Download

Contact

Communication service

E-mail

Annual Report: CIRAD 2008

Designing pesticide-free banana plantations— modelling contributions

To design non-polluting high performance cropping systems, pesticide treatments must be replaced by targeted use of ecological mechanisms. But how can these complex systems, in which there are more biological interactions than in chemical-based systems, be managed? And what system is best for achieving high crop productivity while preserving the environment? These issues could be addressed by using models to integrate all knowledge on agrosystems and predict the effects of innovative practices on its functioning. This approach is being used by CIRAD to design environmentally friendly and economically efficient banana-based systems.

Ecological mechanisms must be implemented at specifically targeted times and points in cropping systems instead of pesticide treatments. For banana plantations, CIRAD has developed modelling tools to integrate knowledge on ecology, agronomy and crop protection and to simulate agrosystem functioning. These tools can predict the impacts of innovative combinations of cropping practices and provide assistance in selecting the most promising production systems from both environmental and economic standpoints. When used in this way, modelling also helps to pinpoint areas in which further knowledge is required for developing new production methods.

Integrating knowledge in prediction tools

Many physical, chemical and biological processes interact within banana plantations, as in all agrosystems. Some of these processes have to be modified when designing innovative cropping systems, which generally leads to changes in the overall functioning of the agrosystem. For instance, adding cover crops leads to major changes in water and mineral flows and in pest and pathogen dynamics. A systemic approach to relationships between the plant, pests and the environment is required to gain insight into and describe the functioning of these new systems, which are more complex than those based on massive use of inputs. Models are a systemic representation of the agrosystem function, which can be parameterized as a function of the value of the studied variables, is a model. Such models are tools that can assist researchers in representing these ecological functions and in designing new cropping systems.

A CIRAD team designed a specific model, called SIMBA, which simulates banana plantation functioning and performance. This model can be parameterized according to different variables, such as climatic data (temperature, radiation, rainfall), soil data (porosity, organic matter level) and initial parasitic nematode population levels (Radopholus similis, Pratylenchus coffeae and Helicotilenchus multicinctus) . The model can simulate banana crop growth and yield, pest development dynamics, soil moisture and nitrogen levels, as well as the growth and impact of cover and weed vegetation. Moreover, it assesses the environmental risks of simulated systems, including nitrogen losses, pesticide water pollution and erosion. This model, called SIMBA, generates agronomic, environmental and economic data (eg gross margins). Output simulations can be used for multicriteria assessment of virtual or real cropping systems.

A special feature of this model is that it accounts for changes in the plant population structure, which becomes less homogeneous over cropping cycles: this key point affects all the system dynamics. Moreover, the SIMBA model includes plant-parasitic nematode dynamics and the interactions with plant growth, soil structure and nematicide treatments; this is a major element in simulating the agroenvironmental performance (sustainability) of simulated banana plantations. Finally, the model can be used for in-depth studies on certain ecological mechanisms, such as host-parasite relationships, and for developing and optimizing new control strategies, such as crop rotations or notill systems based on cover crops.

Plot schemes to reduce epidemics

Inter- and intra-plot spatial organization and spatialization of crop protection resources have long been overlooked, even though these are key elements for pest and disease control. Because of its moderate dispersal, the banana weevil is a good example for studying the effect of the spatial organization of cropping systems on its epidemiology. CIRAD researchers have thus developed a spatially explicit model, COSMOS, which simulates the insect’s movements, reproduction and mortality in interaction with banana trees, crop residues, cover plants and pheromone traps. The COSMOS model can be used to optimize trapping, and proposes landscape plot patterns that hamper the development of banana weevil populations.

Towards simulation of the overall ecological functioning of banana plantations

Current agroecological research shows that it is essential to design new agroecosystem study methods in order to be able to monitor the ecological processes occurring in more complex cropping systems. In this setting, interactions between

Contact

Philippe Tixier
Banana, Plantain and Pineapple Cropping Systems (UPR)
E-mail

Partners

  • Union des groupements de producteurs de bananes de Guadeloupe et Martinique (UGPBAN, France)
  • Centre africain de recherche sur les bananiers et plantains (CARBAP, Cameroon)
  • European Regional Development Fund (ERDF)
  • Direction régionale de la recherche et de la technologie de Guadeloupe et de Martinique (France)

For further information

  • Tixier P., Malézieux E., Dorel M., 2007. A model-based approach to maximise gross income by selection of banana planting date. Biosystems Engineering, 6 : 471-476.
  • Tixier P., Malézieux E., Dorel M., Wery J., 2008. SIMBA: a comprehensive model for evaluation and prototyping of banana-based cropping systems. Agricultural Systems, 97 : 139-150.
  • Tixier P., Salmon S., Chabrier C., Quénéhervé P., 2008. Modelling pest dynamics of new crop cultivars: The FB920 banana with the Helicotylenchus multicinctus-Radopholus similis nematode complex in Martinique. Crop Protection, 27 : 1427-1431.