| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
a USDA-ARS, Mississippi State, MS 39762
b USDA-ARS, Bowling Green, KY 42104
c USDA-ARS, Stoneville, MS 38776
* Corresponding author (htewolde{at}ars.usda.gov).
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: DAP, days after planting • NEE, nitrogen extraction efficiency • NTPA, applied total plant available nitrogen • STD, farm standard fertilization • UAN, urea–ammonium nitrate.
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Loss of litter-derived organic and inorganic N to the immediate environment due to inefficient extraction is a concern when poultry litter is used as a primary row crop fertilizer. A few empirical studies, mostly on forage and pasture grasses, have documented the magnitude of extraction and utilization of litter-derived N. Evers (2002), for example, determined N uptake in a pasture of annual ryegrass (Lolium multiflorum L.) and bermudagrass [Cynodon dactylon (L.) Pers.] fertilized with broiler litter with supplemental commercial N fertilizer. He reported a maximum annual N uptake of 285 kg ha–1 when the ryegrass–bermudagrass pasture was fertilized with 341 kg N ha–1 from litter plus 224 kg N ha–1 from NH4NO3 applied four times during the year. The N uptake was only 84 kg ha–1 by ryegrass and 69 kg ha–1 by bermudagrass for a total of 153 kg ha–1 annual N uptake when litter was the only source of applied N. This uptake represented 45% of the total litter-supplied N. Using container experiments, Chadwick et al. (2000) estimated the total uptake of organic litter-N by perennial ryegrass (L. perenne L.) to range between 16 and 56%, and reported even less uptake of organic N derived from dairy and pig slurries. Brink et al. (2004) measured N uptake by common and six hybrid bermudagrass cultivars fertilized with 11.8 Mg ha–1 yr–1 dry litter that supplied 540 kg total N ha–1 yr–1. Total N uptake when averaged across the cultivars and 4 yr was 316 kg ha–1 on an annual basis, which represents 58% of the total applied litter-N.
Uptake and utilization of litter-derived N by cotton, unlike forage and pasture grasses, is not researched or is not well documented. Past studies on N uptake by cotton pertain to commercial fertilizers that may serve as a guide on the capacity of cotton to extract N from the soil. Maximum uptake of 235 kg ha–1 was reported by Halevy (1976) from research conducted in Israel where two cultivars (Acala 1517-C and Acala 4–42) were grown with irrigation and fertilized with 100 kg ha–1 inorganic N. Other reports show uptake much less than 235 kg ha–1. Bassett et al. (1970) used Acala 4–42 under irrigation in California and reported 
38% less total N uptake (142 kg N ha–1) although they applied 34% more N than Halevy (1976). Mullins and Burmester (1990) reported even less average N uptake (128 kg ha–1) by four cultivars fertilized with 78 kg ha–1 N in the southeastern U.S. cotton production region. Boquet and Breitenbeck (2000) reported an uptake of 160 kg ha–1 N by mature cotton that received 84 kg N ha–1, which supported normal cotton production in the southeastern USA. When they applied twice (168 kg ha–1) as much N, the N in aboveground plant parts increased only slightly to 179 kg ha–1. Olson and Bledsoe (1942), who also conducted their research in the southeastern U.S. cotton production region >60 yr ago, reported an uptake of 94 to 151 kg ha–1 N (depending on the soil type) by aboveground plant parts of cotton fertilized with only 40 kg ha–1 N. These reports suggest the capacity of cotton to extract N from the soil probably depends on the cultivar, soil type, and other conditions that affect plant growth.
The objective of this research was to determine the magnitude and efficiency of N extraction and utilization by cotton fertilized with broiler litter with or without supplemental inorganic N fertilization. The research was part of a larger program with an overall goal of developing management methods and prescriptions for using poultry litter as a primary fertilizer for cotton in the southeastern USA. Lint yield and fiber quality results are presented elsewhere (Tewolde et al., 2007).
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
The 10 treatment combinations within each location were tested in a randomized complete block design with three or four replications. The plots consisted of four 119-m-long rows spaced 1.02 m apart at Cruger and eight 73-m-long rows spaced 0.97 m apart at Coffeeville. Each treatment was applied to the same plot in all 3 yr. Each year, UAN solution (32% N) was applied between square and first flower stage as a sidedress using a commercial liquid fertilizer applicator equipped with coulters that opened slits about 0.15 to 0.20 m away from the row center into which the UAN solution was injected to a depth of 0.1 m. Nitrogen, P, and K fertilizers were applied to the STD at the same rate as adjacent fields as practiced by the respective farm. This treatment received 135 kg N ha–1 as UAN each of the 3 yr at Cruger. At Coffeeville, the treatment received 101 kg N ha–1 in 2002 and 118 kg N ha–1 in 2003 and 2004 as UAN. Phosphorus was applied as triple superphosphate (0–46–0 N–P2O5–K2O) to the STD at 0, 20, and 0 kg P ha–1 at Cruger and 29, 20, and 0 kg P ha–1 at Coffeeville in 2002, 2003, and 2004, respectively. Potassium was applied to the STD as KCl (0–0–60 N–P2O5–K2O) at 140, 98, and 93 kg K ha–1 at Cruger and 56, 75, and 112 kg K ha–1 at Coffeeville in 2002, 2003, and 2004, respectively. All P and K fertilizers were applied to the STD as a broadcast by hand before planting.
Litter was applied 3, 1, and 10 d before planting at Cruger in 2002, 2003, and 2004, respectively. At Coffeeville, litter was applied 22 d before planting in 2002, 25 d after planting (DAP) in 2003, and 9 DAP in 2004. Each year, the litter was broadcast applied with a commercial fertilizer spreader equipped with ground speed–sensing radar, an electronic scale, and a rate-control computer system (Barrons & Brothers, Inc., Gainsville, GA). Litter was soil incorporated in the same day of application under the conventional-till at Cruger and was not incorporated under no-till at Coffeeville. Cotton was planted on 19 Apr. 2002, 16 Apr. 2003, and 19 Apr. 2004 at Cruger and on 21 May 2002, 2 May 2003, and 28 Apr. 2004 at Coffeeville. Additional details on crop management, weather, and litter chemical properties are provided elsewhere (Tewolde et al., 2007).
Nitrogen extraction was determined based on aboveground whole plant samples collected from 0.5 to 0.6 m2 in the center rows of each plot. Plant samples were taken four times during the 2002 and 2003 growing seasons from Cruger and during the 2002 growing season from Coffeeville. The samples from Cruger were taken 68, 97, 112, and 130 DAP in 2002 and 78, 98, 121, and 135 DAP in 2003. The samples from Coffeeville were taken 50, 71, 93, and 112 DAP in 2002. In other years, plants were sampled only at the end of the season 134 DAP in 2004 from Cruger and 137 DAP in 2003 and 138 DAP in 2004 from Coffeeville. Each year, cotton started flowering in late June at Cruger and in early to mid-July at Coffeeville.
Plants sampled were cut at soil level, separated by hand into leaves (leaf blade + petioles), stems (branches + main stem), and reproductive parts (squares + flowers + bolls). Plant parts were dried in a forced-air oven at 80°C to constant weight, weighed, and ground to pass a 1-mm sieve. Reproductive parts were further separated into burs, seed, and lint after drying when bolls were mature enough to make the separation feasible. Lint was separated from seed using a 10-saw gin. Seed samples were delinted with H2SO4 before grinding, as linters on seed made homogenization extremely difficult.
Total N concentration in the different plant parts was determined by an automated dry combustion method using a ThermoQuest (CE Elantec, Inc., Lakewood, NJ) C/N analyzer (Horneck and Miller, 1998). Accumulation of N in each plant part was calculated as the product of N concentration and dry weight of each plant part. Total N accumulated by aboveground plant parts was determined as the sum of N accumulated in leaves, stems, and reproductive parts. Lint N concentration was not analyzed on all samples as the N content of lint is known to be low (2.5 g kg–1 around maturity) and varies little with N nutrition (Fritschi et al., 2004). Only the 2002 lint samples from Cruger were analyzed for N after thoroughly cleaning each sample by hand to remove visible debris. The lint N concentration of these samples was used to determine total lint N accumulation in all years at both locations. The efficiency of cotton to extract applied N, which is sometimes referred to as apparent N recovery (Scholberg et al., 2000), was evaluated by computing extracted N as a percentage of total applied after accounting for N extracted by untreated plants as follows: NEE = 100 x (Nt – Nu)/Na where NEE = nitrogen extraction efficiency, Nt = nitrogen amount extracted by a treatment that received litter-N plus UAN-N, Nu = nitrogen amount extracted by the treatment that received no litter or UAN, and Na = amount of total N applied in litter plus UAN.
Litter and supplemental N effects on N uptake were tested using MIXED model analysis on SAS (Littell et al., 2002). Preliminary analysis of variance was performed for a randomized complete block design with a factorial treatment structure for litter and UAN-N factors. Additional analysis was performed using a trend to describe the litter and UAN-N treatment structure as a response surface model where the full model had three slope parameters that included litter linear (LL) and UAN-N linear (NL) effects and their interactions (LL x NL). The significance of LL and NL were tested by omitting the LL x NL interaction term from the full model when the interaction was not significant. Lack of fit, the term used to describe the treatment effect not explained by the response surface, was identified as a random effect in the analysis. The response surface models were combined over years and locations to test for the interaction of litter and UAN-N treatments with location and year. In this combined analysis, the full model allowed the three slope parameters to be different for each year so that there were nine slope parameters, three for each year. A reduced response surface model was also used by omitting nonsignificant terms for a more robust test of the interactions. This reduced response surface model had two slope parameters (LL and NL) with the same NL slope parameter across the years but separate LL slopes for each year so that there was a total of four slope parameters for all 3 yr combined. The experimental design for data combined over years is a split plot where main plot treatments were the litter and UAN-N treatment combinations as a response surface and the subplot treatment was year as a repeated measure. Random effects for this analysis included two terms for main plot error: lack of fit and replication x main plot treatment. This split plot analysis was also performed combined across both locations. In this analysis location and location x treatment interactions were fixed effects and the errors were pooled across locations. The relationship between N uptake and tissue N concentration with applied total plant available N (NTPA) was tested by regression analysis. The NTPA was estimated by summing litter-N and UAN-N assuming 50% of the total litter N and 100% of the UAN-N become plant available during the cotton growing season. Differences mentioned in the discussion are significant at P 
0.05 unless stated otherwise.
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Analysis of N extraction and tissue N concentration data, after combining for the 3 yr within a location, showed that the litter and UAN-N response surface trends were also not the same over the 3 yr, probably because of cumulative residual effect of the treatments which were not assigned to a new random set of plots each year. A plot received the same treatment each of the 3 yr within a location. As a result, the data were analyzed and are presented separately by year within a location.
Nitrogen Concentration in Plant Parts
Cruger
Nitrogen concentration in bulk leaves averaged across all 10 treatments gradually decreased during the growing season (Table 1). Leaf N concentration decreased 37% (from 45.2 g kg–1 68 DAP to 28.5 g kg–1 130 DAP) in 2002 and 29% (from 34.3 g kg–1 78 DAP to 24.5 g kg–1 135 DAP) in 2003. Bulk leaf N concentration in all treatments in the first two sampling days (68 and 97 DAP) in 2002 exceeded 30 g kg–1. The established sufficiency range based on the youngest fully mature leaf blades taken between early and late flowering is 30 to 45 g kg–1 (Mitchell and Baker, 2000). Leaf N concentration of all treatments except L0N0 and L4.5N0 also exceeded 30 g kg–1 on the first sampling (78 DAP) in 2003. Leaf N concentration began to fall below 30 g kg–1 in treatments that received the lower litter rates with or without 34 kg ha–1 UAN-N starting on the third sampling day (112 DAP) in 2002 and on the second sampling day (98 DAP) in 2003.
|
Within each sampling day, both applied litter and UAN-N significantly affected N concentration in plant parts measured during the first three sampling days in 2002 and in 2003 (Table 1). Nitrogen concentration in plant parts increased linearly with increasing litter rate or UAN-N rate. On the last sampling, litter and UAN-N did not always significantly affect N concentration in plant parts. The interaction between the linear effect of litter (LL) and the linear effect of UAN-N (NL) was not significant on any of the sampling days, which indicates the change in N concentration (slope) with increasing litter rate was the same at any level of the supplemental UAN-N rates. Among all measurements, the STD and L4.5N67 treatments had the largest N concentration in the different plant parts. These two treatments were also consistently among the best yielding treatments (Tewolde et al., 2007). The L6.7N34 treatment was also among the treatments that had the largest N concentration in plant parts.
Nitrogen concentration of plant parts was significantly associated with applied total plant available N (NTPA) (Table 2, Fig. 1). The association was usually linear with a few significant quadratic relationships. Nitrogen concentration in leaves, stems, and reproductive parts increased with increasing NTPA. Seed N concentration increased linearly with increasing NTPA when it was measured 98 and 121 DAP in 2003. Seed N concentration measured on the last day of measurement (135 DAP in 2003 and 134 DAP in 2004) also increased with increasing NTPA but the increase was nonlinear with a tendency to level off at the largest NTPA. Bur N concentration also increased with increasing NTPA. Lint N concentration which was measured only in 2002 was not related with NTPA when measured before maturity (data not shown). At maturity on the last sampling day, however, lint N concentration had a small linear increase with increasing NTPA (Lint N in g kg–1 = 0.0068 x NTPA in kg ha–1 + 1.4247, r2 = 0.87, n = 10).
|
|
4.5 Mg ha–1 litter with no supplemental UAN-N began to fall below 30 g kg–1 on the second sampling day (71 DAP). Only the L4.5N67 treatment maintained leaf N above 30 g kg–1 on the third sampling day. Stem N concentration decreased by 50.4% from an average across treatments of 11.7 g kg–1 50 DAP to 5.8 g kg–1 112 DAP. Reproductive N concentration decreased by 40.3% from 36.0 g kg–1 50 DAP to 21.5 g kg–1 112 DAP. Bur N concentration decreased while seed N concentration sharply increased during the last two measurements. Litter at Coffeeville had a significant (P < 0.10) linearly increasing effect on leaf and stem N concentration in the first three sampling days but had no effect on seed or bur N concentration on any of the sampling days (Table 1). Nitrogen as UAN significantly affected N concentration in leaves, stems, and burs in most of the sampling days but did not affect seed N. Nitrogen concentration of plant parts usually was significantly associated (mostly linearly) with NTPA (Table 2, Fig. 1). Nitrogen concentration in leaves, stems, and reproductive parts during the first three sampling days increased with increasing NTPA.
End-of-Season Nitrogen Extraction
Cruger
The L0N0 treatment, that received no litter or UAN-N, extracted 125 kg N ha–1 in 2002 and 91 kg N ha–1 each in 2003 and 2004 (Fig. 2). Applying litter with or without supplemental UAN-N increased the amount of extracted total N in direct proportion to the litter rate with no significant LL x NL interaction (Table 3; Fig. 2). The L4.5N67 treatment, which consistently yielded as good as or better than the STD (Tewolde et al., 2007), was among the treatments that had the largest N extraction with 196, 242, and 260 kg N ha–1 total extraction in 2002, 2003, and 2004, respectively. These amounts were very similar to the N extraction of the STD with 184, 259, and 278 kg N ha–1 in 2002, 2003, and 2004, respectively.
|
|
|
Coffeeville
The L0N0 treatment at this location extracted 107 kg N ha–1 in 2002 and 100 kg N ha–1 each in 2003 and 2004 (Fig. 2). These amounts may be considered the N supplying capacity of the soil at Coffeeville. Applying litter increased the amount of extracted total N in direct proportion to the litter rate with no significant LL x NL interaction (Table 3; Fig. 2). The L4.5N67 and L6.7N34 treatments, which were among the best yielding treatments at this location (Tewolde et al., 2007), extracted more N than any other treatment. The L4.5N67 treatment extracted 199, 165, and 186 kg N ha–1 and the L6.7N34 treatment extracted 199, 175, and 161 kg N ha–1 in 2002, 2003, and 2004, respectively. The corresponding extraction by the STD was 198, 123, and 168 kg N ha–1.
Plants at Coffeeville, unlike plants at Cruger, usually extracted less N than the total amount applied (Fig. 2 and Table 4). When averaged across the 10 treatments, extracted N at Coffeeville was only slightly more or less than applied N each of the 3 yr. Average applied total N was 150, 145, and 153 kg ha–1 compared with average extracted N of 161, 131, and 140 kg ha–1 in 2002, 2003, and 2004, respectively. But, treatments differed in the amount of extracted N relative to the applied. Extracted N exceeded or nearly equaled applied N when only 2.2 Mg ha–1 litter was applied regardless of the amount of supplemental N. When applied litter was 6.7 Mg ha–1, however, extracted N was less than applied regardless of the level of supplemental UAN-N. The largest discrepancy between extracted and applied N under this treatment was 79 kg N ha–1 in 2002, 97 kg N ha–1 in 2003, and 79 kg N ha–1 in 2004 when no supplemental N was applied. The treatment that received 4.5 Mg ha–1 litter also extracted less N than applied regardless of the amount of supplemental UAN-N. The L4.5N67 treatment, which was also the best yielding treatment (Tewolde et al., 2007), extracted 24, 61, and 41 kg ha–1 less N than applied in 2002, 2003, and 2004, respectively. The STD, which received UAN-N by injection into the soil, extracted 86, 5, and 50 kg ha–1 more N than applied in 2002, 2003, and 2004, respectively, but yielded less than the L4.5N67 treatment.
Nitrogen extraction efficiency averaged across all treatments at Coffeeville was 39, 22, and 24% in 2002, 2003, and 2004, respectively (Table 4). The tendency of NEE to increase with increasing supplemented UAN-N observed at Cruger was also observed at Coffeeville. Nitrogen extraction efficiency at this location significantly and linearly increased with increasing UAN-N rate in 2002 and 2004.
Partitioning of Nitrogen to Plant Parts
Nitrogen extracted by aboveground plant parts a few days before defoliation was partitioned largely to the boll. Nearly 69% of the N at this stage was recovered in seed, lint, and burs. When averaged across all treatments, years, and locations, 57% of the total N accumulated in aboveground plant parts was recovered in the seed alone and an additional 2.4% in lint. The sum of these values (59.4%) represents the proportion of extracted N that would be removed from the field. Plants partitioned an average across treatments of 62% of the total extracted N to seed and lint at Coffeeville and 56% at Cruger.
Partitioning of dry weight to seed was only 25% at both locations when averaged across all treatments and years (data not shown). Dry weight partitioning to lint was an average of 18% at Cruger and 22% at Coffeeville compared with a corresponding N partitioning to lint of only 2.3% and 2.6%, respectively. The average N accumulation in lint did not exceed 3 to 6 kg N ha–1 under either location and in any of the 3 yr.
Nitrogen partitioned to stems was five to six times less than N partitioned to seed (Fig. 2), while dry weight partitioned to stems at both locations was comparable to dry weight partitioned to seed (data not shown). The overall average dry weight partitioned to stem was the same as that partitioned to seed which was 25%. But, while the average N partitioned to seed was 57% of the total, the average N partitioned to stem was only 10%.
Partitioning of N to leaves and burs was more comparable with partitioning of dry weight to these plant parts. Average N partitioned to leaves was 21% compared with average dry weight partitioned to leaves of 12% (data not shown). Average N partitioned to burs was 10% compared with 15% dry weight partitioned to burs. The proportion of N partitioned to leaves was low primarily because the measurements were made late in the season, within a few days before defoliation. At earlier growth stages around the flowering stage, leaves make up a large fraction of the aboveground plant dry weight and become great N sinks. But, as bolls start to form and expand, a large amount of the leaf N is remobilized to developing seeds. This is followed by drastic decrease in N concentration in these leaves (Table 1) and by self-defoliation of the older leaves.
Litter or UAN-N application did not usually significantly affect N partitioning measured as plants approached maturity. When there was a significant treatment effect, the effect in one season or location did not hold the same trend in another season or location.
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Loss of litter-derived inorganic and organic N to the immediate environment due to inefficient extraction is a concern when poultry litter is used as a primary row crop fertilizer. When fertilized with inorganic fertilizers, the largest total N uptake reported in past research is 224 to 235 kg N ha–1 (Halevy, 1976). If poultry litter is applied to meet 100% of the N requirement of cotton, it may be necessary to apply as much as 9.0 Mg litter ha–1 (Tewolde et al., 2007) which may supply as much as 250 to 300 kg total N ha–1. In our research, cotton fertilized with as much as 4.5 or 6.7 Mg ha–1 litter supplemented with 67 or 34 kg ha–1 UAN-N extracted as much as 316 kg N ha–1 which exceeded total applied N (Fig. 2; Table 4).
However, not the entire extracted N is removed from the field in harvested crop. On average, only 59% of extracted N is removed from the field in harvested seed and lint. The most consistently best yielding treatment (L4.5N67) removed an average across years of 124 kg N ha–1 (53% of total uptake) at Cruger and 119 kg N ha–1 (65% of total uptake) at Coffeeville with harvested seed and lint. Fritschi et al. (2004) reported similar N partitioning to seed and lint (an average of 59.1%) in Acala cotton grown in California with inorganic fertilizers. A much earlier research also in California with Acala cotton reported 44 to 58% N partitioning to seed and lint (Bassett et al., 1970). Others, however, reported much lower (42–49%) N partitioning to seed and lint (Halevy, 1976; Mullins and Burmester, 1990).
Up to 47% of the total extracted N at Cruger and up to 35% at Coffeeville was bound in aboveground plant parts that remained in the field after harvest. This amount does not include N extracted during the season and lost in abscised leaves and reproductive parts that can be as much as 26% of the total (Boquet and Breitenbeck, 2000). An additional undetermined amount is also bound in root tissues that can be as much as 14% at the early boll development stage in container-grown plants (Tewolde et al., unpublished data, 2004). Although the N fraction bound in roots, stems, leaves, and burs is not removed from the field, it has little or no risk of becoming released to the immediate environment until the plant parts decompose and organic N mineralizes. This shows the risk of litter-derived N detrimentally affecting the immediate environment when fertilizing cotton with litter supplemented with inorganic N fertilizers is no greater than when fertilizing with 100% inorganic N fertilizers.
Unlike cotton in our research, which extracted more total N than applied at least at Cruger, certain forage and pasture grasses appear to extract much less litter-N than applied. For example, Evers (2002) reported a maximum annual N uptake of 285 kg ha–1 by a ryegrass–bermudagrass pasture that was fertilized with 565 kg ha–1 total N (341 kg ha–1 broiler litter-N plus 224 kg ha–1 N from NH4NO3) applied four times during the year. When litter was the only source of applied N, the N uptake was only 84 kg ha–1 by ryegrass and 69 kg ha–1 by bermudagrass for a total of 154 kg ha–1 annual N uptake, which represented only 45% of the total litter-supplied N. Brink et al. (2004) reported total N uptake of 316 kg ha–1 (averaged across seven cultivars and 4 yr) on an annual basis, which represented only 58% of the total applied broiler litter-N. In pot experiments, Chadwick et al. (2000) estimated the total uptake of organic litter-N by perennial ryegrass to range between 16 and 56% of the initial. They reported even less uptake of organic N derived from dairy and pig slurries. These results suggest cotton may be a more effective crop than, or at least as effective as, forage and pasture grasses—which are harvested several times during the year—in extracting and utilizing litter-supplied N. The N extraction efficiency of cotton is related to its capacity to accumulate a large amount of N in its seed, an average across treatments and years of 54% of the total extraction at Cruger and 60% of the total extraction at Coffeeville. Boquet and Breitenbeck (2000) also described cotton as an exceptionally N efficient crop, in part, because they found the amount of N removed from the field in harvested seed and lint was equivalent to the amount of inorganic fertilizer-N applied for optimal yield. Cotton may still extract applied N in excess of the amount necessary for optimal yield, but the excess is accumulated in other plant parts that are not harvested (Boquet and Breitenbeck, 2000).
Cotton appeared less efficient in utilizing litter-derived N at Coffeeville than at Cruger although the differences cannot be tested statistically (Fig. 2 and Table 4). The N extraction efficiency (NEE) of treatments that received litter with or without UAN-N relative to the STD was also generally less at Coffeeville than at Cruger (Table 4). While it appears cotton at Coffeeville was less able to extract litter-N than cotton at Cruger, the results may simply be a reflection that, before the litter-N becomes available for uptake, a greater fraction of it is lost at different stages of application at Coffeeville under the no-till than at Cruger under the conventional-till. The litter at both locations was applied by surface broadcasting using a fertilizer spreader. The litter was soil-incorporated within 1 d of application at Cruger but was left on the soil surface with no incorporation throughout each season at Coffeeville because of the no-tillage system at this location. Although unincorporated surface applied litter may be an effective method of fertilizing no-till cotton (Tewolde et al., 2007; Mitchell and Tu, 2005), lack of incorporation at Coffeeville may have increased the vulnerability and loss of mineralized litter-N to runoff or volatilization and reduced the NEE by decreasing the amount of applied litter-N that eventually became available for plant uptake. Depending on the conditions of litter and weather, it is possible to lose as much as 24% of the total litter-supplied N as NH3 during a season with the greatest loss occurring within the first week of application (Sharpe et al., 2004). Other factors may also have contributed to the smaller NEE at Coffeeville than at Cruger. Because uptake is a function of total growth, conditions that lead to less plant growth are likely factors that may have caused cotton at Coffeeville to be less efficient than cotton at Cruger in extracting litter-supplied N.
Nitrogen extraction efficiency at both locations usually was greater with than without supplemented UAN-N. This was particularly true with the litter rates of 4.5 and 6.7 Mg ha–1. The lower efficiency of N extraction when litter alone was applied than when litter was supplemented with UAN-N may be a reflection of the nature of the litter-N. Approximately 40 to 50% of litter-N is considered to be unavailable for plant uptake in the first season. This should have contributed to the low NEE when litter was not supplemented with UAN-N (all of which is available for plant uptake in the same season), because NEE was determined based on the total N applied.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
The results also demonstrated plants that received litter and UAN-N combinations that supplied adequate N extracted more N than applied at Cruger and slightly less than applied at Coffeeville. Nitrogen extraction efficiency, sometimes referred to as apparent N recovery, averaged across treatments and years was 28% at Coffeeville and 60% at Cruger. Nitrogen extraction efficiency of the treatments that received 4.5 or 6.7 Mg ha–1 litter usually was greater with than without supplemental UAN-N. Approximately 57% of extracted N was partitioned to seed and an additional 2.4% was partitioned to lint, the sum of which represents an amount that would be removed from the field. Cotton partitioned an average across treatments of 62% of the total extracted N to seed and lint at Coffeeville under no-till and 56% at Cruger under conventional-till.
These results demonstrate that cotton extracts N equivalent to the amount that may be supplied by as much litter as 4.5 to 6.7 Mg ha–1 when supplemented with inorganic N fertilizers. While only about 59% of this amount is removed from the soil in harvested crop, an amount equivalent to the remainder is bound in plant tissues with little or no risk of becoming released to the immediate environment until the plant parts decompose and organic N mineralizes. Our results show the risk of litter-derived N detrimentally affecting the immediate environment when fertilizing cotton with litter supplemented with inorganic N fertilizers is no greater than when fertilizing with 100% inorganic N fertilizers. When compared with reported N uptake of forage grasses, our results suggest that cotton may be a more ideal crop than, or at least as ideal as, forage and pasture grass crops to which litter can be applied as a fertilizer.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Received for publication August 17, 2006.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Applied UAN-N rates in 2004 at Cruger were 67 and 135 kg ha–1 instead of the 34 and 67 kg ha–1 applied in 2002 and 2003. STD = farm standard fertilization.