Nonlinear climatic regulation of flowering phenology in sugarcane wild relatives revealed through mixed- effects and survival modelling

Flowering phenology is a key adaptive trait influencing plant reproduction and responses to climate change, yet the climatic regulation of flowering in tropical perennial grasses remains poorly understood. We investigated the influence of pre-flowering climatic conditions on flowering initiation in 251 sugarcane wild and allied genotypes evaluated over multiple years, representing 1,330 genotype–year observations. Flowering records were integrated with climatic variables during the 30-day pre-flowering period, including growing degree days (GDD30D), cumulative rainfall (RF30D), relative humidity (RH30D), and solar radiation (SRAD30D). Linear mixed-effects models identified relative humidity as the strongest predictor of flowering initiation (P < 0.001), followed by rainfall and solar radiation, whereas GDD showed no significant linear effect. Fixed climatic variables explained 68.2% of the variation, increasing to 88.6% after accounting for genotype and year effects. Interannual environmental variation was approximately 64-fold greater than genotypic variation, highlighting the dominant influence of climatic variability. Generalized additive models revealed pronounced nonlinear responses, while mixed-effects Cox models showed that flowering probability increased with GDD and rainfall but declined with increasing relative humidity and solar radiation. These findings demonstrate that flowering phenology in sugarcane wild relatives is governed by nonlinear climatic regulation, providing new insights into climate–phenology relationships and supporting climate-resilient germplasm conservation and breeding.