Why does this frog come in two different colours?


Study uncovers conditions where diversity flourishes against the odds. Tanya Loos reports.


The dyeing poison frog (Dendrobates tinctorius) of French Guiana has yellow stripes to keep predators away, but some have white stripes instead.

University of Jyvaskyla

A toxic frog with two colour forms is helping researchers unpack the diversity and evolution of anti-predator prey signalling.

Warning colours are used by prey throughout the animal kingdom to warn potential predators that they are dangerous (as in the black-and-yellow stripes of wasps) or toxic to consume (as observed in the diverse array of bright colours and patterns in the poison arrow frog family).

The dyeing poison arrow frog (Dendrobates tinctorius) is toxic to predators, such as birds, due to the high alkaloid content in its skin – and yet the population found in French Guiana has two colour forms or phenotypes: yellow stripes on a black background, or white stripes on black.

This diversity of colour signals goes against the accepted theory that warning signal colouration should be subject to strong, frequency-dependent selection.

In Amazonian butterflies, for example, it has been demonstrated that the fitness of a phenotype increases with its frequency – the more of these colour forms are around, the more chance a predator population learns to understand the signal and avoid the prey. Warning signals that are novel or unusual should be selected against due to their rarity.

And yet these two colour forms persist.

The evolution of warning signals is well studied in mimicry complexes: in Müllerian mimicry, such as in cyanide-ridden, black-and-yellow striped millipedes; and in Batesian mimicry, whereby harmless species avoid predators by imitating dangerous or toxic species as in some butterflies.

Now the Dendrobates tinctorius presents researchers with an opportunity to examine the origin and persistence of colour form variability in a population without mimicry at work.

A team led by JP Lawrence from the University of Mississippi, US, and Bibiana Rojas from the University of Jyvaskylä, Finland, undertook a series of field and laboratory experiments to investigate how two types of poison frog co-exist when we expect only one.

Plasticine clay models of frogs painted variously in white and yellow stripes were placed in the frogs’ home range and assessed for predator responses.

The likely drivers of anti-predator signals in poison arrow frogs are birds, so domestic chickens (Gallus gallus domesticus) and blue tits (Cyanistes caerulus) were used.

Finally, gene flow between populations and alkaloid content of the frogs’ skin were assessed.

“We sought to explore patterns of natural selection by avian predators and infer the conditions necessary to sustain the evolution of novel signals,” the researchers write in a paper in the journal Proceedings of the National Academy of Sciences.

Two mechanisms were uncovered by which a diversity of colour forms might persist in these frogs.

Learning experiments and unpalatability tests show that predators learn to avoid the stronger signals of the more distasteful yellow-striped frogs and generalise this avoidance to white-striped frogs.

In addition, although the stronger defences of yellow frogs should suggest that they will out-compete white frogs in time, genetic analysis revealed virtually no migration or gene flow between the two populations.

“We propose that signals that are easily learned and broadly generalised can protect rare, novel signals, and weak signals can persist when gene flow among populations is limited,” the researchers conclude, adding that this sheds further light on drivers of speciation.

Explore #frogs #evolution
  1. https://amphibiaweb.org/cgi/amphib_query?where-genus=Dendrobates&where-species=tinctorius
  2. https://doi.org/10.1073/pnas.1519216113
  3. https://cosmosmagazine.com/biology/lovely-but-lethal-millipede-discovered
  4. https://cosmosmagazine.com/biology/this-butterfly-is-not-what-it-seems-doubly-so
  5. http://www.pnas.org/cgi/doi/10.1073/pnas.1901872116
Latest Stories
MoreMore Articles