This paper relates the experience of applying the Global Yield Gap Atlas (GYGA) protocols to assess the yield gap for rainfed wheat in Australia (http://www.yieldgap.org/web/guest/methods-upscaling; http://www.yieldgap.org/web/guest/methods-actual-yield). These recently developed protocols have so far been implemented only by GYGA partners. Given GYGA’s aim to globally map the yield gaps for the world’s 8 most important crops, many more teams will need to participate in the project and implement these protocols. Are the GYGA protocols sufficiently robust and explicit to enable teams with access to the requisite tools and skills, but not involved in specifying the protocols, to implement them?

Twenty two reference weather stations (RWS) with a 100 km radius buffer zone were selected within 6 Australian climate zones (CZs). The areas of up to 3 most common soil types within the RWS buffer was ascertained from soil maps. APSIM was used to simulate Yw from weather, soils, best management practice, and varieties data. Fifteen years of data were used to capture Australia’s variable climate to estimate average Yw values for each RWS buffer zone. Ya was estimated from Australian Bureau of Statistics data at SLA (shire/county) level. For years when SLA data were not collected the gaps were filled by using the coarser ABARES SD (regional) data after establishing robust regression relationships between SDs and their component SLAs. The contribution of each SLA to the Ya of each RWS was weighted by its congruency with the defined buffer zone. An average yield gap was thus estimated for 57% of the Australian wheat area. Ya was 1.73 t/ha compared with Yw of 3.52 t/ha.

This research produced a credible yield gap estimate for Australian rainfed wheat and demonstrated that the GYGA protocols were readily adaptable to a new environment by a team not directly involved in their development.