Ratio vegetation indices have the potential to predict extractable protein yields in green protein paludiculture

Paludiculture can be a tool to incentivise rewetting of agricultural peatlands with the option for biomass utilisation in green protein biorefineries. However, the economic feasibility for green protein paludiculture depends on product maximisation. This study explored the potential of a ratio vegetation index (RVI) model, with inclusion of climatic factors relevant for biomass growth, to predict crude protein (CP) contents in green protein precipitates from biorefining Phalaris arundinacea and Festuca arundinacea cultivated under different management intensities on a wet fen. Assessing yields for two years of cultivation, we found that timing of harvest was a key variable for CP extractability using the biorefinery technique. Biomass and protein yields were similar between management treatments and years, but extractability was enhanced in the dryer of the two years. This study highlighted the potential of an RVI model to predict, under varying climatic conditions, CP contents in the protein precipitate with good model performance (R2 = 0.64, NRMSE = 0.23) and accuracy. In 92 % of occurrences, the model was able to predict statistically similar CP contents compared to measured CP in the protein product, with an average deviation between measured and predicted annual values of 1.7 % across species and management intensities. The findings highlight an option for maximising the overall efficiency of green protein paludiculture by determining the optimal timing of harvest, thereby demonstrating an economic potential to incentivise paludiculture farming.