Abstract
The use and development of science-based tools to increase nutrient efficiency without loss of marketable yield is central to the competitiveness of leafy vegetable production in hydroponic systems. At present, guidance on how to pair low-dose biostimulants with defined nutrient reductions, especially for lettuce grown in nutrient film hydroponic systems, is limited. This dissertation addressed that need by (a) establishing an evidence base for biostimulant use in leafy vegetable hydroponics, (b) deriving a phenolic-rich extract from organic pomegranate peel through a data-guided optimisation, and (c) testing that biostimulant extract at full nutrients versus a measured nutrient reduction in lettuce.
Chapter two synthesised the literature using a bibliometric and meta-analysis focused on leafy vegetables in hydroponics with biostimulant applications. Publishing activity clustered around a small set of biostimulant categories and delivery mechanisms, with lettuce and seaweed as model crops and biostimulants, respectively. Pooled effects on fresh weight and total yield were small and highly variable, and context dependent, depending on crop, application method, biostimulant category, and hydroponic system. One constant response emerged: the biostimulants performed best under stress and suboptimal conditions. These findings frame later trials by indicating optimal performance under nutrient reduction and clear value in testing maintenance of yield under reduced nutrients.
Chapter three developed the organic pomegranate peel biostimulant extract. A fractional factorial screening was followed by a Box-Behnken design response surface methodology, which was compared to an artificial neural network (ANN) trained on a conditional generative adversarial network (cGAN) on the original BBD. The ANN outperformed the RSM across all metrics and across all targets (total phenolic content (TPC), extract yield, and a composite (0.8 TPC + 0.2 yield). The operating point selected from the composite (72.8% solvent concentration, 1:55 solid-to-solvent ratio, 46 °C) was validated and used for scale-up.
Chapter four applied the biostimulant in hydroponic lettuce (NFT) at two nutrient regimes, full and at 85%. Under full nutrients, no biostimulant dose (0.1–0.3%) increased fresh weight, but higher doses increased the leaf number, albeit at the expense of reduced greenness. Under the 15% reduction of nutrients, the lowest biostimulant dose B1N85 maintained yield and colour.
Overall, the thesis provides a tested route, from source screening to modelling to application, and an experiment to find where low doses can support nutrient reduction and operational costs.