An Invasive Frog Increases And Decreases Non-Native Mammals In Hawaii
Can a little frog make a difference? This is the question we asked ourselves when a small frog (~3 cm) from one area of the world invaded another. Back in the late 1980s, the coqui frog (Eleutherodactylus coqui) endemic to Puerto Rico was accidentally introduced to Hawaii via nursery plants. People noticed the frog right away because it has a loud mating call, that sounds very much like its common name, coqui (ko-kee). But, ecologists knew that there might be more at stake than dealing with the loud noise.
Back in its native Puerto Rico, the coqui has really high densities (around 20,000 frogs/ha). It was hypothesized that it might have even higher densities in Hawaii (which it does – up to 90,000 frogs/ha) and therefore could threaten endemic species. One of the biggest concerns was that the coqui would interact with and potentially negatively affect Hawaii’s endemic birds. One thought was that coquis might increase non-native mammals, like mongoose (Herpestes auropunctatus) and rats (Rattus), and because these species are known bird nest predators, that this would in turn increase bird nest depredation rates and thus threaten native birds.
In our recent study published in Biological Invasions, we tested these hypotheses across 12 sites on the Island of Hawaii. At each of our study sites, we compared areas where coqui frogs had invaded and neighboring areas that were not yet invaded but had very similar habitats. In these areas, we measured mongoose and rat densities, as well as bird nest depredations rates. We also used camera traps to determine the dominant scavengers of coquis and dominant nest predators.
We found that across the island of Hawaii, coquis were associated with 30% higher mongoose densities – probably because mongoose like to eat coquis, as we observed with our camera traps (Figure 1). Interestingly, coquis were also associated with lower rat densities, particularly the more non-arboreal Pacific rat (R. exulans) that was 17% lower in areas invaded by coquis across the island. We found evidence that Pacific rats might be lower where coquis invade because they compete with coquis for insects, and also perhaps because mongoose densities are higher in these areas and mongoose also consume rats.
But, we did not find any difference in bird nest depredation rates in areas with and without coquis. This might be because both mongoose and rats are bird nest predators, and while mongoose increase with coqui invasions, at the same time, rats decrease. So, these two effects likely cancel each other out, resulting in bird nests not being more threatened where coquis invade.
In summary, invasive frogs do seem like they can influence non-native mammal abundances, but perhaps not bird nest depredation. It appears that this little frog can increase the abundance of an invasive mammal, the mongoose, which many in Hawaii view as a pest. The invasion of coquis might also facilitate the spread of mongoose to areas where they do not currently exist. But, it appears that coquis may have the opposite effect on rats.
What is perhaps most interesting from an ecological perspective is that coquis seem to have inserted themselves into a novel, non-native dominated food web, where they consume and are consumed by other non-native species. The real concern may come if these frogs invade higher elevation forests. It is in these areas, where most of the endemic and rare Hawaiian species remain. So, it is these areas, where we might want to prevent their spread and potential impacts in the future because, as we now know, they can make a difference.
These findings are described in the article entitled Invasive coqui frogs are associated with differences in small Indian mongoose and rat abundances and diets in Hawaii, recently published in the journal Biological Invasions.
- Hill, S.A., K.H. Beard, S.R. Siers, and A.B. Shiels. 2019. Invasive coqui frogs are associated with differences in small Indian mongoose and rat abundances and diets in Hawaii. Biological Invasions. 21(6): 2177-2190. doi: 10.1007/s10530-019-01965-3.