Dataset from a five-season staggered-sowing maize experiment across contrasting rainfed environments in eastern Kenya

Le résumé fourni par la source

This dataset contains curated experimental, rainfall-exposure, phenological, and statistical-analysis outputs from a five-season maize sowing experiment conducted at Machang’a and Kiamaogo in eastern Kenya between March 2009 and August 2011. The two sites represent contrasting rainfed environments: Machang’a is comparatively moisture-limited, whereas Kiamaogo is wetter and has greater soil-water-storage capacity. The split-plot experiment evaluated three sowing strategies (dry sowing, rainfall-triggered wet sowing, and delayed sowing) with three replications and site-relevant maize varieties. The same physical plots were followed across five growing seasons: LR09, SR09, LR10, SR10, and LR11. The repository includes plot- and treatment-level rainfall-exposure metrics, sowing and phenological dates, grain-failure summaries, covariance-structure comparisons, Type III fixed-effect tests, estimated marginal means, planned sowing-treatment contrasts, pairwise variety comparisons, establishment and cob-production analyses, phenological-model outputs, and covariance-sensitivity results. The archived files support the associated agronomic study on how staggered sowing modifies rainfall capture, grain yield, stover production, harvest index, and reproductive failure across contrasting rainfed environments. All datasets are provided as UTF-8 CSV files with descriptive filenames and supporting documentation, including a README. Treatment terminology, dates, units, and interaction notation were standardised during curation, while all numerical estimates and statistical results were retained at their original precision. The repository is intended to support reproducibility, secondary analysis, agronomic modelling, meta-analysis, teaching, and the development of rainfall-informed maize sowing recommendations for mixed crop–livestock systems.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.