Timing of manure application may affect soil N content, crop N use efficiency and crop performance. For example, Hanna et al. (2000) reported that spring application of manure to corn provides the best agronomic response in most instances and manure N use efficiency increases with spring application compared with fall application. However, Loecke et al. (2004) suggested that fall application of solid swine (Sus scrofa) manure to corn increased available N and improved corn yield compared with spring application. In addition, Loecke et al. (2004) reported that fall applied manure produced higher corn grain yields than did spring applied manure due to more timely net N mineralization relative to plant N demand with fall application.
While spring application of manure provided the best agronomic response in most instances, manures are often applied to the soils in the fall, possibly because of greater availability, lower cost, more time to apply and busy spring planting schedules for farmers. In addition, weather risks of applying manure in the spring which allows small windows of opportunity for planting and is less favorable for land preparation (Ruiz Diaz and Sawyer, 2008). In Mississippi, due to wet winter months and in most cases early spring, row crop farmers are forced to apply broiler litter in the fall to prevent delayed manure over a winter fallow is risky because considerable amount of the litter N can be lost to leaching rendering the litter less effective for spring-planted crops. Fall litter application is also associated with potential environmental contamination (Smith and Chambers, 1998; Andraski and Bundy, 2005; Hansen et al., 2004; van Es et al., 2004). Loss of N derived from litter occurs both as leaching and denitrification of NO3–N (Paul and Zebarth, 1997). Owens et al. (1995) reported that most of the N loss occurs as NO3–N leaching during the late fall and early spring months when the soil is fallow (Owens et al., 1995).
The capability of applying litter in the fall is vitally important to row crop growers and poultry producers in the Mid-South and southeastern United States. Many apply broiler litter in the fall as a source of P and K, and then apply a full N rate from inorganic sources in the spring. This practice can lead to nutrient accumulation in the soil and adversely impact the environment. Therefore, management practices that prevent or minimize loss of broiler litter-derived N and enhance the cycling of N applied during the off-season for use by subsequent summer crops offers great practical benefit to growers.
Cover crops have been promoted as a means of maximizing the efficient use of available N to subsequent crops in agricultural systems, potentially enhancing profitability through reduced inorganic fertilizer N requirement (Dekker et al., 1994; Shipley et al., 1992). The use of winter cover crops to capture inorganic N and to reduce leaching losses by recycling NO3–N, especially in areas that have high levels of winter and spring precipitation, is receiving renewed attention. During a 3-yr period, NO3–N leaching losses under the winter-fallow treatment averaged 20 kg N ha−1 yr−1 compared with 1.2 kg N ha−1 yr−1 under the winter cover cropped treatment, when using an ammonium nitrate fertilizer for the preceding corn (McCracken et al., 1994). Winter rye cover crops have been shown to increase the efficient use of available N to subsequent crops in the spring, resulting in reduced NO3–N leaching loss while potentially enhancing profitability through reduced inorganic fertilizer N requirements (Shipley et al., 1992; Dekker et al., 1994; Kowalenko, 1987). The scavenging ability of cover crops, particularly rye, to sequester and reducing N leaching loss has been documented. Wayland et al. (1996) reported that a winter cover crop reduced leached N 65 to 70% compared with winter-fallow soil that received animal manure. Davis and Garwood (1996) found that a rye winter cover crop reduced leaching loss by more than 90% as compared with winter-fallow. McCraken et al. (1994) reported that NO3–N concentration of leachate was almost zero during the fall, winter, and early spring with a rye winter cover crop. Logsdon et al. (2002) showed that rye cover crops in corn–soybean rotation reduced nitrate losses more than 70%. In all of these studies, winter rye cover crop was used to deal with residual N left in the soil from previous spring application of organic or inorganic N fertilizer to row crops. However, the effect of cover crop followed with fall-applied broiler litter on NO3–N sequestration has not been documented. We hypothesized that seeding of a cover crop, such as winter rye, following fall applied broiler litter is an effective practice in reducing leaching losses of litter-derived N during fall, winter, or early spring. The objective of this study was to evaluate the effects of cover crop (winter rye vs. winter fallow) along with broiler litter fertilization timing (fall vs. spring) on cotton, yield, leaching loss, and residual soil NO3–N.
MATERIALS AND METHODS
A 2-yr study was conducted on Leeper silty clay loam soil in 2007–2008. In fall 2006, initial soil samples were taken at the 0- to 15-cm depth and analyzed for physical and chemical characteristics (Table 1 ). Soil pH were determined on 1:1 soil/water suspension with a glass electrode (pH/EC/TDS meter model H19813–0; Hanna, Woonsocket, RI). Total C and total N for soil were determined from air-dried, finely ground soil using an automated dry combustion C/N analyzer (Model NA 1500 NC; Carlo Erba, Milan, Italy). Soil organic matter was calculated from TC (percent of organic matter = TC × 1.72). Soil texture was determined by the method of Day (1965) Soil NO3–N and NH4–N were determined by extracting soil with 2 M KCl (Keeney and Nelson, 1982) and analyzed for inorganic N (NH4 and NO3) using a Lachat instrument (QC 8000 fl ow injection analyzer; Lachat, Loveland, CO). These samples were also extracted using Mehlich-III extractant (Mehlich, 1984) and analyzed for plant-available P, K, Ca, Mg using inductively coupled argon plasma spectrophotometer ICAP 9000 (Thermo Jerrell Ash, Franklin, MA). Soil bulk density was determined using procedure described by Grossman and Reinsch (2002) A typical double-cylinder, hammer-driven core sampler was used for obtaining soil samples for bulk density. A 15-cm ring was placed inside the sampler, inserted by force into the top 0- to 15-cm depth and the whole core was taken. Samples were weighed and oven-dried at 105°C and reweighed again to get a constant weight (Ws). Th e volume of soil assumed equivalent to the volume of the ring that contained soil. Soil Db was calculated by dividing the mass of soil by the volume (Db = Ws/Vs), where Ws = weight of dry soil (g) and Vs. = soil volume (cm3). The experimental design was a randomized complete block design with a split-split plot treatment arrangement and four replications. Main plots consisted of cropping system (cover crop vs. no cover crop), subplots were time of application (fall vs. spring), and sub-subplots were broiler litter rates of 0, 4.5, 9, and 13.4 Mg ha−1 and inorganic N applied only in spring at the recommended rate determined by Mississippi Soil Testing Laboratory. The plots were 4 m wide and 9 m in length with a 3 m alley between the blocks. For the fall application, broiler litter was applied to the soil on 4 Oct. 2006 and 5 Oct 2007 and incorporated into the surface 15 cm within 1 h of application. Winter rye (Secale cereal L.) cultivar Elbon was planted on 5 Oct. 2006 and 6 Oct. 2007 as the cover crop treatment. The rye was drilled at seeding rate of 100 kg ha−1 with a row spacing of 17 cm.
|Organic matter, g kg−1||10.6|
|NH4–N, mg kg−1||3.7|
|NO3-N, mg kg−1||14.1|
|P, mg kg−1||40.3|
|K, mg kg−1||168|
|Ca, mg kg−1||2500|
|Mg, mg kg−1||100|
|Cation exchange capacity, cmolc kg−1||24|
|Bulk density, g cm−3||1.28|
|Sand, g kg−1||10|
|Silt, g kg−1||40|
|Clay, g kg−1||50|
Broiler litter used in the research was obtained from a broiler chicken (Gallus gallus) producer in central Mississippi. Broiler litter for the spring application was taken from the same load used for the fall application. Litter from the load used for fall application was placed in black plastic bags, closed tightly, and stored in a cooler until spring application. Each year, broiler litter samples were taken at the time of application and moisture and nutrient contents of the litter determined (Table 2 ). Because litter in 2008 had less moisture than 2007, the total amount of N available from litter was approximately 25 kg ha−1 greater in 2008.
|Year||Applied litter||Moisture||Total C||Total N||NH4–N||Total P||Total K||C/N ratio||pH|
|Amount applied||Mg ha−1||Mg ha−1||kg ha−1|
To monitor N leaching, after broiler litter application and planting rye, one suction lysimeter was installed at 60-cm depth in each of the plots receiving 0, 4.5, 9, and 13.4 Mg broiler litter ha−1 for winter rye and winter fallow, in two selected blocks for a total of 16 lysimeters. Leachate samples were collected 24 h after each rainfall event during late falls, winter and early spring in both 2006–2007 and 2007–2008. Leached water samples were kept frozen until analysis for NO3–N concentrations. In each year, the quantity of aboveground winter rye biomass was determined by randomly harvesting two 0.5-m2 areas in each plot approximately 6 wk before planting cotton. After weighing and mixing the samples thoroughly, a subsample (≈100 g fresh wt.) was collected for determinations of dry matter yield and N concentrations. Samples were dried at 65°C for 3 d in a forced-air oven, weighed for dry matter production, ground to pass a 1-mm sieve and analyzed for total N by an automated dry combustion method using a ThermoQuest (CE Elantec, Inc., Lakewood, NJ) C/N analyze. After sampling, winter rye cover crops were chemically desiccated using glyphosate [N-(phosphonomethyl) glycine] at an application rate of 0.9 kg a.i. ha−1 in solution with 68 L water ha−1 After the rye was fully desiccated, soil samples were collected from both the winter rye and winter fallow plots to a depth of 30 cm. The samples were air-dried, ground to pass a 2-mm screen, extracted with 2 M KCl (Keeney and Nelson, 1982) and analyzed for NO3–N using a Lachat instrument (QC 8000 flow injection analyzer; Lachat, Loveland, CO).
In the spring, broiler litter was applied on 1 May 2007 and 4 May 2008 at the same rate applied in the fall and incorporated immediately. Cotton cultivar ‘DP 444 BG/RR’ was planted on 96-cm rows at approximately 65,000 seed ha−1 using a conventional four-row planter on 2 May 2007 and 5 May 2008. Dates of selected field operations in each year are presented in Table 3 To monitor plant N status, 20 fully expanded recently mature leaves from the main stem in each plot were taken at peak bloom for N concentrations. Before taking the leaves from the plant, chlorophyll meter readings were taken in each plot using a Minolta hand-held SPAD-502 chlorophyll meter (Minolta, Ramsey, NJ). At the beginning of boll opening, number of bolls was measured from a 1 m2 area.
|Operation||Crop||Fall 2006||Spring 2007||Fall 2007||Spring 2008|
|Litter application||4 Oct.||1 May||5 Oct.||4 May|
|Planting||cover crop||5 Oct.||–||7 Oct.||–|
|cotton||–||2 May||–||4 May|
|Killing||cover crop||–||28 Mar.||–||3 Apr.|
|Defoliation||cotton||–||9 Sept.||–||8 Sept.|
|Harvest||cotton||–||24 Sept.||–||28 Sept.|
Cotton was defoliated with 1590 mL ha−1 of Finish, contains two active ingredient, including cyclanilide [1-(2,4-dichlorophenylaminocarbonyl)-cyclopropane carboxylic acid] and ethephon (2-chloroethylphosphonic acid), on 9 Sept. 2007 and 8 Sept. 2008 and harvested on 20 Sept. 2007 and 21 Sept. 2008 by picking the full length of the middle two rows using a spindle-picker. A subsample of seed cotton from each plot was taken (≈1.0 kg) for measuring ginning percentage. The samples were weighed before and after air-drying to constant weight in paper bags and ginned in a 10-saw tabletop microgin to determine lint percentage. Lint yields were determined by weighing lint and seed collected from each plot and multiplying corresponding seed cotton by the ginning percentage of each plot.
After picking, the cotton stalk were shredded and postharvest soil samples were taken to a depth of 90 cm from each sub-subplot to determine residual soil NO3–N concentration. Soil samples consisting of five 5-cm diam. soil cores were randomly taken from 0 to 15, 15 to 30, 30 to 60 and 60 to 90 cm. Soil samples were combined by depth, placed in plastic bags and kept frozen until analyzed. Soil NO3–N was determined by extracting soil samples with 2 M KCl (Keeney and Nelson, 1982).
All response variables within each year including lint yield, chlorophyll reading, and leaf N concentration, number of bolls, lint percentage, and residual and leaching loss of NO3–N were analyzed by subjecting the data to statistical analysis using MIXED model analysis of SAS (Littell et al., 2006). Both cropping system and broiler litter fertilization timing were used as the fixed factors and replication and its interaction were used as random effect factors. Since the effects of year and year × rate interaction were significant for all dependent variables (Table 4 ), the effect of treatments in each year was evaluated separately. A significance level of P < 0.05 was established.
|Effect||df||Lint yield||Boll number||Lint turnout||Leaf N||Chlorophyll Index||Soil NO3–N||Leached NO3–N|
|P > F|
|Cover crop (CC)||1||0.0037||0.2041||0.1495||0.2788||<0.0001||0.0034||<0.0001|
|Yr × CC||1||0.8176||0.7870||0.1085||0.5565||<0.0001||0.1651||0.4315|
|Yr × MT||1||0. 4114||0.7870||0.0954||0.0086||0.0240||–||–|
|CC × MT||1||<0.0001||0.0033||0.0021||<0.0001||<0.0001||–||–|
|Yr × CC × MT||1||0.3930||1.000||0.1156||0.0050||<0.0001||–||–|
|Manure rate (MR)||3||<0.0001||<0.0001||<0.0001||<0.0001||<0.0001||0.0031||<0.0001|
|Yr × MR||3||0.0002||0.0665||0.0261||<0.0001||<0.0001||<0.0001||<0.0001|
|CC × MR||3||<0.0001||0.0058||0.4072||<0.0001||0.0004||0.0022||<0.0001|
|Yr × CC × MR||3||0.7758||0.8376||0.1994||0.0646||0.0578||0.0834||0.2673|
|MT × MR||3||<0.0001||0.0426||0.0508||<0.0001||<0.0001||–||–|
|Yr × MT × MR||3||0.3796||0.6901||0.2572||0.0043||<0.0001||–||–|
|CC × MT × MR||3||<0.0001||0.01189||0.0911||<0.0001||<0.0001||–||–|
|Yr × CC × MT × MR||3||0.9126||0.6975||0.7195||0.1988||<0.0001||–||–|
RESULTS AND DISCUSSION
The period from broiler litter application in October 2006 until cotton planting in May 2007 was warmer and drier than the 30-yr average (Fig. 1a and 1b ). The late fall in 2007 and winter in 2008, on the other hand, was relatively colder and wetter than the normal. Mean monthly temperature in 2007 cotton growing season (from May–October) was similar to the average except in May and August in which the temperature was warmer than normal. In 2007, the monthly total precipitation during the cotton-growing season was less than the normal. In 2008, mean monthly temperature during cotton growing season in 2008 was similar to the normal, except in October in which the weather was colder than the normal. The 2007 cotton growing season had a different precipitation pattern than the 2008 season. The monthly mean total precipitation during cotton growing season in 2007 was lower than 30-yr average except in September in which the precipitation was greater than normal. In contrast, the 2008 cotton growing season was wetter than normal except in June during which soil was drier than normal.
Cover Crop Yield and Nitrogen Utilization
Winter rye cover crop dry matter yields increased (P < 0.001) with increasing broiler litter rates and quadratic equations best described the cover crop dry matter response to increasing broiler litter applications in 2007 (Y = 828 + 440x −18.2x 2, r 2 = 0.98) and in 2008 (Y = 932 + 655x – 30.6x 2, r 2 = 0.96). At the highest broiler litter rate (13.4 Mg ha−1), winter rye cover crop dry matter yield in 2008 was 19% greater than in 2007 (Table 5 ). The higher dry matter yield in 2008 than 2007 was likely influenced by more precipitation in late fall and early spring, higher temperatures in March and April of 2008 than normal (Fig. 1). In addition, 17% more N was applied in 2008 than in 2007 due to greater broiler litter N concentration in 2008 than in 2007 (Table 2). In both 2007 and 2008, rye total dry matter yields were lower than those reported in literature (Vaughan and Evanylo, 1999; Odhiambo and Bomke, 2001). In Maryland, Clark et al. (1994) reported maximum winter rye biomass as cover crop in the range of 6400 and 7100 kg ha−1 Wagger (1989) reported winter rye cover crop desiccated 2 wk beyond full anthesis had an average total aboveground biomass of 8.2 Mg ha−1, low N concentration, high C/N ratio and supplied less total N into the system for subsequent crops. The reason for lower winter rye aboveground biomass productions in our study is that winter rye cover crop was desiccated before maturity (boot stage) (6 wk before planting cotton) to enhance the potential for net N mineralization during cotton growing season as C/N ratio was kept low. Our results are in agreement with others who reported that killing winter rye cover crop early in the spring would reduce biomass production and C accumulation, increase N concentration, reduce C/N ratio, and accelerate net N mineralization (Huntington et al., 1985; Evanylo, 1991). Huntington et al. (1985) reported that low C/N ratio would facilitate N mineralization during the subsequent crop growing season, increase available N for plant use, and reduce the dependence of the plant's N requirements from inorganic commercial N fertilizer. They showed that chemical desiccation of winter rye cover crop early in spring increased the potential release of cover crop N into the following crop. Synchronization of cover crop N mineralization with N uptake by the subsequent crop is important for efficient crop production. If release and uptake of N are not synchronized, yields may be reduced and risks of high postharvest soil NO3–N concentrations and potential leaching losses may be increased (Meisinger et al., 1991).
|Cropping system||Broiler litter rate||Dry matter yield||N concentration||N uptake|
|Mg ha−1||kg ha−1||g kg−1||kg ha−1|
Similar to cover crop dry matter yield, total N uptake in aboveground biomass of cover crop increased (P < 0.0001) with increasing broiler litter rates in a quadratic manner in 2007 (Y = 11.9 + 6.5x – 0.25x 2, r 2 = 0.97) and in 2008 (Y = 12.5 + 10.2x – 0.44x 2, r 2 = 0.97) but broiler litter had little effects on biomass N concentration (Table 4), indicating total N uptake by cover crop dry matter was primarily a function of dry matter production rather than dry matter N concentration. Averaged across year, cover crop N uptake ranged from 10.8 kg ha−1 for the control to 64 kg ha−1 for the 13.4 Mg ha−1 broiler litter application (Table 4). Winter rye cover crop taken up only 12% and 14% of total N applied with broiler litter at the highest rate of 13.4 Mg ha−1 in 2007 and 2008, respectively. Low winter rye N recoveries in our study was probably related to the low winter rye aboveground biomass, but could be due to the competition for N between plant and microbes. It has been reported that plant root exudation stimulate microbial activity, resulting in stimulated soil organic matter degradation and N immobilization by microbes (Griffiths and Robinson, 1992).
Cotton Lint Yield
Cotton lint yield was greatly affected by cropping system, broiler litter rates and timing as evidenced by a significant interaction of these factors in 2007 (P = 0.0006) and 2008 (P = 0.0001) (Table 6 ). In 2007, cotton lint yield increased with broiler litter applications as compared with the control (Table 7 ). However, no differences in cotton lint yield was obtained among broiler litter rates, except in the absence of cover crop in which spring-applied broiler litter at the rate of 9 Mg ha−1 had the greatest cotton lint yield as evidence by broiler litter timing × litter rate interaction (Tables 6 and 7). Averaged across cropping system and application timing, cotton lint yield increased with increasing broiler litter rates and a quadratic equation best described the cotton yield response to increasing broiler litter applications (Y = 697 + 133x – 6.7x 2, r 2 = 0.97, P < 0.001). In this equation, the rate of litter that resulted in predicted peak lint yield was calculated by setting the first derivative of the equation to zero and solving for the litter rate. Based on this calculation, the predicted lint yield peaked when cotton was fertilized with 9.9 Mg ha−1 The yield-maximizing 9.9 Mg ha−1 litter calculated based on the fitted model was approximately similar to the actual highest-yielding 9.0 Mg ha−1 litter rate (1357 kg ha−1 vs.1291 kg lint ha−1)
|Effect||df||Lint yield||Boll number||Lint turnout||Leaf N||Chlorophyll Index||Leached NO3–N|
|P > F|
|Cover crop (CC)||1||0.0416||0.4927||0.6708||0.2481||0.0500||0.0647|
|Manure time (MT)||1||<0.0001||0.0083||<0.0001||<0.0001||<0.0001||–|
|CC × MT||1||<0.0001||0.0513||0.0071||<0.0001||<0.0001||–|
|Manure rate (MR)||3||<0.0001||<0.0001||<0.0001||<0.0001||<0.0001||<0.0001|
|CC × MR||3||0.0004||0.0872||0.9816||<0.0001||<0.0001||0.0022|
|MT × MR||3||<0.0001||0.2671||0.0566||<0.0001||<0.0001||–|
|CC × MT × MR||3||0.0006||0.4326||0.3754||<0.0001||<0.0001||–|
|Cover crop (CC)||1||0.0403||0.0897||0.0998||0.5254||<0.0001||0.0501|
|Manure time (MT)||1||<0.0001||<0.0001||0.0005||<0.0001||<0.0001||–|
|CC × MT||1||<0.0001||0.0009||0.1428||<0.0001||<0.0001||–|
|Manure rate (MR)||3||<0.0001||<0.0001||<0.0001||<0.0001||<0.0001||<0.0001|
|CC × MR||3||<0.0001||0.0002||0.0132||<0.0001||<0.0001||<0.0001|
|MT × MR||3||<0.0001||0.0560||0.8702||<0.0001||<0.0001||–|
|CC × MT × MR||3||0.0001||0.0317||0.1269||<0.0001||<0.0001||–|
|Cropping system||Timing||Fertilization||Boll no.||Lint turnout||Leaf N||Chlo.§ index|
|Mg ha−1||kg N ha−1||kg ha−1||%||g kg−1|
|P > |t||
|Fall vs. spring||<0.0001||0.0820||<0.0001||<0.0001||<0.0001|
|CC vs. NC||0.0427||0.4929||0.6708||0.2481||0.06100|
In 2008, regardless of cover crop and broiler litter timing, cotton lint yield increased with increasing broiler litter applications in a quadratic fashion (Y = 658 + 152x – 7.6x 2, r 2 = 0.98, P < 0.001). No difference in cotton lint yield was obtained between broiler litter at the rate of 9 and 13.4 Mg ha−1 (Table 8 ), indicating broiler litter at rate >9 Mg ha−1 was not advantageous to cotton lint yield. Application of broiler litter to cotton with more N necessary for maximum yield resulted in more vegetative growth and most likely delayed boll formation and maturity in which bolls were not contributed toward lint yield at the normal time of picking. This agrees with the results of Hunt et al. (1998) who reported that cotton had greater vegetative growth and less bolls with excessive N than required for maximum yield.
|Cropping system||Timing||BL†||CF‡||Growth||Boll no.||Lint turnout||Leaf N||Chlo.§ index|
|Mg ha−1||kg N ha−1||kg ha−1||%||g kg−1|
|P > |t||
|Fall vs. spring||<0.0001||<0.0001||0.0005||<0.0001||<0.0001|
|CC vs. NC||0.0403||0.0897||0.0998||0.5254||<0.0001|
Averaged across broiler litter application rate and timing, significant difference in cotton lint yield was obtained between cover and no cover crop in 2007 (P = 0.0427) and 2008 (P = 0.0403) (Table 6). Cotton lint yield was smaller in the presence of cover crop (1104 and 1136 kg ha−1) than in the absence of cover crop (1134 and 1160 kg ha−1) in 2007 and 2008, respectively (Table 7 and 8). Lower yield in the presence of cover could be related to rapid immobilization of available inorganic N into organic fraction (Green et al., 1995) and delayed release of N for cotton uptake.
The effect of broiler litter application timing on cotton lint yield (averaged across cropping system) was greater for spring than fall application in 2007 (1193 vs. 1046 kg ha−1, P < 0.0001) and in 2008 (1228 vs. 1068 kg ha−1, P < 0.0001, repectively) as evidenced by the significant broiler litter timing × litter rate interaction in both years (P < 00001)(Table 6). In the absence of cover crop, low cotton yield response to fall-applied broiler litter compared with spring application was attributed to leaching losses of nutrients due to typical high precipitation during winter and early spring in the southeastern United States (Tables 7 and 8 and Fig. 2 ). In the presence of cover crop residue, cotton lint yield was also greater for spring than fall application in 2007 and 2008 as evidenced by cover crop × broiler litter timing interactions in 2007 (P < 0.0001) and in 2008 (P < 0.0001) (Table 6). In the presence of cover crop residue, greater lint yield most likely related to this reason that applying broiler litter to the cover crop residue in the spring before planting cotton provided greater available N for microbial activities, enhanced decomposition of cover crop residue, and accelerated the release of plant-available N for subsequent crops (Wilson and Hargrove, 1986; Utomo et al., 1990; Varco et al., 1999). However, fall-applied broiler litter coupled with cover crop, with no supplemental N in spring, resulted in rapid immobilization of available inorganic N (Green et al., 1995) and delayed release of N for cotton uptake. Therefore, in the presence of cover crop residue, spring application of broiler litter could synchronize better with cotton N needs than fall- applied litter. This agrees with the results of Boquet et al. (2004) reported addition of even small amounts of fertilizer N to cover crop residue in spring accelerated mineralization of residues within the zone influenced by fertilization. In agreement with our results, Hanna et al. (2000) reported that spring application of manure to corn provides greater yield and manure N use efficiency than fall application.
Inorganic fertilizer N was applied to cotton only in spring at the recommended rate of 125 kg ha−1 (50 kg ha−1 at planting and 75 kg ha−1 at squaring). Application of inorganic fertilizer to cotton in spring in the presence of cover crop residue resulted in less lint yield than in the absence of cover crop residue (1256 vs. 1392 kg ha−1 in 2007 and 1288 vs. 1436 kg ha−1in 2008) (Tables 7 and 8), suggesting N immobilization by cover crop residues.
Cotton lint components such as number of bolls and cotton lint turnout were also influenced by broiler litter management. Boll number was not affected by cover crops in 2007 (P = 0.4927) (Table 6) and averaged across broiler litter rate and timing, no differences in boll number was obtained between cover crop and no cover crop in 2007 (49 vs. 50)(Tables 7). Only in 2008, similar to cotton lint yield, boll number was affected by cropping system, broiler litter rate and timing as evidenced by a significant interaction of these factors (0.0317)(Table 6). Averaged across cropping system, boll number was greater for spring than fall application (54 vs. 50, P < 0.0001)(Table 6 and 8). In 2008, boll number increased with increasing spring-applied litter and maxiumum boll number was obtained at the rate of 9 Mg ha−1 (Table 8) as evidence by significant broiler litter rate and timing interaction (P = 0.0560)(Table 6). At rate >9 Mg ha−1, boll number was not affected by broiler litter rates (Tables 8). This agrees with the results of Boquet et al. (2004) reported excess available N induces excessive vegetative growth and delayed boll formation and maturity which may not contribute toward cotton lint yield.
Cotton lint yield is determined not only by the number of bolls but also by boll weight and lint turnout. Cotton lint turnout had the highest value in the control plot with no manure and had the lowest magnitude with the highest broiler litter rate (Tables 7 and 8). Cotton lint turnout was not affected by cover crops in 2007 (P = 0.6708) and 2008 (P = 0.0998) (Table 6). Averaged across broiler rate and timing, no differences in cotton lint turnout was obtained between cover crop and no cover crop in 2007 (39.6 vs. 39.5%) and 2008 (39.0 vs. 38.6% (Tables 7 and 8). The effect of broiler litter application timing on cotton lint turnout (averaged across cropping system) was smaller for spring (39.0 and 38.5%) than fall application (39.9 and 38.9%) in 2007 and 2008, respectively. Since spring-applied broiler litter provides more available N for cotton than fall-applied litter, particularly in the absence of cover crop, cotton lint turnout was reduced with increasing spring-applied litter. Our results are in agreement with the work by Sawan (2008) and Tewolde et al. (2007) who reported that lint percentage decreased when N level increased.
Chlorophyll Meter Readings and Leaf Nitrogen Concentrations
Cotton chlorophyll index and leaf N were greatly affected by cover crop, broiler litter application timing and rates as evidenced by a significant interaction of these factors (P < 0.0001)(Table 6). In the presence of cover crop treatment, chlorophyll index increased linearly with increasing broiler litter rates for both fall (Y = 37.6 + 0.58x, r 2 = 0.96, P < 0.001) and spring-applied litter (Y = 37.2 + 0.83x, r 2 = 0.94, P < 0.001) as evidenced by the significant broiler litter rate and timing interaction in 2007 and 2008 (Table 6). In the absence of cover crop, averaged across cropping system, chlorophyll index was higher when litter was applied in spring (44.5 and 46.1) than in the fall (37.5 and 38.3) in 2007 and 2008, respectively as evidenced by the significant cover crop and litter timing interaction (P < 0.0001 (Tables 6, 7, and 8). Similar to chlorophyll readings, leaf N concentrations tended to increase linearly with increasing broiler litter rates for both fall (Y = 30.5 + 0.38x, r 2 = 0.96, p < 0.001) and spring-applied litter (Y = 30.6 + 0.53x, r 2 = 0.98, p < 0.001) in each year. Since chlorophyll readings and leaf N concentrations followed a similar pattern in response to broiler litter application to cotton, our results indicated leaf chlorophyll readings can be used as an indicator of N status in the plant as reported by others (Feibo et al., 1998; Bronson et al., 2003). In addition, Reddy and Kumari (2004) reported that cotton leaf chlorophyll meter reading is a good indicator of plant growth and soil conditions for crop productivity, and is positively correlated with biomass and lint yield of cotton.
Although, cotton lint yield peaked at broiler litter rate of 9 Mg ha−1, leaf chlorophyll index and leaf N concentrations continue to increase to the highest broiler litter rate (13.4 Mg ha−1). This suggests the cotton plant continues to extract soil N beyond that necessary to mature bolls. For example, for the spring application, leaf N concentration at flowering was about (36.3 g kg−1 in 2007 and 39.7 g kg−1 in 2008) when broiler litter was applied at the rate of 9 Mg ha−1, but increased to approximately 39.1 g kg−1 in 2007 and to 42.4 g kg−1 in 2008 if fertilized with 13.4 Mg ha−1 This suggests the cotton plant continues to extract soil N beyond that necessary to mature bolls and optimum yield.
Cropping system did not have significant effect on leaf N concentration (P = 0.2788)(Table 5). Averaged across broiler litter application rates and timing, no differences in leaf N concentrations was obtained between cover crop and no cover crop in 2007 (32.7 vs. 32.9 g kg−1, P = 0.2481) and in 2008 (34.5 vs. 34.6 g kg−1, P = 0.5254) (Tables 7 and 8). However, broiler litter application timing affected cotton leaf N concentrations (P < 0.0001)(Table 6). Averaged across cropping system, cotton leaf N concentration was greater for spring (34.0 and 36 g kg−1) than for fall application (31.4 and 33.2 g kg−1) in 2007 and 2008, respectively as evidenced by the significant manure rate and timing interaction (Table 6). Low leaf N concentration with fall application could be related to the leaching losses of N during fall and winter, particularly where cover crop is not present, and reduce the availability of N for subsequent crop in the spring. This agrees with the results of others who reported that fall application of manure resulted in considerable N losses through both leaching and denitrification NO3–N and was usually associated with potential environmental contamination (Smith and Chambers, 1998; Hansen et al., 2004; van Es et al., 2004; Paul and Zebarth, 1997).
Nitrate Leaching and Soil Residual Nitrate Nitrogen
Soil NO3–N at the 0- to 30-cm depth in spring, after killing cover crop and before planting cotton, was greater for the no cover crop than the rye cover crop each year (Table 9 ). This was possibly related to N immobilization in which inorganic N taken up by winter rye cover crop in the fall and winter converted into organic N pool (Dabney et al., 2001; Parkin et al., 2006). In agreement with our results, Parkin et al. (2006) reported that the winter rye cover crop most likely reduced soil available N by accumulation of N in plant biomass. Kuo et al. (1997) reported that winter rye cover crop was effective at increasing soil organic N because of the great potential of biomass and C input. Similar effects were also observed by Vyn et al. (2000) who reported that winter rye cover crops were associated with lower spring soil NO3–N concentrations where compared with winter fallow.
|Cropping||Rate||Volume||NO3–N leached||Soil NO3–N||Volume||NO3-N leached||Soil NO3–N|
|Mg ha−1||mL||mg L−1||mg kg−1||mL||mg L−1||mg kg−1|
Cover crop resulted in a distinct difference in NO3–N concentration in the leachate samples collected after cover crop establishment (Fig. 2). Nitrate-N concentrations in the leachate were significantly reduced in winter rye cover as compared to winter fallow plots probably due to greater N utilization by the established cover crop. Schroder et al. (1996) reported that the reduction in N leaching was similar to the N content of rye at the time of killing. In both years NO3–N leached was significantly less in zero N plots than in higher rate N plots both with and without a cover crop. In 2007, the mean NO3–N leached across sampling dates was 6.1 and 16.8 mg L−1 in winter rye cover crop and were 6.2 and 36.4 mg L−1 in no cover crop plots for the control and broiler litter application at the highest rate of 13.4 Mg ha−1, respectively. In 2008, NO3–N leached ranged from 7.1 to 26.8 mg L−1 in the cover crop and from 9.1 to 64.6 mg L−1 in the no cover crop plots when applied broiler litter increased from 0 to 13.4 Mg ha−1 (Table 9). The greater NO3–N leached in 2008 were related to the greater amount of total N applied in 2008 with manure that was drier than the one in 2007 (Table 2). Each year the presence of cover crop reduced NO3–N leaching. Averaged across years, cover crop reduced leaching losses of N by 57% as compared to no cover crop (Table 9). In agreement with our results, Kaspar et al. (2007) reported that a rye cover crop reduced 4-yr average flow-weighted nitrate concentrations by 59% compared with no cover crop. Meisinger et al. (1991) reported that a cover crop is known to be an effective strategy and can reduce NO3–N concentrations in leachate from 20 to 80%. Losses of N as NH4 + were not affected by cover crop due to its low mobility in soils or retention on cation exchange complexes. With adequate cation exchange capacity in this Leeper silty clay loam soil NH4 +-N would probably remain largely adsorbed (Dontsova et al., 2005). The total amount of N loss as NO3–N is dependent on leachate concentrations and the volume of water drained from the soil profile. The rye cover crop likely reduced NO3–N leaching through a combination of reduced drainage volume and concentration. The volumes of collected leachate were significantly less for the winter rye cover crop than winter fallow treatment. Averaged across the broiler litter rates and year, winter rye cover crop reduced the volume of leachate by 34% as compared to no cover (206 vs. 313 mL). Under no cover crop, there were no significant differences in the volume of leachate among the broiler litter rates. However, for winter rye cover crop, broiler litter at the low (4.5 Mg ha−1) and medium (9 Mg ha−1) rates had significantly less leachate than at the highest rate of 13.4 Mg ha−1 and the control. Therefore, the percentages of broiler litter N loss were significantly greater with winter fallow than winter rye cover crop where both received the same broiler litter rates.
Winter rye cover crop, relative to the winter fallow, reduced soil NO3–N concentrations by 49 and 51% when measured on soil samples taken at the 0- to 30-cm depth after killing rye cover crop and before planting cotton in early spring (Table 8). Chantigny et al. (2002) reported that mineralization of organic N from fall-applied manure most likely happened during late fall, winter and early spring, increasing the potential of NO3–N loss due to heavy precipitation (Beckwith et al., 1998). Soil NO3–N concentrations in spring before planting cotton was increased with increasing broiler litter applications in both cover and no cover treatments (Table 8). Although living plants have the potential to increase net N mineralization (Griffiths and Robinson, 1992; Dormaar, 1990), the presence of cover crop with no broiler litter application in our study did not significantly increase soil residual NO3–N. The reduction of soil residual NO3–N in the spring for cover crop was attributed to the ability of the winter rye in sequestering soil NO3–N as a component of organic compounds in the plant. Likewise, Sainju et al. (1998) found lower soil NO3–N concentration following a rye cover crop due to its high root density which removes a considerable amount of soil NO3–N. In the presence of winter rye cover crop, no significant difference in soil nitrate N was obtained between broiler litter at the rates of 4.5 and 9 Mg ha−1, however, for winter fallow, soil NO3–N increased with increasing broiler litter applications, particularly at the highest rate of 13.4 Mg ha−1 (P < 0.001) (Table 8).
Postharvest soil residual NO3–N concentrations at the 0- to 15-cm depth were higher for the spring-applied broiler litter than for the fall-applied litter (Fig. 3 ). These results are in agreement with the work by Randall et al. (1999) who reported greater soil NO3–N concentrations for spring-applied manure compared with fall application. The residual soil NO3–N from fall-applied broiler litter in our study was more uniformly distributed in the soil profile in the presence of rye cover crop than in its absence. Further, litter rates did not differ significantly in residual soil NO3–N and cover crop, suggesting that NO3–N content in the soil profile most likely was depleted by the rye cover crop. When the litter was applied in the spring, residual NO3–N at the 0- to 15-cm depth was increased by litter, particularly by the 9 and 13.4 Mg ha−1 rates. This likely is an indication that spring-applied broiler litter to cotton in the presence of cover crop residues increases the mineralization of organic N from crop residue and other sources of organic matter to increase the release of plant available forms of N as Boquet et al. (2004) reported. In the absence of cover crop, both fall and spring applications increased soil residual NO3–N. Applying broiler litter >9 Mg ha−1 in the spring with no cover crop residue appeared to supply N in excess of cotton N need for optimum yield as evidenced by yields and substantial residual soil NO3–N at the top 30-cm soil depth at the highest litter rate.
Cotton lint yield increased with increasing broiler litter applications and maxium lint yield was obtained at the rate of 9 Mg ha−1 Application of broiler litter to cotton at rate >9 Mg ha−1 was not advantageous and showed a tendency to decline. This confirms that application of broiler litter to cotton with more N necessary for maximum yield may result in more vegetative growth and delay boll formation and maturity in which bolls are not contributing toward lint yield at the normal time of picking. Seeding a winter rye cover crop to fall-applied broiler litter did not benefit cotton lint yield but substantially reduced leaching loss of NO3–N. This important work will help fill critical gaps in understanding wether detrimental effect of fall-applied broiler litter can be mitigated by a combination of cover cropping strategies.