Understanding the matrix: Impacts of modification in indigenous grassland landscapes

Conversion of native vegetation to agricultural land is known to have a multitude of effects on ecosystems, including biodiversity loss, habitat fragmentation and alteration of resource availability. The concept of modification due to agriculture is complex and requires quantification in order to successfully understand its impacts. This thesis describes a robust method of quantifying gradients of modification in New Zealand modified tall tussock (Chionochloa spp.) grasslands, which was used to characterise the impact of modification on a focal species, namely the endangered grand skink (Oligosoma grande). A number of other properties were also examined along the gradient of modification, including vegetation attributes, landscape connectivity and resource availability, in order to examine the ways in which increasing agricultural modification may impact these properties and how this might affect the focal species, the grand skink. The gradient of modification included six study sites, ranging from tall tussock reserve-land to simplified pasture. The level of modification within each site was quantified with an index developed using four metrics: vegetation structural diversity, tall tussock biomass, tall tussock fragment length and bryophyte cover. Grand skink abundance and occupancy were determined along the quantified gradient of modification using a photo-resight methodology. Abundance exhibited a negative correlation with the degree of modification, whilst occupancy probability estimates exhibited a high degree of fluctuation along the gradient of modification. As expected, the structure and composition of the vegetation matrix changed markedly along the gradient of modification. Tall tussock cover decreased and fragmentation (measured as number of tall tussock patches) increased with modification. Native plant diversity was found to be relatively consistent as modification increased until a critical threshold was reached, beyond which diversity rapidly declined. Plant community composition changed noticeably as modification increased. Evidence suggested a relationship between simplification of the vegetation matrix and reduced grand skink abundance, potentially due to altered dispersal probability and resource availability. Connectivity for the grand skink was assessed using a network analysis approach. Sites showed differences in the spatial configuration of rock tors (primary grand skink habitat).Results suggested a relationship between the spatial configuration of tors and grand skink occupancy and that the configuration of tors was more important in determining occupancy in the more modified sites, than in the least modified site. This was perhaps a consequence of the matrix in the least modified site being more structurally diverse, allowing for less discriminate dispersal. Essential habitat for maintaining connectivity within each site was highlighted. Resource availability for the grand skink was investigated seasonally along the gradient of modification by surveying aerial invertebrates. Invertebrate biomass and abundance generally increased along the gradient of modification and a significant difference between the least and most modified site was found in most seasons (establishing a link to literature reporting increased grand skink body condition within a pasture matrix). Invertebrate species diversity increased along the gradient of modification in spring and summer, but fluctuated in both winter and autumn. In general, results suggested species responses to modified landscapes are complex and that the vegetation matrix is an important factor that must not be overlooked in conservation. In addition, results suggested that significant biodiversity is retained within intermediately modified tall tussock grasslands. The links between the vegetation matrix, connectivity and resource availability were highlighted as important considerations for conservation management