Plant transpiration and soil hydraulic conditions

Author(s)

Derblai Casaroli

• Document type: Doctoral Thesis
• Press: Piracicaba.
• Institution: Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC)
• Defense date: 2008-05-19

Examining board members

Quirijn de Jong van Lier; Osny Oliveira Santos Bacchi; Durval Dourado Neto; Klaas Metselaar; Joao Carlos Cury Saad

• Advisor: Quirijn de Jong van Lier; Durval Dourado Neto

Abstract

Transpiration lower than potential rates reduce plant productivity and occur in soils with hydraulic conditions that are inadequate to maintain the necessary water flux towards the plant roots. Soil physical models that describe root water uptake are an important tool to estimate this transpiration reduction. The research that resulted in this thesis aimed: to estimate under which hydraulic conditions occurs the onset of transpiration reduction in terms of matric flux potential (Mcrit), water content ( ?crit), pressure head (hcrit) and hydraulic conductivity (Kcrit) for different soil and atmospheric conditions; to demonstrate experimentally that Mcrit is independent of soil properties and only influenced by root characteristics and atmospheric demand, on the contrary of ?crit, hcrit e Kcrit; to compare values of Mcrit obtained in the experiments with Mcrit estimated by a root water uptake model at the same root length density; to estimate the onset of the falling rate phase from Mcrit values; to propose an alternative to estimate ?crit, based on the relationship among M, ? e h. The present study contained two greenhouse experiments with bean plants (Phaseolus vulgaris L.) in pots. A completely randomized design was used with three repetitions in both experiments. The treatments, in the first experiment, formed a factorial 2x2 combination (soils: loamy and sandy; and water regimes: irrigated and water deficit) for plant characteristics and a one-factorial was used for critical hydraulic conditions of the soil (soils: loamy and sandy). In the second experiment a 2x2x2 combination factorial was applied (soils: loamy and sandy; water regimes: irrigated and water deficit; and atmospheric demands: without and with forced ventilation) for plant characteristics and a 2x2 factorial (soils: loamy and sandy; and atmospheric demands: without and with forced ventilation) for the critical hydraulic conditions. The critical point (?crit, hcrit, Kcrit and Mcrit) was identified on the day previous to a relative transpiration TR <1. The TR values as a function of ?, h, K and M observed in the experiments (after occurrence of critical conditions) were compared with values estimated by a model, resulting in a good fit. The experimental values (Mcrit) and the model-estimated values (Mcrit*) were different, with Mcrit > Mcrit*. A model was proposed to estimate the critical hydraulic conditions by heterogeneous distribution of root densities in soil volume fractions. The experimental and model-estimated Mcrit were compared through the RMSE (root medium square error), MAE (medium absolute error) and \"d\" (agreement index of Willmott) indexes. The results showed that Mcrit can be used to identify the onset of the falling rate phase referring to transpiration and liquid photosynthesis for the bean crop. The Mcrit values can be transformed in ?crit and hcrit, which are easily measured at field. The heterogeneous distribution of the root system in the soil caused different values of Mcrit observed experimentally and Mcrit obtained by the model. A good estimation of TR was obtained through M values by using a heterogeneous distribution of root densities in the model. The best fit was obtained when a large soil volume fraction (? 90%) was modeled with a small root density and the remaining 10% with a high root density. (AU)