Conference Agenda

Water availability strongly drives the photosynthetic performance and quality of turfgrass, one of the world's most widely adopted crop systems. To enable early detection and quantification of drought stress, we exposed four cool-season species (Agrostis stolonifera, Festuca rubra commutata, Poa pratensis, and Lolium perenne) to progressive water limitation under climate-controlled conditions. Measurements of leaf water content and relative soil water content provided estimates of physiological water availability. Subsequently, we used a custom-built camera system to spatially acquire phenolic (F450, F520) and chlorophyll (F690, F740) fluorescence emissions within the canopy. Statistical analysis – including temporal stress mapping, Pearson correlation, and random forest classification – identified the earliest and most sensitive drought-stress indicators. Indices combining phenolic and chlorophyll fluorescence revealed an early onset, high correlations (r ≥ -0.9), and high feature importance across species. In contrast, ratios within strictly shorter or longer fluorescence wavelengths provided limited physiological information. A 71.44% reduction in red and far-red signals, accompanied by a 50.57% increase in phenolic fluorescence, largely explained this outcome. These results demonstrate the strong potential of multi-waveband fluorometry for non-destructive drought-stress detection in turfgrass. However, field applications still require disentangling fluorometric responses to differentiate stress types and magnitudes, supporting both greenkeeping and phenotyping.

Functional turf areas can only fulfill their function through regular mowing. However, mowing destroys plant tissue. This, in turn, has an impact on transpiration and water requirements. The aim of the study was, therefore, to investigate the effects of established and newly developed mowing techniques on the water consumption of commercial turf. The results show that the mowing technique has a significant influence on the water balance and individual vitality parameters of commercial turf areas. In a laboratory experiment, grasses cut with a reel mower showed lower average water consumption over four weeks than those cut with a rotary mower. This suggests potential water savings and improved turf quality in the event of water shortages for sports turf and functional turf. In field trials, however, the mowing technique only had a minor influence on root length and dry matter. Nevertheless, Lolium perenne 'Coletta' showed significantly longer roots with optimised reel mowing despite a lower cutting height. Furthermore, we used a laser cutter to achieve an optimised cutting technique. This resulted in precise cut edges and lower water loss, but also reduced turf vitality. It is likely that the grasses react negatively to laser cutting due to recurring thermal stress.

Cantaloupe and onion are major crops in Imperial County, California, covering nearly 6,000 ha with a 2024 farm gate value of about $116 million. Because local agriculture depends entirely on Colorado River water, long‑term drought has increased the need for improved irrigation management and rapid detection of crop stress.

This study evaluated reduced irrigation and nitrogen inputs in bulb onion (Allium cepa L.) and cantaloupe (Cucumis melo). Onion trials ran from October 2021 to May 2022, and melon trials from March to June 2024. Soil moisture sensors monitored water movement, and proximal sensors included the MultispeQ (photosynthesis, chlorophyll, leaf temperature), CCM‑300 (chlorophyll), GreenSeeker (NDVI), SC‑1 porometer (stomatal conductance), and a FLIR E8 thermal camera. Multispectral drone data were collected in the melon trial.

Onion irrigation levels were 40%, 70%, 100%, and 130% ETc with nitrogen rates of skipped N, 84, 168, and 252 kg ha⁻¹. Melon irrigation levels were 60%, 90%, and 120% ETc. Yield and °Brix responded to irrigation in both crops, and onions also responded to nitrogen. Early water stress reduced stomatal conductance, NDVI, and photosynthetic efficiency, while nitrogen deficiency lowered NDVI and chlorophyll in onions. Overall, sensors effectively detected yield‑limiting stress from reduced inputs.

Greenhouse transpiration estimation is essential for precision irrigation, as irrigation control strategies increasingly rely on accurate quantification of plant water use. However, conventional measurement techniques, including leaf-level gas exchange systems (e.g., LI-6800), provide localized measurements that are not suitable for continuous, whole-greenhouse irrigation management. To address the need for plant-level monitoring, sap flow sensing provides a plant-responsive approach by reflecting internal water transport processes. However, absolute transpiration cannot be reliably quantified from sap flow signals alone due to calibration uncertainties and temporal inconsistencies with mass-balance‑derived transpiration and environmental variability. To characterize temporal mismatches between sap flow and environmental variables, a window-based learning framework was developed. This framework segments continuous sensor data into fixed-length temporal sequences, enabling the model to learn characteristic response delays within the crop–environment system. Key environmental variables, including air temperature, relative humidity, light intensity, irrigation supply, drainage, and substrate moisture content, were continuously monitored in a controlled greenhouse. An optimized lag-aware LSTM framework was implemented using sap flow and environmental variables as inputs, with mass-balance-derived transpiration serving as the supervised target. Model performance was evaluated using R², MAE, and RMSE. The proposed lag-aware LSTM framework enables stable, time-aware transpiration prediction for adaptive greenhouse irrigation management.

Orange (Citrus sinensis L.) is a major fruit crop in Spain, where recurrent droughts and tightening water allocations increasingly challenge irrigation management in citrus orchards. Improving regulated deficit irrigation (RDI) strategies therefore requires reliable plant-based indicators capable of detecting water stress under field conditions. This study evaluates the potential of sap flow–derived indicators for monitoring tree water status in orange under RDI. The experiment was conducted in Seville (Spain) during the 2025 growing season in a commercial orchard of mature ‘Navelina’ orange trees. Three irrigation treatments were established in a randomized block design with three replications per treatment: a fully irrigated control, and two RDI treatments with moderate and severe water deficit imposed during summer. Sap flow was continuously monitored using newly commercially available sensors (Treetoscope, Israel), from which a sap flow index (SFI) was derived. Midday stem water potential (Ψstem), measured weekly, was used as a physiological reference. Results showed that sap flow responded sensitively to atmospheric demand and irrigation regime. Importantly, SFI detected water stress even under mild deficit conditions, as indicated by Ψstem values. These results highlight the potential of sap flow–derived indicators for irrigation scheduling and improved RDI management in citrus orchards.

Transpiration is a key regulator of crop water use and a central variable in irrigation management and land–atmosphere interactions. However, most evapotranspiration models still rely on simplified canopy representations, particularly the classical big-leaf assumption. Here, many resistive, capacitive processes within the soil–plant–atmosphere continuum are aggregated into empirical coefficients rather than represented as physically controlled variables. This simplification often limits the model’s ability to capture dynamic feedback between plant hydraulics and atmospheric demand.

Building on recent advances in agrohydrological sensing/modelling developed within the AgrHySMo laboratory, this study investigates the evolution of the Livingston atmometer into a bio-inspired adaptive transpiration monitoring system capable of reproducing the dynamic regulation of plant transpiration.

The proposed concept integrates two complementary control mechanisms, explored here through a proof-of-concept system. A deformable porous interface dynamically regulates vapour diffusion resistance in response to soil water status, while an active humidity modulation system controls vapour concentration within a sub-stomatal-like chamber in response to atmospheric demand. The interaction between these control loops enables the device to emulate key physiological processes governing plant transpiration.

This smart architecture opens new perspectives for transpiration monitoring and enabling next-generation sensing technologies for precision irrigation and climate-resilient water management.