TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

THE ROLE OF SUBSURFACE DRIP IRRIGATION FOR THE SUSTAINABILITY OF ARID-REGION AGRICULTURE
1 1

T.L. Thompson, 2E.C. Martin, and 3R. Tronstad

Dept. of Plant and Soil Science, PO Box 42122, Texas Tech University, Lubbock, TX 79409; Phone: (806) 742-2838; Fax (806) 742-0775; E-mail: thomas.thompson@ttu.edu. 2 Dept. of Agricultural and Biosystems Engineering, Maricopa Agricultural Center, 37860 Smith-Enke Rd., Maricopa, AZ 85239; Phone: (520) 568-2273; Fax: (520) 568-2556; Email: edmartin@cals.arizona.edu. 3Dept. of Agricultural and Resource Economics, University of Arizona, Tucson, AZ 85721; Phone: (520) 621-2425; E-mail: tronstad@cals.arizona.edu

Abstract The adoption of subsurface drip irrigation (SDI) provides a potential solution to the problem of low-water-use efficiency in production agriculture. Other advantages of SDI include reduced NO3 leaching compared to surface irrigation, higher yields, a dry-soil surface for improved weed control, better crop health, and harvest flexibility for many specialty crops. Use of SDI also allows the virtual elimination of crop water stress, the ability to apply water and nutrients to the most active part of the root zone, protection of drip lines from damage due to cultivation and tillage, and the ability to irrigate with wastewater while preventing human contact. Yet, SDI is used only on a minority of cropland in the arid western U.S.A. Reasons for the limited adoption of SDI include high initial capital investment required, need for intensive management, and urbanization that is rapidly consuming farmland in parts of the western USA. The contributions of SDI to increasing yield, quality, and water-use efficiency have been demonstrated. The two major barriers to SDI sustainability in arid regions are economics (i.e. paying for the SDI system), including the high cost of installation; and salt accumulation, which requires periodic leaching, specialized tillage methods, or transplanting of seedlings rather than direct seeding. Key words: irrigation, sub-surface drip irrigation Introduction

Water-use efficiency in agriculture is often low, with plant use accounting for as little as 40-60% of water applied, and the remainder loss due to evaporation, surface runoff, or percolation into the vadose zone or groundwater (Smith, 1995). Low-water-use efficiency in irrigated agriculture results in wasted water resources and potential non-point source pollution of surface and groundwater resources with nitrates, salts, and agrichemicals including pesticides (Oster and Wichelns, 2003).

451

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Agriculture accounts for 80% of water-use in the western states of Arizona, California, Nevada, New Mexico, and Utah (Koniecski, 2004). However, rapid urbanization in these areas threatens water supplies for agriculture. Increasing populations and decreasing supplies of good quality water are causing increased interest in use of water of marginal quality, including sewage effluent, for irrigation (Hills and Brenes, 2001). Use of subsurface drip irrigation (SDI) provides one potential solution to these problems of low-water-use efficiency and decreasing irrigation water quality. Subsurface drip irrigation offers many advantages for crop production. These include increased wateruse efficiency, reduction of nitrate leaching compared to surface irrigation, higher yields, maintenance of a dry soil surface for improved weed control and crop health, the ability to apply water and nutrients to the most active portion of the root zone, protection of drip lines from damage due to cultivation and other operations, and the ability to safely irrigate with wastewater while preventing human contact (Bar-Yosef, 1999; Camp, 1998; Lamm, 2002; Enriquez et al., 2003; Phene, 1995). Recent research has shown that with proper management, crop yields can be optimized with SDI while minimizing pollution through leaching losses of N (Thompson et al., 2000, 2002). Irrigation with SDI allows maintenance of low root-zone salinity, even when using irrigation waters containing appreciable salts (Oron et al., 1999; Oron et al., 2002). Two potential shortcomings of SDI are its high initial cost and the potential for the development of soil salinity. Installation of SDI systems may cost >$1000 ac-1 (Hanson and May, 2003). Conversion from conventional irrigation systems requires high capital inputs and increases in time required for irrigation design implementation and management. Costs of conversion to SDI often prevent producers from seeing the benefits of its adoption.

452

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

However, when amortized over the 10+ year life of the system, actual cost of SDI installation is only $100 acre-1 year-1. Proper management of SDI systems also requires higher time commitment and specialized training, which usually impact the economics of conversion to SDI. Soil salinity is a common problem with SDI. Precise application of water and nutrients below the soil surface is a benefit of SDI, but leads to accumulation of salt near the soil surface. Net upward flow of water from the SDI emitter to the surface and loss of water to evaporation and transpiration can lead to high ECe values at the soil surface (Dasberg and Or, 1999). Management of accumulated salts leads to increased demand on time and money to maintain high levels of productivity. Reduction of salts accumulated at or near the soil surface is often accomplished using sprinklers, which leads to increases in labor and capital inputs (Hillel, 2000). Despite its many advantages for crop production, particularly in arid regions, challenges remain that hinder more complete adoption of SDI. Two challenges, in particular, stand out: paying for the system, and long-term prevention of salt buildup in the soil. Our objective in this paper is to illustrate how our recent research in Arizona has contributed to our understanding of these issues.
Previous Studies

Field experiments with SDI were conducted during 1991-1999 to evaluate various outcomes of production of high-value crops with SDI. The results are described in several published papers (Pier and Doerge, 1995; Thompson and Doerge, 1996; Thompson et al., 2000, 2002). Briefly, each experiment consisted of factorial combinations of soil-water tension (dry to wet) and nitrogen-fertilizer rate (suboptimal to excessive). These two factors

453

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

were chosen because they are the factors most likely to limit crop growth in arid regions. Crop yield and quality were measured at harvest, N fertilizer fate was estimated using the difference method, and net economic return was calculated using standard crop budgets. The objective of these experiments was to evaluate simultaneously the agronomic, economic, and environmental outcomes of crop production with SDI. Spatial analysis techniques were used to evaluate outcomes of SDI production. Briefly, response surfaces for marketable yield, net economic return, and unaccounted N were developed using the SAS RSREG (SAS Institute, 1999) procedure. Optimum ranges on the response surfaces were defined as >95% of marketable yield and net economic return and unaccounted N of <40 kg ha-1. This was estimated to represent the amount of N that could be lost and yet result in drainage water NO3-N of <10 mg L-1. Finally, we determined the ranges of N and soil-water tension where the three optimum ranges overlapped, that is, where agronomic, economic, and environmental production criteria were simultaneously maximized. Sample response surfaces are shown in Figure 1. During 1991-1999, we evaluated agronomic, economic, and environmental outcomes of SDI production for four crops (watermelon, lettuce, broccoli, cauliflower) during eight cropping seasons (Pier and Doerge, 1995; Thompson and Doerge, 1996; Thompson et al., 2000, 2002). During six of the eight seasons, we identified regions where all three production criteria were simultaneously optimized. As shown in Figure 1, we also found that excessive N or water applications resulted in high amounts of N loss, and this N was probably lost from the soil profile. Therefore, good water and fertilizer management is still necessary when using SDI. We concluded from these experiments that production of high-yielding vegetable crops is compatible with environmental protection. These experiments did not, however,

454

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

address some important questions about SDI in arid regions. First, the economic analysis did not include the cost of the SDI system. Second, each experiment utilized a new SDI installation; thus the effects of continued SDI production on soil-salt accumulation were not evaluated. A new project was initiated during 2002 (described below, Long-Term Evaluation of SDI) to address these and other issues important to long-term use of SDI.
Economics of Subsurface Drip Irrigation

The economics of SDI installation and use by growers depend on several factors including 1) cost of system components and installation, 2) system longevity, 3) reduced input or management costs with SDI, 4) water cost, and 5) yield and quality increases with SDI compared to surface irrigation. Doubts about the ability pay for the cost of an SDI system with yield increases or savings in production costs are probably most responsible for inhibiting the adoption of SDI. In Arizona and other states of the U.S. Southwest, rapid urbanization, which is resulting in urban development of farmland also is limiting adoption of SDI. Potential effects of several variables on outcomes of SDI production are illustrated in Figure 2. We assume 33% water savings with SDI compared to surface irrigation. This assumption is likely most valid when SDI is compared to surface-flood irrigation, and less valid when SDI is compared with sprinkler irrigation. The second key assumption is that use of SDI allows growers to save on costs of fertilizer and chemical applications. Application cost savings of $25 ac-1 were used in this analysis. Different system installation costs were used in Figures 2A and 2B. The curves in Figure 2 illustrate the influences of revenue, water cost, and SDI drip tape lifetime on the economic outcomes of SDI compared to surface-flood irrigation. In

455

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

situations with a crop yielding low revenue, with a low water price, or with a short SDI system life, large revenue increases will be needed to pay SDI system costs. Revenue increases may come from yield or price increases. For example, an increase in cotton yield or quality (i.e. micronaire) induced by SDI could result in such revenue increases. For crops such as broccoli, yield increases may not translate into revenue increases because harvest costs are a major component of production costs. Currently (2006), irrigation water in most irrigation districts in central Arizona costs $35 to $45 ac-ft-1. It is evident from Figure 2 that, even with crops yielding high revenues, revenue increases must be achieved in order to pay for SDI systems. Higher water prices (e.g. $80 ac-ft-1 as shown in Figure 2) would make the economics of SDI more favorable relative to surface-flood irrigation. Similarly, as length of system life increases, lower revenue increases are needed to pay for SDI systems.
Long-Term Evaluation of SDI

In 2002, a demonstration/research project (AZdrip, Web site: http://ag.arizona.edu/azdrip) was established at the University of Arizonas Maricopa Agricultural Center. Objectives of the project were to 1) evaluate management practices for efficient and sustainable irrigation using SDI, and 2) provide information on and demonstration of SDI management practices for Arizona growers. The AZdrip project site features five large plots (67' x 405') with SDI or surface-flood irrigation. Large plots allow use of large-scale field equipment for effective demonstration of SDI management techniques. This project features long-term demonstration and evaluation of various aspects of crop production with SDI, in comparison with conventional surfaceflood irrigation. Four of the plots have SDI installed in one of two configurations (Table 1,

456

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Fig, 3), with two different irrigation scheduling treatments (Table 1). With high-frequency" SDI irrigation, soil moisture is kept near field capacity at all times. Five Irrometer transducer-equipped tensiometers located in each high-frequency plot are interfaced with a Campbell Scientific Datalogger to trigger irrigation (0.07") when the average soil-water tension reaches 10 cbar. Non-automated tensiometers in low-frequency plots are used to schedule irrigation events (0.5 - 1.0") when soil-water tension reaches 30-50 cbar. Surface irrigation is scheduled using the AZSched program. These combinations of SDI configuration and management were chosen based upon input from experienced SDI growers in Arizona. The drip tubing is Netafim Typhoon, 13-mil-wall thickness, 12-inch emitter spacing, and 0.18 - 0.25 gal/emitter/hr flow rate. A pump delivers 70 gpm @ 40 psi. Filters are two Netafim 24" sand filters and one Netafim disk filter. Acid, fluid fertilizers, and labeled insecticides are injected with two LMI Milton Roy Electromagnetic Dosing Pumps. Irrigation water for the SDI plots is continuously acidified to pH 6.0. The entire system is flushed and chlorinated twice per season. Filters backflush automatically when pressure differential is >5 psi. This project was established during summer 2002 and is intended to function for at least 10 years. The first crop, broccoli, was planted in October 2002 and harvested in February 2003. Seedless watermelon was planted in March 2003 and harvested in June-July 2003. A second broccoli crop was planted in November 2003 and harvested in March 2004. Another broccoli crop was planted in September 2004 and harvested in January 2005. A crop of watermelon was planted in April 2005 and harvested in July 2005. Barley was planted in November 2005 and harvested in April 2006. Seedless watermelon was planted in April 2006

457

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

and harvested in July 2006. We will continue to grow one to two crops per year. Short-and long-term evaluations allow comparisons of surface vs. SDI and among the various SDI treatments with respect to crop yield and quality, water-use efficiency, fertilizer and pesticide use, and economic returns.
Results of Demonstration Project

High-frequency irrigation with SDI benefited summer-grown watermelon crops, but did not benefit winter-grown broccoli crops compared to low-frequency irrigation with SDI (data not shown). Providing for high-frequency SDI irrigation for summer-grown crops (irrigation at least once/day) to maintain soil-water tension near 10 cbar should result in higher yield of watermelons. The plot design is conducive to high surface irrigation efficiency. However, the combination of generally lower water-use and higher yields with SDI have resulted in substantially higher water-use efficiency (crop produced per unit of water) with SDI than with surface irrigation (Fig. 4). With high-frequency SDI, cumulative water-use efficiency was twice that with surface irrigation. Thus, twice as much crop was produced using the same amount of water with high-frequency SDI than with surface irrigation. We used actual commercial harvest yields determined in the demonstration plots, standard crop budgets (http://ag.arizona.edu/crops/vegetables/econ/vegecon.html ), and actual SDI system installation costs to determine economic outcomes in the AZdrip project. Although all systemsincluding surface-flood irrigationwere profitable, net returns over ownership and variable costs were three to six times higher with SDI than with flood irrigation during the first five seasons (Table 2). This advantage was most pronounced with watermelon crops.

458

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Results of this project to date have demonstrated that use of SDI in Arizona crop production can consistently result in higher water-use efficiencies and higher economic returns than surface-flood irrigation.
Salt Accumulation with SDI

Continued use of SDI in arid regions will inevitably lead to salt accumulations detrimental to crop growth, unless salt accumulation can be minimized. Salt accumulation may pose the single largest constraint to sustainable use of SDI in Arizona. Detrimental salt accumulations can be avoided by a) cultural techniques with some crops (e.g. pre-irrigation followed by removal of the bed cap with cotton), b) periodic leaching with sprinklers, or c) transplanting of high-value crops. In another experiment, we evaluated the effects of germination method (irrigation with SDI or sprinklers), depth of SDI tubing (7 and 10), and irrigation water salinity (1.5 and 2.6 dS m-1) on salt and Br distribution after one growing season. Following only one season, salt accumulation was high enough to significantly reduce the germination and establishment of the next crop. Although sprinklers were needed to achieve 100% establishment of the succeeding crop, timely rainfall can sufficiently decrease salinity and allow germination of a moderately salt-sensitive crop such as cantaloupe. Areas with exceptional high water quality (<0.5 dS m-1) may not require sprinkler pre-irrigation for several years, as shown by Burt et al. (2003). Methods for managing salt without use of sprinklers include transplanting and bed shaping. Using transplants may eliminate the need for sprinklers during establishment, because the root ball is placed from 5 to 10 cm below the zone of highest salt accumulation. However, sprinklers are often used with transplants to prevent desiccation, because several

459

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

hours may be required for water to move from the drip tape to the root zone. Transplants may eliminate the need for sprinklers to manage salts, but require high capital inputs and may not improve the economic sustainability of SDI. Bed shaping has been introduced as a means to manage salt accumulation above the drip tape. This method involves forming the beds to a peak and pre-irrigating to move salts toward the peak. Tops of the bed are then removed into the furrow, leaving behind soil of low ECe. Direct seeding of some large-seeded crops can then occur without inhibition of emergence. Small-seeded crops that require precision planters cannot usually be direct-seeded into moist beds. Bed-shaping procedures may prove effective in some crop rotations by eliminating the need for sprinklers, but the excess water needed to pre-irrigate beds may be less economically feasible, depending on water cost.
Conclusions

To be sustainable, SDI must 1) achieve substantially higher water-use efficiency than surface-irrigation methods, 2) increase crop yield and quality, 3) be economically viable, 4) contribute in a positive way toward environmental protection, and 5) not lead to development of adverse soil properties (salinity, sodicity). Many authors have shown that SDI is superior to most other irrigation methods with respect to points 1, 2, and 4. However, the economic viability of SDI depends on several factors under the control of the grower, including SDI system life, crop produced, water savings, and application cost savings. However, there are other factors, not under the growers control, including water cost and crop prices, that profoundly affect the economics of SDI. In arid regions, there is usually insufficient rainfall to leach salts from the zone of soil above the SDI tape. In this zone, detrimental salinity may result. In most cases, use of SDI in arid regions requires periodic use of sprinklers to leach salts.

460

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Contributions of SDI to increasing yield, quality, and water-use efficiency have been demonstrated. Two major barriers to SDI sustainability in arid regions are economics, including the high cost of installation and low cost of water; and salt accumulation, which demands periodic leaching, specialized tillage methods, or transplanting.
Acknowledgements

The help of colleagues and students, past and present, is gratefully acknowledged. In particular, we acknowledge Tom Doerge, Scott White, Ron Godin, and Trent Roberts. Grant support was provided by the Arizona Water Sustainability Program and other sponsors.
References

Bar-Yosef, B., 1999. Advances in fertigation. Adv. Agron. 65: 2-77. Burt, C.M., O. Al-Amoudi, and A. Paolini, 2003. Salinity patterns on row crops under subsurface drip irrigation on the west side of the San Joaquin Valley of California. ITRC Report No. R 03-004. Camp, C.R., 1998. Subsurface drip irrigation: a review. Trans ASAE 41:1353-1367.Dasberg, S., and D. Or., 1999. Applied agriculture: drip irrigation. Springer-Verlag, New York, NY. Enriquez, C., A Alum, E. Suarez-Rey, C.Y. Choi, and C.P.Gerba, 2003. Survival of bacteriophages MS-2 and PRD-1 in turfgrass irrigated by subsurface drip irrigation. J. Environ. Eng. 129:852-857. Hanson, B.R., and D. May, 2003. Drip irrigation increases tomato yields in salt-affected soil of San Joaquin Valley. California Agriculture. 57:4:132-137.Hillel, D., 2000. Salinity management for sustainable agriculture: integrating science, environment and economics. The World Bank. Washington, D.C.

461

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Hills, D.J., and M.J. Brenes, 2001. Microirrigation of wastewater effluent using drip tape. Appl. Eng. Agric. 17:303-308. Koniecski, A.D., 2004. Water-use trends in five southwestern states, 1950-2000. Arizona Water Resources Supplement. Online at (http://cals.arizona.edu/azwater/awr/julyaugust04/USGS.Supplement2.web.pdf). Verified 24 February, 2005. Lamm F. R., 2002. Advantages and disadvantages of subsurface drip irrigation. http://www.oznet.ksu.edu/sdi/Reports/2002/ADofSDI.pdf. Kansas State University, Colby Kansas. Originally presented at International Meeting on Advances in Drip/Micro Irrigation, Puerto de La Cruz, Tenerife, Canary Islands, December 2-5, 2002. Oron, G., Y. De Malach, L. Gillerman, I. David, and V.P Rao, 1999. Improved saline wateruse under subsurface drip irrigation. Agric. Water Manage. 39:19-33.Oron, G., Y. DeMalach, L. Gillerman, I. David, and S. Lurie, 2002. Effect of water salinity and irrigation technology on yield and quality of pears. Biosys. Engin. 81:237-247.

Oster, J.D., and D. Wichelns, 2003. Economic and agronomic strategies to achieve sustainable irrigation. Irrig. Sci. 22:107-120. Phene, C.J., 1995. The sustainability and potential of subsurface drip irrigation. pp. 359 367. In Freddie Lamm (ed.) Microirrigation for a changing world. Amer. Soc. Agric. Eng. St. Joseph, MI. Pier, J.W., and T.A. Doerge, 1995. Concurrent evaluation of agronomic, economic, and environmental aspects of trickle-irrigated watermelon production. J. Environ. Qual. 24:79-86.

462

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

SAS Institute, 1999. SAS/STAT users guide. Release 8.2 ed. SAS Inst., Cary, NC Smith, M., 1995. Irrigation Scheduling: from Theory to Practice. 1995. Proceedings of the ICID/FAO Workshop on Irrigation Scheduling, Rome, Italy, 12-13 September. Thompson, T.L., and T.A. Doerge, 1996. Nitrogen and water interactions in subsurface tricle-irrigated leaf lettuce. II. Agronomic, economic, and environmental outcomes. Soil Sci. Soc. Am. J. 60:168-173. Thompson, T.L., T.A. Doerge, and R.E. Godin, 2000. Nitrogen and water interactions in subsurface drip-irrigated cauliflower. II. Agronomic, economic, and environmental outcomes. Soil Sci. Soc. Am. J. 64:412-418. Thompson, T.L., T.A. Doerge, and R.E. Godin, 2002. Subsurface drip irrigation and fertigation of broccoli: II. Agronomic, economic, and environmental outcomes. Soil Sci. Soc. Am. J. 66:178-185.

463

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Table 1. Plot treatments for the AZdrip project.

Plot 1 2 3 4 5 SDI Tubing Placement 3 SDI lines per 80" bed 10 beds per plot 1 SDI line per 40" bed 20 beds per plot 3 SDI lines per 80" bed 10 beds per plot None 1 SDI line per 40" bed 20 beds per plot Irrigation Treatment Automated high frequency Automated high frequency Low frequency Surface flood irrigation Low frequency

Table 2. Return over ownership and variable costs ($ ac-1) in the AZdrip SDI demonstration project. Values assume amortizing cost of SDI tubing over 10 years.
Season/Crop 1/broccoli 2/watermelon 3/broccoli 4/broccoli 5/watermelon Total 3 SDI lines/ 80bed highfreq. irrigation 919 1475 150 924 906 4374 1 SDI line/40bed low-freq. irrigation 977 976 -36 752 741 3410 3 SDI lines/80 bed low-freq. irrigation 941 797 164 567 -112 2357 1 SDI line/40 bed low-freq. irrigation 888 183 120 1022 -86 2108 Flood 869 -130 220 323 -561 740

464

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

1994-95

14 12 10 8 6 4 100
14 12 10 8 6 4 100 200 300 400 500
436 kg ha-1

14

Yield
10.3 kPa

12 10 8 6 4

Economic Return

Soil Water Tension (kPa)

200

300

400

500

100

200

300

400

500

14

Unaccounted N

12 10 8 6 4 100 200 300 400 500

Area of Overlap

N Rate (kg ha-1) -1

1995-96

25 20

25

Yield

20 15

Economic Return
13 kPa

Soil Water Tension (kPa)

15 10 5 100 200 300

9.9 kPa 390 kg ha-1

10 5
500

380 kg ha-1

400

100
25 20 15 10 5

200

300

400

500

25 20 15 10 5 100

Unaccounted N

Area of Overlap

200

300

400

500

100

200

300

400

500

Figure 1. Spatial analysis of agronomic, economic, and environmental outcomes of broccoli production with subsurface drip irrigation during 1994-95 (A) and 1995-96 (B). These figures were originally published in Soil Science Society of America Journal 66:178-185 (Thompson et al., 2002). Used with permission.

N Rate (kg ha-1) -1

465

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Figure 2. Revenue increase needed with SDI as a function of annual revenue for SDI systems costing $1,000/ac (A), and $1,400/ac (B). The SDI costs are paid for with water and application cost savings.

466

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

Bed Configurations
Permanent 80 beds with three drip lines:
9 19

80

40 beds with one drip line per bed:

40

Flood irrigation:

40

Figure 3 Bed configurations of the long-term AZdrip SDI sustainability demonstration project.

467

TRACK 4 ENVIRONMENTAL MANAGEMENT Irrigation, Innovation, and the Environment

8000 7000 Water Use Efficiency (kg yield/ha-cm/ha) 6000 5000 4000 3000 2000 1000 0

> 55% higher WUE with SDI compared to flood

2 m beds, 1 m beds, 2 m beds, 1 m beds, High Freq. High Freq Low Freq Low Freq.

Flood

Figure 4. Cumulative water-use efficiency (through five cropping seasons) in the AZdrip SDI sustainability demonstration project.

468