SPATIAL PATTERNS OF GRAPEVINES WITH EUTYPA DIEBACK IN VINEYARDS WITH OR WITHOUT PERITHECIA

Eight vineyards in northern and central California were surveyed durin g three consecutive years, 1989-1991, and the presence or absence of E utypa dieback symptoms was recorded for each vine in contiguous blocks of 1,250-3,150 vines. The vineyards were located in areas with differ ent levels of mean annual rainfall; some vineyards contained inoculum sources (perithecia) of Eutypa lata; others did not. The spatial patte rns of infected vines were examined by ordinary runs, two-dimensional distance class, spatial autocorrelation, and geostatistical analyses. Disease incidence ranged from 3.4% in 1989 to 81.5% in 1991. During th e study, disease incidence more than doubled in five of the vineyards. Vineyards with perithecia had higher disease incidence. A disease gra dient or edge effect was detected in two vineyards that did not contai n inoculum sources; one of these was found to be adjacent to a vineyar d with E. lata perithecia. The different analyses consistently describ ed the relative randomness of the patterns of diseased vines among the vineyards. Those vineyards that contained perithecia had a higher pro portion of vineyard rows with nonrandom disease patterns according to runs analysis. Two-dimensional distance class analysis showed that vin eyards with perithecia contained clusters of diseased vines or other n onrandom patterns. Vineyards with perithecia also consistently had mor e significant spatial autocorrelation coefficients and semivariograms that indicated spatial dependence at distances up to 25 m. A nonrandom pattern was consistently found in one vineyard that was not near any known inoculum source. Three other vineyards with no known inoculum so urces nearby were consistently considered to have random patterns, acc ording to the spatial pattern analyses, in these vineyards, there was no evidence that would indicate disease spread by means other than air borne ascospores from distant sources.