Measuring accuracy of land cover data and content of cover classes: a reply to Brockerhoff et al. (2008)

In a response to our recent paper (Walker et al. 2006), Brockerhoff et al. (2008) reiterate the need to assess ongoing loss of habitat for indigenous species, and for more reliable data to support assessment of status and trends in indigenous biological diversity. However, they question our conclusion, based on comparison of national land cover databases (LCDBs), that ‘plantation forestry remained one of the major causes of indigenous cover loss’ between 1996/07 and 2001/02. In this reply, we address matters of land cover database accuracy and the content of cover classes.