Analysis of Spatial Patterns of Virus-Diseased Tobacco Plants

Publisher Website
Google Scholar
Abstract
Epidemics caused by tobacco etch virus (TEV) and tobacco vein mottling virus (TVMV) were monitored in six experimental fields of tobacco [Nicotiana tabacum] in Kentucky [USA] from 1983 to 1985. Aggregation of virus-diseased plants was determined by dividing fields into contiguous quadrats and using point pattern (e.g., variance-to-mean ratio and Lloyd''s patchiness [m*/m]) and spatial autocorrelation analyses. Spatial distribution of diseased plants was neithr solely clustered nor random, but changed with time during the epidemics. All indices of aggregation indicated a random pattern at the beginning of the epidemics if the first disease assessment was early enough. Patchiness increased to a maximum (m*/m > 2) and then declined throughout the remainder of the epidermics. In many fields, patchiness also indicated randomness (.apprx. 1) by the last assessment time. First-order autocorrelations (rw, ra) increased throughout most epidemics, eventually indicating clustering ( > 0.23). Autocorrelations often exceeded 0.5 at their maxima. When mean disease density approached 100% incidence, autocorrelations declined at the end of the epidemics. Spatial correlograms suggested a first-order autoregressive process. Iwao''s regression o mean crowding (sensu Lloyd) on mean virus disease density indicate a random pattern of clusters (slope.simeq.1). The sequence of diseased plants per row also was found to change with time based on separate ordinary runs analyses. Percentage of tobacco rows with a clustered pattern increased during most of the epidemics from a low initial level ( < 10%); the percentage declined as disease incidence approached its maximum. In agreement with spatial autocorrelation analysis, maximum percentage of rows with a clustered pattern was reached before the last assessment date if disease density was high.