Thursday, 27 January 2011

Sometimes, it does matter where you're from.
One of the major thrusts of the research I've involved with in the last few years is in the world of "domestic introgression." Some times, when we move animals around, there will be a pre-existing wild form of that species. For reindeer transplanted to Alaska, it's the endemic caribou. Some mink farms are in areas that already have a wild mink population. And a dog is nothing, if not a very funny looking, strangely behaving wolf. The concern is often that our tame species - the domestic species - will breed with the native wildlife, introducing genes that have more to do with living with humans than in the wild. That's what we call introgression - broadly speaking, it's when genes from a species or subspecies sneak into a new species. In many cases, natural selection can do its thing and purge these genes, but with the right conditions, even these very poorly-adapted traits do filter in to the wild population.

Red foxes have been extensively bred for farm fur, not unlike mink. The sort that have ended up in these farms are a total hodgepodge of various foxes bred together for the best fur characteristics, while still being tame. You might remember the tame, Russian foxes I talked about earlier, with their collie like appearance and their cheery disposition. Some foxes escaped in Californian fur farms in the mid-1900s, where they established a growing invasive population in the Sacramento Valley over the ensuing time. From here, they've come into contact with the native foxes - foxes that have long been there, and are well adapted to the local environment.

Enter Benjamin Sacks and colleagues, working out of the University of California, Davis. They took DNA from foxes throughout the Sacramento Valley, and looked at a variety of markers - some mitochondrial, some microsats, and some SNPs. In addition to the typical analyses (For HWE, Linkage, and so on), they assessed whether there were domestic introgression into the wild foxes using STRUCTURE (a package that assigns individuals to populations when you don't know the number or placement of the clusters) and BayesAss, which assigns individuals to populations based on prior knowledge of the populations. Finally, they used Migrate-N to assess how much the geneflow there was between the domestic foxes and the native foxes.

From the traditional stats, they had a microsat HE of 0.65 and 0.69 - something I would think is low given an average number of 6.1 and 6.6 alleles per locus. They didn't have any HWE issues. Additionally, it was clear that the wild and the introduced groups of foxes assorted with themselves spatially - that is, there was a region of native fox, and a region of introduced fox, and not the two intermixed in the same area. You can check out the figure to the right to get the general feel for the lay out - the caption has a lot of useful information in reading it.

When they looked at information about potential migrants, and admixture, they found a small number of individuals who didn't match either the 'native' or the 'introduced' groups that they resided within. When they looked at the mitochondrial DNA with the nuclear DNA, it was clear that all of them were of clear hybrid origin - not migrants, but the first generation cross between a migrant and the local group of fox. However, the authors also argue that there's a region of individuals who are primarily of hybrid origin. Finally, the analysis of Migrate-N showed a generally low level of introgression between the groups, of about 1.31 and 0.91 effective migrants per-generation.

Taken together, Benjamin Sacks and colleagues argue that there is some sort of selection preventing the groups from intermixing. Something about the inherent ecology of the wild or domestic foxes (or both) is preventing too much of the domestic genes from bleeding into the native foxes, and vice versa. They suggest that the hybrids they do find are due to a low density area, and but that the mating system of red foxes makes it difficult for introduced foxes to interbreed with the native foxes. I remain somewhat skeptical of this argument, and I'd like to see some experimental evidence to back the claim.

The study was generally well done, but I do have a few general critiques. The sample size from their "hybrid zone" is very low indeed. I would really prefer to see additional data from there, to actually elucidate the strength (or lack there of) of the hybridization. Getting samples isn't always easy, but drawing inferences about those regions on the basis of n=3 and n=3 seems perilous to me. I'd like to see a better treatment of hybridization using BayesAss. Ideally, there should be 'reference' populations to check against. I recognize this isn't available for one or both groups sometimes (oh, what I wouldn't give for reference populations in some of my own work), but results must then be interpreted in light of the lack of reference populations. There could be previous introgression that we don't see, because it's gone to fixation in one or both groups. It is unlikely that this has happened here, given the high Fst and that hybrids seem to be selected against, but it's something I keep in the back of my mind.

SACKS, B., MOORE, M., STATHAM, M., & WITTMER, H. (2011). A restricted hybrid zone between native and introduced red fox (Vulpes vulpes) populations suggests reproductive barriers and competitive exclusion Molecular Ecology, 20 (2), 326-341 DOI: 10.1111/j.1365-294X.2010.04943.x

Figure reproduced from the above cited publication under a fair-use rationale.

No comments: