soil water components based on capacitance probes in a sandy soil

Soil Water Components Based on Capacitance Probes in a Sandy SoilA. Fares and A. K. Alva*

ABSTRACTUnderstanding soil water movement is needed to manage irrigation

to minimize water drainage, nutrient leaching below the root zone,and contamination of groundwater. We hypothesized that soil watercontent determined by capacitance probes can be used for irrigationscheduling and estimating soil water components. Objectives of thisstudy were (i) to evaluate the performance of capacitance probes foroptimizing irrigation management for 'Hamlin' orange trees [Citrussinensis (L.) Osb.] on Swingle citrumelo [Citrus paradisi Macf. XPoncirus trifoliata (L.) Raf.] rootstock on a Candler fine sand soil(hyperthermic, uncoated, Typic Quartzipsamment) in Central Floridaand (ii) to determine soil water balance components. Irrigation levelswere determined based on available soil water (ASW) and tree growthstage. The soil water data measured at finite time interval by capaci-tance probes were used with irrigation and rainfall data to calculatedaily evapotranspiration (ET) and drainage rates. Daily ET ratesshowed strong seasonal patterns and varied from 0.4 mm d ' inJanuary to 5 mm d ' in July and August. The annual ET in 1997 was920 mm or 53% of the total water input (irrigation and rainfall). Thecumulative annual drainage in 1997 was 890 mm, or 47% of the totalwater input. Furthermore, 82% of the cumulative annual drainagewas contributed by rainfall. Irrigation based on monitoring soil watercontent using capacitance probes minimized water drainage belowthe root zone in a system where rainfall contributed substantiallyto drainage.

MAINTENANCE OF ADEQUATE soil water contentthrough most of the crop growing period is neces-

sary to support optimum plant growth and yields. Inmost growing regions, soil water is at optimum levelonly for a short portion of the growing season; hence,irrigation is needed to maintain adequate soil water

A. Fares, University of Florida, IFAS, Citrus Research and EducationCenter, 700 Experiment Station Rd., Lake Alfred, FL 33850; andA.K. Alva, USDA-ARS-PWA, 24106 N. Bunn Rd., Prosser, WA,99350. Contribution of the Citrus Research and Education Center.Florida Agricultural Experiment Station Journal Series no. R-06632.Received 1 Oct. 1998. *Corresponding author ([email protected]).

Published in Soil Sci. Soc. Am. J. 64:311-318 (2000).

availability to support optimal production and quality.The purpose of well-managed irrigation is to optimizewater spatial and temporal distribution, to promote cropgrowth and yield, and to enhance citrus economic re-turns. Hence, the aim is not necessarily to obtain thehighest yields per unit area of land, or per unit volumeof water, but to maximize net returns.

When water is applied on the soil surface, assuminglittle or no surface runoff, a portion of water is utilizedby the plants, or retained in the soil, and the excess waterdrains through the vadose zone into the groundwater,which contributes to aquifer recharge. This excess watermay contain agricultural chemicals and soluble nutri-ents. Irrigation best management practices (BMP) aredesigned to (i) minimize water and nutrient leachingbelow the root zone to maintain adequate irrigationwater within the rooting depth, (ii) minimize non-pointsource pollution of groundwater, and (iii) reduce pro-duction costs associated with water and nutrient lossesby leaching.

Supply of water to crops must be based on a clearunderstanding of the soil water dynamics. The watercycle of an agricultural field comprises evapotranspira-tion, irrigation, rainfall, runoff, and drainage losses be-low the root zone. Under-tree sprinklers and drip irriga-tion systems are designed to deliver water at rates lowenough so that the soils can retain the water withoutcontributing to losses by runoff or excessive downwarddrainage through the soil.

Using these low-volume irrigation systems, it is possi-ble to schedule small but frequent irrigation events tomaintain an optimum water content in the root zonewhile decreasing water losses below this depth. Undersuch conditions, it is possible to minimize the waterdrainage below the root zone. This is in marked contrast

Abbreviations: ASW, available soil water; BMP, best managementpractices; ET, evapotranspiration; ET0, daily potential evapotranspi-ration.

312 SOIL SCI. SOC. AM. J., VOL. 64, JANUARY-FEBRUARY 2000

with high drainage rates following low-frequency andhigh-volume irrigation events (Hillel, 1990).

Given the close relationship between excess irrigationand water losses, an understanding of the dynamic com-ponents of the soil water balance is important to manageirrigation properly. A water budget requires the ac-counting of all water input and output components, in-cluding rainfall, irrigation, drainage below the root zone,runoff, evapotranspiration, and changes in soil waterstorage. One-dimensional analytical models (Alva et al.,1999; Darusman et al., 1997; Chopart and Vauclin, 1990)and one dimensional numerical models (Lotse et al.,1992; Jabro et al., 1995) were used to estimate fieldwater drainage fluxes below the root zone.

The objectives of this study were (i) to evaluate theuse of capacitance probes to optimize citrus irrigationand (ii) to use collected data to estimate the componentsof soil water balance under young citrus tree growingconditions.

MATERIALS AND METHODSThe experiment was conducted in a research farm of the

University of Florida Citrus Research and Education Centerat Lake Alfred (280°01' R, 81°055' W), Polk county, Florida,using a 4-yr-oId citrus grove of Hamlin orange trees on Swinglecitrumelo rootstock planted (at 7.6 by 4.6 m spacing) in aCandler fine sand in September, 1993. This soil is a well-drained sandy soil with no restricting soil layer and, therefore,a deep water table that will not influence the hydrology ofthis system. The trees were irrigated with under-tree low-volume sprinklers (80% efficiency) using one emitter per tree,with a delivery rate of 50 L h"1 covering an area of 7.3 m2,which resulted in an irrigation rate of 7 mm h"1 (=20% of thetotal grove area). The soil water release curves for the differentsoil horizons were presented by Alva et al. (1999). The meanbulk density values were 1.59,1.52,1.51,1.61, and 1.55 g cm"3,and saturated hydraulic conductivities were 5.21, 9.48, 8.52,7.27, and 6.96 m d"1, respectively, for the soil horizons at 0-to 20-, 20- to 50-, 50- to 100-, 100- to 130-, and 130- to 200-cm depths (Alva et al., 1999). The rainfall events were mea-sured using two rain gages placed in two separate locationson the grove; however, irrigation rates were determined basedon timing and flow meter readings. The rainfall and irrigationevents during 1997 are shown in Fig. 1. The annual rainfallfor 1997 was 1470 mm, which was =22% greater than the long-term mean annual rainfall of 1200 mm for this location. Arecommended fertilizer program (Tucker et al., 1995) wasfollowed. In addition, a standard herbicide program (Knapp,1994) was followed to maintain a 210-cm-wide herbicide bandon both sides of the trees. The herbicide band was maintainedweed-free throughout the year.

A Real-Time Soil Water Monitoring SystemThe soil water content through the soil profile was continu-

ously monitored using three EnviroSCAN capacitance probes(EnviroSCAN, Sentek PTY Ltd., South Australia).1 Sensorsfor each of the probes were installed within the emitter wettingarea under canopy along the tree drip line of randomly selectedtrees at 10-, 20-, 40-, 70-, and 110-cm depths. These probeswere connected to a solar-powered data logger, and soil water

readings were recorded at 10-min intervals. The first threedepths represent the depth of rooting (A.K. Alva, 1999, un-published data), while the last two depths represent the soilprofile below the rooting depth. A schematic of the locationof a probe is shown in Fig. 2. Each probe was installed underthe tree drip line 50 to 60 cm away from the low-volumesprinkler to assure a good irrigation coverage of the sensors.

EnviroSCAN systems have been extensively tested underlaboratory (Mead et al., 1995; Paltineanu and Starr, 1997) andfield conditions (Buss, 1993; Starr and Paltineanu, 1998). Theyhave been used for soil water content monitoring and irriga-tion scheduling of different crops in the USA (Fares et al.,1998; Starr and Paltineanu, 1998) and in Australia (Buss,1993). Soil water content data were recorded on a data loggerlocated up to 500 m away from the access tube (Fig. 3).

The capacitance method measures the apparent dielectricconstant of the soil surrounding the sensor, which reflects thewater content of the soil-water-air mixture. The dielectricconstant of a medium depends on the polarization of its mole-cules in an electric field. Because the dielectric constant ofwater (80) is large compared with that of the soil matrix (<10)or air (1), small changes in soil water content strongly influencethe dielectric constant of the soil-water-air mixture. This rela-tionship between water content change and dielectric constantof the medium depends on soil type and frequency range ofthe measuring apparatus. Because of their high frequencyrange (=150 Mhz), measurements by EnviroSCAN sensorsare relatively insensitive to changes in fertilizer salts overtypical concentrations used in agricultural cropping systems(Gardner et al., 1991; Paltineanu and Starr, 1997). Paltineanuand Starr (1997) studied the response of the capacitance sen-sors subjected to a wide range of air and water temperatures.They found that errors due to temperature variations between10 to 30°C were less than the root mean square error fortheir calibration curve. Measured frequencies are convertedto water content using the calibration curve supplied by themanufacturer. These sensors are capable of measuring volu-metric water content values ranging from a saturated soil toalmost oven-dry soil with a resolution of 0.1% (Buss, 1993).

Irrigation SchedulingEnviroSCAN water content readings through the soil pro-

file were downloaded every other day to schedule irrigationor process data. The target refill points for optimal irrigationscheduling were based on an evaluation of the allowable soilwater depletion within the rooting depth of 40 cm. The ASWwas determined as:

ASW = 6FC - 6P [1]

1 Brand name and distributor are given to provide specific informa-tion and do not constitute endorsement of the product by the authorsor by the University of Florida

where 9FC and 9PWP are the volumetric water content (cm3

cm"3) at field capacity and at the permanent wilting point(equivalent to soil matric potential of -1.5 MPa), respectively.The field capacity term was determined based on the workof Fares and Alva (2000). They found that Candler fine sandsoil holds between 0.08 to 0.10 (cm3 cm"3) of water 1 d aftersaturation, with little or no surface evaporation. Thus, thevolumetric water content at field capacity was taken as 0.09(cm3 cm"3). The permanent wilting point (0.015 cm3 cm"3)was selected based on laboratory measurements (Sodek et al.,1990). Therefore, the ASW is 0.075 (cm3 cnT3). The ASW perunit area for the top 40 cm, which is the effective rootingdepth, is 0.075 by 40 cm, or 3 cm.

Optimal citrus production requires maintaining soil watercontent above the 33% depletion of the ASW during theperiod from February to May to avoid potential adverse effectsof water stress on flowering and fruit set (Koo, 1969). How-

FARES & ALVA: USE OF CAPACITANCE PROBES TO DETERMINE CITRUS EVAPOTRANSPIRATION 313

Rainfall and Irrigation Events for 1997

_g>

70

60

50

40

30

20

10

0

50

40

.± 30

20

10

0

Rain

B Irrigation

0 30 60 90 120 150 180 210 240 270 300 330 360Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Calendar DayFig. 1. Daily (A) rainfall and (B) irrigation events during 1997 for the study site.

ever, during the remaining part of the growing season, ASWcan be allowed to deplete by 67% before replenishment ofthe soil water back to field capacity. Depletion of 33 and67% of the ASW to a soil depth of 40 cm corresponds withequivalent water content of 2.6 and 1.6 cm, respectively. Thegoal of each irrigation event was to deliver an adequateamount of water to replenish the deficit in the top 40 cm tofield capacity. This target point is defined as the full point,which is equivalent to 3.6 cm (40 cm x 0.09 = 3.6 cm) forthe target depth of irrigation (40 cm). Thus, irrigations wereapplied when the total water content in the root zone reached2.6 and 1.6 cm during the first period from February to Mayand from June to January, respectively.

Citrus tree

Sprinkler coveragearea

Data Processing Using a Utility ProgramA utility program, SENTPROC, was developed which con-

verts the water content data stored in the data logger intospreadsheet format. SENTPROC was coded using the FOR-TRAN programming language (a copy of the executable fileof SENTPROC can be obtained from the authors). Using aspreadsheet and knowing the depth of the root zone, the usercan calculate the soil water content for the entire soil depthof probe installation and also for within and below the rootzone depths (Fig. 4). These data are useful for determiningirrigation scheduling and for calculating the different waterbalance components (i.e., evapotranspiration and drainage be-low the root zone). Data downloaded from the data loggerwere processed using SENTPROC.

Calculation of Drainage and EvapotranspirationThe field water balance of a citrus grove grown on a sandy

soil can be defined based on the conservation of mass asfollows:

AS = R + I - D - ET [2]

Solar Panel

Fig. 2. Schematic of the location of sprinkler and the EnviroSCANcapacitance probe under the tree canopy.

no*

Fig. 3. Schematic of multiple EnviroSCAN capacitance probes layoutin the field.


14

12

g lO

¥ 8S§ •O<5 14*j

*«O5 100)I •E 50£45§,40

S »^^

(O 30jjj 25

5 20

70cm110cm

r/-

— •• Rootzone— — Below rootzone

- — -__X

— ..J

02-11 02-13 02-15 02-17Time (February 10-18)

Fig. 4. A sample data showing (A and B) soil water content at eachdepth and (C) depth-integrated water content within the root zone(0-40 cm) and below the root zone (40-110 cm) of 4-yr-old Hamlinorange trees on a Swingle citrumelo rootstock in a Candler finesand.

where, AS is the change in water storage in the root zone, Ris the rainfall, / is the irrigation, D is the drainage at thebottom of the root zone, and ET is the evapotranspirationduring a given period of time Af. All of these variables areexpressed as millimeters. Runoff component was not consid-ered in Eq. [2] since most of Florida sandy soils have highhydraulic conductivities and very little slope, resulting in negli-gible runoff. Irrigation and rainfall were directly measured byrain gages and flow meters, respectively. Water storage withinand below the root zone was calculated using the soil watermeasurements by the capacitance probes (Eq. [6]). The re-maining unknown terms are D and ET (Fig. 5).

Drainage below the root zone was estimated using Darcy'sequation since the water content at the bottom and below therooting zone are known at short time intervals. The water fluxacross the plane at depth z can be computed using Darcy'sflow equation:

D =Az

Ar P]

where, /c(6) (cm d"1) is the effective unsaturated hydraulicconductivity at the water content 6 (cm3 cm"3) of the soil layerbelow the rooting zone; Af is the time for which the drainagewas computed (1 h was used); A/z (cm) is the pressure headgradient between the bottom of the rooting zone depth andnext depth in the profile where the water content is monitored;Az (30 cm) is the distance between the bottom of the rootingzone (40 cm) and the next depth in the soil profile where thewater content is known (70 cm).

van Genuchten (1980) derived an analytical expression that

Fig. 5. Flow chart explaining the steps taken during the calculationof ET, drainage, and storage variations through the soil profile.

relates the unsaturated hydraulic conductivity to the watercontent. The relationship between the unsaturated hydraulicconductivity and the water content is:

where, Se = (0 - 6r)/(9s - 0r), 9S (cm3 cm"3) is the water contentat saturation; 6r (cm3 cm"3) is the residual water content; 6(cm3 cm"3) is the water content at which k is being calculated;m is a fitting parameter; and fcs (cm d"1) is the saturatedhydraulic conductivity.

Knowing the water content at a given location of the soilprofile, the pressure head at that same location can be esti-mated by the following equation (van Genuchten, 1980):

|*|= ' «<.a \0S - 9r,

1 [5]

where, a (cm"1) and n are fitting parameters.The variation in soil water storage (AS) between two depths

(zi = 0 cm and z2 = 40 cm) for a given period of time (Af =fi - f2; i.e., 1 h was used) was calculated based on measuredwater content readings by the capacitance probes using thefollowing equation:

e(z,fe)dz [6]Variation in water content below the root zone is due to

water redistribution into this depth from the soil profile aboveit. Thus, any net increase in the water content in the soilhorizon immediately below the root zone depth can be usedas a first approximation of water losses below the root zoneof the soil profile.

RESULTS AND DISCUSSIONSoil Water Storage

Volumetric soil water contents measured using thecapacitance probes were used to calculate the watercontent for the entire profile. Figure 6 shows the meanwater content (of three capacitance probes randomlyplaced in the field) for the 0- to 40-, 40- to 110-cm, and0- to 110-cm depth profile. Irrigation was scheduledbased on the recommended refill points depending onthe tree growth stages. Accordingly, the soil water con-tent in the root zone was maintained within 33% deple-tion of the ASW during the critical period from Febru-ary to June and at 67% depletion of the ASW duringthe rest of the year. From Fig. 6 it is evident that the


water content in the root zone was maintained withinthe target refill points during the entire growing season,except for few occasions during February to June whenthe water content in the root zone dropped slightlybelow the refill point just before the subsequent irriga-tion event. The sharp increase and decrease in the watercontent following each irrigation or rain event clearlydemonstrates the low water-holding capacity and highhydraulic conductivity of this soil. Fares and Alva (2000)reported that water content in this Candler fine sandsoil decreased to =0.09 cm3 cm"3 within a day aftersaturation as a result of a large irrigation or rainfall.

The increase in water content in the soil immediatelybelow the root zone is due to water drainage from theroot zone. During the first 90 d of the year, the watercontent below the root zone varied around 40 mm, ex-cept on 19 January when it increased to 60 mm as aresult of 42 mm of irrigation applied as a precautionarymeasure to protect against forecasted freezing tempera-tures. In April, the water content below the root zoneshowed large increases in response to heavy rainfallevents. This was due to a net total rainfall of 174 mm,which was nearly fivefold greater than the long-termaverage. May and June were drier than the average.This explains the decrease in the water content belowthe root zone as compared with its level during April.The remaining portion of the year was in the rainyseason, and an unusually wet fall resulted in excess waterdrained below the root zone.

DrainageThe daily drainage varied considerably through the

study period (Fig. 7). The mean drainage values wereas high as 47 mm following a 58-mm rainfall event on28 May. In general, the daily drainage rates were lowduring January to March, which represents the drymonths when the optimal irrigation is critical for thetrees. The low drainage below the root zone duringthis dry period demonstrates that irrigation schedulingbased on near-continuous monitoring of the water con-tent within the root zone minimized drainage. All ofthe irrigation events were managed to add only enoughwater to replenish the soil to the full point. The rainfallin April 1997 was fivefold greater than the long-termaverage and accordingly resulted in several large drain-age events as shown in Fig. 7. Greater than 40% of theannual cumulative drainage occurred during Octoberto December (Fig. 7), which were unusually wet months.

In 1997, the annual cumulative drainage below theroot zone was 890 mm (Fig. 7) from the 1470 mm ofrainfall plus 340 mm irrigation. The 890 value is veryclose to the upper limit of the range (i.e., 625 and 860mm) reported by Rogers and Bartholic (1976) for acitrus grove in Central Florida. They calculated the an-nual water balance based on daily weather data andwater storage in three different soils using irrigationinput that is 50% lower than that used in our work.

Daily EvapotranspirationThe daily ET values calculated based on Eq. [2] for

each calender date for 1997 are shown in Fig. 8. The

—— Water content— - - Refill Point II (67% depletion of ASM)— - Refill Point I (33% depletion of ASM)


Calendar Days (1997)Fig. 6. Depth-integrated soil water content (A) within and (B) below

the root zone and for (C) the entire monitored soil profile depthduring 1997. The refill points indicate the soil water content atwhich irrigation was scheduled to replenish the water deficit. Theallowable depletion of the available soil water (ASW) was 33% inFebruary to June, and 67% during the rest of the year.

seasonal variations in daily evapotranspiration show anorder of magnitude difference during the growing sea-son (i.e., <0.4 mm d~' on some days in January toalmost 5 mm d"1 on some days in June through August).Weekly citrus ET estimated by Rogers and Bartholic(1976) using a modified Blaney-Criddle procedure fol-lowed a similar seasonal pattern. Their data were foran average mature citrus grove and under an averageyearly rainfall (1335 mm) and irrigation (304 mm). Theiraverage daily ET values based on a weekly estimationwere as low as 0.3 mm d"1 during the winter months(December-February) and as high as 4.2 mm d"1 duringthe summer months (June-August).

Daily evapotranspiration based on total monthly val-ues from a citrus grove with grass understory cover inpoorly drained soils in the East coast of Florida rangedbetween 1.9 and 5.0 mm d"1 (Rogers et al., 1983). Thesedaily averages based on monthly totals mask the largevariability of daily ET in every month as is shown inour work (Fig. 8) and in the data by Rogers and Bar-tholic (1976) and Sumner (1996).

Daily pan evaporation data measured at a weatherstation in the vicinity of this field experiment site wereused with a crop coefficient 0.55 to estimate daily poten-tial evapotranspiration (ETP). The above crop coeffi-cient value was adapted from data reported by Rogerset al. (1983). Their work showed that the crop coefficientof young citrus trees increased with tree growth, from0.51 for 1-yr-old trees to 0.65 for 8-yr-old trees. In our


45

40

~. 35E.§, 300)

_cS 20•o

1 1510

5

0

4

fl) 1

Q 3

2 2OT 1

0

-

-

_

Daily drainage

ill1 1lUwt \r,i1

— '

I J

J

,l 1

--,-

,

s -^

J

_Cumulative drainage y'' -

^j-

\

/

X

-

^

-

i-r.< . i . . . . i . . . . i . . . . i . . ' . . i . . . . i . . . . i . . . . i . . . . i . . . . . . . . i . . . . , .

_

,1.1 JLi.J. 1 11 . , .k I, ii. 1 1, ui1 , „ , ,,u

1 • '

j '

_

llilll j j

-

900^

750 £0)O)

600 jjj

2450 '

.>

300 *5E3

150 °

0

4

3

2

1

00 30 60 90 120 150 180 210 240 270 300 330 360

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov DecCalendar Days

Fig. 7. Calculated daily mean, standard deviation, and cumulative mean values of water drainage below the root zone of 4-yr-old Hamlin orangetrees on Swingle citrumelo rootstock grown in a Candler fine sand during 1997.

study the trees were 4 yr old. The data in Fig. 9 showa strong linear relationship (r2 = 0.91) between (ETP)and ET with some degree of variability (standard errorcoefficient = 0.03) below and above the 1:1 regressionline. This may be due, in part, to spatial variability inthe soil water content, which consequently could influ-

ence the calculated ET and drainage data. Consideringsoil variability, some variation in soil properties is ex-pected even within an area as small as 1 m2 of a relativelyhomogeneous soil profile (Gardner, 1986). Taylor (1955)reported coefficients of variation of 17 and 20% forgravimetric soil water content of samples from soil from

5 h

I4

o1 3Q.(A

i,a 2§UJ

nQ 1

r- 1000Cumulative ET

Mean ET

«**AS^^

800EE,

o600 2

'5.

400

- 200

- o

Q.reUJ»

i3EO


Calendar Days (1997)Fig. 8. Calculated daily mean, standard deviation, and cumulative values of evapotranspiration for 4-yr-old Hamlin orange trees on Swingle

citrumelo rootstock grown in a Candler fine sand during 1997.


E

I4

S'5.£

I 3a3UJ>, 2'5a

Y = 0.987 Xr2 = 0.91SEC = 0.03

.siBa-Tv'.1 2 3 4 5

Daily Potential Evapotranspiration (mm)Fig. 9. Linear regression analysis between calculated daily evapotranspiration and potential evapotranspiration for 4-yr-old Hamlin orange trees

on Swingle citrumelo rootstock grown in a Candler fine sand during 1997. SEC is standard error coefficient.

field plots under furrow and sprinkler irrigation meth-ods, respectively, for a relatively uniform Millville(coarse-silty, carbonatic, mesic Typic Haploxeroll)loam soil.

Yearly Cumulative EvapotranspirationThe cumulative ET in 1997 for 4-yr-old citrus trees

was 920 mm (Fig. 8). The annual ET value from thisstudy is within the range of values reported for citrusin a deep sandy, well-drained soil in central Florida(Koo and Harrison, 1965; Smajstrla et al., 1986) orpoorly drained soils with a perched water table in theEast coast of Florida (Rogers and Bartholic, 1976; Rog-ers et al., 1983), in Texas (Wiegand and Swanson, 1982),and in Arizona (Erie et al., 1965). The ET value in thisstudy is 140 mm lower than that reported by Rogers etal. (1983), 200 mm lower than that of Koo (1963), and520 mm lower than that of Hoffman et al. (1982) formature citrus in Arizona. This high ET data for citrusin Arizona compared with Florida may be due to thegreater evaporative demand (potential evapotranspira-tion) in Arizona compared with that of Florida. Thegreater ET reported by Koo (1963) compared with ourstudy can be attributed primarily to the difference intree age (i.e., 4-yr-old trees in our study vs. mature treesin Koo, 1963) and variation in climatic conditions duringdifferent years. Rogers et al. (1983) reported a linearrelationship between citrus tree age and annual ET anddemonstrated a 13% increase in ET for the 8-yr durationof their study.

Estimated annual ET for a deforested area in theLake Wales Ridge, FL (Sumner, 1996) reached 680 mm.The low ET values reported by Sumner (1996) com-pared with the ET for commercial citrus in this study

can be attributed to the lack of irrigation and shallow-rooted natural vegetation. Irrigation in commercial cit-rus groves produces a wet soil, and more water will beavailable for plant uptake, which causes a higher actualET than the nonirrigated native vegetation. Hoffmanet al. (1982) reported a 48% greater annual citrus ETfor a frequently irrigated grove compared with that fora less frequently irrigated grove used by Erie et al.(1965). Actual ET is strongly related to evaporativedemand (potential ET) and ASW for plant uptake. Con-sequently, a given crop species with a given potentialET has a greater ET in a wet soil than in a dry soil.

SUMMARY AND CONCLUSIONSResults of this experiment demonstrated that moni-

toring of soil water using capacitance probes can beused to optimize irrigation scheduling for citrus groveson a sandy soil. Given the knowledge of soil water char-acteristic curves, effective rooting depth, and recom-mended depletion of ASW content depending on thecrop growth stages, the root zone soil water can bereplenished to its optimum level while minimizing drain-age and avoiding plant stress. The data provided bythe capacitance probes can be used to determine thecomponents of citrus water balance. The water balancemethod was used to determine the daily evapotranspira-tion. Cumulative annual evapotranspiration and drain-age were 920 and 890 mm, respectively. Most of thedrainage occurred during the summer months and theunusually wet fall. Daily evapotranspiration varied sea-sonally, ranging from 0.4 mm d"1 in January to =5.0mm d~' in July and August. These daily ET values werewell within the daily ET values reported for citrus treesin Florida.


ACKNOWLEDGMENTSThis study was made possible by partial funding from the

Florida Citrus Production Research Advisory Council.

soil water components based on capacitance probes in a sandy soil

Documents