Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → 2009 → Improving the nitrogen use efficiency and crop quality in the Khorezm region, Uzbekistan

Rheinischen Friedrich-Wilhelms-Universität Bonn (2009)

Improving the nitrogen use efficiency and crop quality in the Khorezm region, Uzbekistan

Kienzler Kirsten Maren

Titre : Improving the nitrogen use efficiency and crop quality in the Khorezm region, Uzbekistan

Auteur : Kienzler Kirsten Maren

Université de soutenance : Rheinischen Friedrich-Wilhelms-Universität Bonn

Grade : Doktor der Agrarwissenschaften (Dr. agr.) 2009

Résumé
In the irrigated agriculture of Central Asia, low nitrogen (N) fertilizer use efficiency in cotton (Gossypium hirsutum L.) and winter wheat (Triticum aestivum L.) decreases yields and farm income. Current N-fertilizer use is based on recommendations from Soviet times when fertilizer supply was subsidized to maximize production at all costs. Modern N management needs to enable farmers to obtain stable crop yields of good quality and preserve the environment. The present study, based on field experiments conducted 2004-2006 in the Khorezm region, Uzbekistan, intended to (i) establish cotton and wheat yield and quality responses to N fertilization ; (ii) evaluate N-fertilizer use efficiency of officially recommended N use and farmers’ practice ; (iii) simulate soil N dynamics and yields under varying N rates, irrigation water quantities and groundwater levels with CropSyst ; and (iv) determine the financial feasibility of different N practices. The study included labeled N fertilizer (15N) experiments in 2005 to quantify the fate of the applied N fertilizer. Although N was the most limiting nutrient, the N response curve of cotton and wheat yield to increasing N rates was rather flat with a yield maximum at 120 and 180 kg N ha-1, respectively. This can be attributed to unaccounted N supplements from ground- and irrigation water of around 31 kg ha-1. The official N recommendations of 200 and 180 kg N ha-1, for cotton and wheat respectively, corresponded well with both the measured and simulated N uptake at yield maximum. However, at this rate, the opening of cotton bolls was delayed beyond the period during which the ginneries offer the highest prices for cotton. Total N-use efficiency was very high for both crops (81-84 %). The large share of soil-15N (48 and 47 %, respectively) indicates that immobilization processes and/or pool substitution strongly influenced recovery rates. Farmers’ N fertilization practice gave highest cotton yields, but around 22 % lower total 15N recovery rates (64 %). For wheat, an additional late N application at the heading stage yielded highest total 15N recovery rates (52 and 53 % in plant biomass and the soil, respectively). N fertilization with diammonium phosphate before seeding showed the highest N-use efficiency for wheat and cotton as compared to urea fertilizer. Cotton fiber quality was of lowest grade (i.e. 31 mm length, 25 g tex-1 strength, and 4.08 micronaire) and remained unaffected by N treatments, timing of applications or N-fertilizer types. Fertilized with the recommended N amount, protein and gluten content of wheat kernels (12.3 and 23.0 %, respectively) met the criteria of only satisfactory to good wheat filler and low to medium flour thickener. Increasing N rates enhanced kernel protein (15 % at 300 kg N ha-1), but not gluten content (25.0 %). Protein content and yield were negatively related, showing the need for breeding or introducing wheat varieties with narrower quality and yield potential suitable for irrigated conditions in Uzbekistan. The cotton-generic routine developed for the CropSyst model predicted the experimental yields with a high accuracy (RSME 1.08 Mg kg-1). Simulations show that gaseous N losses can be reduced by lowering the groundwater level. Increasing cotton yields without increasing N losses seems possible when matching water demand and supply more closely. For cotton, returns to N investments were highest (1,069,332 UZS ha-1 net benefit) for the farmers’ N practice and for N rates below 120 kg ha-1, which encouraged fast maturation of cotton bolls at pick 1 and 2. The economic optimum thus diverged from the plant-N demand and recommendations of 200 kg ha-1. The economically most promising wheat treatments were those fertilized with the recommended N rate of 180 kg ha-1 and those receiving additional N just before anthesis (340,669 UZS ha-1 net benefit). However, the present reimbursement system at the mills lacks attractive quality-based incentives to encourage high quality production. Overall, the N management and N-use efficiency in irrigated cotton and wheat production can be improved by changing the payment system of the ginneries and mills to encourage sustainable N practices and increase crop quality. Wheat quality can be further enhanced through late N application, or by (breeding for) better varieties. CropSyst could demonstrate the impact of different agricultural practices on cotton yields and soil parameters and thus can help identifying changes in the current management system.

Présentation

Version intégrale (8 Mb)

Page publiée le 31 décembre 2015, mise à jour le 31 décembre 2018