was not clear wheth­er sig­nals that are physiolo­gic­ally cheap to pro­duce could be hon­est, and we have shown that social poli­cing is an effect­ive way of guar­an­tee­ing that.

EL: How do we know that these fish see sig­nals such as col­our and pat­tern in the same way we do?

HG: For all we know, Prin­cess cich­lids don’t see col­ours as we do, but it is fas­cin­at­ing, if not tricky to ima­gine how they see them. Let’s start with human col­our vis­ion. In our ret­ina there are cells respons­ible for col­our vis­ion (the cones), which have three molecules (the opsins) that are most sens­it­ive to three dif­fer­ent col­ours (blue, green and red). Actu­ally they are sens­it­ive to dif­fer­ent wavelengths of light and the brain even­tu­ally trans­lates that into col­ours. People with red-green col­our blind­ness have dif­fi­culty dis­tin­guish­ing between reds, greens, browns and oranges because of shif­ted opsin sens­it­iv­ity (http://www.colourblindawareness.org/). So the same wavelengths are seen dif­fer­ently by dif­fer­ent people depend­ing on their opsin set. Put­ting it in dif­fer­ent terms, col­ours are an indi­vidu­al per­cep­tion, a brain con­struct, and each one of us sees them slightly dif­fer­ently (in gen­er­al, men tend to do poorer in col­our dis­crim­in­a­tion than women).

What we know about col­our vis­ion in these fish is that they are most likely tri­chro­mats like us, the exquis­ite aspect being that they are sens­it­ive to ultra­vi­olet and not very sens­it­ive to reds. There­fore, to be able to say some­thing about how the fish see col­ours we mod­elled their vis­ion tak­ing into account sev­er­al things like cone types, opsin sens­it­iv­it­ies, ambi­ent light under­wa­ter and so forth. For instance, if you’re a diver, you know that reds become grey­ish the deep­er you go because this light wavelength becomes atten­u­ated under­wa­ter. Even if you don’t dive you must have exper­i­enced the effect that arti­fi­cial light has on col­ours and how we some­times have dif­fi­cult telling which col­our it is. We could determ­ine that the facial pat­tern we stud­ied is rather con­spicu­ous to the fish, pos­sibly like it is to us, even if they per­ceive the col­ours some­what dif­fer­ently. Using this visu­al mod­el­ling approach we also determ­ined that Prin­cess of Bur­undi cich­lids per­ceive changes in the intens­ity of the hori­zont­al stripe, which we inde­pend­ently con­firmed in our manipulations.

EL: How aggress­ive are these fish under nor­mal cir­cum­stances? Do they fight fre­quently in the wild and how import­ant is this to oth­er aspects of their life?

HG: For a fish named Prin­cess they are quite feisty. Any­one who has kept them as pets would con­firm they can take over a large aquar­i­um for them­selves. They will not hes­it­ate to attack an intruder to their ter­rit­ory, espe­cially con­spe­cif­ics. The lar­ger fish in a fam­ily will also chase away poten­tial pred­at­ors to pro­tect its young­er mem­bers, as long as the pred­at­or is not so large as to endanger them. Unfor­tu­nately we do not know how fre­quently they openly fight in nature. We do know they are able to indi­vidu­ally recog­nise fam­ily mem­bers and even neigh­bours from adja­cent ter­rit­or­ies, who are tol­er­ated if they do not pose a threat. There­fore, aggres­sion is not indis­crim­in­ate but it can escal­ate very fast and hence the sig­nalling of intent to fight can be import­ant in pre­vent­ing unne­ces­sary injury.

EL: A key find­ing of your study is that these cich­lids can detect and pun­ish liars in the pop­u­la­tion, almost like their own ver­sion of a neigh­bour­hood watch! How effect­ive is this poli­cing likely to be – is it ever pos­sible for cheats to prosper?

HG: This is per­haps the single most import­ant piece of inform­a­tion we found in our study. It means that even sig­nals that are cheap to pro­duce and carry little or no physiolo­gic­al costs can be hon­est as long as there is a poten­tial social cost to liars. This does not mean that sig­nalling has to be reli­able at all times, but it could be rather dan­ger­ous for a fish to be caught lying. From our data it seems that bluffers would have a harder time to get by than Tro­jans, which is con­cord­ant with pre­vi­ous math­em­at­ic­al sim­u­la­tions. Hence we can­not exclude that this sig­nalling sys­tem can be replaced by a non-sig­nalling sys­tem driv­en by Tro­jans. We have yet to test the con­sequences of lying when there are by-stand­ers, but if I was to spec­u­late, I would say that this would only make social selec­tion a more effi­cient mech­an­ism in main­tain­ing sig­nalling hon­esty. Cich­lids are known to be extremely good at extract­ing social inform­a­tion from oth­er indi­vidu­als they inter­act with. When you con­sider that these fish live in neigh­bour­hoods, pop­u­lated by both fam­ily mem­bers and adja­cent fam­il­ies, being caught lying and get­ting their repu­ta­tion des­troyed could have dire consequences.

To find out more, read the Hugo’s study, now freely avail­able to read and down­load from Evol­u­tion Let­ters: http://onlinelibrary.wiley.com/doi/10.1002/evl3.24/full