• BACI experiment;
  • biodiversity;
  • disturbance;
  • ecologically sustainable forest management;
  • Eucalyptus diversicolor;
  • karri;
  • logging


  • 1
    Increasing concern over the sustainable management of forested landscapes and the extent of forest clearance world-wide has led to a growing interest in the impacts of logging and associated habitat disturbance on biodiversity.
  • 2
    We conducted an experimental study of the impact of clearfelling on birds of the karri Eucalyptus diversicolor forest in south-west Western Australia over a 17-year period, and a retrospective study of both clearfelled and naturally regenerated karri stands aged from 0 to 146 years.
  • 3
    One-third of species still had significantly reduced abundance 14 years after disturbance ceased, although all affected species made limited use of regenerating forest.
  • 4
    Multivariate analysis of changes in bird community structure showed that the effects of disturbance were still evident 14 years after clearfelling. Clearfelling may also have produced some temporary changes in community structure in adjacent unlogged forest.
  • 5
    Species richness and total abundance of birds declined by 58% and 96%, respectively, in the first year after clearfelling, and 14 years after logging were still 17% and 55% below levels in adjacent undisturbed forest. During this early successional phase both measures increased as a simple function of stand age. Species richness of regrowth reached that of old-growth at 30–50 years and total abundance of all bird species in regrowth was similar to that of old-growth after approximately 70 years.
  • 6
    Several bird species offer potential as indicators of the ecological sustainability of karri forest management. These species nest in large hollows in standing live trees (two cockatoo species) or have been slow to recolonize immature regrowth karri forests (six species).
  • 7
    Post-hoc power analysis showed that even the long-term and intensive sampling employed in this study failed to detect declines in abundance of less than 80–90% for most bird species. For many uncommon species, trying to estimate changes in abundance is problematic and likely to require replication which is difficult to achieve in field situations where logged and unlogged forests are compared. Despite this, the present study identified some key impacts of forest clearfelling on bird communities, with implications both for the consequences of clearfelling of forests and the criteria for sustainable forest management.