Larry Curtis, Charles Burmester, David Harkins, and Chet Norris

Farmers may not be able to control rainfall, but cotton producers in North Alabama have new options and management strategies to quench their crops’ thirsts thanks to an irrigation research project underway at the Tennessee Valley Research and Extension Center (TVREC) in Belle Mina.

Cotton continues to be a major agricultural commodity in the Tennessee Valley region of North Alabama with more than 200,000 acres produced in a five-county area along the Tennessee River. Annual yield fluctuations are quite common and often these fluctuations are related to drought or irregularly distributed rainfall. With financial and technical support from the Tennessee Valley Authority, an irrigation research and demonstration facility was constructed at the TVREC facility in 1994-95 to evaluate the potential for enhancement of water resources in that region and to conduct research related to irrigated cotton production.

Aerial view of irrigation reservoir at Tennessee Valley Research and Extension Center.

A nontraditional off-stream storage reservoir was first constructed at the TVREC to harvest abundant water available in the winter and spring for irrigation during the summer. Using this water source, cotton irrigation research has been underway since 1996 with data reported here for 1997 through 1999.

A number of different cotton irrigation experiments are in place. One of the first experiments evaluated irrigation water requirements and irrigation scheduling. This study was designed to determine the minimum irrigation system design flow rate capability needed to produce optimum yields. Three irrigation systems were tested that applied one inch in 5.4 days, one inch in 2.5 days, and a very wet treatment where the soil moisture deficit in the root zone was not allowed to exceed 0.6 inch.

Yields from these management systems were compared to yields of nonirrigated (dryland) cotton, and yield data from 1997-1998 (Figure 1, top) indicate that all three irrigation strategies increased cotton yields. (In all cases the variety planted was NuCotn 33B and bales per acre based on 38% yield from seed cotton with 500-pound lint bales.) Yield differences among the irrigated treatments were small; however, some yield suppression was noted for the extremely wet treatment. In 1999 an experiment was developed to evaluate a wider range of irrigation application capabilities focusing on the minimum design flow rate for irrigation systems that would still maintain optimum yields. The growing season in 1999 was extremely dry and yield data indicated a progressive increase in yields with treatments ranging from no irrigation to a capability of more than two inches per week (Figure 1, middle).

Fig. 1. Top, Sprinkler yield data for 1997-98; middle, sprinkler yield data for 1999; bottom, soil moisture as affected by rain and irrigation in 1999.

Even with the maximum irrigation capability some water mining (a progressive reduction of available soil moisture) was observed from the deeper soil profile (Figure 1, bottom).

The growing season in 1999 was likely the worst case scenario in terms of drought severity. While wet through June and early July, only 1.3 inches of rain was received from July 13 until crop maturity (late August to early September). In all cases irrigation was scheduled using MOISCOT, a spreadsheet-based software package for scheduling cotton irrigation developed by Auburn University researchers, Ted Tyson and Larry Curtis of Biosystems Engineering.

A second major area of study underway at the Research Center involves use of subsurface drip irrigation (SDI) for cotton production. In one experiment, irrigation drip tubing with emitters located every two feet along the tubing was buried permanently at a depth of 15 inches between every-other-row (controlled traffic) (see illustration) and perpendicular to rows (random traffic) with the same number of outlets and same amount of tubing on a per-acre basis.

Diagram of subsurface drip irrigation between every other row.

Irrigation was applied daily based on calculated pan evaporation provided by Alabama Weather Information Service (AWIS) for that location. Irrigation amounts equaled 30, 60, and 90% of pan evaporation after full crop canopy with corresponding percentage adjustments prior to full canopy. Pan evaporation is a measurement of daily evaporation from a standardized U.S. Weather Service device and relates closely to daily water use by growing plants. These rates result in amounts of approximately 0.1 inch per day, 0.2 inch per day, and 0.3 inch per day, respectively, during periods of peak water use. Significant yield differences occurred in 1998 and 1999 with the most dramatic differences occurring in 1999 (Figure 2, top).

A second subsurface drip irrigation study initiated in 1998 is comparing five different drip irrigation tape products with a fertigation component included. This study was installed in an area where continuous crops have been produced for many years. Each of the five tapes in the study have emitters spaced two feet along the tape and tape buried 15 inches between every other row. Rows 340 feet in length were used to better simulate field conditions. Each tape product is being evaluated with conventional fertilizer and fertigated treatments (see table). A tape product also is used on the surface with the conventional fertilizer treatment.

In 1998 little difference between fertility treatments was observed (Figure 2, middle). In 1998 sufficient rainfall occurred late in the growing season so that fertilizer in the upper layers of the soil was more readily available. In 1999 significant yield differences were observed for the fertigated versus conventional plots (Figure 2, bottom). Extremely dry conditions in the upper layers of the soil profile made conventionally applied fertilizer less available, resulting in significant yield reduction compared to fertilizer applied through the irrigation system. Significant yield differences were observed each year between nonirrigated plots and tape plots with fertility treatments. To date only minimal differences have been observed between the different drip irrigation tape products. Results of these two studies indicate that, while yield results for SDI are attractive, questions remain related to the cost, longevity, and suitability of products currently available for the rolling farmland in the Tennessee Valley. More research will be needed to answer these questions.

Fig. 2. Top, 1998-99 yield results from subsurface drip placement and irrigation scheduling; middle, 1998 yield results from subsurface drip products; bottom, 1999 yield results from subsurface drip products.