**2.2.3 Component 3: The control of the inoculum pressure in the field**

It is achieved through prophylactic leaf removal coupled with relevant crop management, such as plant density and water and fertilizer management.

The experience of BLSD control in Gabon (see §2.2) has clearly shown that when plantain fields were isolated into a forest environment and far from infected plots, preventing them from external contamination, the disease control was strongly facilitated.

Thus, keeping the sources of inoculum at a very low level is very important to ensure the success of the chemical control strategy. Where extensive spotting is present, new infections will develop quickly because chemical sprays do not remove the disease from spotted leaves and the only solution is to remove them mechanically from the banana tree. Therefore, such necrotic leaf removal must be a regular practice in banana growing countries where BLSD is present. It should be based on a risk management approach, taking into account that it could affect the production beyond certain level.

Experimental leaf removal has shown that bunch weight, the main yield component, is poorly affected as long as 5 to 7 leaves remain on the plant from flowering to harvest (Ramsey et al., 1990; Daniells et al., 1994; Vargas et al., 2009). A stronger leaf removal could affect the final yield. In addition a lack of resources during this phase, due to severe spotting could compromise the yield of the next ratoon (Eckstein et al., 1995; Dens et al., 2008).

Leaf removal is particularly needed in critical areas where regular sprays are not well performed due to natural obstacles or to regulatory measures such as in buffer areas where aerial spraying is prohibited (e.g. 50 m from houses, gardens, rivers and roads in the FWI). A strong communication effort should be done towards the farm managers in order to improve the situation, because the chemical control strategy will not be successful if the leaf removal is not done properly. Such leaf removal might be mandatory on BLSD susceptible banana plants that cannot be part of the spraying agenda: in home gardens, near houses, along roads or rivers. In such situations the best solution would be to replace the susceptible bananas by resistant varieties. It requires to take into account social and political components, such as the acceptability of these varieties by the consumers in order to encourage voluntary behavior. In some cases like in Australia, more coercive measures were used with an objective of complete eradication which was successfully achieved (Henderson et al., 2006).

In addition it is essential to destroy systematically all abandoned plantations and isolated roadside banana plants that are potential reservoir of disease inoculums.

In areas where fungicide sprays are done, it is essential to realize necrotic leaf removal before any fungicide spray in order to lower the inoculum pressure and thus enhance the efficiency of the fungicide and to minimize the exposure of the pathogen to the fungicide and thus the risk of appearance of resistant strains to the fungicide.

An Integrated Approach to Control the Black Leaf Streak

the Leaf Spot Diseases control

(part 2)

Disease (BLSD) of Bananas, while Reducing Fungicide Use and Environmental Impact 209

**Box 5b -** Technical Sheet used in the FWI for the mechanical sanitation of banana for

**Box 5a -** Technical Sheet used in the FWI for the mechanical sanitation of banana

for the Leaf Spot Diseases control

(part 1)

An Integrated Approach to Control the Black Leaf Streak

Bellaire et al., 2010b).

fungicide.

**2.2.5 Component 5: Organization of the control** 

grouped in an association that would perform the control strategy.

mode of organization.

fungicide amended media.

The success story of SLSD control in Guadeloupe and Martinique is particularly due to the

Then new methods are presently developed from conidia to overcome these drawbacks.

Centralization of decisions and operations is essential and the banana growers should be

Disease (BLSD) of Bananas, while Reducing Fungicide Use and Environmental Impact 211

**Box 6. Resistance monitoring:** The basic methodology relies on the comparison of the sensitivity to the different fungicides in fungal populations (50-100 spores) sampled in commercial farms (treated with the fungicides) and fungal populations sampled in untreated locations. The monitoring of sensitivity is based on germination tests: the germination of spores grown on agar media added with different concentrations of fungicides is compared with the germination of spores grown on agar (de Lapeyre de




In the methodology approved by the FRAC (see FRAC website, monitoring methods (MYCOFI)), fungal populations consist in ascospores obtained from necrotic leaf samples (20-25 plants are sampled in each location). At laboratory, these necrotic leaf samples are bulked and incubated in a moist chamber for 48h and then leaf pieces are used for ascospore discharge on petri dishes enriched with the targeted fungicide.

This method has several drawbacks : (i) ascospore production is very fluctuable and the population analyzed in the Petri dishes might be very different from the population initially sampled in the field; (ii) in certain cases sporulation does not occur and the test cannot be carried, especially in the dry season; (iii) this method does not allow to use a predetermined sampling design; (iv) the populations analysed on each fungicide concentration and the control are always different; (v) other ascospores belonging to the genus *Mycosphaerella* might be confused with the ascospores of *M. fijiensis*, especially on

germtube as compared with the control on ager medium.

resistant if germ tube length or GI is over a threshold value.

#### **2.2.4 Component 4: The fungicide sequence management**

The forecasting strategies enable BLSD control with a limited use of fungicides. In such strategies, the cost of control is lower but, above all, the environmental impact is limited.

Such strategies should be preferred to systematic strategies that are conducive to significant environmental impact and they should be combined with other measures such as sanitation (leaf removal), drainage, under-cover irrigation and good fertilization practices. Forecasting strategies have faced some drawbacks in some countries, especially in Cameroon, because of the emergence of fungicide resistance to systemic fungicides. It is the reason why such strategies should be implemented under specific conditions in the related banana industry: (i) banana areas with no or few fungicide resistance to available fungicide families and particularly those where adverse climatic conditions exist during a part of the year (dry season). This is still the case in some parts of Ivory Coast, Ghana and the Caribbean; (ii) availability of curative fungicides with different mode of actions that can be used in alternation and offering alternatives in case of fungicide resistance is detected in one family; (iii) logistic capacity to operate when necessary without any delay and in the best conditions (window); (iv) strong involvement of the plantation managers and cooperation of the citizens and administration to contribute to the sanitation of banana plants (see §1.4).

When new plantations are established in new areas of a country where fungicide resistance is present, special attention should be paid to germplasm movement in order to prevent the introduction of resistant strains through planting material. Nurseries for planting material must be established in areas where the presence of resistant strains has not been demonstrated. This is, for instance, the case in Cameroon and Ivory Coast. In such situations, the sustainability of the forecasting strategy would be a significant challenge.

It is clear that a key component of a good fungicide sequence management relies on a regular monitoring of fungal populations (see box 6).

A special attention should be paid to the management of fungicide resistance that might develop following the repetitive use of curative fungicides that have a monosite action. Alternation between different groups of fungicides, or mixtures with contact fungicides, is essential to reduce selection pressure and delay the emergence of such resistance.

In several countries where such attention was not so high, the regular and intensive use of systemic fungicides has resulted in an important development of resistant strains in *Mycosphaerella fijiensis* populations. These data are well documented inside the banana working group of FRAC (Fungicide Resistance Action Committee). The resistance risk is considered as severe for strobilurins and antimitotics, and important for IBS of group 1.

The current status of fungicide resistance to these chemicals is a key information to monitor control strategies for BLSD based on a forecasting strategy because these strategies rely on the use of fungicides having a strong curative effect, i.e. systemic fungicides. It is also important to note that the cost of systemic fungicides is higher than the protectants, and that their use is then justified if they really enable a good control of BLSD with a lower number of applications.

It is the reason why a regular resistance monitoring is of prime importance in an integrated disease management system for BLSD (box 6)

The forecasting strategies enable BLSD control with a limited use of fungicides. In such strategies, the cost of control is lower but, above all, the environmental impact is limited.

Such strategies should be preferred to systematic strategies that are conducive to significant environmental impact and they should be combined with other measures such as sanitation (leaf removal), drainage, under-cover irrigation and good fertilization practices. Forecasting strategies have faced some drawbacks in some countries, especially in Cameroon, because of the emergence of fungicide resistance to systemic fungicides. It is the reason why such strategies should be implemented under specific conditions in the related banana industry: (i) banana areas with no or few fungicide resistance to available fungicide families and particularly those where adverse climatic conditions exist during a part of the year (dry season). This is still the case in some parts of Ivory Coast, Ghana and the Caribbean; (ii) availability of curative fungicides with different mode of actions that can be used in alternation and offering alternatives in case of fungicide resistance is detected in one family; (iii) logistic capacity to operate when necessary without any delay and in the best conditions (window); (iv) strong involvement of the plantation managers and cooperation of the

citizens and administration to contribute to the sanitation of banana plants (see §1.4).

When new plantations are established in new areas of a country where fungicide resistance is present, special attention should be paid to germplasm movement in order to prevent the introduction of resistant strains through planting material. Nurseries for planting material must be established in areas where the presence of resistant strains has not been demonstrated. This is, for instance, the case in Cameroon and Ivory Coast. In such situations, the sustainability of the forecasting strategy would be a significant challenge.

It is clear that a key component of a good fungicide sequence management relies on a

A special attention should be paid to the management of fungicide resistance that might develop following the repetitive use of curative fungicides that have a monosite action. Alternation between different groups of fungicides, or mixtures with contact fungicides, is

In several countries where such attention was not so high, the regular and intensive use of systemic fungicides has resulted in an important development of resistant strains in *Mycosphaerella fijiensis* populations. These data are well documented inside the banana working group of FRAC (Fungicide Resistance Action Committee). The resistance risk is considered as severe for strobilurins and antimitotics, and important for IBS of group 1.

The current status of fungicide resistance to these chemicals is a key information to monitor control strategies for BLSD based on a forecasting strategy because these strategies rely on the use of fungicides having a strong curative effect, i.e. systemic fungicides. It is also important to note that the cost of systemic fungicides is higher than the protectants, and that their use is then justified if they really enable a good control of BLSD with a lower number

It is the reason why a regular resistance monitoring is of prime importance in an integrated

essential to reduce selection pressure and delay the emergence of such resistance.

**2.2.4 Component 4: The fungicide sequence management** 

regular monitoring of fungal populations (see box 6).

disease management system for BLSD (box 6)

of applications.

**Box 6. Resistance monitoring:** The basic methodology relies on the comparison of the sensitivity to the different fungicides in fungal populations (50-100 spores) sampled in commercial farms (treated with the fungicides) and fungal populations sampled in untreated locations. The monitoring of sensitivity is based on germination tests: the germination of spores grown on agar media added with different concentrations of fungicides is compared with the germination of spores grown on agar (de Lapeyre de Bellaire et al., 2010b).


In the methodology approved by the FRAC (see FRAC website, monitoring methods (MYCOFI)), fungal populations consist in ascospores obtained from necrotic leaf samples (20-25 plants are sampled in each location). At laboratory, these necrotic leaf samples are bulked and incubated in a moist chamber for 48h and then leaf pieces are used for ascospore discharge on petri dishes enriched with the targeted fungicide.

This method has several drawbacks : (i) ascospore production is very fluctuable and the population analyzed in the Petri dishes might be very different from the population initially sampled in the field; (ii) in certain cases sporulation does not occur and the test cannot be carried, especially in the dry season; (iii) this method does not allow to use a predetermined sampling design; (iv) the populations analysed on each fungicide concentration and the control are always different; (v) other ascospores belonging to the genus *Mycosphaerella* might be confused with the ascospores of *M. fijiensis*, especially on fungicide amended media.

Then new methods are presently developed from conidia to overcome these drawbacks.

#### **2.2.5 Component 5: Organization of the control**

The success story of SLSD control in Guadeloupe and Martinique is particularly due to the mode of organization.

Centralization of decisions and operations is essential and the banana growers should be grouped in an association that would perform the control strategy.

An Integrated Approach to Control the Black Leaf Streak

2-4 kg/ha/year in the former forecasting system)

rarely used (de Lapeyre de Bellaire et al., 2010b)

both economic and environmental – way.

**Box 7.** 

**Box 8.** 

Cameroon:

supply the urban markets of Libreville

"dilution" of fungicide resistance

Disease (BLSD) of Bananas, while Reducing Fungicide Use and Environmental Impact 213

Thus, in 2006, despite a continuing effort to drive the chemical control from the observation of biological descriptors, about 40 treatments were performed on most plantations. This increase in the number of treatments resulted in an increase in the cost of the control, but also in environmental risks, as contact fungicides are applied at higher rates than systemic fungicides (de Lapeyre de Bellaire et al.*,* 2009). So, this evolution has led to an important increase of negative environmental effects since 30-40 kg/a.i/ha/year are now applied (vs.

Recent observations of the latest monitoring show there is a decrease in resistance levels in some commercial plantations, especially since the systemic fungicides are no longer or

This trend suggests that the phenomena of resistance to fungicides may be reversible (see Box 7 and 8) and thus that new treatment strategies can be redefined in a more sustainable -

In the horizon of 2 or 3 years, it seems possible to recover one or two fungicides and reintegrate them with newly available chemical family as part of an integrated strategy that would reduce

Resistance to benzimidazoles and strobilurins has actually dropped since they are no longer used. But we still need a background of resistance that can even advise their use. However there is every reason to be optimistic about the possibility of reuse, at the horizon of a few years in some areas under warning strategies that would also benefit the registration of systemic fungicides with new mode of action, like succinate deshydrogenase inhibitors (SDHI) that will be soon released. With regard to the triazoles, the sensitivity deteriorates steadily mainly because they still are used in most

Several mechanisms may be behind this recent drop of resistance levels observed in

neighboring plantains), to commercial plantations (because of fungicide applications populations have lower size in commercial treated farms) that could cause a progressive


fungicides, fungicide resistance has a fitness cost (Karaoglanidis *et al*, 2001)


**2.4 BLSD control in Gabon in small scale plantain production for domestic markets**  Here are briefly presented the activities conducted on the agro-industrial plantation of plantains at N'toum (100 ha, 60 km from Libreville, Gabon) created in the late 1980s to

the number of fungicide applications and the amount of active ingredient spread.

plantations. Their use should be postponed in order to recover sensitivity.

Since ascospores are transported by wind over long distances, the control strategy should be the same in all banana plantations to prevent any disruption. The organization of the treatments is more efficient when a centralization of the decision is performed by a unique technical service operating according to rational guidelines rather than if each grower implements his own strategy.

#### **2.3 BLSD control in Cameroon in the banana industry for export**

In Cameroon, *M. fijiensis* was first reported in 1981. In the late 80s, a warning method using biological descriptors has been developed based on the strong experience gained with the control of SLSD in the FWI (Ganry & Laville, 1983; Bureau & Ganry, 1987; Bureau et al., 1982) and on results obtained in Gabon with BLSD on plantain (Fouré 1982a, 1982b, 1983, 1984, 1985; Fouré & Grisoni, 1984, Fouré et al., 1984). It was successfully applied, thus limiting the number of applications to 12-14 per year. This rational control by warning relied heavily on the use of systemic fungicides with a high curative effect (Fouré, 1988a, 1988b, 1988c; Fouré & Mouliom Pefoura, 1988; Fouré & Moreau, 1992).

This situation was sustainable for 10 years, but disease control then became unsuccessful due to logistic failures (shortage of airplanes), which led to the more intensive use of systemic fungicides.

As a consequence, since 1996 the emergence of strains resistant to systemic fungicides resulted in the progressive replacement of this rational control strategy by more frequent applications of systemic fungicides. As a result, resistance to site-specific strobilurins has developed particularly swiftly. Systemic fungicides have been progressively abandoned and replaced by contact fungicides, with chlorothalonil the most frequently used (Fig. 4). The contact fungicides do not cause the emergence of resistant strains, but have not the curative effect required for a prolonged action, as in the case of systemic's.

Fig. 5. History of fungicide use for BLSD control in a representative commercial banana farm in Cameroon from 1985 to 2006 (de Lapeyre de Bellaire et al., 2009).

Thus, in 2006, despite a continuing effort to drive the chemical control from the observation of biological descriptors, about 40 treatments were performed on most plantations. This increase in the number of treatments resulted in an increase in the cost of the control, but also in environmental risks, as contact fungicides are applied at higher rates than systemic fungicides (de Lapeyre de Bellaire et al.*,* 2009). So, this evolution has led to an important increase of negative environmental effects since 30-40 kg/a.i/ha/year are now applied (vs. 2-4 kg/ha/year in the former forecasting system)

Recent observations of the latest monitoring show there is a decrease in resistance levels in some commercial plantations, especially since the systemic fungicides are no longer or rarely used (de Lapeyre de Bellaire et al., 2010b)

This trend suggests that the phenomena of resistance to fungicides may be reversible (see Box 7 and 8) and thus that new treatment strategies can be redefined in a more sustainable both economic and environmental – way.

In the horizon of 2 or 3 years, it seems possible to recover one or two fungicides and reintegrate them with newly available chemical family as part of an integrated strategy that would reduce the number of fungicide applications and the amount of active ingredient spread.

#### **Box 7.**

212 Fungicides for Plant and Animal Diseases

Since ascospores are transported by wind over long distances, the control strategy should be the same in all banana plantations to prevent any disruption. The organization of the treatments is more efficient when a centralization of the decision is performed by a unique technical service operating according to rational guidelines rather than if each grower

In Cameroon, *M. fijiensis* was first reported in 1981. In the late 80s, a warning method using biological descriptors has been developed based on the strong experience gained with the control of SLSD in the FWI (Ganry & Laville, 1983; Bureau & Ganry, 1987; Bureau et al., 1982) and on results obtained in Gabon with BLSD on plantain (Fouré 1982a, 1982b, 1983, 1984, 1985; Fouré & Grisoni, 1984, Fouré et al., 1984). It was successfully applied, thus limiting the number of applications to 12-14 per year. This rational control by warning relied heavily on the use of systemic fungicides with a high curative effect (Fouré, 1988a, 1988b,

This situation was sustainable for 10 years, but disease control then became unsuccessful due to logistic failures (shortage of airplanes), which led to the more intensive use of

As a consequence, since 1996 the emergence of strains resistant to systemic fungicides resulted in the progressive replacement of this rational control strategy by more frequent applications of systemic fungicides. As a result, resistance to site-specific strobilurins has developed particularly swiftly. Systemic fungicides have been progressively abandoned and replaced by contact fungicides, with chlorothalonil the most frequently used (Fig. 4). The contact fungicides do not cause the emergence of resistant strains, but have not the curative

Fig. 5. History of fungicide use for BLSD control in a representative commercial banana

farm in Cameroon from 1985 to 2006 (de Lapeyre de Bellaire et al., 2009).

**2.3 BLSD control in Cameroon in the banana industry for export**

1988c; Fouré & Mouliom Pefoura, 1988; Fouré & Moreau, 1992).

effect required for a prolonged action, as in the case of systemic's.

implements his own strategy.

systemic fungicides.

Resistance to benzimidazoles and strobilurins has actually dropped since they are no longer used. But we still need a background of resistance that can even advise their use. However there is every reason to be optimistic about the possibility of reuse, at the horizon of a few years in some areas under warning strategies that would also benefit the registration of systemic fungicides with new mode of action, like succinate deshydrogenase inhibitors (SDHI) that will be soon released. With regard to the triazoles, the sensitivity deteriorates steadily mainly because they still are used in most plantations. Their use should be postponed in order to recover sensitivity.

#### **Box 8.**

Several mechanisms may be behind this recent drop of resistance levels observed in Cameroon:



#### **2.4 BLSD control in Gabon in small scale plantain production for domestic markets**

Here are briefly presented the activities conducted on the agro-industrial plantation of plantains at N'toum (100 ha, 60 km from Libreville, Gabon) created in the late 1980s to supply the urban markets of Libreville

An Integrated Approach to Control the Black Leaf Streak

Fig. 6. Geographic expansion of BLSD in the Americas

management decisions (Marin, 2003)

**Belize** 

**2.6 BLSD control in Latin America in the banana industry for export: The case of** 

The forecasting system described in previous chapters, has been adapted and implemented in several countries in Latin America (Marin, 2003). In some situations in Central America, the commercial adaptation of the early warning system resulted in a significant reduction in the number of fungicide applications. However, due to various factors related to weather patterns and high resistance levels, there was a "come-back" to systematic sprays with contact fungicides. Some commercial programs still use the system only to help

In Belize, BLSD is controlled on an area of 2600 Ha through aerial spraying of fungicides. The control of the disease is centrally managed by the Banana Growers Association (BGA). The Sigatoka service of BGA is accountable for the number of cycles, timing of applications and for the type of fungicides used during the campaign for BLSD control on each farm belonging to a same pedo-climatic area. At the beginning of 2006, the Sigatoka service of BGA was in charge of weekly disease assessment in the different farms, based on the evaluation of the youngest leaf with visible streaks from the ground (YLS) and of the total number of functional leaves ( see §1.1 ). Some farms had started to implement disease

assessments for a biological forecasting (de Lapeyre de Bellaire, 2006 & 2007).

Disease (BLSD) of Bananas, while Reducing Fungicide Use and Environmental Impact 215

For the first time in Central Africa, the BLSD was detected in this area of Ntoum in 1980 (Frossard, 1980). Given the pathogenic activity of the fungus, a research program was implemented in this country. It was based on the study of certain aspects of the biology and the epidemiology of the causal agent of the BLSD, *Mycosphaerella fijiensis* and the development of a warning method for a rational chemical control of the disease (Fouré, 1983; 1984, Fouré et al., 1984). Continuous analysis of biological descriptors (observation of various stages and progression of the disease on the foliage) and the use of systemic fungicides has produced very satisfactory results in this intensive plantation of plantain and effectively controlled the BLSD on 100 ha of plantain in Ntoum (ten fungicide sprays / year; alternation of systemic fungicides). One of the reasons why it was possible to reach such good results was related to the situation of these plantain fields which were very isolated into a forest environment, preventing from external contamination.

This program was an important contribution in the adaptation of a biological warning method for BLSD control.

### **2.5 BLSD control in the French West Indies**

Up to recently, the French West Indies islands were still free of the BLSD**.** 

The SLSD has been controlled effectively and at lower economic and environmental costs through the implementation of a pest management strategy based on a warning system over more than 30 years (5-7 treatments / year). It has been already described in the §1.

Recent developments in the phytosanitary regulations in France (withdrawing of most fungicides and difficulty for registration of new active ingredients) resulted in a sharp decrease of the 'in vitro' susceptibility of the fungicides that are still registered for the control against the SLSD. Thus, the number of treatments performed each year has increased very recently (10-12 treatments per year). In addition, there was also a change in French legislation on aerial spraying and the establishment of untreated buffer zones in a distance of 50 m from houses, gardens, rivers, roads…. Ultimately, it is possible that the aerial treatments will be prohibited.

Recent developments in the spread of the BLSD in the Greater Antilles and more recently in the Lesser Antilles (Fig. 5) suggested that its arrival in Martinique (and probably a later deadline in Guadeloupe) was inevitable in a more or less short term.

Inexorably, BLSD was detected in Martinique in September 2010 and since then it has spread very fast, making it unfeasible any eradication attempt.

Thus, the unique solution is to apply an integrated disease management approach based on the key principles inherited from the lessons learned with BLSD in other areas, and taking into account the excellent know-how of the banana industry in control of the SLSD through a forecasting system.

The effectiveness of pest management is also based on the common management and centralized mode of organization that is working out in Martinique and Guadeloupe for over 30 years. This type of organization is undeniably a major asset for the successful implementation of this strategy. Nevertheless, the implementation of this strategy might be hampered by the limited number of fungicides registered in the FWI for BLSD control and by the effective regulation of BLSD in the 50 m buffer areas.

For the first time in Central Africa, the BLSD was detected in this area of Ntoum in 1980 (Frossard, 1980). Given the pathogenic activity of the fungus, a research program was implemented in this country. It was based on the study of certain aspects of the biology and the epidemiology of the causal agent of the BLSD, *Mycosphaerella fijiensis* and the development of a warning method for a rational chemical control of the disease (Fouré, 1983; 1984, Fouré et al., 1984). Continuous analysis of biological descriptors (observation of various stages and progression of the disease on the foliage) and the use of systemic fungicides has produced very satisfactory results in this intensive plantation of plantain and effectively controlled the BLSD on 100 ha of plantain in Ntoum (ten fungicide sprays / year; alternation of systemic fungicides). One of the reasons why it was possible to reach such good results was related to the situation of these plantain fields which were very isolated

This program was an important contribution in the adaptation of a biological warning

The SLSD has been controlled effectively and at lower economic and environmental costs through the implementation of a pest management strategy based on a warning system over

Recent developments in the phytosanitary regulations in France (withdrawing of most fungicides and difficulty for registration of new active ingredients) resulted in a sharp decrease of the 'in vitro' susceptibility of the fungicides that are still registered for the control against the SLSD. Thus, the number of treatments performed each year has increased very recently (10-12 treatments per year). In addition, there was also a change in French legislation on aerial spraying and the establishment of untreated buffer zones in a distance of 50 m from houses, gardens, rivers, roads…. Ultimately, it is possible that the

Recent developments in the spread of the BLSD in the Greater Antilles and more recently in the Lesser Antilles (Fig. 5) suggested that its arrival in Martinique (and probably a later

Inexorably, BLSD was detected in Martinique in September 2010 and since then it has

Thus, the unique solution is to apply an integrated disease management approach based on the key principles inherited from the lessons learned with BLSD in other areas, and taking into account the excellent know-how of the banana industry in control of the SLSD through

The effectiveness of pest management is also based on the common management and centralized mode of organization that is working out in Martinique and Guadeloupe for over 30 years. This type of organization is undeniably a major asset for the successful implementation of this strategy. Nevertheless, the implementation of this strategy might be hampered by the limited number of fungicides registered in the FWI for BLSD control and

more than 30 years (5-7 treatments / year). It has been already described in the §1.

into a forest environment, preventing from external contamination.

Up to recently, the French West Indies islands were still free of the BLSD**.** 

deadline in Guadeloupe) was inevitable in a more or less short term.

spread very fast, making it unfeasible any eradication attempt.

by the effective regulation of BLSD in the 50 m buffer areas.

method for BLSD control.

**2.5 BLSD control in the French West Indies** 

aerial treatments will be prohibited.

a forecasting system.

Fig. 6. Geographic expansion of BLSD in the Americas

#### **2.6 BLSD control in Latin America in the banana industry for export: The case of Belize**

The forecasting system described in previous chapters, has been adapted and implemented in several countries in Latin America (Marin, 2003). In some situations in Central America, the commercial adaptation of the early warning system resulted in a significant reduction in the number of fungicide applications. However, due to various factors related to weather patterns and high resistance levels, there was a "come-back" to systematic sprays with contact fungicides. Some commercial programs still use the system only to help management decisions (Marin, 2003)

In Belize, BLSD is controlled on an area of 2600 Ha through aerial spraying of fungicides. The control of the disease is centrally managed by the Banana Growers Association (BGA). The Sigatoka service of BGA is accountable for the number of cycles, timing of applications and for the type of fungicides used during the campaign for BLSD control on each farm belonging to a same pedo-climatic area. At the beginning of 2006, the Sigatoka service of BGA was in charge of weekly disease assessment in the different farms, based on the evaluation of the youngest leaf with visible streaks from the ground (YLS) and of the total number of functional leaves ( see §1.1 ). Some farms had started to implement disease assessments for a biological forecasting (de Lapeyre de Bellaire, 2006 & 2007).

An Integrated Approach to Control the Black Leaf Streak

high (de Lapeyre de Bellaire, 2007).

deterioration caused by the disease.

country.

**Panama** 

aircraft.

sprays

Disease (BLSD) of Bananas, while Reducing Fungicide Use and Environmental Impact 217

In such conditions, satisfactory control could be achieved through a systematic use of contact fungicides, such as mancozeb and chlorothalonil. However, this strategy does not reduce the environmental impact since the number of fungicide applications/year remained

Nevertheless, as observed in Cameroon, this strategy could create the opportunity for a possible reversion of resistance, especially for IBS of group 1 and give the possibility to implement in the future the biological forecasting system developed by CIRAD in this

In Panama, as in other Central American countries, a very significant decrease in the production of plantains was observed after the emergence of BLSD. The production in this country was divided by three in five years, from 100,910 tons in 1979 to 31,134 tons in 1984 (Diaz, 1986). The effects were very sensitive on the supply of domestic markets as well as on the prices that have almost doubled over this period. In addition all exports, while expanding (661 tons in 1980, 2338 tons in 1982) were stopped due to a very high quality

It was therefore considered interesting to use the experience gained with the control of the

The warning system has been implemented on two types of production: small farms of 4-5 ha on average, with a low technicality and family labor and medium farms of 10 to 30

In the first case the control was done through ground sprays and in the second case by

The results were very promising with a very good control of BLSD in both situations with nine applications a year in family farms and 6.5 applications in medium-sized farms. Such promising results have been achieved through a rigorous execution giving its full preventive character to the system. It is interesting to notice that under these conditions, a very good relationship was found between the evaporation Piche and the duration of treatment efficacy. It appears that the aerial sprays are much more efficient than ground

Apart from Panama, the forecasting system was also used on plantain in other countries such as Costa-Rica. In this latter country the forecasting system has been simplified (Marin, 1992) as well as combined with climatic factors to develop a bioclimatic forecasting system

Beyond the biological and technical components there are economic, political, logistical, environmental and social issues that are key points to consider in a holistic approach,

Sigatoka Disease in banana plantations of the French West Indies (Bureau, 1990).

hectare with a good technical level, hired labor and more productive.

**3. Lessons drawn from various experiences and perspectives**

for plantain (Jimenez et al., 1995; Lescot et al., 1998)

otherwise leading to unpredictable failures.

**2.7 BLSD control in Latin America in plantain production for export: The case of** 

The centralization of BLSD control was very suitable for disease management because it guarantees that a same technical guideline is used over the whole banana area. Nevertheless, an indirect negative consequence of this centralization was that banana growers were less involved directly in BLSD, especially for leaf removal.

Three airplanes and two pilots were available for spraying, which is the minimum to control the disease over 2600 ha in order to optimize spraying during the best conditions (small window in the morning).

Fungicides used for BLSD control fall in the 2 categories described earlier: protectants, and systemic fungicides (Box 3).

The spraying program was generally based on the use of contact fungicides during the dry season (February to May) and systemic fungicides being mainly used in the rainy season.

Many fungicides are registered for BLSD control in Belize (Table 3), but despite the intensive use of systemic fungicides in recent years (Fig. 6), control of the disease was not successful at that time, and significant losses were registered (de Lapeyre de Bellaire, 2006). High disease outbreaks resulted in bunch reject, bunch weight losses and strong quality problems linked with early ripening and heterogeneity of ripening in ripening rooms.

Fig. 7. History of fungicide use for BLSD control in a representative commercial banana farm in Belize (1995-2006) (from de Lapeyre de Bellaire, 2006)

In 2006, a fungicide resistance monitoring campaign was achieved thanks to the contribution of CIRAD's laboratory in Montpellier. The results of this monitoring analysis showed a very worrying situation in terms of fungicide resistance in all the commercial banana farms. For the three chemical groups evaluated (strobilurins, antimitotics and IBS group 1) high levels of resistance have been observed (de Lapeyre de Bellaire, 2006), which explained the poor control obtained with systemic fungicides in this country.

The centralization of BLSD control was very suitable for disease management because it guarantees that a same technical guideline is used over the whole banana area. Nevertheless, an indirect negative consequence of this centralization was that banana

Three airplanes and two pilots were available for spraying, which is the minimum to control the disease over 2600 ha in order to optimize spraying during the best conditions (small

Fungicides used for BLSD control fall in the 2 categories described earlier: protectants, and

The spraying program was generally based on the use of contact fungicides during the dry season (February to May) and systemic fungicides being mainly used in the rainy season.

Many fungicides are registered for BLSD control in Belize (Table 3), but despite the intensive use of systemic fungicides in recent years (Fig. 6), control of the disease was not successful at that time, and significant losses were registered (de Lapeyre de Bellaire, 2006). High disease outbreaks resulted in bunch reject, bunch weight losses and strong quality problems linked

Fig. 7. History of fungicide use for BLSD control in a representative commercial banana

explained the poor control obtained with systemic fungicides in this country.

In 2006, a fungicide resistance monitoring campaign was achieved thanks to the contribution of CIRAD's laboratory in Montpellier. The results of this monitoring analysis showed a very worrying situation in terms of fungicide resistance in all the commercial banana farms. For the three chemical groups evaluated (strobilurins, antimitotics and IBS group 1) high levels of resistance have been observed (de Lapeyre de Bellaire, 2006), which

farm in Belize (1995-2006) (from de Lapeyre de Bellaire, 2006)

growers were less involved directly in BLSD, especially for leaf removal.

with early ripening and heterogeneity of ripening in ripening rooms.

window in the morning).

systemic fungicides (Box 3).

In such conditions, satisfactory control could be achieved through a systematic use of contact fungicides, such as mancozeb and chlorothalonil. However, this strategy does not reduce the environmental impact since the number of fungicide applications/year remained high (de Lapeyre de Bellaire, 2007).

Nevertheless, as observed in Cameroon, this strategy could create the opportunity for a possible reversion of resistance, especially for IBS of group 1 and give the possibility to implement in the future the biological forecasting system developed by CIRAD in this country.

#### **2.7 BLSD control in Latin America in plantain production for export: The case of Panama**

In Panama, as in other Central American countries, a very significant decrease in the production of plantains was observed after the emergence of BLSD. The production in this country was divided by three in five years, from 100,910 tons in 1979 to 31,134 tons in 1984 (Diaz, 1986). The effects were very sensitive on the supply of domestic markets as well as on the prices that have almost doubled over this period. In addition all exports, while expanding (661 tons in 1980, 2338 tons in 1982) were stopped due to a very high quality deterioration caused by the disease.

It was therefore considered interesting to use the experience gained with the control of the Sigatoka Disease in banana plantations of the French West Indies (Bureau, 1990).

The warning system has been implemented on two types of production: small farms of 4-5 ha on average, with a low technicality and family labor and medium farms of 10 to 30 hectare with a good technical level, hired labor and more productive.

In the first case the control was done through ground sprays and in the second case by aircraft.

The results were very promising with a very good control of BLSD in both situations with nine applications a year in family farms and 6.5 applications in medium-sized farms. Such promising results have been achieved through a rigorous execution giving its full preventive character to the system. It is interesting to notice that under these conditions, a very good relationship was found between the evaporation Piche and the duration of treatment efficacy. It appears that the aerial sprays are much more efficient than ground sprays

Apart from Panama, the forecasting system was also used on plantain in other countries such as Costa-Rica. In this latter country the forecasting system has been simplified (Marin, 1992) as well as combined with climatic factors to develop a bioclimatic forecasting system for plantain (Jimenez et al., 1995; Lescot et al., 1998)
