A thirteen year time-series (1994 to 2006) of gross annual pasture production (GAPP; representing both pasture and crop) was created for the Mediterranean-climate area in the southwest of Western Australia (SWWA) using a light-use efficiency model, incorporating NOAA-AVHRR and NASA- MODIS images in combination with climate data. Trends across the GAPP time-series were quantified by aggregating pixels to spatial regions (called a partition, unit, or spatial support) so that the effects of local spatial noise were minimized. We compared the GAPP analysis using the three spatial partitioning schemes (precipitation zones, Interim Biogeographic Regionalisation for Australia (IBRA) eco-regions, and Statistical Local Areas), and showed that the aggregation unit's size & shape impacted on the analysis. Our results demonstrate trends in GAPP that may be indicative of broader trends in climate change for the SWWA.