With increasing political pressure to produce more food whilst being environmentally and socially considerate, alongside the need to cope with climatic extremes and financial instability, farming needs to become more sustainable. To monitor and improve understanding of sustainable agriculture, farmers will need additional tools to illustrate the impacts of their business decisions. However, current tools to monitor the sustainability of agriculture require measurement of variables that are rarely readily available. Moreover these tools exclude farmers in their development and interpretation. This paper suggests a pragmatic approach to creating a farm-based monitoring tool. We propose that farm-level indices of sustainability are initially based only on data that is readily available. Whilst this would increase its appeal to farmers and therefore participation rates, it may initially have little immediate value as a measure of sustainability. Therefore a 'design-action-design' cycle-the basis of adaptive co-management-must be employed to allow the tool to evolve. Starting from this pragmatic, bottom-up perspective, as data collection systems improve, more theoretically driven (i.e. top-down) site-specific variables of sustainability can be included to provide a more comprehensive tool. This paper illustrates the principles involved by (i) calculating a farm-specific composite sustainability index (CSI) for a commercial farm based on readily available data and (ii) emphasising the need to establish better data collection systems.

Lorne Peak Station, near Garston in Northern Southland is a 5650 ha station, which until recently, has been run as an extensive store sheep and beef property. Major constraints for this hill country property, like many others in the region, are long winters with occasional snowfall, low pasture growth rates, dry summers, unpredictable autumns and limited cash flow to develop and improve the land. The intensification of New Zealand's hill country farms is a critical aspect in ensuring financial viability. Since 2009, Lorne Peak has undergone a rigorous intensification program, initially aided by income produced from wintering dairy cows on the property. With careful selection of crops and pastures, farm profitability has increased, through selection of different stock breeds/classes and high utilisation of feed supply over the year. FARMAX® modelling software has been used retrospectively to analyse these changes, and compare it with the current scenario. The analysis clearly shows that the use of lucerne for grazing, fodder beet for beef production, subdivision and increasing soil fertility have been real game changers for Lorne Peak. A 292% increase in farm profit before tax is testament to this achievement. However, an increase of 26% for total farm working expenses, highlights the need for investment to be planned to build resilience into the farming operation. Keywords: hill country, intensification, dryland, lucerne, fodder beet, sheep, beef, FARMAX®, profitability

Decision Support Tools (DSTs) and models are becoming increasingly important to make informed decisions. The aim of this review was to understand how the scientific assumptions behind related pasture-based DSTs used in New Zealand interrelate. Using modelled feed intake of grazing sheep as a case study, we reviewed the metabolisable energy (ME) assumptions of relevant New Zealand models. Although the DSTs were largely dependent on a common model, there were some discrepancies between them. We raise two concerns: 1. In New Zealand, pasture-based livestock models to predict ME intake are highly dependent on equations developed in Australia, not validated in New Zealand. 2. The minor discrepancies between models make collaboration difficult - this is particularly pertinent to farmer-targeted DSTs. Closer collaboration between DST companies and research institutions is required to agree on consistent, robust equations aligned with the latest scientific knowledge.