Search

Soil tillage and cover crops impact spontaneous vegetation in agrosystems. Soil tillage destroys seedlings and can affect the soil seedbank. Cover crops compete with naturally occurring spontaneous species. Our aim was to gain insight into how soil tillage (here, the false seedbed technique) and cover crops impact the spontaneous vegetation in fallows preceding banana crops. In a fallowed field in Guadeloupe (French West Indies), two factors were combined in a split-plot design: the soil tillage modality before cover crop sowing (with or without the false seedbed technique) and the cover crop species (Crotalaria spectabilis, Pueraria phaseoloïdes, a 50/50 mix of Brachiaria decumbens + Brachiaria ruziziensis, a spontaneous vegetation control). On a monthly basis, in three permanent quadrats per sub-plot, all spontaneous species were identified and their numbers counted, while calculating the species richness and Shannon's diversity index. The spontaneous vegetation dynamics contrasted depending on the treatment. From 7 months post-sowing the species density, richness and Shannon index were significantly lower in treatments with B. decumbens + B. ruziziensis and P. phaseoloïdes, while in treatments with C. spectabilis, the species density and richness were significantly higher with than without the false seedbed technique. These effects were probably related to differences in the competitive ability of the cover crop used and, to a lesser extent, to the false seedbed effects on the soil seedbank. This is the first study to show that under a tropical climate, sown cover crop species combined with different soil tillage modalities have substantial impacts on the species density, richness and diversity of the spontaneous plant community over a 1-year period. These results could help farmers in their weed control strategies by providing them with a deeper understanding of the effects of practices on the spontaneous plant community.

The data presented in this data paper describe the early growth of cover crop cultivated in growth chamber under non-limiting conditions. Seventeen species of four botanical groups were described after one month of growth. Traits related to plant growth and leaf area development were measured (five traits) and calculated (eight traits). This data set is made available to enable comparisons between dataset, extended analysis and meta-analysis on cover crop traits. The data presented in this article were used on the research article entitled “Leaf area development strategies of cover plants used in banana plantations identified from a set of plant traits’ [1]. Keywords: Cover crops, Plant growth, Leaf area, Leaf traits, Plant traits

Black leaf streak (BLS), caused by the phytopathogenic fungus Pseudocercospora fijiensis, leads to leaf necrosis and often to a substantial reduction in fruit yield. It also causes the fruit to ripen early by reducing the time between harvest and the beginning of natural ripening, called green life (GL). The usual method of control is by spraying fungicides, but this type of management is challenged worldwide due to the high costs, regulation of products, environmental concerns and the development of resistance. Integrated pest control is an alternative management method based on the combination of agronomic practices that can slow down the development of the epidemic, and limit damage to the fruit. Two experiments were conducted in the Dominican Republic and Martinique, to evaluate the performance of an innovative prototype cropping system with no chemical control. This system is based on optimal growth management, to reduce stress and enhance the leaf emission rate, combined with the systematic removal of necrotic parts of the leaves, which strongly limits ascopore sporulation and reduces the effect of BLS on GL. Depending on the conditions and cycles, the reduction in bunch weight with the prototype system versus chemical control ranged from O to 15%. Bananas were harvested at the same physiological age in the prototype and in the control. The bananas harvested from plants with as low as zero leaves left at harvest retained a good aptitude for conservation, with up to 40 days of GL, on average. The limited impact on fruit weight observed suggests that physiological compensation phenomena through remobilization of resources inside the plant (stem, bunch, suckers) occur. Above all, the removal of necrotic leaves prevented a significant negative effect on early fruit ripening when the bananas were harvested at the correct physiological age, and enabled export of the fruit despite the small number of leaves remaining on the plant at harvest. These results offer perspectives for the development and transfer of alternative practices to chemical control.

Cover crop-based cropping systems have gained considerable attention in recent years because of the recognized benefits of increasing agroecosystem diversity to enhance ecological functions and agroecosystem services through positive effects between species. While agronomic tools and methods to assess these services in multispecies plant communities barely exist, trait-based approaches, originally developed in the field of comparative functional ecology, provide an appropriate framework to study the effects of plant diversity on agroecosystems. We present our theoretical and applied research conducted over recent years using a trait-based approach to help improve cover crop choice to sustain agroecosystem services in banana agroecosystems of the French West Indies. We first provide a conceptual framework for assessing expected agroecosystem services from cover crops in banana agroecosystems, i.e., weed control, plant-parasitic nematode control, improvement of the nutrient cycling and avoidance of interference for resource acquisition in banana (dis-service), based on plant effect traits. Then, we experimentally characterized a collection of cover crop species to identify their resource acquisition strategies and the potential trade-offs they reveal. More practically, we characterized the cover crop species according to their potential to deliver four agroecosystem services expected from banana/cover crop systems. We constructed functional profiles of the cover crop species, and compared these profiles to cover crop ideotypes to select the best-suited species for each service. Finally, we used "engineering traits" to help design and optimize design and management of cover crop species mixtures. These examples from our research demonstrate that functional, trait-based approaches have strong potential to tackle agronomic issues and to resolve fundamental and applied questions related to the design of multispecies cropping systems that optimize multiple services.

The limitations of conventional farming systems become apparent when confronted with the challenges of sustainability. Conventional agricultural practices tend to degrade soil fertility and result in environmental damage and health risks through their various activities, including tillage, water use and management, and the use of fertilizers and pesticides, thus making these practices no longer acceptable from the economic and social points of view. CIRAD and its partners are therefore working on an agroecological transition founded on new ecological intensification practices that primarily consist of the elimination of tillage to help revive the soil’s biological activity, use of cover crops to boost nutrient recycling and the organic-mineral management of crops, maintenance of a permanent cover to improve water efficiency and to control erosion, and the recourse to biological control (functional biodiversity and cropping practices) of pests to limit the damage caused by them. Implementing these practices implies reviewing cropping system design methods and processes to better integrate the dimensions of actors, space, and time. It is indeed essential to involve or engage producers and other local actors in a more comprehensive manner, take different levels into account (cropping systems, production systems, markets, global changes, land use, etc.) as well as the temporal aspects of this agroecological transition, promote trajectories of progressive adaptation, and envisage modalities of providing support to actors. This process can give rise to technical solutions that can help enhance the sustainability of agricultural production systems.

Cover crops introduced into cropping systems can lead to chemical input reductions and pollution mitigation because they enhance ecological functions. The choice of the best cover crops for a specific cropping system is, however, difficult because of the large range of potential cover species. A promising method involves functional traits as simplified indicators of plant functions. In banana cropping systems, cover crops are used especially to control weeds by development of their leaf area to boost competition for light. The aim of this study was to seek trait-based leaf area development strategies among tropical cover species, based on four plant traits chosen because of their mathematical link with leaf area development: specific leaf area (SLA), aboveground leaf mass fraction (LMFa), seed mass (SM) and aboveground relative growth rate (RGRa). We measured trait values and leaf areas of 17 tropical cover species grown for 1 month in a growth chamber. Strong positive and negative covariations were observed between SM, LMFa and RGRa, revealing a “syndrome” of traits and suggesting trade-offs between traits. Four groups of species were identified based on PCA and cluster analyses and were characterized by significantly different sets of trait values. They showed four leaf area development strategies: species that allocate a large part of biomass to leaf area (G1), species that develop large biomass and leaf area at emergence (G2), species with rapid biomass growth and low biomass investment in leaves (G3) and species with a non-specialized strategy (G4). After 1 month, species of groups G1 and G2 had higher leaf area, although not significantly, than species of groups G3 and G4. Comparisons between this functional classification and the taxonomic monocot/dicot classification showed that the functional classification captured a larger part of the variability in traits involved in leaf area development than the taxonomic monocot/dicot classification. This encourages the use of such a classification to describe plant functioning, to understand plant roles in plant–plant interactions and guide the choice of the best cover species to introduce into cropping systems.

Dessert banana production worldwide is based on just a few highly productive, easy to grow and international market compatible cultivars from the Cavendish sub-group (AAA genome). However, these cultivars are hampered by pests and diseases, which has led to extensive pesticide use. New pest and disease resistant varieties that are productive enough to meet international market demand are now needed to reduce pesticide use and thus the detrimental environmental impact. Hybrid bananas bred by the CIRAD banana breeding program are resistant to many diseases, particularly Sigatoka diseases, but exhibit morphological traits and yield components which differ markedly from those of Cavendish bananas. After an analysis of traits related to plant morphology and yield in 530 banana hybrids, the effects of modifying the level of competition between bunches and suckers through sucker pruning management practices were compared on the new hybrid CIRAD 925 and on a reference variety from the Cavendish sub-group ‘Grande Naine’. Hybrids from the CIRAD banana breeding program tend to have thin stems, small bunches and a short crop cycle. Total sucker pruning until mother plant harvest significantly increased the fruit diameter, first cycle bunch weight, second cycle duration and decreased the plant size. However, the effect on bunch weight was greater for Grande Naine, while the effects on second cycle duration and pseudostem girth were greater for CIRAD 925. The level of apical dominance of the central meristem on suckers seemed to explain the varietal differences in the effects of sucker pruning management. Our results showed that traits related to plant morphology and yield components could be modified by sucker pruning management and that a tailored management of mother-plant/sucker competition could be a way to improve the suitability of new hybrids for the banana industry.

Publication Inra prise en compte dans l'analyse bibliométrique des publications scientifiques mondiales sur les Fruits, les Légumes et la Pomme de terre. Période 2000-2012. http://prodinra.inra.fr/record/256699; Nematodes are major pests for crops, including banana. Environmentally friendly methods for managing plant-parasitic nematodes have to be developed, such as organic material application. Our study focuses on the impacts of several organic amendments on banana plants, considering mainly their effect on soil nitrogen supply and soil microbial biomass, and the consequences on plant-parasitic nematode impacts on the plants. A microcosm experiment for 13 weeks was conducted to evaluate four organic materials: sugarcane bagasse, sugarcane sludge, plant residues, and sewage sludge, compared to a control without organic amendment. Input of organic materials led to an important change on nitrogen resource, and plants grew better when the N availability was the highest, but better growth conditions did not necessarily reduce parasitic nematodes impacts on the roots. Damage on the roots depended on plant-parasitic nematode abundance. Three of four tested amendments exhibited a regulator effect on plant-parasitic nematode populations (bagasse, sugarcane sludge, and plant residues). Root growth was not the explanatory factor for this regulation. Only sugarcane sludge led to an overall positive effect on the plant, increasing its growth and reducing its parasitism pressure. The other organic materials exhibited an antagonism between the promoted plant growth and the reduced nematode populations.

1 The banana weevil Cosmopolites sordidus Germar (Coleoptera: Curculionidae) is the most serious pest of banana and plantain in most production areas, including the West Indies. 2 During a period of 2 years, we assessed the efficacy of a pheromone masstrapping control strategy of C. sordidus in field conditions in Guadeloupe at different cropping stages, both in sanitation fallows and in different ratoon banana crops. 3 In the fallows, catches peaked 3 months after beginning trapping and then decreased to zero after 9 months. By contrast, for the new plantations, the catches of C. sordidus increased after the 11th month and, in the older banana fields, there was no decrease in C. sordidus catches. The C. sordidus catches increased in the neighbouring banana plots, whereas they decreased in the fallows, and these catches decreased with the distance from fallow. 4 In conclusion, mass trapping with synergized pheromone traps within fallows should allow better sanitation of banana plantations. Yet, within the farms, fallows must not be located next to new plantations to avoid massive damage to the young plants. More generally, the present study emphasizes that the control of this insect should be managed at the farm scale and not at the field scale, with special attention being paid to the location of fallows. (Resume d'auteur)

Introduction. Aux Antilles francaises, les planteurs considerent generalement qu'il est necessaire d'effectuer periodiquement un travail du sol pour regenerer la porosite du sol et maintenir un fort rendement. Cependant, dans les bananeraies perennes d'altitude, le sol qui n'est jamais travaille presente de meilleures proprietes physiques et biologiques que dans bananeraies conduites de maniere conventionnelle. Pour determiner si le travail du sol avant la plantation est necessaire au bon fonctionnement d'une culture de bananiers et pour evaluer l'effet du labour sur la qualite du sol, une plantation de bananiers apres travail du sol classique a ete comparee a la plantation directe de bananiers sur un paillis de residus de recolte en parcelle experimentale. Materiel et methodes. La qualite du sol a ete evaluee par des indicateurs tels que la porosite, le statut organique, la biomasse microbienne et la structure des populations de nematodes. Le fonctionnement de la culture a ete evalue par la croissance de la plante, la distribution des racines, et la disponibilite de l'eau et de l'azote du sol. Resultats. Nous avons constate que le travail du sol a reduit la biomasse microbienne du sol et le nombre de guildes fonctionnelles de nematodes. Le travail du sol a eu seulement un effet a court terme sur la porosite du sol et n'a pas permis un enracinement plus profond. Bien que la mineralisation de l'azote organique du sol ait ete plus elevee avec le travail du sol conventionnel, la nutrition en azote des bananiers n'a pas ete meilleure, probablement parce que la fertilisation azotee elevee a compense les variations de la disponibilite de l'azote d'origine organique. Nous avons constate que la croissance des bananiers etait meilleure dans le traitement sans travail du sol. Cela a pu etre explique par un plus faible dessechement du sol du au paillis de residus de recolte laisse sur la surface du sol pour le traitement sans travail du sol. Conclusion. Par rapport au travail du sol classique, une plantation sans labour a ameliore la qualite du sol et les performances de la culture.

International audience; Farmers are inundated with advertisements about many innovations that are supposed to increase their yields or reduce environmental impact. However, the benefit of these innovations depends on the farming context. Here we present the ad hoc adaptation of the crop model SIMBA and a method to evaluate 16 innovations in six types of farms previously selected through a typology of the banana farming systems in Guadeloupe. The innovations include regulation of pesticide use, rotations and fallows, intercropping, conditional application of pesticides, resistant cultivars, and integrated systems. Our results show that, for a given innovation, the yield and pesticide reduction vary widely with different farm types. We show that environmentally friendly innovations often cause a greater decrease in yield in more productive farm types. Nevertheless, despite an apparent trade-off between yield and pesticide use, some innovations address both production and environmental issues, e.g., rotation with fallows improved with cover crops, regular fallows, and rotations with pineapples for the most intensive farm types. Our modelling study confirms the importance of innovation-farm type interactions and the usefulness of models for assessing large numbers of technological innovations among a wide range of biophysical and technical contexts.

Banana plantlets (Musa acuminata cv Grande Naine) cultivated in hydroponics take up silicon proportionally to the concentration of Si in the nutrient solution (0-1.66 mM Si). Here we study the Si status of banana plantlets grown under controlled greenhouse conditions on five soils developed from andesitic volcanic ash, but differing in weathering stage. The mineralogical composition of soils was inferred from X-ray diffraction, elemental analysis and selective chemical/mineralogical extractions. With increasing weathering, the content of weatherable primary minerals decreased. Conversely, clay content increased and stable secondary minerals were increasingly dominant: gibbsite, Fe oxides, allophane, halloysite and kaolinite. The contents of biogenic Si in plant and soil were governed by the reserve of weatherable primary minerals. The largest concentrations of biogenic Si in plant (6.9-7 g kg(-1)) and soil (50-58 g kg(-1)) occurred in the least weathered soils, where total Si content was above 225 g kg(-1). The lowest contents of biogenic Si in plant (2.8-4.3 g kg(-1)) and soil (8-31 g kg(-1)) occurred in the most weathered desilicated soils enriched with secondary oxides and clay minerals. Our data imply that soil weathering stage directly impacted the soil-to-plant transfer of silicon, and thereby the stock of biogenic Si in a soil-plant system involving a Si-accumulating plant. They further imply that soil type can influence the silicon soil-plant cycle and its hydrological output.

In banana plantations of the Caribbean, nitrogen (N) fertilization widely exceeds nutrient outputs after harvest. Under wet tropical climate, leaching results in considerable waste of N. Fertilization management aims at maintaining soil mineral N at the optimal level for banana nutrition throughout the year but it does not take into account variations in crop N demand or N supply through mineralization of crop residues. The dynamics of crop N demand and crop residue supply depend on the structure of banana populations, which become asynchronous with time. We designed the SIMBA-N model to simulate N dynamics in successive crop cycles of banana. The model calculates the N balance weekly, including N uptake by banana, N leaching, and N supply by organic matter mineralization. We validated the model using data from a field experiment comparing five levels of fertilization. Results showed SIMBA-N provides reliable indicators to support banana fertilization management taking into account N flows in the soil and change in N demand related to banana population structure.

Several plant species accumulate silicon, which is taken up by roots in soil solution. The Si concentration in soil solution can be governed by silicate dissolution and formation, and thus soil constitution. Here, we study the Si leaf content of mature banana plants (Musa acuminata cv Grande Naine) cropped on soils derived from andesitic ash in Guadeloupe through standard foliar analysis. The soils strongly differ in weathering stage and total Si content. The most desilicated soils (Andosol-Nitisol Ferralsol) occur in the wettest areas, on the Eastern slopes (Es) of the volcano exposed to rain bearing winds. Least weathered soils (Andosol-Cambisol) occur on Western slopes (Ws). The average leaf Si concentration ranges from 2.7 to 3.9 g kg(-1) for bananas cropped in Es soils, and from 7.7 to 9.6 g kg(-1) in Ws soils. The leaf Si concentrations are lowest for the Es gibbsite-rich Andosols and Ferralsols. The leaf Si concentration is positively correlated with soil CaCl2-extractable Si content, soil Si content and total reserve in weatherable minerals. The silicon content of banana leaves thus reveals the weathering stage of volcanic ash soils in Guadeloupe.

Correspondance: tixier@cirad.fr; International audience; Banana monocultures (Musa spp., AAA, Cavendish sub-group cv. Grande Naine) can have a detrimental impact on the environment. In these agro-systems, pesticide treatments can lead to surface and groundwater pollution, as is the case in the tropical insular conditions of the French West Indies. Using models to design alternative cropping systems is of growing interest but most of the research work has been concentrated on annual crops and most often in temperate climate. A specific model called SIMBA was built to assess environmental risks under a large range of cropping techniques and to help design more sustainable cropping systems. SIMBA simulates banana-cropping systems at field level over several cropping cycles. It includes sub-models that simulate soil structure, water balance, root nematode populations, yield, and economic outputs with a sound balance between representing the major phenomena well and keeping the model simple to reduce the parameterization costs in a large range of conditions. Agro-environmental indicators generated by the model make it possible to assess the major potential environmental impacts. The model has been developed and calibrated in Guadeloupe and Martinique and is used to draw up practical recommendations for farmers and for virtual experiments of agro-technological innovations or field management strategies. The structure of SIMBA is presented and a methodology is proposed for designing sustainable banana-based cropping systems using the model. SIMBA has been evaluated in a broad range of cropping systems in Guadeloupe by comparing model estimates to data collected in field experiments and surveys. Simulations lead to trends in rotation-based cropping systems characterized by systems that can be considered as intensive for profit evaluation, and combinations of frequent replanting, low nematicide application, no ploughing, and low fertilization level, for environmental evaluation. Simulations performed to optimize the replanting decision rule showed that relatively frequent replanting is good for profit while low frequency replantations (over four banana cycles) give a better environmental evaluation

Publication Inra prise en compte dans l'analyse bibliométrique des publications scientifiques mondiales sur les Fruits, les Légumes et la Pomme de terre. Période 2000-2012. http://prodinra.inra.fr/record/256699; International audience; Radopholus similis is a worldwide endoparasitic nematode that greatly hampers banana (Musa acuminata, Cavendish subgroup) productivity. Earthworms are known to closely interact with above-ground and under-ground soil biota and particularly with plants and microfaunal communities. This study was aimed at investigating, under greenhouse conditions, the effects of the earthworm Pontoscolex corethrurus on banana growth and nutrient uptake, and assessing the influences of this earthworm on the development of an inoculated population of R. similis. Six-week-old tissue culture banana plants were submitted to four treatments: with P. corethrurus, R. similis, P. corethrurus+R. similis, and a control with no earthworms or nematodes. At the end of the experiment, the P. corethrurus treatments showed significantly higher leaf surface areas, shoot dry root weights, and root fresh weights than those without earthworms. This root growth enhancement probably contributed to the evident but non-significant decrease in the density of nematodes in the roots, even though earthworms did not reduce the total number of nematodes per whole root system. Moreover, the presence of earthworms slightly alleviated the severity of root damage. N bioavailability in the soil, along with N, Ca, and Mg content of banana plants, were also significantly increased in the presence of earthworms. Our results demonstrated that banana plant growth and nutrition were positively influenced by earthworms. Cropping practices that boost the development of earthworm communities in soil should therefore be promoted to enhance sustainability and to naturally alleviate nematode impact.

E-mail Addresses: tixier@cirad.fr, dorel@cirad.fr, malezieux@cirad.fr; International audience; The harvest dynamic model (SIMBA-POP) was used to choose the best planting date of a banana field in order to optimise the gross income. The proposed method takes into account the seasonal variation of the selling price of bananas. The gross income was calculated weekly by multiplying the simulated number of bunches harvested weekly by the mean weight of bunches, and by the current selling price. The weekly gross income over the simulated period was then determined. The Study was carried out in Martinique (French West Indies) for climatic conditions corresponding to 50, 350, and 650 m in altitude and for different field management strategies (number of cropping cycles before replanting). The result of our simulations showed that the optimal planting month enables a 10-36% increase in income over the worst month

Like many intensive monocultures, some features of banana-based cropping systems may be detrimental to the environment. Pesticide use is a major cause of surface water and groundwater pollution. The risk of water pollution due to pesticides is very high in the insular areas of the French West Indies (FWI). In order to assess these risks and help design more sustainable cropping systems, we propose using an indicator to assess the risk of pesticide pollution over time (Rpest). Rpest provides dynamic assessments through a linkage with a cropping system model that simulates environmental factors and agricultural practices. An expert validation check was conducted and demonstrated that Rpest can rank cropping systems by risk as well as experts. A sensitivity analysis highlighted that the indicator can take the properties of active ingredients and pedoclimatic data into account in assessments. Rpest helps to pinpoint high pollution risk periods. It can also be used to test alternatives and compare systems. This tool can be integrated into a model-based, global, prototyping methodology used for more sustainable cropping systems.

Cover plant-based cropping systems have gained considerable attention over the last few years because of the recognized benefits of increasing the diversity of agrosystems to sustain services other than those pertaining to production. While agronomical tools and methods to assess these services in multispecies plant communities barely exist, trait-based approaches, originally developed in the field of comparative functional ecology, provide an appropriated framework to study the effects of plant diversity on agrosystems. The aim of this article is to discuss how a trait-based approach can be used to assess the services delivered in cover plant-based cropping system. We focus on semiperennial cropping systems based on cover plants, exemplified by the banana cropping systems. These systems are described in a first section to identify the services targeted. A second section, based on a literature survey, analyzes the processes associated with each service and identifies related effect traits of cover plants. A shortlist of markers to be used to assess agrosystem processes is proposed and hypothesis on trade-offs and synergies among services are formulated. In a third section, issues related to the scaling-up from plant traits to the services delivered by plant agricultural communities are discussed. Such an approach, originated in the field of ecology, appears highly promising to tackle agronomic issues and its application to agrosystems constitutes a challenging test of their genericity. It raises specific questions and stresses the need for new methodological developments, which could be part of a research agenda at the interface between the two disciplines.

Suppression of plant-parasitic nematodes and decomposition of organic matter are important soil food web functions. The availability and nature of carbon substrates in soil drive the structure, dynamics, and activities of soil food webs and may alter the way they function in the ecosystem. In this study, we assessed the effects on nematode communities of incorporating litter from three cover crops with contrasting biochemical characteristics into soil taken from a banana field. The litter type influenced nematode taxa and functional guilds during the 90-day assay. The cellulose- and hemicellulose-rich litters of Crotalaria zanzibarica and Paspalum notatum induced similar nematode successions, causing bacterivorous and fungivorous nematode populations to increase early in the study period and promoting a fungal decomposition pathway at the end of the trial. These two litters were the most effective at decreasing the soil’s nematode infective potential with regard to major banana plant-parasitic nematodes, such as Pratylenchus coffeae , Helicotylenchus multicinctus , and Rotylenchulus reniformis . C. zanzibarica showed good potential as a green manure plant, as it increased the soil mineral nitrogen content during the 45 days following incorporation of litter. By contrast, litter from Acacia auriculiformis , which had high lignin content, induced immobilization of soil nitrogen and did not significantly increase microbivorous nematodes before day 45. This treatment failed to reduce the soil’s nematode infective potential for H. multicinctus and R. reniformis populations. Our study also showed that the dynamics of nitrogen mineralisation in soil were related to the C:N ratio of litters but not to the decomposition pathway. The biochemical characteristics of litters appear to be critical drivers of nematode ecological successions and decomposition pathways following plant litter amendments to soil taken from a banana field.

Cover plants are one of the means to increase the functional biodiversity of fields and to enhance the ecological functions of the communities. However, the design of cropping systems including cover plants relies largely on expert knowledge. There is a lack of methods to select the best suited cover plants according to their role in the agrosystem. We propose to use functional traits to select cover plants suited to sustain ecological services in the banana agrosystems of the French West Indies. Our objectives were (i) to characterize a collection of cover plants on a trait basis, according to their potential to provide the services expected in a banana agrosystem and (ii) to discuss the potential use of this characterization for the design of innovative multi-species banana cropping systems. In these systems, four main services are targeted, i.e. controlling weeds, controlling plant-parasitic nematodes, improving soil fertility and particularly N availability, and avoiding competition with banana for resource acquisition. Three experiments were conducted, under field and controlled conditions, to evaluate the potential of a collection of 28 tropical cover plants to achieve the functions related to these services. For each cover plant, a functional profile was constructed from a combination of plant traits that are easy to assess experimentally. It described plants’ potential to achieve the functions expected in a banana agrosystem. Functional profiles required for cover plant usages were also identified. The comparison of the plant functional profiles and the functional profiles required for their usages enabled us to select the best suited plants for each usage. However, these functional profiles rarely corresponded, meaning that a cover plant is rarely sufficient to achieve all the functions required for a usage. Functional complementarities obtained by the mixture of different species of cover plants are thus often required. Compared to classical approaches of innovative cropping system design based on the experimental testing of spatial and temporal plant combinations, such a trait-based approach saves time by focusing on a shortlist of cover plants best suited according to their functions in the agrosystem.

We compare the effects of minimum and conventional tillage on banana growth, under irrigated and rain-fed conditions. Measurements of leaf growth, root length density and soil water depletion were made. We used a crop simulation model to understand and explain the differential behavior of banana plants in the experiments. Our approach consisted of (i) an adaptation of the STICS model, initially developed for temperate crops, to the banana crop grown under tropical conditions and (ii) use of the integrative ability of the model to identify the respective roles of irrigation and tillage. The adaptation of the model concerned the phasic development, and leaf and root growth. Some specific characteristics of the tropical environment were also introduced. The model points out that under irrigated conditions crop water requirements were always satisfied whatever the tillage depth. In that case, the difference in terms of leaf growth and yield between shallow and deep tillage treatments can be explained by the nitrogen deficit due to a poor root system. Rain-fed crops suffered from water shortage, but the increase in soil exploration, allowed by conventional tillage, compensated for the absence of irrigation. This work improves knowledge of the agro-physiology of the banana crop. It demonstrates the fundamental part played by the root system, which appears as an important limiting factor for a correct water and nutrient supply. It provides an opportunity to elaborate a global model for the banana crop. (Resume d'auteur)

Publication Inra prise en compte dans l'analyse bibliométrique des publications scientifiques mondiales sur les Fruits, les Légumes et la Pomme de terre. Période 2000-2012. http://prodinra.inra.fr/record/256699; International audience; To reduce chemical inputs while maintaining crop yield, disturbed ecosystem functions must be restored, for example by cover-crops. In these cropping systems, because of competition between species, soil resources must be carefully managed. Dedicated tools and models are needed that account for the adverse effects of the cover-crop on the cash crop in terms of resource availability. Besides classical agronomic calculations of stress indices, which are difficult to generalize and require numerous parameters, recent work in ecology has related plant traits, like the biomass accumulated, to the resource availability during the whole plant cycle. Following such a simple approach, the objectives of this study were (i) to determine the effects of water and nitrogen limitations on banana growth and development and to test whether simple integrated traits can highlight the effects of these stresses on banana growth, (ii) to parameterize a simple generic module of soil water and nitrogen availability linked to SIMBA GROW, the growth module of the banana crop model SIMBA, (iii) to assess the ability of the model to simulate banana growth and development in the environmental conditions of a banana/cover-crop system, with particular attention on the effects of changes in plot temperature on model outputs. Three experiments were conducted on whole production cycles to investigate the effects of different water regimes, different nitrogen fertilization levels, and the cover-crop Neonotonia wightii on banana growth, nutritional status, and date of flowering. Results showed low nitrogen availability affected banana growth only weakly but decreased leaf nitrogen content and delayed flowering. Low water availability delayed flowering and decreased banana growth. In both cases, the delayed flowering allowed longer banana growth, which balanced the negative effect of low availability on the growth rate. The cover-crop modified both the rooting depth of the banana, and thus the amount of resources accessible to the banana roots, and the plot microclimate, especially air temperature. The model correctly reproduced the differences of date of flowering and leaf area index at flowering for a first cycle of production between a bare-soil and a cover-crop system, provided air temperature was reduced by 2-3% due to the cover-crop. To conclude, this study showed that in fairly constant environmental conditions (temperature, radiation and rainfall) a simplified model using resource availability over the growing period and integrated plant traits satisfactorily simulates banana growth in an intercropped system.

Principale production vegetale sur 9 000 ha en Martinique, la culture bananiere d'exportation 'Cavendish' peut presenter un risque de degradation pour l'environnement (erosion, pollutions aquatiques) en raison de ses pratiques culturales, du relief accidente des zones de production et de fortes pluies. Recemment, une etude en parcelles experimentales de 100-200 m² sur un sol brun tropical volcanique a demontre que les systemes intensifs a gestion de paillis protegeaient le sol de l'erosion hydrique a l'oppose de la culture traditionnelle d'ananas billonne, entrant en rotation avec les bananeraies. Afin d'approfondir ces resultats, une campagne de simulations de pluies cycloniques (19 tests sur des micro-parcelles de 1m² arrosees par une pluie artificielle d'intensite de 100 mm.h-1 durant trois heures) a donc ete realisee in situ avec pour objectif de mieux comprendre le fonctionnement hydrique du sol, l'efficacite du paillage et de rechercher des liens hydrodynamiques remarquables entre les parcelles experimentales de 200 m² sous pluviosite naturelle et les micro-parcelles 1m² sous simulations de pluies. Les resultats des pluies simulees sont bien conformes aux mesures observees en parcelles d'erosion sous pluies naturelles, notamment a travers les hydrogrammes obtenus sur les differents traitements testes et a travers les parametres de ruissellement. Le sol brun tropical volcanique presente une forte capacite d'infiltration et meme apres trois heures de simulation de pluie cyclonique, la saturation du sol superficielle n'est jamais atteinte (Hp

Silicon (Si) is known to reduce the incidence of pathogens on many plants. Little information is available on the potential positive effects of Si on the susceptibility of banana (Musa acuminata) to pathogens. Root-rot fungi of the genus Cylindrocladium have been reported, along with endoparasitic nematodes, to be the causal agent of toppling disease and severe yield loss. The objective of this study was to determine the effects of Si supply on Cylindrocladium spathiphylli infection on banana. Plantlets inoculated by dipping the root system in a conidial suspension of the pathogen were grown on a desilicated ferralsol and amended, or not, with 2 mM of soluble Si under greenhouse conditions in Guadeloupe. The root lesion severity was evaluated using the image analysis program WinRHIZO 7, 14 and 21 days after inoculation. A reduction of about 50% of root necrosis was observed 14 days after inoculation for the Si-supplied plants compared with those not supplied with Si. The Si amendment also alleviated growth reduction caused by the pathogen. These results suggest that Si could have a positive effect on banana resistance to C. spathiphylli and provide an environmentally friendly alternative to pesticides for the integrated control of an important crop disease.

Publication Inra prise en compte dans l'analyse bibliométrique des publications scientifiques mondiales sur les Fruits, les Légumes et la Pomme de terre. Période 2000-2012. http://prodinra.inra.fr/record/256699; Plant-parasitic nematodes are injurious crop pests that have been managed mainly by chemical nematicides. However, safe and alternative methods such as those based on organic materials need to be developed. Our study has evaluated (i) the effects of four organic amendments with different biochemical compositions that are abundantly produced in the study area (Guadeloupe, French West Indies) on soil nematode communities and (ii) some of the suppression mechanisms of banana parasitic nematodes, especially those involving the soil food web. This study is based on a microcosm experiment comparing sugarcane bagasse, sugarcane sludge, plant residues and sewage sludge. All amendments except sewage sludge decreased the root abundances of plant-parasitic nematodes, by 96% in the case of sugarcane bagasse. For this treatment, soil densities of carnivorous nematodes were six times higher than the treatments without organic amendment. Plant residues and bagasse were mainly composed of materials that are difficult to decompose, namely cellulose and lignins. These organic materials favored a fungal decomposition pathway and permitted development of carnivorous nematode populations and increased the Channel Index (CI). Pratylenchus coffeae control after sugarcane refinery sludge application remains unexplained. Lastly, sewage sludge, composed mainly of easily degradable compounds, did not permit nematode control, and only bacterivorous nematode populations were enhanced by this treatment. (C) 2011 Elsevier B.V. All rights reserved.

The monoculture of banana can have a serious detrimental impact on the environment as pesticide treatments can lead to surface and groundwater pollution. Different approaches based on IPM (integrated pest management) have been developed in the French West Indies to reduce the use of the pesticides in banana cultivation. These methods have been developed using management techniques that prevent the build-up of banana pathogens and also eliminate them, non-chemical techniques, such as cultural practices and biological control, and resistant cultivars. As a result, new crop management systems have led to a 65% decrease of pesticide use over the last 10 years. Major results of the research already undertaken and future research that is being considered to reduce pesticide use in banana plantations are reviewed. (Resume d'auteur)

In the French West Indies, the productivity of export banana plantations is adversely affected by plant-parasitic nematodes (PPNs) including the endoparasitic species Radopholus similis (Pratylenchidae). In the last decades, control of PPNs was mainly based upon repeated applications of carbamate or organophosphate nematocides that are potentially toxic for human health and the environment. This paper describes a prophylaxis-based strategy, combining soil and plant sanitation that was developed in recent years to reduce dependence on chemical nematocides. Soil sanitation was implemented through a cleansing system based on glyphosate injection of banana plants before up-rooting. In addition, as a decision support tool, soil cleansing assessment biotests were developed to evaluate the effectiveness of the method before planting new banana crops. Crops were initiated using tissue culture-derived plants of "Grande Naine" (AAA, Cavendish subgroup). In association with soil sanitation, this resulted in a reduction of 60% in nematocide use. The plant sanitation system is being modified by i/ selecting R. similis non-host rotational crops, such as perennial soybean (Neonotonia wightii), siratro (Macroptilium atropurpureum), and forage grasses like Digitaria decumbens and Brachiaria humidicola; ii/ exploiting the existing variation in susceptibility to R. similis within Cavendish clones; and iii/ identifying R. similis resistant or weakly susceptible clones among improved banana hybrids developed by the CIRAD breeding programme. New initiatives to further enhance the PPN control method are also discussed. They include the promotion of techniques designed for improving management of crop residues, the identification of nematotoxic plants showing allelopathic effects against R. similis and the benefits from the preservation of soil biodiversity in banana cropping systems. Most of these approaches are being carried out in the framework of a European Commission-funded network of excellence, named ENDURE. (Resume d'auteur)

E-mail Addresses: tixier@cirad.fr, risede@cirad.fr, malezieux@cirad.fr Publication Inra prise en compte dans l'analyse bibliométrique des publications scientifiques mondiales sur les Fruits, les Légumes et la Pomme de terre. Période 2000-2012. http://prodinra.inra.fr/record/256699; International audience; This article describes the biological background, the model-building methodology and some applications of SIMBA-NEM, a new model to simulate the population dynamics of two major plant-parasitic nematode species of banana, Radopholus similis and Pratylenchus coffeae. For each species, each generation is represented by one cohort. Cohorts of the same species form a chain representing the developmental stages of nematodes. A logistic function describes population growth in relation with: (i) an environmental carrying capacity (K) that depends on available banana root biomass, (ii) an intrinsic growth rate (c) and (iii) competition between nematode species. Soil water content and the quantity of nematicides used are considered to be the main variables influencing the intrinsic population growth rate of each species. SIMBA-NEM was calibrated and validated using datasets from banana cropping systems in Guadeloupe (French West Indies). By analysing the sensitivity of the model to the main parameters and performing simulations of validation for various cropping systems (banana monoculture with or without nematicide applications use and a banana/sugarcane rotation) we were able to test the ability of the model to predict nematode population dynamics under a range of conditions. SIMBA-NEM is able to predict long-term nematode population size, while taking interspecific competition into account. It also helped to define knowledge gaps in nematology and modelling. SIMBA-NEM was used to optimise the effect of nematicide applications. SIMBA-NEM can already be a very helpful tool for designing sustainable and more environment-friendly banana cropping systems. In the SIMBA global crop-modelling environment, SIMBA-NEM is a key sub-model which provides essential information concerning the sustainability of the simulated system and thus permits planning environmentally friendly cropping systems

Banana crops represent a collection of individual plants that vegetatively propagate at their own rhythm, with stabilised but unsynchronized production of inflorescences over time. Such agrosystems cannot be simulated with existing crop models due to the unsynchronized behavior of individual plants. A new simulation model (SIMBA-POP), based on the cohort population concept, was built to predict phenological patterns of the population and harvest dynamics in banana cropping systems. The model was calibrated and validated for #Musa# spp., 'AAA' group, cv. 'Cavendish Grande Naine' with field data from the French West Indies (Guadeloupe and Martinique). It can quite accurately predict temporally-varying banana harvesting dynamics (date and number of harvested bunches). The harvesting peak is predicted with a precision less than 3 weeks for the first 3 cropping cycles. The model structure is based on two linear chains of cohorts characterized by both physiological age (heat unit accumulation) and development-stage dispersion in the banana population due to flowering, harvesting and sucker selection. This model is a valuable tool for both farmers (field management) and crop scientists (to simulate and design cropping systems). This approach is a first step towards long-term simulation in non-synchronized agrosystems that cannot be simulated with existing crop models.

Volcanic ash soils are generally recognized as soils with excellent and stable physical properties. Here we characterized the porosity and water properties of volcanic ash Andosols and Nitisols from Guadeloupe in contrasting banana systems: (1) perennial crop without mechanization, (2) mechanized and regularly replanted crop. Desiccation from 1 kPa to 1550 kPa moisture tension leads to significant shrinkage in the Andosol, representing a 50% reduction of the void space. The clayey Nitisol exhibited limited shrinkage. Soil clods from the mechanized plots had a significantly smaller macroporosity than that from perennial plots. The soil hydraulic conductivity was also drastically reduced in the compacted layers of the mechanized plots. However, Nitisols appeared to be less affected than Andosols. Laboratory compression tests showed that both soils were susceptible to compaction at soil moisture close to field capacity. The shrinkage properties of the Andosol were due to micro-aggregation of non-crystalline components upon drying. The relative stability of the macroporosity in the Nitisol was probably related to the presence of stable microaggregates made of halloysite and iron oxide. Two major processes promote soil structure degradation in the Andosol under mechanized banana cropping, surface desiccation and soil compaction. They are both induced by repeated tillage after clearing.

Wound anthracnose is a post-harvest disease which develops during storage and ripening of bananas. In the French West Indies, it mainly occurs on fruits coming from plantations situated on soils at low-altitude, during the second half of the year. It is caused by a pathogenic fungus, Colletotrichum musae . A diagnostic survey was carried out on 106 plots representative of all the soil/climatic conditions and techniques in Guadeloupe in order to assess the variability of fruit susceptibility to wound anthracnose. Secondly, the effect of mineral nutrition on this susceptibility was analysed for the soil/climatic zone where the anthracnose problems are most serious. For this purpose, 54 plots on halloysitic and ferrallitic soils were chosen by including in the selection plots from all cultural situations. This study has brought to light a wide variation in the susceptibility of bananas to Colletotrichum musae . Fruits from high-altitude plantations are the least susceptible. On low-altitude soils, where the most variability is observed, a relationship was found between the Mn content of fruit and susceptibility to anthracnose; the plants producing the most susceptible fruit had high foliar Mn concentrations and low Ca concentrations, and had grown on rather acid soils. Hypotheses for the physiological mechanisms involved in the sensitisation of the fruit are discussed.

Soils developed on voIcanic parent materials have many intrinsic qualities favourable to cropping. Howerer, fertility decreases dramatically when they are badly managed. A short review and case studies from Reunion and Guadeloupe highight the special characteristics of these soils, and their response to management. The interplay of cropping systems and physical characteristics of Andisols is first considered through the example of #Pelargonium# and food crop systems in Reunion. Progressive decrease in production and cropping potential shows in falling yields as well as in the overall decline of the system. The example of banana production in Guadeloupe highights the increase in inputs needed to realise the land's potential and to maintain yields, in particular more tillage and pest treatment. In both cases, these trends are connected to the co-evolution of soil characteristics and cropping systems. They lead to an increase of risks with less security and less scope in the choice of cropping systems. Technical solutions in the form of erosion-control measures, rotation and planting techniques have been developed and prove to be relevant and consistent in their benefit. (Resume d'auteur)

Increased plant diversity may alter soil food web structure and soil function in agroecosystems. This thesis sets the basis for an experimental approach in order to select some plant species in regard to several functional traits and to assess their effect on soil food web and soil functioning. We focused on cover-crops able to induce regulation of the banana parasitic nematodes Radopholus similis and Pratylenchus coffeae and to promote decomposition of organic matter during the fallow period in banana plantations in the French West Indies (FWI). We used functional analysis of nematode communities as a soil food web bio-indicator. Bibliography studies show us that “plant host status” with respect to plant-parasitic nematodes, “biochemical characteristics of litters” and “primary productivity of plants” are functional traits involved in plant species effects on soil food web and on the ecosystemics service we targeted. We therefore use those traits to perform a typology of cover-plants. Furthermore, to study the effects of litter inputs with contrasted biochemical characteristics we pick out the plant species Paspalum notatum, Crotalaria zanzibarica and Acacia auriculiformis which show differential hemicellulose, cellulose, lignin and nitrogen content. A microcosm assay shows that biochemical characteristics of litter determine ecological successions in nematode communities and promote, in a contrasted manner, the decomposition pathway of organic matter. We discussed the mechanisms of plant parasitic nematodes suppression involved by those litter inputs. We also confirm that the culture of those three plant species can diminish the population of R.similis in soils. We also assessed the effects of plant covers composed of P. notatum and C. zanzibarica and a mix of those two plant species on nematodes communities during a field trial. In order to do so, we cultivated those cover-plants during a nine months period and then destroyed them to restitute plant litter in soils surface. Our results showed that cover-plant alters differentially plant-parasitic nematodes and the infectious potential of soils. Cover-crop cultivation also induced bottom-up and top-down regulations in plant parasitic nematode communities. After litter restitution, we observed high abundance of nematodes involved in litter decomposition within the soil surface layer. The latter provides underlying soil layers with mineral nitrogen and improves omnivore and predator nematode abundances. It results in the formation of a structured and diversified nematode community concomitantly with R. similis and P. coffeae regulation in soils. We observed greater effects with the cover-crop composed of C. zanzibarica and the mix of C. zanzibarica and P. coffeae. This study confirms that the functional traits of “host status”, “biochemical composition of litter” and “primary production” of plant species are decisive in the selection of cover-crop for the fallow period in banana plantation.; L’introduction de biodiversité végétale dans les agroécosystèmes peut modifier la structure des réseaux trophiques du sol et les fonctions écosystémiques qu’ils assurent. Cette thèse fixe les bases d’une démarche expérimentale pour sélectionner des espèces végétales à partir de différents traits fonctionnels et pour évaluer leurs effets sur les réseaux trophiques et le fonctionnement des sols. Nous avons plus particulièrement cherché à sélectionner des plantes de services pouvant induire la régulation des nématodes parasites du bananier Radopholus similis et Pratylenchus coffeae et favoriser la décomposition des matières organiques durant l’interculture dans les bananeraies antillaises. C'est-à-dire la période entre la destruction de la bananeraie et l'implantation d'une nouvelle. Au cours de ce travail, les réseaux trophiques du sol ont été appréhendés par une analyse fonctionnelle des communautés de nématodes. Une étude bibliographique nous a permis de montrer que les traits fonctionnels « statut d’hôtes » vis-à-vis des nématodes phytoparasites, « composition biochimique des litières » et « productivité primaire des espèces végétales » contribuent aux effets des plantes sur les réseaux trophiques du sol et sur les deux services écosystémiques étudiés. Nous avons donc effectué une typologie des plantes de services à partir de ces traits. Puis pour étudier les effets d’apports de litières de composition biochimique contrastée sur les communautés de nématodes du sol, nous avons sélectionné les espèces végétales Paspalum notatum, Crotalaria zanzibarica et Acacia auriculiformis dont les litières présentaient des teneurs variables en hémicellulose, cellulose, azote, et lignine. Une étude en microcosmes a montré que la composition biochimique des litières détermine les successions écologiques au sein des communautés de nématodes et stimule de manière contrastée les voies de décomposition de la matière organique dans les sols. Nous avons discuté des mécanismes de régulation des nématodes parasites du bananier induits par ces apports de litières. Nous avons également montré que la culture de ces trois espèces induit une diminution des populations de R. similis dans les sols. Enfin, nous avons évalué les effets au champ de couverts de P. notatum, de C. zanzibarica et d’un couvert associant ces deux espèces. Pour cela, nous avons cultivé ces couverts durant neuf mois avant de les détruire et de restituer leurs litières à la surface des sols. Nos résultats montrent que ces couverts végétaux influencent différemment l’abondance en nématodes phytoparasites et le potentiel infectieux du sol. Ils induisent également des régulations bottum-up et top down dans leurs communautés. Après la restitution des litières, nous avons observé de fortes abondances de nématodes liées à la décomposition des litières dans la couche de surface du sol. Celle-ci fournit les couches sous-jacentes du sol en azote minéral et favorise le développement des nématodes omnivores et prédateurs. Il en résulte la formation de communautés de nématodes structurées et diversifiées concomitantes à la réduction des populations de R. similis et P. coffeae. Ces résultats étaient plus marqués pour les couverts monospécifiques de C. zanzibarica et les couverts mixtes. Nos travaux confirment que les traits fonctionnels « statut d’hôte », « production primaire », « composition biochimique des litières » sont déterminant pour sélectionner des plantes de services pour l’interculture en bananeraies.

Banana agro-ecosystems in the French West Indies have relied for a long time on intensive monocultures. A shift to agroecological management practices have been made since the “Plan Banane Durable” program in 2009. To achieve the aim of reducing the use of chemical inputs, the introduction of cover plants to deliver agroecosystem services under study to reconcile agronomic and environmental issues. These services include the biological control of weeds by cover plants which requires to introduce cover plant species with a good ability to compete for light towards weeds. However the introduction of cover plants should not cause a dis-service by reducing the production of bananas through a competition for mineral and water soil resources. A better knowledge of aboveground and belowground competition process by cover plants is then essential to reason their use in cropping systems. We chose an approach based on functional traits to characterize plant species from this very anthropized agroecosystem because an exhaustive study of the ecophysiological processes underlying competition is not suitable for a large panel of potentially useful cover plant species. We assume that a combination of values of functional traits related to the acquisition and use of resources reflects a strategy of competition. The objectives of this thesis were 1) to characterize a panel of potential cover plant and weed species as well as two banana cultivars according to functional traits related to the acquisition and use of aboveground and belowground resources, 2) identify competition strategies on the basis of the combinations of trait values obtained for each species and 3) comparing the competitive effects of the species as a function of their competition strategies. This study is based on a bibliographic work to identify traits of interest, a first field experiment to characterize the competition strategy of the different species by measuring functional traits on plants grown individually and in non-limiting resources conditions and a second field experiment to characterize the competitive relationships between species according to their respective competitive strategies. The analysis of the results allowed us to build two classifications of species according to their strategy, one based on competitive traits for light and the other one based on the competitive traits for soil resources. Our results allowed to distinguish four aboveground competition strategies for light ("shade", "obstruction," "smothering low”," strong smothering ") and four belowground competition strategies for soil resources (" exploration "," occupation "" exploitation, "" neutral "). The species of the panel were replaced on the main axes of differentiation between species in ecology opposing resource conservation and capture strategies. The results do not show links between aboveground and belowground strategies. These functional classifications can be used to identify the most qualified species, a priori, to be used in banana cropping systems of the French West Indies. The use of functional traits in an agronomic context enabled us to acquire a better knowledge of the species competition abilities and opens perspectives for the establishment of a decision-making tool for the choice of species to include in multi-species cover.; Les agroécosystèmes bananiers aux Antilles, longtemps menés en monocultures intensives, ont pris le virage d’une gestion agro-écologique depuis la mise en place du Plan Banane Durable en 2009. Pour atteindre l’objectif de diminution de l’utilisation des intrants chimiques, l’introduction de plantes de services (PdS) destinées à accomplir des services écosystémiques est à l’étude pour concilier au mieux les enjeux agronomiques et environnementaux. Parmi ces services, la régulation biologique des adventices par les PdS nécessite d’introduire des espèces ayant une bonne aptitude à la compétition pour la lumière vis-à-vis des adventices. Néanmoins la réalisation de ce service ne doit pas engendrer un dis-service vis à vis de la production agricole par les bananiers en provoquant une compétition pour les ressources minérales et hydriques du sol. Une bonne connaissance des processus de compétition aérienne et souterraine mis en place par les PdS est alors indispensable pour raisonner leur utilisation dans le système de culture. Nous avons choisi une approche par traits fonctionnels pour caractériser les espèces de cet agrosystème très anthropisé car l’étude exhaustive de processus écophysiologiques n’est pas adaptée pour le grand panel de plantes de services potentiellement utilisables. Nous faisons l’hypothèse qu’une combinaison de valeur de traits fonctionnels liés à l’acquisition et l’utilisation des ressources du milieu reflète une stratégie de compétition. Les objectifs de cette thèse sont de 1) caractériser un panel d’espèces de PdS potentielles et d’adventices ainsi que deux cultivars de bananiers selon des traits fonctionnels liés à l’acquisition et l’utilisation des ressources aériennes et souterraines, 2) identifier des stratégies de compétition sur la base des combinaisons de valeur de traits obtenus pour chaque espèce et 3) comparer les effets compétitifs des espèces en fonction de leur stratégie de compétition. Cette étude repose sur un travail bibliographique pour identifier les traits d’intérêts, une première expérimentation au champ pour caractériser la stratégie de compétition des différentes espèces par la mesure de traits fonctionnels sur des plantes isolées et en situation de ressources non limitantes et sur une deuxième expérimentation au champ pour caractériser les relations de compétition entre espèces en fonction de leur stratégie respective. L’analyse des résultats a permis de construire deux classifications des espèces en fonction de leur stratégie, l’une basée sur les traits de compétition pour la lumière et l’autre basée sur les traits de compétition pour les ressources du sol. Nos résultats ont permis de distinguer 4 stratégies de compétition aérienne pour la lumière (« ombrage », « obstruction », « étouffement faible », étouffement fort») et 4 stratégies de compétition souterraine pour les ressources du sol (« exploration », « occupation », « exploitation », « neutre »). Les espèces du panel ont été replacées sur les grands axes de différenciation d’espèces de l’écologie opposant des stratégies de conservation et de capture des ressources. Les résultats ne mettent pas en évidence de liens entre stratégies aériennes et souterraines. Ces classifications fonctionnelles peuvent être utilisées pour identifier les espèces les plus adaptées, a priori, au contrôle des adventices dans le système de culture bananier aux Antilles. L’utilisation de traits fonctionnels dans un contexte agronomique a permis d’acquérir une meilleure connaissance des aptitudes des espèces à la compétition et ouvre des perspectives quant à la création d’un outil d’aide au choix des espèces composant des couverts multi-espèces.

Banana agro-ecosystems in the French West Indies have relied for a long time on intensive monocultures. A shift to agroecological management practices have been made since the “Plan Banane Durable” program in 2009. To achieve the aim of reducing the use of chemical inputs, the introduction of cover plants to deliver agroecosystem services under study to reconcile agronomic and environmental issues. These services include the biological control of weeds by cover plants which requires to introduce cover plant species with a good ability to compete for light towards weeds. However the introduction of cover plants should not cause a dis-service by reducing the production of bananas through a competition for mineral and water soil resources. A better knowledge of aboveground and belowground competition process by cover plants is then essential to reason their use in cropping systems. We chose an approach based on functional traits to characterize plant species from this very anthropized agroecosystem because an exhaustive study of the ecophysiological processes underlying competition is not suitable for a large panel of potentially useful cover plant species. We assume that a combination of values of functional traits related to the acquisition and use of resources reflects a strategy of competition. The objectives of this thesis were 1) to characterize a panel of potential cover plant and weed species as well as two banana cultivars according to functional traits related to the acquisition and use of aboveground and belowground resources, 2) identify competition strategies on the basis of the combinations of trait values obtained for each species and 3) comparing the competitive effects of the species as a function of their competition strategies. This study is based on a bibliographic work to identify traits of interest, a first field experiment to characterize the competition strategy of the different species by measuring functional traits on plants grown individually and in non-limiting resources conditions and a second field experiment to characterize the competitive relationships between species according to their respective competitive strategies. The analysis of the results allowed us to build two classifications of species according to their strategy, one based on competitive traits for light and the other one based on the competitive traits for soil resources. Our results allowed to distinguish four aboveground competition strategies for light ("shade", "obstruction," "smothering low”," strong smothering ") and four belowground competition strategies for soil resources (" exploration "," occupation "" exploitation, "" neutral "). The species of the panel were replaced on the main axes of differentiation between species in ecology opposing resource conservation and capture strategies. The results do not show links between aboveground and belowground strategies. These functional classifications can be used to identify the most qualified species, a priori, to be used in banana cropping systems of the French West Indies. The use of functional traits in an agronomic context enabled us to acquire a better knowledge of the species competition abilities and opens perspectives for the establishment of a decision-making tool for the choice of species to include in multi-species cover., Les agroécosystèmes bananiers aux Antilles, longtemps menés en monocultures intensives, ont pris le virage d’une gestion agro-écologique depuis la mise en place du Plan Banane Durable en 2009. Pour atteindre l’objectif de diminution de l’utilisation des intrants chimiques, l’introduction de plantes de services (PdS) destinées à accomplir des services écosystémiques est à l’étude pour concilier au mieux les enjeux agronomiques et environnementaux. Parmi ces services, la régulation biologique des adventices par les PdS nécessite d’introduire des espèces ayant une bonne aptitude à la compétition pour la lumière vis-à-vis des adventices. Néanmoins la réalisation de ce service ne doit pas engendrer un dis-service vis à vis de la production agricole par les bananiers en provoquant une compétition pour les ressources minérales et hydriques du sol. Une bonne connaissance des processus de compétition aérienne et souterraine mis en place par les PdS est alors indispensable pour raisonner leur utilisation dans le système de culture. Nous avons choisi une approche par traits fonctionnels pour caractériser les espèces de cet agrosystème très anthropisé car l’étude exhaustive de processus écophysiologiques n’est pas adaptée pour le grand panel de plantes de services potentiellement utilisables. Nous faisons l’hypothèse qu’une combinaison de valeur de traits fonctionnels liés à l’acquisition et l’utilisation des ressources du milieu reflète une stratégie de compétition. Les objectifs de cette thèse sont de 1) caractériser un panel d’espèces de PdS potentielles et d’adventices ainsi que deux cultivars de bananiers selon des traits fonctionnels liés à l’acquisition et l’utilisation des ressources aériennes et souterraines, 2) identifier des stratégies de compétition sur la base des combinaisons de valeur de traits obtenus pour chaque espèce et 3) comparer les effets compétitifs des espèces en fonction de leur stratégie de compétition. Cette étude repose sur un travail bibliographique pour identifier les traits d’intérêts, une première expérimentation au champ pour caractériser la stratégie de compétition des différentes espèces par la mesure de traits fonctionnels sur des plantes isolées et en situation de ressources non limitantes et sur une deuxième expérimentation au champ pour caractériser les relations de compétition entre espèces en fonction de leur stratégie respective. L’analyse des résultats a permis de construire deux classifications des espèces en fonction de leur stratégie, l’une basée sur les traits de compétition pour la lumière et l’autre basée sur les traits de compétition pour les ressources du sol. Nos résultats ont permis de distinguer 4 stratégies de compétition aérienne pour la lumière (« ombrage », « obstruction », « étouffement faible », étouffement fort») et 4 stratégies de compétition souterraine pour les ressources du sol (« exploration », « occupation », « exploitation », « neutre »). Les espèces du panel ont été replacées sur les grands axes de différenciation d’espèces de l’écologie opposant des stratégies de conservation et de capture des ressources. Les résultats ne mettent pas en évidence de liens entre stratégies aériennes et souterraines. Ces classifications fonctionnelles peuvent être utilisées pour identifier les espèces les plus adaptées, a priori, au contrôle des adventices dans le système de culture bananier aux Antilles. L’utilisation de traits fonctionnels dans un contexte agronomique a permis d’acquérir une meilleure connaissance des aptitudes des espèces à la compétition et ouvre des perspectives quant à la création d’un outil d’aide au choix des espèces composant des couverts multi-espèces.