Integrated on-farm crop diagnosis of upland rice yields in northern Thailand

Van Keer Koen, Trébuil Guy. 2003. Integrated on-farm crop diagnosis of upland rice yields in northern Thailand. In : Conférence académique annuelle de l'Institut de recherche sur le riz de Thaïlande, Chonburi, 7-9 mars 2003. s.l. : s.n., 15 p.

Communication sans actes

Prévisualisation

Version publiée - Anglais
Utilisation soumise à autorisation de l'auteur ou du Cirad.
document_513721.pdf
Télécharger (310kB) | Prévisualisation

Résumé : In northern Thailand, swidden cultivation of upland rice (Oryza sativa L.) is still an important component of remote highland farming sytems. Under no-external-input cropping systems, yields in farmers' fields are generally low (1-1.5 t ha-1), but quite variable. Yields of 3 and 4 t ha-1 for early- and late-maturing varieties, respectively, are observed in pockets in farmers' fields. But such yields are seldom achieved because of multiple and interacting permanent or transient limiting factors. Identifying and ranking these limiting factors is a prerequisite to setting research and extension priorities for improving current fanners' practices. An onfarm diagnostic survey was therefore carried out in a remote highland village of Fang District, Chiang Mai Province (600-800 mas1), to identify, date, and rank the environmental and cropping system variables causing major yield limitations in upland rice under actual farmers' circumstances and management practices. Data on the crop population status for two early- and late-maturing types of local cultivars, crop environmental conditions, and cropping practices were obtained by monitoring 432 squares in 63 Lahu fanners' fields on deep granitic Acrisol soils with a clay-loamy texture during four cropping seasons. An empirical model of yield buildup and phase realization indices were used to date and rank the periods of yield differentiation and principal component analyses with instrumental variables were carried out to identify the factors affecting yields and their components. For the reproductive phase, the analysis of rice yield buildup processes pinpointed panicle formation and spikelet differentiation as key periods of yield differentiation. During the vegetative period, poor crop biomass accumulation was more important than low plant density. Vegetative biomass accumulation per plant also influenced panicle and spikelet formation. The analysis of pooled data revealed strong negative relationships for (a) no. of panicles plant-1 and no. of spikelets panicle-1 vs rice root aphid infestation; (b) plant density vs late weed competition and late-maturing cultivars (sown at lower densities); and (c) percentage of filled spikelets and 1000-grain weight vs 1995 wet season, which was characterized by dry spells during the vegetative and the reproductive phases. Weak negative relationships were found for no. of panicles plant' vs slope angle, erosion, no. of successive upland rice crops, minimum tillage, and early weed stress. A weak positive correlation between the older secondary forest type of fallow and upland rice grain yield was also observed. Brown spot was the only disease having some limited effect on crop productivity, especially for late-maturing varieties. Rice root aphid infestation was identified as the single major limiting factor with a strong and consistent effect on final grain yields. Weed competition was found to be the second most important yieldlimiting factor, followed by soil erosion in sloping fields after shallow tillage. These results show that attention should be paid to the management of soil-borne pests when setting priorities for improving upland rice-based cropping systems in northern Thailand highlands.

Mots-clés Agrovoc : riz pluvial, Oryza sativa, pratique culturale, composante de rendement, facteur du milieu, facteur de rendement, rendement des cultures

Mots-clés géographiques Agrovoc : Thaïlande

Classification Agris : F01 - Culture des plantes

Auteurs et affiliations