Use of stem diameter variations to detect plant water stress in tomato

The sensitivity of stem diameter variations (SDV) measured with linear variable transducer (LVDT) sensors as indicators of plant water status in tomato was evaluated. Two tomato crops were grown sequentially in a sandy loam soil in an unheated plastic greenhouse. These were an autumn-winter tomato crop (autumn crop) and a spring-summer tomato crop (spring crop). One drying cycle of 61 days was imposed to the autumn crop in winter at 92 days after transplanting (DAT). Two drying cycles, each of 29 days, were applied to the spring crop, to young (58 DAT) and mature plants (121 DAT). For each drying cycle, four replicate plots did not receive irrigation, and four were well watered. During each drying cycle, LVDT sensors continuously measured SDV, and daily measurements were made of leaf (Psi(leaf)) and stem water potential (Psi(stem)). SDV data was interpreted using the SDV-derived indices, maximum daily shrinkage (MDS) and stem growth rate (SGR). The response of SDV-derived indices to water deficit differed with (1) climatic conditions during stress imposition and (2) crop age. In the winter drying cycle of the autumn crop, the responses of the SDV-derived indices to soil drying were relatively small and slower than Psi(leaf) and Psi(stem). Under warmer conditions, the SDV-derived indices were much more responsive to soil drying. In rapidly growing young plants, where SDV was characterized by high SGR and small MDS, SGR was the most sensitive SDV-derived index. In more mature plants with little stem growth, MDS was the most sensitive SDV-derived index. In mature plants grown in warm to hot conditions, MDS (1) responded at a similar time or earlier than Psi(leaf) and Psi(stem) and (2) had larger "signal" values (ratio of values from unwatered to control plants) than Psi(leaf) and Psi(stem). However, there was appreciably more "noise" (coefficient of variation, CV) associated with the SDV-derived indices, giving lower "sensitivity" values, determined from "signal" to "noise" ratios, than for Psi (leaf) and Psi(stem). Regression analysis between MDS of well-watered plants and climatic variables gave best results for a linear relationship between MDS and daily maximum vapor pressure deficit. There were strong linear relationships between MDS and Psi(leaf) for each drying cycle. The slopes of these relationships differed with crop age indicating that there was no constant relationship between MDS and Psi(leaf) for a whole season. Overall these results demonstrated that MDS and SGR can be sensitive indicators of the water status of tomato crops under conditions of moderate to high evaporative demand. However, the variability associated with the SDV-derived indices and the changing MDS-Psi(leaf) relationship with crop age represent major issues regarding the development of irrigation scheduling protocols for tomato.