Galápagos Racers: answers to your questions about the BBC Planet Earth II iguana chase scene

One of the few phylogenies to include Galápagos Racers From Pyron et al. 2013

Broadly, Pseudalsophis is nested within a large clade of Caribbean, Central, and South American xenodontine snakes including, among numerous others, the genus Alsophis, which once contained Galápagos Racers and after which their current genus is named. They have been in a variety of genera since their description, especially Dromicus, which is no longer in use, from 1876 to 1997.

In 1973, herpetologist Charles Myers wrote: "The classification of colubrid snakes in general, and of South American colubrids in particular, is in a notoriously unsatisfactory state." Unfortunately, we are not that much better off today when it comes to Galápagos Racers. It seems pretty clear that the nearest relative of P. biserialis, P. dorsalis, and P. occidentalis is Pseudalsophis elegans, the only species in the genus found on the mainland (in Ecuador, Peru, and extreme northern Chile). Beyond that, there isn't a lot of clarity about their next-closest relatives. They are possibly most closely related to obscure South American "groundsnakes" in the genus Psomophis, or to the even more obscure genus Saphenophis, which was described by Myers as "quite lacking in peculiar or unique features" and so named "in allusion to one incontrovertible fact about these snakes...from the Greek saphenes (evident truth, clear) + ophis (a serpent), meaning 'clearly a snake'". We don't really have a great hypothesis about how the different lineages of Galápagos Racers are related to one another, or even if they are all descended from a single common ancestor, because we only have DNA from one of them so far.

Hypothesized scenario for the evolution of Pseudalsophis snakes So far, we have no DNA evidence that would support or refute this model From Ali & Aitchison 2014

Two reviews based on morphology addressed this question in the late 1990s. The first (Thomas 1997) focused exclusively on Galápagos Racers and suggested that P. biserialis, P. dorsalis, and P. occidentalis are descended from a shared common ancestor with P. elegans, but that P. hoodensis is more closely related to the mainland species Philodryas chammissonis, and that P. slevini and P. steindachneri are most closely related to Caribbean species. The other study (Zaher 1999), which looked at hemipene morphology over a much larger group of snakes, disagreed, finding a shared derived character—an inflated papillate ridge, placed far medially, on the medial surface of the lobes—linking the Galápagos Racers together with the mainland species P. elegans. Statements that Galápagos Racers have “very similar hemipenes” notwithstanding, Zaher was criticized for not describing the specific characters uniting the Galápagos species to the exclusion of others.

Maglio (1970) noted that the tooth counts and arrangement and the and shape of the premaxilla bone was most similar among the three Galápagos species that he examined (P. biserialis, P. dorsalis, and P. slevini), and different from the West Indian species that Taylor later suggested are P. slevini's closest relatives. More recently, a study led by Grazziotin claimed that they "unequivocally support...Zaher' s (1999) hypothesis based on morphology that continental Pseudalsophis elegans is closely related to the Galápagos Island species of Xenodontinae (herein represented by Pseudalsophis dorsalis), rather than to West Indian Alsophis and Antillophis, and mainland Philodryas (Thomas, 1997)." However, they obviously didn't read Thomas's paper very carefully, because he also hypothesizes that P. dorsalis is closely related to P. elegans, and the Grazziotin paper didn't sequence any DNA from P. slevini, P. steindachneri, or P. hoodensis, and therefore didn't test any hypotheses about them.

As for whether or not the snakes in Planet Earth II should be called P. occidentalis or P. dorsalis occidentalis, that's really a lumper/splitter question. But, both the IUCN and the 2014 edition of Snakes of the World recognize P. occidentalis as a full species; it was originally described as such by Van Denburgh in 1912, sunk to a subspecies of P. dorsalis by Mertens in 1960, and re-elevated to a full species in a 1999 paper by Zaher that was not primarily concerned with taxonomy and appears to have subsequently been neglected. The Reptile Database is currently a holdout for the subspecies designation, which has not been disputed but which is also not explicitly supported by unambiguous data. Perhaps wisely, the official webpage of Galápagos National Park chooses not to use scientific names and refers to the Fernandina racers as the "western subspecies". The truth is that, until more research is done, we won't be able to settle on an accurate taxonomy for these snakes.

Thanks to Andy Kraemer and Jim Moulton for the use of their photographs.