[Jordan and Evermann, 1896-1900, p. 523.]
The smelt is distinguishable from all other fish common in our waters by its slender form, combined with a long pointed head, large mouth, a small but evident adipose fin standing above the rear part of the anal, and a deeply forked tail. The location of its dorsal fin above the ventrals instead of in front of them, and its much larger mouth and small eye separate it from the argentine. The large, fang-like teeth on the smelt's tongue, its larger scales (of which there are about 75 along each row on the sides, all alike in the two sexes), its shorter adipose fin, its narrower pectoral fins, that its lower jaw projects only slightly beyond the upper and its scales slip off very easily, obviate any danger of confusing [page 136] it with the capelin. The body of the smelt is only about one-fifth as deep as long (exclusive of caudal fin), with broadly rounded back but flattened enough sidewise to be egg-shaped in cross section. It is deepest about its mid-length, tapering thence toward the head and toward the tail (at least in fat fish), whereas the capelin is of nearly uniform depth from gill opening to anal fin (p. 134). Its mouth gapes back of the eye.
Printed accounts of the smelt usually credit it with a peculiar "cucumber" odor, and smelt fishermen often speak of a trace of this, but it is so faint that we have never noticed it though we have caught and handled many.
Transparent olive to bottle green above, the sides are of paler cast of the same hue but each with a broad longitudinal silvery band. The belly is silvery, while the fins and body are more or less flecked with tiny dusky dots. This color pattern is shared by another slender little fish, the silverside (p. 302), but the latter has two large dorsal fins, so there is no danger of confusing the smelt with it.
Smelt grow to a maximum length of about 13 or 14 inches. Few, however, are more than a foot long, and adults run only about 7 to 9 inches. Smelt weigh from 1 to 6 ounces according to size and fatness.
The smelt is an inshore fish, confined to so narrow a zone along the coast that none has ever been reported more than a mile or so out from the land, or more than two or three fathoms in depth, while many spend the whole year in estuarine situations.
Young smelts certainly, and old ones probably, travel in schools, which are mostly composed of fish of a size, hence probably are the product of one year's hatching, and they live pelagic, not on the bottom, though confined to shoal water.
Most authorities describe the smelt as feeding on small crustaceans, which is correct as far as it goes, for shrimp (both decapods and mysids) and gammarids are probably its favorite food, and shrimp were long considered the best smelt bait. But it has been found that pieces of "sea worms" (Nereis) are more attractive to the larger smelt (at least in Massachusetts Bay). Small fish also form an important item in the diet. We have, for example, found smelts taken in the Sheepscot River in May packed full of young herring, and have caught many with small mummichogs (Fundulus) as bait; while cunners, anchovies, launce, sticklebacks, silversides, and alewives have been identified from smelt stomachs at Woods Hole. The Woods Hole diet list also includes shellfish, squid, annelid worms (Nereis), and crabs, but even as greedy a fish as the smelt ceases to feed during its spawning visits to fresh water. Young smelt depend chiefly on copepods and on other minute pelagic crustaceans. Smelt fishermen are familiar with the fact that a smelt approaches a bait slowly, then stops, and appears to suck it in. If the smelt take their living prey in this same way, it is somewhat of a mystery how they succeed in capturing animals as active as shrimps and small fish.
Smelt, like alewives, shad, and salmon, make their growth in salt water, but run up into fresh water to spawn.
The summer habitat of the smelt varies off [page 137] different parts of the coast of the Gulf, depending on the summer temperature of the water and perhaps on the food supply. Most of them desert the harbors and estuaries of the Massachusetts Bay region and along the southern coast of Maine during the warmest season. But it is probable that they move out only far enough to find cooler water at a slightly greater depth, and a few may be found in harbors through the summer. Smelt, for instance, are caught then in Cohasset Harbor in some years, but not in others; and east of Penobscot Bay, where the surface temperature does not rise so high as off Massachusetts, smelt are to be found in the harbors, bays, and river mouths all summer, and are sometimes taken in numbers then in the weirs.
Adult smelt gather in harbors and brackish estuaries early in autumn, where smelt fishing with hook and line is in full swing by October. The schools then tend to move into the smaller harbors on the flood tide, and out again on the ebb, especially if the tidal current is strong, as it is in Cohasset, a locality with which we are familiar. But some smelt remain over the ebb in the deeper basins. And some of them have run as far as the head of tide by the time the first ice forms in December. Most of them winter between the harbor mouths and the brackish water farther up; the maturing fish commence their spawning migration into fresh water as early in the spring as the ice goes out of the streams and the water warms to the required degree.
Temperature observations by the Massachusetts Commission show that the first smelt appear on the spawning beds in Weir River, a stream emptying into Boston Harbor, when the temperature of the water rises to about 40-42°. This may take place as early as the first week in March or as late as the last, about Massachusetts Bay, depending on the forwardness of the season and on the particular stream. The chief production of eggs takes place in temperatures of 50-57°, and spawning is completed in Massachusetts waters by about the 10th or 15th of May, year in and year out. East of Portland, smelt seldom commence to run before April, and continue through May. In the colder streams on the southern shores of the Gulf of St. Lawrence they do not spawn until June. On the other hand, they may commence spawning as early as February along the southern New England coast west of Cape Cod.
As a rule smelt do not journey far upstream; many, indeed, go only a few hundred yards above tidewater, whether the stream be small or large. Thus Dr. Huntsman informs us that the smelt that enter the estuary of the Stewiacke River, Nova Scotia (a tributary of the lower Shubenacadie, near the head of the Bay of Fundy) spawn only in the tidal part. And some spawn in slightly brackish water in certain ponds back of barrier beaches (e. g., Straits Pond, Cohasset, Mass.). But flooding with salt water, which sometimes happens, kills the eggs.
The adult smelts return to salt water immediately after spawning to spend the summer either in the estuary into which the stream in which they spawn empties or in the sea close by. On the Massachusetts coast north of Cape Cod all the spent fish have left fresh water by the middle of May, earlier in some years. On the Maine coast, too, a good proportion of the spent fish are in salt water by the first weeks in May; thus we have seen a bushel of large smelt taken in a weir at Cutler (near the mouth of the Grand Manan Channel) as early as May 4.
The eggs average about 1.2 mm. (0.05-inch) in diameter and they sink to the bottom, where they stick in clusters to pebbles, to each other, or to any stick, root, grass, or water weed they chance to touch. According to the Manual of Fish Culture a female weighing as little as 2 ounces will produce between 40,000 and 50,000 eggs; The eggs of the closely allied European smelt (Osmerus eperlanus) hatch in 8 to 27 days, according to temperature, and the incubation period of the American fish is the same probably, for smelt eggs are reported as hatching in 13 days at the Palmer (Mass.) hatchery.
The smelt has proved a favorable fish for artificial hatching and large numbers of fry are so produced yearly in Massachusetts, the eggs being [page 138] taken in Weir River, just mentioned, and it has proved possible to re-establish smelt by introducing the eggs or fry into streams from which it has been extirpated. For example, good smelt fishing was reported in "Poorhouse Brook," Saugus, a tributary of Boston Harbor, three years after the stream was stocked with eggs, and attempts have been similarly successful on Long Island, N. Y. Maintenance of the stock is a question either of providing accessible spawning grounds of sufficient extent, or of making up for lack of such by artificial propagation.
The precise season when young smelt go down to the sea in the Gulf of Maine streams is yet to be learned; probably early in summer. We seined several hundred fry, 1¼ to 13/5 inches long, October 1, 1924, on a beach of Mount Desert Island, evidence that the rate of growth is about the same for our smelt during its first summer and autumn as for the European, i. e., to a length of 1¾ to 2¾ inches.
Most of the smelt evidently do not spawn until they have passed an autumn, a winter, a summer, and a second winter in salt water.
East coast of North America from eastern Labrador, Strait of Belle Isle, and the Gulf of St. Lawrence southward regularly to New Jersey, and reported to Virginia; running up streams and rivers to spawn. Smelt, also, are landlocked naturally in many lakes and ponds in New Hampshire and in Maine, also in Lake Champlain, and in various Canadian lakes.
The smelt is a familiar little fish around the entire coast of the Gulf of Maine, but varies greatly in abundance from place to place according to the accessibility of streams suitable for spawning, from which it seldom wanders far alongshore. Smelt are plentiful, still, all around the inner parts of Massachusetts Bay and its tributary harbors, though many of the local streams are barred to them now; thence northward and eastward all along the coast of Maine; tolerably so in the region of Passamaquoddy Bay (catch for Charlotte County, New Brunswick, 7,400 pounds in 1945), and more so along the western shore of Nova Scotia (60,100 pounds for Yarmouth County in 1945). But they are less plentiful passing inward along the Nova Scotia shore of the Bay of Fundy, as illustrated by catches in 1945 of 20,100 pounds for Digby County, but only 7,600 pounds for Kings County, 2,000 pounds for Hants, and 1,800 for Colchester (covering the Minas Basin region). So few smelt exist along the New Brunswick side of the Bay, inward from the Passamaquoddy region, that none at all were reported for that stretch of coast in any year during the period 1939-1945. Doubtless this scarcity up the Bay is "due to absence of streams suitable for spawning, and the general turbidity of the water," as Jeffers has remarked.
Smelt once were so plentiful in the Back Bay at Boston (now mostly filled in) that "distinguished merchants of lower Beacon Street might be seen, at early hours, eagerly catching their breakfast from their back doors." Those happy days, however, are long since past, and smelt certainly are not so numerous as they were even 50 years ago, around the Massachusetts shoreline of our Gulf, where various streams either have been closed to them, or have been rendered uninhabitable by pollution. But enough still remain to provide sport for thousands of anglers, and we still hear of an occasional catch there of many dozens by some one lucky enough to hit a run of fish at the right time and tide.
In 1938, when a special effort seems to have been made to gather smelt statistics, the reported catch for the inner part of Massachusetts Bay and northward to the New Hampshire line was 25,900 pounds, or some 300,000 fish, if they ran about a dozen to the pound. The yearly catch reported for the coast of Maine, added to that of the Passamaquoddy area (which form one faunal unit so far as the smelt is concerned) averaged about 644,000 pounds during the period 1937 to 1946, or perhaps some 8,000,000 fish; about 61,000 pounds for Digby and Yarmouth Counties, Nova Scotia, combined, which covers most of the catch for the Gulf, north and east of New Hampshire.
The catches of smelt that are made along the coasts of Maine, New Brunswick, and Nova Scotia may seem impressive if taken by themselves. [page 139] But Miramichi Bay, alone, on the southern shore of the Gulf of St. Lawrence yields yearly between three and four times as much smelt as does the entire coastline of the Gulf of Maine.
Catch records do not suggest any striking alteration in the abundance of smelts during the past 10 years or so for Maine or for the Canadian shores of the Gulf. But they seem to have been somewhat more plentiful along the Maine coast previous to the early 1900's, for catches of 1,125,268 to 1,279,550 pounds there in 1887, 1888, and 1902 have not been equaled since then, the nearest approach being 968,300 pounds in 1945.
We are often asked what effect the disappearance of the eel grass (Zostera) from our coasts has had on the abundance of the smelt. Unfortunately, the statistics of the yearly catch do not yield any clear answer. Neither can we offer any convincing explanation for the violent fluctuations that take place from year to year in the abundance (or availability?) of smelts at one point or another. Fishermen report, for example, that they were far less plentiful in Massachusetts Bay and in the Great Bay region, N. H., in 1950 than they were in either of the two previous years.
The smelt also has a great recreational value, smelt fishing being a favorite pastime for home consumption. As many as 2,326 people, for instance, have been counted fishing at one time about Houghs Neck in Boston Harbor, and this same sort of thing is to be seen up and down the Massachusetts coast in harbors and stream mouths in autumn. Many smelt are caught in Great Bay, N. H., in good years, through the ice for the most part. And this applies equally to many localities along the coast of Maine. So plentiful are the fish on occasion and so greedily do they bite, especially on the flood tide, that it is usual to number the catch about Massachusetts Bay by the dozens rather than by the individual fish. Sea worms (Nereis) are generally considered the best bait, especially for the larger smelt, shrimp the second best, small minnows or clams a poor third. Smelt have also been taken with a small red artificial fly in the Gulf of St. Lawrence, and perhaps elsewhere.
 The European smelt (O. eperlanus) smells so strong that it is not held in very high esteem as a food-fish.
 Kendall (Bull. U. S. Bur. Fish., vol. 42, 1927, p. 244) has given a detailed account of the habits, distribution, and catches of the smelt of the New England coast, also of the landlocked populations.
 This method of feeding seems first to have been described in print by "Grif" (Forest and Stream, vol. 54, No. 8, Feb. 24, 1900, p. 151).
 Atkins (Fish. Ind. U. S., sect. 5, vol. 1, 1887, pp. 690-693) gives much information on the smelt in Maine.
 Kendall (Bull. U. S. Bur. Fish., vol. 42, 1927, pp. 231-233) summarizes these observations and gives additional information for streams on the coast of Maine.
 Rept. U. S. Fish Comm., 1897, p. 188.
 The European smelt is landlocked in many lakes in northern Europe.
 Ann. Rept. Biol. Board Canada, (1931) 1932, p. 27.
 Mass. Rept. for 1870, p. 23.
 Kendall (Bull., U. S. Bur. Fish., vol. 42, 1927, pp. 244-249) gives many interesting details as to catches in Massachusetts.
 Smelt fishing has long been restricted to hook and line along this part of the coast.
 Maximum 675,700 pounds in 1945, minimum 316,400 pounds in 1939. No data are available for Maine for the years 1941 or 1942.
 Average reported Catch for Northumberland County 1937-1946, was 2,258,030 pounds.