NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Committee on Atlantic Salmon in Maine. Genetic Status of Atlantic Salmon in Maine: Interim Report from the Committee on Atlantic Salmon in Maine. Washington (DC): National Academies Press (US); 2002.

Cover of Genetic Status of Atlantic Salmon in Maine

Genetic Status of Atlantic Salmon in Maine: Interim Report from the Committee on Atlantic Salmon in Maine.

Show details

3Current State of Atlantic Salmon in Maine


Maine has the last of the wild Atlantic salmon populations of the United States. Historically, 300,000 to 500,000 adults probably entered U.S. rivers each year (Stolte 1981, Beland 1984). An analysis of the distribution of U.S. Atlantic salmon by the Biological Review Team (1999)—using zoogeographic information to construct ecological provinces, including aquatic ecological units (Bailey 1995, Maxwell et al. 1995)—suggests that historic Atlantic salmon populations were divided into at least three distinct groups of populations: one in Long Island Sound, with eight major river populations, including the Connecticut River; one in Central New England, including the Merrimack River; and one in the Gulf of Maine, including the eight current DPS rivers.

The Long Island Sound populations were extirpated by the early 1800s (Meyers 1994), and the central New England populations by the mid-1800s (Stolte 1981, 1994). The remaining Gulf of Maine populations might have produced 100,000 adults per year, but these numbers have not been seen since the late 1800s. Since the late 1960's, the number of returning adults has been only about 5,000 or fewer. Fewer than 1,000 adults returned each year during the beginning of the period and in 1999 and 2000 (Figure 3). The estimated total return for the eight streams comprising the Gulf of Maine DPS was only 75–110 adults in 2000 (John Kocik, National Marine Fisheries Service, presentation to NRC committee, June 12, 2001).

FIGURE 3. Documented adult Atlantic salmon returns to all Maine streams (rod and trap catches combined).


Documented adult Atlantic salmon returns to all Maine streams (rod and trap catches combined). These numbers represent minimum adult returns. Source: E.Baum, Atlantic Salmon Unlimited, unpublished material, 2001. Printed with permission of the author. (more...)


Augmentation of wild populations with hatchery releases began soon after the first major declines in Atlantic salmon in the 1870s. It is convenient to divide Maine's stocking history into three phases: historical (1871 to 1970); recent (1970 to 1992); and contemporary (1992 to present) (Figure 4). Details can be found in Baum (1997) and the report of the Biological Review Team (1999).

FIGURE 4. Number of Atlantic salmon stocked in Maine rivers, 1871–1995.


Number of Atlantic salmon stocked in Maine rivers, 1871–1995. Source: Adapted from Baum 1997.

Historical Period (1871–1970)

This period began with the release of parr from Ontario into the Sheepscot River (probably from Lake Ontario1 ). As the import cost was high, a hatchery was established at Craig Brook, and the Penobscot River became the primary source of eggs for artificial production for the next 50 years. As runs began to decline in the Penobscot River, during the 1920s and 1930s, Canadian populations (primarily from the Miramichi and Gaspe rivers) became the main sources of eggs for the Craig Brook Hatchery. By the 1940s, Canadian eggs were largely replaced by eggs from the Machias and Penobscot rivers and, to a lesser extent, from the Dennys River. In the 1950s and 1960s, declines in Penobscot runs once again resulted in the use of Canadian Miramichi salmon, supplemented with eggs from the Machias and Narraguagus brood stocks. Throughout this period, hatcheries stocked only fry and parr, and the results were poor. By the late 1960s, stocking had switched largely to smolts, which was followed by the recovery of the salmon run in the Penobscot. The last of the Miramichi-origin salmon were released in Maine in 1968, when the Penobscot population became the main supplier, but Maine salmon could have been infiltrated with Canadian genetic material by then.

Recent Period (1970 to 1992)

This period was typified by an emphasis on producing and releasing smolts, assisted by the Green Lake National Fish Hatchery, and a focus on using Maine sources, primarily the Penobscot but also the Narraguagus and Machias rivers. There was growing recognition that wide geographic movements of genetic material were less than ideal, often resulting in poor returns and creating problems for local adaptation (Ricker 1972, Hindar et al. 1991, Waples and Do 1994, NRC 1996, Utter 1998). By the mid-1970s, the Penobscot brood stock supported essentially all the artificial propagation needs of Maine. During this recent period, Atlantic salmon throughout Maine were managed as a single population for stock-release purposes.

Contemporary Period (1992 to present)

Since 1991, stocking has been river-specific and based on a conservation-hatchery program. The Craig Brook National Fish Hatchery was converted from a single-stock smolt-production facility to a multiple-stock fry-production facility (BRT 1999). Brood-stock collections began with returning adults in 1991, but insufficient numbers led to parr collections in 1992 (Buckley 1999). The current protocol involves raising (river-caught) parr to adulthood, mating them according to approved protocols of the Maine Technical Advisory Committee (Beland et al. 1997, Copeland et al. 1998), rearing the embryos, and releasing the fry (usually before they begin feeding). Fry stocking began in 1992, and stocking levels had reached target density levels by 1997. There is no indication of a successful increase in returning adults as yet. The shift to a river-specific management and propagation system was based on the premise that naturally spawning fish from even more local waters would provide the best hope for success.

Consequences of Stocking Protocols

Two points about the stocking protocol are important. First, despite 130 years of stocking, using a variety of life stages (fry, parr, smolts, and even adults) and releasing about 120 million Atlantic salmon, the systematic decline in run sizes has not been reversed. That raises the question of whether hatchery stocking has ever had a substantial impact on populations of Atlantic salmon in Maine. Second, the consequences of extensive stocking on the genetic makeup of Maine's salmon are unknown. Has the release into Maine's rivers of large amounts of genetic material, some from outside the region and always with the potential for genetic change through hatchery selection, overwhelmed the gene pool of the aboriginal Gulf of Maine populations?

The Biological Review Team (1999) and Baum (1997) concluded that the hatchery fish have not displaced the local gene pool because of the poor success of historical hatchery stocking and the likelihood that Canadian fish were poorly adapted to Maine streams. In addition, there is now considerable evidence that stocked fish do very poorly (Lieder et al. 1990, Fleming and Gross 1993, Skaala et al. 1993, Fleming et al. 1997, Petersson and Järvi 1997). A review of 31 studies of incursion of hatchery genetic material into wild populations (Fleming and Petersson 2001) reported that 14 studies showed little or no evidence of incursion, despite prolonged hatchery releases. Many of those studies involved anadromous populations. In contrast, 16 of the 17 studies showing an incursion involved nonanadromous populations, suggesting that anadromous populations are more resistant to introgression (see also Utter 2000). If that is true, genetic infiltration of Maine salmon by hatchery releases would be minor. In addition, from 1970 to the present, an average of 84% of stocked Atlantic salmon were of natural origin in the eight Maine DPS rivers (USASAC 1999). The Gulf of Maine populations might have maintained their genetic integrity largely through natural processes. Although hatchery introductions must have had some impacts, they may have been small.


The second-most dramatic change involving Atlantic salmon in Maine (the collapse of wild populations being the first) is the production of farm-raised Atlantic salmon through aquaculture. Although Maine produced only about 16,400 metric tons (t) of farm salmon in 2000, compared with a global production of more than 800,000 t (Working Group on North Atlantic Salmon 2001), its production was zero as recently as 1986. Five freshwater hatcheries across the state provide smolts for about 600 net-pens floating in sheltered areas, where the fish grow on food pellets broadcast into the pens and reach market size in about 18 months.

If production grows in Maine, additional farm sites in the protected waters along Maine's coast will be needed. The current sites apparently do not have room for additional production. Production is currently concentrated in Washington and Hancock counties—referred to as Downcast Maine—an area that includes five of the eight DPS rivers. Aquaculture in the area could potentially be affected by ESA provisions.

The genetic issue for wild runs is that salmon escape from aquaculture and enter the rivers to spawn (Hansen et al. 1991, Hindar et al. 1991, Carr et al. 1997, Youngson et al. 1997, Gross 1998). These farmed fish are of a different genetic makeup, because they include normative strains, because of directed selection by the breeders for traits valuable to the industry (e.g., growth rate, fat content, disease resistance, and delayed maturity), and because of the inadvertent selection by the novel environment (e.g., reduced fright response, disease resistance, and altered aggressive behaviors) (Gross 1998, Fleming and Einum 1997, Johnsson et al. 2001). Those same traits might not be adaptive in the wild.

It is difficult to know what genetic lineages are being used in Maine aquaculture. The Biological Review Team (1999) concluded that there were three brood-stock lines: Penobscot (Maine), Saint John (Canada), and Land-catch (Scotland). The latter is composed primarily of Norwegian strains. The predominant strain was developed from populations from 41 rivers and locations, and was supported in development by the Norwegian government (Gjedrem et al. 1991). Imported sperm of Norwegian origin, via Iceland, also has been used in recent years. Baum (1998) suggested that European genetic material permeates approximately 30–50% of farm salmon in Maine. The genetic blend can be expected to evolve as time passes.

Farm salmon escape from containment at all life stages, from embryos through adults, despite efforts to design and maintain escape-proof containers. No accurate data are available on escapement in Maine, but data on the numher of individuals entering rivers as adults and some data on hatchery escapes are available. Gross (2002) estimated 3% escapement from similar facilities in British Columbia and elsewhere (British Columbia Environmental Assessment Office 1997). Based on the number of fish being raised in Maine waters, 3% escapement in Maine would translate into about 180,000 escapees per year from net pens. Continued improvements are being made that will reduce, but not eliminate, the number of escapees. An escape rate as low as 0.17%, which would be impressive, would still pro vide 10,000 escapees per year, 100 times the number of adults that returned to spawn in Maine's eight DPS streams in 2000.

These escapees might not have much impact on healthy wild populations, because farm (pen-raised) salmon have shown competitive inferiority in the wild (Fleming et al. 1996, 2000). However, because of the low numbers of wild adults returning to spawn in recent years, farm salmon represent a large proportion (100% in some years) of the adults entering the rivers to spawn. The effect might be ameliorated to some extent by the precocious wild parr in the streams and by the low reproductive success of farm adults (Fleming et al. 1996). In experimental facilities, farm males had only 1–3% of the success of wild males, and farm females had only 20–40% of the success of wild females, with most matings involving wild males (Fleming et al. 1996). In addition, Fleming et al. (2000) showed that farm salmon introduced experimentally into a wild population had only 16% of the success of wild salmon in producing recruits. Thus, it is possible for wild populations to “resist” genetic infiltration by farm fish, but that potential drops as the number of wild fish becomes small, relative to the number of farm fish. Even a 10:1 adaptive advantage for wild salmon might not be sufficient to overcome a 100:1 numerical advantage for aquaculture escapees. It remains unclear to what degree farm salmon have infiltrated wild populations genetically, or conversely, how resistant wild salmon have been to genetic infiltration. Based on samples taken in 1994–1998, genetic infiltration of farm fish into wild Maine populations was minimal (King et al. 1999). However, if salmon farming in Maine expanded further, the numerical impact (among likely spawners) of aquaculture escapees would have the potential to become significant.



The Atlantic salmon in Lake Ontario, extirpated by about 1900, might have consisted of two runs, those that remained in the lake (landlocked) and those that went to the ocean (anadromous). They might have spawned in the same or different tributaries of the lake (Kendall 1935, Parsons 1973, Scott and Crossman 1973). It is not known whether the fish used for stocking purposes in Maine were lake- or ocean-run fish or whether these two possible types were genetically different from each other.

Copyright 2002 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK223898


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (700K)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...