irrigation scheduling: advantages and pitfalls of plant-based methods

Irrigation scheduling: advantages and pitfallsof plant-based methods

Hamlyn G. Jones*

Plant Research Unit, Division of Environmental and Applied Biology, School of Life Sciences,University of Dundee at SCRI, Invergowrie, Dundee DD2 5DA, UK

Received 12 November 2003; Accepted 27 May 2004

Abstract

This paper reviews the various methods available for

irrigation scheduling, contrasting traditional water-

balance and soil moisture-based approaches with

those based on sensing of the plant response to water

deficits. The main plant-based methods for irrigation

scheduling, including those based on direct or indirect

measurement of plant water status and those based on

plant physiological responses to drought, are outlined

and evaluated. Specific plant-based methods include

the use of dendrometry, fruit gauges, and other tissue

water content sensors, while measurements of growth,

sap flow, and stomatal conductance are also outlined.

Recent advances, especially in the use of infrared

thermometry and thermography for the study of sto-

matal conductance changes, are highlighted. The rela-

tive suitabilities of different approaches for specific

crop and climatic situations are discussed, with the

aim of indicating the strengths and weaknesses of

different approaches, and highlighting their suitability

over different spatial and temporal scales. The poten-

tial of soil- and plant-based systems for automated

irrigation control using various scheduling techniques

is also discussed.

Key words: Dendrometry, sap-flow, stomatal conductance,

thermography, water balance.

Introduction

Irrigation scheduling has conventionally aimed to achievean optimum water supply for productivity, with soil watercontent being maintained close to field capacity. In manyways irrigation scheduling can be regarded as a matureresearch field which has moved from innovative science

into the realms of use, or at most the refinement, of existingpractical applications. Nevertheless, in recent years therehas been a wide range of proposed novel approaches toirrigation scheduling which have not yet been widelyadopted; many of these are based on sensing the plantresponse to water deficits rather than sensing the soilmoisture status directly (Jones, 1990a).

The increasing worldwide shortages of water and costs ofirrigation are leading to an emphasis on developingmethods of irrigation that minimize water use (maximizethe water use efficiency). The advent of precision irrigationmethods such as trickle irrigation has played a major role inreducing the water required in agricultural and horticulturalcrops, but has highlighted the need for new methods ofaccurate irrigation scheduling and control. In recent years ithas become clear that maintenance of a slight plant waterdeficit can improve the partitioning of carbohydrate toreproductive structures such as fruit and also controlexcessive vegetative growth (Chalmers et al., 1981), givingrise to what has been termed by Chalmers et al. (1986) as‘regulated deficit irrigation’ (RDI). Achievement of suc-cessful RDI depends on accurate soil moisture or plant‘stress’ sensing, and requires an ability to irrigate ‘little andoften’ on demand. A disadvantage of RDI is that it requireswater status to be maintained accurately within a rathernarrow tolerance; any excess application loses the advan-tage of the regulated deficit and can cost more in terms ofwater used, while any under-application can lead to severeyield or quality losses. An alternative recent innovation toachieve the same measure of growth control has been thedevelopment of partial root-zone drying (PRD), whereirrigation is supplied alternately to different parts of the rootsystem (Dry and Loveys, 1998; Stoll et al., 2000b). Apotential advantage of this method is that precise irrigationcontrol is probably less critical for success than it is forRDI, as plants can always obtain adequate water from the

* Fax: +44 1382 34275. E-mail: [email protected]

Journal of Experimental Botany, Vol. 55, No. 407, ª Society for Experimental Biology 2004; all rights reserved

Journal of Experimental Botany, Vol. 55, No. 407,

Water-Saving Agriculture Special Issue, pp. 2427–2436, November 2004

doi:10.1093/jxb/erh213 Advance Access publication 30 July, 2004

at University of T

exas at Austin on June 23, 2014

http://jxb.oxfordjournals.org/D

ownloaded from

http://jxb.oxfordjournals.org/

well-watered side of the root system and the drying sideprimarily provides a signal to modify growth and stomatalaperture (Stoll et al., 2000a).The range of crops to which RDI and PRD methods have

been applied is increasing all the time, but their greatestsuccesses have been in high-value horticultural and fruitcrops, usually those where the harvested part of the plant isits reproductive organ. Applications of such techniques toextensive arable crops are in their infancy, although thereare some exciting preliminary reports (Kang et al., 2000,2003). At present it is much less clear whether PRD or RDIwould be so valuable for vegetative crops, althoughappropriate application can be used to restrict growth, asis required for high quality in some ornamental crop species(RS Harrison-Murray, personal communication).The choice of irrigation scheduling method depends to

a large degree on the objectives of the irrigator and theirrigation system available. The more sophisticated sched-uling methods generally require higher-precision applica-tion systems; nevertheless even less sophisticated systemssuch as flood irrigation scheduling can benefit fromimprovements in irrigation scheduling as outlined here.The pressures to improve irrigation use efficiency and touse irrigation for precise control of vegetative growth, as inRDI, both imply a requirement for increased precision inirrigation control, maintaining the soil moisture statuswithin fine bands to achieve specific objectives in cropmanagement. Such objectives can only be met by precisionirrigation systems such as trickle irrigation that can applyprecise amounts of water at frequent intervals (often severaltimes per day). Effective operation of such systems equallyrequires a sensing system that determines irrigation need inreal time or at least at frequent intervals; this rules out large-scale manual monitoring programmes for such purposesand indicates a need for automated monitoring systems.

Basics of irrigation scheduling

The main methods that are used for irrigation scheduling, orthat have the potential for development in the near future,are summarized in Table 1. Irrigation scheduling is con-ventionally based either on ‘soil water measurement’,where the soil moisture status (whether in terms of watercontent or water potential) is measured directly to de-termine the need for irrigation, or on ‘soil water balancecalculations’, where the soil moisture status is estimated bycalculation using a water balance approach in which thechange in soil moisture (Dh) over a period is given by thedifference between the inputs (irrigation plus precipitation)and the losses (runoff plus drainage plus evapotranspira-tion). Soil moisture measurement techniques have been thesubject of many texts and reviews (Smith and Mullins,2000; Dane and Topp, 2002) and will not be addressedhere. Similarly, the detailed methods for estimating evapo-transpiration and calculation of crop water requirements for

different crops and different climates, as required in thewater balance calculation, have been reviewed in detail byAllen et al. (1999). Although the water balance approachis not very accurate, it has generally been found to besufficiently robust under a wide range of conditions.Nevertheless it is subject to the serious problem that errorsare cumulative over time. For this reason it is oftennecessary to recalibrate the calculated water balance atintervals by using actual soil measurements, or sometimesplant response measurements (as outlined below). Some ofthe main advantages and disadvantages of the differentirrigation scheduling approaches are outlined in Table 1.

A potential problem with all soil-water based approachesis that many features of the plant’s physiology responddirectly to changes in water status in the plant tissues,whether in the roots or in other tissues, rather than tochanges in the bulk soil water content (or potential). Theactual tissue water potential at any time therefore dependsboth on the soil moisture status and on the rate of water flowthrough the plant and the corresponding hydraulic flowresistances between the bulk soil and the appropriate planttissues. The plant response to a given amount of soilmoisture therefore varies as a complex function of evap-orative demand. As a result it has been suggested (Jones,1990a) that greater precision in the application of irrigationcan potentially be obtained by a third approach, the use of‘plant ‘‘stress’’ sensing’. For this approach irrigationscheduling decisions are based on plant responses ratherthan on direct measurements of soil water status; some ofthe possible physiological measurements and responses thatcan be used are discussed in the following section.

Plant-based methods for irrigation control

Introduction

If soil water-based measures are to be replaced by plant-based measures it is important to consider what measuresmight be most appropriate for irrigation scheduling pur-poses. Possible measures include direct measurements ofsome aspect of plant water status as well as measurements ofa number of plant processes that are known to respondsensitively to water deficits. One might expect that a directmeasure of plant water status should be the most rigorousand hence the most useful indicator of irrigation require-ment, although the question remains as to where in the plantthat quantity should be measured. In practice, as has beenargued strongly by Jones (1990b), most plants exercisesomemeasure of autonomous control over their shoot or leafwater status, tending to minimize changes in shoot waterstatus as the soil dries or as evaporative demand increases(Bates and Hall, 1981; Jones, 1983). In the long term, thiscontrol is achieved through changes in leaf area and rootextension, and in the shorter term through changes in leafangle, stomatal conductance, and hydraulic properties of the

2428 Jones

at University of T



ownloaded from


Table 1. A summary of the main classes of irrigation scheduling approaches, indicating their main advantages and disadvantages

Comments that relate to all methods in a section are not repeated in subsections.

Advantages Disadvantages

I. Soil water measurement(a) Soil water potential (tensiometers,

psychrometers, etc.)Easy to apply in practice; can be quite precise;at least water content measures indicate ‘howmuch’ water to apply; many commercialsystems available; some sensors (especiallycapacitance and time domain sensors) readilyautomated

Soil heterogeneity requires many sensors (oftenexpensive) or extensive monitoring programme(e.g. neutron probe); selecting position that isrepresentative of the root-zone is difficult;sensors do not generally measure water statusat root surface (which depends on evaporativedemand)

(b) Soil water content (gravimetric;capacitance/TDR; neutron probe)

II. Soil water balance calculations(Require estimate of evaporation and rainfall) Easy to apply in principle; indicate ‘how much’

water to applyNot as accurate as direct measurement; needaccurate local estimates of precipitation/runoff;evapotranspiration estimates require goodestimates of crop coefficients (which dependon crop development, rooting depth, etc.);errors are cumulative, so regular recalibrationneeded

III. Plant ‘stress’ sensing(Includes both water status measurement andplant response measurement)

Measures the plant stress response directly;integrates environmental effects; potentiallyvery sensitive

In general, does not indicate ‘how much’ waterto apply; calibration required to determine‘control thresholds’; still largely at research/development stage and little used yet for routineagronomy (except for thermal sensing in somesituations)

(a) Tissue water status It has often been argued that leaf water status isthe most appropriate measure for manyphysiological processes (e.g. photosynthesis),but this argument is generally erroneous (as itignores root–shoot signalling)

All measures are subject to homeostaticregulation (especially leaf water status), thereforenot sensitive (isohydric plants); sensitive toenvironmental conditions which can lead toshort-term fluctuations greater than treatmentdifferences

(i) Visible wilting Easy to detect Not precise; yield reduction often occurs beforevisible symptoms; hard to automate

(ii) Pressure chamber (w) Widely accepted reference technique; mostuseful if estimating stem water potential (SWP),using either bagged leaves or suckers

Slow and labour intensive (therefore expensive,especially for predawn measurements);unsuitable for automation

(iii) Psychrometer (w) Valuable, thermodynamically based measure ofwater status; can be automated

Requires sophisticated equipment and highlevel of technical skill, yet still unreliable inthe long term

(iv) Tissue water content (RWC, leafthickness [c- or b-ray thicknesssensors], fruit or stem diameter)

Changes in tissue water content are easier tomeasure and automate than water potentialmeasurements; RWC more directly related tophysiological function than is total waterpotential in many cases; commercialmicromorphometric sensors available

Instrumentation generally complex or expensive,so difficult to get adequate replication; watercontent measures (and diameter changes)subject to same problems as other water statusmeasures; leaf thickness sensitivity limited bylateral shrinkage

(v) Pressure probe Can measure the pressure component of waterpotential which is the driving force for xylemflow and much cell function (e.g. growth)

Only suitable for experimental or laboratorysystems

(vi) Xylem cavitation Can be sensitive to increasing water stress Cavitation frequency depends on stressprehistory; cavitation–water status curve showshysteresis, with most cavitations occurringduring drying, so cannot indicate successfulrehydration

(b) Physiological responses Potentially more sensitive than measures oftissue (especially leaf) water status

Often require sophisticated or complexequipment; require calibration to determine‘control thresholds’

(i) Stomatal conductance Generally a very sensitive response, except insome anisohydric species

Large leaf-to-leaf variation requires muchreplication for reliable data

– Porometer Accurate: the benchmark for research studies Labour intensive so not suitable for commercialapplication; not readily automated (though someattempts have been made)

– Thermal sensing Can be used remotely; capable of scaling up tolarge areas of crop (especially with imaging);imaging effectively averages many leaves;simple thermometers cheap and portable; wellsuited for monitoring purposes

Canopy temperature is affected byenvironmental conditions as well as by stomatalaperture, so needs calibration (e.g. using wetand dry reference surfaces)

Irrigation scheduling 2429

at University of T



ownloaded from


transport system. In extreme cases, plants with good endog-enous control systems maintain a stable leaf water statusover a wide range of evaporative demand or soil watersupplies; these plants are termed ‘isohydric’ (Stocker, 1956),and include especially plants such as cowpea, maize, andpoplar (Bates and Hall, 1981; Tardieu and Simonneau,1998). This is by contrast with those species such assunflower or barley which appear to have less effectivecontrol of leaf water status and have been termed ‘anisohy-dric’. In practice the distinctions between isohydric andanisohydric behaviour are often not clear-cut; even differentcultivars of grapevine have been shown to have contrastinghydraulic behaviours (Schultz, 2003).The choice of which plant-based measure to use depends

on their relative sensitivity to water deficits. The definitionof sensitivity, however, is somewhat problematic. Therelative sensitivities of different physiological processeswere reviewed in some detail by Hsiao (1973), whoidentified cell growth as being most sensitive to tissuewater deficits, closely followed by wall and protein syn-thesis, all of which could respond to water deficits of lessthan 0.1 MPa (Fig. 1). Hsiao reported that stomatal closurewas only rarely affected when tissue water potential fell by0.2–0.5 MPa, with decreases of 1.0 MPa or more beingrequired for stomatal closure in many cases. Althoughphotosynthesis was classified as moderately sensitive byHsiao, largely as a result of its dependence on stomatalaperture, some component processes such as electrontransport are now known to be particularly insensitive(Massacci and Jones, 1990). It is now believed that Hsiao’s(1973) classification is somewhat misleading, and under-estimates the true sensitivity of the stomata, as it is based onobserved responses to leaf water potential alone andignores the internal root–shoot signalling that is nowknown to play a major part in controlling stomatal aperture(Davies and Zhang, 1991).The error arising from a reliance on leaf water status is

readily apparent when one considers that many plantsoperate when optimally watered with the leaf waterpotential at around �2 MPa, yet the stomata may close asthe soil dries by only a few tens of Pa, with little change inleaf water potential (Bates and Hall, 1981). A furtherconsideration is that any attempt to relate stomatal apertureto leaf water potential in a long-term drought experiment

can also be misleading, because with slowly developingstress the plant adapts by decreasing leaf area; as a resultstomatal conductance and photosynthesis rate per unit leafarea may remain fairly stable as soil dries (Moriana andFereres, 2002). Nevertheless, over shorter time-scales it stillappears that stomata are a particularly sensitive earlyindicator of water deficits.

In principle, water status is not ideal as a measure ofwater deficit as it is already subject to some physiologicalcontrol, and indeed, as has been outlined above, leaf waterpotential generally shows some homeostasis. Nevertheless,changes in water status somewhere in the plant system areassumed to be a prerequisite for any physiological adapta-tion or other response. All that a homeostatic system can dois to minimize, not eliminate, the changes in water status;indeed for a feedback system of stomatal control it is nottheoretically possible for such a system to stably eliminatechanges in shoot water status if that is the variable thatactually controls the stomata (Jones, 1990b; Franks et al.,1997).

In general, the use of any plant-based or similar indicatorfor irrigation scheduling requires the definition of referenceor threshold values, beyond which irrigation is necessary(Table 1). Such reference values are commonly determinedfor plants growing under non-limiting soil water supply(Fereres and Goldhamer, 2003), but obtaining extensiveinformation on the behaviour of these reference values asenvironmental conditions change is an important stage inthe development and validation of such methods. Another

Reduction in tissue ψ (MPa) required to affect process

Process affected 0 1 2

Cell growthWall synthesisProtein synthesisProtochlorophyllideNitrate synthaseABA accumulationCytokinin concentrationStomatal openingPhotosynthesisRespirationProline accumulationSugar accumulation

Fig. 1. Generalized sensitivities of plant processes to water deficits(modified with permission from Hsiao, 1973).

Table 1. Continued

Advantages Disadvantages

– Sap-flow sensors Sensitive Only indirectly estimates changes inconductance, as flow is also very dependenton atmospheric conditions; requires complexinstrumentation and technical expertise; needscalibration for each tree and for definition ofirrigation control thresholds

(ii) Growth rate Probably the most sensitive indicator of waterdeficit stress

Instrumentation delicate and generallyexpensive

2430 Jones

at University of T



ownloaded from


general limitation to plant-based methods is that they do notusually give information on ‘how much’ irrigation to applyat any time, only whether or not irrigation is needed.

Plant water status

Perhaps the first approach to the use of the plant itself as anindicator of irrigation need, and one that is still frequentlyadopted today, was to base irrigation on visible wilting.Unfortunately, by the time wilting is apparent a substantialproportion of potential yield may already have been lost(Slatyer, 1967). More rigorous and more sensitive measuresof plant water status are therefore required. Althoughrelative water content (RWC) (Barrs, 1968) is a widelyused measure of water status that does not require sophis-ticated equipment, it is often argued that water potential,especially of the leaves (wleaf) is a more rigorous and moregenerally applicable measure of plant water status (Slatyer,1967; Jones, 1990b). In spite of this, RWC has theadvantage that it can be more closely related to cell turgor,which is the process directly driving cell expansion, than itis to the total water potential (Jones, 1990b).

The fact that plant water status, and especially leaf waterstatus, is usually controlled to some extent by means ofstomatal closure or other regulatory mechanisms, arguesagainst the use of such measures, especially in stronglyisohydric species. A further problem with the use of leafwater status as an indicator of irrigation need was pointedout by Jones (1990b), who noted that even though there wasoften homeostasis of leaf water potential between differentsoil moisture regimes, rapid temporal fluctuations are oftenobserved as a function of environmental conditions (such aspassing clouds). This makes the interpretation of leaf waterpotential as an indicator of irrigation-need doubly unsatis-factory. Nevertheless, in spite of the concerns with the useof leaf water status that have been outlined above, it hasbeen reported that leaf water potential can, when correctedfor diurnal and environmental variation, provide a sensitiveindex for irrigation control (Peretz et al., 1984).

As a partial solution to the variability of leaf water status,various workers have proposed that a more useful and morerobust indicator of water status is the xylem water potentialor stem water potential (SWP, measured by using a pressurechamber on leaves enclosed in darkened plastic bags forsome time before measurement and allowed to equilibratewith the xylem water potential; McCutchan and Shackel,1992). As a more stable measure of water status, othershave even recommended that measurements should bemade on pre-equilibrated leaves from root suckers (Jones,1990a; Simonneau and Habib, 1991). These methods arethought to be preferable largely because they approachmore closely the soil water status than does the value of leafwater potential, although as a result they therefore miss outon the potential advantages of plant-based methods.

Perhaps an even better estimator of the soil waterpotential is the predawn leaf water potential (as wleaf should

largely equilibrate with wsoil by dawn). Unfortunately this isoften found to be rather insensitive to variation in soilmoisture content (Garnier and Berger, 1987). Further, thisis not very convenient for irrigation scheduling as routinemeasurements predawn are expensive to obtain, and at bestcan only be obtained daily. As yet another alternative, Jones(1983) suggested the indirect estimation of an effective soilwater potential at the root surface of transpiring plantsbased on measurements of leaf water potential and stomatalconductance during the day, and argued that this shouldhave significant advantages over predawn measurements.Such an approach has been successfully tested by Lorenzo-Minguez et al. (1985).

None of the above plant-based methods are well adaptedfor automation of irrigation scheduling or control becauseof the difficulties of measurement of any of the variablesdiscussed. Although it may be possible to use automatedstem or leaf psychrometers (Dixon and Tyree, 1984), theseinstruments are notoriously unreliable. In conclusion, it isapparent from the above discussion that the favoured wayto use plant water status is actually as an indicator of soilwater status; this negates many of the advantages ofselecting a plant-based measure! Indeed soil water potentialcan be measured directly, thus avoiding the need for anyplant-based measurement, although it is worth noting thatthis does not necessarily give a good measure of theeffective water potential at the root surface during activetranspiration (Jones, 1983).

Several indirect methods for measuring or monitoringwater status have been developed as alternatives to directmeasurement. The general behaviour of a number of suchmethods have been compared by McBurney (1992) andSelles and Berger (1990). In general, these indirect methodssuffer from the same disadvantages as do the directmeasurements of leaf water status, but in certain circum-stances have been developed into commercial systems.Some of these approaches are reviewed below.

Leaf thickness: A number of instruments are available forthe routine monitoring of leaf thickness, which is known todecrease as turgidity decreases. Approaches include directmeasurement using linear displacement transducers (e.g.LVDTs [Burquez, 1987; Malone, 1993] or capacitancesensors [McBurney, 1992]) or through measurements ofleaf ‘superficial density’ using b-ray attenuation (Jones,1973). Unfortunately, leaf thickness is frequently even lesssensitive to changes in water status than is leaf watercontent because, especially with younger leaves, a fractionof leaf shrinkage is often in the plane of the leaves ratherthan in the direction of the sensor (Jones, 1973).

Stem and fruit diameter: Stem and fruit diameters fluctuatediurnally in response to changes in water content, and sosuffer from many of the same disadvantages as other waterstatus measures. Nevertheless, the diurnal dynamics ofchanges in diameter, especially of fruits, have been used to


at University of T



ownloaded from


derive rather more sensitive indicators of irrigation need,where the magnitude of daily shrinkage has been used toindicate water status, and comparisons of diameters at thesame time on succeeding days give a measure of growthrate (Huguet et al., 1992; Li and Huguet, 1990; Jones,1985). Although changes in growth rate provide a particu-larly sensitive measure of plant water stress, such dailymeasurements are not particularly useful for the control ofhigh-frequency irrigation systems. Nevertheless, severalworkers have achieved promising results for low-frequencyirrigation scheduling by the use of maximum daily shrink-age (MDS). For example, Fereres and Goldhamer (2003)showed that MDS was a more promising approach forautomated irrigation scheduling than was the use of stemwater potential for almond trees, while differences inmaximum trunk diameter were also found to be particularlyuseful in olive (Moriana and Fereres, 2002). The use ofsuch dendrometry or micromorphometric techniques hasbeen developed into a number of successful commercialirrigation scheduling systems (e.g. ‘Pepista 4000’, DeltaInternational, Montfavet, France); these are usually appliedto the study of stem diameter changes. Selles and Berger(1990) reported that variations in trunk diameter or stemwater potential were more sensitive as indicators of irriga-tion need than was the variation in fruit diameter. This wasprobably a result of the poor hydraulic connection betweenfruit tissue and the conducting xylem. There is currentlymuch interest in evaluating such techniques for irrigationscheduling, with a number of relevant papers presented atrecent meetings (e.g. the International Society for Horticul-tural Science 4th International Symposium on Irrigation ofHorticultural Crops, 1–5 September 2003, Davis, CA, USA[as yet unpublished], and Kang et al., 2003).

c-ray attenuation: A related approach to the study ofchanges in stem water content was the use of c-rayattenuation (Brough et al., 1986). Although this was shownto be very sensitive, safety considerations and cost havelargely limited the further application of this approach.

Sap flow

The development of reliable heat pulse and energy balancethermal sensors for sap-flow measurement in the stems ofplants (Granier, 1987; Cohen et al., 1981; Cermak andKucera, 1981) has opened up an alternative approach toirrigation scheduling based on measurements of sap-flowrates. Because sap-flow rates are expected to be sensitive towater deficits and especially to stomatal closure, manyworkers have tested the use of sap-flow measurement forirrigation scheduling and control in a diverse range ofcrops, including grapevine (Eastham and Gray, 1998;Ginestar et al., 1998a, b), fruit and olive trees (Ameglioet al., 1998; Fernandez et al., 2001; Giorio and Giorio,2003; Remorini and Massai, 2003) and even greenhousecrops (Ehret et al., 2001).

Although the changes in transpiration rate that sap flowindicates are largely determined by changes in stomatalaperture, transpiration is also influenced by other environ-mental conditions such as humidity. Therefore changes insap flow can occur without changes in stomatal opening.Even though rates of sap flow may vary markedly betweentrees as a result of differences in tree size and exposure, thegeneral patterns of change in response to both environ-mental conditions and to water status are similar (Easthamand Gray, 1998). Appropriate sap-flow rates to use as‘control thresholds’ may be derived by means of regularcalibration measurements, especially for larger trees. Al-ternatively, it is at least feasible in principle to derive anirrigation scheduling algorithm that is based on an analysisof the diurnal patterns of sap flow, with midday reductionsbeing indicative of developing water deficits (though ofcourse diurnal fluctuations in environmental conditions canmimic such changes). Another potential problem with sapflow for precision control is that it tends to lag behindchanges in transpiration rate owing to the hydrauliccapacitance of the stem and other plant tissues (Wronskiet al., 1985).

It follows that, although sap-flow measurement is welladapted for automated recording and hence potentiallyautomated control of irrigation systems, it can be a littledifficult to determine the correct control points for any crop.

Xylem cavitation

It is generally accepted (Steudle, 2001) that water in thexylem vessels of transpiring plants is under tension; aswater deficits increase, this tension is thought to increase tosuch an extent that the water columns can fracture, or‘cavitate’. Such cavitation events lead to the explosiveformation of a bubble, initially containing water vapour.These cavitation events can be detected acoustically in theaudio- (Milburn, 1979) or ultrasonic-frequencies (Tyreeand Dixon, 1983), and the resulting embolisms may restrictwater flow through the stem. Substantial evidence, thoughlargely circumstantial, now indicates that the ultrasonicacoustic emissions (AEs) detected as plants becomestressed do indeed indicate cavitation events and that AErates can be used as an indicator of plant ‘stress’ (Tyree andSperry, 1989). Nevertheless, there remain many uncertain-ties as it seems that at least a proportion of the AEs detectedas woody tissues dry out may not be related to xylemembolisms. For example, the large numbers observed bySandford and Grace (1985) as coniferous stems dried outwere substantially in excess of the number of conductingtracheids present, thus suggesting a major contribution toobserved AEs by the non-conducting fibres (Jones andPena, 1986). Although the measurement of AEs has provedto be a powerful tool for the study of hydraulic architecturein plants, there has been little progress in adapting thismeasure as an indicator for irrigation scheduling. Note,however, the recent report by Yang et al. (2003), who

2432 Jones

at University of T



ownloaded from


implemented a control algorithm based on the associationof AE rate with transpiration rate for the precision irrigationof tomato. It is likely that the main reasons for the lack ofuptake include the fact that the relationship between thenumber of AEs and water status changes with successivecycles of stress, and the fact that cavitation events aremostly observed during the drying phase, not duringrewetting, and so cannot provide an indicator of whenirrigation has been sufficient to replenish the soil watersupply.

Stomatal conductance and thermal sensing

As outlined above, it appears that changes in stomatalconductance are particularly sensitive to developing waterdeficits in many plants and therefore potentially providea good indicator of irrigation need in many species. It is inthis area that most effort has been concentrated on thedevelopment of practical, plant-based irrigation schedulingapproaches. Although stomatal conductance can be mea-sured accurately using widely available diffusion porome-ters, measurements are labour-intensive and unsuitable forautomation. The recognition that leaf temperature tendsto increase as plants are droughted and stomata close(Raschke, 1960) led to a major effort in the 1970s and1980s to develop thermal sensing methods, based on thenewly developed infrared thermometers, for the detectionof plant stress (see reviews by Jackson, 1982; Jones andLeinonen, 2003; Jones, 2004).

An early method of accounting for the rapid short-termvariation in leaf temperature as radiation and wind speedvary in the field was to refer leaf temperatures to air tem-perature and to integrate these differences (e.g. the StressDegree Day measure of Jackson et al., 1977); significantelevation of canopy temperature above air temperature wasindicative of stomatal closure and water deficit stress. Themethod was transformed into a more practical approachfollowing the introduction of the crop water stress index

(CWSI) by Idso and colleagues (Idso et al., 1981; Jacksonet al., 1981), where CWSI was obtained from the canopytemperature (Tcanopy) according to

CWSI = ðTcanopy � TnwsÞ=ðTdry � TnwsÞ ð1Þwhere Tnws is a so-called non-water-stressed baseline tem-perature for the crop in question at the same atmosphericvapour pressure deficit, and Tdry is an independently derivedtemperature of a non-transpiring reference crop (Fig. 2). Inthis approach all temperatures are expressed as differencesfrom air temperature so that standard relationships for Tdryand Tnws can be used. Although this approach was found tobe useful in the clear arid climate of Arizona where themethod was developed, it has proved to be less useful inmore humid or cloudy climates where the signal-to-noiseratio is somewhat smaller (see Fig. 2; Hipps et al., 1985;Jones, 1999). In spite of its deficiencies, there has beenwidespread use of infrared thermometry as a tool inirrigation scheduling in many, especially arid, situations(Jackson, 1982; Stockle and Dugas, 1992; Martin et al.,1994), especially with the development of ‘trapezoidal’methods involving the combination of temperature datawith a visible/near infrared vegetation index (Moran et al.,1994).

In order to improve the precision of the approach in morehumid or low-radiation environments, Jones (1999) in-troduced the approach of using physical dry and wetreference surfaces to replace the notional Tdry and Tnwsrequired for equation 1. A number of recent papers haveshown that this approach can give reliable and sensitiveindications of stomatal closure (Diaz-Espejo and Verhoef,2002; Jones et al., 2002; Leinonen and Jones, 2004) andhence has the potential to be used for irrigation scheduling.The most important recent advances in the application ofthermal sensing for plant ‘stress’ detection and irrigationscheduling, however, have been provided by the introduc-tion of thermal imagery (Jones, 1999, 2004; Jones et al.,

Fig. 2. (A) Illustration of the calculation of Idso’s CropWater Stress Index: CWSI=(Tcanopy�Tnws)/(Tdry�Tnws), showing the dependence of Tnws (– – –)and Tdry (- - -) on air vapour pressure deficit (vpd, kPa). (B) Illustration of the effect of a given experimental ‘noise’ (for example resulting frommeasurement errors and variations in irradiance), indicated by the double-headed arrow, showing that the signal-to-noise ratio decreases markedly as thevpd decreases from levels found in hot and arid/semi-arid climates to values typical in humid or maritime climates.


at University of T



ownloaded from


2002), although their expense has meant that such systemshave yet to be widely used.In addition to the use of the absolute temperature rise as

stomata close, it has also been proposed that use may bemade of the fact that the variance of leaf temperatureincreases as stomata close (Fuchs, 1990). Indeed, this maybe a more sensitive indicator of stomatal closure than isthe temperature rise (Jones, 2004). Again, the introductionof thermal cameras now makes the wider use of suchapproaches feasible, especially when combined withautomated image analysis.

Automation

Themost widespread use of automated irrigation schedulingsystems is in the intensive horticultural, and especially theprotected cropping, sector. In general, the automated sys-tems in common use are based on simple automated timeroperation, or in some cases the signal is provided by soilmoisture sensors. For timer-based operation many systemssimply aim to provide excesswater to runoff at intervals (e.g.flood-beds or capillary matting systems), although some atleast attempt to limit water application by only applyingenough to replenish evaporative losses (often calculatedfrom measured pan evaporation; Allen et al., 1999). Muchgreater sophistication is required if an objective is to improvethe overall irrigation water use efficiency or to apply an RDIsystem. Most of the remaining automated systems currentlyin operation base control on soil moisture sensing; at leastthis approach has the potential for greater precision andimproved water use efficiency.Applications of automated plant-based sensing are largely

in the developmental stage, partly because it is usuallynecessary to supplement the plant-stress sensing by addi-tional information (such as evaporative demand). In princi-ple, with high-frequency on-demand irrigation systems onecould envisage a real-time control system where watersupply is directly controlled by a feedback controlleroperated by the stress sensor itself, so that no informationon the required irrigation amount is needed. For such anapproach carewill be necessary to take account of any lags inthe plant physiological response used for the control signal.The use of expert systems (Plant et al., 1992), which

integrate data from several sources, appears to have greatpotential for combining inputs from thermal or other cropresponse sensors and environmental data for a water budgetcalculation to derive a robust irrigation schedule.Among the various plant-based sensors that have been

incorporated into irrigation control systems are stem di-ameter gauges (Huguet et al., 1992), sap-flow sensors(Schmidt and Exarchou, 2000) and acoustic emissionsensors (Yang et al., 2003), though there has been mostinterest in the application of thermal sensors. For example,Kacira and colleagues (Kacira and Ling, 2001; Kaciraet al., 2002) have developed and tested on a small scale an

automated irrigation controller based on thermal sensing ofplant stress. Similar approaches have been applied in thefield: for example, Evans et al. (2001) and Sadler et al.(2002) mounted an array of 26 infrared thermometers(IRTs) on a centre pivot irrigation system which they usedto monitor irrigation efficiency, but had not developed thesystem to a stage where it could be used for fully automatedcontrol. Colaizzi et al. (2003) have tested another systemthat includes thermal sensing of canopy temperature ona large linear move irrigator (where the irrigator movesacross the field). In another approach to the use of canopytemperature that makes use of the ‘thermal kinetic win-dow’, Upchurch et al. (1990) and Mahan et al. (2000) havedeveloped what they call a ‘biologically identified optimaltemperature interactive console’ for the control of trickleand other irrigation systems based on canopy temperaturemeasurements. In this direct control system, irrigation isapplied as canopy temperature exceeds a crop-specificoptimum. The development of thermal infrared imagingmethods of irrigation control will be aided by the recentdevelopment of automated image analysis systems forextraction of the temperatures of leaf surfaces from thermalimages, including shaded and sunlit leaves, soil, and othersurfaces (Leinonen and Jones, 2004).

Conclusions

This review has briefly considered the current state of theart and potential opportunities for use of plant-based stresssensing as the basis for irrigation scheduling and control.The advantages and disadvantages of each of these ap-proaches are summarized in Table 1. Although plant-basedsensing has several potential advantages, including a greaterrelevance to plant functioning than soil-based measures,these have been offset by a number of practical difficultiesof implementation that have thus far limited the develop-ment of commercially successful systems. However, pres-sures for enhanced water use efficiency and for greaterprecision in irrigation systems are likely to provide a realimpetus for the development of new precision irrigationscheduling systems that take account of the irrigation needof individual plants, and may well involve greater use ofplant-based sensing systems.

Acknowledgements

The author is grateful to sponsors of various aspects of the work pre-sented, who include the European Commission (projects: WATER-USE, contract EVKI-2000–22061, and STRESSIMAGING, ContractHPRN-CT-2002–00254).

References

Allen RG, Pereira LS, Raes D, Smith M. 1999. Crop evapo-transpiration. FAO Irrigation and Drainage Paper 56. Rome: FAO.

2434 Jones

at University of T



ownloaded from


Ameglio T, Archer P, Cohen M, Valancogne C, Daudet FA,Dayau S, Cruiziat P. 1998. Significance and limits in the use ofpredawn leaf water potential for tree irrigation. Plant and Soil 205,155–167.

Barrs HD. 1968. Determination of water deficits in plant tissues. In:Kozlowski TT, ed.Water deficits and plant growth, Vol. I. Londonand New York: Academic Press, 235–368.

Bates LM, Hall AE. 1981. Stomatal closure with soil waterdepletion not associated with changes in bulk leaf water status.Oecologia 50, 62–65.

Burquez R. 1987. Leaf thickness and water deficits in plants: a toolfor field studies. Journal of Experimental Botany 38, 109–114.

Brough DW, Jones HG, Grace J. 1986. Diurnal changes in thewater-content of stems of apple trees as influenced by irrigation.Plant, Cell and Environment 9, 1–7.

Cermak J, Kucera J. 1981. The compensation of natural temper-ature gradient at the measuring point during sap flow-rate de-termination in trees. Biologia Plantarum 23, 469–471.

Chalmers DJ, Burge G, Jerie PH, Mitchell PD. 1986. Themechanism of regulation of Bartlett pear fruit and vegetative growthby irrigation withholding and regulated deficit irrigation. Journal ofthe American Society of Horticultural Science 111, 904–907.

Chalmers DJ, Mitchell PD, Vanheek L. 1981. Control of peach treegrowth and productivity by regulated water supply, tree densityand summer pruning. Journal of the American Society of Horti-cultural Science 106, 307–312.

Cohen Y, Fuchs M, Green GC. 1981. Improvement of the heatpulse method for determining sap flow in trees. Plant, Cell andEnvironment 4, 391–397.

Colaizzi PD, Barnes EM, Clarke TR, Choi CY, Waller PM,Haberland J, Kostrzewski M. 2003. Water stress detection underhigh frequency sprinkler irrigation with water deficit index.Journal of Irrigation and Drainage Engineering 129, 36–42.

Dane JH, Topp GC (eds). 2002. Methods of soil analysis. Part 4:Physical methods. Soil Science Society of America Books Series:5. Madison, WI: Soil Science Society of America.

Davies WJ, Zhang J. 1991. Root signals and the regulation ofgrowth and development of plants in drying soil. Annual Review ofPlant Physiology and Molecular Biology 42, 55–76.

Diaz-Espejo A, Verhoef A. 2002. Use of IR thermography forassessing the canopy exchange processes in sunflowers. In:Villalobos FJ, Testi L, eds. Proceedings of VII European Societyof Agronomy Congress, Cordoba. European Society of Agronomy,265–266.

Dixon MA, Tyree MT. 1984. A new stem hygrometer, corrected fortemperature-gradients and calibrated against the pressure bomb.Plant, Cell and Environment 7, 693–697.

Dry PR, Loveys BR. 1998. Factors influencing grapevine vigour andthe potential for control with partial rootzone drying. AustralianJournal of Grape and Wine Research 4, 140–148.

Eastham J, Gray SA. 1998. A preliminary evaluation of thesuitability of sap flow sensors for use in scheduling vineyard irriga-tion. American Journal of Enology and Viticulture 49, 171–176.

Ehret DL, Lau A, Bittman S, LinW, Shelford T. 2001. Automatedmonitoring of greenhouse crops. Agronomie 21, 403–414.

Evans DE, Sadler EJ, Camp CR, Millen JA. 2001. Spatial canopytemperature measurements using center pivot mounted IRTs.Proceedings of the 5th International Conference on PrecisionAgriculture, Bloomington, Minnesota, July, 2000, 1–11.

Fereres E, Goldhamer DA. 2003. Suitability of stem diametervariations and water potential as indicators for irrigation schedul-ing of almond trees. Journal of Horticultural Science and Bio-technology 78, 139–144.

Fernandez JE, Palomo MJ, Diaz-Espejo A, Clothier BE, GreenSR, Giron IF, Moreno F. 2001. Heat-pulse measurements of sap

flow in olives for automating irrigation: tests, root flow anddiagnostics of water stress. Agricultural Water Management 51,99–123.

Franks PJ, Cowan IR, Farquhar GD. 1997. The apparent feed-forward response of stomata to air vapour pressure deficit: infor-mation revealed by different experimental procedures with tworainforest trees. Plant, Cell and Environment 20, 142–145.

FuchsM. 1990. Infrared measurement of canopy temperature and de-tection of plant water stress. Theoretical and Applied Climatology42, 253–261.

Garnier E, Berger A. 1987. The influence of drought on stomatalconductance and water potential of peach trees growing in the field.Scientia Horticulturae 32, 249–263.

Ginestar C, Eastham J, Gray S, Iland P. 1998a. Use of sap-flowsensors to schedule vineyard irrigation. I. Effects of post-veraisonwater deficits on water relations, vine growth, and yield of Shirazgrapevines. American Journal of Enology and Viticulture 49,413–420.

Ginestar C, Eastham J, Gray S, Iland P. 1998b. Use of sap-flowsensors to schedule vineyard irrigation. II. Effects of post-veraisonwater deficits on composition of Shiraz grapes. American Journalof Enology and Viticulture 49, 421–428.

Giorio P, Giorio G. 2003. Sap flow of several olive trees estimatedwith the heat-pulse technique by continuous monitoring of a singlegauge. Environmental and Experimental Botany 49, 9–20.

Granier A. 1987. Evaluation of transpiration in a Douglas fir standby means of sap-flow measurements. Tree Physiology 3, 309–319.

Hipps LE, Ashrar G, Kanemasu ET. 1985. A theoretically-basednormalization of the environmental effects on foliage temperature.Agricultural and Forest Meteorology 35, 113–122.

Hsiao T. 1973. Plant responses to water stress. Annual Review ofPlant Physiology 24, 519–570.

Huguet J-G, Li SH, Lorendeau JY, Pellous G. 1992. Specificmicromorphometric reactions of fruit trees to water stress andirrigation scheduling automation. Journal of Horticultural Science67, 631–40.

Idso SB, Jackson RD, Pinter PJ Jr, Reginato RJ, Hatfield JL.1981. Normalizing the stress-degree-day parameter for environ-mental variability. Agricultural Meteorology 24, 45–55.

Jackson RD. 1982. Canopy temperature and crop water stress.Advances in Irrigation Research 1, 43–85.

Jackson RD, Idso SB, Reginato RJ, Pinter PJ Jr. 1981. Canopytemperature as a drought stress indicator. Water Resources Re-search 17, 1133–1138.

Jackson RD, Reginato RJ, Idso SB. 1977. Wheat canopy temper-ature: a practical tool for evaluating water requirements. WaterResources Research 13, 651–656.

Jones HG. 1973. Estimation of plant water status with the beta-gauge. Agricultural Meteorology 11, 345–355.

Jones HG. 1983. Estimation of an effective soil water potential at theroot surface of transpiring plants. Plant, Cell and Environment 6,671–674.

Jones HG. 1985. Physiological mechanisms involved in the controlof leaf water status: implications for the estimation of tree waterstatus. Acta Horticulturae 171, 291–296.

Jones HG. 1990a. Plant water relations and implications forirrigation scheduling. Acta Horticulturae 278, 67–76.

Jones HG. 1990b. Physiological aspects of the control of water statusin horticultural crops. HortScience 25, 19–26.

Jones HG. 1999. Use of infrared thermometry for estimation ofstomatal conductance in irrigation scheduling. Agricultural andForest Meteorology 95, 135–149.

Jones HG. 2004. Application of thermal imaging and infraredsensing in plant physiology and ecophysiology. Advances inBotanical Research (in press).


at University of T



ownloaded from


Jones HG, Leinonen I. 2003. Thermal imaging for the study of plantwater relations. Journal of Agricultural Meteorology 59, 205–214.

Jones HG, Pena J. 1986. Relationships between water stress andultrasound emission in apple (Malus 3 domestica Borkh.).Journal of Experimental Botany 37, 1245–1254.

Jones HG, Stoll M, Santos T, de Sousa C, Chaves MM, GrantOM. 2002. Use of infrared thermography for monitoring stomatalclosure in the field: application to grapevine. Journal of Experi-mental Botany 53, 2249–2260.

Kacira M, Ling PP. 2001. Design and development of an automatednon-contact sensing system for continuous monitoring of planthealth and growth. Transactions of the American Society ofAgricultural Engineers 44, 989–996.

Kacira M, Ling PP, Short TH. 2002. Establishing crop water stressindex (CWSI) threshold values for early, non-contact detection ofplant water stress. Transactions of the American Society ofAgricultural Engineers 45, 775–780.

Kang SZ, Davies WJ, Shan L, Cai HJ (eds). 2003. Water-savingagriculture and sustainable use of water and land resources. Xian,PR China: Shaanxi Science and Technology Press.

Kang SZ, Liang ZS, Pan YH, Shi PZ, Zhang JH. 2000. Alternatefurrow irrigation for maize production in an arid area. AgriculturalWater Management 45, 267–274.

Leinonen I, Jones HG. 2004. Combining thermal and visibleimagery for estimating canopy temperature and identifying plantstress. Journal of Experimental Botany 55, 1423–1431.

Li SH, Huguet JG. 1990. Controlling water status of plants andscheduling irrigation by the micromorphometric method for fruittrees. Acta Horticulturae 278, 333–341.

Lorenzo-Minguez P, Ceulemans RJM, Gabriels R, Impens I,Verdonck O. 1985. Effect of short-term water stress on leaf gasexchange processes and plant water potential in Schefflera.HortScience 20, 932–934.

Mahan JR, Burke JJ, Upchurch DR, Wanjura DF. 2000.Irrigation scheduling using biologically-based optimal temperatureand continuous monitoring of canopy temperature. Acta Horticul-turae 537, 375–381.

Malone M. 1993. Rapid inhibition of leaf growth by root cooling inwheat: kinetics and mechanism. Journal of Experimental Botany44, 1663–1669.

Martin DL, Wehner DJ, Throssell CS. 1994. Models for predictingthe lower limit of the canopy-air temperature difference of two coolseason grasses. Crop Science 34, 192–198.

Massacci A, Jones HG. 1990. Use of simultaneous analysis of gas-exchange and chlorophyll fluorescence quenching for analysingthe effects of water stress on photosynthesis in apple leaves. Trees:Structure and Function 4, 1–8.

McBurney TA. 1992. The relationship between leaf thickness andplant water potential. Journal of Experimental Botany 43, 327–335.

McCutchan H, Shackel KA. 1992. Stem water potential asa sensitive indicator of water stress in prune trees (Prunusdomestica L. cv. French). Journal of the American Society forHorticultural Research 117, 606–611.

Milburn JA. 1979. Water flow in plants. London: Longmans.Moran MS, Clarke TR, Inoue Y, Vidal A. 1994. Estimating cropwater deficit using the relation between surface–air temperatureand spectral vegetation index. Remote Sensing of Environment 49,246–263.

Moriana A, Fereres E. 2002. Plant indicators for schedulingirrigation of young olive trees. Irrigation Science 21, 83–90.

Peretz J, Evans RG, Proebsting EL. 1984. Leaf water potentials formanagement of high frequency irrigation on apples. Transactionsof the American Society of Agricultural Engineers 27, 437–442.

Plant RE, Horrocks RD, Grimes DW, Zelinski LJ. 1992. CALEX/Cotton: an integrated expert system application for irrigation

scheduling. Transactions of the American Society of AgriculturalEngineers 35, 1833–1838.

Raschke K. 1960. Heat transfer between the plant and the environ-ment. Annual Review of Plant Physiology 11, 111–126.

Remorini D, Massai R. 2003. Comparison of water status indicatorsfor young peach trees. Irrigation Science 22, 39–46.

Sadler EJ, Camp CR, Evans DE, Millen JA. 2002. Corn canopytemperatures measured with a moving infrared thermometer array.Transactions of the American Society of Agricultural Engineers45, 581–591.

Sandford AP, Grace J. 1985. The measurement and interpretation ofultrasound from woody stems. Journal of Experimental Botany 36,298–311.

Schmidt U, Exarchou E. 2000. Controlling of irrigation systems ofgreenhouse plants by using measured transpiration sum. ActaHorticulturae 537, 487–494.

Schultz HR. 2003. Differences in hydraulic architecture account fornear-isohydric and anisohydric behaviour of two field-grown Vitisvinifera L. cultivars during drought. Plant, Cell and Environment26, 1393–1405.

Selles G, Berger A. 1990. Physiological indicators of plant waterstatus as criteria for irrigation scheduling. Acta Horticulturae 286,87–100.

Simonneau T, Habib R. 1991. The use of tree suckers to estimateroot water potential. Plant, Cell and Environment 14, 585–591.

Slatyer RO. 1967. Plant water relationships. London and NewYork: Academic Press.

Smith KA, Mullins CE (eds). 2000. Soil and environmentalanalysis: physical methods, 2nd edn. New York: Marcel Dekker.

Steudle E. 2001. The cohesion-tension mechanism and the acquisi-tion of water by plant roots. Annual Review of Plant Physiologyand Plant Molecular Biology 52, 847–875.

Stocker O. 1956. Die Abhangigkeit des transpiration von denunweltfaktoren. In: Ruhland W, ed. Encyclopedia of plant physi-ology, Vol. 3. Berlin: Springer-Verlag.

Stockle CO, Dugas WA. 1992. Evaluating temperature-based in-dices for irrigation scheduling. Irrigation Science 13, 31–37.

Stoll M, Dry P, Loveys B, Stewart D, McCarthy M. 2000b. Partialrootzone drying: effects on root distribution and commercialapplication of a new irrigation technique. Australian and NewZealand Wine Industry Journal 15, 74–76.

Stoll M, Loveys B, Dry P. 2000a. Hormonal changes induced bypartial rootzone drying of irrigated grapevine. Journal of Exper-imental Botany 51, 1627–1634.

Tardieu F, Simonneau T. 1998. Variability among species ofstomatal control under fluctuating soil water status and evaporativedemand: modelling isohydric and anisohydric behaviours. Journalof Experimental Botany 49, 419–432.

Tyree MT, Dixon MA. 1983. Cavitation events in Thuja occidenta-lis L.? Ultrasonic emissions from the sapwood can be measured.Plant Physiology 66, 397–405.

Tyree MT, Sperry JS. 1989. Vulnerability of xylem to cavitationand embolism. Annual Review of Plant Physiology 40, 19–38.

Upchurch DR, Wanjara DF, Mahan JR. 1990. Automating trickle-irrigation using continuous canopy temperature measurements.Acta Horticulturae 278, 299–308.

Wronski EB, Homes JW, Turner NC. 1985. Phase and amplituderelations between transpiration, water potential and stem shrink-age. Plant, Cell and Environment 8, 613–622.

Yang SF, Qian DP, Huo XJ, Okushima L, Sase S. 2003.Computerized precision irrigation system based on estimation ofcrop water stress with acoustic emission technique. In: Kang SZ,Davies WJ, Shan L, Cai HJ, eds. Water saving agriculture andsustainable use of water and land resources. Xian, PR China:Shaanxi Science and Technology Press, 480–483.

2436 Jones

at University of T



ownloaded from


irrigation scheduling: advantages and pitfalls of plant-based methods

Documents