Modeling strawberry dry matter partitioning, and yield estimation in greenhouse hydroponic production

Abstract

Precise modeling of dry matter (DM) accumulation captures the biomass production and allocation patterns influencing total strawberry yield in greenhouse hydroponic systems. Current models often lack robust integration of canopy development, DM partitioning, and yield formation over multiple production years and cultivars. In this study, a deterministic, modeling framework based on Euler’s method was developed to simulate temporal patterns of aboveground DM accumulation, partitioning, and yield in day-neutral strawberries under greenhouse conditions. The model was parameterized and validated using ‘Albion’ and ‘San Andreas’ cultivars grown in three consecutive production cycles (2022–2025) in Auburn, Alabama, USA. The model captured the relationships between light interception and leaf area index (LAI), across all experimental years (R² = 0.70–0.89, RMSE =0.06–0.18). Simulated vegetative and total DM followed a sigmoidal pattern consistent with observations, and model performance across plant organs (leaves, crowns, flowers, fruits,) showed high agreement (R² = 0.88–0.99). Over three years, marketable yield was comprised of 60–75% of total yield. In Years 1–3, 'Albion' had a marketable yield of 63, 60, and 75%, while 'San Andreas' had 61, 72, and 74%, respectively. Although economic value is determined by marketable yield, the model simulates the total yield potential where DM allocation reflects growth processes rather than fruit quality determinants of marketability. Higher yields observed in the third production year were associated with increased allocation of DM to reproductive organs. Fruit and flower DM were positively correlated with yield, whereas vegetative and total DM showed a negative association, indicating that reproductive partitioning plays a key role in yield variability. Sensitivity analysis identified light-use efficiency (LUE) as the primary driver of simulated yield across cultivars and years. The extinction coefficient k had a secondary influence under conditions of incomplete light interception. Overall, the model provides an integrated framework linking canopy development, DM accumulation and distribution, and total yield in controlled environments. Further validation across diverse conditions and genotypes and incorporating crop development partitioning will enhance its broader applicability.