Contrasting dynamics in two subpopulations of a leech metapopulation over 25 year-classes in a small stream


  • J. M. Elliott

    Corresponding author
    1. Freshwater Biological Association, The Ferry House, Ambleside, Cumbria LA 22 0LP, UK
      J. M. Elliott, Freshwater Biological Association, The Ferry House, Far Sawrey, Ambleside, Cumbria LA22 0LP, U.K. E-mail:
    Search for more papers by this author

J. M. Elliott, Freshwater Biological Association, The Ferry House, Far Sawrey, Ambleside, Cumbria LA22 0LP, U.K. E-mail:


  • 1The leech, Erpobdella octoculata (L.), occurred as two local populations at sites A (deep section with macrophytes) and B (shallow stony section) in a small stream. A June census estimated densities of young leeches (R1 m−2) and mature leeches (R2 m−2) towards the end of the first and second years of the life cycle, respectively. An August census estimated the density of cocoons laid by the mature leeches before they died in autumn. Cocoon density multiplied by the mean number of young per cocoon estimated the density of young produced at the end of each year-class (R3 m−2) and also the density at the start of the next year-class (S m−2). Information was obtained for 25 year-classes (1966–90).
  • 2The spatial distribution of the leeches and their cocoons was aggregated and similar at both sites in spite of the marked differences in habitat. Survivor density at three life stages (R1, R2, R3) was related to the density of young at the start of each year-class (S), the density-dependent relationship being described by a dome-shaped Ricker curve that differed markedly between sites. A few year-classes were excluded from the analyses because their density was reduced by a severe spate and three summer droughts. A stability analysis showed that the subpopulation at site A was essentially stable around the equilibrium density, whereas that at site B was strongly cyclic and was approaching chaotic behaviour.
  • 3Loss-rates were fairly constant in the second year of the life cycle, but those in the first year were very variable and related positively to the density of mature leeches. The latter ate newly hatched leeches, even those still in cocoons. Cannibalism appeared to be the most obvious mechanism responsible for the density-dependent relationships, its impact being greatest at site B. Loss-rates were not dependent on the density of the young leeches themselves (i.e. direct density dependence) but on the density of mature leeches that ate them (i.e. delayed density dependence). This case study illustrates how subpopulations in a metapopulation can have both similar (spatial distribution) and contrasting (density fluctuations) dynamics.