DAILY TEXAN (U of T, Austin, Texas) 15 June 06 Jumping and jiving with poison frogs - UT researchers studying toxicity, color patterns, diet find interesting links (Scott Solomon)
The forests of Central and South America are home to a group of perhaps the most elaborately colored animals on earth. Some have brilliant red backs with deep blue legs. Others are a nearly iridescent blue contrasting with irregular black splotches. Yet another is a smooth, pastel green interrupted by jet black spots. These are just some of the eye-popping color patterns found in a group known as the poison frogs.
As their name implies, these frogs are toxic, which is the reason they sport such incredible exteriors. Predators who attempt to eat one of these frogs will not soon forget the awful taste, or in some cases, severe illness caused by the poison.
A group of researchers at UT has been investigating the evolution of these frogs, known as Dendrobatids. They are now beginning to understand the origin of the poisons and how being toxic has affected their evolutionary history.
"For years, people that kept [poison frogs] as pets have known that after a while, they tended to lose their toxicity when they were in captivity. No one really understood why," said David Cannatella, an assistant professor of Integrative Biology at UT.
This loss of toxicity hinted that something in the frog's natural environment was responsible for the toxins. Experiments in which the frogs were fed purified extracts of their own poisons, known as alkaloids, suggested that something in the diet might be the source of the toxins.
"It really took a while for people to consider the possibility that, like a number of butterflies and caterpillars, they're actually sequestering the toxins from their diet," said Cannatella.
Poison frogs mostly eat insects. Careful examination of wild frogs' stomach contents has produced data on which insects in particular are eaten by different poison frog species.
A graduate student in Cannatella's lab discovered the source of the toxins. Juan Carlos Santos began studying poison frogs during his undergraduate studies at Pontificia Universidad La Católica in his native country of Ecuador. He became hooked on poison frogs after his first field experience with the charismatic amphibians in the tropical forests near his hometown.
"I just went ahead and tried to grab some of the leaflitter," he said of his first attempt at catching one of the elusive frogs.
"When I got the leaflitter, I felt something inside my hand. I was really excited," he said. "It was this wonderful frog: tiny, with dark and yellow blotches; extremely bright coloration."
Santos honed his frog-catching skills by traveling around the small South American country collecting as many poison frogs as he could. He also captured non-toxic relatives of the poison frogs, which tend to be more camouflaged. He later brought his collection to Cannatella's lab at UT where he extracted and compared the DNA of each species.
The results were surprising: poison frogs did not form one distinct group as previously believed. Instead, the data suggested that bright coloration, as well as toxicity, had evolved independently at least four times.
"Before Juan's work, it was generally assumed that the toxic species, which were also the brightly colored ones, formed one evolutionary lineage; they were all each others' closest relatives, because they had these complex alkaloid compounds. The assumption was, for these complex alkaloids to evolve more than once seemed sort of unlikely," said Cannatella.
Santos' genetic data had another interesting surprise: when he and Cannatella matched the phylogeny depicting the evolutionary relationships of each species with their known diets, there was a clear pattern.
"What you see is that the species that are toxic eat largely a diet of ants and maybe a couple of other groups of things including some mites. The species that are not toxic are fairly generalist," said Cannatella.
This result suggested that the toxic frogs, which have more specialized diets, are acquiring the poisonous alkaloids from the ants or other invertebrates in their diet. But not all the toxic frogs are equally poisonous, and who is to say what constitutes bright versus non-bright coloration? Another of Cannatella's graduate students, Cat Darst, set out to quantify these vague properties.
To measure the degree of toxicity of a frog species, Darst, who defended her Ph.D. dissertation in April, extracted the alkaloids from the frog's skin, purified them, and injected them into laboratory mice. The amount of time it took the mouse to recover from the toxin could be used as an indication of how potent the substance was. She then used a device called a spectrophotometer to measure the precise colors of each species.
Darst was able to show that, in general, toxic species tend to have brighter colors, as the researchers predicted. However, she found that the correlation was not perfect. That is, individual poison frog species may be very brightly colored and only moderately toxic, or vice versa. Darst wondered whether these strategies are equivalent when it comes to avoiding being eaten.
"The problem," Darst said, "is that no one knows what eats poison frogs."
As a first pass, Darst tried feeding the frogs to snakes. It seemed to work at first, but she was unable to collect enough snakes to conduct the experiment. Next she tried chickens, which she was able to purchase in high numbers from Ecuadorean villagers.
"I learned a lot about chickens," she said.
After discovering that adult chickens would not cooperate, especially fighting cocks, Darst settled on young chicks to perform the experiment. She first placed a toxic frog species in a clear glass dome in front of the chicks so that the chicks could observe it. She then lifted the dome to allow the chicks, which had presumably never encountered a poison frog before, to eat the toxic species.
"Most chickens would taste the frog and then spit it out," Darst said.
The next time Darst placed a poison frog of the same species under the glass dome, the chicks wanted nothing to do with it once the dome was removed. They had learned to avoid the frogs based on their first bad experience, and their avoidance was presumably based on the bright color patterns of the frogs.
Darst demonstrated that both strategies - having very high toxicity with moderately bright coloration, or very bright coloration with moderate toxicity - confer equal protection from the predator.
Darst also discovered that one of the brightly colored frog species, Epipedobates zaparo, was not toxic. This species, which resembles two poisonous species which it co-occurs with in the Amazon rainforest, was a mimic. E. zaparo had the bright coloration but not the toxicity to back it up.
Darst conducted further experiments with the chickens as surrogate predators, this time to determine how experience with a truly toxic frog affects a predator's avoidance of a mimic species. She first allowed the chicks to eat the poison frog, as before, but then she presented them with a perfectly edible mimic species. The chicks wouldn't touch the mimic. Apparently, the chicks were able to generalize their avoidance of frogs with a color pattern similar to the toxic species they had previously encountered. But how similar does the mimic have to be?
Based on her field observations in the Amazonian lowlands of eastern Ecuador, Darst knew that the mimic species occupied a relatively large range. But the model species - those that are toxic which the mimic resembles - have separate ranges. One species, Epipedobates bilinguis occurs in the north, and Epipedobates parvulus in the south. As expected, the mimic species always resembles the model with which it co-occurs, as is necessary to successfully deceive the predators.
However, Darst was able to locate a narrow zone of overlap between the two model species. Interestingly, in this zone where the mimic could resemble either of the two models, Darst found that all the mimic species looked like the model that was less toxic and less abundant.
This was exactly the opposite of what was expected. According to theory, a mimic species should always resemble the most common model species, since predators will only learn to avoid a color pattern if they frequently encounter it. If instead, a predator eats a mimic species, it will not associate the pattern with something that is poisonous.
Darst solved this enigma with the help of the chickens. It turned out that when a chick's first experience with a poison frog was with one that is very highly toxic, they would avoid anything that even marginally resembles the toxic species. In contrast, the chicks that were presented with only a moderately toxic frog learned to avoid only that particular pattern. Darst termed the phenomenon "toxicity-mediated avoidance generalization."
The same phenomenon might apply to someone who is allergic to peanuts.
"If they will kill you, you might not eat any nuts, but if it will just give you hives, you might just avoid peanuts," Darst said.
Her finding explained why the mimic species, E. zaparo, only resembles the less toxic model species when both models are present. Since predators that encounter the more toxic species will avoid any frog that looks remotely similar, E. zaparo need not resemble that species very closely. In contrast, by closely mimicking the less toxic species, E. zaparo guarantees that it will be ignored by predators that have encountered the toxic model.
Despite the team's advances in understanding poison frogs and their evolutionary history, there is still more to learn. Cannatella explained that they have noticed behavioral differences between the toxic and non-toxic frog species. The toxic species, for instance, tend to be more active than their non-toxic relatives. They also tend to have higher active metabolic rates. Could these two things be linked, and could they be a result of the frogs not needing to worry about predators?
"To be honest, it's not quite clear yet how all this stuff connects," said Cannatella.
http://media.www.dailytexanonline.com/media/storage/paper410/news/2006/06/15/LifeArts/Jumping.And.Jiving.With.Poison.Frogs-2087307.shtml?sourcedomain=www.dailytexanonline.com&MIIHost=media.collegepublisher.com
