Responses of flora and fauna in wet eucalypt production forest to the intensity of disturbance in the surrounding landscape

Abstract

Abstract In disturbance-dependent forest ecosystems, species associated with mature forest may maintain their ranges both by persisting in patches of mature forest and by recolonising patches regenerating after fire, harvesting or other disturbances. The null hypotheses are that these processes are both independent of the intensity of disturbance in the landscape surrounding those patches. These were tested in a 1120 km2 region of tall, wet eucalypt forest in southern Tasmania, which provides a landscape-scale gradient in disturbance intensity shaped by past wildfires and post-European land-use. For all 50 × 50 m cells in the region, a Landscape Disturbance Intensity (LDI) index was calculated based on the vegetation in the surrounding 0.5, 1.0 and 2.0 km-radius landscapes. The occurrence of species in three taxonomic groups – birds, vascular plants and flighted beetles – was then sampled in 28 plots (50 × 50 m) across the range of LDI in each of two forest age-classes – mature eucalypt forest (MAT) and 25–50 years-old silvicultural regeneration arising from clearfell harvesting (SILV). In both MAT and SILV plots, the occurrences of many common species showed significant relationships with LDI. Most (>80%) of those relationships reflected species that declined in occurrence with increasing LDI, i.e. were Landscape Disturbance-Sensitive (DSL). However, there were contrasting patterns between the forest age-classes as well as among the three taxonomic groups in the proportion of common species that were DSL. More plant and bird species were DSL in SILV plots (15/49 and 9/27, respectively) than in MAT plots (4/49 and 1/27, respectively). The dominant fractions of the DSL plant and bird species in SILV plots were classed as rainforest trees or shrubs in previous studies and birds that were associated with MAT plots rather than SILV plots in this study, i.e. species that were also sensitive to disturbance at the patch-scale. A similar number of the common beetle species were DSL in MAT plots (29/234) as in SILV plots (35/234). Beetle species that were Patch Disturbance-Sensitive (DSP) in this study were over-represented in the subset of DSL species in MAT plots (9/29), but not in SILV plots (5/35). We conclude that many DSP species inhabiting mature forest stands are resilient to disturbance in the surrounding landscape, but their ability to recolonise regenerating forest stands can be diminished by disturbance at the landscape-scale. This highlights the role of vegetation structure at both patch and landscape-scales in the long-term viability of mature-forest-dependent species in production-forest landscapes.