H13E-0967
Effect of plant-uptake representation on the water-optimal root depth
The depth of roots depends on a variety of conditions, including soil properties, plant type, nutrient availability, and climate. A focus on water enables the determination of a water-optimal root depth by equating the marginal carbon cost of deeper roots with the benefit of those roots to continued transpiration and carbon assimilation. Calculation of the transpiration benefit requires the mathematical representation of plant uptake as a function of root depth and soil moisture. This work compares the effect of two bounding representations of plant uptake on the water-optimal root depth and the response of that depth to changes in precipitation. Soil-moisture dynamics are driven by precipitation events that arrive as a Poisson process and are characterized by a mean frequency and depth. Infiltration and drainage are instantaneous, filling the root zone up to a maximum field capacity. Plant uptake is represented in one case as a step function of soil moisture; transpiration proceeds at a potential rate until the wilting-point saturation is reached when uptake drops to zero. Until that critical threshold, soil moisture has no effect on transpiration. In the second case, transpiration decreases linearly from its potential at field capacity to zero at the wilting point; soil moisture exerts a continuous and gradual influence on plant uptake throughout the drying cycle. With both the linear and step-function representations, the water-optimal root depth is more sensitive to changes in precipitation depth than frequency under dry conditions and more sensitive to precipitation frequency when the climate is wet. Under wet conditions, optimal root depths predicted with the step function show a greater sensitivity to climate than do those based on the linear model. Under dry conditions, the reverse is true; the water-optimal root depth is slightly more sensitive to changes in precipitation when the linear model is employed than when the step function is used. For all climate conditions, the water-optimal root depth is deeper and average transpiration is lower when plant uptake is represented as a linear function of soil moisture.
H13E-0968
Application of Normal Distribution Model to Estimate Root Water Uptake Profile by an Isotopic Approach
To confirm usefulness of a diagnostic model for estimating root water uptake profile by an isotopic approach, isotopic measurements of plant xylem water, soil water and groundwater were conducted at seven Japanese red pine forest sites and then the model was applied to the measured results. The model assumes that depth profile of relative uptake rate can be approximated by the normal distribution function, and xylem water isotopic composition is computed from interpolated depth profile of isotopic composition of subsurface waters. The peak depth and distribution range of water uptake zone for a given species at a given site are inversely determined by direct search method (assuming depth interval of 5 cm up to 2 m) so as to minimize root mean square error throughout observation period. Estimated water uptake profiles showed that in six sites the uptake zone of Japanese red pine (Pinus densiflora) ranged from 5 to 60 cm depth, while it was changed to deeper depths in the other site where Quercus myrsinaefolia and Pleioblastus chino coexist. On the other hand, Populus sieboldi and Malus sieboldii take up water from depths deeper than those for Pinus densiflora within a community, although the two species are usually considered as shallow rooted plants. These results indicate water source partitioning under inter-species competition, and we conclude that the present model is capable of making clear the plant water use strategy. Estimated water uptake zone also provides useful information for improving/calibrating prognostic, physical models of root water uptake.
H13E-0969
Root Growth and Uptake Dynamics Under Different Drip-irrigation Strategies: New Insights Using ERT
Root uptake of water and nutrients is a dominant and crucial component in the design of efficient irrigation and fertigation practices for applications ranging from conventional irrigation to highly-advanced high frequency irrigation practices, as well towards other unique uses of land and water resources. Nevertheless, root water and nutrient uptake is often ignored or oversimplified when simulating soil water flow and solute transport. This is primarily so because of the a-priori unknown dynamic processes of root uptake, especially when coupled with spatially heterogeneous soil water and nutrient distributions. In this research we investigate the spatial and temporal patterns of root water uptake, and the way these patterns are influenced by environmental conditions. We consider the soil-root system as continuum. Our greenhouse setup includes bell pepper grown in sand under three different irrigation schemes, differing in the rate at which water is applied (high rate, small rate, and pulses). For each scheme we have two cylindrical growing chambers equipped with 96 ERT electrodes (one with and one without a plant), similar chambers with TDR probes, continuous weighting of chambers and of drainage, and 12 equal dimensions sacks for bi- weekly mapping of root presence. The experimental set-up enables the quantification of the dynamics of the root system, the total water uptake, the water regime within the growing medium, and the spatial and temporal distribution of the uptake function within the root zone by means of the ERT data, for each of the irrigation schemes. Preliminary results indicate significant difference in the root development and functionality for the different environmental conditions applied, up to 100 percents difference in uptake.
H13E-0970
Spatial Root Distribution and Clonal Connectivity in Trembling Aspen Along a Hillslope Catena; Implications for Above and Below Ground Water Regulation
Spatial exploitation of available soil water by tree roots is considered an important mechanism governing the ability of trees to survive in water limited environments. However, there is limited information on horizontal and vertical root distribution, particularly in mature trees growing in natural environments. Even less is known about morphological characteristics in clonal tree species, such as trembling aspen (Populus tremuloides), that likely affect intra- and inter-clonal water transport, uptake and transpiration dynamics. As part of a larger study that investigates above/below ground controls over water uptake, transpiration and hillslope dynamics in aspen, we measured and described the vertical and horizontal distribution of fine and coarse roots, and the occurrence of intra- and inter-clonal root grafts in several aspen clones along three belt transects along a hillslope catena. Each transect included an aspen clone at the base of the slope and a second clone located near the top of the slope. Root systems were excavated to a depth of 30cm in three 10cm layers using a fire pump. Most roots (79% of total) were found in the upper 20cm of the soil profile. Few roots were found below 30cm except for tap-roots, which did not exceed a depth of 1m. Fine roots (< 1cm dia.) comprised 52% of the total root surface area within the top 30cm of the soil profile, whereas medium (1-2.5cm dia.) and coarse (> 2.5cm dia.) roots comprised only 28% and 20% of total root system surface area, respectively. While fine roots dominated both the root mass per unit area and the root surface area in the upper 20cm, root size classes were more uniformly distributed at the lower soil layer. Although intra-clonal root grafts were very common, no inter-clonal grafts were observed suggesting that inter-clonal transport of water through root grafts to the upper hillslope is not likely. However, it was observed that the aspen clones located on the top of the hillslope had shallow roots that extended to the lower slope positions (up to 16m in length), presumably to access the more easily available water. This may be an important mechanism for water translocation uphill to the more water limited areas.
H13E-0971
Advances towards a low complexity model for calculating vertical root water uptake profiles
In current soil-vegetation-atmosphere transfer (SVAT) models, root water uptake is regarded as a sink term for 1-dimensional (vertical) Richards Equation. Thereby, root water uptake is allocated according to the product of rooting density distribution and a water stress factor. Where some SVAT models compensate water stress effects by balancing reduced uptake from dryer soil layers to wetter layers, none of these models take the nonlinear radial flow towards a single root into account. Also, none of these concepts consider the uptake mechanism in the root network. We develop a simplified model, which captures the main features of plant-water uptake but is computationally fast. This model is intended for implementation in SVAT schemes and for testing hypothesis on optimal root behaviour in different environments. Based on nonlinear models of water flow towards a single root, we identified different phases during water uptake, one where the water flow towards the root is almost linear (moist soil), and one where the flow is increasingly nonlinear (very dry soil). During the linear phase the uptake does not depend on root abundance (hence rooting density) but on other factors, like root radial and axial resistances, and rooting depths. If theses resistances are comparatively low, the uptake is uniform over all layers. During the non- linear flow phase the rooting density distribution becomes more and more important in determining the uptake profiles. We conclude that for modelling of root water uptake profiles, root system properties have to be included in the framework. Therefore, we present our mixed dimensional root architecture based water uptake model (aRoot), where a 3-dimensional root network is embedded in a 1-dimensional vertical soil column. In this framework the water flow is modelled along a network of resistances from the bulk soil over a chain of root segments up to the root collar. The non-linear water flow towards the root is modelled with a simplified equation that covers the dynamics of the non-linear Richards model, especially close to the roots. Our overall aim is to investigate the role that root profiles play for root water uptake in different regimes/phases as well as to define potential constraints when neglecting the root network.
H13E-0972
Neutron Radiography of Root Water Uptake
Water flow from soil to roots presents still important open questions: which parts of the roots are more active in water uptake? How do the soil properties affect the root uptake? In particular, which are the properties of the soil near the roots , i.e. the rhizosphere? We used neutron radiography and tomography to image the water content distribution in soils during root uptake. Rectangular (quasi 2D) and cylindrical containers were filled with sandy soil and planted with lupins. Three weeks after planting, the samples were equilibrated at -10 and -30 hPa and have been radiographed for 9 days at intervals of 6 hours. A region of water depletion formed around the tap root and the more proximal parts of the lateral roots. As the soil dried up, water was stored around the more distal parts of the lateral roots and it moved into the roots. When the soil was irrigated, steep gradients in water content formed around the roots, indicating a quick root uptake. High water content near roots and quick uptake after irrigation show that the soil near the roots is a region with specific hydraulic properties where fast fluxes and high gradients occur. We expect that the properties and dynamics of this soil region control the root water uptake.
H13E-0973
Relative Water Uptake as a Criterion for the Design of Trickle Irrigation Systems
Previously derived analytical solutions to the 2- and 3-dimensional water flow problems describing trickle irrigation are not being widely used in practice because those formulations either ignore root water uptake or refer to it as a known input. In this lecture we are going to describe a new modeling approach and demonstrate its applicability for designing the geometry of trickle irrigation systems, namely the spacing between the emitters and drip lines. The major difference between our and previous modeling approaches is that we refer to the root water uptake as to the unknown solution of the problem and not as to a known input. We postulate that the solution to the steady-state water flow problem with a root sink that is acting under constant, maximum suction defines un upper bound to the relative water uptake (water use efficiency) in actual transient situations and propose to use it as a design criterion. Following previous derivations of analytical solutions we assume that the soil hydraulic conductivity increases exponentially with its matric head, which allows the linearization of the Richards equation, formulated in terms of the Kirchhoff matric flux potential. Since the transformed problem is linear, the relative water uptake for any given configuration of point or line sources and sinks can be calculated by superposition of the Green's functions of all relevant water sources and sinks. In addition to evaluating the relative water uptake, we also derived analytical expressions for the steam functions. The stream lines separating the water uptake zone from the percolating water provide insight to the dependence of the shape and extent of the actual rooting zone on the source- sink geometry and soil properties. A minimal number of just 3 system parameters: Gardner's (1958) alfa as a soil type quantifier and the depth and diameter of the pre-assumed active root zone are sufficient to characterize the interplay between capillary and gravitational effects on water flow and the competition between the processes of root water uptake and percolation. For accounting also for evaporation from the soil surface, when significant, another parameter is required, adopting the solution of Lomen and Warrick (1978).
H13E-0974 TI: Spatial structure of soils and hydraulic properties in desert piedmonts of interspersed shrub and bare soil mosaic affects our ability to understand large scale water budget and other processes important to desert environments and ecosystems. Recent studies indicated that spatial structure of soil hydraulic properties exists at local scale and is highly correlated and scalable to shrub size. This study investigates the implications of distinct hydraulic properties structures of under canopy and interspace in a desert environment based on the observed variability and spatial correlation of hydraulic properties using high- resolution characterizations of unsaturated hydraulic conductivity from the shrub and interspace mosaic. The impact on the large scale water budget of spatial variability of desert shrubs coverage is also investigated with different ranges of correlation length for the shrub distributions. Given the distribution of the shrub coverage, two distinct fields for the hydraulic parameters, based on the statistics obtained form field characterizations are generated. The first field is the under canopy hydraulic parameters, and the second one represents the interspace parameters. We combine these two fields to produce one complete spatial distribution for the hydraulic parameters. The hydrological process is then calculated for the local scale and then aggregated to large scale water flux partitions. The impact on the moisture redistribution and large scale fluxes of both hydraulic property variability and the spatial structure of shrub coverage is addressed and discussed.
H13E-0975
A Coupled Formulation for Vadose Zone Transport of Multiple Gas Species With Plant Exchange Under Variable Gravity
Most plants require a balance between water availability and oxygen availability in the rooting zone. Procedures for raising plants under microgravity conditions, such as might be encountered in long-term space missions, face a special challenge: water redistribution is not affected by gravity, leading to difficulty in maintaining both water and oxygen levels in the rooting zone because flow is dominated by capillary properties. The plant substrate used for microgravity conditions is typically a coarse material that drains extremely rapidly under the fluctuating gravity conditions (0 to 1.8 G) experienced on KC-135 aircraft during flight parabolas. To evaluate control strategies for meeting plant water uptake and respiration needs under microgravity and to characterize flow redistribution under fluctuating gravity, a single formulation considering partial to full saturation was developed to cover this range of conditions. The fully coupled system of equations considers N>1 gaseous species, including water, that are all constituents in the liquid phase and in equilibrium between the gas and liquid phases where both phases are present. Plants are considered as separate quasi steady continua. Plant uptake and respiration, when considered, are defined using (possibly age dependent) transfer functions characterized by root length density. The formulation avoids complexities arising from switching variables when going from very dry to saturated conditions by using variables that are continuous throughout the domain: liquid pressure and N-1 mass fraction variables, expressed as partial capillary pressure. The mass fractions of all species in both phases are recovered from the standard equilibrium conditions used to define the partial capillary pressures. The use of partial capillary pressure state variables is inspired by mass balance considerations near saturation (where capillary pressure is almost zero), because mass balance convergence rates are dominated by the phase balance between gas and liquid. Because the capillary pressure/saturation relationship for the coarse plant medium has a cusp at a saturation of 1, a further step was needed to enable fluctuating-gravity simulations to converge as a water table forms after the onset of 1.8 G conditions. Typically the retention function is smoothed near the cusp to improve convergence; a simple alternative mass-conservative linearization in the updating scheme eliminates the convergence issue without altering the retention relationship.
H13E-0976 INVITED
Discriminating Mechanistic Models of Root Water Uptake Under Stress Conditions
A variety of mechanistic models have been proposed to describe the effect of drought and salinity stresses on the uptake of water by plant roots. Examples from the literature demonstrate that, in practice, it has been difficult to discriminate among these models owing to the difficulty of measuring relevant soil and plant parameters under a range of stress conditions. A new greenhouse installation aimed at obtaining data relevant to modeling uptake under stressed conditions will be discussed and preliminary data presented.
H13E-0977 INVITED
Modeling and measuring 3-D root water uptake from the plant to the field scale
Root water extraction from soil is controlled by the local water potential gradient between soil and roots. However taking this local effect in consideration is challenging given the lack of real data on root architecture, on the distribution of the water potential at the root-soil interface, or on the local hydraulic properties of soil and roots. In order to numerically investigate the water uptake processes at the plant scale, we built a 3-D model, called R-SWMS, which combines water flow in conducting vessels of a plant and within the soil matrix. We use light transmission imaging techniques to monitor two-dimensional soil water distribution and validate our modeling approach. We show the importance of combining architecture with hydraulic models to obtain an integrated and functional characterization of root-soil interactions. At the field scale, the spatial variability between plant development and root water uptake cannot be neglected. We show how these processes can be assessed at the field scale thanks to geophysical techniques.
H13E-0978 INVITED
Plant water uptake at the single plant scale: experiment vs. model
This study tested the hypotheses that the soil is the main resistance to the extraction of water by the plant roots, owing to a combination of low root length density (unit length of root per unit volume of soil), low soil water diffusivity at low soil water content. To test this hypothesis wheat plants were grown in undisturbed and repacked clay-loam and repacked sand. The plants were kept in a controlled environment where they were challenged with a range of evaporative demands, first rising and then falling, and the transpiration rate, E, and the null measurement of the xylem water potential, B, were measured non-destructively and continuously. The experimental measurements were compared to the output of a mathematical model that solves the radial diffusion equation for the flow of water to a single plant root, assumed to represent all roots. For the repacked clay-loam and the repacked sand, the model could match the data during the rising phase of E, if it was assumed that only 10% of the roots were taking up water and that the soil water diffusivity was constant and low. However it could not match the data during the falling phase of E, unless it was assumed that there had been a significant rise in the hydraulic resistance of the plant, or perhaps more likely, that an additional, yet constant, interfacial resistance had developed when E was high and B was rapidly increasing. That the slope of B(E) during the falling phase of E, for the repacked clay-loam and the repacked sand, was essentially constant suggests that the radial flow of water through the soil generated only minor gradients in soil suction and therefore that neither low soil water diffusivity nor low root length density was inhibiting the extraction of water from the soil by the plant roots. For the undisturbed clay-loam soil, the radial-flow model did not agree with the experimental data even when various combinations of soil water diffusivity and root length density were tried. This disagreement may have been due to a skewed distribution of roots in the cores, for few if any roots were seen at the base of any core. However, even when the roots were assumed to be confined to the top 50 or 75% of the total soil volume the model and experimental data did not agree. This work provides evidence that the flow of water to the plant roots, as encapsulated in the radial-flow model, is not inducing large gradients in suction close to the plant roots growing in the three soil types used in the experiments reported here. The clear disagreement between the experimental data and the model suggests that something else is generating the large hydraulic resistances evident between the soil and the leaves of the plants.