Peanut production covers just over 322,000 ha in the tri-state area of Alabama, Florida, and Georgia with only 28% of these acres being irrigated (USDA, 2009). Only 10% of the total irrigated hectares in Georgia were irrigated using drip, trickle or micro-sprinkler while Florida has over 220,000 ha using some type of drip or trickle irrigation (USDA, 2009). Due to the expense of drip system installation, it is assumed that most of these drip systems are on high value vegetable crops. It is unknown, if or how many, of these drip or trickle systems may be used to grow peanut or other traditional row crops such as cotton (Gossypium hirsutum L.) or corn (Zea mays L.).

Economic simulations showed that subsurface drip irrigation (SSDI) would be more profitable for small areas (<30 ha) because of its lower investment per unit land area and lower pumping costs compared to fixed or towable center-pivot systems. As emphasized by Bosch et al. (1998) and O'Brien et al. (1998), SSDI systems have a near-static cost per hectare compared with overhead sprinkler systems (center pivots), where per hectare cost decreases as the length of the system increases covering more area. Overhead sprinkler irrigation systems are the most common in the tri-state area, because they are easy to assemble, durable, do not require elaborate filtering systems, and familiarity with operation and maintenance. One major concern with overhead sprinkler systems is that once water exits an overhead sprinkler system, its fate may be affected by environmental conditions such that water may not hit the intended target but be lost due to wind and evaporation before it reaches the soil surface and becomes available for crop use. Thus, subsurface drip irrigation has the potential to provide consistently high yields while conserving soil, water, and energy. Some of the major benefits of drip irrigation include precise placement of water and chemicals, low labor requirements, and reduced runoff and erosion compared with overhead sprinkler system. These SSDI systems have the capability of frequently supplying water to the root zone thereby reducing the risk of cyclic water stress typical of other irrigation systems. Research has indicated that crop yield and quality may be increased and that SSDI can be used on cotton (Bucks et al., 1988; Henggeler, 1988, Nuti et al., 2006, Dougherty et al., 2009), and corn (Mitchell, 1981; Mitchell and Sparks, 1982; Powell and Wright, 1993).

These SSDI systems are adaptable to various field sizes and shapes making them an important economic consideration, especially in the southeast. This economic advantage is further evident when considering the option to design a SSDI system to effectively cover an irregularly shaped field that would not be totally covered with a sprinkler type system (Bosch et al., 1998). With proper SSDI designs these systems can provide sufficient water to different fields according to the area, soils, and crop species.

Drip tube laterals have been installed at 0.2- and 0.3-m soil depths (Bucks et al., 1986; Tollefson, 1985; Phene et al., 1987; Camp et al., 1989) on cotton, corn, fruits, and vegetables. Drip laterals have been spaced at 1, 2, and 3 m apart with yields decreasing as lateral spacing increased to greater than 2 m (French et al., 1985; Lamm et al., 1992; Powell and Wright, 1993; Camp et al., 1997; Enciso et al., 2005). Drip tubing was buried or laid on the soil surface at various lateral spacings, i.e., every row or alternate row furrows, in continuous cotton or cotton-corn-peanut rotations (Camp et al., 1993; Camp et al., 1997; Dougherty et al., 2009; Sorensen et al., 2008). In continuous cotton with alternate row lateral spacing, there was year to year variability due to climatic patterns, but irrigated cotton yields were greater than nonirrigated yields especially in dry years (Dougherty et al., 2009). A comparison of alternate row spacing versus every-row lateral spacing indicated no yield differences with either continuous cotton, or cotton-corn-peanut rotations (Camp et al., 1993; Camp et al., 1997; Sorensen et al., 2008).

With increasing concern for water conservation in the tri-state region (Alabama, Georgia, and Florida), the use of SSDI due to the greater irrigation efficiency of these systems, may be of great interest to individual growers, water and environmental conservancy agencies, and policy making agencies. There is little long term peanut yield response data with SSDI in the southeast to make management recommendations. Therefore, the objectives of this research were to determine the long-term yield response of peanut to: 1) three irrigation rates, 2) two lateral spacings, and 3) five crop rotations using SSDI over a 10-year period.

The research site was located in Terrell County near Sasser, GA on a Tifton sandy loam soil (Fine-loamy, kaolinitic, thermic Plinthic Kandiudults) with 2 to 5% slope. A SSDI system was installed in 1998 on non-irrigated farmland that consisted of three irrigation levels, five crop rotations, two drip tube lateral spacings, and three replications for a total of 90 individual plots. Cotton had been planted two years prior to installing the SSDI system. Land ownership had changed such that long term crop rotations were not available and the current owner had not raised peanut since 1993. A 6.8 ha area rectangle was divided into three equal areas referred to as tiers. There were alley-ways (12.2 m minimum) between tiers, at the sides, and crop row ends for equipment turn areas. Each SSDI tier (38 m by 274 m) was randomly assigned an irrigation level. A SSDI tier consisted of three blocks (replications), five crop rotations, and two thin-wall drip lateral spacings for a total of 30 plots per tier. The irrigation levels were 100%, 75% and 50% of estimated crop water use (Sorensen et al., 2001).

The five crop rotations included continuous peanut (PPP), cotton-peanut (CP), corn-peanut (MP), cotton-corn-peanut (CMP), and a cotton-corn-corn-peanut (CMMP) (Table 1). All crops were planted on a 0.91-m row spacing in a single row pattern. The two drip tube lateral treatments had drip tubes installed underneath each crop row (narrow, 0.91-m) and in alternate crop row furrows (wide, 1.83-m). Each narrow row subplot had six crop rows with one drip tube lateral installed under each row and was replicated three times across each tier (replication per block). Each wide row subplot had 10 crop rows with five drip tube laterals installed in alternate crop row middles and replicated three times, one replication in each block. Sorensen et al. (2001) describes in detail the treatments, irrigation system design criteria, and irrigation control. The thin-wall drip tube (Super Typhoon, Netafim Irrigation, Inc., Fresno, CA; www.netafim-usa.com) had a wall thickness of 0.254 mm and emitters spaced every 46 cm with a flow rate of 1.5 L/h per emitter. All thin-wall drip tubing was buried approximately 30 cm deep using a modified ripper shank.

Irrigation water was applied daily based on replacement of estimated crop water use for peanut (Table 2). Air temperature (maximum, minimum and average), relative humidity, total solar radiation, and precipitation were recorded daily. From 1998 to 2003, meteorological data were collected using programmable logic control (PLC) modules. This system worked, but was vulnerable to lightning. In spring 2004 this PLC system was replaced with more reliable datalogger system (Campbell Sci., Inc, Logan, UT; CR23X). Daily potential evapotranspiration (ET_o) was estimated using the modified Jensen-Haise equation adjusted for local conditions (Jensen and Haise, 1963). Daily crop coefficients, K_c, were determined by dividing the estimated daily mean peanut (Stansell et al., 1976), cotton (Harrison and Tyson, 1993), and corn (Lambert et al., 1988) water use values by the daily estimated archived ET_o data for the same time period. Daily ET_o was then multiplied by the daily K_c to estimate the daily water replacement for each crop (estimated ET_a) which is identified as the 100% irrigation level. The other two irrigation levels were determined by multiplying the 100% irrigation level by 75 and 50%, respectively. Length of irrigation time for each irrigation treatment was calculated on estimated daily ET_a to apply the desired depth of water. An irrigation event was not applied if precipitation exceeded the estimated crop water use for that day. A maximum of 10 mm precipitation would be used as a “carry over” to stop irrigation for a short time span following a precipitation event. Daily ET_a values were subtracted from the “carry over” until its value was zeroed, then irrigation events would resume. The 10-mm “carry” is about 25% soil moisture depletion for this soil type.

Lime was applied on all plots and in all years as determined by soil test to maintain soil pH to approximately 6.5. Seed-bed preparation for all crops consisted of one to two passes (once in the fall and once in the spring) with experimental tillage equipment (USDA-ARS-National Peanut Research Laboratory) that would essentially till the top 10 to 15 cm of soil. This equipment would reshape the soil into one single planting bed that was about 1.4 m wide. This equipment also provided the opportunity for controlled-traffic such that no wheeled equipment ran over the buried lateral positions. A small field cultivator was used to break any soil crust, incorporate herbicides, and for weed control prior to planting any crop. After harvest, the crop residue would be mowed (cotton and corn), lightly tilled with a disc harrow, and then re-bedded as described previously.

The only peanut cultivar used, “Georgia Green” (Branch, 1996), was planted between May 1^st to 12^th (depending on weather conditions) with a vacuum type planter (Monosem, ATI., Inc., Lenexa, KS) at about 20 seeds m⁻¹ on a 0.91-m row spacing (Table 2). Treatments in each respective year received the same weed, insect, and disease control management applications following general recommendations outlined by individual product labels or University of Georgia Agricultural Extension Service recommendations (Prostko, 2004). Harvest dates were based on the optimum crop maturity determined by the hull scrape method (Williams and Drexler, 1981). Yield rows were dug with a 2-row inverter and harvested with a 2-row combine. Sample weights were recorded and subsequently divided such that a 4 to 7-kg sub-sample was collected from each plot sample. Each sub-sample was graded and shelled to determine farmer stock grade and kernel size distribution, respectively. Pod yield was based on total sample weight adjusted to 7% moisture. Farmer stock grade and kernel size distribution were determined using procedures specified by the USDA (USDA, 1998). Gross revenue was determined using the average market price for 2008 of $0.45 kg⁻¹ of farmer stock grade peanut (Georgia Dept. of Ag, 2009).

Due to restricted amount of land area, not every crop rotation was planted every year. Consequently, not every combination of peanut rotation could be analyzed by rotation every year. A factorial design of general analysis of variance procedure was used to analyze peanut yield and grade data (Statistix9, 2008) with respect to irrigation rate (tiers), crop rotation, and lateral spacing. Crop yield and grade data were analyzed by individual years, irrigation treatment, crop rotation, and lateral spacing within and across years if applicable. Differences between crop yield and quality means were determined using Tukey's HSD multiple comparison when ANOVA F-test showed significance (P ≤ 0.05).

The use of deep subsurface drip irrigation is feasible for peanut and the associated crop rotations. Average peanut yields across the life of the project indicate 75 and 100% irrigation level had similar yield values and both were numerically higher than the 50% irrigation level. In addition, there was no yield reduction when 75% of the recommended water was applied compared with full (100%) recommended amount, implying a possible 25% water savings for the same yield.

There was a numerical yield benefit of 150 kg/ha with laterals installed underneath each crop row compared with alternate row middles. However, the gross revenue returned from the area where the tubing was placed under each row may not offset the cost of the extra tubing compared with the alternate row middles. It would take over 5 years of constant yield increase to pay for the extra tubing. Therefore, it is recommended that on this soil series and environmental conditions, laterals may be spaced in alternate row middles for highest economic yield and possible economic return.

There was no clear evidence supporting either corn or cotton prior to peanut in an alternate year rotation. Peanut yield did increase as time between peanut crops increased. There was a 27, 43, and 56% increase in peanut yield with 1, 2, and 3 year between peanut crops compared with continuous peanut crops, respectively. It is recommended to have the longest time frame between peanut crops as possible for highest peanut yield with major emphasis on holistic farm planning for maximum economic returns.

Kernel size distribution was most effected by crop rotation and not by irrigation rate or lateral spacing. Typically, as percentages of jumbos decreased, percentages of mediums and ones increased. Also, as both jumbos and mediums decreased then percentages of ones increased. There was no clear evidence of any one crop rotation negatively affecting kernel distribution compared to another except for continues peanut treatment where jumbo- and medium- sized kernels tended to decrease and number one sized kernels tended to increase. This implies that shorter time periods between peanut crops in a rotation would negatively affect kernel size but not necessarily the peanut grade or TSMK.