Intermittent Irrigation Effects on Barnyardgrass Weed Control and Rice Yield

Transcription

1 PEST MANAGEMENT: WEEDS Intermittent Irrigation Effects on Barnyardgrass Weed Control and Rice Yield E.F. Scherder, R.E. Talbert, J.D. Branson, M.L. Lovelace, and E.D. Vories ABSTRACT An experiment was conducted at Stuttgart, AR, in 2000, 2001, and 2002 to evaluate weed control and rice yield components with intermittent flooding (48, 41, 34, 27, or 21% volumetric water contents prior to reflooding) versus a permanent flooding system in dry-seeded rice. The treatment structure was a split-plot design with the main plot consisting of six moisture regimes and the subplot factor consisting of seven herbicide treatments. Rice was dry-seeded, and plots remained unflooded until the time of the permanent flood at a 5- to 6-leaf rice growth stage. The flood was introduced in each moisture regime and then allowed to decline naturally through evapotranspiration, except in those bays designated as permanent flood. When soil reached the targeted volumetric water content in each bay, the flood was reintroduced. This cycle of flooding and draining was continued until maturity was reached. Observations for the experiment were total water usage, annual grass control, and various crop parameters. The intermittent irrigation system reduced the amount of water used, with total water savings ranging from 2 to 4% for the 48% moisture regime compared to total water usage in the permanent floods, to a 22 to 41% savings in water with the 20% moisture regimes. The amount of water saved had no effect on chlorosis or injury observed by any of the treatments or water regimes. Propanil-susceptible barnyardgrass control was generally >96% for herbicide programs containing quinclorac or clomazone. Water management, rather than herbicide treatment, influenced rice yield more. No significant differences were observed in predicted rice yield among the permanent flood, 48, and 41% moisture regimes. The 20% moisture regime, however, gave a severe reduction in rice yield. 156

2 B.R. Wells Rice Research Studies 2002 INTRODUCTION Rice production in the United States has relied on the use of irrigation from rivers, reservoirs, and underground aquifers to provide a flood for rice production. In 2002, 1.4 million acres of rice were planted in Arkansas, which comprised 40% of the United States total production 1. Numerous crops are grown in Arkansas, including cotton, corn, soybean, and rice. All of these crops require season-long irrigation or timely rainfall to be a profitable commodity in Arkansas. Irrigation water in Arkansas has been provided mainly through irrigation wells connected to two main aquifers, the Alluvial aquifer and the Sparta/Memphis sand aquifer (Arkansas Soil and Water Conservation Commission, 1995). In 1965, the average amount of water withdrawn from the Alluvial was 1067 gallon days -1 and from the Sparta/Memphis sand aquifers was 192 million gallons day -1, respectively (Arkansas Soil and Water Conservation Commission, 1995). In 1994, 4221 and 192 million gallons day -1 were withdrawn from the Alluvial and Sparta/Memphis sand aquifer, respectively (Arkansas Soil and Water Conservation Commission, 1995). The increased use over these 30 years equates to a 3.5 times increase in irrigation water withdrawn, primarily due to increased irrigation in the Mississippi Delta region of Arkansas. The main commodity using this irrigation water in the Mississippi Delta is rice, which has been estimated to use about 70% of the total water needed for irrigation in the state of Arkansas (Scott et al., 1998). To improve water management efficiency, different approaches for irrigation in rice production have been tried: sprinkler irrigation, furrow irrigation, and delayed flooding and/or early removal of flood water. Sprinkler irrigation and furrow irrigation are widely used throughout the central and mid-south regions of the United States in soybean [Glycine max (L.) Merr.] (Wang et al., 2000; Purcell et al., 1997;), cotton (Gossypium hirsutum L.) (Pringle and Tupper, 1998; McConnell et al., 1999), and corn (Zea mays L.) (Paltineanu and Starr, 2000). Sprinkler irrigation research has shown a potential water savings of 50% compared to traditional rice-producing practices (Ferguson and Gilmour, 1977). Weed control, seeding rates under non-flooded condition, and the ability to provide adequate water at key times of rice growth are three key factors that have prevented this approach of irrigation from being adapted. Furrow irrigation is also being evaluated in rice production for its benefits in decreasing water usage and reduced pumping time, which relate to less fuel used, a longer life span of pumping equipment, and a decrease in irrigation labor (Chang et al., 1991). The disadvantages of furrow irrigation are similar to those of sprinkler irrigation, with increased weed competition due to lack of flood, the potential for delays in rice maturity, and the possibility of yield reductions (Smith and Fox, 1973). The objectives of this field research included reduction of water use through intermittent season-long irrigation and the potential for grass control using current soil-applied preemergence herbicides. The effects that intermittent irrigation and herbicides have on rice harvest index, milling quality, yield, and yield components will be determined through end-of-season measurements

3 AAES Research Series 504 GENERAL METHODS Field experiments were conducted in 2000, 2001, and 2002 at the Rice Research and Extension Center at Stuttgart, AR, on a DeWitt silt loam (fine, smectitic, thermic Typic Albaqualfs) that contains 1% organic matter and has a ph of 5.3. The experimental design was a split-plot with four replications in each year. The main-plot factor consisted of six water irrigation treatments with a subplot factor of six herbicide treatments and an untreated check. Main-plot dimensions were 7.5 by 18 m with subplot dimensions of 4.8 by 1.5 m. The rice cultivar Wells was drill-seeded in eight rows 25 cm apart at 135 kg/ha in plots with dimensions of 1.5 m by 4.8 m. Broadleaf signalgrass and propanil-resistant and -susceptible barnyardgrass were sown (two rows per species) perpendicular to the drilled rows of rice at the time of seeding to insure a weed infestation for rating. Main-plot treatments included: permanent flood and five bays represented by volumetric soil water content (θ v ) of 0.48, 0.41, 0.34, 0.27, and 0.20 m 3 water m 3 soil -1. Subplot treatments include: untreated check, clomazone 2 at 0.34 and 0.66 kg ai ha -1 applied preemergence (PRE), quinclorac at 0.42 kg ai ha -1 (PRE), pendimethalin at 1.12 kg ai ha -1 applied delayed preemergence (DPRE), thiobencarb at 4.48 kg ai ha -1 DPRE, and a standard comparison treatment of propanil at 4.48 kg ai ha -1 applied early postemergence (EPOST) followed by (fb) propanil at quinclorac at 0.42 kg ha -1 preflood. Due to the potential degradation of the soil-applied herbicides, fenoxaprop + isoxadifen was applied as needed up to panicle initiation if weed control fell below 80% for a given grass species. All irrigation treatments were flushed at the same time if deemed necessary for rice growth up until the establishment of the permanent flood which was at a 5- to 6-leaf stage of rice growth. At this time all main plots were flooded to an 8-cm depth, and the amount of water was recorded for each individual bay using a 10-cm McCrometer propeller flowmeter (Irrigation Flow Meter, Hemet, CA). The main-plot treatments included a conventional standard in which the permanent flood was maintained seasonlong, represented by θ v of 51%. The five remaining bays were flooded to an initial depth of 8 cm at the same time, and the flood water was allowed to dissipate naturally through evapotranspiration. These five bays were then reflooded using five different soil volumetric water contents to determine when the reflooding would occur. When the targeted volumetric soil water contents had been reached, an additional two-day waiting period was implemented before the bay was reflooded to mimic the time required for flooding on a large-scale basis. On average, it takes at least 2 to 3 days for a grower to achieve a permanent flood on half or more of a typical rice field. This pattern of flooding and dissipation continued throughout the growing season until maturity. Visual estimates of injury, rice chlorosis, and rice growth reduction were made 7, 14, 21, 28, 35, 42, 56, 72, and 95 days after emergence (DAE). Efficacy ratings were taken at these timings for control of propanil-resistant and -susceptible barnyardgrass and 2 Command 3 ME was used for the applications of clomazone. 158

4 B.R. Wells Rice Research Studies 2002 broadleaf signalgrass. Control was rated on a scale of 0 to 100%, with 0 = no control and 100 = complete control. Prior to mechanical harvest, ten panicles were harvested from randomly chosen plants in each plot. These panicles were used to evaluate total florets panicle -1 and the number of fertile and infertile florets panicle -1. Rice heights were also measured from five randomly chosen plants in each plot. Measurements were taken from the soil surface to the last floret of the upright panicle. Harvest index was calculated using a 1- m row of rice from one of the four center rows. Harvest index consisted of the economic yield (oven-dried rice florets) divided by biological yield (oven-dried vegetative growth) (Gardner et al., 1985). Milling quality was also taken using two 1-m row samples from the center rows of each plot. Grain was dried at 25 C for 12 hours, cleaned mechanically, and evaluated for milling quality at Riceland Foods at Stuttgart, AR. All grain yields were adjusted to 12% moisture for calculating final yield. Weed control data were subjected to analysis of variance with yield and other cropping parameters subjected to linear and non-linear regression, with 95% confidence intervals computed for each cropping parameter. RESULTS AND DISCUSSION In this experiment, soil moisture did not have an effect on weed control and only the main effects and interaction of year and herbicide were significant. Barnyardgrass control over a 3-year period was consistently greater than 98% 28 DAE with the soilapplied applications of quinclorac and clomazone at 0.67 kg/ha (Table 1). This control was attained prior to flood establishment or the flooding cycles of the intermittent irrigation system. The four remaining herbicide treatments gave less consistent control over the three-year period with control ranging from 60 to 100% 28 DAE. At 70 DAE, approximately 40 days after intermittent irrigation had been implemented, control was greater than 96% for treatments that used quinclorac PRE or PREFLD and clomazone at 0.67 kg/ha PRE. The three remaining treatments generally had to have an over-the-top application of fenoxaprop + safener each year to maintain control to an acceptable level, 84 to 100%. This option of using an effective postemergence application gave greater than 90% control, three-fourths of the time and 80 to 90% control one-fourth of the time. This research demonstrates that, given the current herbicide technologies available on the market, weed control can be attained in an intermittent irrigation system, with moisture having little effect on weed control. Intermittent irrigation reduced the total water usage when compared to the current production method of maintaining a permanent flood season-long, represented by the 51% moisture regime (Table 2.). Water usage followed a linear relationship over the 3-year period with a year-by-moisture interaction observed along with the main effects of year and moisture. The 48% moisture regime gave a 2.2 to 4.0% reduction in total water usage compared to the 51% moisture regime, although this was not statistically different from the 51% moisture regime. The 41% moisture regime, however, gave a significant reduction in water over the 3 years, with savings ranging from 7.5 to 13.3% 159

5 AAES Research Series 504 compared to the continuous flood. When the soil was allowed to dry to a 20% moisture level between flooding, a 23.3 to 41% reduction in water usage could be observed, depending on the year. Other research in irrigation systems has shown a savings in water; however, these savings usually resulted in a direct or indirect yield loss. In this research no significant differences were observed between predicted yields for the moisture regimes of 41 to 51% for all 3 years, regardless of herbicide treatment (Table 3). Allowing the soil to dry to a moisture level less than 27% gave a severe reduction in rice yield. Even though the 20 and 27% moisture regimes gave the highest reductions in water usage, the reduced yields more than offset the savings in water, therefore making intermittent irrigation systems at these lower moisture regimes impractical for the grower. Water-use efficiency (WUE), which is the ratio of total rice yield to the amount of water used to achieve that yield followed a non-linear regression relationship. When rainfall is included with the total irrigation water used, no differences were observed among moisture regimes for the predicted values for clomazone at 0.67 kg/ha over the 3-year period (Table 4). When evaluating our five remaining herbicides, the 34 and 41% moisture regimes were shown to give the highest predicted WUE; however, this was similar to WUE of the continuous permanent flood, represented by 51%. The WUE for the 20% moisture regime was less than the 34 to 51% moisture regime. ACKNOWLEDGMENTS These studies were supported in part by rice growers checkoff funds through the Arkansas Rice Research and Promotion Board. We gratefully acknowledge the men and women at the Rice Research and Extension Center for their help and guidance in these studies. LITERATURE CITED Arkansas Soil and Water Conservation Commission Groundwater protection and management report December. Chang, C.C., B.R. Eddleman, and B.A. McCarl Potential benefits of rice variety and water management improvements in the Texas gulf coast. West. J. Agric. Econ. 6: Ferguson, J.A. and J.T. Gilmour Center-pivot sprinkler irrigation of rice. Ark. Farm Res. 26(2):12. McConnell, J.S., E.D. Vories, D.M. Oosterhuis, and W.H. Baker Effect of irrigation termination on the yield, earliness, and fiber qualities of cotton. J. Prod. Agric. 12: Paltineanu, I.C. and J.L. Starr Preferential water flow through corn canopy and soil water dynamics across rows. Soil Sci. Soc. Am. J. 64:

6 B.R. Wells Rice Research Studies 2002 Pringle, H.C. III and G.R. Tupper Cotton yield response to sprinkler irrigation and in-row subsoiling. Proc. Beltwide Cotton Conf. 1:682. National Cotton Council of America, Memphis, TN. Purcell, L.C., E.D. Vories, P.A. Counce, and C.A. King Soybean growth and yield response to saturated soil culture in a temperate environment. Field Crop. Res. 49: Scott, H.D., J.A. Ferguson, L. Hanson, T. Fugit, and E. Smith Agricultural water management in the Mississippi delta region of Arkansas. Ark. Agric. Exp. Stat. Res. Bull. 959:98. Smith, R.J. Jr. and W.T. Fox Soil water and growth of rice and weeds. Weed Sci. 21: Wang, D., M.C. Shannon, C.M. Grieve, and S.R. Yates Soil water and temperature regimes in drip and sprinkler irrigation, and implications to soybean emergence. Agric. Water Manag. 43: Table 1. Propanil-susceptible barnyardgrass control in an intermittent irrigation system in 2000, 2001, and 2002 at Stuttgart AR. Propanil-resistant barnyardgrass control Application 28 DAE y 70 DAE Herbicide treatment z timing (%) Clomazone 0.34 kg/ha PRE x Clomazone 0.67 kg/ha PRE Quinclorac PRE DPRE Thiobencarb DPRE Propanil fb EPOST propanil + quinclorac PREFLD LSD (0.05) comparison within year 3 3 LSD (0.05) comparison across years 3 4 z Fenoxaprop + safener was applied once over-the-top at a preflood or 1 week postflood timing to pendimethalin and thiobencarb in 2000, 2001, and 2002 along with clomazone at 0.34 kg/ha in y DAE: days after emergence. x PRE = preemergence; DPRE = delayed preemergence; EPOST = early-postemergence; and PREFLD = preflood. 161

7 AAES Research Series 504 Table 2. Regression equations, predicted values, and 95% confidence intervals for total water usage as affected by intermittent irrigation in 2000, 2001, and 2002 at Stuttgart, AR. Moisture level (θv) y Year Regression equation z Confidence interval [Total water usage (cm)] Y = X Upper Predicted Lower Y = X Upper Predicted Lower Y = X Upper Predicted Lower z Regression equations intercepts and slopes were combined across year when t-test analysis allowed. y Moisture (θv): The volumetric water content in the soil during the non-flooded portion of the intermittent irrigation system with the permanent flood represented by 51%. 162

8 B.R. Wells Rice Research Studies 2002 Table 3. Regression equations, predicted values, and 95% confidence intervals for rice yield for 2000, 2001, and 2002 at Stuttgart, AR. Regression Confidence Moisture level (θv) y Year Herbicide z equation interval [Rice Yield (kg/ha)] Clomazone 0.67 kg/ha Y= X X 2 Upper Predicted Lower Clomazone 0.34 kg/ha Y= X X 2 Upper Thiobencarb Predicted Quinclorac Lower Clomazone 0.67 kg/ha Y= X X 2 Upper Predicted Lower Clomazone 0.34 kg/ha Y= X X 2 Upper Thiobencarb Predicted Quinclorac Lower Clomazone 0.67 kg/ha Y= X X 2 Upper Predicted Lower Clomazone 0.34 kg/ha Y= X X 2 Upper Thiobencarb Predicted Quinclorac Lower z Herbicide intercepts and slopes were combined within year when t-test analysis allowed. y Moisture θv: The volumetric water content in the soil during the non-flooded portion of the intermittent irrigation system with the permanent flood represented by 51%. 163

9 AAES Research Series 504 Table 4. Regression equations, predicted values, and 95% confidence intervals for water use efficiency including rainfall in 2000, 2001, and 2002 at Stuttgart, AR. Regression Confidence Moisture level (θv) y Year Herbicide z equation interval [WUE x (kg/ha/cm water)] Clomazone 0.67 kg/ha Y= X X 2 Upper Predicted Lower Clomazone 0.34 kg/ha Y = X X 2 Upper Thiobencarb Predicted Quinclorac Lower Clomazone 0.67 kg/ha Y= X X 2 Upper Predicted Lower Clomazone 0.34 kg/ha Y = X X 2 Upper Thiobencarb Predicted Quinclorac Lower Clomazone 0.67 kg/ha Y= X X 2 Upper Predicted Lower Clomazone 0.34 kg/ha Y = X X 2 Upper Thiobencarb Predicted Quinclorac Lower z Herbicide intercepts and slopes were combined within year when t-test analysis allowed. y Mositure v: The volumetric water content in the soil during the non-flooded portion of the intermittent irrigation system with the permanent flood represented by 51%. x WUE = water use efficiency. 164