So a couple of weeks ago, the Journal of Biogeography published a paper arguing that humans and orangutans are sister taxa.
This week, the journal has published a paper on the biogeography of Sasquatch.
Yes, that’s correct. The same journal that published the updated “red ape” paper has now published a Bigfoot paper.
OK, Journal of Biogeography, I’ll be your straight man. I mean, I bow to no one in my ability to snark on human evolution, but this is like some sort of karmic singularity.
Besides, this Bigfoot paper is really good. The authors intend it as a “tongue-in-cheek” example, and it works on that basis, as an illustration of the “garbage in, garbage out” principle. Biogeographers have been using increasingly sophisticated computer algorithms to predict the “ecological niche” of species. The algorithms take information about sightings or recorded incidences of a species, find commonalities among those sightings against maps of other ecological data (rainfall, forest type, presence of other species, etc.), and spit out an ecogeographic distribution for the target species.
Of course, the algorithm will come to some result, regardless of how well each piece of data in the system is known. And the algorithms are complex enough that the creeping effects of errors may be hard to evaluate:
While the value of publicly available sample locality data is not questioned, the consequent introduction of errors in the accuracy of specimen identity and georeferencing could be problematic for developing ENMs [that is, "ecological niche models"] from public data sources (Graham etal., 2004; Sobern & Peterson, 2004). Although georeferencing inaccuracies can be identified in databases from qualitative or quantitative accuracy thresholds (e.g. http://manisnet.org/GeorefGuide.html), poor taxonomy and/or misidentification may be less detectable. This issue may be particularly problematic, for example, with cryptic species or subspecies that are morphologically similar but may have very distinct ecological requirements and geographic distributions, or for those data sources that contain indirect observations rather than references only to physical specimens.
What better way to illustrate the problem, than by applying the analytical methods straight to a “species” whose existence is, shall we say, “questionable.”
The authors additionally raise one particular element of confusion that may enter into ecological niche modeling – sightings of similar species may confound each other. They consider the example of black bears – another large North American mammal that occurs in many of the areas where Bigfoot sightings are reported. They run the same analysis on black bears as they did on Sasquatch, finding a large overlap in distributions. But interestingly, they have fewer observations for bears than they do for Bigfoot – leading them to an underestimate of the actual range of black bears. They reflect on the possible interpretations of the overlap:
Thus, the two 'species' do not demonstrate significant niche differentiation with respect to the selected bioclimatic variables. Although it is possible that Sasquatch and U.americanus share such remarkably similar bioclimatic requirements, we nonetheless suspect that many Bigfoot sightings are, in fact, of black bears.
From my perspective, this paper is important for two reasons, neither of them really having to do with large North American primates. First is a social and legal angle. Increasingly, the habitat distributions of endangered or threatened species are evaluated on the basis of similar computer models of ecological niche. In particular, the changes in species distributions under scenarios of climate change are modeled in this way. This computer modeling has the appearance of objectivity, and it certainly allows reams of data to be statistically simplified into human-readable maps. That makes the results of such analyses really valuable to cases where political and legal units need to make decisions about how to comply with threatened species regulations.
But if the data going into the model aren’t correct, then the predictions of the models won’t reflect reality. The question is, how much will they be wrong?
In this case, the limited black bear dataset leads to a substantial underestimation of the black bear habitat niche. And possible confusion between Sasquatch and black bear sightings raises the possibility that any rare species will be significantly overrepresented in ecological niche modeling when such confusions are possible. Neither of these outcomes tells us how bad such errors are likely to be, but they point to real weaknesses in the maps generated by these computer algorithms.
I expect that smart lawyers will be finding ways to use this Bigfoot paper a lot.
Second, I think the paper is important at this moment in paleoanthropology. Late last year, I wrote about a paper that evaluated ecological niche models for Neandertals and people who made the Aurignacian (“‘Competitive exclusion’ and the extinction of Neandertals: should we believe it?”). The paper, by William Banks and colleagues, had used the observed distribution of archaeological sites between certain radiocarbon date intervals to estimate an ecological niche model for the two hominid groups.
I think that paper was very good work, but it obviously is subject to uncertainties in the initial observations – just as the Sasquatch data are. The archaeological observations add even more uncertainties of dating and certainty of association between biology and archaeology. And the Sasquatch data set is much, much larger in terms of sighting numbers; meaning that the archaeological cases ought to have more error, when it comes to evaluating niche flexibility.
For the purposes of the Banks et al. (2008) paper, I think their conclusions are pretty secure. They tested the hypothesis that the reduction of Neandertal occurrences over time could be explained by climate change; they were able to reject that hypothesis by showing the ecological niche breadth of earlier Neandertals included paleoenvironments that were very widespread across Europe throughout the period when they were declining. It’s a nice demonstration.
But the Sasquatch example shows that we have to evaluate the ability of ecological niche modeling to test each hypothesis, on the basis of the data that are likely to be available. Can the model show that the spread of Aurignacian people caused the Neandertals to decline? That depends on our confidence about the dating and biological associations at early Aurignacian sites. The computer algorithm gives a structured way to reduce information that is already manifest in the data.
Lozier JD, Aniello P, Hickerson MJ. 2009. Predicting the distribution of Sasquatch in western North America: anything goes with ecological niche modelling. J Biogeogr (early online) doi:10.1111/j.1365-2699.2009.02152.x
Banks WE, dErrico F, Peterson AT, Kageyama M, Sima A, et al. (2008) Neanderthal Extinction by Competitive Exclusion. PLoS ONE 3(12): e3972. doi:10.1371/journal.pone.0003972