And now for some more research blogging! Here’s a paper comparing the roles of geography and ecology in the early stages of speciation. I’m incredibly excited about this study because it begins to get around some of the major difficulties inherent in studying reinforcement and ecological speciation. But first let me back up and explain some background.
We define species (at least when we’re talking about sexually reproducing organisms) as reproductively isolated populations—that is, groups of individuals that can’t successfully interbreed with members of other groups. But this leads to a paradox, namely that natural selection should never favour a loss of the ability to breed. Why should new species ever evolve?
There are, of course, several ways around this paradox, but I’ll only mention the two that are relevant to this paper. First, the inability to interbreed (aka reproductive isolation) could evolve as a byproduct of adaptation to two different environments. For example, male guppies in high predation environments have duller colours so they’re less visible to predators, but then should they encounter females from low predation environments they’ll be less attractive.
A second mechanism is reinforcement. Two populations that don’t currently exchange migrants begin to diverge (by adaptation to different environments or some other process), then come back into contact. Hybrids between the two populations are less fit in some way, so selection favours individuals that recognize their own population and avoid mating with the other.* Critically, reinforcement can only occur between populations that exchange migrants.
The scientists who conducted this study looked for evidence for these two processes in guppies—not different species of guppy but populations of the same species that are incipiently divergent. The populations live in three river systems in Trinidad; populations that live higher up each watershed have lower predation than downstream populations because waterfalls prevent larger predatory fish from getting there. This is a convenient scenario for comparing the effects of reinforcement and byproducts of adaptation. If females prefer males from the same predation regime (high vs. low) regardless of what watershed they come from (guppies from different watersheds will never encounter one another in the wild), ecology has the stronger effect (i.e. reproductive isolation is evolving as a byproduct of adaptation to predation). Two types of geographical patterns are possible: females could prefer males from their own river, or from different rivers. The latter case would be reinforcement—females learn or evolve discrimination against the “wrong” males only if their populations are in contact. The former case could mean several things: byproduct effects of adaptation to different watersheds (ecological factors other than predation) or simply shared evolutionary history (populations within a watershed are more closely related and thus haven’t had enough time to evolve reproductive isolation).
The authors measured female preference in two different experiments. One was a no-choice (i.e. one male and one female guppy in an aquarium) laboratory experiment that took into account both the intensity of the females’ courtship display and whether mating occurred. The other looked at only high-predation females placed in large stream enclosures with both low- and high-predation males and used genetic markers to determine the paternity of embryos (guppies are livebearers, so the mother’s identity was known).
In the no-choice experiment, low predation females preferred low predation males regardless of whether they came from the same watershed—evidence that diverging ecology caused reproductive isolation as an incidental byproduct. High predation females, however, discriminated against low predation males from the same watershed but not from different watersheds—evidence for reinforcement. This finding was paralleled in the enclosure experiment, where high predation males sired most of the offspring when they were competing with low predation males from the same watershed, but statistically equal numbers of offspring compared to low predation males from other watersheds.
The authors looked for correlations of male colouration—area and intensity of orange patches—with predation regime and with mating success, but, although low predation males were consistently more brightly coloured, results were mixed. However, many other traits are involved in mate choice and/or predator avoidance.
This paper clearly shows that reproductive isolation between diverging populations can arise for different reasons. Most importantly, however, it takes advantage of a geographic situation (parallel adaptation to different predation regimes in different rivers) to disentangle the effects of local adaptation and reinforcement, two processes that can produce very similar patterns when you only have one or two populations to look at. I doubt that this sort of geographic pattern is particularly common, so it may be hard to apply this approach in many other places. Nevertheless, this study is a fresh look at how ecological divergence and reinforcement might act in nature when new species are formed.
*These are not by any means mutually exclusive processes. In fact, in order for reinforcement to proceed there already has to be some reproductive isolation between the populations (hybrids have lower fitness); this could be a byproduct of adaptation or it could arise through some other process. But that’s another story.
Schwartz, A., Weese, D., Bentzen, P., Kinnison, M., & Hendry, A. (2010). Both Geography and Ecology Contribute to Mating Isolation in Guppies PLoS ONE, 5 (12) DOI: 10.1371/journal.pone.0015659