Thierry Doré 1 Maniével Sène François Pellissier Christiane Gallet 1 INA P-G, UMR d'agronomie lnra-lNA P-G, BP 01 , F-78850 Thiverval-Grignon <dore@inapg.inra .fr> 2 Agence nationale de conseil agri cole et rural, Route des Pères Maristes, Hann BP 10307, Dakar Liberté Sénégal 3 Laboratoire dynamique des écosystèmes d'altitude, CISM Université de Savoie, 73376 Le Bourget-du-Lac cedex 4 Laboratoire dynamique des écosystèmes d'altitude, CISM Université de Savoie, F-73376 Le Bourget-du-Lac cedex

L’allelopathie est l’objet d’un nombre croissant de recherches. Des progres importants ont ete realises dans la comprehension des mecanismes a l’origine des phenomenes observes, et dans « l’etablissement de la preuve » que ces phenomenes n’etaient ni dus a un artefact experimental, ni confondus avec d’autres phenomenes. Cependant, peu de tentatives ont ete menees dans l’optique d’une maitrise agronomique de ces phenomenes. Une meilleure connaissance des relations entre pratiques agricoles et allelopathie serait necessaire afin de valoriser cette derniere dans des strategies de protection integree des cultures, et/ou de mieux maitriser les « effets precedent » des cultures. Un recensement des connaissances actuellement disponibles est effectue, insistant notamment sur les facteurs responsables des variations de potentiel phytotoxique, et sur leur controle en vue de l’application agronomique de l’allelopathie. Plusieurs orientations pour une approche agronomique de l’allelopathie sont proposees, en particulier : replacer les mecanismes de l’allelopathie dans le contexte du fonctionnement du champ cultive, identifier les mecanismes-cles qui varient en fonction des pratiques, et etudier les effets d’une gamme de pratiques agricoles sur ces mecanismes ; verifier l’occurrence du phenomene en parcelles agricoles en s’appuyant sur des etudes analytiques ; apprehender simultanement l’allelopathie et les autres dimensions agronomiques majeures de l’activite agricole, en particulier l’elaboration du rendement des cultures.

CE145–66 is an improved early-maturing grain sorghum genotype, increasingly grown by farmers in the Sahelian part of Senegal. This genotype is known to have negative effects on the following groundnut crop, because of the release of allelopathic phenolic compounds into the soil. We have assessed the synthesis of phenolics in sorghum vegetative parts and the variations in synthesis between years and sites. Total phenols and phenolic acids in the aerial parts and roots of flowering sorghum plants from 52 farmers' fields at two sites (Sagnanème and Médina) in Senegal in 1996 and 1997 were measured. Thirty-eight soil samples, collected after the sorghum harvest, from sorghum rows and interrows also were analyzed for their phenolic content. Total phenols reached 1.1–1.5% of root dry weight and 1.1–2.2% of aerial parts dry weight, with little variation between sites, and large variability between years, presumably due to climatic conditions. Eight phenolic acids and three associated aldehydes were identified by HPLC, with p-hydroxybenzoic, p-coumaric, and ferulic acids the most abundant. Their totals reached 2.9–3.2 mg/g in 1996 and 2.6–2.8 mg/g in 1997 for the aerial part; and 3.3–3.6 mg/g in 1996 and 2.8–3.3 mg/g in 1997 for roots. In soils under sorghum rows, the mean water-soluble total phenols increased from 4.6 in 1997 to 6.7 μg/g in 1998 in Sagnanème, and from 3.8 in 1997 to 5 μg/g in 1998 in Médina. The concentrations of total phenols and phenolic acids were higher in rows than in interrows. All the phenolic monomers identified in vegetative parts were recovered in associated soil samples, with vanillic and p-hydroxybenzoic acids the most abundant. Finally, variability in plant phenolic content seemed more due to climatic than to cropping or soil factors, as differences between years appear more important than differences between or within sites.

Sorghum bicolor is an allelopathic crop that reduces the yield of succeeding crops. We have assessed its effect on the germination, emergence, and seedling growth of Arachis hypogea sown in soil that had had a prior sorghum cropping. A. hypogea was sown on rows and interrows of a previous sorghum crop in 1997 and 1998 in Senegal. Seedling establishment (germination rate and seedling weight) was better between rows than on rows of the previous crop. The highest concentrations of phenolic compounds occurred in the rows in 1998, while contents of row and interrow soils were similar in 1997. Vanillic acid was the main component of the six chemicals found in 1997 soils, whereas the 1998 soil samples contained mainly p-hydroxybenzoic acid, p-hydroxybenzaldehyde, vanillic, and p-coumaric acids (10 phenolics identified). The germination of peanut seeds in water (control), soil water extracts, and mixtures of pure phenolics (equivalent to those in 1997 and 1998 soil samples) was tested. All extracts inhibited germination compared to controls, but there was no significant difference among treatments, i.e., the inhibition was the same for seeds in soil solutions and those in the respective phenolic mixtures. Similarly, there were no significant differences among the germination rates in soil water extracts of rows and interrows or in the pure phenolic mixtures of rows and interrows. We propose a geometrical sowing pattern for peanuts between the rows of the previous sorghum crop to escape the latter's "allelopathic heritage."