Quantification of environmental-economic trade-offs in nutrient management policies

Nitrogen losses from agricultural are a key source of human impacts on the environment, and many countries have adopted policies to reduce nitrogen losses. Policy in New Zealand is being developed at the national and regional levels to address nitrogen losses and water quality. Several policy options were explored using a multi-agent simulation model of the Southland region of New Zealand in order to quantify the trade-off between the economic value of agricultural production and nitrogen losses from farming. It estimated the relative effectiveness and efficiency of alternative nitrogen mitigation policies while taking into account the heterogeneity of soil vulnerability to nitrogen leaching, land management options, and farmer behaviour. It used a hybrid modelling technique, assembling a multi-disciplinary model from outputs of other, specialised models, and using an agent-based approach to model land-use change. The policy options included uniform limits on nitrogen losses that applied across all farms, as well as differentiated policies that took into account either the propensity of a farm to leach nitrogen, past dairy conversion, or the type of land use. After 25 years, the impacts on dairy land area, nitrogen losses, agricultural production, and farm gross margin were compared with a baseline of no policy. The results suggested that policies worked better when they took account of the heterogeneity of agriculture practices and the environment. Those policies could be more effective at reducing nitrogen losses from farms, in term of the total mitigation in the region. They were also more efficient across the policies modelled: per kilogram of nitrogen mitigated, they produced the lowest economic costs. Choosing the right policy approach would be some combination of the absolute level of mitigation required, the historical patterns of land use, the variability of the absorptive capacity of the environment, the ability to spread the economic or environmental impacts across many farms and people, and the ability to specify required input or outputs. Most importantly, hybrid multi-agent simulation modelling provided a tool for examining the potential impacts of policies before they are implemented.