Evaluation de la productivité du blé dur (Triticum durum) en relation avec les pertes à la récolte induites par certains paramètres incidents en zone semi-aride de Sétif Djoudi, Mohamed Badr Islam

Abstract

Improving the productivity of durum wheat (Triticum durum Desf.) to semi-arid climatic conditions is a crucial challenge for food security and the sustainability of cereal-based systems in Algeria. This study was conducted over two cropping seasons (2020-21 and 2021-22) at three agroecological sites in the Sétif province (ITGC, Ain Oulmene, and Beni Fouda), evaluating three different genotypes: Boussellam (BOS), Oued El Bared (OB), and GTA-dur (GTA), using a completely randomized block design. The main objective was to identify the determinants of productivity in semi-arid zones through a combined analysis of agro-morpho-physiological performance, grain-filling dynamics, and harvest-related losses. Analysis of variance revealed significant effects of genotype, site, and year on several parameters, including chlorophyll content, canopy temperature, membrane stability, and earliness, highlighting genotype × environment interaction. Boussellam exhibited good overall performance, with no significant differences in grain yield. Multivariate analyses emphasized the importance of the number of spikes per m², grains per spike, biomass, and plant height, with a key role played by thousand-kernel weight. The study of grain-filling dynamics showed a sigmoid dry matter accumulation curve, with parameters influenced by both genotype and environment. OB stood out for its high grain-filling rate (TRG) and high grain weight, whereas GTA exhibited a longer duration but lower final grain weight. Grain-filling rate (TRG) and the absolute filling intensity (IAR) were the best predictors of maximum grain weight, while duration (DRG) and maximum filling rate (VMR) had indirect effects. The evaluation of technological quality and harvest losses revealed significant differences between genotypes in protein content (15.38-17.35%) and test weight (75-78.84 kg.hL-1). A strong correlation was established between post-maturity moisture decline and yield loss, effectively modeled (62% of variance explained), with an average estimated loss of 1.0151 q.h-1 for every 1% decrease in moisture content. Additionally, significant instability in the harvester's performance was observed in 2020-21, reflected by high variance and a regression coefficient (bi) > 1, indicating considerable mechanical losses.