Get access

Molecular and morphological analysis of secondary contact zones of Cottus gobio in Fennoscandia: geographical discordance of character transitions


Finnish Environment Institute, Nature Division, PO Box 140, FIN-00251 Helsinki, Finland. E-mail:


Northern Europe was postglacially colonized from different directions by distinct phylogeographical lineages of the bullhead Cottus gobio L. (Pisces: Scorpaeniformes). These lineages have then come into contact in coastal habitats of the currently brackish Baltic Sea and in the freshwaters north of it. We studied the patterns of intergradation in the contact zones in four morphometric and six molecular characters. In the north, intergradation between the western (W) and eastern (E) bullhead lineages is found both among rivers (west-to-east) and along individual rivers (south-to-north). The locations of the transition zones probably relate to the timing of the initial contact, subsequent Baltic shoreline displacement (i.e. emergence of the lower river reaches), and dispersal barriers caused by variations of coastal salinity. The transitions (clines) in different characters are, however, not geographically coincident. Mitochondrial DNA clines are generally found upstream and to the east of the other transitions, and GPI-1 allozyme clines are mostly shifted downstream in the rivers, and west of the other transitions on the broader scale of the Baltic Sea. The location of the mtDNA clines may best reflect the initial contact between lineages, and the displacement of the other clines could result from dispersal being overall asymmetric (predominantly downstream) and sex-biased (stronger in males). Alternatively, the non-coincidence might reflect selection against deleterious cytonuclear character combinations. No clear evidence of reproductive incompatibility between the lineages was seen in local population structures; no remaining genetic correlations were observed locally among traits. In another transition area, a coastal transect in southern Finland, clinal patterns similar to those in the northern contact zone were recorded, but the population compositions could not be explained by simple in situ mixing of any of the putatively pure, invading refugial lineages. Probably, the bullhead stocks that initially came into contact in this southern study area already represented mixtures of the invading lineages. © 2004 The Linnean Society of London, Biological Journal of the Linnean Society, 2004, 81, 535–552.