Summary of Committee’s Interim Report (Excerpted from NRC 2002a)
Atlantic salmon in Maine, once abundant but now seriously depleted, were listed as endangered under the federal Endangered Species Act (ESA) in November 2000. The listing covers the wild fish in eight Maine rivers as a single “distinct population segment.” The controversy in Maine that accompanied the listing led Congress to request the National Research Council’s (NRC’s) advice on the science relevant to understanding and reversing the declines in Maine’s salmon populations. The charge to the NRC’s Committee on Atlantic Salmon in Maine included an interim report focusing on the genetic makeup of Maine Atlantic salmon populations. This is the interim report. Understanding the genetic makeup of Maine’s salmon is important for recovery efforts, because the degree to which populations in Maine differ from adjacent populations in Canada and the degree to which populations in different Maine rivers and tributaries differ from each other affect the choice of recovery options that are most likely to be effective. This report focuses only on questions of genetic distinctiveness. The committee’s final report will address the broader issues, such as the factors that have caused Maine’s salmon populations to decline and the options for helping them to recover.
Naturally reproducing populations of Atlantic salmon occur in rivers and streams from southwestern Maine to northwestern Europe. Historically, they were found in the Hudson River in New York and north and
east to the Canadian border but today are found only in Maine, from the lower Kennebec River to the Canadian border. The populations have declined drastically, from perhaps half a million adults returning to U.S. rivers each year in the early 1800s to about 1,000 in 2000.
Salmon spawn in freshwater, where the young hatch and grow for a year or 2 before migrating to sea. At sea, they grow faster in the rich marine environment and then return to the rivers where they hatched (called natal streams) to spawn. Most fish die after spawning, but some return to the ocean, and some of those return to spawn again. Adults return to their natal streams; only about 2% stray to other (usually nearby) streams.
The occasional straying is probably important evolutionarily, because it allows recolonization of a stream if the local population dies out and provides for small infusions of new genetic material for evolutionary adaptation to changing conditions. Their homing provides an opportunity for the salmon to adapt to environmental conditions in their natal streams. This complex life history pattern makes salmon vulnerable to environmental disruptions both at sea and in fresh water. It also complicates the understanding of the genetic makeup of salmon populations because of the relationship between local adaptations and exchange of genetic material through occasional straying.
HATCHERIES AND AQUACULTURE
Augmentation of wild populations of Maine salmon with hatchery releases began in the early 1870s. At first, young fish were obtained from Lake Ontario, and then the Craig Brook Hatchery, using eggs from Penobscot River fish in Maine, was the stocking source. By the 1920s, Canadian eggs were being used, followed in the 1940s by eggs from the Machias, Penobscot, and Dennys rivers of Maine. In the 1950s and 1960s some eggs of Canadian origin were used again, but by the late 1960s, eggs from Maine’s Machias, Narraguagus, and Penobscot rivers were used. Fish reared in hatcheries derived from Penobscot River fish were used until late 1991, when the practice of river-specific stocking was adopted. The protocol used since involves catching young, actively feeding fish (parr) in the river, rearing them to maturity in the hatchery, mating them, and releasing the resulting fry into the rivers before they start to feed.
In addition to stocking, which at least until 1992 added to rivers many fish (and eggs) whose genotypes did not reflect adaptation to the local environment, aquaculture (farming) of Atlantic salmon began in Maine in the 1980s, the first fish for market being produced in 1987. Derived in part from European Atlantic salmon, the genetic strains used for fish farming are even more different from native strains than hatchery strains. Farm
fish escape at all life stages, despite the efforts of producers to prevent escapes. In some years and in some rivers, more escaped farm fish return to spawn than wild fish. The impact of escapees on the genetics of wild populations is not well documented in Maine, but both hatchery- and pen-reared fish compete poorly in rivers with wild fish in other areas that have been studied. However, because there are so many escaped farm fish compared with wild fish in some rivers, some impact is likely to have occurred, especially as farm production has increased in recent years.
The addition of so many nonwild genotypes from hatcheries and possibly from aquaculture escapees has led some to conclude that the fish returning to spawn in Maine’s rivers could not possibly represent anything more than some nonnative mix of genotypes from Europe, Canada, and Maine. If that were true, then options for conservation might be considerably different from those that might be undertaken if the wild fish in Maine were distinct, and that is why it is important to understand the genetic makeup of the wild salmon populations in Maine.
THE DATA ON GENETICS OF MAINE SALMON
The committee’s focus in this interim report is on assessing how Maine salmon populations differ from other Atlantic salmon populations and among themselves. The committee has addressed the question at three levels. First, are North American Atlantic salmon genetically different from European salmon? Second, are Maine salmon distinct from Canadian salmon? Third, to what degree are salmon populations in the eight Maine rivers mentioned in the ESA listing distinct from each other?
Much of the evidence on genetic distinctiveness is based on laboratory analyses of variations in the gene products (proteins) and in the genetic material (DNA) itself. The preliminary evidence indicated distinctiveness at all three levels, and that indication led to the ESA listing. However, the evidence has been questioned on statistical, methodological, biological, and other grounds, and so it bears close evaluation.
The committee evaluated the original evidence, including technical reports, as well as newly published information. It reviewed earlier studies and studies of similar situations involving other locations and some other species of fishes in the salmon family and considered the questions raised about the evidence on Maine salmon. In addition, the committee considered the effect that overlapping generations1 of salmon might have on the evidence.
The evidence on distinctiveness of Maine salmon includes statistical studies on a variety of protein and DNA markers. The statistical significance of the results is so strong and the departures from random expectations are so large that the committee judged the results to be persuasive. Many appropriate questions have been raised about the evidence, and the most recent studies have benefited from criticisms of earlier work. Those criticisms could still be used to improve future work, but the general conclusions are so strongly supported by the evidence that they are not invalidated by imperfections in the data collections or analyses.
The committee concludes that North American Atlantic salmon are clearly distinct genetically from European salmon. In addition, despite the extensive additions of nonnative hatchery and aquaculture genotypes to Maine’s rivers, the evidence is surprisingly strong that the wild salmon in Maine are genetically distinct from Canadian salmon. Furthermore, there is considerable genetic divergence among populations in the eight Maine rivers where wild salmon are found.
The heavy stocking of salmon in Maine’s rivers and streams has included periods of heavy Canadian stocking, interspersed with strictly Maine stocking. Exactly how much Canadian genetic material has infiltrated Maine salmon populations is impossible to judge at this date.
It is thus appropriate to ask whether wild salmon in Maine reflect only (or mainly) the result of decades of hatchery stocking. That seems unlikely, because if that were so, Maine salmon should be more similar to Canadian salmon than they are. In addition, if their genetic makeup were largely due to stocking of non local salmon broadly across Maine’s rivers, salmon populations within Maine would be genetically much more similar than they are. A related question is whether the genetic differences among the fish in the various Maine streams reflect natural processes that occur in watersheds that are connected in networks. More specifically, the issue concerns the relative importance of natural selection over long periods, which influenced the differentiation of Maine’s original salmon populations, versus recent genetic drift (sampling effects) caused by small populations. This question cannot be answered at present, but the pattern of genetic variation seen among Maine streams is similar to patterns seen elsewhere in salmon and their relatives where no stocking has occurred. Maine streams have salmon populations that are genetically as divergent from Canadian salmon populations and from each other as would be expected in natural salmon populations anywhere else in the Northern Hemisphere.