[NPS Arrowhead] U.S. Dept. of Interior National Park Service Archeology Program
Quick Menu Features
* Sitemap * Home
common ground

History Through a Pinhole
Fall/Winter 1996, vol. 8(3/4)

Online Archive

*  The [In]Significance of Atlantic Salmon

(image) Salmon.

"There is an undefinable but compelling sensation one has upon coming across, in a clearing or along a barren road, a structure that was once a thriving enterprise."

Thomas Harding

by Catherine C. Carlson

Common folklore in New England holds that Atlantic salmon (Salmo salar), a highly prized game and food fish in both Europe and North America, were once so abundant in the rivers that early colonists could walk across the backs of the fish as they ran up the rivers in spring to spawn. There are tales that people became so tired of eating them that a law was passed requiring poor servants and laborers to be fed the fish no more than twice a week.

Based on such accounts, restoration biologists have written that "the Atlantic salmon rivaled the cod as an important and reliable source of protein to the early New England colonists."1 The anthropologist Erhardt Rostlund argued that "there is theoretical reason for thinking that Atlantic salmon, per unit area, was at least as plentiful as Pacific salmon."2

When Europeans first arrived in northeastern North America, the Atlantic salmon was reportedly found in every river not barred by impassable falls, from northeastern Labrador to the Housatonic River, and possibly into the Hudson River.3 John Smith commented in 1616 that "on the western shores of the Atlantic, it [salmon] is found from Greenland to the Hudson, but is exceedingly rare in the latter river, and never penetrates farther south."4 Common belief holds that at the turn of the 19th century, increasing pollution in the rivers (sewage, mills, etc.), weir fishing at the mouths of the rivers, and the construction of large main stem dams across the rivers (for example, at South Hadley and Turners Falls on the Connecticut River in 1794 and 1798) caused salmon to become extinct in the rivers of southern New England and severely depleted in northern New England. Fisheries biologists contend, on the basis of their interpretations of historical accounts, that the Atlantic salmon resource today is a mere remnant of the fishery prior to the introduction of dams and pollution in the rivers; for this reason, restoration programs to "bring back the salmon" have been, and continue to be, an extensive and ongoing effort supported by an effective sports fishermen's lobby.

To an anthropologist, the importance of the reportedly dense salmon runs of New England in the past is the valuable food resource that the fish would have provided the native peoples of the area long before the Europeans arrived. In the Pacific Northwest, where vast runs of Pacific salmon have survived up to the present day, aboriginal peoples harvested, preserved, and stored enormous quantities of the fish such that it enabled them to free their time from the everyday subsistence activities typical of most hunter-gatherers without agriculture. This contributed to the development of highly complex cultural and social institutions, art, ritual ceremonies, sophisticated technologies, trade networks, and permanent villages; it also supported high population densities. The basic cultural pattern of the Northwest Coast aboriginal peoples was impacted by Europeans so much later than in the New England region that most of their traditional culture survived to be described by ethnographers as late as the turn of the 20th century.

This is unfortunately not so for the Atlantic seaboard where introduced European diseases had such a devastating impact on the aboriginal peoples as early as the 15th century that little remained of their way of life, culture, and population.5 Hence, it is mostly only through archeology that we can attempt to reconstruct the cultural traditions in this region. The possibility that the New England aboriginal cultures may have had access to a salmon resource comparable to that in the Pacific Northwest is therefore of interest in archeological reconstructions, and the past presence of a salmon resource has been assumed by numerous archeologists working in the region.6

In 1980 the author began a study of the prehistoric fisheries of the Boothbay region of the Maine coast through archeological analysis of the fish bones excavated from prehistoric middens. The initial study involved the analysis of the fish remains from 21 archeological sites located in the estuaries and along the coast of Maine adjacent to the Sheepscot and Damariscotta rivers.7 Fish bone had never been analyzed for New England. (It is an uncommon specialty in zooarcheology.) One of the main components of archeological faunal analysis is to identify the bones to species, a rather technical process, but one that produces, minimally, a species list and relative abundances, eventually providing an understanding of the relative importance and abundances of certain fish species to the diet of prehistoric aboriginal peoples. Analysis of 30,000 fish bones revealed a lack of salmon bones in the site assemblages, an unusual circumstance in view of the quantity and dominance of salmon bones at similar sites in the Pacific Northwest (British Columbia). A possible explanation was that the small drainages of the Sheepscot and Damariscotta rivers did not support salmon runs (even though there were historical accounts of runs in the Sheepscot), and that cod, rather than salmon, were the principal seafood resource for the aboriginal peoples of the region.

However, the lack of salmon in the Maine sites remained an unanswered question requiring a broader regional approach to the study of prehistoric fishing.8 As research proceeded, it became apparent that there were no salmon bones in site after site in New England, although bones of numerous other species were recovered. Eventually, the analysis or review of bone remains from over 75 New England sites9 revealed only two possible salmon vertebrae at Kidder Point and Lindquist10 and possibly two at Frazer Point11 (all of which may be trout). How could this be, given that the historic accounts describe such vast quantities of the fish? Did aboriginal peoples not catch them, perhaps because they lacked suitable fishing gear, or because they did not like them as a food item? Do the bones not survive in the soil conditions of New England? Were the historical accounts of salmon grossly inaccurate, embellished fish tales? Did the salmon runs not really exist in the rivers of New England, or were they so minimal as to be undetected archeologically? If there really were no salmon, then what were the implications for archeological reconstructions based on analogies with the Pacific Northwest?

Essentially the problem was one of a discrepancy between historical accounts of vast quantities of salmon in New England rivers and the archeological record that showed virtually a complete absence of the fish. The possibility that aboriginal peoples lacked suitable technology for harvesting salmon, or that they found them disagreeable as a food item, could not be supported.12 Likewise, the suggestion that the bones do not survive in the soils of New England was also discredited because of the fact that so many other fish species with equally fragile bones have been preserved, and salmon certainly have been preserved in great quantities in archeological sites in the Pacific Northwest. How then to account for the absence?

One part of solving the problem was to review critically the primary historical documents about fish. The sources claiming vast quantities of salmon were 19th- and 20th-century syntheses and compilations made long after fishing events, and frequently based on hearsay, and were therefore subject to bias or error in interpretation due to their derivative nature. Anthony Netboy's unreferenced statement that in colonial New England, salmon "were sometimes so thick in the rivers that they overturned small boats,"13 or A. G. Huntsman's report of an 1879 account in Lake Ontario that the salmon were once so abundant that women "seined them with flannel petticoats"14 are undoubtedly examples of embellishment—the classic "fish story"—that cite nothing other than hearsay. Likewise, the story that poor laborers should not be made to eat salmon more than twice a week because of its cheap abundance was investigated by Newton Brainard and discredited: "Let us review the old story of the apprentice agreements which were supposed to have protected the poor by a clause stipulating that he was not to be required to eat salmon more than twice a week. This story was intended to show how plentiful and cheap salmon was here [Connecticut River], and has been generally accepted as true. As a matter of fact, it is an English or Scotch tradition which is not true, even in the land of its origin. As long ago as 1867 the London Field offered a reward of five pounds to anyone who would produce one of these agreements. The reward was withdrawn a year later, unclaimed."15

It was clear that the question of the abundance of Atlantic salmon in New England prior to its demise around A.D. 1800 had essentially never been thoroughly investigated through a critical analysis of primary documents.

Numerous problems can also exist for primary documents, and all historical written sources are not equally reliable. M. J. Ingram and others have described how even in the case of first editions of published manuscripts, there can be error. Authors of natural history accounts "often copied earlier [unpublished] manuscripts, mostly without acknowledgment, frequently misunderstanding and distorting the earlier materials . . . . Legends, rumors and downright fabrications were on occasion included to swell the story."16 One of the major problems in attempting an analysis of the relative abundances of salmon in the New England rivers during the colonial period is that the primary accounts are not quantitative (i.e., measured or counted systematically). Therefore the task was to make quantitative-like interpretations from highly subjective qualitative accounts that were influenced by personal and cultural biases. Perception of environmental phenomena can vary among different societies and individuals. In addition, they can change over time in relational terms, that is, in comparison with other changing environmental and/or social conditions. Therefore, in attempting to make quantitative estimates of salmon abundance from qualitative sources, it was necessary to evaluate (1) why the material was originally documented; (2) how the phenomena were categorized, and how the categories fit into modern ones; and (3) what the significance is of qualitative terms of degree, as, for example, such statements as "once salmon runs were as great as . . . ," given that many accounts are biased toward recording extreme events, such as the one good run of salmon that everyone remembers years later as the norm in the "good old days." People throughout the ages have thought that the fishing in earlier times was better than in their present day. For example, as early as 1753, Peter Kalm noted that in New England "many old people said that the difference in the quantity of fish in their youth in comparison with that of today was as great as between day and night."17

A further problem relates to the use of language because early descriptions of fish drew on a variety of vernacular terms applied before the Linnean system of binomial classification came into use after 1735. There was often a lack of vocabulary to describe particular North American species, and attempts were made to relate them to familiar Old World fish. For example, two 17th-century explorers' lengthy accounts of fish in the region—those of John Josselyn18 and James Rosierl9—cite "white salmon," which were undoubtedly shad. The latter were probably mistaken for salmon by early explorers and colonists with some frequency, creating "salmon inflation" in early and later derivative accounts. By the time that major systematic study of the natural history of the fishes of North America began in the 19th century, the Atlantic salmon runs of southern New England had long since disappeared.

Could the accounts of salmon also have been subject intentionally to embellishment? This is highly likely because the earliest writers were in reality "promoters" who would be biased in having strong motives for presenting to the folks back in the old country a considerably brighter image of New England as a place of natural abundance than was necessarily the case. Since salmon was a much esteemed fish at home, its inclusion and description was important. Atlantic salmon was a status fish to the English, and any amount of salmon occurring in New England would be praised and potentially embellished. It was esteemed by both gourmets and sports fishermen among the gentry. R.W. Dunfield remarks that Izaak Walton's The Compleat Angler began "the campaign to set both angler and salmon apart from common man and common fish."20 Walton accounted salmon the "king of fresh-water fish."21

To evaluate objectively the issue of salmon abundances, a survey was made of the primary historical documents of the 17th and 18th centuries, the time prior to extensive dam construction reputedly responsible for the salmon's demise. Since actual figures on numbers of salmon are nonexistent in these accounts, the statements regarding salmon abundance were compared with those of other fish species in order to achieve a general impression of relative abundance within the larger context of fish abundances—an issue that secondary historical syntheses have not addressed since they were reading the accounts from the perspective of assumed great salmon abundances.

In general, the primary accounts reviewed included entries by explorers and merchants ("promoters") and miscellaneous diaries and travelers' descriptions. The evidence indicated that while a number of accounts demonstrate that some salmon were present historically (and that is certainly a quantitative leap over the prehistoric record, both archeological and paleontological), they do not support the notion of abundant salmon runs in New England in the way that they are often made out to do.

For example, when the species of fish are listed or described, salmon, if mentioned at all, tend to fall towards the middle or end of a species list, suggesting their lesser significance. Some of the accounts go into considerable detail in describing each particular species of fish, and all are much more brief in their references to salmon than to other species. Furthermore, it was interesting to note that a number of sources did not even mention salmon.

Put within the context of cod, or shad, or alewives, or sturgeon, salmon appears to have been quite minor. It was not even commercially marketable, as was the case for the Pacific salmon for which a major industry was developed in the Northwest. William Douglass in 1749 reported that "this salmon [of the Merrimack and Connecticut rivers] is not of a good quantity and is not so good quality and is not so good for a market as the salmon of Great Britain and Ireland."22

Of all the secondary accounts of salmon, only one, Brainard's two page article in the Connecticut Historical Society Bulletin presents the idea that salmon were never more than a minor species. He wrote that "the sparse early records fail to indicate any excessive number of salmon in the Connecticut River, even in the early days."23 It is likely that the romantic folklore of once vast salmon runs is in fact myth and legend, a tall fish tale that has influenced all thinking about salmon in this century.

While a conclusion that salmon was not a major, but a minor, resource is interesting, it only goes part way to explain why the prehistoric archeological record of fish indicated its virtual absence. One hypothesis is that salmon did not begin to colonize New England streams until the historic period, corresponding to a more favorable period of climatic cooling known as the Little Ice Age (A.D. 1550-1800).24 At the end of this period, the climatic warming created less favorable environmental conditions for salmon, and hence their range retracted. Salmon are basically a cool water species that have a very narrow temperature tolerance range for developing eggs and smelts, and New England is the southern extent of its range. The idea that initial colonization did not occur until this time, and then only as a temporary range expansion, explains (1) the lack of salmon in prehistoric sites, (2) the apparent limited abundances of salmon historically, and (3) the extinction/depletion of the fish at the end of the 18th century. Since this is fundamentally a natural climatic explanation for both salmon appearance and disappearance as opposed to an anthropogenic one (dams and pollution), its implications for the modern salmon restoration programs should not be ignored. Fish biologists maintain that the resource can be restored by improving salmon habitat in the rivers through pollution control and construction of fish ladders.

In order to investigate a climatically induced hypothesis for salmon appearance and disappearance, environmental and climatic factors affecting range shifts and the mechanisms of migration in salmon were studied. Harsh glacial conditions during the Pleistocene (the last period of the great Ice Ages ending 12,000 years ago) resulted in an environment not conducive to salmon survival until about 10,000 years ago when modern warm Holocene environmental conditions began. Salmon must have migrated from Europe after the end of the Pleistocene, across Atlantic currents, during periods of suitable ocean conditions of temperature and food. Immediately prior to the Little Ice Age, the medieval warming period known as the Little Climatic Optimum (A.D. 900-1300) diminished sea pack ice around Iceland and Greenland.25 It is possible that salmon may have migrated during this time to Davis Straits between Labrador and Greenland, an area that today is still an important feeding ground for both European and American salmon populations. This got them to the shores of North America; then as the medieval warming waned, and the Little Ice Age set in, cooler conditions south of the Labrador coast initiated salmon range expansion into the waters of the New England region. Unfortunately, the mechanisms by which Atlantic salmon colonize new streams are poorly understood and there is little reported research on the subject. Nevertheless, salmon do have the ability, despite their innate programming, to return to the stream of their birth to spawn, to colonize new streams. Research on salmon in Swedish rivers indicates a "rate of strays" of around 2 percent, suggesting that colonization of new drainages can occur relatively rapidly in suitable environmental conditions.26

Paleontological fossil specimens of fish add empirical evidence by extending the record of northeastern fish further back in time than archeological specimens and provide evidence of the fish fauna in the region during the end of the Pleistocene. Many fossil fish specimens come from the Green Creek nodules in glacial Lake Champlain deposits near Ottawa, Ontario.27

The fossils provide information on what fish species survived the harsh glacial conditions of the Pleistocene. To date, smelt, cod, sculpin, whitefish, lake trout, lump fish, stickleback, and sturgeon are the predominant species; there is no evidence from paleontology that salmon were present during the Pleistocene, which supports the later archeological record of the Holocene (post-glacial conditions).

The issue of landlocked salmon, those populations of fish that remain in inland lakes throughout their life cycle without migrating to the sea, also required investigation. There has been some suggestion that they may have become introduced to inland lakes at the end of the Pleistocene when sea levels were higher as the ice sheets were melting and then subsequently have been trapped or "landlocked" as sea levels dropped, making them what biologists call "glaciomarine relicts," a possibility that would override the negative paleontological record. Only in four lakes in Maine (Sebago, Green, Sebec, and Grand lakes) are there natural indigenous landlocked populations; the rest have been introduced through fish stocking. In all four lakes, the fish had free access to the sea prior to the construction of dams. These fish are therefore considered to be voluntarily landlocked, a natural process poorly understood, but nevertheless, not the result of Pleistocene sea level changes.

Recent research by geneticists on the composition of European and Atlantic salmon stocks indicates no genetic markers differentiating the two geographical populations, supporting the idea that the evolutionary divergence of the two stocks is recent. The possibility that salmon colonized the rivers of New England only in the last 600 years cannot be refuted by the genetic data that support a recent origin of the fish to North America.28

In summary, Atlantic salmon are likely to be very recent colonizers to North America, particularly to New England, and their presence short and relatively insignificant. Their initial colonization and subsequent retreat may have been due largely to climatic fluctuations over the last 1,000 years from the medieval-period Little Climatic Optimum to the Little Ice Age and to the modern 19th and 20th centuries, that controlled habitat conditions in both the marine and riverine environments for migration, stream colonization, and range retraction.

It is fashionable in western culture today to view human impact on the natural environment as often the major contributing factor in environmental change. Faunal and floral extinctions, ecological "imbalance" due to exotic species introductions, deforestation, greenhouse gases for climate change, and even in the archeological literature blaming the extinction of the North American Pleistocene megafauna on over-hunting by Paleoindian hunters—are just a few examples. It is also fashionable to suggest that science can "fix" or undo anthropogenic environmental change. While humans unarguably have impacts on the natural environment, and have done so for as long as their four-million-year evolutionary history on the earth, the environmental ideology that grants omnipotence to humans over the environment can blind us from recognizing that natural environmental fluctuations in climate and species distributions or extinctions are probably more substantive on a long-term scale than human induced ones. One needs only to look at the paleoenvironmental, paleobotanical, and zooarcheological records of the past to fully appreciate this.

This article ultimately is concerned with how the issue of disappearing Atlantic salmon in southern New England, and its considerable depletion in the north, is an example of too great a focus on anthropogenic environmental change. Today's fish and wildlife managers appear to have largely ignored the paleoenvironmental databases that present long-term records of climatic change in concert with animal and plant species range changes, and even total extinctions, because of their preoccupation with the effects of industrialization. While biologists such as D. W. Lufkin have stated that "the circumstances causing the demise of Salmo salar are relatively simple to identify. . . [as] dams, pollution, logging practices, and over-fishing,"29 this article argues that causes behind its demise are more complex, with ecological and climatological bases. If pollution and dams were the major cause of their extinctions, then why were the runs not made extinct on the Penobscot, a heavily dammed and polluted river in Maine? Also unaccounted for is why salmon runs became extinct downstream of the dams on the Connecticut River. The general lack of success in salmon restoration programs over the last two centuries, despite fish ladders and habitat improvement, suggests a more fundamental ecological cause for impoverished salmon runs in New England than an anthropogenic one.

We also need to examine more closely how social and cultural values can fashion a natural creature, in this case the salmon, in ways that identify it with high-ranked social positions such that it unwittingly influences our thinking in everything from the establishment of fisheries societies and restoration facilities to archeological reconstructions of prehistoric societies. The romantic allure of the king of fish colors the visions of prestigious sportsmen, biologists, and politicians. The political correctness of "environmental awareness," in which salmon has become the symbol for clean rivers, whether justified or not, becomes a factor in the judgments being made. The lowly codfish appears to be a more appropriate fish symbol for New England and one that is presently environmentally threatened; the politicians supporting salmon restoration have apparently forgotten about the "great cod" that hangs in the halls of the Massachusetts statehouse and its historical significance.

Catherine C. Carlson is an assistant professor of archeology at University College of the Cariboo, Kamloops, British Columbia. Her research specialties include ichthyo-zooarcheology and historical archeology of the native peoples of North America. For more information, she may be contacted at University College of the Cariboo, Department of Anthropology, Kamloops, BC, V2C 5N3 Canada, e-mail ccarlson@cariboo.bc.ca.

The author is indebted to Neal Salisbury, Boyd Kynard, H. Martin Wobst, the Department of Anthropology at the University of Massachusetts, George Nicholas, and especially Dena Dincauze for guidance and assistance with this article and the doctoral dissertation on which it is based. Any errors, omissions, or limitations are strictly the author's own.

This article originally appeared in New England's Creatures: 1400-1900, volume 18 of the Annual Proceedings of the Dublin Seminar for New England Folklife, published by Boston University in June of this year. Reprinted with permission. Copies of this volume and others can be obtained by writing Boston University Scholarly Publications, 985 Commonwealth Ave., Boston, MA 02215. The price of the volume is $15 plus $1.50 for shipping and handling.


1. Dan C. Kimball and Lawrence W. Stolte, "Return of the Atlantic Salmon," Water Spectrum (Fall 1978): 1.

2. Erhardt Rostlund, Freshwater Fish and Fishing in Native North America (Berkeley: University of California Publications in Geography, vol. 9, University of California Press, 1952), p. 25.

3. Henry B. Bigelow and William C. Schroeder, Fishes of the Gulf of Maine (Bulletin of the U.S. Department of Fisheries 40, Washington, D.C., 1953).

4. John Smith, A Description of New England (1616; reprinted in Peter Force, Tracts and Other Papers, vol. 2, Washington, D.C., 1838), p. 137.

5. Catherine C. Carlson, George J. Armelagos, and Ann Magennis "Impact of Disease on the Precontact and Early Historic Populations of New England and the Maritimes," in Disease and Demography in the Americas, ed. John W. Verano and Douglas H. Ubelaker (Washington: Smithsonian Institution Press, 1992), pp. 141-54.

6. Catherine C. Carlson, "The Atlantic Salmon in New England Prehistory and History: Social and Environmental Implications." (Ph.D. diss., Anthropology, University of Massachusetts at Amherst, 1992), p. 4. This paper forms part of a much larger study documented in this dissertation.

7. Catherine C. Carlson, "Maritime Catchment Areas: An Analysis of Prehistoric Fishing Strategies in the Boothbay Region of Maine" (MS. thesis, Institute for Quaternary Studies, University of Maine at Orono, 1986).

8. Catherine C. Carlson, "Where's the Salmon?: A Reevaluation of the Role of Anadromous Fisheries to Aboriginal New England," in Human Holocene Ecology in Northeastern North America, ed. George P. Nicholas (New York: Plenum Press, 1988), pp. 47-80.

9. Carlson, "The Atlantic Salmon in New England Prehistory and History."

10. Arthur Spiess and Mark Heddon, "Kidder Point and Sears Island in Prehistory," Occasional Publications in Maine Archaeology 3 (Augusta: Maine Historic Preservation Commission, 1983); David Yesner, Nathan Hamilton, and Richard Doyle, Jr., "‘Landlocked' Salmon and Early Holocene Lacustrine Adaptations in Southwestern Maine," North American Archaeologist 4, no.4 (1983): 307-33.

11. Catherine C. Carlson, "Report on the Analysis of the Frazer Point Site Fish Remains" (Department of Anthropology, University of Maine at Orono, 1981).

12. There are numerous archaeological sites and ethnohistorical accounts with evidence of aboriginal fishing technology suitable for salmon fishing, such as weirs, fish spears, and canoes. See Carlson, "The Atlantic Salmon in New England Prehistory and History," pp . 131-33.

13. Anthony Netboy, The Salmon: Their Fight for Survival (Boston: Houghton Mifflin, 1974), p. 521.

14. A. G. Huntsman, "Why Did Ontario Salmon Disappear?" Transactions of the Royal Society of Canada 5, ser. 3, vol. 38 (1944): 83-102.

15. Newton C. Brainard, "Connecticut River Salmon," Connecticut Historical Society Bulletin 16, no.2 (1951): 9-11.

16. M. J. Ingram, D. J. Underhill, and G. Farmer, "The Use of Documentary Sources for the Study of Past Climates," in Climate and History, ed. T. M. Wigely, M. J. Ingram, and G. Farmer (Cambridge: Cambridge University Press, 1981), p. 195.

17. Peter Kalm, Travels into North America (Stockholm, 1753; revised from the Swedish and reprinted in Peter Kalm's Travels in North America, vol. 1, ed. A. B. Benson (New York: Wilson-Erickson, 1937), p. 155.

18. John Josselyn, An Account of Two Voyages to New England (London, 1675; reprinted in Collections of the Massachusetts Historical Society 3, ser. 3, 1833).

19. James Rosier, A True Relation of the Most Prosperous Voyage Made this Present yeere, 1605, by Captaine George Waymouth (London, 1605; reprinted in Early English and French Voyages, 1534-1608, ed. Henry S. Burrage, New York: Charles Scribner's Sons, 1932), pp. 353-94.

20. R. W. Dunfield, "The Atlantic Salmon in the History of North America," Canadian Special Publication of Fisheries and Aquatic Sciences 80 (Ottawa: Department of Fisheries and Oceans, 1985):4.

21. Izaak Walton, The Compleat Angler (1653; reprint ed., New York: Weathervane Books), p. 127. Author's italics.

22. William Douglass, A Summary, Historical and Political, of the first Planting, Progressive Improvements, and Present State of the British Settlements in North America, Volume 1 (London: Rogers and Fowle, 1760; reprint ed., New York: Arno Press, 1972).

23. Brainard, "Connecticut River Salmon," p. 10.

24. Carlson, "The Atlantic Salmon in New England Prehistory and History," p. 200.

25. Hubert H. Lamb, "Climatic Variation and Changes in the Wind and Ocean Circulation: The Little Ice Age in the Northeast Atlantic," Quaternary Research 11, no. 1, (1979): 1-20.

26. G. Stahl, "Genetic Differentiation among Natural Populations of Atlantic Salmon (Salmo salar) in Northern Sweden," in Fish Gene Pools, ed. N. Ryman (Ecological Bulletin 34, Stockholm, 1981), p. 102.

27. Donald E. McAllister, Stephen L. Cumbaa, and C. R. Harington, "Pleistocene Fishes (Coregonus, Osmerus, Microgadus, Gasterosteus) from Green Creek, Ontario, Canada," Canadian Journal of Earth Sciences 18 (1981): 1356-64.v

28. William S. Davidson, Timothy P. Birt, and John M. Green, "A Review of Genetic Variation in Atlantic Salmon, Salmo salar L., and Its Importance for Stock Identification, Enhancement Programmes and Aquaculture," Journal of Fisheries Biology 34 (1989): 547-60.

29. D. W. Lufkin, "Environmental Quality," New England Atlantic Salmon Restoration Conference; IASF Special Publication Series 6 (St. Andrews, New Brunswick, 1975):21.