Temporal changes in vertebrates during landscape transformation: A large-scale "Natural Experiment"

Abstract

Plantation development is a significant form of landscape change worldwide. We report findings from a large-scale longitudinal natural experiment that quantified changes in Australian vertebrates as a former grazing landscape was transformed to one dominated by a radiata pine (Pinus radiata) plantation. The study included four main "treatments": woodland remnants surrounded by emerging radiata pine (52 sites, termed "woodland treatments"), stands of radiata pine (10 sites, "pine controls"), woodland remnants where the surrounding landscape remained unchanged (56 sites, "woodland controls"), and paddocks with scattered woodland trees that surrounded the 56 woodland remnants (10 sites, "paddock controls"). In our study region, woodland is distinguished from forest by differences in tree height, tree spacing, bole length, and canopy development. Between 1998 and 2006, occupancy rates of "woodland treatments" by most mammals and reptiles increased linearly. Similar trends occurred in the "woodland controls," suggesting that species had increased landscape-wide, rather than displaying year × treatment interaction effects. We cross-classified birds according to the statistical significance and nature of time trajectories. Groups included those that: (1) declined in woodland treatments in comparison with woodland controls, (2) decreased within woodland treatments but increased in woodland controls, (3) declined across the entire study area, (4) increased within woodland treatments in comparison with woodland controls, (5) increased within woodland treatments but declined in woodland controls, and (6) increased across the entire study area. Attributes of woodland treatments significantly associated with temporal changes in bird occupancy included: (1) age of surrounding pine stands; (2) number of boundaries with surrounding pines; (3) size of the woodland patches; (4) dominant vegetation type of woodland patches; and (5) temporal changes in vegetation structure in the woodland treatments. Bird species associated with open country and woodland environments were disadvantaged by landscape transformation, whereas those that benefited were forest taxa and/or habitat generalists capable of inhabiting pine stands and adjacent woodland patches. Beyond this generalization, an unanticipated finding was a lack of association between life history attributes and landscape transformation. We suggest that several key processes are likely drivers of change at multiple spatial scales. Recognition of such processes is important for conservation in landscapes transformed by plantation expansion.