Phenotypic dataset from early- and late-flowering pearl millet landraces

This phenotypic database is used for phenotype-genotype association analyses in early- and late-flowering pearl millet landraces in Senegal. Data are generated from three field experiments performed in the 2016 rainy season in Senegal and in 2017 in both Niger and Senegal. In Senegal, trials were conducted at the Institut Sénégalais de Recherche Agricole (ISRA) field station in Bambey (14°70′N, -16°47’W). In Niger, the trial was conducted at the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) field station in Sadoré (13°14′N, 2°17′E). The trials included three repetitions fully randomized. Eight and 10 individuals per accession for each repetition were sown in Niger and Senegal, respectively. Spacing between each hill was 0.9 m × 0.9 m in the Bambey trial, and 1 m × 0.8 m in the Sadoré trial. To avoid side effects, two rows of cultivated pearl millet were used to border the plots. The sowing dates were 2 August 2016, 21 July 2017 in Bambey and 17 July 2017 in Sadoré. The trials were conducted under rainfall conditions with supplementary sprinkler irrigation when necessary. The Eperon fungicide (3.88% metalaxyl-M + 64% mancozeb) was used at the seedling stage to prevent mildew attacks. Thinning was done to two plants per hill two weeks after sowing. All trials were fertilized using the micro-dosing technique (6 g NPK – 15–15–15/hill, corresponding to 93 kg ha^-1) applied at planting, followed by a 50 kg ha^-1 urea topdressing after thinning. A total of 9,290 plants were phenotyped for 11 traits associated with plant morphology and fitness: heading date (i.e. number of days from sowing to heading), main stem length, main stem diameter, main panicle length, main panicle diameter, main panicle weight, total seed weight and 1,000 seed weight of the main panicle, total number of tillers and total number of productive and non-productive tillers. For each repetition, the mean trait value was calculated from 6.5 individuals on average after elimination of the minimal and maximal measures.