The Proceedings of the National Academy of Science used to be a prestigious journal. And still is. Getting a paper in PNAS is a mark of approval from the scientific community at large. It is seen as a mark of approval for one’s work.
Therefore, if that journal publishes an asinine preposterous sloppy piece of foolish fluff, it must mean that the scientific community sees asinine preposterous sloppy foolish fluff as good.
Our peer-reviewed PNAS paper is “Systemic racism alters wildlife genetic diversity” by Chloe Schmidt and Colin J. Garroway. I don’t know, but I’m guessing with those names the Canada-based authors are white.
The title announces its falsity. We know that we must be in for a piece of science fantasy fiction, because we know there is no such thing as “systemic racism” against blacks. Indeed, all observation shows at all official levels, public and private, the opposite is true.
From the Abstract, my paragraphifications, my emphasis:
…it is not known whether varying habitat structures and natural resource availability associated with racial segregation affect the demographics and evolution of urban wildlife populations.
To address this question, we repurposed and reanalyzed publicly archived nuclear genetic data from 7,698 individuals spanning 39 terrestrial vertebrate species sampled in 268 urban locations throughout the United States.
We found generally consistent patterns of reduced genetic diversity and decreased connectivity in neighborhoods with fewer White residents, likely because of environmental differences across these neighborhoods.
The strength of relationships between the racial composition of neighborhoods, genetic diversity, and differentiation tended to be weak relative to other factors affecting genetic diversity, possibly in part due to the recency of environmental pressures on urban wildlife populations.
However, the consistency of the direction of effects across disparate taxa suggest that systemic racism alters the demography of urban wildlife populations in ways that generally limit population sizes and negatively affect their chances of persistence.
Our results thus support the idea that limited capacity to support large, well-connected wildlife populations reduces access to nature and builds on existing environmental inequities shouldered by predominantly non-White neighborhoods.
Those italics tell us the data did not support their contention that whites were evil, but, darn it, whites are evil “racists” anyway because everybody knows whites are evil “racists”.
I suppose an alternate explanation is that blacks scare squirrels.
Our authors whine about the long-past strategy of “red-lining”, and insist that lingering effects from it is what accounts for their meager signal. Even though it’s far from clear that practice existed at each of the points they sampled. And they say nothing about the voluntary separation of whites and blacks, accelerated in the late 1960s (and afterwards) because of rioting and crime.
Take Detroit, which used to be called the Paris of the Midwest. After the late-1960s riots, non-elite whites ran out to the suburbs and black leadership took hold in the city, cheered on by white elites. Detroit is no longer called the Paris of the Midwest.
Funny, though, because all indications are that the diversity of animals in Detroit has only increased, given that large portions of the city were outright abandoned, entire blocks being left to sink back into the soil. There are now even reports of coyotes roaming abandoned lots!
Incidentally, don’t we expect cities to be free of most kinds of critters? Isn’t that usually seen as desirable? Do we expect deer to strut down Fifth Avenue?
I can’t tell from their paper whether Detroit made the cut in their database. They have a picture listing the 268 sites they used (Fig. 1), but it’s difficult to make out (and no details are in their Supplement). Huge cluster of cities around, I guess, Chicago.
At least in some cities, like Detroit, there are more blacks than whites. And so it would be natural to find more animals around where whites lives, because fewer whites live in those cities.
To then say that “systemic racism” causes lack of genetic diversity would be, well, dumb.
Anyway, they only looked at “8 amphibian, 14 bird, 15 mammal, and 3 reptile species” and used, it seems, models of models to estimate genetic diversity. Not just counts you understand: not something like there are more animals outside cities. No. Genetic diversity of those animals sampled in the areas. But, of course, if there were fewer animals inside cities, we’d expect greater genetic diversity outside cities. So again, this study is absurd.
We could spend some time criticizing details of the model, which are unnecessarily complicated, critiques which would boor you. Let’s instead note that, according to their model, and using the usual “95% criterion”, there was only one species with “significantly” greater genetic diversity in white neighborhoods: desmognathus fuscus. Salamanders. Gussie Fink-Nottle would be thrilled.
All the rest, according to their “betas”, could not differentiate, using the normal 95% windows, any difference (in their Supplement). Which is why, perhaps, they lowered that 95% to 90%, such was their desire to discover “racism”.
It comes down to this: perhaps genetic diversity of salamanders of a certain sort are found more often by whites, the remaining species not caring.
Yet they claim “It is clear that systemic racism is altering the demography of urban wildlife populations on a national scale in ways that can shape the evolutionary processes acting on them and the probability of long-term persistence in cities.”
Sigh.
Subscribe or donate to support this site and its wholly independent host using credit card click here. Or use the paid subscription here. For Zelle, use my email: matt@wmbriggs.com, and please include yours so I know who to thank.
Chloe and Colin surely did not collect the samples themselves. Maybe it is easier to find volunteers for such work in "white" neighbourhoods?
Absurd ''science'' indeed. It is hard to believe that a statistically qualified person would approve of such confusing and idiotic modelling. I catch some things just by skimming through: What is the scientific model on which they base their statistical model? What is the justification behind informative priors, and what is meant by not strongly affected? A random effects model assumes linear relationships and that you observe and account for all (important) variables, where is the justification for this assumption? I am sure there are many other things.
It makes me sad to see that such incompetent research gets published.