Rainbow smelt

Smelt biology and life history , Paul Bentzen

Rainbow smelt (Osmerus mordax) is a small, primarily anadromous fish found along the eastern coast of North American from Labrador to New Jersey. Smelt spend the majority of their time in estuaries or the ocean but during the spring they ascend small freshwater streams to spawn, usually just above the head of tide. The adhesive eggs develop for ~20 days before hatching to produce larvae that immediately drift down to the estuary. Smelt usually mature at an age of 2-3 years.

Smelt also occur in landlocked populations that carry out their entire life cycle in fresh water. These lacustrine (lake) forms have repeatedly evolved into two ecotypes: a �dwarf� and a �large� morph form. Apart from the difference in mature body size, a number of morphological traits tend to distinguish the two ecotypes, all of which are related to differences in their trophic (feeding) ecology. A remarkable feature of the freshwater trophic ecotypes is that both forms co-occur in a few lakes. In these lakes, the two forms show very little evidence of interbreeding; therefore, they appear to be well along in the process of evolving into separate biological species.

The fishery for rainbow smelt has traditionally been recreational with limited, concentrated commercial fisheries (e.g. Miramichi, NB). However, the primary importance of this species lies in its ecological role as an important forage species for a number of larger, commercially important marine and freshwater fishes.

An excellent model species for research

Rainbow smelt has a number of attributes that make it an ideal species in which to study a number of ecological and evolutionary processes. It has numerous populations that are relatively easy to sample during spawning periods. Although smelt are technically anadromous, spawning occurs very close to the ocean (or lake), and stream residence is limited to the period in which eggs are incubating. Therefore, smelt can be regarded as an essentially marine (as opposed to anadromous) fish, but with spawning locations that are much more discrete and easily sampled than those of most marine species.


Smelt are also well suited to the study of local adaption, because of the large number of populations that occur in different environments (wide variety of estuarine and marine habitats, as well as freshwater lakes) as well as the dwarf/large ecotypic variation in fresh water. The dwarf/normal ecotypes that occur together in the same lakes provide an opportunity to study ecological speciation.

Finally, smelt are known to have survived past cycles of glaciation in at least two refugia: an �Atlantic� refugium along the eastern coast of North America near the southern end of the current species range, and an �Acadian� or �Northeast Banks� refugium on the Grand Banks off Newfoundland, parts of which were ice free and above sea level during the Pleistocene Epoch. Contemporary populations of smelt are the result of colonization from these genetically distinct sources; therefore, contemporary genetic structure of smelt populations is the result of both historical and contemporary dispersal.

Current research

Smelt-related research in our laboratory addresses a number of topics: marine connectivity, local adaptation, ecological speciation and disentangling historical and contemporary patterns of gene flow in populations. Former PhD student/current post-doc, Ian Bradbury and current PhD student Mark Coulson are studying population structure and the factors that influence connectivity (gene flow) among anadromous smelt populations on different spatial and temporal scales, different geographic regions, and using a variety of approaches including tagging, otolith microchemistry and molecular genetic data (microsatellites, mitochondrial DNA [mtDNA] and amplified fragment length polymorphism [AFLP]). Work to date suggests very different patterns of connectivity among different regions along the mainland part of the species range, as well as between Newfoundland and the mainland. The difference in the degree of genetic structuring of Newfoundland and mainland smelt populations is comparable to that seen among different species of Pacific salmon, such as chinook and pink salmon, which have very different life histories and fidelity of homing.

An important goal of our research is to understand the factors underlying the differences in connectivity and population structuring that occur among regions. Factors that are likely to be important include marine hydrographic features and the spacing of suitable spawning sites, but also potentially important is the degree (if any) to which smelt home to their natal streams to spawn. In addition it is important to disentangle historical effects from contemporary dispersal. As noted above, rainbow smelt in Atlantic Canada derive from a mixture of two glacial races that evolved in isolation from each other during glacial advances; therefore some aspects of smelt population structure may be due, in part, to historical factors, as well as contemporary processes. We are currently re-examining the phylogeography of smelt with the use of mtDNA sequencing, single nucleotide polymorphism (SNP) and AFLP analyses.

Other projects are examining genetic differentiation among co-occuring ecotypes of freshwater smelt and investigating potential pre- and/or post-zygotic mechanisms preventing the formation of hybrids.

Smelt publications

See Publications page for abstracts.

Bradbury, I.R., M.W. Coulson, S.E. Campana, E. Baggs and P. Bentzen. 2008. Post glacial recolonization and the loss of anadromy in rainbow smelt (Osmerus mordax Mitchill) from coastal Newfoundland. Proceedings of ANACAT II �Challenges for Diadromous Fishes in a Dynamic Global Environment. (in press, Transactions of the American Fisheries Society ).

Bradbury, I.R., S.E. Campana, and P. Bentzen. 2007. Low genetic connectivity in an estuarine fish with pelagic larvae. Canadian Journal of Fisheries and Aquatic Sciences. (in press)

Bradbury,I.R., K.Gardiner, P.V.R. Snelgrove, S.E. Campana, P. Bentzen, L. Guan 2006. Larval transport, vertical migration behaviour, and localized recruitment in anadromous rainbow smelt (Osmerus mordax). Canadian Journal of Fisheries and Aquatic Sciences. 63:2822-2836.

Coulson, M.W., I.R. Bradbury and P. Bentzen. 2006. Temporal differentiation: continuous vs. discontinuous spawning runs in anadromous rainbow smelt (Osmerus mordax). Journal of Fish Biology 69 (supplement C):209-216.

Bradbury, I.R., M.W. Coulson, S.E. Campana and P. Bentzen. 2006. Morphological and genetic differentiation in anadromous smelt (Osmerus mordax Mitchill): disentangling the effects of geography and morphology on gene flow. Journal of Fish Biology 69 (supplement C):95-114.

Coulson, M.W., I.G. Paterson, A. Green, R. Kepkay, and P. Bentzen. 2006. Characterization of di- and tetranucleotide microsatellite markers in rainbow smelt (Osmerus mordax). Molecular Ecology Notes 6:942-944.

Bradbury, I.R., S.E. Campana, P. Bentzen and P.V.R. Snelgrove. 2004. Synchronized hatch and its ecological significance in rainbow smelt Osmerus mordax in St. Mary�s Bay, Newfoundland. Limnol. Oceanogr. 49(5): 2310-2315.

Taylor, E.B. and P. Bentzen. 1993a. Evidence for multiple origins and sympatric divergence of trophic ecotypes of smelt (Osmerus) in northeastern North America. Evolution 47:813-832.

Taylor, E.B. and P. Bentzen. 1993b. Molecular genetic evidence for reproductive isolation between sympatric populations of smelt, Osmerus, in Lake Utopia, southwestern New Brunswick, Canada. Molecular Ecology 2:345-357.

©2017 Paul Bentzen [email protected] Dalhousie University •
Web-design by Peter Feige