Spatial Analysis Based on Variance of Moving Window Averages |
| |
| Authors: | B M Wu K V Subbarao F J Ferrandino and J J Hao |
| |
| Institution: | Author's addressess: Department of Plant Pathology, University of California, Davis, c/o US Agricultural Research Station, 1636 East Alisal Street, Salinas 93905, CA;;Plant Pathology and Ecology Department, The Connecticut Agricultural Experiment Station, New Haven, CT 06504;;Department of Plant Pathology, University of California, Davis, CA 95616, USA (correspondence to K. V. Subbarao. E-mail: ) |
| |
| Abstract: | A new method for analysing spatial patterns was designed based on the variance of moving window averages (VMWA), which can be directly calculated in geographical information systems or a spreadsheet program (e.g. MS Excel). Different types of artificial data were generated to test the method. Regardless of data types, the VMWA method correctly determined the mean cluster sizes. This method was also employed to assess spatial patterns in historical plant disease survey data encompassing both airborne and soilborne diseases. The results obtained using the VMWA method were generally different from those obtained with Lloyd's index of patchiness and beta‐binomial distribution methods, were in partial agreement with the results from spatial analysis by distance indices, and were highly consistent with the results from semivariogram and spatial autocorrelation analysis methods. Results demonstrated that the VMWA method can be applied to many types of data, including binomial diseased or healthy plant counts, incidence, severity, and number of diseased plants or pathogen propagules although directional and edge effects may limit its application. |
| |
| Keywords: | aggregation index spatial dependency |
|
|
