SUBSURFACE DRIP IRRIGATION OF LEAF LETTUCE
AND BROCCOLI II. WATER BALANCE
Thomas L. Thompson and Kepi L. Maki
Abstract
The objective of this research was to estimate a season -long water balance under one
subsurface trickle- irrigated plot each of lettuce (Lactuca sativa L. var. Waldmann's Green) and
broccoli (Brassica olearacea L. var. Claudia). One lettuce plot during 1992 -93 and one
broccoli plot during 1993 -94 were intensively instrumented with automated tensiometers.
Tensiometer readings and estimates of evapotranspiration were used to estimate seasonal water
contents in the crop root zone, and water losses due to leaching. For the monitored portion of
the 1992 -3 growing season, 19.1 an of irrigation water was applied, 12.5 an of rainfall fell,
and ET, was 11.5 cm. Estimated deep percolation was 60% of total water applied (irrigation
plus rainfall). Leaching was periodic, and was mostly associated with rainfall events. During
the monitored portion of the 19934 season, 21.2 an of irrigation water were applied, 8.0 an
of rainfall fell, and ET, was 21.9 cm. Estimated deep percolation was 28% of total water
applied. Almost all of this leaching was associated with one major rainfall event. Water
stored in the root zone (top 50 cm) was relatively constant at 12 -14 an water /50 an soil except
after rainfall.
Introduction
The potential for greater efficiency of water and nitrogen (N) fertilizer application afforded by
subsurface drip irrigation (SDI) has led to its increasing use in recent years. Subsurface drip -irrigated acreage
in the U.S. has increased from 200,000 ha in 1985 (Hall, 1985) to nearly 1 million ha in 1993 (Anonymous,
1994). Subsurface drip irrigation offers significant savings of water over other irrigation methods by decreasing
the portion of the soil surface that is wetted and susceptible to evaporation (Bogle et al., 1989; Sammis, 1980).
Because small amounts of water can be applied by SDI with greater efficiency than other methods, considerable
savings in water may be achieved, particularly early in the season when crop water requirements are minimal
(Bernstein and Francois, 1973). Less moisture near the soil surface also limits the growth of weeds which
compete with crops for water and nutrients. Subsurface drip irrigation has an additional advantage over
traditional surface irrigation because it does not rely on the root zone as a reservoir to store irrigation water
(Phene and Beale, 1976). With SDI water is applied frequently, commonly on a daily basis. The frequent
water applications associated with SDI are necessitated by the development of large water and nutrient gradients
when irrigations are too infrequent. Such gradients are produced as a result of the relatively low volume of the
soil wetted in the use of trickle irrigation (Bar -Yosef et al., 1980). The frequent application of water with SDI
keeps soil moisture at a more constant level than most surface irrigation methods in which the soil is
periodically flooded. By maintaining constant soil moisture levels during the growing season, SDI may also
reduce plant water stress (Phene et al., 1982), and improve crop yields.
More information is needed on soil water status under SDI. Therefore, the objective of this research
was to estimate a season -long water balance under one subsurface trickle- irrigated plot each of lettuce (Lactuca
50
sauva L. var. Waldmann's Green) and broccoli (Brassica olearacea L. var. Claudia).
Materials and Methods
A complete description of the experimental methodology is given in the accompanying paper
(Thompson and Maki, 1995). A simple water balance equation was employed:
L = OD; - BD; +, + I + P - ET
(1)
where L is leaching (i.e., deep percolation) losses from the root zone, 8D; and BD; +I are the amounts of water
in the root zone at the beginning and end of the period, respectively, I is the amount of irrigation water applied,
P is precipitation, and ET, is actual crop evapotranspiration. The amount of irrigation water applied was
calculated from readings taken directly from meters attached to the main water lines. Evapotranspiration was
calculated as follows:
ET, = K, * ET.,
(2)
where KÇ is an empirically determined crop coefficient (dimensionless) and ET is the reference crop
evapotranspiration. The ET, values were provided by the Arizona Meteorological Network (AZMET) station at
the Maricopa Agricultural Center located about 100 m from the study site. Values for K, were estimated using
AZSCHED (Fox et al., 1992), an irrigation scheduling program which employs the soil water balance method
of irrigation scheduling. The heat unit accumulation limits used were 4.44 and 21.1 °C.
To estimate the amount of water in the root zone we determined the relationship between soil water
content ( 8) and SWT. Ten intact soil cores were removed from the monitored soil profile after the 1992 -3
harvest. The cores were collected at depths according to the soil horizon analysis, with depths corresponding to
the approximate vertical midpoint of each horizon. The hanging water column method (Klute, 1986) was used
to derive soil moisture release curves for each core. Data from both desorption and sorption experiments were
fitted to van Genuchten's equation (van Genuchten, 1980). We derived one h-8 relationship for the entire root
zone in order to facilitate conversion of SWT to 8, to be used in the water balance. To obtain one van
Genuchten curve for the entire root zone, a weighted average was calculated. SWT and 8, values from a given
core were weighted according to the percentage of the root zone occupied by the horizon they represented.
Contour plots of SWT throughout the profile were constructed for midnight every day of the growing
The
season. This time was chosen because it represents about the midpoint in time between two irrigations.
kriging method was used to generate the contours. The soil water tension within any contour interval was
assumed to be the average of the two contour limits, e.g., the portion of the root zone between 9.0 and 10.0
kPa was assumed to be 9.5 kPa. This average SWT value was then converted to 8, by use of the single
weighted - average van Genuchten equation discussed above. By multiplying the various 8 values by the
percentage of the root zone they occupied, a single 8 value for the portion of the profile within the root zone
was obtained. For both seasons the maximum rooting depth was about 50 cm. Therefore, the water balance
was calculated only for the top 50 cm of soil.
Results and Discussion
The estimated water balance for the 1992 -3 growing season is presented in Fig. 1. Water stored in the
root zone varied little during the season, despite the higher than average rainfall recieved during this season.
The large "spikes" in water applied were due entirely to rainfall events. It is evident from Fig. i that much of
the leaching was induced by rainfall events. Total leaching losses were estimated to be 60%. This is higher
than values from some earlier studies. For example, experiments involving furrow irrigation gave leaching
values of 25 % for corn (Linderman et al., 1976) and 9, 19, 22 and 33% for dry to wet treatments, respectively,
for soybean ( Carvallo et al., 1975). Leaching losses from sprinkler- irrigated orchardgrass were 24, 40 and 49
% for dry to wet treatments (Watts et al., 1991). Chopart and Vauclin (1990) reported leaching losses of 53,
47 and 46% for sprinkler- irrigated soybean plots of low to high sowing density, respectively. For the
monitored portion of our growing season, 34.4 cm of water were applied (12.5 cm was rainfall), while ET, was
51
only 11.5 cm. Leaching losses determined in this research were particularly high early in the season, when
crop evapotranspirative demand was significantly less than the amount of water required to maintain optimal
SWT levels.
Unfortunately, none of the studies reporting deep percolation losses discussed above utilized SDI or
irrigation scheduling according to optimal SWT. During 1992 -3, optimum lettuce yields occurred at 6 to 7 kPa
SWT, approximately the same as the target tension. This SWT level was also the level at which profitability
was maximized.
Estimated daily leaching losses generally declined as the growing season progressed. Undoubtedly the
magnitude of the estimated seasonal leaching loss was due in large part to the unusually heavy rains during the
1992 -3 winter growing season. Deep percolation values increased considerably during and after precipitation.
Removing data from rain -affected days (Dec. 18 and 27 -30, and Jan. 3-4 and 6 -20) results in 42% estimated
leaching losses for the whole season. This value is more representative of the actual deep percolation losses
due to irrigation only, independent of rainfall.
The estimated water balance for 1993 -4 is given in Fig. 2. This water balance does not cover the
entire monitored portion of the growing season because of a loss of data beginning on Dec. 28. As in 1992 -93,
soil water storage was constant (12 -13 cm/ 50 cm soil), except for one period of heavy rainfall (Julian days 314318). This period represents the only significant rainfall received during the season, and the only major
leaching event. Leaching for the entire monitored portion of the growing season was 28 %. This value includes
the effects of rain. The total depth of deep percolation was estimated to be 8.3 cm, whereas the depth of
rainfall during this same period was 8.1 cm and irrigation was 21.2 cm. The estimated ET, for this period was
21.9 cm. Excluding all rain - affected days (October 7 and November 11 -19), leaching losses were < 1%.
Therefore, the majority of leaching was likely rain- induced.
Conclusions
An in -situ soil water balance was calculated for subsurface drip irrigated leaf lettuce during 1992 -93
and broccoli during 1993 -94. Plots were irrigated to maintain an average soil water tension of about 7 kPa at
30 cm depth near the drip tubing. One plot during each year was intensively instrumented with automated
tensiometers. Tensiometer data and a nearby weather station were used to calculate changes in soil water
storage, evapotranspiration , and to estimate deep percolation below 50 cm depth. For the monitored portion of
the 1992 -3 growing season, 19.1 cm of irrigation water was applied, 12.5 cm of rainfall fell, and ET, was 11.5
cm. Estimated deep percolation was 60% of total water applied (irrigation plus rainfall). Leaching was
periodic, and was mostly associated with rainfall events. During the monitored portion of the 1993 -4 season,
21.2 cm of irrigation water were applied, 8.0 cm of rainfall fell, and ET, was 21.9 cm. Estimated deep
percolation was 28 % of total water applied. Almost all of this leaching was associated with one major rainfall
event. The target tension of 7 kPa prior to irrigation was closely matched both seasons.
Leaching from SDI systems will depend on climatic conditions and the crop grown. Rainfall events
will trigger deep percolation events, and crops with higher water use requirements will result in less leaching
overall. In general, leaching decreases later in the season when the crop water use is higher.
52
References
Anonymous. 1994. 1993 Irrigation Survey: Renewed growth results from weather and economy. Irrig. J.
44:24 -41.
Bar -Yosef, B., C Stammers and B. Sagiv. 1980. Growth of trickle irrigated tomato
volume and uptake of N and water. Agron. J. 72:815 -822.
as related to rooting
Bernstein, L. and L.E. Francois. 1973. Comparisons of drip, furrow, and sprinkler irrigation. Soil Sci. 115:7386.
Bogle, C.R., T.K. Hartz and C. Nuñez. 1989. Comparison of subsurface trickle and furrow irrigation on
plastic -mulched and bare soil for tomato production. J.Amer. Soc. Hort. Sci. 114:40 -43.
Carvallo, H.O., D.K. Cassel and A. Bauer. 1975. Water losses from an irrigated soybean field by deep
percolation and evapotranspiration. Water Resour. Res. 11:267.
Chopait, J.L. and M. Vauclin. 1990. Water balance estimation model: Field test and sensitivity analysis. Soil
Sci. Soc. Amer. J. 54:1377 -1384.
Fox, F.A., Jr., T. Scherer, D.C. Slack and L.J. Clark. 1992. Arizona Irrigation Scheduling (AZSCHED
Version 1.01). The University of Arizona Cooperative Extension Service, Agricultural and Biosystems
Engineering. January 1992.
Hall, B.J. 1985. History of Drip/Trickle Irrigation. p. 1 -7. In Proc. Third Intl. Drip/Trickle Irrig. Congr.,
Fresno, CA. 18 -21 Nov. 1985.
Klute, A. 1986. Water retention: Laboratory methods. p. 635 -662. In A. Klute (ed.). Methods of soil analysis.
Part 1. 2nd ed. Agron. Monogr. 9. ASA, Madison, WI.
Linderman, C.L., L.N. Mielke and G.E. Schuman. 1976. Deep percolation in a furrow- irrigated sandy soil.
Trans. ASAE 19:250 -253.
Phene, C.J. and O.W. Beale. 1976. High- frequency irrigation for water nutrient management in humid regions.
Soil Sci. Soc. Amer. J. 40:430 -436.
Phene, C.J., R.A. Radulovich, J.L. Rose and M.F. Blume. 1982. The effect of high- frequency trickle irrigation
on water stress of tomatoes (Lycopersicon esculentum). Papers of the ASAE Microfiche collection (fiche no. 822521).
Sanunis, T.W. 1980. Comparison of sprinkler, trickle subsurface and furrow irrigation methods for row crops.
Agron. J. 72:701 -704.
Thompson, T.L., and K.L. Maki. 1995. Subsurface drip irrigation of leaf lettuce and broccoli I. Spatial
distribution of roots and soil water tension. 1994 University of Arizona Vegetable Report.
van Genuchten, M.T. 1980. A closed -form equation for predicting the hydraulic conductivity of unsaturated
soils. Soil Sci. Soc. Amer. J. 44:892 -898.
Watts, D.G., G.W. Hergert and J.T. Nichols. 1991. Nitrogen leaching losses from irrigated orchardgrass on
sandy soils. J. Environ. Qual. 20:355 -362.
53
auoz
N
ui paamS (wo) INLrn
O
(wo) Jawn jo tpdaQ
54
00
Water stored in root zone (cm)
N
N
Cr)
M
N
M
11
M
M
N
00
O
O
M
O
N
O
(uze) aam3o udaQ
N
w
--
55