A natural capital approach to agroforestry decision-making

Agroforestry systems are well known to enhance provision of multiple ecosystem services (e.g. wind speed reduction, habitat provision, carbon sequestration). These ecosystem services deliver both public and private benefits at a range of scales. For example, well-designed agroforestry systems enhance livestock productivity at the paddock scale by providing shelter from extreme temperatures, while also contributing to biodiversity conservation and climate change mitigation at broader scales. Despite growing global awareness of these benefits, adoption, particularly in some temperate industrialised agricultural landscapes including those in Australia, remains constrained. To encourage adoption of agroforestry and inform optimal design of agroforestry systems, new approaches are required to measure and communicate the broad range of benefits provided by agroforestry. Communication of farm-scale benefits is particularly important, as decisions relating to adoption and design of agroforestry in these landscapes generally fall to individual landowners. Natural capital accounting (NCA), a process of accounting for natural resources underpinned by ecosystem service valuation, could provide the framework for such an approach. While NCA is currently effectively applied at regional and national scales, this thesis examined whether NCA concepts can be usefully and practically applied at finer scales, such as individual farms or paddocks, to improve farm-scale decision-making and encourage adoption of agroforestry. The first of three research questions asked: can measurement and valuation of multiple ecosystem services, within the NCA framework, be usefully and practically applied for the purpose of encouraging adoption of agroforestry? To answer this question, existing concepts and methods for ecosystem service measurement and valuation were reviewed to assess their suitability for application in this context (Chapter 2). The review found that while some existing concepts (e.g. ecosystem service classification systems) are readily transferrable to agroforestry, established methods for ecosystem service measurement and valuation require adaptation (namely in how they deal with scale and choice of beneficiary) in order to be usefully applied to encourage adoption of agroforestry and inform decision-making. A discrete choice experiment (DCE) survey was also conducted with farmers from areas of Australia that hold potential for expansion of temperate agroforestry (Chapter 6). Results of this survey highlighted the importance of demonstrating 'value for money', or a return on investment in agroforestry, and revealed that farmers value a wide range of ecosystem services in agroforestry systems. These findings confirm the need for agroforestry valuation methods that incorporate multiple ecosystem services and acknowledge different forms of 'value'. Ecosystem service valuation within the NCA framework provides this flexibility and is therefore useful in an agroforestry context. The second research question explored the issue of scale: which scale is most appropriate when applying NCA concepts for the purpose of encouraging agroforestry adoption and informing decision-making, and what implications does this choice of scale have for ecosystem service measurement and valuation? A review of the literature (Chapter 2) showed that because farmers are the key decision makers, NCA concepts should be applied at the farm or paddock scale and focus on farmers (or their investors) as the key beneficiaries. Chapter 2 also identified that there are significant gaps in the evidence base for measurement of ecosystem services by agroforestry assets at fine scales, particularly in the case of biodiversity-related services. Field experiments conducted in shelterbelt systems in the Midlands region of Tasmania, Australia (Chapters 3, 4, and 5), addressed several of these gaps. Chapter 3 showed that paddock-scale provision of a range of ecosystem services (wind speed reduction, wood production, carbon sequestration, and habitat provision) can be predicted based on measurement of vegetation structural characteristics. Chapters 4 and 5 found that shelterbelts support invertebrate communities that differ significantly from those in open pasture, and that they improve potential for pollination in adjacent paddocks. These findings contribute to the expanding evidence base for ecosystem service provision in agroforestry, at scales that are useful for informing decision-making by farmers. The third research question focused on the critical NCA concept of 'condition', asking: is vegetation structure a suitable and practical metric for assessing the condition of agroforestry assets? Results from the DCE (Chapter 6) showed that farmers had strong preferences for particular agroforestry design attributes (e.g. tree species selection) and further, that farmers are motivated to make design choices that will maximise benefits in alignment with their objectives. Many of the ecosystem services identified as being important to farmers (e.g. wind speed reduction) are influenced by vegetation structure (e.g. shelterbelt porosity). Chapter 3 examined whether vegetation structure offers a practical solution to the issue of condition assessment, by quantifying the impact of shelterbelt tree species selection (Pinus radiata, Eucalyptus nitens, and mixed native species) on specific structural attributes (e.g. height, porosity, and floral diversity) that determine fine-scale provision of key ecosystem services. Species selection, in addition to shelterbelt age, was found to significantly affect structure, and therefore delivery of services and benefits at fine scales. However, Chapters 4 and 5 found that shelterbelt tree species selection, and therefore structure, had relatively low influence on composition of invertebrate communities within and adjacent to shelterbelts. Findings from Chapters 3-5 show that while structure serves as a practical condition metric that is useful for informing agroforestry design, tracking flows of some services (e.g. biodiversity) may require consideration of broader landscape-scale condition metrics such as connectivity and complexity. Overall, this thesis demonstrates that with some adaptation, NCA concepts (e.g. condition assessment, ecosystem service measurement, and valuation) can be practically and usefully applied to agroforestry systems at the farm scale. The thesis shows that NCA concepts can be used to highlight the wide range of values provided in agroforestry systems, thereby building a more holistic business case for agroforestry that will appeal to a wider range of farmers and investors. Ultimately, this will assist in increasing adoption of agroforestry by farmers. The thesis also shows that application of NCA concepts to agroforestry will assist in informing design decisions (e.g. tree species selection), enabling farmers to design systems which maximise benefits that are important to them. Expanding this approach to consider both private and public values could also be useful in guiding development of policies aimed at enhancing broader benefits of agroforestry, such as biodiversity conservation and climate change mitigation.