Authors: Josh Gottlieb, Dvir Ochman, Cheng-Wei Huang, Jean-Christophe Domec, Nimrod Schwartz, Samantha Hartzell
Categories: Perspective, Agricultural science, Earth sciences, Process in plant, Soil hydrology, Soil science
Source: iScience
Authors: Josh Gottlieb, Dvir Ochman, Cheng-Wei Huang, Jean-Christophe Domec, Nimrod Schwartz, Samantha Hartzell
Salt stress has a detrimental impact on crop yield and survival rates, which salt-tolerant cultivars can resist through numerous adaptive mechanisms. Most models of salt stress impacts on productivity and water use employ empirical or simplified schemes to represent salt-adaptive traits. However, with an increased understanding of these physiological tolerance mechanisms and emergent measurement techniques for monitoring key salinity dynamics, the potential for developing mechanistic agrohydrological models of the soil-plant-atmosphere continuum has grown. This perspective highlights strategies for modeling salt tolerance mechanisms, including root system architecture adaptation, salt filtration, adaptation of plant hydraulics, ion compartmentalization, and stomatal responses, to improve model representation and prediction. Incorporating these mechanisms into dynamic models can help inform management strategies and biotechnological cultivation, increasing long-term salt stress resilience within salt-affected agricultural systems.
Salinity issues are prevalent across many agricultural systems, affecting one-fifth of the world’s irrigated lands and resulting in $27.3 billion in lost crop production annually.^1^ Soil salinity results in a significant negative change in total plant water status (via a decrease in soil water potential), and stymies crop productivity through increased osmotic stress, declining photosynthetic capacity, and reduced stomatal conductance.^2^ Commonly exacerbating this issue is co-occurring drought stress, which disproportionately affects salinized lands.^3^ Lack of available fresh water for irrigation adds to anthropogenic salinization of soil and groundwater, causing alterations to agroecosystems and leading to long-term unsustainability for both annual and perennial crops. While general relationships between soil salinity and agricultural salt stress have been well studied, the responses of various agricultural crops to soil salinization remain complex, mediated by many factors at both the soil and plant level (see Figure 1). Soil-plant-atmosphere continuum (SPAC) models incorporating both crop-specific responses and salinity dynamics can be a tool for predicting the effects of salt-stress adaptations on crop productivity.Figure 1Translating soil salinity to agricultural salt stress requires information regarding intermediate crop-specific processes and traitsThese include root system architecture (RSA) adaptation to salt stress, salt filtration (FE), acclimation of plant hydraulic properties (root radial conductance (kr,radial), root axial conductance (kr,axial), and stem xylem conductance (ks,plant)), ion compartmentalization, and responses of stomatal conductance (gs), stomatal-mesophyll conductance (gsm) and photosynthetic machinery to salinization. Water and salt transport are driven by gradients in water potential (ψ) between each component, which are functions of dynamically changing relative water content (θ(t)) and salt concentration, (c(t)). Resulting water use (TrG) and carbon assimilation (An) depend on salinity adaptations across the soil-plant-atmosphere continuum.
Salinity tolerance varies widely across plant species. Halophytes, which maintain productive growth in saline soils, have developed adaptive strategies to maintain function in conditions of high osmotic stress,^4^ caused by lowered soil water potential, and frequently, ionic stress, caused by salt buildup which leads to cellular dysfunction and photosynthetic reduction.^5^ Many important agricultural crops including barley, cotton, sugarbeets, olives, and quinoa show varying degrees of halophytic capacity, maintaining productivity up to specific thresholds of soil salinity.^6^ Plant traits and mechanisms contributing to salt tolerance response at the cellular, tissue, and macroscopic level have been described in a multitude of review articles.^4^^,^^5^^,^^7^ Traits such as ability to modify root system architecture (RSA) through expansion of fine and lateral roots,^8^ adjustment of xylem structure, and vacuolar capacity to compartmentalize ions explain salt tolerance variation between species and settings.^6^ Traits which impact salt stress responses vary widely between major crops and have less frequently been incorporated as mechanisms within distinct components of SPAC hydraulic models. In the cases when these mechanisms have been modeled, the models have frequently been simplified in ways that limit applicability to agricultural scenarios. For example, plant osmoregulation has been incorporated into SPAC models on a subdaily-to-daily timescale through steady-state analysis. These models have treated salt filtration efficiency as a model parameter impacting transpiration and plant water storage.^2^^,^^9^ However, the dynamic variability of crop traits including filtration efficiency have not been incorporated into longer simulations (e.g., growing season timescale) over which significant changes in plant salt balance may occur.
Numerical SPAC models of saline systems have largely focused on soil-root zone dynamics. Many models have aimed to capture the relationship between root zone salinity, a varying number of plant parameters, meteorological conditions, and transpiration.^6^^,^^10^^,^^11^^,^^12^^,^^13^ Sophisticated one, two, and three-dimensional models, such as HYDRUS, SOTE, and SALTMED have been applied to understand the impact of salinity on soil conductivity, root water uptake, and the evolution of soil moisture temporally and spatially.^7^ Plant parameters used in these models have included leaf and root area index, root zone depth, and crop-specific coefficients used to fit a salinity and water stress reduction functions for transpiration. However, plant transpiration (often calculated using crop coefficients) is frequently used as a boundary condition in these models, which do not typically represent the complexity of stomatal response, salt filtration, root plasticity, and plant water storage dynamics in saline contexts.
The impact of salt stress on important agricultural indicators, such as productivity and water use efficiency, could be more thoroughly understood using mechanistic agrohydrological models which incorporate crop-specific salt tolerance traits and their dynamic adjustments over agriculturally-significant timescales (Figure 2). Treating salinity adaptations as evolving mechanisms that influence salt mass and water transport can expand the limits of previous models, which mostly assume crop parameters are static. In the following sections, we highlight the physiological dynamics involved in key halophytic mechanisms, namely dynamic root structure adaptation, filtration of salt ions in plant roots, dynamic modification of plant hydraulic properties and biomass allocation to alter water transport efficiency, osmoregulation within xylem and leaf water storage, and modulation of stomatal apertures to balance carbon assimilation and water conservation. For each mechanism, we suggest numerical methods for incorporating salt stress dynamics into common SPAC model frameworks, taking into account trait variation, and measurement considerations for model improvement.Figure 2Salt impacts and physiological response mechanisms of salt-toleranceSalt impacts (light blue) and responses (dark blue) within key parts of the soil-plant-atmosphere continuum vary over timescales (onset, growing season, perennial crop life cycle). The osmotic and ionic phases of salt stress are shown, with the osmotic stress phase lasting through the first part of the growing season before ionic stress dominates.
Root water uptake (RWU) of halophytes can be adversely impacted by exposure to salinity through salt accumulation at the root surface and within the root xylem.^14^ Tolerant species can counterbalance losses through one or more mechanisms of RSA adjustment, including increasing their root-shoot ratio,^8^ increasing the density of fine roots to expand soil-root contact area,^15^ expanding lateral roots to tap into broader areas of soil, and halotropism, the responsive root growth into soil patches with favorable osmotic conditions. These mechanisms provide halophytic crops more favorable soil moisture access within heterogeneous soil horizons.^16^
The influence of salinity on soil hydraulic properties has been described in previous reviews.^7^ Modeling efforts have indicated that saturated hydraulic conductivity losses may be larger in soils hosting halophytes, suggesting root-based tolerance mechanisms may increase risk of soil degradation.^6^ Radial and axial root conductances, which influence the rate at which water can pass from soil through the root membranes, and through the xylem of the root, can be modulated by salt-tolerant crops.^17^ Radial root conductance, typically the controlling conductance, is a function of the root surface area which is exposed to soil moisture.^18^ The more fine root surface area in contact with available moisture, the greater the potential for RWU. Root conductance can thus be calculated as a function of root surface area, which is itself a function of root length density, rooting depth, and the average diameter of fine roots.^18^ In a saline context, salt tends to build up around the surface of fine roots as a function of ionic filtration. Osmotic potential becomes increasingly negative, reducing the soil-root potential gradient. For crops with adaptive root plasticity, maintenance of RWU through dynamic root growth can be modeled in multiple ways. Because only the soil moisture directly adjacent to fine root surfaces exerts direct osmotic pressure, halotropism can be represented through a modification of the bulk soil water potential over time.^19^
In addition to halotropism, the increase in root surface area exhibited by many salt tolerant plants can be captured by treating root surface area as a dynamic biomass variable, influenced by time of exposure, and salinity at the root surface. Shifts in biomass allocation affect root and stem hydraulic conductive tissues differently as plants allocate carbon to roots or shoots adaptively,^20^ suggesting that the correlation between above- and below-ground hydraulic traits varies across contexts. Salinity can cause a significant shift in hydraulic resistance away from the stem to the roots, possibly because of differential biomass allocation, impacting the contribution of root to whole-plant hydraulic conductance.^21^ Roots are the first organs to be affected by salt stress,^22^ and may serve as “hydraulic fuses” that uncouple shoots hydraulically from transpiring surfaces and soil. For example, tree seedlings have been shown to change their root to shoot ratio depending on salinity.^23^^,^^24^ In one study, salinity caused reductions of 25% in whole-plant hydraulic conductance accompanied by a decline of more than 50% in stomatal conductance.^24^ Quantifying the relative contribution of root hydraulic efficiency to whole-plant hydraulic conductance seasonally and how it varies with salinity is thus essential for understanding how salt-stress influences transpiration and climatic sensitivity. Dynamic numerical schemes representing root surface area evolution within vertical and horizontal layers of the soil-root zone can be informed by empirical observations of root expansion in response to salt exposure.
Compared to RSA plasticity, which is multigenic and controlled by physiological processes which are incompletely understood, the mechanisms responsible for filtration of salt ions are well-resolved. The traits responsible for these filtration mechanisms have been described in many papers, and have been successfully targeted for engineering salt-tolerant crop cultivars.^8^ Salt-filtration is controlled by the production and optimization of selective and non-selective membrane transport structures in endodermal root cells. The suberin lamellae within the root endodermis acts as a semipermeable membrane, which excludes most salt from entering the roots, inducing an osmotic potential gradient between soil and root, impacting to RWU.^25^ Two main halophytic salt filtration strategies have been observed in various salt-accumulation in order to maintain productive leaf water potential and enhanced salt-exclusion to prevent ion toxicity.^26^ While a certain amount of salt uptake and transport through root-shoot xylem into leaf cells is necessary to facilitate osmotic adjustment and retention of a transpiration-conducive water potential gradient, too little filtration leads to insurmountable long-term stress on photosynthetic processes, especially for species without capacity to compartmentalize ions.
Even as halophytic crops mitigate internal stresses through root filtration, hydraulic dynamics in the rhizosphere may impose transpirational limitations. Studies have shown that salt filtration, also referred to as exclusion, increases root zone salinity buildup over time.^5^^,^^27^ Salt buildup leads to osmotic stress, weakening the driving force for plant water uptake^5^^,^^28^ and decreasing productivity. Recent work^15^ compared the halophyte Suaeda salsa to the glycophyte Beta vulgaris L. in a salt exposure experiment and found that more salt accumulated around the roots of the halophyte. The degree of water potential change within the rooting system is often significantly different than that in the bulk soil due to vastly differing levels of species-specific root filtration efficiency.^29^^,^^30^^,^^31^ Mechanistic SPAC models have the potential to illuminate these dynamics, particularly the interplay between productivity maintenance and long-term soil viability, factors heavily influenced by salt filtration.
Filtration has been incorporated parsimoniously into ecohydrologic models, but long-term impacts on SPAC elements have not been thoroughly explored. Perri et al.^2^ analyzed the steady-state impact of filtration efficiency on halophyte transpiration. Taking a minimalist approach, xylem water potential, ψxylem, was calculated as a function of filtration efficiency, FE, and soil salt concentration using a modified form of the Van’t Hoff equation. Filtration efficiency was modeled as a linearly increasing function of soil salinity, following experimental evidence for a particular species. Plant water storage potential and soil salt concentration in this framework were considered static, whereas a dynamic model would need to account for the interaction of these variables in order to realistically represent their evolution over the long term. Bioenergetic and conceptual soil-root models have been developed to quantify the impacts of salt filtration on plant growth and plant-water relations, but attempts at numerical modeling of both short- and long-term implications of salt filtration have scarcely been attempted.^19^^,^^32^ Given the complexity and productivity trade-offs involved, implementing mass balances of both water and salt ions in a SPAC model may be necessary to assess the contributions of filtration efficiency to crop salt tolerance. Filtration efficiency could be used to model the uptake of salt into the xylem over time (Δcx) using Δcx=(1−FE)(RWU)(cs), where cs is the salt concentration in the soil water, and RWU is RWU.
In addition to root-based adaptations, recent research has been crucial in distinguishing the whole-plant hydraulic strategies of different plant functional groups for coping with salinity. Plant responses to salinity have typically been viewed through the lens of hydraulic stress equivalent to extreme drought conditions.^33^^,^^34^ These studies have revealed that water flow in the SPAC is determined by the hydraulic conductance of soil and plant tissues (Kplant), which characterizes the structural capacity of the whole plant to move water.^35^ Thus, it seems reasonable to predict that many of the traits most likely to predict responses to salinity will be traits related to hydraulic capacity.
To maintain adequate carbon gain yet avoid desiccation, stomatal conductance (gs) must be balanced to respond to a suite of environmental and biological variables. In the gas phase and under steady state conditions,^36^ the water demand for transpiration on a leaf area basis (TrG) can be expressed mathematically in relation to leaf water potential (ψl), vapor pressure deficit (VPD), and atmospheric pressure (Pa) (Equation 1)TrG(ψleaf)=1.6gs(ψleaf)VPDPa.At a given planting density, stand phenology will be influenced by adjustments in leaf area (Aleaf), crop height (hc), average Kplant, and xylem-to-leaf area ratio (Axylem/Aleaf). Hence, in a bulk representation form, water supply or transpiration rate in the liquid phase TrL is driven by soil to leaf water potential difference ψsoil−ψleaf, i.e.,(Equation 2)TrL(ψleaf)=Kplant(ψleaf)(ψsoil−ψleaf),where Kplant is related to plant specific hydraulic conductivity, ks,plant, as,(Equation 3)Kplant(ψleaf)=ks,plant(ψleaf)(AxylemAleaf)1hc.
Combined, Equations 1 and 2 indicate that whole-plant hydraulic conductance and stomatal conductance must interact to control transpiration. From Equation 3, if well-watered conditions prevail (i.e., ψleaf ≪ ψsoil), and maximum TrL occurs when leaf water potential is near its “safety value” (generally −1.5 to −2.5 MPa in crops^37^^,^^38^^,^^39^), then increasing crop height logically leads to decreased TrL and thus TrG, unless other hydraulic adjustments mediate them. Under high salinity with ψsoil decreasing and becoming close to ψleaf, the adjustments needed to maintain TrL would need to include a decrease in leaf water potential (thereby making plants more susceptible to turgor loss and embolism-induced cavitation), an increase in Kplant, and/or an increase in Axylem/Aleaf. During the growing season an increase in newly formed (non-embolized) xylem tissues can compensate for some of the embolism-induced loss of hydraulic conductivity. However, since neither Axylem/Aleaf nor crop height vary significantly over a short time span (of hours to a day), rapid changes in Kplant and thus in TrL reflect changes in ks,plant due primarily to embolism-induced loss of hydraulic conductivity.^40^^,^^41^ Published studies demonstrate that stomatal conductance and plant hydraulic conductance are indeed tightly coordinated in trees grown under various salt treatments.^24^^,^^42^^,^^43^^,^^44^ Therefore, maintaining the integrity of the xylem hydraulic continuum from roots to leaves requires that the stomata be highly responsive in coordinating transpiration with dynamic variation in Kplant. In agricultural plants, several studies have reported strong relationships between stomatal conductance and whole-plant hydraulic conductance that are consistent with this scenario in crops including sugarcane,^39^ squash,^45^ corn,^46^ and fava bean.^47^
Maintaining a sufficient water supply to leaves is challenging because the mechanism driving water flux during transpiration places xylem under tension, making it vulnerable to cavitation-induced embolism.^35^ The bulk of research on xylem embolism involves woody plants, with few reports on crop plants for which xylem embolism was shown to cause more than 25% loss of Kplant at a leaf water potential less than −1.0 MPa.^48^^,^^49^^,^^50^^,^^51^^,^^52^ The water potential resulting in a 50% loss of ks,plant due to embolism (ψ50) has been widely used to compare species’ vulnerability to drought,^53^ and model plant mortality under extreme water stress. Like drought stress, soil salinity can lead to a decrease in leaf osmotic potential, and the loss of turgor pressure in the leaves.^23^ This osmotic adjustment decreases leaf water potential and therefore can potentially make embolism more likely even when water is fully available. While osmotic stress acts in a manner similar to drought stress, ionic stress could reduce ks,plant and ψ50 irreversibly. Ionic stress would affect cell growth, in particular decreasing pit membrane porosity and increasing membrane stiffness, thus reducing the probability of air-seeding and the vulnerability to cavitation-induced embolism.^54^
However, aside from a few studies on mangrove trees,^55^^,^^56^ there are limited reports on the effect of salinity on plant-specific hydraulic conductivity and we only know of two studies that looked at the effect of salinity on cavitation resistance in perennial species.^57^^,^^58^ Furthermore, only one study investigated annual or agricultural plants.^59^ Results from these studies showed that plants could adapt to salinity by becoming more resistant to cavitation-induced embolism, compensating for decreases in leaf water potential. In addition, salt stress reduced xylem hydraulic capacity and increased embolism, because the change in ψ50 did not fully compensate for the decrease in leaf water potential. This response may not be true in all species and scenarios, however, considering that a study measuring only loss of hydraulic capacity under field conditions did not find differences in the degree of embolism in salt-stressed olive varieties compared to irrigated ones.^21^ In maize, analysis of the impact of simultaneous changes in hydraulic conductivity and stomatal conductance on leaf water supply, TrL(ψleaf), and demand, TrG(ψleaf), illustrates the significance of these factors (Table 1; Figure 3). Under a VPD of 1 kPa, salinization decreases the maximal transpiration rate, Tr, from 6.6 to 4.6 mmol/m^2^/s as leaf water potential declines from −0.58 to −0.85 MPa (Figure 3A). When atmospheric dryness increases to 5 kPa, salinization causes maximal transpiration to decrease further, from 8.7 to 5.6 mmol/m^2^/s, as leaf water potential shifted from −0.85 MPa to −1.05 MPa (Figure 3C). While a reduction in transpiration due to lowered stomatal conductance may prevent catastrophic embolism and limit salt buildup in crop leaves, this can conjointly reduce carbon assimilation rates, resulting in lowered productivity.Table 1Model parameterization for maize with (s) and without (c) salinity treatmentModel parameterValueUnitsAxylem/Aleaf1/1740a–hc0.45bmpa101.325kPaψsoil,s−0.45cMPaψsoil,c−0.2cMPakpmax,c0.56ckg/m/s/MPakpmax,s0.45ckg/m/s/MPabc1.16cMPabs1.36cMPacc4.16c–cs3.61c–gs,max,c0.30cmol/m^2^/sgs,max,s0.29cmol/m^2^/sdc0.95cMPads1.0cMPaec2.4c–es3.4c–cs420ppmgres0mol/m^2^/sExpressions for kplant(ψleaf) and g1An(ψleaf) were fit to sigmoidal curves such that kplant(ψleaf)=kpmax∗exp(−(−ψleafb)c) and g1An(ψleaf)=gs,max∗exp(−(−ψld)e).aEstimated at the jointing stage following Liu et al.^60^bEstimated at the jointing stage following Qiu et al.^61^cFit to data from Li et al.^59^Figure 3Water supply and demand curves(A–C) Supply (TrL, solid lines) and demand (TrG, dashed lines) curves are shown as a function of leaf water potential, ψleaf, for VPD*=* 1 kPa (A), 3 kPa (B), 5 kPa (C). Curves are plotted for both control (c, green) and salinized (s, orange) conditions by incorporating the response of plant xylem hydraulic conductivity and stomatal conductance to leaf water potential in maize grown under varying salinity levels (data adapted from^59^) in (Equation 1), (Equation 2), (Equation 3). Points where the supply and demand curves intersect (filled circle) represent the maximum possible transpiration (Tr), while thicker lines represent the envelope of possible operating. Salt treatment concomitantly reduced maize leaf water potential, xylem sensitivity to water stress, and stomatal conductance sensitivity to water stress, thus allowing the maintenance of significant water flow at more negative leaf water potentials.
Ion compartmentalization is an osmotic adjustment mechanism, which allows many halophytes to accumulate salt ions without losing hydraulic and photosynthetic functioning. Crops relying more on salt filtration to manage stress (e.g., tolerant cultivars of wheat) may compartmentalize ions more extensively in their roots, while crops more reliant on osmotic adjustment (e.g., cotton, barley) may tend to compartmentalize more extensively in their leaves.^4^^,^^62^ Compartmentalization occurs when unfiltered ions, deposited at the end of the transpiration stream within leaf cell cytosol, are pumped through the vacuolar membrane. Vacuolar capacity within leaf cells is often a dominant tolerance trait, though some compartmentalization may also occur in dead or dying tissue.^5^^,^^63^^,^^64^ In concert with compartmentalization, crops with this adaptation produce osmoprotectants—solute molecules conducive to cellular functioning (e.g., proline) to match salt-driven decreases in vacuolar osmotic potential. Some species also upregulate transport structures to maintain the balance of K^+^ and other ions with compartmentalized Na^+^.^63^ The production of transporter proteins requires energy, and thus must be balanced with stomatal adaptations, which tend to lessen photosynthesis.^65^
Ionic compartmentalization as a tolerance mechanism in saline agriculture could be incorporated into dynamic models following Bartlett et al.,^66^ which described osmotic adjustment patterns during drought for various crops. The total water potential within plant storage compartments (ψstorage) can be represented as the sum of osmotic potential (ψo) and turgor pressure (ψp) when turgor is maintained, and solely ψo when relative water content (RWC) is below the turgor loss point, RWCtlp:(Equation 4)ψstorage={ψo+ψp,Θtlp<Θ≤Θftψo,Θr<Θ≤Θtlp.In this formulation θft, θtlp, and θr represent the volumetric water content at full cell turgor, the turgor loss point, and the point of residual water content (apoplastic fraction), respectively. The osmotic and turgor components of stored water potential are modeled as functions of the bulk elastic modulus of cells (ε), the osmotic potential at full turgor (π0), and the RWC at full turgor (RWCft). In the following equations, π0 and ψ0 are always negative, while ψp is positive or equal to (Equation 5)ψo=π0(θftRWCft−θr)(θ−θr),and(Equation 6)ψp=|π0|−ε(θftRWCft−θ)(θftRWCft−θr).
RWC at the turgor loss point has strong evidence as an indicator of drought stress tolerance, and is itself a function of these same crop-specific parameters (RWCtlp=π0+εε).^66^ In a salt-stressed context, ion compartmentalization as a tolerance mechanism might be effectively represented by allowing π0 to become more negative over time as osmoprotectants accumulate. This decline in π0 can be modeled using a crop-specific function based on experimental measurements of plant water storage potential and RWC and their changes over time. Additionally, certain halophytes secrete salt out of the cells onto the leaf surface through hydathodes. This adaptation could be modeled dynamically as a sink term in the leaf salt balance, reducing the leaf water salt concentration (cl) over time, making π0 less negative. Figure 4 shows the change in the relationship between RWC and ψstorage as π0 becomes more negative.Figure 4Pressure-volume curves showing the effects of modifying the osmotic potential at full turgorOsmotic potential at full turgor (π0) represents accumulation of salt ions in plant water storage compartments (ψstorage representing ψleaf or ψxylem). Replacing ions with osmoprotectants in the cell cytosol allows halophytes to maintain lower leaf water potential without incurring tissue damage, and maintain turgor at lower water content.
Bartlett et al.^66^ found that for numerous crop cultivars, plasticity in RWCtlp (as a function of π0) was a significant factor in drought tolerance. To our knowledge, the same hypothesis has not been tested in models of salinity tolerance. Using π0 to capture the process of ion compartmentalization may further the understanding of this tolerance mechanisms’ interaction with other salt tolerance mechanisms, which have been discussed. Methods for measuring osmotic adjustment (which ion compartmentalization strongly influences) have been reviewed in previous papers.^67^ Advancements in monitoring methods, which can provide more frequent measurements of osmotic adjustment will be beneficial in ground truthing models of this adaptation in salt-tolerant crops.
While impacts of water stress on plant productivity are widely included in gas exchange models using approaches including the Ball-Berry-Leuning model, the Medlyn stomatal optimization approach, and the Jarvis multiplicative approach,^68^ understanding and modeling the impacts of salinity stress through numerous pathways can be less straightforward due to the numerous, varied, and crop-specific salinity induced adaptations mentioned previously.^38^^,^^44^^,^^63^ Osmotic stress impacts productivity and water use through responses of stomatal and mesophyll conductance to declining leaf water potential, driven by lowered soil osmotic potential. Such an effect of salinity-induced changes in leaf water potential may be incorporated into already established water stress functions which impact stomatal conductance, gs, in Equation 1,^2^^,^^69^ for (Equation 7)gs(ψleaf)=1.6(1+g1/VPD)An(ψleaf)ca+gres,where ca is the CO2 concentration at the leaf surface and gres is the residual conductance to water, frequently taken to be zero. In such approaches, both the direct response of gas exchange to leaf water potential, embedded in An(ψleaf), and the slope of the relationship between gs and An, reflected in g1, are impacted by salinization, with g1 showing an almost linear relationship with soil salinity.^59^^,^^69^
A direct impact of salinity on photosynthetic capacity through ionic stress has been shown experimentally,^70^ with additions of NaCl having been shown to increase oxidative stress in chloroplasts, damaging the structure of thylakoid membranes and reducing their capacity for photon capture and electron transport.^71^ However, these impacts of ionic stress are difficult to model without estimates of salt concentration within plant tissue, the primary driver of ionic stress. Furthermore, the relationship between photosynthetic capacity and bulk tissue salt concentration can be disrupted by ion compartmentalization, which protects photosynthetic machinery from ionic stress. Current methods commonly use simultaneous measurements of soil or irrigation water salinity and crop productivity to infer the overall impacts of such competing processes.^2^^,^^69^ Modeling work assuming steady-state conditions of salt accumulation in the soil and plant tissue has endeavored to partition the effects of salt stress on photosynthesis among altered stomatal conductance, mesophyll conductance, and photosynthetic capacity by using stomatal optimization theory to compute parameters related to photosynthetic capacity and marginal water use efficiency.^33^ Other work has developed empirical reduction functions, which can be applied directly to the photosynthetic rate or stomatal conductance based on assumed values of tissue salinity.^2^ Such approaches are largely unable to capture the significant temporal dynamics involved in ionic stress, as they cannot account for the slow accumulation of Na^+^ in plant leaves. Pairing models of photosynthetic machinery and stomatal conductance with ion compartmentalization schemes as suggested in the previous section has the potential to isolate ionic stress responses of photosynthetic machinery to salinity. This would allow the development of more consistent frameworks for predicting productivity responses across longer timescales and in the presence of dynamically varying soil salinity and plant salt accumulation.
We propose that new techniques for measuring heterogeneity of soil salinity around the rooting system^72^ as well as assessing the plant leaf and stem osmotic potential in the field^73^ can be used in tandem with traditional salt-manipulation experiments. This will help to bridge the complex gap between soil salinization and agricultural productivity, as well as to better understand tradeoffs in plant hydraulic traits responses to salinization. While emerging experimental methods come with practical limitations (including limited temporal and spatial resolution and additional costs), they show promise for improving understanding of the process-based effects of heterogeneity and trait variability. Because soil salinity is heterogeneous under field conditions, it is important to compare greenhouse studies designed to screen plants for salinity tolerance to field responses.^4^ We also recommend the utilization of split-rooting experiments, in which salinity levels in various areas of the root zone can be precisely monitored and controlled,^74^ to better understand the impact of horizontal and vertical heterogeneity of soil salinity on productivity.
Combining modeling efforts with emerging technologies, such as electrical resistivity tomography,^72^ holds great potential for advancing our understanding of root-soil interactions under saline conditions. Measuring salinity at the root-soil interface scale is challenging due to the fine spatial resolution required and the dynamic nature of salt accumulation near active roots. Magnetic resonance imaging (MRI) offers a promising alternative for non-destructive 3D mapping of parameters such as sodium concentrations at the root-soil interface.^75^^,^^76^ To effectively track salt filtration in the roots, it is critical to simultaneously measure plant salt content and soil salinity. Plant osmotic potential may be measured by extracting samples of xylem sap or destructive sampling of leaves for analysis in an osmometer.^73^^,^^77^^,^^78^ Using plant sap salt content data can help parameterize the extent to which species perform both salt filtration and ionic compartmentalization in cell vacuoles, leading to clearer understanding of how salt tolerant plants adapt in agricultural settings. Measurements of plant sap ionic concentrations can be incorporated into one or multiple storage components within a resistance-capacitance or porous media SPAC model.
Improving model predictions of salt effects on gas exchange requires simultaneous measurements of photosynthesis, leaf water potential, stomatal conductance, and tissue ion concentrations. Accounting for temporal dynamics of Na^+^ accumulation in the plant tissue, as well as acquiring sufficient data for fitting species-specific parameters, remain primary challenges. Due to the effects of ion filtration and osmoregulation, relationships between total leaf water potential, carbon assimilation, and transpiration can be expected to respond differently to salinity based on the salt adaptation strategy and changes in plant hydraulics (see Figure 3). Given the overall goal of linking crop traits to above-and below-ground productivity and carbon cycling, it is necessary to rigorously test the hypothesized links between traits related to whole-plant hydraulic conductance and carbon gain, and between above- and below-ground traits. Most critically, we need to assess whether observed trait-trait relationships are consistent across salinity exposures. A series of combined measurements under varying salinity and drought levels should be undertaken to investigate the responses of structural (plant biomass, growth, specific leaf area, and specific root length) and physiological traits (leaf water potential, whole plant hydraulic conductance, water potential at 50% loss of hydraulic conductivity, osmotic potential, stomatal conductance, and leaf photosynthesis).^2^ Measuring hydraulic, anatomical and additional traits along with overall measures of growth and yield will determine, which traits best predict how stomatal conductance responses to drought are altered by salinity.
To supply accurate predictions regarding the impact of salinization on future agricultural productivity, it will be increasingly critical to incorporate the key processes and parameters governing plant salinity responses into dynamic mechanistic and process-based models. Such an approach has historically proven difficult due, on one hand, to the plethora of crop adaptations to salinity, and on the other hand, to the temporal dynamics of salt exposure, with salt exhibiting cumulative and long-term impacts over the growing season, as well as inducing changes to xylem and stand structure over the crop life cycle.^79^ When considering perennial and long-lasting agricultural products, such as orchard crops, the ability to predict long-term damage and plant mortality, in addition to short-term impacts on yield, is critical for evaluating lasting viability.
One major difficulty in the development of mechanistic models is the complexity of biochemical adaptations to salt stress, which may not fit neatly into traditionally considered physiological variables. In particular, osmotic regulation and ion sequestration are difficult to capture within traditional agrohydrological models. These features disrupt pressure-volume curves, modeled water potentials, and modeled flows of moisture within the soil-plant-atmosphere continuum. For example, hydraulic redistribution has been suggested to be distorted in salinized systems with respect to modeled values,^14^ most likely because of the complications of salt distribution within the plant xylem and rooting system. Assessing the responses of many physiological variables, such as stomatal conductance, photosynthetic capacity, and hydraulic conductivity, to co-occurring water and salt stress is also frequently confounded by co-occurring ion transport and sequestration. Munns^80^ analyzed the energetic costs associated with salt filtration in the roots and ion compartmentalization, and noted the tradeoff between investing in the two processes, which leads certain crops to filter more salt from entry (excluders) where other species rely more on ion compartmentalization to maintain low leaf water potential and sustain transpiration. Variability in the degree to which filtration efficiency and ion compartmentalization are employed has even been observed between subvarieties of crops such as bread wheat and durum wheat.^80^ Such conflicting responses could mean that certain crops with similar short-term responses to soil salinization could exert very different controls on the long-term evolution of an agroecosystem.
To resolve these key processes and parameters on a macroscopic scale, it will be critical to develop tandem modeling and measurement approaches for agricultural salinization. More complex porous media models using enhanced resistor-capacitor representation in analogy with Ohm’s law—including a representation of salt exclusion and vacuolar sequestration—may be used to drive hypothesis testing and sensitivity analysis to suggest future data acquisition efforts. Such approaches are sorely needed to predict the response of leaf gas exchange to co-occurring changes in salinization and climate. Both experimental and modeling work on a limited number of species suggests a positive effect of increasing atmospheric CO2 on agricultural productivity in the presence of salt stress,^33^^,^^41^ although modeled sensitivity of assimilation to atmospheric CO2 decreased significantly as soil salinity increased, suggesting that CO2 fertilization effects may only be relevant on mildly to moderately salt affected lands. In future climate regimes, dynamic agrohydrological models may strengthen predictions of combined drought and salinity impacts on crop productivity and water use requirements. This will aid in implementing effective management techniques, including crop selection, irrigation, and targeted soil remediation, across salinized agricultural lands.
S.H. acknowledges support from NSF-EAR-2423295 NSF-CBET 2139003, and DE-SC0023468. J.-C.D. acknowledges support from DOE DE-SC0023309.
J.G.: writing – original draft, investigation, visualization, writing – review and editing. D.O.: writing – original draft, investigation, writing – review and editing. C.-W.H.: conceptualization, funding acquisition, writing – review and editing. J.-C.D.: writing – original draft, investigation, writing – review and editing. N.S.: writing – original draft, investigation, funding acquisition, writing – review and editing. S.H.: writing – original draft, investigation, visualization, funding acquisition, supervision, writing – review and editing.
The authors declare no competing interests.