Saturday, November 7, 2015

Tortoise Splitters II

Totally for real this time.
This here article, facebook'd to me by Chris Clarke, is pretty cool. Marshalling new and old genetic evidence, they show convincingly that the Galapagos island of Santa Cruz (aka Indefatigable) hosts not one but two genetically and geographically distinct populations of tortoises. This means one species (or subspecies, depending purely on opinion) (Chelonoidis porteri is the species version) gets Officially Split, with the newly identified Eastern population dubbed Chelonoidis donfausti (or C. nigra donfausti).
'kipedia link

The coolest part of the article to me, though, was their Fig. 2, showing a phylogenetic tree for all specimens of Galapagos tortoises that have been sequenced to date, including lots of museum specimens:

Which, ignore the nighbor-joining expercise on the right and check out the beeyootiful tree on the left there. Clades marked in red and orange are the two populaitons from Santa Cruz, and you can see how different they are in context.
As I sez to Chris on Facebook I sez "why no map of the islands with that tree mapped on there?" I sez and then I sez "Now I have to do it myself" I sez and then "also shell shapes oughta be mapped on that tree there too" sez I.
So here's my map [click to N-large]:
Where the amoebas enclose nested genetic clades according to the tree. Island by island, there have evidently never been tortoises on Genovesa and Marchena, hence the Xs. Tortoises are extinct on Floreana, Santa Fe, Pinta, and absent-and-maybe-extinct-but-maybe-never-really-there on Rabida (small) and Fernandina (new), hence the slashes.

Here's a simplified version of the big detailed tree, showing populations by island with color coding corresponding to the map [clickable]:
 ...where X marks extinct taxa and the circled S, I, and D tags refer to Saddle, Intermediate, and Dome-shaped carapaces, as per the Intro from the article at hand here.
For reference, slightly different lists of recognized species (or subspecies) are  here ('kipedia text and photos, here (text), and here (map) [differ mainly in recognizing or not the populations on the 4 other volcanoes of Isabela].

And so but if we naively and foolishly but heuristically take the genetic data at face value, then the tree plus map means that:
1. There are two main groups of GTs; one (1) comprises populations from Santiago and Volcano Wolf, the northernmost volcano of Isabela (where the new pink iguana species was found not long ago...uh...  link), and the second (2) is everybody else (green lines on map). That split represents a very early dispersal event. The northwestern Wolf/Santiago clade is in relatively new territory, but seem to represent an ancient lineage. (Galapagos tortoises probably evolved on islands that are long gone to the southeast).
2. Clade 2 separates nicely into two clades corresponding to eastern (2E) and western (2W) areas. The convincing thing about the article at hand is that the two populations on Santa Cruz belong to separate branches of this split. The newly named species (or subspecies) is marked with a star.
2A. Dome-shaped carapaces look to have evolved separately in the two Santa Cruz species (or subspecies). That, or saddle shapes evolved at least 6 times independently. Maybe a nice example of eviolutionary convergence.
3. Within the eastern clade 2E, samples referred to Santa Fe are most distinct. That's interesting because Santa Fe's purported population is listed in the references linked above as "disputed" and "of doubtful existence," apparently deriving from but a few bone fragmen ts. But if the DNA is that different, one's doubt dissapates a bit. The green coding of the speciment numbers in Fig. 2 of the article at hand means data from museum specimens that were previously reported; I have not dug back to find where.
4. Pinta island tortoises were most closely related to the once-relict-but-now-a-thriving-conservation-success-story torts from Espanola, way the hell down at the totally other end of the archipelago. [the last Pinta Island tortoise, the well-known Lonesome George, is specimen # abiLG in the tree. I have photos of him alive in 2009 someplace.]
5. Over in the western clade 2W, tortoises from the little island of Pinzon are most distinct. One of the 'kipedia articles linked above sez that there is yet a third distinct tortoise population on Santa Cruz, in the northwest, and that they are closest to Pinzon animals, which would be cool.
6. The disputed population from tiny Rabida looks like a transplanted Isabela tortoise, and the specimens attributed to the various volcano-based species (or subspecies) of Isabela are all pretty well jumbled together in this analysis. Maybe should be lumped back into C. vicina as on the linked map.
7. The "doubtful" population from Fernandina is interesting. It's from a single specimen (1906) and no other evidence of tortoises has ever been found there. Here it looks genetically distinct, but closest to the tortoises from southwest Santa Cruz. Except that the Santa Cruz animals are dome-shaped and the speciment allegedly from Fernandina is said to be saddled. And Santa Cruz is clear over on the other side of Isabela and Pinzon from Fernandina. WTF.

All in all, some satisfylingly tidy stories of vicariance, and some puzzles too. It's complicated.