A frog pollinating a flower? Not so fast!

May 10, 2023 • 11:30 am

A frog pollinating a flower? That would be remarkable, and a paper supposedly describing the phenomenon was recently published. It got a lot of attention, including a large piece in (guess where?) Scientific American.  I was prepared to write about it based on the publicity, but I needed to see the paper first. When I read it, I was sorely disappointed. The evidence for frog pollination, which would be the very first described case of an amphibian effecting pollination—was as thin as a piece of paper.

The relevant paper, from the journal Food Webs, is not widely available, even through the University of Chicago library. Fortunately, a kind reader somehow got hold of the pdf (you can too, through judicious inquiry), and just a bit of it is online, which you can see by clicking the link below.

I’ll be brief (well, I’ll try). This group of investigators from Brazil report that the tropical treefrog Xenohyla truncata was observed eating fruit, petals of flowers, and sipping nectar during one four-hour observation period.  This itself is unusual because frogs are mostly carnivorous and insectivorous. This species (and one congeneric relative) were reported earlier to be omnivorous, eating both fruits and invertebrates.

The popularized result? One (count it, one) frog was found with pollen on its back after sticking itself into a flower to eat. Was it observed pollinating another flower? No. We don’t even know if the plant is self-compatible, so that a frog could even effect cross-pollination on the same plant. Did the frog visit more than one plant, so that cross pollination was possible? No.

The upshot s that all the publicity given to this frog is comes from the observation of a single individual exiting a flower with pollen on its back. That says virtually nothing about whether it is a pollinator, and even less about whether it’s an important pollinator.

Here are the data described in the paper:

We conducted in situ observations of a breeding population of X. truncata on 15 December 2020, in a Restinga vegetation area in the municipality of Búzios, state of Rio de Janeiro, southeastern Brazil (224613.94”S, 41574.47” W; WGS84; 2 m a.s.l.), for approximately four hours (from 6:00 to 10:00 pm). Air temperature was 25.8 C. We observed five individuals of X. truncata in feeding activity on two plant species between 7:00 and 9:00 pm.

. . .Around 8:00 pm we observed other X. truncata individuals leaving bromeliads and climbing a Brazilian milk fruit tree [JAC: Cordia taguahyensis] full of fruits and flowers. Three individuals (sex undetermined) clustered around a ripe fruit and began a dispute to get close to the fruit, pushing each other away as they tried to bite the fruit (Fig. 1D; Video S2). After approximately five min, two individuals gave in and remained perched on branches close to the fruit, while the third began to nibble the fruit, increasing a pre-existing hole to gain access to the pulp (Fig. 1D). While this individual fed on the fruit, the others no longer disturbed it. The same individual remained nibbling and sucking the fruit pulp for about 10 min, the others eventually approaching to feed. x

. . . On the other side of the same tree, we observed a X. truncata individual that climbed a branch and entered an open flower (Fig. 1E), where it remained for approximately 5 min performing suction-like movements (Video S3). Upon leaving, pollen grains were adhered to its back (Fig. 1F).

This is the first report of a frog species actively feeding on nectar and flowers in nature and the first evidence that it may act as pollinator.

Here are the two pictures mentioned above, along with their captions from the paper:

X. truncata within a Cordia taguahyensis flower (E)


. . . and coming out of it with pollen grains (red arrow) on the back (F). Photographs by C. H. de-Oliveira-Nogueira.

The frog was also observed feeding on an “alien” (THEIR WORD) species, the beareded Iris, Iris x germanica (see video below; the “x” indicates the species is of hybrid origin).

So what’s the evidence that this frog actually effected any pollination? None that I can see.  What is the evidence that it’s a regular and important pollinator of the native milk fruit tree? None, nada, zippo.  Now it is possible that this frog could still pollinate other flowers on the same tree (if it’s self-compatible) or on other trees, but we don’t have that evidence. What we have is the authors’ speculation, eagerly and breathlessly snapped up and regurgitated by the press.  More from the paper (my emphasis)

As mentioned, the relationship between X. truncata and the native C. taguahyensis is remarkable. The flower structure of C. taguahyensis allows X. truncata to enter and exit the flower, and to carry pollen grains after the visit. In this case, X. truncata could act as a pollinator of this species, or even of other plant species with similar floral structure. However, to play the pollinator role of C. taguahyensis, this frog should visit another flower or another plant individual on the same night. We lack information about the breeding system of C. taguahyensis, but some Cordia species are self-compatible, whereas others are self-incompatible (Opler et al., 1975; Machado and Loiola, 2000; Mcmullen, 2011; Wang et al., 2020). As X. truncata wanders from one plant to another before it settles in a bromeliad for daytime shelter (our pers. obs.), it is likely that the above mentioned scenario about its pollinator role actually occurs. 

How likely is it? We don’t know.

Species in the genus Cordia are visited by a wide variety of invertebrates, such as bees, butterflies, beetles, wasps and flies (Opler et al., 1975; Machado and Loiola, 2000; Lopes et al., 2015), as well as vertebrates such as bats (Alvarez and Quintero, 1970) and birds (Opler et al., 1975; Dalsgaard, 2011; Wang et al., 2020). Thus, C. taguahyensis is likely pollinated by multiple animal species and the treefrog X. truncata is now a potential pollinator candidate.

“May”, “might”, “seems likely”, and so on it goes. To show pollination, you need to show pollination, not speculate that it’s likely. One way to do this is dust a flower with fluorescent pigment, like those used in making black-light posters, and then see if any fluorescent pollen makes its way to another flower (and, of course, that the flowers are reproductively compatible). This wasn’t done (we used this method to mark Drosophila in the wild.) Ergo, while we have new evidence that X. truncata is omnivorous and eats petals and nectar, and that pollen adheres to its back, that’s all we have.  It’s possible that a cross pollination occurs occasionally, but even that is speculation, and doesn’t show that the frog, as opposed to the many insects that visit the plant, is of any importance as a pollinator. As the authors say, “the treefrog X. truncata is now a potential pollinator candidate.”

Well, at least there’s a video of the cute little frog eating from a flower (but the “alien” plant, not the milkfruit flower); this comes from IFL Science. Cute little bugger, isn’t it?  The nectar and plant material may be a valuable supplement to the diet of this frog.

The journals that mentioned this paper as a possible case of amphibian pollination include Science, the New York Times, IFL Science, and other places. But the first place that comes up when you google “pollinating frog” is Scientific American.

Sofia Quaglia at the NYT gives the most critical take on this paper, saying this:

But more observations are needed to say the frogs really are pollinating plants.

“We cannot say that these frogs are actually pollinators,” said Felipe Amorim, a pollination ecologist at São Paulo State University who was not involved in the research. “They are flower visitors, they are flower-visitor frogs. We have a lot to learn about this novel interaction.”

For instance, the mucus secreted by the frog’s skin needs to be tested to confirm it doesn’t spoil the pollen before it gets to its destination. Scientists also need to work out whether the pollen is ever delivered to other flowers and if it does successfully fertilize and germinate them. It’s also still unclear why this frog species has developed a liking for flora over fauna in the first place.

And at least Sci. Am. mentions some of these problems. But really, the publicity given this observation, which has an interesting part (some frogs eat flowers and nectar) and a not-so-interesting part (one frog got pollen on its back) far exceeds its scientific novelty. This is what happen when either a university publicity machine goes into action or journalists copy each other’s content.  Two colleagues to whom I sent the paper both found the claims of possible pollination by the press (and by the authors, too) wildly exaggerated.

As Kurt Vonnegut said, “So it goes.”

13 thoughts on “A frog pollinating a flower? Not so fast!

  1. Thanks for checking that out. The claim (pollinating frogs) seemed based on very flimsy evidence, but all I had to go on was what was in the press and very little that can be read online is believable.

  2. Yes, that’s my assessment too. I suspect this is mostly the fault of the university or journal PR people, but the authors also failed to put their observations in perspective. Among the wide range of known pollinators for this flower, the frog would make a vanishingly small contribution to its pollination, even if the number is not exactly zero.

    Practically ANYTHING can be an accidental pollinator of any flower

    1. Having said that, the news that there is a fruit- and flower-eating frog is very interesting and does raise the possibility that a frog may be an important pollinator of some species of plant somewhere in the world.

      1. I was also going to say something along these lines. I think the authors were right to point out this possibility since it is indeed interesting, but they should have been much more open with all the caveats of this single observation.

  3. These conclusions would not have been given a passing grade in an undergraduate Biology course. How such a paper could be published is beyond my comprehension.

  4. Thanks PCC(E) and everyone – I suppose this still means there are good reasons to “F’ing love” science – even more so – the peer review, the “Nullius in Verba” of it, is how it works.

    How about :

    I Nullius in Verba Love Science

    Not sure….

  5. Brazilian indigenous people know a lot about frogs (poison arrows etc.). If they don’t have indigenous knowledge of this frog as a pollinator then I think it’s not a pollinator. At least those are the rules eh?

  6. Thanks indeed to PCC(E) for digging in on this. A very nice dissertation on the operation of the scientific method.

  7. I must admit that I imagined a frog that had evolved to be dependent on these flowers and the flower to be dependent on the frog. The reality did not live up to the billing.

  8. in a Restinga vegetation area in the municipality of Búzios,

    I’m not sure what a “restinga” area is precisely, but it sounds like some sort of relic of the original landscape and vegetation surrounded by recent city expansion.

    That isn’t a recipe for the wildlife and vegetation to display natural behaviour. As indeed the fact that the frog is seen interacting with an “alien” (i.e. human introduced) plant species.
    Totally off topic, I wonder if any of the CSS options for blockquote work (i.e., survive WP’s input sanitisation)? “Cite”? (survives input, but has no effect) “style”? (does not survive sanitisation)

    1. That is a good point, human-altered habitats are no good places to try to understand the evolutionary forces underlying observed behaviors. But restingas are special native vegetation formations in southeastern South America, on sandy or acidic nutrient-poor soils. These are valid stable natural ecosystems.

Leave a Reply