The influence of environmental drivers and restoration intervention methods on postmine restoration trajectories

Mining activity causes major changes in site topography, soil physical and chemical properties, hydrology, and biological assemblages. Reassembling plant communities in these highly altered environments is dependent on interactions between environmental drivers and management intervention techniques through time, yet the long-term effects of these interactions have rarely been quantified. This study examined a 19-year-old, postmining woodland restoration chronosequence in southwest Australia to understand plant community compositional development in postmine restoration sites. We tested the effects of environmental drivers (rainfall regimes, site aspect, slope) and management intervention techniques (substrate ripping and material properties) in terms of four restoration criteria: species richness, plant density, vegetation cover, and compositional similarity to reference sites. Irrespective of environmental drivers or management intervention techniques, vegetation cover increased through time, while plant density and species richness declined. Bray–Curtis similarity to reference communities remained unchanged. Within these trends, ripping and first-year rainfall significantly affected restoration criteria outcomes; species richness and plant density were greatest when rainfall in the first winter immediately following site restoration was low but followed by a high summer rainfall. The most effective ripping depth was dependent on rainfall, deep-ripped sites performed best when rainfall was high, and nonripped sites performed best under low-mean rainfall conditions. Measured restoration criteria had not reached the levels of the target reference community after 19 years, which may be attributable to the still-developing vegetation structure. This emphasizes the importance of assessing postmining restoration outcomes over longer time frames (>>5 years) with implications for determining appropriate, time-dependent completion criteria.