Plant communities of fragmented agricultural landscapes, are subject to variation in configuration, composition, and distance from one another. These spatial factors affect biological processes. However, connectivity and spatial structuring are rarely considered together in the investigation of biological processes. Spatially optimised predictor variables that are based on informed measures of connectivity, offer a solution to untangling multiple processes that drive biodiversity. A novel spatial optimisation method was used to estimate the scale of effect of biotic and abiotic factors that distinguish plant communities of study sites. We tested whether different hypotheses of connectivity among sites was important to measuring diversity, and whether spatially optimised variables of species abundance and the abiotic environment were consistent with diversity parameters in distinguishing habitat types. The spatially optimised variables indicated significant positive and negative autocorrelation of species abundance at fine and broad spatial scales, which depended on the measure of connectivity. These findings showed that variation in community diversity parameters does not necessarily correspond to underlying spatial structuring of species distributions. The spatially optimised method is extendable to multiple variables of interest. Spatially-optimised variables with appropriate measures of connectivity might be better than species diversity parameters in explaining functional differences among communities.