ISHS Acta Horticulturae 489: VIII International Workshop on Fire Blight
FIFTEEN YEAR SUMMARY OF THE EFFICACY OF THE MARYBLYT PREDICTION SYSTEM ON APPLE IN WEST VIRGINIA (1984 – 1998) |
| Authors: | G.W. Lightner, T. van der Zwet, P.W. Steiner |
| Abstract:
MARYBLYT, a computer model to predict and manage fire blight, was developed initially in 1987 (Lightner and Steiner, 1990; Steiner, 1990). This model identifies the conditions that are conducive to the development of four separate types of fire blight symptoms: blossom, canker, shoot, and trauma blights. During the blossom period, the model identifies dates with risk of infection, provides a qualitative assessment of risk (low, moderate, high, and infection), and predicts the date of the first visible symptom. The earliest symptom consists of small white to amber ooze droplets, visible on the young ovary or on the blossom stem (Figure 1). The MARYBLYT prediction system version 4.3 was evaluated for the past three years (1996–98) on apple at the Appalachian Fruit Research Station. In all three years, we recorded one infection period each, all resulting in negligible blossom blight seasons. During the last 15 years, fire blight blossom infections were observed eleven times. The MARYBLYT system accurately predicted the occurrence of blossom blight symptoms (BBS) in ten years to within 24 hours of actual symptoms observed in the orchard and within 48 hours in the other year (1989). In five of these ten years, BBS dates matched observations exactly, i.e. blossom blight symptoms were observed precisely on the predicted dates. (Table 1). These observations included those published in 1994 (van der Zwet, et al., 1994). Attempts to quantify blossom infection severity using the calculated epiphytic inoculum potential (EIP) alone are invalid. The MARYBLYT system uses four parameters for determining infection: open blossoms, cumulative degree hours >18.3C, wetness >=0.25mm, and average daily temperature >=15.6C. These are minimum thresholds that indicate only that some blossom infections are likely to occur. The degree to which any one or more of these thresholds is exceeded can be used as a subjective indication of likely severity, but severity indices have not been quantified. It appears that moderate or severe fire blight seasons result from the occurrence of multiple infection periods. Although light fire blight seasons may have more than one infection period, the moderate blight season (1990) contained two infection periods and severe blossom blight seasons were observed only in years with three or more infection periods (Table 2). | |
Download Adobe Acrobat Reader (free software to read PDF files) | |
