The majority of fungi, aerial nematodes and bacteria that cause plant diseases require liquid “free” water on the plant surfaces before they can infect the plant. Free water, sufficient for disease infection,  commonly occurs  in the form of rain, fog, and dew.  Or it can even occur after a sprinkler irrigation , syringing , or pesticide spray. Moreover, many fungi need high relative humidity to produce spores, and abundant free water can contribute to high relative humidity. Dew formation is triggered when the surface temperature of the plant surface drops below the dew point temperature of the surrounding air (See previous blog post).  This typically occurs at night in greenhouses that are not ventilated and heated properly, or outside on calm clear nights. Plant pathogens have different leaf wetness requirements for infection to occur (as seen below).   

Many diseases will get worse with longer leaf wetness periods. In the table below you can see that the severity of gray mold (Botrytis cinerea) on rose increases with longer wet periods. 

Commercial disease prediction models exist for apple scab, cedar apple rust, potato late blight, tomato early blight, strawberry anthracnose, botrytis fruit rot, citrus brown spot, lettuce downy mildew, grape powdery mildew, among others.  Leaf wetness sensors help quantify leaf wetness periods, and models predict disease risk  These systems can reduce the number of sprays that are needed for disease control.  It has been recently suggested that disease models instead use leaf wetness data based on a simple empirical model using relative humidity. Relative humidity sensors can be standardized and calibrated more easily than leaf wetness sensors.

Many current greenhouse control systems can help collect and organize data from leaf wetness, relative humidity and temperature sensors. Alternatively, a simple environmental monitoring system can be pieced together for an outdoor nursery or greenhouse using readily available sensors and data loggers from various companies (e.g. , Campbell Scientific Inc (Logan, UT), Onset (Bourne, MA),  Spectrum Technologies Inc. (Aurora, IL).   Most disease risk models have not been tested in ornamental crops but there is no reason why they cannot be wholly or partly used for disease risk monitoring in ornamental crops. Botrytis models have been intensively studied in other crops and should be one of the first to try in ornamental crops.  Empirical evaluation of these models in the field is a first step to confirm their usefulness.   Models that predict high disease risk could improve scouting efficiency by targeting more intensive scouting during these periods, help reduce fungicide applications by predicting optimal timing of fungicides before infection occurs, and target periods when de-humidification cycles are needed in greenhouses.