Did historical tree removal promote woody plant encroachment in Australian woodlands?

Question Woody plants have increased in density in many ecosystems, but the factors promoting encroachment are often debated. Since European colonization, C allitris glaucophylla has recruited abundantly in many E ucalyptus- C allitris woodlands in eastern Australia following changes to disturbance regimes analogous to changes in many other ecosystems globally. We used a dynamic stand model to disentangle effects of disturbances on Callitris encroachment and asked, to what extent was C allitris encroachment enhanced by historical removal of pre-settlement trees and subsequent thinning of regrowth? Location Woodlands dominated by E ucalyptus sp. and C . glaucophylla, which originally occupied ca. 100 000 km2 in eastern Australia. Methods We used a process-driven stand simulation model to simulate tree demography, with growth, survival and recruitment mediated by annual rainfall, competition and disturbance. Following parameter calibration and model testing, we orthogonally manipulated historical removal of pre-settlement trees and thinning of regrowth to identify how both processes may have influenced structural changes over 120 yr of European settlement. Results Removal of pre-settlement trees had little effect on modelled encroachment, as trees increased to comparable densities whether pre-settlement trees were retained or removed. In unthinned scenarios, C allitris regeneration formed 'locked' stands of high density but low basal area. Thinning promoted growth of retained C allitris, increased total stand basal area, and together with direct removal of large pre-settlement E ucalyptus and thinning of E ucalyptus regrowth, transformed stands from E ucalyptus to C allitris dominance. Conclusions Removal of pre-settlement trees does not appear to have been a necessary precursor to modelled encroachment in E ucalyptus- C allitris woodlands in eastern Australia, perhaps because initial tree cover was low and C allitris can regenerate beneath isolated trees. Manual thinning was required to effect major structural change because C allitris self-thins extremely slowly; thinning will be less important in other ecosystems that self-thin more rapidly. The impact of historical tree removal on encroachment is likely to vary according to initial tree cover, with greatest impacts in dense ecosystems with high tree cover. These results highlight the value of simulation models for disentangling the effects of multiple disturbances on tree encroachment and other ecosystem dynamics. [ABSTRACT FROM AUTHOR]