Complementarity of indigenous flora in shrublands and grasslands in a New Zealand dryland landscape

Succession from fire-induced grassland to secondary shrubland vegetation is occurring in parts of deforested dry eastern South Island, New Zealand, but little is known about how the change from herbaceous to woody vegetation alters the indigenous biota. We predicted that development of secondary shrublands would exclude few grassland-specialist plant species and increase indigenous plant ‘species occupancy’ (the extent to which indigenous species capable of living in dryland landscapes are present) at a landscape scale. We also predicted that taller and denser indigenous plant canopies would facilitate indigenous but not exotic plant species at multiple scales, for example by reducing exotic grass competition or through positive plant–plant interactions. We tested our predictions by sampling vegetation in 100-ha blocks representing a gradient from grassland to mixed ‘intermediate’ grassland and shrubland to shrubland vegetation at three sites in Central Otago. We found indigenous plant species in shrubland were complementary to those in grassland (additional to the existing set) within and across our study sites, which led to higher indigenous species occupancy in this dryland landscape. Grassland vegetation did not support significant numbers of indigenous plant species complementary to those in shrubland, and we found no evidence that secondary succession displaced indigenous plant species. Shrubland had higher diversity of indigenous plant species than grassland at multiple scales (across sites, within 100-ha blocks and within 144-m2 plots), and had higher indigenous dominance of composition across sites, but not at smaller (within-block and within-plot) scales. Indigenous dominance of composition within blocks and plots was related to dominance of the physical structure of plant communities by indigenous woody plants or tussock grasses, rather than to development of secondary shrubland alone. Reduced densities of indigenous plant species were related to higher frequencies of exotic grass only where indigenous structural dominance was low, indicating that competitive exclusion of indigenous plants by exotic grasses may depend on degradation of indigenous structural dominance. We conclude that secondary succession from structurally-depleted grassland to mixed indigenous–exotic shrubland is likely to assist the conservation of indigenous plant species in this dryland landscape.