Methods to study agricultural systems

Modern agriculture faces complex and ever-evolving challenges. Productive, environmental and social requirements are to be met while fulfilling the needs of numerous stakeholders across a wide array of conditions. To better meet these challenges, researchers study agricultural systems using a myriad of methods, across varied disciplines and contexts. To help connect and orientate these research efforts, an overview is required to assess and categorize the diversity of approaches and methodologies being used to study agricultural systems. Whilst a plethora of specialized studies are available, broad-scope methodological reviews are lacking. Here we review methods used in Australia and New Zealand to study farms, farmers and their broader environment. Both quantitative and qualitative studies were included across a particularly wide range of publications while retaining a high level of methodological detail. An original overarching framework was produced that coherently summarized, described and categorized the diversity of methods encountered. This included defining classification criteria that can be conveniently applied to compare methods, assess their relative use, and identify linkages between approaches.

To achieve this, a new interpretative approach was employed. From a diverse pool of 400 items, 92 publications were reviewed in detail. Three overarching approaches were identified, with publications focusing on “characterizing agricultural situations” (64%), “identifying relationships” (59%), or “retracing processes” (42%). A total of 28 method groups were identified, clustered in 11 broader methodological families that ranged from simple indicators, statistics and narratives to elaborate participatory research, system modelling and mathematical programming. These categories were used to appraise Australian methodological practices. This included highlighting the connectivity between the high number of quantitative methods encountered (used by 72% of publications) and their usually clear, well-established protocols. In contrast, qualitative studies (41%) frequently suffered from inconsistent terminologies and demonstrated a lower diversity of methods in use. The review also noted that details and justifications about descriptive methods were often neglected, in spite of their impact on the entire analysis process. In addition, relatively few studies made use of long-term historical information to generate new knowledge and insights about future directions. Mixed methods were little used (only in 13% of publications), as well as some frameworks including those pertaining to social-ecological systems, agrarian systems or evolutionary economic geography.

Overall, these results show that agricultural research in Australia remains highly quantitatively oriented, favoring analytical details over data origin, and eclipsing a range of qualitative methods and emerging frameworks. Likely explanations include enduring disciplinary and topical divides (notably social/technical), lack of awareness, and lack of expertise. To remedy this, under-used methods would benefit from being pro-actively promoted and taught, and from concerted efforts to build shared, stable nomenclatures about meanings and protocols. Better documentation of data sources is also required.