Debate Continues Over Partial Reproductive Isolation

Last year, scientists described how partial reproductive isolation between two sockeye salmon populations had evolved at the astonishingly rapid rate of about 13 generations. This was stunning to many biologists, who think of reproductive isolation as a process that evolves over tens of thousands, or even millions of years, but certainly not decades.1 Researchers led by Andrew Hendry, a postdoctoral scientist at the University of Massachusetts, Amherst, used microsatellite markers and morphological data to piece together the recent evolutionary history of two neighboring populations of sockeye salmon in the Pacific Northwest. One fish group spawns on a lakeside beach, the other in one of the lake's major tributaries. Both populations originated about 60 years ago from the same ancestral stock, when then-barren Lake Washington had salmon reintroduced to it from a nearby hatchery. Hendry's team observed that male river sockeyes are slimmer than their lake counterparts (to better deal with the river's strong current), while the female river sockeye are larger than their beach counterparts. River females need to bury their eggs deep into the gravel, which can be disturbed or destroyed by high water flows. Hendry has never claimed to have found a different species, only two reproductively isolated populations that do not interbreed freely anymore. "This is the same process by which new species might arise," he says. Despite the "might," Hendry's paper has spawned much discussion. The findings "cross two broad areas of interest, salmon conservation and speciation," Hendry says, "and I've been challenged on both fronts, almost independently." Some support for his theory comes from evolutionary biologist David Reznick, University of California, Riverside, who says the study "serves as an important and different way of how speciation might occur." Courtesy of Andrew Hendry, University of Massachusetts River Sockeye Salmon After Hendry's results were published, a group of evolutionary biologists responded by describing how these researchers had "failed to make the case that the two populations are indeed reproductively isolated."2 And several biologists from the National Marine Fisheries Service in Seattle responded with a letter outlining how "Hendry et al.'s conclusions, however, depend on several assumptions."3 Currently, some Seattle-based researchers are taking a closer look at the key assumptions. "Because there are such important conservation implications," says Jerry Johnson, a postdoctoral scientist at the Northwest Fisheries Science Center in Seattle, "it's important to get this right." The spawning difference between these two sock-eyed salmon populations is one criterion for identifying demographically independent populations. Opponents to salmon conservation argue that "If you can create a beach-spawning population from a river-spawning population in 13 generations," then there is no need to protect them as separate stocks or populations, says Johnson's colleague, conservation biologist Rick Gustafson. "People are using the study as a reason why salmon recovery is not as crucial as scientists here have been claiming it is." But the danger of thinking like this, says evolutionary biologist Mike Kinnison, University of Maine, is that evolution, whether fast or slow, can't happen without sufficient genetic variation. Wiping out all of a population's genetic variation by not practicing good conservation eliminates the evolutionary opportunity. Notes Kinnison, "It is still more efficient to preserve that old variation." Hendry agrees, "I don't think my results argue against the conservation of distinct populations." In a letter to American Scientist, Hendry wrote, "Although rapid from the perspective of evolution," still the rate of change "is glacially slow, relative to the interests of humans."4 Conservationists Take a Closer Look The conservationists' main scientific concern is that the genetic differences upon which Hendry and his collaborators base their argument may simply reflect an already existing variation in the original hatchery population, not something that has recently and rapidly evolved. These salmon have a "very convoluted history" says Gustafson. "There were both beach-spawning and river-spawning life-history types in the original mix of fish that were ultimately used to transplant into Lake Washington." Hendry agrees that the original population was mixed, but he says that the fish were propagated "as a single-mixed stock for up to eight generations" before they were introduced into the lake. According to Hendry, this was enough time for all of the fish in the hatchery to become genetically homogeneous. But, "that in itself is something that needs to be tested," says Kerry Naish, an evolutionary biologist at the University of Washington. Gustafson feels even more strongly: "I don't agree with the eight generations, and I don't agree that they were homogenized over that time." Adds biologist Steven Kalinowski, National Marine Fisheries Service, "Even if they did undergo eight generations of mixing, it's hard to say whether there would have been a complete genetic blending. There may still be a fraction that are pure beach-spawners and a fraction that are pure river-spawners." Hendry understands his peers' vantage points. "We can never know for sure what happened," he says, "and so the determination of whether the hatchery rearing homogenized things becomes personal opinion. Future work will be needed to resolve this issue." The Seattle scientists also question whether the salmon came from the same source, as Hendry and colleagues claim. Naish's graduate student, Ingrid Spies, has unpublished microsatellite data suggesting that the salmon in the Lake Washington drainage may have come from many different geographic locations, not just the hatchery, Naish says. Some salmon may have been in the lake before the reintroduction, adds Kalinowski. If so, again, the genetic divergence that Hendry and his collaborators detected may have already existed long before 13 generations ago. The Seattle team is planning a series of studies to help them tease apart the pre- and post-transplantation history of the salmon. And they're quick to point out that, despite their criticisms of Hendry's study, this is not a case of "us versus Hendry." Just a few weeks ago, Hendry visited Seattle to show the researchers the spawning sites, share his data, and point out potential problems. "This whole thing really is a good example of checks and balances in science ensuring that we get to the truth," says Johnson. Evolutionists Respond As Well The crux of Hendry and his colleagues' argument for reproductive isolation is that the genetic differences between the two populations are too great for them to be interbreeding, given the number of fish who migrate freely back and forth between the two habitats. "If true, this finding would deserve considerable attention," a group of evolutionists wrote in response to Hendry's work. Even though reproductive isolation is a crucial, initial step in the speciation process, very little is known about how it works, especially in nature. However, the authors, led by evolutionary biologist Daniel Howard, New Mexico State University, describe several crucial problems with the study, most notably that the genetic differentiation that Hendry's team found is in fact too small to suggest that any sort of reproductive isolation has evolved between the two populations.2 Indeed, "part of it was just boldness on his part," Reznick says of Hendry's desire to home in on this very early phase of the speciation process. Much of what is known about speciation comes from studying the fossil record, looking at the tail end of speciation, and asking how did it happen. But evolutionary rates based on the fossil record are very slow, and "people have been deceived by thinking that the fossil record is a model for how evolution occurs," Reznick says. When calculating evolutionary rates from the fossil record, researchers divide a measure of phenotypic change, such as snout length, by a very large amount of time, Kinnison explains. But when someone calculates evolutionary rates in real, or contemporary time, the rate will often be much faster because it requires dividing a similar change by a much smaller amount of time, (up to orders of magnitude smaller). Furthermore, evolutionary change is not linear, according to Kinnison. If the focus is on the beginning stages of evolutionary change, evolutionary change seems to "pop up very quickly," Kinnison says, but it slows down over time. Hendry isn't the only researcher to have found evidence of rapid speciation. Over the last 15 to 20 years, biologists have been discovering more evidence of rapid speciation, such as insect species having evolved in as few as 200 generations. In light of this, much knowledge of speciation is now being reevaluated, says Howard. But "in our excitement to rush to a new paradigm we have to be very careful about the evidence," he says. Hendry is taking the controversy in stride. "I won't be disappointed if my conclusions ultimately prove to be wrong, which I don't think will happen. At least it will stimulate people to look for reproductive isolation over shorter time scales." Leslie Pray (lpray@nasw.org) is a freelance writer in Leverett, Mass. References 1. A.P. Hendry et al., "Rapid evolution of reproductive isolation in the wild: evidence from introduced salmon," Science, 290:516-8, 2000. 2. D.J. Howard et al., "Examining evidence of reproductive isolation in sockeye salmon." Science, 291:1853, March 9, 2001. 3. R.G. Gustafson et al., "Evolution of sockeye salmon ecotypes," Science, 291:251-2, January 12, 2001. 4. A.P. Hendry, "Something fishy," American Scientist, 89:293-4, July-Aug. 2001.

Debate Continues Over Partial Reproductive Isolation

Interested in reading more?

Become a Member of