Biological forecasting system to control black leaf streak

Biological forecasting system to control black leaf streak

The biological forecasting system to control black leaf streak disease (BLSD) monitors the development of the disease in order to control it with a minimum of fungicide applications. The aim of the method is to block the development of the disease towards necroses. Since the efficiency of the treatments relies on a strong curative effect, systemic fungicides are preferred to contact fungicides. Adding mineral oil to the fungicide will also increase its curative effect. Originally developed for Sigatoka leaf spot disease, the method is not suitable for areas that receive rain all year round and works best when the logistics and decision-making is centralized and done by specialists. Besides lowering the cost of controlling leaf spot diseases, it also minimizes the impact on the environment.


The timing of the decision to apply fungicides takes into account the development of the disease and of the plant, both of which are climate driven. The first step consists in setting up a field-based monitoring system. The main parameter that needs to be assessed is the stage of evolution of the disease (SED). The SED is also a good indicator of the efficiency of the fungicide applications. The complete protocol is published in Fruits1 .

Observation plot

In general, 10 banana plants that have between 8 to 10 leaves are used to monitor the progress of the disease. The 10 plants should be randomly selected two months after planting and labeled. The observation of symptoms is preferably done on non-flowering plants once a week. It is recommended to plant 20 new plants every 3 to 4 months to ensure that non-flowering plants are available for observation after the first ones have flowered.

The number of banana plots will depend on the area covered by the warning system and the presence of micro-climates. One plot will cover an area between 20 and 200 ha, depending on its uniformity.

Data collection

Every week, note down the number of banana leaves on each plant and the stage of unfolding of the cigar leaf, using Brun's scale. The stages of development of the cigar leaf are scored as follows: stage A=0, stage B=2, stage C=4, stage D=6 and Stage E=8.

Every week, note down, for each leaf (leaves 2, 3 and 4) of each plant, the most advanced stage of the disease, using Fouré's scale. Also note down the lesion density as "–" if there are less than 50 lesions (irrespective of the development stage of the disease) and by "+" if there are more than 50. Put 'none' if the leaf doesn't show any disease symptom.

Estimation of the SED

Determining the SED requires the calculation of a coefficient, which represents the speed at which the disease develops, and of the foliar emission rate (FER).

Calculation of the coefficient

The stage of the disease, the density of lesions and the leaf number (counting from the youngest leaf whose leaf number is 1) are used to determine the value of the coefficient  (see table below). For a given leaf number, the more advanced the disease stage, the faster the disease develops. For a given disease stage, the younger the leaf, the faster the disease develops.

Stage of the disease

Density of lesions

Leaf number




1 - 60 40 20
+ 100 80 60
2 - 100 80 60
+ 140 120 100
3 - 140 120 100
+ 180 160 140
4 - 180 160 140
+ 220 200 180
5 - 220 200 180
+ 260 240 220
6 - 260 240 220
+ 300 280 260
A corrected coefficient for each stage of the disease is determined for each leaf number. For a given leaf number, the number of sampled plants that have the same coefficient is multiplied with the value of that coefficient. For example, if leaf 3 on 6 of the sampled plants are at stage 1 and have less than 50 lesions (-), the coefficient, 40, is multiplied by the number of plants, 6, for a corrected coefficient of 240 for stage 1 on leaf 3. For a given leaf number, the corrected coefficients for each of the disease stages are added. This total is added to the total for leaves 3 and 4 to obtain the Sumcoef that will be used in the calculation of the SED.

Calculation of the FER

Calculate the FER for each sampled plant as follows:
FER = ((Current number of leaves) + (0.1 x Current cigar stage)) - ((Previous number of leaves) + (0.1 x Previous cigar stage))

Add the FER of each sampled plants to obtain the SumFER.

Calculate the FER for 10 days as follows:
FER10d = (SumFER x 10) / (Number of banana plants x Number of days between observations)

Calculate the FER of the current week as follows:

Calculation of the weekly SED


Timing of fungicide applications

Plot the SED over time. A significant increase in the SED can trigger the decision to apply a fungicide, but should also take in other factors, such as weekly rainfall, and other disease parameters2 . The time between the decision to treat with fungicides and their application should not exceed 2 days and the entire area should be covered on the same day. If aerial spraying is used to apply the fungicides, care must be taken that the climatic conditions are suitable.


Example of the timing of fungicide applications (the different colour of arrows corresponds to different fungicides) based on SED and rainfall data. (Source Ganry et al. 20122 )


A biological forecasting system is best used in regions where there is a dry period (even as short as one month), the disease pressure is low and fungicide resistance has not been observed. The disease will be better controlled if the data collecting, the decision to treat and the logistics of treatment are handled by a well-trained technical unit, instead of being left to the individual producers. Growers, however, can help keep inoculum level low through regular deleafing, especially before a scheduled treatment.

If the SED doesn't decline after an application, it means that the treatment hasn't been effective. It could be due to various factors, such as poor coverage, the choice of fungicide(s), the fungus is losing its susceptibility to the fungicide or the inoculum level is too high.

Using mineral oil as a carrier for the fungicide will improve the efficiency of the treatment. Only fungicide formulations that are compatible with pure oil should be used. The phytotoxic effect of the oil should also be known to determine the concentration to use.  

Depending on the type of fungicide used, and the length of time the fungus is exposed to the fungicide, there is always a risk of fungal populations developing resistance to the fungicide. Monitoring this indicator and taking appropriate action, such as changing fungicides or alterning them, will reduce the selection for resistant fungal strains.


This forecasting system was developed in Guadeloupe and Martinique in the 1970s to control Sigatoka leaf spot disease. It was then adapted to BLSD in the 1980s, first on plantains in Gabon and then on export bananas in Cameroun, Ivory Coast and Ecuador. In places where this forecasting system has been used, BLSD has been controlled with an average of 10 to 15 applications a year 3 . In the 1990s, a study in Colombia showed a 50% reduction in the number of fungicide applications and a 40% decrease in the cost of controlling BLSD4 .


2 Ganry, J., Fouré, E., De Lapeyre De Bellaire, L. and Lescot, T. 2012. An integrated approach to control the Black Leaf Streak Disease (BLSD) of bananas, while reducing fungicide use and environmental impact. p193-226. In: Dhanasekaran, D., Thajuddin, N. and Panneerselvam, A. (eds.). Fungicides for Plant and Animal Diseases. InTech, Rijeka (HRV).
3 Ganry, J. 2001. Maîtrise de la culture du bananier pour une production raisonnée face aux nouveaux défis. Comptes Rendus de l'Académie d'Agriculture de France 87(6):119-127.

Also on this website

Musapedia pages on pesticide-reducing practices:

Further reading

Adaptation of the forecasting system to control black leaf streak disease of banana in the specific conditions of Dominican Republic in Fruits, 2014, volume 69:261-278.

This portal was initiated by a grant from the Ministère français de l'agriculture, de l'agroalimentaire et de la forêt to the World Banana Forum.White