Estimating wheat partitioning coefficient using remote sensing and its coupling with a crop growth model

Context

Accurately estimating the partitioning of daily photosynthetic assimilates among different plant organs is crucial for understanding crop growth and yield formation. However, challenges in field measurements, especially in assessing belowground biomass, hinder precise evaluation of the partitioning process.

Objective

This study developed a novel approach to estimate time series of partitioning coefficient (PC) using unmanned aerial vehicle (UAV) images.

Methods

Firstly, UAV-based remote sensing data was utilized to estimate leaf biomass growth (G_leaf), aboveground biomass growth (G_above), leaf area index (LAI), and leaf chlorophyll content (LCC). Next, total wheat growth (G_total) was estimated by integrating LAI and LCC into a photosynthesis model. Finally, the leaf partitioning coefficient (LPC) and aboveground partitioning coefficient (APC) were calculated by combining G_leaf, G_above, and G_total.

Results

The proposed method effectively captured the variability of partitioning coefficients (PCs) across different phenological stages and treatments, with a relative root mean square error (RRMSE) of 24 % between the estimated and measured average LPC (ALPC). The theoretical RRMSE for the estimated average APC (AAPC) derived from a synthetic dataset was 29 %. By incorporating the estimated PCs into a crop model, the simulation accuracy for aboveground biomass (AGB) and leaf dry matter weight (LDW) improved, achieving RRMSEs of 12 % and 11 %, respectively, while simulations based on default PCs in the APSIM model resulted in overestimation.

Conclusions

This study achieved a high-throughput estimation for the wheat partitioning coefficient.

Implications

The proposed approach holds promise for advancing our understanding of photo-assimilate partitioning.