Examining the spatial influences of natural capital in the Australian agricultural landscape

Abstract

Overall, this thesis seeks to explore – using three case studies - the environmental and economic influences and outcomes of on-farm natural capital in the Australian agricultural landscape over space and time. In particular, it explored: 1) the spatial influences on the adoption of certified organic farming (which is used as a proxy indicator of natural capital conservation technologies) at a regional level in Australia using agricultural census data from 2010/11 and 2015/16; 2) the association between the presence of certified organic farming and regional biodiversity at the postcode level over sixteen years in South Australia; and 3) the association between farm land value and natural capital in the forms of native woody vegetation coverage and climate in South Australia over sixteen years. The first case study focused on Australia as a whole and modelled farmers’ adoption behaviour of certified organic farming (using it as a proxy for sustainable agriculture technologies to conserve on-farm natural capital such as soil, water, and biodiversity). Spatial diffusion of organic farming represents an interesting case study, given the large amount of skills and knowledge regarding management of natural resources that organic farmers need to apply/learn for their farms’ viability. Although farmers’ adoption and diffusion behaviour is well studied in the literature, modelling of the role of spatial spill-over effects on diffusion intensity, especially in regards to the adoption of organic farming, is not well known. This thesis uses national Australian agricultural census data from 2010/11 and 2015/16 and a SLX Tobit model (N=2,134) to model the influences on the intensity of the diffusion of organic farming (namely percentage of organic land holding) in regional areas, and found statistically significant local spatial spill-over effects from neighbouring regions’ characteristics. In addition, a higher share of organic farmland in regions is associated with regional characteristics such as: larger irrigated farms; lower stocking rates; increased proportion of grazing and horticultural land; increased labour supply; increased green vegetation; rural areas with low human population density; and higher community income. The second study explored the associations between farmers’ land use behaviour (i.e. the extent of certified organic farming in a region) and regional biodiversity outcomes (vascular plant and bird species richness) at the postcode level. This study put together a new dataset on certified organic farming presence and locations in South Australia, using databases from organic certifiers. The spatial association between biodiversity indicators and organic farming was analysed using a spatial Durbin error model, while controlling for the effects of landscape attributes, human population footprint, climate and productivity from 2001 to 2016 (N=5,456) in South Australia. The results found that increased organic farming presence in postcode areas had a statistically significant positive association with vascular plant species richness, but little to no statistically significant association was found for bird species richness. Environmental heterogeneity in terms of land cover diversity, elevation range, and plant productivity seems to be the other prime determinants of plant and bird species richness. The third study focused on the association between native woody vegetation on agricultural properties and their economic values in South Australia, using both sales and valuation prices of agricultural properties from 1998 to 2013 (N=10,513). Findings from the spatio-temporal Durbin model revealed that the presence of native woody vegetation on agricultural properties significantly increased the per hectare market price (i.e. price sold in the market), but at a decreasing rate as the proportion of vegetation increased. The marginal return of vegetation was highest for small size properties and lowest for larger properties. In addition, the direct effects of increased annual rainfall, increased soil natural productivity, increased market accessibility, proximity to locational amenities, smaller size properties, availability of irrigation, and higher commodity price were also positively capitalised into sales prices. On the other hand, increased drought and high soil erodibility significantly reduced per hectare sales prices. Comparing valuation price models with sales price models, it was found that the valuation prices seem to undervalue the presence of native vegetation on agricultural properties and hence provide weaker evidence of the value of on-farm natural capital in the South Australian context.