Wednesday, April 22, 2009

Anthracothorax Mangos in Western Panama

What's going on with Mango hummingbirds in western Panama?

Current taxonomy recognizes three species of Mango hummingbird in Central America. Most experts agree with the following schematic for the distribution of these species.

The Green-breasted Mango (Anthracothorax prevostii) is found from southern Mexico to northern Costa Rica, and also Bocas del Toro Panama. Black-throated Mango (A. nigricollis) is found throughout much of the tropical lowlands in South America but can also be found in eastern Panama westward until the Canal Zone. A third species Veraguan Mango (A. veraguensis) has been described from the Pacific lowlands of western Panama and adjacent Costa Rica.

It is important to note that this is a short-hand simplification of a known taxonomic problem. What is actually going on is much more complicated, and largely unknown. Such problems are quite common in most taxonomic groups, but much less so in birds, because of the historical efforts of bird specimen collectors to document bird variation throughout their range. Ornithologists have solved the vast majority of these problems. But lack of specimens and/or research effort has left a few. Sometimes, new tools such as DNA sequencing and/or comparisons of bird vocalizations have provided  important clues. A good, recent example from North America are the two Empidonax species that until the 1989 AOU checklist were considered "Western Flycatcher".

In the case of the Mangos, University of New Mexico PhD student Michael Lelevier and I are looking at the plumage and DNA of the Mangos from western Panama at the same time to try to understand species limits in this area. We have found a correlation with Veraguan vs. Green-breasted type plumages and very shallow, subtle DNA variation in the VERY, VERY limited number of samples that we have available.

In the above picture, we compare a known male Green-breasted Mango (from Mexico) to three males from Chiriqui. We can see that two of the three Chiriqui birds have a deep indigo patch on the throat that extends to the upper breast. Such a patch is shared with the Green-breasted Mango. In contrast, the third individual does not show such a blue/black patch (note that the darkness in the extreme upper throat is an effect of shadowing, and under better light is glittering green; the colors in these hummingbirds are from structural pigments so the angle of the observation affects the color observed). This third individual has plumage characteristic of Veraguan Mangos. It should be noted that all three of these Chiriqui birds were collected on the same day at the exact same site over a period of two hours in a windbreak near Remedios; thus any differences are not due to geography.

When we classify all existing specimens of Mangos from western Panama that have a museum skin voucher specimen and an associated tissue allowing DNA-based studies (a total of 17 individual birds collected during the last 30 years), we find the smallest possible variation between the two plumage groups (All Chiriqui birds were collected by Michael and me, and classified to plumage by Michael; birds from Bocas were collected in the late 1980
s by Smithsonian scientists from
 Washington DC, who provided the plumage-based classification. In fact the variation (two mutations, see the big red arrow) is the same as what we find within individuals of the same plumage type. In fact, there is one Chiriqui bird with Green-breasted plumage that varies by 5 mutations from other Green-breasted Mangos. Also note that the two shared mutations that help to discriminate Green-breasted from Veraguan plumages are found in geographically remote populations in Mexico and Belize.

What does this mean? If the two are separate species, they are incredibly young. Most sister species of birds are between 2% - 10% different from one another using the DNA marker that Michael and I are using here. That would mean 20 - 100 mutations (since our DNA fragment is about 1000 bases long). Therefore, if we have two different species, we are talking an order of magnitude less difference than what is "normal". Also, it means that there is no genetic barrier to inter-breeding between the two forms. Up to about 15% differences, different bird species can readily hybridize, and much below 4%, they are so compatible that it is possible for one species' genes to swamp another. This is currently going on in Blue-winged and Golden Winged (Vermivora) Warblers in North America. Any reproductive isolation would have to involve mating displays or other behavior since the two plumage types ARE NOT geographically seperated (contra the earlier schematic). 

Alternatively, speciation was interrupted. Speciation might take a million years or more; the genetics tells us that the split between Veraguan and Green-breasted Mangos is 100,000 years or likely much less (since we expect 2% variation per million years, however you should subtract the level of within population variation to correct for natural variation that occurred...doing so would leave you with an estimate of 0 years since splitting. The true date is somewhere much less than 100,000 years ago). If this is right, than it isn't fair to characterize the Mangos of western Panama as either Veraguan or Green-breasted, they actually would be a little of both. (I know that this is troubling for twitching life lists...and future posts are going to address this with Jacanas and many other birds from Panama. I hope that ultimately this information makes time spent watching Panama's birds more fulfilling, even if it costs a few twitches).

How do we seperate between the two possibilities? First, it would be great to have more specimens. The taxonomy of these guys is based on the 17 birds that we sequenced and only a dozen or two more that were collected before the days of DNA science. We will need to punt until we get more material. If we do have more material, we can use other DNA tools to more finely characterize gene flow between the forms. At the same time, field observations of mating behavior are perhaps equally critical. Ideally we'd geneically characterize birds attempting to mate to measure the extent of mate choice, but if you've spent any time watching these birds, you know how operationally difficult that will be. Additional genetic data should point us in a better direction.

I hope to provide a similar snippet of my on-going research using this blog on a weekly basis. Up coming notes are being worked up for Selasphorus hummingbirds of western Panama, Snowy-bellied Hummingbirds, Mionectes flycatchers and Schiffornis, as well as I hope some quite novel data on Red-capped and Golded-headed Manakins from eastern Panama.