Alternative stable states and the role of fire-vegetation-soil feedbacks in the temperate wilderness of southwest Tasmania

Two ecological models have been put forward to explain the dynamics of fire-promoting and fire-sensitive vegetation in southwest Tasmania: the alternative stable states model of Jackson (in Proc Ecol Soc Aust 3:9-16, 1968) and the sharpening switch model of Mount (in Search 10:180-186, 1979). Assessing the efficacy of these models requires high resolution spatio-temporal data on whether vegetation patterns are stable or dynamic across landscapes. We analysed ortho-rectified sequences of aerial photography and satellite imagery from 1948, 1988 and 2010 to detect decadal scale changes in forest and non-forest vegetation cover in southwest Tasmania. There was negligible change from forest to non-forest (< 0.05%) and only a modest change from non-forest to forest over the study period. Forest cover increased by 4.1% between 1948 and 1988, apparently due to the recovery of forest vegetation following stand-replacing fire prior to 1948. Forest cover increased by 0.8% between 1988 and 2010, reflecting the limited ability of forest to invade treeless areas. The two models include interactions between vegetation, fire and soil, which we investigated by analysing the chemical (phosphorus, nitrogen) and physical properties (clay, silt) of 128 soil samples collected across 34 forest-non-forest boundaries. Phosphorus in the upper horizon was typically lower in non-forest vegetation compared to forest vegetation, which is consistent with proposed fire-vegetation-soil feedbacks. Mineral horizons were dominated by sand, with low levels of clay under all vegetation types. Available field evidence lends support to the Jackson (1968) alternative stable states model as the most suitable model of vegetation dynamics on nutrient poor substrates in southwest Tasmania although modifications of the timeframes for transitions toward rainforest are required.