Résumé
Lodging is a major yield-reducing factor in rice systems, particularly under intensified cultivation using high-yielding cultivars. It is usually triggered by rainstorms during grain filling. The study of natural lodging is difficult due to unpredictable weather. Proxy traits for lodging resistance such as stem morphology and breaking moment have limited validity because lodging is complex. We present a new mobile, low-cost field methodology called Blaster. It uses a wind turbine creating a 1 m wide and 6 m deep wind path in the field with wind speed adjustable to 30, 45 and 60km h(-1). Water injected into the air stream provides for wetting of the crop. A measurement cycle takes ca. 10 min per plot. The lodging response of 20, mostly high-yielding but diverse rice genotypes was studied in 2013, 2014 and 2015 under irrigated transplanted conditions in the Philippines. Genotypic responses were reproducible between seasons and were predictive of natural lodging observed in the wet season for a subset of 8 genotypes. Cultivars PARAO and CT 5805 from Latin America consistently stood out for superior lodging resistance. Across genotypes, morphological traits such as plant height, biomass or stem diameter were not or poorly predictive of lodging induced by Blaster. The best single proxy trait, stem-base bending moment at breakage (BM), explained about 40% of Blaster-induced lodging, and the Lodging Resistance Index composed of plant height, wet biomass and BM explained about 70%. We propose that Blaster is a reliable tool to evaluate crop lodging resistance as affected by genotype, cultural practices or season.