14 April 2006

Aplications of Evolution 4 - Migration and Sex

In recent years, the costs associated with genetic studies have dropped dramatically. That isn't much of a surprise. When the two main tools used in modern genetic studies, the polymerase chain reaction (PCR) and DNA sequencing, were first developed, they were extrordinarily labor intensive. Initially, PCR required the scientist to spend hours and hours moving the samples from one water bath to another. DNA sequencing required the researcher to determine the order of bands on a gel by eye - determining a single sequence could easily take a whole day. Later, methods were developed to automate these processes, and they became much less labor intensive. Instead, they were simply very, very expensive.

Fortunately, technology tends to become cheaper over time, and this has proven to be as true for molecular genetics as it was for CDs and DVDs. In recent years, the costs have dropped dramatically. This means that using genetic techniques has now become feasible for many more researchers, and that they are being used to address many more questions. In many cases, the addition of genetic tools to a study has led to some amazing and sometimes unexpected results.

One interesting example just appeared in the "online early" section of papers for Molecular Ecology. (This means that the paper has been reviewed and accepted by the journal, and will appear in print in a future issue.) In this paper, a group of wildlife biologists used both radio-tracking and genetic techniques to look the effect of the Ventura Freeway on the Los Angeles area populations of bobcats and coyotes. The authors examined two types of data: they looked at the movement of individual animals that had been captured and radio collared, and they looked at the genetic structure of the populations. What they found is that many animals migrate from one side of the freeway to the other, but that few of the migrants are reproductively successful.

The authors used three study locations. Two were on the north side of the freeway, but separated from each other by a major secondary road. The second was south of the freeway, across from the northeastern site. This experimental setup let the authors compare the effect of the freeway to the effect of the secondary road. Over the seven years of the study, they radio collared and tracked 110 coyotes and 87 bobcats. Genetic samples were collected from all of the animals they captured.

The radio collar data allowed the researchers to see where the animals were moving. This technique has been around for a while, and is still used today. They found that only a small percentage (4.5% of coyotes; 11.5% of bobcats) of the animals they studied were observed to cross the freeway. A much larger percentage (~50%) of the animals crossed major secondary roads.

OK, there's nothing surprising about that. Anyone whose ever watched Wile E Coyote and the Roadrunner square off can tell you that coyotes and highways aren't a real good mix. The bigger the road, the bigger the problem for the animal. That makes sense. However, the road is not an absolute barrier to the migration, and it only takes about one migrant per generation to keep populations from drifting apart, so it looks like the highway won't significantly fragment the bobcat and coyote populations. That's good news, because a fragmented population is usually at a greater risk of extinction than an intact one.

If the radio collar data does a good job at predicting the migration rate, the genetic data should show that there aren't big differences between the animals at the three study sites. To examine this, the authors used a series of statistical tests to determine whether the individuals sampled should best be considered to be a single population, or separate populations. They found that the coyotes grouped into two distinct subpopulations, and that the bobcats grouped into three. A second series of tests was used to determine how distinct the populations were from each other. This demonstrated that there were statistically significant differences between the populations. These results indicate that there is a very low level of migration between the populations.

So what's going on here? The radio collar data seems to point to a much higher migration rate than the genetic data demonstrates. What could cause such a difference?

It all comes down to that perennially popular topic - sex. It looks like physical migration across the freeway is relatively common. The genetic data actually supported this, too. Some individuals with the genetic signature of the south of the freeway populations were found north of the freeway, and vice versa, indicating that there are migrants. The reason that the populations show so much genetic divergence in the face of all that migration is simple: the migrants aren't reproductively successful. Their genes aren't winding up in the gene pool, and from a genetic perspective any migrant that doesn't reproduce doesn't count.

This is one of those experiments that screams to be repeated. If the pattern is real and holds true for a wide range of taxa, it implies that populations might be more easily separated than suspected, which might, in turn, imply that evolutionary divergence is also even easier than suspected. Even if the findings only apply for territorial animals such as bobcats and coyotes, they might require substantial changes in how those species are managed.

No comments: