Conserving Water Through Deficit Irrigation of Alfalfa in the Intermountain Area of California

Steve Orloff, Farm Advisor, Cooperative Extension, Siskiyou County, University of California, Yreka 96097; and Blaine Hanson, Extension Irrigation and Drainage Specialist, Department of Land, Air and Water Resources, University of California, Davis 95616

Abstract

Alfalfa is the dominant crop in the intermountain area of northern California, annually using about 28% of the irrigation water supply. Thus, interest exists in mid-summer deficit irrigation (no irrigation) of alfalfa as a source of water for river enhancement and lake level management for fisheries in times of limited water supplies. This mid-summer deficit irrigation strategy maintains the relatively high yields of the first part of the year and eliminates irrigations during the summer when yields are small and quality is poor. Experiments conducted from 2003 to 2006 examined the impacts of deficit irrigation strategies on alfalfa yield in commercial fields of the Klamath Basin (KB) (Torripsammentic Haploxerolls, Mollic Andaquepts, and Andaquept Haplaquolls) and Scott Valley (SV) (Fluvaquentic Endonaquolls), near Yreka, CA. Irrigation treatments were full irrigation described as the growers’ normal irrigation practice, no irrigation after the first cutting, and no irrigation after the second cutting. Another experiment investigated the effect of deficit irrigation on alfalfa varieties. Deficit irrigation imposed after the first cutting reduced yields by 0.6 to 2.2 ton/acre, while no irrigations after the second cutting reduced yields by 0.31 to 1.23 ton/acre. Yield differences among varieties were significant, but the ranking of varieties varied from year to year.

Introduction

Alfalfa is the dominant crop in the intermountain area of northern California, using about 28% of the water applied for irrigation. Thus, interest exists in mid-summer deficit irrigation of alfalfa as a source of water for river enhancement and lake level management for fisheries and for non-agricultural uses in times of limited water supplies. This mid-summer deficit irrigation strategy consists of full irrigation to meet crop demand during the early part of the crop season and no irrigation (deficit irrigation) during mid-summer. This approach maintains the relatively high yields in spring that contribute to most of the seasonal yield, and reduces yields during the summer when production generally is low and quality is poor. For example, where three harvests per year are made, nearly 75% of the seasonal yield occurs in the first two harvests. In four-harvest situations, the first two harvests contribute to about 63% of the seasonal yield (7). The sale of the water not used during the mid-summer period could compensate the grower for the reduced revenue from lower alfalfa yields caused by mid-summer deficit irrigation.

Past studies have shown that mid-summer termination of irrigation reduced alfalfa yield. Yields under mid-summer deficit irrigation near Yuma, Arizona (USA) were negligible and did not recover at a site with a sandy soil, but summer termination of irrigation at a Maricopa site had a less dramatic effect where soil type was a sandy loam (6). Mid-summer deficit irrigation on a fine sandy loam soil resulted in no yield at a site in Cyprus (5). At a site in the San Joaquin Valley near Fresno, CA yields of a mid-summer irrigation termination treatment (described as no irrigation after June until the following spring) were 65 to 71% of those of a fully irrigated treatment in a two-year study (1). Yields recovered during a third year when all treatments were fully irrigated.

Mid-summer deficit irrigation (with no water applied in July, August, and September) in the Imperial Valley of California reduced yields from 53 to 64% of fully-irrigated alfalfa (9). Mid-summer deficit irrigation reduced yield to 46% of the fully-irrigated treatment in the Palo Verde Valley of southern California (8). Significant yield reductions occurred due to deficit irrigation in a three-year study in Nevada, but yields recovered with adequate irrigation in the fourth year (3).

Mid-summer deficit irrigation treatments resulted in yields ranging from 11 to 49% of the fully irrigated treatment for two fields with clay loam soil near Davis, CA from 2003 to 2006 (4). Total yield reductions were smaller for deficit irrigation in July and August followed by a September irrigation compared to deficit irrigation in July thru September. During the last two years, measurements of crop water use showed seasonal evapotranspiration (ET) of the fully irrigated alfalfa to be similar to the historical value, but the seasonal ET of the deficit-irrigated alfalfa was 224 to 239 mm lower than that of the fully irrigated treatment (4).

These studies showed a variety of alfalfa yield responses to deficit irrigation. The main contributing factors appeared to be soil texture and climate. Because of these site-specific yield responses, we conducted a study in the Intermountain region of northern California to determine the effect of different deficit irrigation strategies on alfalfa yield. The reduction in applied water as a result of deficit irrigation and the carry-over effect of the deficit irrigation on the subsequent cutting the following year were evaluated. While this study focused on the Intermountain region of northern California, it has implications for other areas in the northwest because of increasing urban and environmental demands on limited water supplies (2,10).

Commercial Fields Experiments

Experiments were conducted in commercial fields in the Klamath Basin (41°57’N, 121°28’W) located in southern Oregon and northern California and in Scott Valley (41°33’N, 122°49’W) near Yreka, CA from 2003 to 2005 to examine the effect of deficit irrigation strategies on alfalfa yield. Site nomenclatures, site locations, and site soil types are:

• Malin (Klamath Basin) – Fordney loamy fine sand (Torripsammentic Haploxerolls);

• Tulelake1 (Klamath Basin) – Capjac silt loam (Mollic Andaquepts);

• Tulelake2 (Klamath Basin) – Tule Basin mucky silty clay loam (Andaqueptic Haplaquolls) – The Tulelake 2 site is on the University of California Intermountain Research and Extension Center (IREC), at Tulelake, CA;

• Etna (Scott Valley) – Settlemeyer loam (Fluvaquentic Endonaquolls);

• Ft. Jones (Scott Valley) – Stoner gravelly sandy loam (Typic Haploxerepts).

The elevation of the Scott Valley is approximately 2,700 ft, while that of the Klamath Basin is about 4,000 ft. About 80% of the Klamath Basin is underlain by shallow ground water within 3 to 5 ft of the soil surface during the alfalfa growing season.

A randomized block experimental design was used with four replications. Plot dimensions were 90 by 60 ft. Irrigation treatments were (i) fully irrigated alfalfa; (ii) no irrigation after the first harvest; and (iii) no irrigation after the second harvest. The fully irrigated treatment was the growers’ normal irrigation practice that generally consisted of two irrigations per harvest with 12 hour irrigation times. All locations were irrigated with wheel-line sprinkler systems except at Tulelake2 which was irrigated with a travelling boom system. Irrigation treatments were imposed by plugging three consecutive nozzles on the wheel-line systems so that each irrigation treatment was applied to a plot approximately 60 by 120 ft. Only the center area was evaluated to avoid areas where lateral movement of water may have occurred or areas where irrigation water may have drifted onto the plots.

Fields were fertilized according to grower practice, which was an application of 60 to 75 lb of P₂O₅ per acre as 11-52-0 in one annual application or twice that amount every two years. According to soil tests, no other nutrients were needed and all soils were near pH 7 so no lime was applied.

The fields were harvested three to four times depending on location and the growers’ management practices. Yield was determined for each harvest using a Carter flail-type forage harvester (Carter MFG Co. Inc., Brookston, IN) after the irrigation termination treatments were imposed. Alfalfa stand density was assessed the following spring by visually estimating percent plant cover and by counting the number of stems per square foot. Yield of the first harvest the following spring was used to determine any carry-over effects from deficit irrigation into the subsequent year.

The Tulelake2 2005 deficit irrigation trial was conducted differently. Relatively small areas, 20 by 50 ft, were irrigated with a 70-ft buffer between plots using a travelling irrigation boom with a hose reel (Fig. 1). This enabled us to apply more precise amounts of water uniformly and, by only irrigating small areas of the field, concerns were minimized about any lateral movement of irrigation water into the plots via the perched water table.

Fig. 1. Traveling boom sprinkle irrigation systems used at the Intermountain Research and Extension Center.

Water applied for the fully irrigated treatment at each site was determined by measuring the sprinkler nozzle discharge rates (q) for each treatment and then calculating the average application rate (I) using Equation 1 and the nozzle discharge rate, the sprinkler spacing along the lateral (S_l), and the lateral spacing along the mainline (S_m).

I (inches/hour) =	96.3 × q (gal/min)	[1]
	S_l (ft) × S_m (ft)

The application rate was multiplied by the irrigation time to determine the amount of applied water.

Soil water content was measured with a neutron probe moisture meter, calibrated for each site. Measurements were made at 1-, 2-, and 3-ft depths on sites with shallow ground water and at 1-, 2-, 3-, and 4-ft depths for locations with no shallow ground water. Measurements were periodically made throughout the deficit irrigation time to determine trends.

Variety-Irrigation Experiment

An experiment was conducted from 2004 through 2006 at the IREC to determine yield response of varieties to deficit irrigation. Alfalfa was row seeded with a cone planter at 20 lb/acre to a depth of ½ inch on 8 May 2002. Plots were uniformly fertilized prior to planting with 80 lb of P₂O₅ per acre as 11-52-0. Imazethapyr was applied on 4 June 2002 at 0.063 lb ai/acre for weed control during the establishment year. A combination of metribuzin and paraquat at 0.5 and 0.375 lb/acre, respectively, was applied in early March each subsequent year to control winter annual weeds. Alfalfa weevil treatment was only needed in one year and cyfluthrin was applied at 0.03 lb ai/acre on 23 May 2004. Plots were cut three times the first year and uniformly irrigated throughout the season so that the alfalfa was well established when the irrigation treatments were first imposed in the spring of 2004.

Plots were arranged in a randomized split-plot design with three replications. The main treatments were six levels of irrigation and the 5-ft by 20-ft subplot treatments were 14 alfalfa varieties. Treatments were full irrigation, irrigation amounts equal to 66% of the fully-irrigated treatment applied throughout the irrigation season, irrigation amounts equal to 33% of the fully-irrigated treatment applied throughout the irrigation season, no irrigation after 1 June, no irrigation after 15 July, and no irrigation during the crop season.

Yield and stand data from the commercial field experiments were subjected to analysis of variance using the statistical software package MSTAT-C (MSU, East Lansing, MI) and means were compared using LSD (0.05). A t-test was used to statistically evaluate differences among soil moisture levels at different depths. A split-plot analysis of variance was used to test for statistical significance of treatment effects and interactions in the variety-irrigation experiment using MSTAT-C. Means for annual total forage yield were separated using LSD procedures when the effect of treatment or the interaction was significant (P < 0.05).

Water application rates for the fully-irrigated treatment were determined using reference crop evapotranspiration (ETo) from the California Irrigation Management Information System station located at the IREC. Soil moisture was monitored using Watermark electrical block soil water tension sensors (Irrometer Inc., Riverside, CA) so that water application rates could be adjusted downward to account for the contribution from the perched water table to assure that the crop was not over-irrigated.

Deficit Irrigation in Commercial Alfalfa Fields

Average monthly maximum temperatures for 2003 through 2006 at Ft. Jones (Scott Valley) were higher than those at Tulelake (Klamath Basin), while little difference was found between the average minimum temperatures of the two locations (Fig. 2A). The average monthly precipitation for the same period was higher in late fall and winter for Ft. Jones than for Tulelake (Fig. 2B), but differences in average monthly precipitation from April through October were minimal. The average annual precipitation was 22.9 inches (standard deviation = 3.7) at Ft. Jones and 11.3 inches (standard deviation = 2.9) at Tulelake (10).

Fig. 2. (A) Average maximum and minimum temperatures at Ft. Jones and Tulelake, and (B) average precipitation at Ft. Jones and Tulelake. Values are the average monthly readings for 2003, 2005, and 2006. Error bars are one standard deviation.

Yield (commercial fields). Yields of the deficit irrigation treatments compared to the normal yields at the Malin site in 2003 were 65 and 83% with no irrigation after the first or second harvests, respectively (Fig. 3). This corresponded to respective yield reductions of 0.60 and 0.29 ton/acre (Table 1). However, yields lower than 0.5 ton/acre are not economical to harvest based on grower experience. Thus the practical yield reduction was 0.82 and 0.67 tons/acre. The practical yield reduction is the difference between the total yield of the full-irrigation treatment and the total yield of the deficit irrigation treatment excluding yield of harvests lower than 0.5 ton/acre. Differences were statistically significant for the third harvest and the total yield.

Fig. 3. Fully irrigated and deficit irrigated alfalfa at the Malin site. The fully irrigated alfalfa is above the yellow line; the deficit irrigated alfalfa is below the yellow line.

Table 1. Effect of deficit irrigation on alfalfa yield for the Malin 2003 and Malin 2004 sites in the Klamath Basin.

Irrigation treatment	Yield (tons/acre)
Irrigation treatment	2nd harvest	3rd harvest	Total yield	Actual yield reduction
Malin 2003
Harvest date	Jul 22	Aug 30	—	—
Fully-irrigated	1.03	0.67	1.70	—
No irrigation after 1st harvest	0.88	0.22	1.10	0.60
No irrigation after 2nd harvest	1.03	0.38	1.41	0.29
LSD (0.05)	NS	0.18	0.34	—
Malin 2004
Harvest date	Jul 20	Aug 30	—	—
Fully-irrigated	1.53	1.33	2.86	—
No irrigation after 1st harvest	1.02	0.38	1.40	1.46
No irrigation after 2nd harvest	1.41	0.71	2.12	0.74
LSD (0.05)	0.45	0.40	0.34	—

NS = not significant.

For the Malin site in 2004, total yields of the respective deficit irrigation treatments were 49% and 74% of the full-irrigation yield (Table 1). Actual yield reductions were 1.46 and 0.74 tons/acre, respectively. Yield differences were statistically significant for both harvests and for the total yield.

Deficit irrigation reduced the respective total yields to 81 and 86% of the full-irrigation total yield at Tulelake1 in 2003 and 83 and 88% in 2004 (Table 2). Smaller yield reductions occurred at the Tulelake2 site with respective deficit irrigation yields of 86 and 93% of the fully-irrigated yield in 2005 (Table 2). Total yields of the two respective deficit irrigation treatments at the Etna site in 2004 were 50 and 73% of the full-irrigation yield (Table 3). Alfalfa yield responses to deficit irrigation were 55 and 75% of the full-irrigation total yield, respectively, at the Ft. Jones site in 2005 (Table 3) (Fig. 4).

Fig. 4. Fully irrigated and deficit irrigated alflafa at the Ft. Jones (Scott Valley) site.

Table 2. Effect of deficit irrigation on alfalfa yield for the Tulelake1 2003, Tulelake1 2004, and Tulelake2 2005 sites in the Klamath Basin.

Irrigation treatment	Yield (ton/acre)
Irrigation treatment	2nd harvest	3rd harvest	4th harvest	Total yield	Actual yield reduction
Tulelake1 2003
Harvest date	Jul 4	Aug 5	Sep 11	—	—
Fully-irrigated	1.24	1.36	1.21	3.81	—
No irrigation after 1st harvest	0.95	1.13	1.01	3.10	0.71
No irrigation after 2nd harvest	1.22	1.20	0.86	3.28	0.53
LSD (0.05)	NS	NS	0.33	0.66	—
Tulelake1 2004
Harvest date	Jun 29	Jul 27	Aug 30	—	—
Fully-irrigated	1.14	1.60	1.26	3.99	—
No irrigation after 1st harvest	0.89	1.47	0.94	3.30	0.69
No irrigation after 2nd harvest	1.20	1.48	0.82	3.50	0.49
LSD (0.05)	NS	0.10	0.20	0.32	—
Tulelake2 2005
Harvest date	Jul 20	Aug 23	Oct 12	—	—
Fully-irrigated	2.28	1.55	0.82	4.65	—
No irrigation after 1st harvest	2.00	1.40	0.59	4.00	0.65
No irrigation after 2nd harvest	2.30	1.38	0.67	4.34	0.31
LSD (0.05)	NS	0.13	NS	0.49	—

NS = not significant.

Table 3. Effect of deficit irrigation on alfalfa yield for the Etna 2004 and Ft. Jones 2005 sites in Scott Valley.

Irrigation treatment	Yield (ton/acre)
Irrigation treatment	2nd harvest	3rd harvest	4th harvest	Total yield	Actual yield reduction
Etna 2004
Harvest date	Jul 28	Sep 6	—	—	—
Fully-irrigated	1.73	1.06	—	2.80
No irrigation after 1st harvest	1.19	0.20	—	1.39	1.41
No irrigation after 2nd harvest	1.71	0.33	—	2.04	0.76
LSD (0.05)	0.37	0.37	—	0.61	—
Ft. Jones 2005
Harvest date	Jul 13	Aug 18	Sep 30	—	—
Fully-irrigated	2.56	1.33	1.03	4.92	—
No irrigation after 1st harvest	2.10	0.48	0.14	2.72	2.20
No irrigation after 2nd harvest	2.71	0.79	0.19	3.69	1.23
LSD (0.05)	0.33	0.33	0.17	0.66	—

These results are similar to other studies in that deficit irrigation reduced yield. In general, the degree of yield reduction we observed was less than in previous studies. This is likely due at least in part to the mild summer temperatures in the intermountain area of northern California and southern Oregon compared with the more arid and hot desert areas of central and southern California and Arizona. We also found that under shallow groundwater conditions, the groundwater contributed substantially to the crop ET and yield losses were minimal. This condition did not exist in previous studies. Yield reductions in this study and previous studies, expressed as a ratio of the deficit irrigation yield to full irrigation yield, varied depending on location, so it is difficult to state that mid-summer deficit irrigation generally reduces yield by a given amount. This underscores the importance of understanding local site conditions to predict the impact of deficit irrigation even within a given geographic area.

It is conceivable that an alfalfa crown could survive the effects of deficit irrigation but be weakened and produce fewer stems per crown. Therefore, we assessed alfalfa stand density at each site the following spring of each year by visual ratings of plant cover and by counting the number of stems occurring within an 18 inch diameter ring tossed at random three times per plot. We found no difference in visual stand ratings or stem numbers between fully irrigated and deficit irrigated plots (data not shown). In addition, first harvest yields the following year were the same (data not shown) for all treatments indicating no residual effect from the deficit irrigation treatments.

Soil water content. No trend in soil water content over time was found for the full irrigation in Scott Valley (Fig. 5A), while soil water content decreased over time after the 1 June cutoff (Fig. 5B). Significantly lower (P < 0.05) water contents occurred at the 1-ft depth for the cutoff treatment than for the other depths. At this site, no contributions of ground water to ET occurred. Because of the contribution of the shallow ground water at the 2- to 3-ft depth at the Tulelake site, no trend in soil water content over time was found for all treatments (Fig. 6).

Fig. 5. Volumetric soil water content for (A) fully irrigated and (B) deficit irrigated (no irrigation after the first cutting) treatments for the Ft. Jones (Scott Valley) site.

Fig. 6. Volumetric soil water content for (A) fully irrigated and (B) deficit irrigated (no irrigation after the first harvest) treatments for the Tulelake site (Klamath Basin).

Water savings. The water saved by use of mid-summer deficit irrigation was equal to the amount of water applied to the fully irrigated treatments minus the water applied to the deficit irrigation treatments. Water savings from the deficit irrigation treatments varied among sites depending on the growers’ irrigation practices (Table 4). Irrigation termination after the first harvest (typically no irrigation after 1 June) resulted in a water savings of between 11 and 23 inches. When no irrigation was applied after the second harvest (usually equated to no irrigation after 15 July), the water savings was between 6 and 17 inches. These amounts represent a considerable reduction in the total seasonal water application, since most alfalfa fields in the Intermountain area are only irrigated one or two times before first harvest.

Table 4. Water savings due to mid-summer deficit irrigation.

Treatment	Malin 2003	Malin 2004	Tulelake1 2003	Tulelake1 2004	Tulelake2 2005	Etna 2004	Ft. Jones 2005
Treatment	Water savings (inches)
No irrigation after 1st harvest	11.0*	11.0	21.0	16.8	15.1	16.0	22.9
No irrigation after 2nd harvest	5.5	5.5	16.8	8.4	6.2	6.4	13.1

* Water savings equals the amount of water applied to the fully-irrigated treatments minus the water applied in the deficit irrigation treatments.

These results suggest that voluntary water transfers may be feasible without excessive yield loss or a reduction in stand density. The water price needed to make up for the yield reductions due to deficit irrigation varied considerably between sites depending both on the degree of the yield loss and the amount of water applied to achieve full yield. On average, a water value of $119/acre-ft was needed to cover the loss in alfalfa yield assuming an alfalfa hay price of $120/ton. The value ranged from $49 to $240/acre-ft depending on the site. These values are calculated based on gross returns and are not discounted for the reduction in inputs that would occur if yield was reduced due to deficit irrigation (examples include lower harvest cost, less fertilizer required, perhaps reduced pesticide inputs, etc.).

Variety-irrigation experiment. The variety-irrigation experiment at the IREC showed statistically significant differences in yields among varieties (P < 0.0001) (Table 5). There was a significant year by variety interaction (P < 0.0001) indicating that the relative ranking of varieties differed across years. The fall dormancy 2 varieties were lower yielding than the less dormant varieties. However, there was no consistent difference among the fall dormancy 3 to 5 cultivars.

Table 5. Effect of irrigation treatment and alfalfa variety on average annual yield in 2004 through 2006 on Tule Basin mucky silt clay loam (Andaqueptic Haplaquolls) at the Intermountain Research and Extension Center, Tulelake, CA.

Variety	FD	Irrigation treatment
		Full irrig.	66% of full	33% of full	Jun 1 cutoff	Jul 15 cutoff	No irrig.	Variety mean^w
		Yield (ton/acre)^yz
Dura 512	5	8.31	8.31	7.83	7.78	8.03	7.60	7.97
Abilene+Z	5	8.24	8.21	8.08	7.71	7.99	7.67	7.98
WinterGold	4	8.55	8.39	8.09	7.87	8.27	7.67	8.14
HybriForce 400	4	8.22	8.25	8.11	8.10	8.30	7.92	8.15
Reward II	4	8.36	8.05	8.23	8.07	8.11	7.83	8.11
Y54Q25	4	8.43	8.17	8.07	8.10	8.18	7.88	8.14
Masterpiece	4	8.29	8.49	8.15	7.83	7.85	7.60	8.04
LegenDairy YPQ	3	8.59	8.26	8.12	8.07	8.01	7.88	8.16
Magnum V Wet	3	8.63	8.17	8.06	8.43	8.05	8.05	8.23
Blazer XL	3	8.34	8.32	8.00	7.98	7.92	7.74	8.05
WL 325 HQ	3	8.33	8.41	8.41	8.24	8.38	8.21	8.33
Ladak DL	3	8.07	7.87	7.91	7.73	7.93	7.43	7.82
Rennovator	2	7.70	7.83	7.60	7.45	7.86	7.49	7.66
Vernal	2	7.68	7.80	7.50	7.14	7.47	7.28	7.48
Irrigation mean^x	—	8.27	8.18	8.01	7.89	8.03	7.73	—

^w LSD (P = 0.05) for comparison of 14 variety means is 0.23.

^x LSD (P = 0.05) for comparison of 6 irrigation means is 0.35.

^y LSD (P = 0.05) for comparison of variety subplot means for the same irrigation treatment is 0.57.

^z LSD (P = 0.05) for comparison of variety means at same or different irrigation treatment is 0.61.

The irrigation treatments had a highly significant effect on yields (P < 0.0001) (Table 5). There was also a significant interaction between variety and irrigation at a level of significance equal to 0.05 (P = 0.0372). In general, yield decreased with decreasing amounts of applied water with the highest yields occurring for the full irrigation treatment and the smallest yields for the no-irrigation treatment. However, there were exceptions such as the highest yield in 2005 occurred for the 66% treatment and the smallest yield in 2006 occurred for the 1 June cutoff treatment. Varieties differed somewhat in their response to deficit irrigation with deficit irrigation having more of an effect on some varieties than others. The response did not appear to be associated with fall dormancy.

Applied water amounts ranged from 10.5 to 15.9 inches for the full treatment (Table 6). The applied water amounts of the two cutoff treatments reflect the amount applied before cutoff. In 2006, because of the wet spring, no irrigations occurred before the 1 June cutoff. Yield differences between the full and no irrigation treatments averaged 0.54 (0.70, 0.27, and 0.64 ton/acre for 2004, 2005, and 2006, respectively) considered to be small given the differences in irrigation amounts. These small differences reflect the effect of the contribution of the shallow ground water to the crops’ ET.

Table 6. Annual water application rates for the six irrigation treatments in 2004-2006.

Irrigation treatment	2004	2005	2006
Irrigation treatment	Water application rates (inches)
Full irrig.	10.8	10.5	15.9
66% of full	6.8	7.6	10.6
33% of full	3.4	4.2	5.3
June 1 cutoff	3.5	0.6	0.0
Jul 15 cutoff	7.3	3.9	5.7
No irrigation	0.0	0.0	0.0

Conclusions

Deficit irrigation generally reduced total yields compared to full irrigation. The later the implementation of deficit irrigation, the smaller was the effect on total yield. The degree of yield reduction varied considerably among sites depending on factors such as depth to the water table and soil type. Other possible contributing factors include age and productivity of the stand, number of total cuttings, and grower irrigation practices. Klamath Basin sites had relatively high water table levels with depths ranging between about 3 to 3.5 ft, whereas, the water table at the Scott Valley sites was inaccessible to the alfalfa roots. Therefore, deficit irrigation had a greater effect on yield at the Scott Valley sites than at the Klamath Basin sites (Malin and Tulelake). The yield reduction at the sites in the Klamath Basin was also usually greater at sites with more sandy textured soil than at sites with the high organic matter silt and silty clay loam soils. In fact, the yield per cutting in deficit irrigated plots at the Tulelake sites with high organic matter soil (Tulelake1 and 2) never fell below 0.5 ton/acre, the amount assumed to be necessary to warrant harvest. The yield reduction was greater at sites that were adequately irrigated. Even the fully irrigated treatments were under-irrigated at some sites so the full difference in yield may not have been realized at these locations.

Deficit irrigation of alfalfa during the mid-summer shows promise as a strategy for dealing with water shortages for the climatic conditions at these sites by allowing growers to harvest the more valuable and higher yielding spring cuttings and then deficit-irrigate during the period of lower yields and poor quality. Thus, while total alfalfa yield for the season is reduced, a significant proportion of the annual production is still obtained. We found no reduction in alfalfa stand density or a negative carryover effect on yield the following year in these studies.

The quantity and value of transferable water may depend on the site and its hydrology, source of irrigation water, and the fate of water applied in excess of crop ET. In areas where surface runoff and deep percolation from agricultural fieamount of transferable water might equal the ET differences between fully-irrigated and deficit-irrigated alfalfa. In areas where the surface runoff and deep percolation from a field cannot be reused and is not stored or reused at some later time, the amount of transferable water might equal the amount of the irrigation that would be applied during the mid-summer. In either case, alfalfa is well suited to deficit irrigation in summer for potential water transfers due to its drought tolerance and, unlike many other crops, considerable production is still obtainable even with partial irrigation.

Literature Cited

1. Frate, C. A., Roberts, B.A., Marble, V. L. 1991. Imposed drought stress has no long-term effect on established alfalfa. Calif. Ag. 45:33-36.

2. Griggs, T. C., and Riggs, W. 2002. Emerging issues for alfalfa and other forages in the Great Basin. Pages 9-12 in: Proc. of the 32nd Western Alfalfa and Forage Conf., 11-13 December 2002, Reno, NV. Dept. of Agron. and Range Sci. Coop. Ext., Univ. of Calif., Davis.

3. Guitjens, J. C. 1993. Alfalfa irrigation during drought. J. Irrig. Drain. E. 119:1092-1098.

4. Hanson, B., Putnam, D., and Snyder, R. 2007. Deficit irrigation of alfalfa as a strategy for providing water for water-short areas. Agr. Water Manage. 93:73-80.

5. Metochis, C., Orphanos, P.I. 1981. Alfalfa yield and water use when forced into dormancy by withholding water during the summer. Agron. J. 73:1048-1050.

6. Ottman, M. J., Tickes, B. R., Roth, R. L. 1996. Alfalfa yield and stand response to irrigation termination in an arid environment. Agron. J. 88:44-48.

7. Putnam, D. 2208. Intermountain alfalfa variety trials data. Online. California Alfalfa Workgroup, Alfalfa and Forages, Univ. of Calif., Davis.

8. Putnam, D., Takele, E., Kallenback, R., and Graves, W., 2000. Irrigating alfalfa in the Low Desert: Can summer dry-down be effective for saving water in alfalfa? Report submitted to the Bureau of Reclamation (USDI), Yuma, AZ.

9. Robinson, F. E., Teuber, L. R., and Gibbs, L. K. 1994. Alfalfa Water Stress Management During Summer in Imperial Valley for Water Conservation. Desert Res. and Ext. Center, El Centro, CA.

10. Shewmaker, G., Fransen, S., and Hannaway, D. 2002. Emerging issues with alfalfa in the Pacific Northwest. Pages 1-9 in: Proc. of the 32nd Western Alfalfa and Forage Conf., 11-13 December 2002, Reno, NV. Dept. of Agron. and Range Sci. Coop. Ext., Univ. of Calif., Davis.

11. Western Regional Climate Center. 2008. Online. National Regional Climate Centers, Desert Research Institute, Reno, NV.