Silvicultural impacts in jarrah forest of Western Australia: synthesis, evaluation, and policy implications of the FORESTCHECK monitoring project of 2001-2006

This paper, the final in a series of ten papers that report the impact of silvicultural treatments (harvesting and associated burning) in jarrah (Eucalyptus marginata) forest, reviews these papers and explores similarities and disparities. More than 2500 species were processed, dominated by macro-invertebrates, vascular flora and macrofungi. Few significant impacts were evident, and most species groups were resilient to the disturbances imposed. Regeneration stocking did not meet specified standards on two gap release and seven shelterwood grids subjected to silvicultural treatment in the period 1988-2002. Six treated grids had a retained basal area of more than 18 m(2) ha(-1), which obviated the need for further regeneration. More than 50 y may be needed for biological processes to reverse the increase in bulk density of soil caused during harvesting. Cryptogams (especially lichens) were the species group most sensitive to disturbance, although recovery of species richness was nearly complete 10 y after disturbance. For cryptogams and vascular flora, species recorded in only one grid (singletons) were more likely to occur on reference grids than on silviculturally treated grids. For all species groups studied, the imprint of harvesting 40 or more years earlier on species composition had become indistinguishable from that on grids never harvested. Soil nutrient status correlated with species rich-ness for fungi on wood (negatively), light-trapped inverte-brates (positively), birds (positively) and terrestrial vertebrates (frogs, reptiles and mammals, negatively). Silvicultural disturbance (timber harvesting and associated burning) correlated with species richness for fungi on wood (positively), terrestrial vertebrates (positively) and cryptogams (negatively). Time since the last (prescribed) fire did not correlate with any species group. Plant disease decreased species richness of light-trapped invertebrates by about 35%. Very few taxa were sufficiently widespread or sufficiently responsive to silvicultural disturbance to be of value as bio-indicators, demonstrating the superiority of biodiversity monitoring over bio-indicator monitoring. It is recommended that FORESTCHECK be expanded into a biological survey of the lower south-west of Western Australia.