Abstract
<jats:p>Microalgae and cyanobacteria have emerged as promising resources for sustainable agriculture; however, integrated methodologies for the rational design of crop-specific agricultural formulations remain scarce. This study proposes an integrated framework that combines biomass production, species characterization, nutrient characterization, mixture design, nutrient profile estimation, biological validation, and statistical optimization for the rational development of agricultural formulations based on microalgae and cyanobacteria. As a proof of concept, the proposed framework was applied to formulate consortia composed of C. vulgaris, Scenedesmus sp., and A. platensis using common bean (Phaseolus vulgaris L., ecotype 'Sangre Toro') and rainfed rice (Oryza sativa L., cv. 'Fedearroz 2020') as model crops. The experimentally determined nutrient composition of the individual biomasses was integrated into a simplex-lattice mixture design coupled with response surface methodology and desirability analysis to identify crop-specific optimal formulations and estimate their nutrient profiles. The cubic model provided the best fit (P < 0.05), showing high predictive performance and a non-significant lack of fit. The optimal bean formulation consisted of 31.6% C. vulgaris and 68.4% Scenedesmus sp., whereas the optimal rice formulation comprised 62.3% A. platensis and 37.7% C. vulgaris, demonstrating distinct crop-specific responses. The optimized bean formulation exhibited higher estimated concentrations of calcium, phosphorus, iron, and zinc, whereas the rice formulation showed higher estimated potassium and Kjeldahl nitrogen contents. These findings demonstrate the feasibility of developing crop-specific microalgal formulations and highlight that different crops may require distinct formulations rather than a universal approach. The proposed approach offers a reproducible, integrated framework for the development and optimization of next-generation agricultural formulations for sustainable crop production.</jats:p>