Une typologie basée sur les pratiques maraîchères a été réalisée à partir d’enquêtes auprès de 300 producteurs de dix périmètres maraîchers en milieu urbain, semi-urbain et rural de la région de Bobo-Dioulasso au Burkina Faso, afin d’évaluer le potentiel d’évolution des exploitations vers des systèmes de culture plus écologiques. Aucun agriculteur ne pratique actuellement une production exclusivement biologique ou agroécologique. La typologie a permis de définir quatre types d’exploitations sur base de leur localisation, de leur superficie, de l’intensité d’utilisation des pesticides et des engrais minéraux et organiques. Ces exploitations présentent des atouts en commun, comme la pratique généralisée de la rotation et de l’association des cultures et de la fertilisation organique, mais aussi des défis spécifiques à relever en termes d’usage des pesticides de synthèse et de fertilisation raisonnée pour une transition agroécologique. Les producteurs ont généralement une connaissance faible et imprécise des effets de leurs pratiques sur l’environnement, tout particulièrement en milieu urbain où la sensibilisation aux impacts sanitaires et environnementaux des pratiques maraîchères doit être renforcée.

Participatory development of sustainable soil fertility (SF) management practices requires building bridges between science and indigenous knowledge. Whereas scientific SF indicators are reasonably standardized, indigenous SF indicators are site‐specific and system‐specific. Furthermore, the process of indigenous SF evaluation is very time consuming. This study therefore aimed at investigating the compatibility between local and analytical SF indicators, and at optimizing the process of indigenous SF assessment. Thirteen cotton plots were evaluated by famers on a scale from 0 (fertile) to 5 (degraded) and sampled for physico‐chemical characterization. Farmers identified 13 SF indicators related to soil, crop status, biological activity and management. Of these, 10 were retained to calculate a local SF index (LFI). Clay and silt content, pHwater and P‐Bray explained 88% of the total variance in LFI, indicating good yet incomplete compatibility between the indigenous and scientific approaches. Because of redundancy among local indicators, it was shown that the LFI could be estimated using one pedo‐indicator, one bio‐indicator and one phyto‐indicator without significant loss of accuracy. However, a minimum local indicator dataset composed of six indicators was recommended to allow for data cross‐checking. There was a trade‐off between the number of indicators and the number of farmers needed to estimate the LFI. There was a reasonable agreement between local and analytical fertility indicators, even though farmers' perception appeared rather more based on the effect and consequences of SF than on its actual status. Although compatible, analytical indicators should be translated into variables similar to local indicators in order to facilitate communication. Copyright © 2015 John Wiley & Sons, Ltd.

No abstract

Urban and peri-urban agriculture plays a key role by providing many goods and services. In particular, it provides diversified food and employment for vulnerable groups (youth and women). However, it often involves negative externalities due to non-conventional soils fertility management practices. This study aimed to investigate the chemical quality of soils over six (06) sites of the market gardening area of Bobo-Dioulasso (Burkina Faso) as affected by fertilizers uses and sites location. Thirty (30) representative market gardening farms, located in urban, semi-urban and rural areas, were randomly selected from a baseline survey database. Within each farm, composite soil samples made up of 3 individual cores were taken over the 0 -15 cm soil depth for determining soils carbon, total nitrogen, available phosphorus contents and pH-water. These data were normalized and summarized to compute a synthetic Soil Fertility Index (SFI). The data processing was focused on a Principal Component Analysis and an Ascendant Hierarchical Classification in order to make a typology of the vegetable farms. Fertilizers management effects on soils quality were compared through Variance Analysis (ANOVA) following a GLM procedure in Rstudio software. As main results, soils chemical parameters, except for available K, were affected by the location of the sites. Soils in urban farms are less acidic (on average pH = 6.9), while semi-urban and rural sites (Samadeni, Nakaguana) have more acidic soils. However, the latter site had the highest values of C and N. Moreover, the long-term application of organic matter sources results in improving of the chemical quality of the market garden soil.