VALIDATION OF TOMATO POWDERY MILDEW FORECASTING MODEL USING METEOROLOGICAL DATA IN EGYPT

Tomato (Lycopersicon esculentum Mill.) is very important vegetable crop in Egypt for both local consumption and exportation purposes. Powdery mildew, caused by Leveillula taurica (L ́ev.) Arn., is the most important disease for tomato in both greenhouses and open fields. The anamorphic stage of Leveillula taurica found in tomato crop is known as Oidium taurica (L ́ev.) Salmon [1]. Besides tomato, powdery mildew infects other vegetable crops such as pepper, eggplant, and cucumber [2]. The most common symptoms of tomato powdery mildew are yellow and necrotic spots on the upper leaf surface and powdery white spots on the lower side of the leaves [3].

The computerized model (Egy.Tom-PM) simulates the development of powdery mildew on the basis of short-term observation of microclimate factors. The input variables to the model are: temperature, Relative humidity, leaf wetness, wind speed and global radiation for the disease infection potential (DIP). The objective of this research was to validate Egy.Tom-PM, the first Egyptian computerized forecasting model for better disease control with minimum fungicide use. It works by forecasting the most appropriate spray intervals to manage effective control of tomato powdery mildew, compared with routine fungicide application.

Materials and Methods Disease Forecasting Model to Avoid Epidemics of Tomato's Powdery Mildew caused by Leveillula taurica Weather Monitoring
Automated portable and solar powder in-crop weather stations with high technology named (Adcon Telemetry model A733add wave, [ Fig-1] was used to monitor the microclimatological conditions. Every 15 minutes, digital data for temperature, relative humidity, wind speed, leaf wetness and global radiation within the canopy of tomato plants were transmitted via PC using radio waves up to 20 km in distance with no geographic obstacles. The meteorological data are transmitted automatically by radio from the weather station via data collection center (receiver) to receive and save the collected data for up to 50 days and advantage software which allows simple interpretation of the readings using a graph and detailed list of chosen items [Fig-2].

Tomato Powdery Mildew Forecast Model
A computerized simulation software model named by the authors Egy.Tom-PM [ Fig-3 Validation of Tomato Powdery Mildew Forecasting Model using Meteorological Data in Egypt (DDIP) of tomato powdery causal agent Leveillula taurica, to be calculated. Then, the model outputs a daily announcement as a warning message (spray or don't spray) to guide the fungicides application for perfect disease control in the appropriate time. The model evaluation follows the basic rules of system analysis to identify events (A and B). The model starts after the emergence phase is activated, and it tries to detect the event A, which is defined as: Event A is triggered when the model detects at the least X accumulated dynamic summation hours of RH >75 and temperature between Y and Z according to the data tabulated in [Table-1]. Then the model looks for event B which is continues hours of RH > 85 with temperature of the same value between Y and Z. The B event must happen within the 24h of the A event, but not later than 24h after an A event. Moreover, using the model system analysis brings an extension to the model that applies the same rules not only to identify the first critical phase of the season, but also to issue warnings through the whole growing season (DIP).  The following three treatments were tested: (i) a full-schedule fungicide program, in which plants were sprayed every 10 days; (ii) a full-schedule fungicide program, in which plants were sprayed every 7 days; (iii) spraying when nominal use of forecasting system indicated that a spray application was needed and at least 10 days had elapsed since the previous fungicide application. In addition, sulphur 250 gm/100 L water and protectant fungicide (sumi-8) 35 cm3/100 L water were used alternatively with some specific guidance (sprays should cover tomato plants, particularly on the under surface of the foliage and the lower plant canopy and should not be used during very warm, sunny weather to prevent phytotoxicity) were used for disease control as a recommended effective chemicals. Weekly tomato field inspection was initiated to ensure clear detection of the early signs of powdery mildew. Microclimate weather data were automatically collected 24h a day by the advanced telemetry weather station (Adcon A 733), established within the tomato field canopy.

Disease Assessment
The percentage of disease severity was calculated using the equation suggested by Townsend & Heuberger [17] as follows: where : P = percentage of disease severity. a = number of leaves in each category. b = numerical value of each category. N = total number of leaves in sample. K = numerical value of the highest category. A scale of scores was used in accordance with the leaf area affected, proposed by Ullasa, et al [18] of which: 1. Resistant (no symptoms). 2. Moderately resistant (10% of the leaf area affected). 3. Moderately susceptible (11-20% of the leaf area affected). 4. Susceptible (21-50% of the leaf area affected). 5. Highly susceptible (51% or more of the leaf area affected).

Results and Discussion
The results of entire study indicated that Egy.Tom-PM program (Powdery mildew early warning computerized model) for tomato powdery mildew, which designed using the roles of system analysis, was successfully validated in Egypt under tomato field conditions allowing the best timing for fungicide application to reduce the number of fungicide applications for effectively control of powdery mildew compared with full schedule-fungicide program (routine application) with both (7D) and/or (10D) schedule. At the same time, the message of early warning model can forecast the probability of powdery mildew outbreak therefore avoid the exorbitant crop loss in such case with a consideration of powdery mildew epidemics had been recorded in deferent locations around the world. The results are expressed in an early warning message (spray and do not spray) in agreement with weekly field inspection. Data in [  As a matter of fact, in this study the application of tomato powdery mildew disease forecast and early warning model, using a designed computerized model adopted with role of system analysis has been done for the first time not only in Egypt but also all over Arabian and African countries. According to the literature, it is also the first description of powdery mildew all over the Middle East area. Results in this point showed that using advanced last generation of agrometerological remote weather station to collect the daily microclimate elements (the disease forecast and early warning input data) was agreement with Hstgaard [19] who reported that a fully equipped weather station with wireless data transmission, the Hardi Metpol-eReg., and its presentation software developed for use by farmers. The Hardi MetpoleReg, was equipped with sensors for collecting wind velocity, surface wetness, global radiation, air temperature, relative humidity and rainfall data at various heights above the ground. Data were transmitted by radio waves to a receiver at the farm and presented using the computer software. All data are transferable to ASCII-format files for use in computerized monitoring and forecasting systems. Powdery mildew of tomato caused by L. taurica in considered one of the most limiting diseases for tomato production and could be controlled effectively in Egypt for the first time with fewer fungicide applications than a slandered calendar -based schedule with system of disease forecast using computerized model Egy -Tom -PM. This model resulted in a noticeable reduction in fungicide application and consequently reduced the number of sprays. During the 2006 season there were four warning messages in September, October, November and December individually. On the other hand, in 2006 season there were one, two, one and one warning messages in September, October, November and December, respectively. The previous results may be due to the favorable weather conditions, especially under optimum temperature (20)(21)(22)(23)(24)(25) and R.H. ranging between 75-90% in which tomato powdery mildew (L. taurica) needs to development and progress.   [14], Guzman-Plazola [15] and Seem [23] found that effective control of tomato powdery mildew could be achieved in California when protectant fungicides were applied according to a spray forecast model. Moreover, disease symptoms were never observed when the model did not signal a successful infection, even though the starting date for fungicide applications was accurately determined.

Conclusion
The results obtained have emphasised how for tomato protection applications, the use of disease forecasting program for tomato powdery mildew was applied after program validity test during two successive seasons 2006 and 2007 at El-Khawagat village, Yousef El-Seddik County, Fayoum governorate, Egypt. Powdery mildew forecast program reduced the number of application times of fungicides compared with control. Simulation models based on agrometeorological variables and the energy balance could therefore be an important tool for operational applications. Their integration with the development of disease simulation models is of great importance regarding the application time of fungicide sprays and reduce the number of total applications numbers [24] while maintaining a high quality of production with a low impact on the environment. As there was a good agreement between model outputs and actual disease severity, the model can be considered as a satisfactory simulator of the effect of environmental conditions on the progress of powdery mildew epidemics. These approaches have proved useful in describing control effectiveness and, in some cases, optimizing or changing control practices [25,26].