Sampling and monitoring plant diversity in natural landscapes requires careful measurements of species richness, identifying rare species habitats, and detecting exotic species invasions. Since only a small portion of any landscape can be affordably surveyed, mapping and protecting biodiversity requires extrapolating information from the small survey sites to the much larger unsampled area with known precision and accuracy. Land managers collect field data with multi-scale plots (different plot sizes) across various environmental gradients. However, understanding the effects of using sampling designs, plot sizes, and issues of scale and patterns of plant diversity remains a difficult challenge for many researchers in landscape ecology. These issues can be addressed with current advancements in computer technology, fast computer computational capabilities, and the availability of diverse satellite remotely sensed data with high spatial and spectral resolution. We present a simulation of different sampling designs approach to investigate the effects of different plot sizes on the scale and patterns of plant species richness and foliar cover with examples from semi-arid landscape environments. We present diversity indices, spatial autocorrelation, cross correlation, and geospatial statistical analyses on field data collected from multi-scale, nested plots of 1-m², 10-m², 100-m², and 1000-m². Our methods integrate field data, geospatial information (e.g., remotely sensed data, geographic information system, and global positing systems) with spatial statistical modeling and mapping approaches.  Our techniques that use sampling designs, geospatial quantitative modeling, and thematic map products also can be used for ecological forecasting. Combined, our approach can be used by others to investigate patterns of biodiversity at different scales and to predict patterns and spatial relationship of plant species richness and invasive species.