Denitrification Losses from Kentucky Bluegrass Sod

Denitrification may represent an important mechanism in the fate of N applied to turf. Denitrification losses were directly measured from fertilized 'Baron' Kentucky bluegrass (Poa pratensis L.) sod samples sealed in acrylic chambers using the acetylene inhibition technique. Losses were correlated with soil texture, percent soil sat· uration (SAT), and temperature. Losses from turf on a Hadley silt loam soil and Hadley silt soil (both coarse-silty, mixed, nonacid, mesic Typic Udifluvents) incubated at 2rC did not exceed 0.4 and 0.1%, respectively, of the applied potassium nitrate fertilizer (4.5 g N m-2) when soil water levels were less than 75% saturated. Soil saturation increased denitrification losses from the silt loam and silt soils to 2.2 and 5.4% of the applied N, respectively. The relationship between percent soil saturation and denitrification loss was quadratic and highly significant for both soils. The equations are: milligrams of N20-N m2 10 d' = 1432.50 38.96 (percent SAT silt soil) + 0.26 {percent silt soil)2 or 130.80 5.40 (percent SAT silt loam soil) + 0.05 (SAT silt loam soil)l. A linear relationship [milligrams of N20 m-2 10 d-' = 0.49(°C) 9.70] existed between denitrification losses and soil temperatures between 22 and 30°C in the silt soil at 75% of soil saturation. Soil temperatures of 30°C or greater coupled with saturated soil conditions resulted in the greatest losses, equiv­ alent to 44.6 and 92.6% of the applied N to the silt loam and silt soils, respectively. Denitrification losses did not increase at soil tem­ peratures above 30°C. These results indicate that denitrification loss from fertilizers applied to turfgrasses may not be a serious problem unless the soils are saturated and at higher soil temperatures. Additional index words: Nitrous oxide, Nitrogen fertilizer, Poapra­ tensis 1., Turfgrass. D ENlTRIFICATION loss after N fertilizer applica­ tions to field and vegetable crops (1,5,15,20) and grasslands (4,5,7,12,14,21) has been studied exten­ sively. No research data is available concerning de­ nitrification losses from t ighly managed turfgrass areas. Conditions that may promote denitrification losses include light, frequent irrigations (4,5,12,14), fertilizer N applications (4,14,21), incorporation of organic matter into soils (4,12), and higher soil temperatures (4,7,10). These conditions are found in high-quality turfgrass areas, with the organic matter being contrib­ uted by the turfgrass root system. Soil temperatures under closely mowed turf may be high (2). Denitrification losses from irrigated vegetable crops can be as high as 233 kg N ha' yrI (52% of applied N)(14,15). Losses from small grains range from 1 to 19 kg N ha' ye l (1,5,20). Perennial ryegrass (Lolium C.F. Mancino, Dep. of Plant SeL, Univ. of Arizona, Tucson, AZ 85721; W.A. Torello, Dep. of Plant and Soil ScL, Univ. of Massa­ chusetts, Amherst, MA 01003; D.J. Wehner, Horticulture Dep., Univ. of Illinois, Urbana, IL 61801. Journal Paper no. 2841 ofthe Massa­ chusetts Agric. Exp. Stn., Amherst, "'Corresponding author. Re­ ceived II May 1987. Published in Agron. J. 80:148-153(1988). perenne L.) swards may lose from 1 to 29 kg N haI yrI (12,14). Losses from the grasses are small despite N applications as high as 500 kg N ha' yr1 (14). Although peak effluxes from soils under grasses and other crops may be similar in magnitude, the duration is shorter in the presence of grasses due to the grass root system's ability to quickly deplete soil N03'-N levels (14). Turfgrass root systems can be shallow due to light, frequent irrigation and close mowing. Peak N20-N effluxes (20-50 mg N m-2 dI) tend to occur during the first few days after an irrigation or rainfall event (6,17,18). These effluxes are associated with wetter soil conditions (more than 20% w/w) and higher N03'-N levels (more than 5 IJ,g N per gram of soil) (14). These losses are short lived but can be sig­ nificant in terms of their contribution to total annual denitrification loss. Soil water and texture influence the denitrification process by affecting gas exchange and the diffusion of O2 to active microbial sites (12,17,21). Small, frequent irrigations have been found to stimulate denitrifica­ tion (6,7,18), while larger, more frequent irrigations may not (16). A small water input into a finer textured soil may result in greater denitrification losses than from coarser textured soils (18,21). Increased soil temperature increases denitrification rates (3,4,10). Many factors, however, influence the quantitative effects of temperature on denitrification rate. These include diversity and availability of soil organic matter, and the selective effects oftemperature on different bacterial species and denitrifying popu­ lations (8,9). Bremner and Shaw (3) found that max­ imum soil denitrification rates occurred in vitro at 30°C in six soils amended with glucose and N03'-N. The objectives of this study were to determine total denitrification losses from Kentucky bluegrass sod as influenced by soil texture, soil water content (percent saturation), and soil temperature. MATERIALS AND METHODS 'Baron' Kentucky bluegrass sod samples (30.5 by 30.5 by 7.5 cm) from Hadley silt and Hadley silt loam sites (Table 1) were installed into clear acrylic chambers constructed of 1.3-cm-thick clear acrylic, and had internal dimensions of 30.5 by 30.5 by 13.0 cm. Each chamber was airtight when its lid was sealed on using a lithium paste. Five perforated copper tubes were installed into the bottom ofeach chamber, which permitted the injection of acetylene (CZH 2) into the sod soils through a septum-sealed sidearm that extended through the chamber wall. The first and second experiments described in this study were conducted in a growth room while the third experiment was done in growth chambers. 149 MANCINO ET AL.: DENITRIFICATION FROM KENTUCKY BLUEGRASS Effect ojSoil Water Content The initial soil water content of the silt soil was 91.7% of soil saturation (48.4% water, wIw). Soil water levels were determined gravimetrically. Six sod samples were allowed to air dry until the soil was approximately 50% saturated. Reagent grade potassium nitrate (KNO),3.02 g per sample) was dissolved in 1.5 L of water and applied to each of two samples. Final soil water content of these samples was 100% of saturation. Six hundred milliliters of water were used to apply the same amount ofKNO) to two more samples, bring­ ing the soil water contents to approximately 75% of satu­ ration. The remaining two samples were allowed to dry to a soil water content of 37% of saturation before 400 mL of water containing KNO) was applied, yielding a final soil water content of approximately 50% of soil saturation. Following N fertilizer and water applications, the chamber lids were sealed onto each chamber. One hundred milliliters of C2H2 were injected into each box, giving a final soil C2H2 concentration of approximately 1% (vIv). The chambers re­ mained sealed for 12 h before headspace gas samples were removed. The chambers were then opened and fresh air was circulated around the sod and through the thatch using a hand-held vacuum cleaner. The chambers were again sealed, injected with C2H 2, and allowed to stand for another 12 h. This process was repeated for 10 d, with 12-h intervals cor­ responding to light (19.1 W m-)/dark periods. Soil tem­ perature was maintained at 22 c C. After sampling, the headspace gas was analyzed for N20­ N content using a Varian Series 2700 moduline gas chro­ matograph (Varian Associates, Inc., Palo Alto, CA) equipped with an electron capture detector (TPH). Nitrous oxide was separated from other gases using a Varian Porapak Qcolumn (80/100 mesh, 2 m by 3.2 mm). Column temperature was 75°C, injector port temperature was 195°C, and detector temperature was 220 C. The carrier gas was N2 at a flow rate of 14 mL min-t. By utilizing the acetylene-inhibition technique, N20-N loss represented total denitrification loss (16). After 10 d the sod samples were replaced with six fresh samples from the silt soil site. Three additional samples were also used for a total of nine sod samples. Initial soil water content of the sad was 66% of soil saturation. Potassium nitrate was applied in quantities of water ranging from 400 mL to 2 L, with treatments varying by 200-mL increments. Final soil water contents ranged from 72 to 106% of soil saturation. The above study was repeated with nine sod samples from the silt loam soil site (Table I). Initial soil water content was 93.5% of saturation (40% water, w/w). After soil water levels fell to approximately 68% of soil saturation, seven sod sam­ ples received KNO) (3.02 g per sample) applied in water quantities ranging from 400 to 900 mL. The two remaining sod samples were dried to 45% of soil saturation and the fertilizer was applied in 500 mL of water. Final soil water levels ranged from 33 to 10I% of soil saturation. Soil tem­ perature was maintained at 22 c C. Total denitrification loss was calculated on a daily basis as well as on a lO-d basis. Within each soil type, a stepwise regression analysis was used to test for a significant rela­ tionship between total denitrification loss from KN03-N treated sod samples and the level of soil saturation. Controls were not used during this study. Preliminary results had shown that total denitrification loss from nonfertilized sod incubated at 22°C was insignificantly low, and often unde­ tectable, even under saturated soil conditions. Effect of Temperature In this study, six sod samples from the silt soil were in­ stalled in the acrylic chambers and incubated at 22 C. Initial soil water content was 66% of soil saturation. Potassium Table 1. Characteristics of Hadley silt and Hadley silt loam.