Lin-An Zhang, Xuan Dong, Kai Liu, Nicolas Gonzalez, Jose Manuel Estevez, Wen-Xue Li, Xing Wang Deng, Feng Yu, Yunbi Xu
Climate change demands accelerated plant adaptation and de novo domestication. Yet current enviromics focuses disproportionately on external environments, neglecting internal dynamics-gene expression, metabolic flux, and signal transduction-within predictive envirotyping frameworks. This gap constrains plant-environment adaptation research and crop improvement. Integrating multi-scale envirotyping with plant-environment interaction networks could catalyze a paradigm shift from empirical selection to mechanism-informed design breeding. Four challenges remain: (1) constructing adaptive multi-dimensional networks, (2) engineering transgenerational epigenetic reprogramming, (3) scaling domestication pipelines, and (4) predicting adaptive trajectories. Future efforts should converge on five domains: high-throughput microprobe envirotyping arrays, spatiotemporally resolved multi-omics, decoding epigenetic memory carriers, artificial intelligence (AI)-guided genome design, and phenotype prediction models. Ultimately, advancing from multi-omics dissection and mechanistic interpretation to targeted de novo design will enable the precise engineering of crop adaptive responses to environmental change.