"and I want no other fame": The tale of the butterfly novelist--Vladimir Nabokov

Lolita may be a story for adults only, but Vladimir Nabokov told another story that can be enjoyed by all.  It appears to be an amateur's triumph, based on a correct guess that was before its time.  No definitive data could have been assembled to tell the tale while Nabokov was alive.  But thanks to the very regular, clocklike way that DNA accumulates variation among descendants over time, the guess seems to have been confirmed with current methods.

Polyommatous blue, by Lilly M.,
Wikimedia Commons

The immediate story is told by Carl Zimmer in the NY Times, but is based on a paper just published in the Proceedings of the Royal Society.  If Lolita was about a man in pursuit of a (too-)young person he obsessively adores, the story of the Polyommatous blue butterflies is one of a man chasing a group of old species that he obsessively adores.

It has long been widely known that Nabokov was a persistent, knowledgeable, dedicated if technically amateur butterfly enthusiast.   It was also long known that he did extensive work on the structures and relationships of the species that came years ago to be known as 'Nabokov blues'.  You can read about this, with maps and illustrations, in a fine book by Kurt Johnson and Steve Coates, Nabokov's Blues: The Scientific Odyssey of a Literary Genius, Zoland Books, 1999.  Johnson is a lepidopterist, expert in these butterflies.

They write "Where South American temperate life-forms had come from became a compelling question from the earliest stages of the continent's exploration." Much of their book, which is a readable, popularized narrative, tells of the efforts by Nabokov and others then and since, to understand the taxonomy--species relationships--among the South American blues and their relatives elsewhere. 

The traits of species available to Nabokov and others did not neatly correlate with their locations in the Andes or down into warmer environments.  That meant that some nearby species seemed too different to have had an evolutionarily recent common ancestor.  And, if they did not, and for the group as a whole, where were their origins then?

This issue had arisen out of the work of many naturalists going back at least to the early 1800s.  From an evolutionary point of view a couple of explanations were plausible.  One is polyphenism: species can have very different morphology or behavior depending on the individuals' local environment or genotypes.  And there is mimicry: distantly related species can come to resemble each other by natural selection. But that will only affect the mimicked trait, leaving the species' other traits less similar.  But in any case how did these butterflies get into the Americas?  There were various suggestions.  One possibility that Nabokov entertained was that the species had expanded into the Americas from Asia, by way of the Bering land bridge--that is, into South America from the north, what is now the Arctic.  The North American ancestral species may then have died out, leaving only their South American descendants.

The story in the Times, like news stories tend to do, makes this seem as if Nabokov got this 'blues' story out of the blue, so to speak, and was insistent on this view.  Indeed, Vila et al. in the Proceedings of the Royal Society paper, do the same:

The radiation of Polyommatus blues in the New World was first appreciated by the famous writer Vladimir Nabokov when he was working as curator in the Museum of Comparative Zoology at Harvard in the early 1940s.

And:

Our results show that Nabokov’s inferences based on morphological characters (primarily of the male genitalia) were uncannily correct in delineating not only species relationships but also the historical ordering of these five key events in the evolution of New World blues.

But even brilliant discoveries occur in a context, and people rarely take immovable positions when the evidence is ambiguous. Johnson and Coates go to great lengths to describe the long history of South American biogeography--the distribution of species over space and how it gets that way.  Ideas of species rafting across vast oceans had been suggested, and when continental drift was discovered, land connections between South America and Asia and Africa provided a possible explanation.  But climate change and a northern connection was another possibility.

There's no taking away from Nabokov that he was diligent and insightful, but others were thinking similar kinds of things, about the biogeography of many different species besides butterflies.  As he said in our title quote, he wanted to be known for this work, and properly so.  But at least Johnson and Coates present the history as one of many widely discussed possibilities by many different investigators--including the Northern origin of South American species generally.  Nabakov appears to have weighed different possibilities, and his preferred hunch, given the set of specimens and knowledge available to him, appears to have been right. 

The new paper uses extensive DNA sequence to resolve these tales in detail, data that of course was unavailable in the past.  The authors draw species-relationship trees based on the degree of sequence identity.  The times of splitting among the species were estimated by the number of DNA sequence differences that had arisen.  Many of the branches were  deep--large numbers of sequence variants, suggesting ancient splits, relative to other nearby species in South America.  Since the nearest relations in these instances were in Asia, the American species must more than once--in different expansion waves--have come from Asia.

Paleoclimatology and continental drift, and known climate tolerance patterns in the butterflies made it possible for the butterflies, over many generations, to expand here from Asia and survive in the warmer climes that existed episodically over an estimated period of 10 million years.  DNA analysis can even work when there has been selection for polyphenism or mimicry.  That's because such selection would distort relationships only at the genes involved in those particular traits, but genomewide variation will accumulate in a pattern that corresponds to the species' histories.  Indeed, DNA evidence could provide evidence for it, by showing that though some species looked similar, they really were not very close relatives overall (genomewide).

Here is an application of genetics that is entirely appropriate, that makes it possible to draw convincing inferences, when morphological or behavioral analysis may not.  A tree of descendant genomes accumulates variation probabilistically, so little can be said from observation of just a few nucleotides.  But these erratic patterns even out when thousands of nucleotides are compared, as these authors did.

Even so, let's keep in mind that genetic analysis is not exactly a scientific miracle.  If genes cause traits, then traits will diverge over evolutionary time in ways that generally reflect underlying genetic divergence.  That is why, in effect, Nabokov was indirectly using genomewide genetics to draw his conclusions--he just couldn't see the genes directly.  This is in essence how Darwin did what he did, too, without a good understanding of what inheritance really was.  But what we now have is explicit genetics to replace Nabokov's inferences from aggregate, implicit genetics. And this can get around problems of phenotypes that don't fit the history.

DNA is more specific and in the sense of time estimation much more rigorously useful than morphology and behavior.  As in forensic applications of genetics, that match sequences to their owners as in crime investigations, analysis that relies on the clear properties of genomes can tell stories that are as powerful as the compelling novels that Vladimir Nabokov wrote.  And he is justly famed for both!