High-throughput phenotyping of bioethanol potential in cereals by using UAV-based multi-spectral imagery

Bioethanol production obtained from cereal straw has aroused great interest in recent years, which is leading to the development of breeding programs to improve the quality of lignocellulosic material in terms of biomass and sugar content. This process requires the analysis of genotype-phenotype relationships, and although genotyping tools are very advanced, the phenotypic tools are not usually capable of satisfying the needs of massive evaluation of potential characters for bioethanol production in field trials. However, the unmanned aerial vehicle (UAV) platforms are demonstrating their capacity for efficient and non-destructive acquisition of crop data with application to high-throughput phenotyping. This work shows the first evaluation of UAV-based multi-spectral images for estimating bioethanol-related variables (total biomass dry weight, sugar release and theoretical ethanol) of several accessions of wheat, barley, and triticale (234 cereal plots). The full procedure involved several stages: 1) acquisition of multi-temporal UAV images with a six-band camera along different crop phenology stages (94, 104, 119, 130, 143, 161 and 175 days after sowing), 2) generation of ortho-mosaicked images of the full field experiment, 3) image analysis with an object-based (OBIA) algorithm and calculation of vegetation indices (VI), 4) statistical analysis of spectral data and bioethanol-related variables to predict a UAV-based ranking of cereal accessions in terms of theoretical ethanol. The high variability observed in the field trials was captured by the UAV-based system over time. Of the seven VI studied, the Near-infrared-based VIs were more appropriate to estimate crop biomass, meanwhile the visible-based VIs were suitable for crop sugar release. The temporal factor was very relevant to achieve better estimations. The results obtained from single dates (i.e., temporal scenario 1, TS-1) were always lower than those obtained in TS-2 (i.e., averaging the values of each VI obtained during plant anthesis), mainly for estimating sugar release, and in TS-3 (i.e., averaging the values of each VI obtained during the full crop development), mainly for estimating crop biomass and theoretical ethanol. The Normalized Difference Vegetation Index (NDVI) reported the highest correlation in this study (R2 of 0.66), which served to provide a ranking of cereal accessions in terms of theoretical ethanol.