SEARCH

SEARCH BY CITATION

Keywords:

  • cline;
  • colonisation–extinction;
  • equilibrium–non-equilibrium;
  • patch occupancy;
  • Salvelinus malma

Summary

1. We investigated the occurrence (presence or absence) of stream-dwelling Dolly Varden (Salmonidae: Salvelinus malma) in the 78 tributaries of the Sorachi River system, Hokkaido, Japan to examine evidence for metapopulation structure.

2. We hypothesised (i) that the occurrence of Dolly Varden would be less likely in small tributaries because of the higher extinction risk and (ii) that occurrence would be less likely in isolated tributaries because of reduced rates of dispersal and colonisation. In addition to these two general predictions, we hypothesised (iii) that the occurrence would be more likely in spring-fed tributaries, because Dolly Varden prefer cold water as present in these systems.

3. When all the tributaries were included in a multiple logistic regression analysis, habitat size and habitat type (spring-fed versus non-spring-fed tributaries) were selected for the best model, whereas isolation was not important. However, when tributaries in the lower Sorachi basin (an area likely influenced by human activities) were excluded from the analysis, the effect of isolation became significant.

4. Our results indicate there is a potential metapopulation structure of the Dolly Varden in the Sorachi River basin. Furthermore, it is possible that metapopulation dynamics follow a clinal pattern. In the upper and middle reaches of the Sorachi River basin, an equilibrium between extinction and colonisation of Dolly Varden populations is a possibility. In the lower portion of the basin, isolation appeared to be less important, and extinction rate may be driving patterns of occurrence toward systematic extinction of local populations.

5. Our results illustrate the possibility of spatial variation in the processes influencing occurrence of Dolly Varden within a single river basin, and the potential influences of habitat alteration on extinction and colonisation processes.