Research on potato late blight forecasting in Cuba was initiated by Mayea et al. (9) when they evaluated the Beaumont (1), Wallin (12) and Hyre (8) models from 1971 to 1974 in the Remedios and Santa Clara municipalities in the province of Villa Clara, situated in the central part of the island. Their results indicated that rain was the most important factor for disease development and that a method using rain and temperature was most convenient for disease forecasting in Cuba.

In the late 1970´s, the short term forecast model of Naumova (13) was implemented in all the Plant Protection Stations (ETPP) located in potato growing areas. This model considers a critical period when the mean and minimum relative humidity are greater or equal to 84 % and 60 %, respectively, the minimum temperature is greater than 11 ˚C and the maximum temperature is lower than 25 ˚C during two consecutive days. The incubation period (the time between infection and symptom appearance) can be evaluated using the Naumova Nomogram, which is calculated taking into account three days after the first critical day occurrence, i.e., a critical period and the following day. The arithmetical averages of maximum, minimum and mean temperatures are then determined. A transparent paper with two perpendicular lines is placed on Naumova Nomogram (Figure 1). The upper part of the vertical line is placed on the mean (average) temperature axis and the right and left sides of the horizontal line are placed on maximum and minimum temperatures axes, respectively: The lower part of the vertical line indicates the duration of the incubation period where it intersects on the scale. Finally, the date of appearance of the first symptoms is determined.

Fig. 1 Naumova Nomogram

This model was later modified because light outbreaks of the late blight could appear without the occurrence of the critical period. The revised model, commonly called “Modified Naumova” (3), included another favorable period known as the alert period to forecast these light outbreaks. The alert period differed from the critical period in that the maximum temperature was set between 25˚C and 28˚C.

In early 1990´s, as a consequence of the meteorological event “El Niño-South Oscillation” (ENOS) in Cuba (10), the relationship between the presence of ENOS, late blight severity and the accumulated rainfall records from January to March was studied for a 60 year period (1929–1990). A clear relationship between ENOS events and the occurrence of severe to moderate potato late blight epidemics was observed. Consequently, the use of predictions of ENOS was recommended for long-term forecasting of late blight epidemics (4).

In 1993–1994, the disease was extremely severe, mainly in the provinces of Havana and Matanzas, where 836.36 ha were lost. In 1994–1995 different steps were taken to avoid the losses produced by the disease and daily warnings were issued to farmers and decision-makers (5). During 1994–1995 the “Rain Threshold” model (11), which had not been used in production areas before, was employed. According to this model, epidemics are predicted after an accumulation of 38 mm of rainfall in a four-week period and weekly average temperatures under 24 ˚C. The prediction is restricted to the initiation of epidemics.

The efficacy of both “Rain Threshold” and “Modified Naumova” models was evaluated in different locations and provinces for 20 years (5, 7) and the percentages of correct predictions reached 80 % and 71 %, respectively. The inefficiency of both methods was also made apparent during this study.

About fifty percent of the cold fronts coming to the island bring about favorable conditions for the development of potato late blight. Most of these favorable condition periods occur in January, and, in decreasing order, December, February and March. It is not surprising then, that the first outbreaks of blight are generally detected in January, when potato plants are 30–50 days old (6). At least two favorable periods must occur before the first symptoms of the disease are visible (5). As the cold fronts move eastwards from Havana, they get weaker and a decrease in favorable periods is observed in the central and eastern regions of Cuba.

The detailed analysis of the effectiveness of the models described above, the causes of their inefficiency, and epidemiological studies led to the design of a new model called “Late Blight Risk Index” (IRTT), which applies the best from both the “Modified Naumova” and “Rain Threshold” models (5).

The IRTT is expressed in equation 1:

IRTT = K i - 1 + [m (∑ αi pi ) + 1] [n β + 1]     1)

K i = [m (∑ αi pi ) + 1] [n β + 1]     2)

Where:

K i = Accumulative weekly constant calculated as in equation 2

K o = 0 (first week, December 1–7)

i = Number of the week in the year

α = Number of favorable periods in that week

p = Type of favorable periods in that week

(p = 3: critical; p = 2 : alert; p = 1: critical-alert, where one day is critical and the other alert)

β = 0, when the accumulated rainfall in the week assessed is less than the threshold value

β = 1, when the accumulated rainfall in the week assessed is equal to or higher than the threshold value

m and n are monthly constants:

in December___________ m = 1; n = 3

in January_____________ m = 3; n = 3

in February____________ m = 2;n = 2

in March______________ m = 1; n = 1

IRTT can be calculated daily or weekly from December 1st, taking into account the threshold values of daily and weekly-accumulated rainfall.

This model allows the emission of warnings or disease alerts of likely disease appearance and an estimate of the disease risk so that proper actions can be taken to protect the fields. The risk can be obtained comparing the weekly values of the analyzed period when potato is grown with tables of IRTT mean values when epidemics are light, moderate and severe.

The IRTT was also used for making long- term predictions. A forecast about the impact of global climatic changes was made using both the scenario IS92a* and the HADCM2 Global Climatic Model** with mean sensitivity for the years 2010, 2030, 2050 and 2100 (2). For this comparison a baseline was developed based on IRTT for December to February for the years 1978–1994. For 2010 the situation was similar to that obtained in the baseline, i.e. the appearance of the disease every year and a 50 % frequency of severe epidemics, mainly in the western part of Cuba. A decrease in disease severity might be expected for 2030, 2050 and 2100 because the IRTT values are below those found in the baseline.

The occurrence of both more frequent and more severe epidemics that have caused serious damage to the potato crop in the western region of country since 1993–1994 and the presence of different, unknown inoculum sources raised great interest and led to a project aimed at improving late blight integrated management. This project, supported by Cuban Agricultural Ministry, started in 2001 and is expected to finish in 2005. The population of Phytophthora infestans has been characterized phenotypically and genotypically. Presently, efforts are underway to correlate the results of these population studies with evaluations of new fungicides, the epidemiology of the disease in different years, current control strategies and the forecasting system.

Guadalupe Gómez and Keren Hernández

Plant Health Research Institute. Calle 110 # 514 Esq. 5ta B,

Playa, Ciudad de La Habana, Cuba.

Tel: 537 202 6788, Fax: 537 24 0535

Email: ggomez(at)inisav.cu

Literature Cited:

1.  Beaumont, A. 1994. The dependence on weather of dates of outbreaks of potato late blight epidemics. Transactions of the British Mycology Society 31:45–53.

2.  Centella, A. et al. 1998. Climatic change scenarios for impact assessment in Cuba. Segundo Taller sobre “Evaluación rápida del impacto del cambio climático en Cuba”. INSMET, La Habana, 55 pp.

3.  Gómez, Guadalupe et al. 1999. Naumova modificado: ajuste de un método de pronóstico para el tizón tardío de la papa y el tomate en Cuba. Fitosanidad 3:95–100.

4.  Gómez, Guadalupe et al. 1999. Influencia del Evento Meteorológico “El Niño-Oscilación Sur” sobre epifitótias de tizón tardío de la papa en Cuba. Fitosaniddad 3:21–26.

5.  Gómez, Guadalupe et al. 1999. Sistema de pronóstico para el tizón tardío de la papa causado por Phytophthora infestans (Mont) de Bary en Cuba. Tesis presentada en opción al Grado Científico de Doctor en Ciencias Agrícolas. INISAV, MINAG, Cuba.

6.  Gómez, Guadalupe et al. 2001. Pronóstico del tizón tardío (Phytophthora infestans (Mont) de Bary) de la papa en Cuba. I. Análisis de la edad del cultivo e intervalo de tiempo óptimo para el surgimiento de los primeros brotes de la enfermedad. Fitosanidad 5:23–28.

7.   Gómez, Guadalupe et al. 2002. Pronóstico del tizón tardío (Phytophthora infestans (Mont) de Bary) de la papa en Cuba. II. Evaluación de la efectividad del modelo Naumova Modificado. Fitosanidad 6:35–39.

8.  Hyre, R. A. 1959. Progress in forecasting late blight of potato and tomato. Plant Disease 43:245–253.

9.  Mayea, S. S. et al. 1975. Algunos métodos para el pronóstico del tizón tardío de la papa y las posibilidades de su utilización en Cuba. Centro Agrícola 2:77–85.

10.  Meulenert, P. A. and García, V. G. 1992. Efecto del Evento Meteorológico “El Niño-Oscilación Sur” sobre el estado del tiempo en Cuba y su influencia sobre la agricultura. La Habana: ACC. INSMET, 30 pp.

11.  Padrón, S. J. 1982. Umbrales de lluvia para el pronóstico del tizón tardío en Cuba. Ciencia y Técnica en la Agricultura. Serie de Protección de Plantas 5:77–85.

12.  Wallin, J. R. 1951. Forecasting Tomato and Potato Late Blight in North-Central Region (Abs.). Phytopathology 47:37–38.

13.  Zchumakov, A. 1970. Metodología para descubrir, diagnosticar y señalizar las enfermedades de la papa. Metodichieskiie Ukasaniia.

* A simple climatic model of greenhouse effect of released gases that estimates the mean range of future emissions.

 ** A general circulation model taking into account the greatest warming and the driest conditions on Earth .