(M.J. Dadswell in McAllister and Coad, 1974). First a glacial lake (Lake Belleville-Fort 

 Ann) formed in the Ottawa Valley and neighbouring regions. This was followed by the 

 Champlain Sea, a marine invasion caused by sea levels rising over the land still depressed 

 by the weight of the glaciers. This sea then turned estuarine, lacustrine (Lampsilis Lake) 

 and finally fluvial. We suggest that pygmy smelt, unknown to be anadromous, invaded 

 southern Quebec either during the freshwater Lake-Belleville-Fort Ann or the Lampsilis Lake 

 stage. But it cannot be excluded that pygmy smelt retained tolerance to sea water and 

 dispersed along the coast to Maine, New Brunswick, and via the Champlain Sea to the Gatineau 

 Valley. Salinity tolerance tests of pygmy smelt and lacustrine rainbow smelt would be 

 interesting. 



When rainbow smelt invaded the Ottawa Valley region is less clear. Fossils probably 

 referable to Osmerus mordax dating from approximately 10,000 BP were reported from Green 

 Creek near Ottawa, in the Ottawa Valley by McAllister, Cumbaa and Harington (1981). While 

 the size of the two fossils, about 110 mm SL and 117 mm SL suggests the lacustrine form of 

 the rainbow smelt rather than the anadromous form, it is quite possible that these are 

 subadult anadromous smelts. Whatever the form, the fossils do indicate that smelt had 

 invaded the Ottawa Valley within one or two thousand years of deglaciation. 



MANAGEMENT IMPLICATIONS 



That there are two species of smelt, one small, one large in eastern North America has 

 several management implications. 



If both species occur in the same lake, then separate catch statistics should be kept for 

 each species if the populations are to be regulated. Habitat protection must take into 

 account spawning grounds, spawning times and habitat requirements. Beach and stream 

 spawners are more subject to disturbance, pollution and overfishing than sublittoral lake 

 s pawners. 



The larger smelt, Osmerus mordax , is probably more sought after and subject to capture by 

 the ice fishing angler. But both species would be accessible to capture at spawning time by 

 dip net or seine in populations that spawned after break-up in streams or on beaches. 



The interactions of the two smelt species with other species (and between themselves) 

 needs to be considered in managing existing populations or in making introductions. The 

 pygmy smelt is unlikely to feed on the young of other sport fishes or their prey, unlike 

 adult rainbow smelt. Larger pygmy smelt would compete directly with adults of only a few 

 sport and commercial fishes such as ciscos and whitefishes of the genus Coregonus . 

 Occurrence of adult pygmy smelt largely in the hypolimnion would preclude their competition 

 with larvae of those sport fishes that are found in the epilimnion. The small size of the 

 pygmy smelt makes them an ideal forage fish for those fishes which share their habitat. 

 Kendall (1927) stated that most lakes naturally inhabited by landlocked salmon were also 

 inhabited by smelt and that there were no instances of successful stocking of any lake with 

 landlocked salmon when smelts were not also introduced. Lake trout, brook trout and other 

 sport fishes also feed on smelts (Kendall, 1927). 



27 



