Effects of secondary shrublands on bird, lizard and invertebrate faunas in a dryland landscape

Succession from anthropogenic grassland to secondary woody plant communities in New Zealand’s eastern South Island dryland zone has potential to alter animal communities. We compared indigenous and exotic birds, terrestrial invertebrates and ground-dwelling lizards in 100-ha blocks representing vegetation at three woodiness levels (grassland, mixed grassland–shrubland, and shrubland) at three sites in Central Otago. We predicted that shrublands would support invertebrate taxa and indigenous bird species complementary (additional) to those in grasslands, thus increasing the diversity of taxa present in the landscape, and that shrublands would have higher indigenous dominance of bird composition than grassland. We also expected that vegetation woodiness would affect the taxonomic composition of assemblages of indigenous and exotic birds and of ground-dwelling invertebrates. We predicted shrubland would have little net effect on indigenous dryland ground-dwelling lizards, which we expected to be regulated by other microhabitat characteristics. To test our predictions we used rarefaction-based analyses of faunal complementarity, and canonical correspondence analysis (CCA) and generalised mixed models of animal response variables in relation to vegetation and environmental predictors. Our results showed that shrubland blocks supported ground-dwelling invertebrate orders and indigenous forest bird species complementary to those present in grasslands. Block-level woodiness was a primary predictor of indigenous, but not exotic, bird assemblages. At smaller scales of transects within blocks, the indigenous dominance of bird species assemblages in shrubland blocks was not significantly higher than elsewhere, but was related to plant-community-specific variations in more numerous and diverse exotic birds. Terrestrial beetle-family assemblages within plots varied significantly with local woodiness, but invertebrate-order assemblages did not. Shrubland development had little effect on common ground-dwelling lizards, which showed strong microhabitat preferences and favoured habitats dominated by indigenous plant species. We conclude that sizeable (i.e. >1 km2) areas of shrubland vegetation enhanced the diversity of indigenous birds and of ground invertebrate orders present at a landscape scale. Both grassland and shrubland habitat patches in drylands supported indigenous ground-dwelling lizards, but lizards may be vulnerable to habitat modifications that reduce indigenous plant dominance.