Fantastic Bridges

This was pointed out to me by someone else, it's an odd glitch in google maps.

  The map shows south Brooklyn. Manhattan Beach is on the east-end of a  bit of land on the southern base of Brooklyn, with Coney Island at the west-end. The long strip at the bottom of the screen capture is a barrier island, made up of  Breezy Point and the Rockaways.  This map shows a bridge extending from Breezy Point into Manhattan Beach. There's no such bridge.

Here we're zoomed in on the north end of the 'bridge', and it actually is shown running into the CUNY Kingsborough Community College campus.
In fact a big chunk of the campus here is mysteriously shown as under water.
There's nothing like any of this in the satellite view, shown below.

Outside of this being a weird map rendering issue---and the error is not normally there, I've checked this map before---my first thought was, maybe it's a projection involving potential sea-level rise. But that wouldn't explain the "bridge", which is also rather wide for a bridge. Plus, the campus has a high sea-wall running all around it, so if the south end of it was flooded, the whole campus should be flooded. Water could be coming in to the west, off campus, and run into campus, but I'd think if that was the case then there'd be more flooding overall. Anyway, it's probably not the case that 'someone' at Google accidentally put in data from some imaginary flood scenario into the standard Google Map.

I actually wish this bridge did exist, it'd be better taking the Southern Parkway and driving through the Rockaways, and then over this bridge, than taking the dread Belt Parkway.

Colin Groves - Carl meets Karl: The case for testability in Primate Taxonomy

Dr. Colin Groves is a well known primatologist from the Australian National University. Amoung other things, he was one of the people who named Homo ergaster. His books include the textbook: Primate Evolution, 2001's Primate Taxonomy, and 2011's Ungulate Taxonomy. He's been called "primatology's latter-day Linnaeus" [Rosenberger 2009 (pdf)], which figures in the talk's topic.

The ~hour long talk---"Carl meets Karl: The case for testability in Primate Taxonomy"---was held at SUNY Stonybrook's Anatomic Sciences Seminar room (HSC T8-025) Monday July 1st at noon. Dr. Jungers did the introduction.

The two Carls are Linneaus and Popper, as might be guessed from the issues of testability and taxonomy.  

A little Background

Dr. Groves was making an argument for the use of the Phylogenetic Species Concept (PSC) in Primatology/Anthropology, in lieu of the Biological Species Concept (BSC). In brief, species concepts are attempts to define what we actually mean by species, which is no small task. Effectively there is no one universal species concept today, and debate on the issue can get quite heated.

The BSC defines a species as a population capable of interbreeding: if you have two populations that can't produce fertile offspring, you have two separate species. This is the species concept of highschool biology courses, but of course it was advocated by no less a figure than Ernst Mayr. 

There are three different types of Phylogenetic Species Concepts (Wilkins, 2009) :

1. Hennigian: species are unbroken & unbranched lineal segments of an evolutionary tree
2. Phylogenetic Taxon: the smallest unit in a phylogenetic tree is the species
3. Autapomorphic: species are set by diagnostic associations of unique characters

The bulk of the Talk

Groves says that the BSC is not testable because in most cases where we wonder "are these two species or one?", the populations have an allopatric distribution. They're geographically separated and there's no test of reproductive isolation. The BSC then falls outside the limits of Popper's demarcation criterion (that science must be testable). In the few instances where we can test inter-fertility (when the populations are in sympatry or the populations are housed together in, say, a zoo) interbreeding between putative good species often occurs, the BSC fails the test.

Groves explained that some good primate species, like Trachypithecus pileatus, have been shown to be hybrids (via Y-chromosome and mtDNA studies). He also showed a map of the distribution of baboon species across Africa, and then showed an African baboon mtDNA haplotype map, which broke up some of these species and, he said, gave evidence for hybridization between some species.

So, for Dr. Groves at least, the BSC is either unscientific (and thus should be rejected) or often falsified (and thus should be rejected).

Dr. Groves reviewed a few alternatives to the BSC, such as:

1.  Ghiselin 1996 - Species as individuals
2. Van Valen's Ecological Species Concept
3. Paterson's Recognition/Fertilization System Concept
4. Templeton's Cohesion Species Concept
5. Mayden's Genetic Distance Species Concept

But Groves ultimately rejects them all in favour of the PSC. Groves, following the standard usage, attributes the PSC to Eldridge & Cracraft. Groves also notes that Darwin anticipated some aspects of the PSC (that species are just like sub-species, but with wide gaps between them as intermediate forms go extinct), and that Gerrit Miller (who recognized that the Piltdown jaw was actually from a non-human ape) felt that if there are no morphological intermediates between two populations, then they must rank as separate species.

Groves defines the PSC as "the smallest population or aggregation of populations which has fixed, heritable differences from other such populations."
There are, apparently, many species of Mangebey (Cercocebus) monkeys, with different people sorting them differently. Groves performed a discriminant analysis which he claims showed absolute differences between the populations, and that thus each of these groups is a good, phylogenetic, trait-based species.

So Groves's major points is that the biological species concept at best fails when tested, and that the phylogenetic concept is more objective and testable; therefore anthropologists should explicitly use it.

In the brief question session, Groves was asked about what he meant by "fixed", he said he was talking about a binary 'yes/no short/long' difference between populations, but at the same time seemed to say that there was a good amount of wiggle room there.
Another important questions was: how far should someone take this method? Taxonomy & Systematics has long been divided over the 'lumper/splitter' issue. Groves rather strongly felt that the default should be that an operational taxonomic unit (OTU) is a species, and that one would need to demonstrably show that it was a subspecies otherwise.

My reactions

So one thing to note here is that Groves is using the Autapomorphic "flavour" of the phylogenetic species concept, that a species is something that is 'diagnosable', and that species are the terminals of a cladogram. Some people treat this concept as meaning that the pattern of characters is real, but that pattern doesn't tell us anything about how the groups of species evolved. Thus there's a pattern vs process arguement amoung advocates of the Autapomorphic concept (and often this is called an problem of ontogeny-epistemology). 

Wilkins (2009) reminds us of a case where workers had identified the species of seals and were now considering   whether mtDNA haplotype data supported them. But under the PSC definition, each haplotype has to be considered a separate and distinct species. The authors in that study ended up resorting to biogeography and breeding data to justify their original choices of species, but that invalidates using the phylospecies concept in the first place.
This problem is taken to an extreme by Vrana & Wheeler (1992), who argue that the terminal taxa in a cladistic analysis should be individuals, not species. The only way to group individuals in a cladistic analysis (under this concept) would be if all members of the group were identical for all characters considered. Vrana & Wheeler also note that this would be a major problem for conservation biology, because the taxonomic units that people try to conserve aren't real.^*

Groves believes that his usage of the PSC works in the modern day, and for any given 'slice' through paleontological/evolutionary time. But a problem here is highlighted by some of his listed antecedents. In the past, intermediate forms and slight gradations would connect what today are separate species with 'fixed' characters. So what's a species, based on morphology, today, somehow wasn't a species, based on the same morphology, in the past. Groves thus has, ironically, speciation by extinction. 
A question actually came up for what Groves and the PSC means for paleontologists, since Groves was using soft-part anatomy (like pelt coloration) to mark new species. But it should be obvious that the PSC can be applied to paleo-species, it's just that the characters are going to be preservable ones. Of course, the major problem here is that a paleontologist today, and a hypothetical neotologist in the distant evolutionary past, would practically never identify the same species.
This all weakens the falsifiability claim that Groves makes for the PSC, the concept becomes extremely subjective in his rendering, much more so that, say, testing for reproductive isolation (the hallmark of the biological species concept).

Further, does the biological species concept really fail in the instances cited? In the case of two allopatrically separated 'obviously good' species that are brought into captivity, mate, and produce fertile offspring, what has been tested? A particular hypothesis (that there are two species) has been tested, and it failed, there's really one species in this scenario, according to the biological concept, which wasn't itself tested.

The existence of hybrids also is certainly not a new challenge to the BSC, plant species apparently easily hybridize, producing fertile offspring, in fact plant species can sort of self hybridize, they can fertilize themselves and instantly produce a new species with twice the chromosome number of the parental species. 

Further, the phylospecies is distinguished from the morphospecies when the characters have undergone a cladistic analysis: that's how the autapomorphies are identified as such (as unique, derived traits specific to the group).
There was very little discussion of phylogeny in Groves talk and no cladistic analysis presented. Perhaps this is just because he was giving a condensed talk to a small group and he's explained this elsewhere.

The question over species concepts has been around for an extremely long time and isn't likely to be settled anytime soon. Coyne and Orr's [2004] recommendation that workers adopt a definition that best addresses their particular question is probably where the issue stands today, which isn't much help.


* Vrana & Wheeler continue by stating that, even if the 'species' isn't real, this could still mean, if a cladistic analysis reveals it, that there's structure within that 'species', and so a conservationist now has an argument for preserving each lineage within that, an appeal for preserving "cladistic diversity"


References

Coyne, J. A., & Orr, H. A. (2004). Speciation (p. 545). Sunderland, MA: Sinauer Associates.


Rosenberger, A.L. 2009. History of Primatology: The Alpha Taxonomist's View. Book review of: Extended Family: Long Lost Cousins. A Personal Look at the History of Primatology. Colin Groves. Arlington VA, Conservation International. Evolutionary Anthropology, 18: 79.

Wilkins, J. S. (2009). Defining species: a sourcebook from antiquity to today.  American University Studies. Series V. Philosophy. Vol. 203. Peter Lang Publishers.