SSSAJ Grow Your Career with SSSA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 11 April 2005
Published in Soil Sci Soc Am J 69:718-728 (2005)
DOI: 10.2136/sssaj2003.0340
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fritschi, F. B.
Right arrow Articles by Hutmacher, R. B.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Fritschi, F. B.
Right arrow Articles by Hutmacher, R. B.
Agricola
Right arrow Articles by Fritschi, F. B.
Right arrow Articles by Hutmacher, R. B.
Related Collections
Right arrow Nutrients
Right arrow Other Crop Management
Right arrow Other Cropping Systems
Right arrow Cotton
Right arrow Nutrient Cycling

Nutrient Management & Soil & Plant Analysis

Recovery of Residual Fertilizer-N and Cotton Residue-N by Acala and Pima Cotton

Felix B. Fritschia,*, Bruce A. Robertsc, D. William Rainsb, Robert L. Travisb and Robert B. Hutmacherd

a 9611 S. Riverbend Ave., Parlier, CA 936438
b Dep. of Agronomy and Range Science, Univ. of California, One Shields Ave., Davis, CA 95616
c Dep. of Plant Science, California State Univ. Fresno, 2415 E. San Ramen Ave., Fresno, CA 93740
d Univ. of California, Shafter Research and Extension Center, 17053 N. Shafter Ave., Shafter CA 93263

* Corresponding author (ffritschi{at}fresno.ars.usda.gov)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Little is known about the availability of residual fertilizer-N and cotton (Gossypium spp.) residue-N to subsequent crops. This study determined the effects of two soil types and N fertilization rates on the recovery of residual 15N-fertilizer and 15N from labeled cotton residue in the soil and plants of subsequent cotton crops. Microplots in the San Joaquin Valley, California, on a Panoche clay loam [fine-loamy, mixed (calcareous), thermic Typic Terriorthents] and a Wasco sandy loam (coarse-loamy, mixed, nonacid, thermic Typic Torriorthents) were used to trace the fate of 15N from labeled aboveground cotton residue separately from the 15N in roots and soil. Total 15N-fertilizer recovery in the second year after application averaged 5.8% for Acala (G. hirsutum L.) and 2.9% for Pima (G. barbadense L.) cotton. In the third year after application, total 15N-fertilizer recovered by Acala averaged 2.0% on the clay loam soil and 3.3% on the sandy loam. Most of this recovered 15N-fertilizer was cycled through soil pools and roots and only small amounts originated from labeled aboveground residue. Virtually all of the 15N applied in the form of labeled aboveground residue and recovered in the soil was found in the top 0.3 m. The 15N-fertilizer which cycled through belowground pools was found mainly in the top 0.6 m of the soil. Recovery of residual 15N-fertilizer appears to contribute little to total cotton N uptake in the second and third crop after application.

Abbreviations: CEC, cation exchange capacity


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
EFFICIENT USE AND CONSERVATION of nutrient resources is critical to sustain agricultural production capacities, to protect the environment, and to profitably produce agricultural crops. Nitrogen management is critical to the successful production of cotton and may be quite challenging, particularly since N availability tends to be highly variable. Nitrogen fertilization recommendations are often based on yield goal and mineral-N available in the soil at the beginning of the growing season. Nitrogen from crop residues and soil N mineralization potential are usually not taken into consideration when fertilizer applications are made. Decomposing crop residues may release significant amounts of N and influence the availability of N by affecting mineralization–immobilization processes in the soil (Hood et al., 2000; Trinsoutrot et al., 2000). The quantity of N derived from crop residues by a succeeding crop is highly variable and depends largely on residue characteristics and the synchronization between N release and crop N uptake. To develop cotton production practices which use N resources efficiently, it is necessary to understand the processes, interactions, and the pools that determine N availability to a cotton crop.

Most of the cotton grown in the San Joaquin Valley, California, is of the Acala Upland type. However, considerable amounts of American Pima is grown in the San Joaquin Valley. In fact, California is the leading U.S. producer of American Pima. Pima is an extra-long staple cotton that generally yields less than Upland cotton and is more sensitive to delays in planting and excessive N fertility, which can result in greater vegetative growth and delayed maturity (Unruh and Silvertooth, 1996; Tewolde et al., 1995; Kittock et al., 1981; Silvertooth et al., 1995). Because the vast majority of cotton grown in the USA is G. hirsutum, research conducted on G. barbadense is limited to the former major producing states of Arizona and New Mexico. Because of its current importance to California cotton growers, production of American Pima merits research attention.

Applications of fertilizer-N can increase cotton biomass production, increase tissue N concentrations, change crop C/N ratios, and may change the soil microbial population size and activity (Fritschi et al., 2003, 2004a, 2004b; Entry et al., 1996; Conti et al., 1997; Ladd et al., 1994). Thus, N application rates could affect the decomposition of cotton residues and in turn the extent of residue N made available to the next crop. Decomposition and nutrient release are influenced by chemical and physical characteristics of crop residues (Müller et al., 1988; Fox et al., 1990; Vanlauwe et al., 1996) and a number of other factors, including residue placement and soil type (Hubbard and Jordan, 1996; Whitmore and Groot, 1997; Egelkraut et al., 2000; Thomsen et al., 2001). Although numerous studies have been conducted to determine N recoveries from a variety of crop litters and much is known about residue decomposition in general, information on the release of N from decomposing cotton residues is sparse. However, to efficiently manage soil fertility and maximize the benefits from cotton residues, it is important to understand how nutrients become available as these residues decompose. Furthermore, efficient use of N resources not only requires understanding the contributions by cotton residues (aboveground) to the next crop, but also those of previous year's fertilizer-N remaining in the soil (including root residues). Currently it is not clear how much of the N that is made available to the succeeding crop is from incorporated residue and how much originates from mineral and organic soil N and roots.

Fertilizer applied to a crop is subject to many fates, including plant uptake, mineralization–immobilization processes, and loss to the environment. The quantity of N subject to each pathway is determined by a multitude of factors. Depending on the weather conditions prevalent between growing seasons, variable amounts of mineral-N not taken up by the plants may be lost during the winter months. Plant uptake depends largely on the synchronization between N availability and plant demand and uptake capacity. The quantities of fertilizer-N taken up by the plants or incorporated into soil organic pools are in part a reflection of the competitiveness of plants and microbes for mineral-N. Microbial activity can immobilize fertilizer-N, which may subsequently lead to the incorporation of this N into organic forms that may be relatively resistant to mineralization (Kelley and Stevenson, 1996). Collectively, the processes involved in stabilization and release of mineral-N into and from organic forms largely define the availability of N for plant uptake and thus critically influence the recovery of fertilizer-N in the years following application. Determination of fertilizer-N recovery in crop and soil after multiple growing seasons provides an integrative measure of these soil N dynamics and allows evaluation of the efficiency of fertilizer-N applied to cotton.

Previously, we reported fertilizer use efficiencies between 43 and 49% for Acala and Pima cotton grown in the San Joaquin Valley, California (Fritschi et al., 2004b). The data presented in this study were collected in the same fields, and complement the series of studies on N dynamics in irrigated Acala and Pima cotton (Fritschi et al., 2003, 2004a, 2004b; Hutmacher et al., 2004). The purpose of this study was (i) to determine the contribution of N from aboveground cotton residues and from root and soil to subsequent cotton crops, (ii) to estimate the recovery of fertilizer-N in cotton plants grown one or two years after fertilizer application, and (iii) to examine N rate and soil type effects on objectives i and ii.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Site Characteristics and General Management Strategies
This study was conducted in the San Joaquin Valley, California, at the University of California West Side Research and Extension Center in Fresno County and at an on-farm site in Kings County during 1998, 1999, and 2000. The soils were characterized as Panoche clay loam at the Fresno County location and as Wasco sandy loam at the Kings County location. Selected soil properties of the surface 0.3 m were as follows. Panoche clay loam: 244 g kg–1 sand, 443 g kg–1 silt, 313 g kg–1 clay, 26.9 cmol kg–1 cation exchange capacity (CEC), pH 7.8, 7 g kg–1 organic matter; Wasco sandy loam: 658 g kg–1 sand, 272 g kg–1 silt, 70 g kg–1 clay, 12.6 cmol kg–1 CEC, pH 6.9, 8 g kg–1 organic matter (Sheldrick and Wang, 1993; Janitzky, 1986; U.S. Salinity Laboratory Staff, 1954; Nelson and Sommers, 1982). Unusually high rainfall and cool spring temperatures were characteristic for 1998, and resulted in a late start (planting in 1998: 24 and 28 April; 1999: 16 and 19 April; 2000: 3 and 8 April) and slow initial development of cotton. Weather conditions in the two following years were closer to normal. Additional information on soil and weather characteristics for these locations and years can be found in Fritschi et al. (2003).

Acala cotton ‘Maxxa’ was grown in all 3 yr at both locations and Pima cotton ‘S-7’ was grown in 1999 and 2000 at the Panoche clay loam site. Cotton was planted in rows 0.96 (Wasco sandy loam site) or 1.01 m (Panoche clay loam site) apart in main plots four to 12 rows wide and 80 to 170 m long (lengths of irrigation runs). Four N treatments consisting of 56, 112, 168, and 224 kg N ha–1 (N-56, N-112, N-168, and N-224) were established by subtracting prefertilization soil NO3–N (in the top 0.6-m soil layer) from these target rates and banded application ({approx}0.15 m deep, {approx}0.2 m away from the row on both sides of the plants) of the difference in the form of urea-N. A single application of urea–N was made before the first in-season irrigation once plants had developed three to five true leaves (4–5 wk after planting). To establish prefertilization soil NO3–N levels, samples were collected within 9 to 17 d after planting at six locations per main plot. Samples were taken in 0.3-m increments to a depth of 0.6 m, composited within each replicate, air dried at 35 to 40°C, and NO3–N was determined on 2 M KCl extracts (Keeney and Nelson, 1982). The treatments were set up in a randomized complete block design with either four (Acala) or three (Pima) replications. Cotton was planted and grown using cultivation, irrigation, weed, and pest management strategies typical for the region.

Labeling of Cotton Residue
In the first year of the trials (Acala, 1998; Pima, 1999) microplots were established in the N-56 and N-168 main plots. Microplots measured 7 m by seven rows in the Acala trials and 4 m by six rows in the Pima trial. These plots were not fertilized when N treatments in the main plots were established, but were fertilized with matching rates of 15N-enriched urea [9.5 (Pima) and 11 (Acala) atom% excess 15N-labeled urea in N-56 and 0.9 (Pima and Acala) atom% 15N-labeled urea in N-168] within 1 wk of the N application in the main plots. The 15N-enriched urea was dissolved in water and injected (self-refilling repetitive syringe attached to an injection rod) at every 0.1 m in a band 0.2 m from the plants along the cotton rows to a depth of 0.15-m to simulate the large scale application in the main plots (for more details see Fritschi et al., 2004b).

In the microplots, leaves that dropped to the ground were collected after defoliation but before machine harvest and stored in an empty greenhouse until use after harvest. Once seed cotton was harvested, aboveground plant material was shredded with a flail shredder. To keep 15N-labeled plant material from microplots separate from nonlabeled residue of the main plots, nonlabeled plant material was removed from borders around the microplots before shredding of labeled plant material. Shredded residue from microplots was collected immediately and weighed. Subsamples of shredded residue and previously gathered leaf residue were collected, dried at 65°C to constant weight, combined based on the ratio of the collected residue mass, ground, and analyzed for C, N, and 15N (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Selected characteristics of the 15N-labeled cotton residue applied to Acala or Pima microplots. Values in parentheses represent standard errors of four (Acala) or three (Pima) replications.

 
Experimental Procedure
To substitute labeled with nonlabeled crop residue in the microplots where 15N fertilizer was applied, shredded cotton residues were collected in the same N treatment from nonlabeled areas of the main plots where leaves had been collected when the labeled leaves were gathered in the microplots. Within 2 wk of harvest, labeled cotton residues and previously collected labeled leaves were then exchanged with nonlabeled residue and leaves and applied at equal rates (on a dry matter basis) to microplots (Acala, 6 m long by five rows wide; Pima, 3 m by four rows), resulting in two microplot series. One series, consisting of the original set of microplots with labeled belowground (soil and root) components, was amended with nonlabeled crop residue, while in the second series, a new set of microplots, labeled residue collected from the original microplots, was applied instead of nonlabeled residue. The plant residue was carefully spread in each microplot, leaf samples from storage first and then shredded residue, and incorporated when the main plots were disked. To trace the exact location of each microplot, metal markers were buried below the plow layer. These markers were pinpointed throughout the duration of the study using triangulation measurements and a metal detector.

Acala cotton planted in 1999 and 2000 was grown with the N treatments established in the same way and in the same main plots as in 1998. However, microplots were fertilized at the same time as the main plots and did not receive any 15N-labeled fertilizer. After seed cotton harvest in 1999, plant residues were shredded and returned to the same microplots. Since the Pima trial was established 1 yr after the Acala trials, the 2000 growing season represented the first year after switching the labeled and nonlabeled residues of the microplots.

Aerial portions of the cotton plants were sampled five or six times every year from the center of the microplots as previously described (Fritschi et al., 2004b). Briefly, samplings of selected developmental stages (early square, early bloom, peak bloom, just before defoliation corresponding to >60% open bolls, and physiological maturity) were matched as closely as possible for the 1999 and 2000 seasons and for the two species. At each sampling, plants from the 0.5 to 1.0-m row length were cut 25 mm below the cotyledonary node and separated into stems (including branches, petioles, squares, and flowers), leaves, and bolls. As bolls approached maturity, they were separated into burs (carpel walls), seed, and lint. Burs were combined with immature bolls into one fraction, while seed and lint made up the other two separate fractions. Plant tissues were dried at 65°C until constant weight, ground, and pulverized in a ball mill. After cotton harvest and shredding of the crop residues, soil samples were taken either in eight 0.3-m increments to a depth of 2.4 m or in four 0.3-m increments for the 0- to 1.2-m depth and then two 0.6-m increments for the 1.2- to 2.4-m depth increment. Triplicate samples per depth increment were collected {approx}0.25 m from the plant rows using a power-driven soil core sampling device fitted with a 0.05-m-diam. tube. The samples were air dried, crushed, sieved through a 2-mm screen, and pulverized for analyses. Natural abundance (background) 15N levels were determined on samples collected in nonlabeled control plots and averaged 0.3686% for plant material sampled just before defoliation and 0.3691% for soil samples (0- to 0.3-m depth). Total N and 15N content of plant and soil samples were determined using a Europa Scientific Integra Mass Spectrometer (PDZ Europa Ltd., Crewe, UK).

Percentage 15N-fertilizer or labeled residue recovery (%15NR) in crop and soil was calculated according to:

where a is the atom% excess (above background) in plant tissue or soil, f is the atom% excess in fertilizer or residue applied, TN is the total amount of N in the aboveground plant or in the soil (kg ha–1), and AN is the amount of fertilizer or residue N applied (kg ha–1). To calculate the percentage of N derived from fertilizer (%Ndff) and residue (%Ndfr), the following equation was employed:

where a is the atom% excess in aboveground plant tissue and s is the atom% excess in fertilizer or residue applied.

Analysis of variance was conducted using the SAS software package (SAS Institute, 1999). To determine treatment effects across time, mixed model ANOVA for a repeated measures design was used. Mean separation was determined at the 0.05 probability level using Fisher's LSD.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Recovery of Nitrogen-15 from Aboveground Residue
On both soil types, uptake of aboveground residue–15N by Acala cotton was 1.4 to 2.6 times greater in the first year than in the second year after application (Table 2). The majority of the residue–15N was recovered between early square and peak bloom ({approx}70–130 d after planting; Fig. 1), coinciding with the time of greatest total plant N uptake. In the first year after residue application, maximum uptake rates were observed between early square and early bloom and in the second year either between early square and peak bloom or early bloom and peak bloom. Recovery of 15N from Pima aboveground residue was only studied for one year after application. Similar to the results for Acala, recovery of aboveground residue–15N was very low and much of it occurred in the period between early square and early bloom (Fig. 1).


View this table:
[in this window]
[in a new window]
  		Table 2. Recovery of residue N in cotton plants at defoliation and soil after harvest as determined by 15N tracer method.

 


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1. Recovery of fall incorporated (Acala in 1998; Pima in 1999) 15N-labeled aboveground cotton residue N by Acala and Pima cotton in the next growing season (Acala: 1999; Pima: 2000) or the second growing season since residue application (Acala: 2000) as affected by N treatment. Bars indicate standard errors of three (Pima) or four (Acala) replicates.

 
In the first year after residue application (1999), Acala aboveground residue–15N recovery was not significantly different between the Wasco sandy loam and the Panoche clay loam. However, a tendency toward greater residue–15N recovery on the Wasco sandy loam was observed at the time of defoliation (P = 0.053). At that time, an average of 3.1% of the aboveground residue–15N was recovered by the Acala plants on Panoche clay loam and 5.9% by those on Wasco sandy loam. In the second year (2000) after residue application, Acala aboveground residue–15N recovery was significantly greater (P < 0.05 across all sampling dates) on the Wasco sandy loam (3.2% at defoliation) than the Panoche clay loam (1.7% at defoliation). However, the percentages of 15N derived from aboveground residue in the two growing seasons after application did not differ between locations or N treatments at either location and averaged 2.3% in 1999 and 1.6% in 2000 (Table 3). On a mass basis, labeled aboveground residue applied in 1998 contributed 1.5 to 6.6 kg ha–1 to the total N assimilated in 1999 and 1.1 to 4.6 kg ha–1 in 2000.


View this table:
[in this window]
[in a new window]
  		Table 3. Nitrogen in aboveground cotton biomass derived from fertilizer, residue, and soil at the time of defoliation.

 
Since Acala and Pima were planted in separate experiments, no direct statistical comparisons between the two were made; however, means for residue–15N recovery and percentages of N derived from residue in the next crop were lower for Pima than for Acala (Fig. 1, Table 3). Less than 1% (<0.5 kg N ha–1) of the N taken up by the Pima crop was derived from aboveground residue applied in the previous year.

Low recovery of 15N from aboveground residue in the first crop, and even lower recovery in the second crop after application was associated with minimal contributions to total plant uptake in these seasons. Percentages of N derived from cotton residue determined in this study were similar to the 1% reported by Rochester et al. (1993) and about 5% reported by Constable et al. (1992) for cotton grown on heavy clay in Australia. However, 15N recoveries were lower than those from many other plant residues. Recoveries of legume residue-N by a subsequent crop are often between about 10 and 35% (Ladd et al., 1981; Ta and Faris, 1990; Harris et al., 1994; Ranells and Wagger, 1997; Stevenson et al., 1998). Recoveries from nonleguminous residues are frequently, but not always lower than those from leguminous residues. For instance, in a comparative study, Norman et al. (1990) observed that 3% of 15N added in rice (Oriza sativa L.) straw, 37% of 15N added in wheat (Triticum aestivum L.) straw, and 11% of 15N added in soybean [Glycine max (L.) Merr.] residue was recovered in a subsequent rice crop. Rannells and Wagger (1997) observed that 4% of applied rye (Secale cereale L.) residue-N and 21% of applied crimson clover (Trifolium incarnatum L.) monoculture residue-N was recovered by corn (Zea mays L.). Both of these studies cite that the N recovery from the different residues appeared to be, at least in part, a function of the residue C/N ratio. In fact, the C/N ratio of plant residues is among a number of characteristics often cited to control the rate of plant residue decomposition (Taylor et al., 1989; Vanlauwe et al., 1996; Trinsoutrot et al., 2000). Other factors including N, polyphenol, and lignin concentrations and their ratios have also been shown to closely relate to the decomposition of plant material (Müller et al., 1988; Fox et al., 1990; Constantinides and Fownes, 1994; Trinsoutrot et al., 2000). With C/N ratios observed for residue applied in this study (Table 1), decomposition would be expected to progress rather slowly and result in low rates of residue N recovery in the seasons following application. It is probable that leaf, stem, and bur fractions decomposed at varying rates as a result of differences in composition. For instance, C/N ratios of selected leaf and stem samples were about 23 and 46 on Wasco sandy loam and about 31 and 73 on Panoche clay loam, respectively. On the basis of these C/N ratios, the leaf fraction would be expected to decompose more quickly than the stem fraction, and its relative contribution to uptake in the first year was likely greater than that of the stem fraction. Additional support for this proposition is provided by Egelkraut et al. (2000), who found different N mineralization dynamics for cotton leaves and stems. In their incubation study, they observed net mineralization from leaf residues to occur as soon as 18 d after residue incorporation, while a net N immobilization was observed for stems over the entire duration of their study (179 d). Thus, recovery of 15N from stem and bur residues was probably minimal in the first season, but may have contributed more in the second season after application. Besides chemical residue characteristics, differences in particle size sometimes result in differential decomposition rates of plant fractions (Ambus and Jensen, 1997; Angers and Recous, 1997; Vestergaard et al., 2001). Therefore, decomposition rates were predicted to vary for leaves that easily crumbled and stems and burs which were coarsely chopped by the shredding operation. In addition to decomposition rates, uniformity of 15N labeling of different plant tissues also affects 15N recovery. In 1998, 15N analysis of the different plant fractions confirmed nonuniform tissue labeling in this study (P < 0.01). A comparison across locations and treatments showed that the 15N label in leaves was significantly ({approx}10% across all treatments and both locations) lower than in stems, seed, burs, and fiber. Combined with greater decomposition rates, this reduced 15N label of leaves relative to other residue fractions resulted in a slight underestimation of residue N recovery.

Averaged across N treatments, the recovery of residue–15N by cotton grown on Wasco sandy loam compared with Panoche clay loam was almost double. This was not surprising, since enhanced mineralization of the residues from Wasco sandy loam would be expected simply based on their lower C/N ratios (Table 1), and a number of studies indicate that soil texture may influence N mineralization from organic material. Previously, Thomsen (1993) reported greater recovery of ryegrass (Lolium multiflorum Lam.) shoot-N by barley on a coarse sand than on a sandy loam soil. Egelkraut et al. (2000) found that soil texture influenced N mineralization from cotton residues. They reported that, of four different soils, those with greater clay concentrations exhibited longer periods of initial N immobilization and lower amounts of N mineralized from added cotton leaves and stems during a 179-d incubation. Wagger et al. (1985) observed that soil texture affected the mineralization of sorghum (Sorghum bicolor L.) residue but not that of wheat residue, and Van Veen et al. (1985) reported slower N immobilization and mineralization rates in a clay loam compared with a sandy loam. Proposed mechanisms for the differences in organic matter decomposition with soil texture include protection of organic matter in pores or aggregates which are not accessible to microorganisms, protection by adsorption on clay minerals, and effects of soil structure on the microbial turnover (Van Veen and Kuikman, 1990). In this study, differences in texture may have affected soil temperature dynamics and consequently organic matter decomposition.

In 1999, {approx}60% of the aboveground residue–15N recovered by the Acala plants was removed from the field in seed and lint on both soil types (Table 2). In the second year after 15N-residue application, partitioning of recovered residue–15N into the harvested portion of Acala plants was again about 60% on Panoche clay loam, but only about 44% on Wasco sandy loam. Combined across the two growing seasons, only 2.7 to 5.2% of the initially applied aboveground residue–15N was removed from the field as seedcotton (Table 2). After harvest of the 2000 crop, 48 to 86% of the 15N applied in form of aboveground cotton residue in 1998 was recovered in the soil, and between 12.9 and 50.9% of residue N was not accounted for (Table 2). In Pima, only about one-third of the aboveground residue–15N was recovered in the soil, <1% was removed from the field, and 62 to 75% was not accounted for at the end of the 2000 season (Table 2).

Because of variation within replications, neither locations nor treatments within locations differed in total recovery of residue–15N in the soil (Table 2). The wide range in soil recovery may in part be ascribed to difficulties in collecting representative samples in microplots amended with coarsely chopped cotton residue incorporated under standard agricultural practices. These shredding and incorporation practices, although random, may have produced pockets of residues, and hence a patchy decomposition pattern within a microplot.

In part, a consequence of the variation in residue–15N recovery in the soil, the range in residue–15N not accounted for after 2 yr was large (Table 2). The N missing in the balance may have been lost from the system in pools not measured in this study (e.g., volatilization, leaching), and might in part be associated with the challenges related to the collection of representative samples due to residue pockets. It is likely that a considerable fraction of residue–15N remained on the field in coarse, partially decomposed stem particles that were not analyzed in this study.

Recovery of Nitrogen-15 from Soil and Roots
Tracing the fate of residual fertilizer–15N in belowground pools, that is, 15N in soil and root components from 15N-fertilizer applied to Acala cotton in 1998, revealed that only small percentages of fertilizer–15N were recovered by the crop in the second and third year after application (Fig. 2). At the time of defoliation, average fertilizer–15N recoveries from soil and roots of 4.6 and 1.7% were measured for Acala grown on Panoche clay loam in the second and third year after fertilizer application, respectively. The corresponding recoveries observed on Wasco sandy loam were 5.3 and 2.7%, respectively. While differences were not significant in 1999, a greater, although agronomically insignificant proportion (corresponding to {approx}1 kg N) of residual fertilizer–15N was recovered by cotton grown on Wasco sandy loam than on Panoche clay loam in 2000 (P < 0.01). As discussed above, numerous studies have reported greater mineralization rates on lighter- than on heavier-textured soils, hence, greater recovery on the sandy loam was not surprising (Van Veen et al., 1985; Wagger et al., 1985; Egelkraut et al., 2000).



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 2. Recovery of fertilizer–15N applied in the spring of 1998 (Acala) or 1999 (Pima) and remaining in soil and root component or cycled back in aboveground cotton residue (residue) at the end of the season, by cotton grown in the 1999 (Acala) and 2000 (Acala and Pima) seasons. Bars indicate standard errors of three (Pima) or four (Acala) replicates.

 
In both years and at both sites, recovery of fertilizer–15N from the soil-root component proceeded in a similar pattern to that of total plant N uptake (Fig. 2). Between early square and peak bloom, plants took up {approx}70% of the fertilizer–15N recovered from the belowground pool (2-yr average across both locations and N treatments). In the second Acala crop after 15N-fertilizer application, 15N-recoveries were greater in the high-N treatments than in the low-N treatments on the Panoche clay loam (P < 0.01) but not on Wasco sandy loam. In the third crop after 15N-fertilizer application, the N recovery was greater in the low-N than in the high-N treatment on the Panoche clay loam (P < 0.01), while the two treatments again did not differ on the Wasco sandy loam. Pima recovery of 15N-fertilizer cycled through the belowground pools in the second season after application (2000) averaged 2.7% at the time of defoliation (Fig. 2). The contribution of labeled 15N from the belowground pool to the next Pima crop was minimal (<2.6%; <3.1 kg N ha–1) and lower for the N-56 than the N-168 treatment (Table 4). Independent of location, the low recoveries of 15N from labeled belowground pools in the second and third crop indicated that much of the 15N-fertilizer not taken up by the first crop may have been incorporated into more stable organic N pools. The contributions of such pools to the N uptake by plants depends on their size and rate of turnover, and were of limited agronomic importance in this study.


View this table:
[in this window]
[in a new window]
  		Table 4. Fertilizer-N derived from 15N-labeled soil-root component or residue component in cotton plants sampled at defoliation 2 and 3 yr after fertilizer application, and 1 and 2 yr after residue application.

 
Contributions of 15N-fertilizer cycled through soil and roots to Acala N assimilation were smaller on Wasco sandy loam than on Panoche clay loam (P < 0.05), and smaller in the low-N than the high-N treatments (Table 4). This was expected because of greater total N uptake on the Wasco sandy loam and only minimal amounts of fertilizer-N added to the N-56 treatments (Tables 4 and 5). In 1999, the 15N-fertilizer cycled through soil and roots contributed up to 7 kg N ha–1 and in 2000 up to 3 kg N ha–1 to the total N assimilated by Acala cotton.


View this table:
[in this window]
[in a new window]
  		Table 5. Fate of fertilizer-N applied in spring 1998 to Acala and spring 1999 to Pima cotton.

 
Total Recovery of Fertilizer-Nitrogen-15 Applied to the First Crop in the Second and Third Crop
To obtain total recovery of fertilizer–15N cycled through the soil and subsequent crops, results from microplots amended with labeled cotton residue were combined with those from microplots with the labeled belowground component. Of the fertilizer–15N applied in the previous year, Acala cotton assimilated 4.6 to 6.7% in the aboveground biomass at the time of defoliation in 1999 (Table 5). Location did not affect the recovery of residual 15N-fertilizer in 1999 (second crop after 15N-fertilizer application). However, when analyzed across both locations, fertilizer–15N recoveries in Acala shoots of the N-56 treatments were lower than in the N-168 treatments (P = 0.053). Roughly 60% of the fertilizer–15N recovered in the Acala crop, was removed from the field as yield in 1999. In 2000, plants took up 1.9 to 3.3% of the 15N-fertilizer applied in 1998. The recovery was significantly greater in plants grown on Wasco sandy loam than on Panoche clay loam (P < 0.01). This resulted in the removal of a greater percentage of fertilizer-N from the field on Wasco sandy loam, even though proportionately less 15N was found in seed compared with seed from Panoche clay loam plots (P < 0.05). However, neither total fertilizer–15N recovery in the plants nor the proportion removed from the field was affected by N treatment. Of the 15N-fertilizer applied in the previous year, Pima recovered 2.5% in the N-56 treatment and 3.2% in the N-168 treatment (Table 5). This is about half as much as the recovery observed for Acala. Possible reasons for this difference between Acala and Pima may include the greater C/N ratio of the Pima residue than the Acala residue (Table 1), discrepancies in residue characteristics not determined in this study (i.e., concentrations of lignin, cellulose, hemicellulose, polyphenol), and variations in weather conditions between 1998–1999 and 1999–2000.

After three growing seasons, about one-third of the fertilizer–15N applied to Acala in the first of those seasons was removed from the field as seed and lint (Table 5). The majority of the 37 to 46% recovered in the field after harvest in 2000 remained in different soil N pools, while a small fraction remained in the form of aboveground residue. Thus, after three years, 22 to 31% of the labeled fertilizer-N remained unaccounted for. At the end of the second Pima growing season, about 40% of the labeled fertilizer was recovered in the soil and cotton residue remaining on the field, and only about 27% had been removed as yield in 1999 and 2000. Approximately one-third of the fertilizer-N was not accounted for (Table 5). As for the Acala trials, it is assumed that some of this N was lost from the system (e.g., volatilization, leaching) while some remained in the field in cotton roots which were not quantified in this study and possibly in cotton residues from previous years that were not entirely decomposed at the time of soil sampling.

Recovery data and apparent losses reported in this study, fall in the range observed by Raun et al. (1999) for a similar experiment. In their winter wheat system, they found total recoveries of 15N fertilizer after three growing seasons of 46.7 to 90.5% and recoveries in the soil of 24.6 to 58.2%, depending on site and N fertilization treatment.

In both 1999 and 2000, fertilizer–15N recoveries were greater in microplots with labeled belowground component than those amended with labeled aboveground residues (Fig. 2). Averaged across both years and locations, Acala plants derived 5.7 times the amount of 15N from the fertilizer–15N cycled through the soil and roots than through aboveground residue. In part, this was probably due to the approximately twofold (average across all Acala treatments) greater amounts of 15N-fertilizer remaining in the soil compared with the amount reapplied in residue after harvest in 1998. However, these results also suggest differences in availability between 15N in roots and soil and 15N from aboveground residue. Similar results were reported by Hubbard and Jordan (1996), who found that N in wheat roots and soil was more readily available to corn than N from wheat straw. Furthermore, Eagle et al. (2001) found that residual fertilizer-N in soil pools was a more important N source to the rice crop than N derived from incorporated straw.

Treatment effects on N recovery from 15N soil pools observed in the second crop after fertilizer application were probably the consequence of distinct mineralization–immobilization dynamics (Fig. 2). Higher fertilizer application rates and possibly greater soil mineral-N concentrations as a result of application rates in the previous cropping season may have resulted in shorter periods of N immobilization and greater net N mineralization in the N-168 than the N-56 treatments, and hence an increased recovery of residual fertilizer–15N. Potential treatment differences in root tissue composition (not determined) may have affected recovery from the soil pools as well. Similar recovery percentages for the two N treatments in the third cropping season after fertilizer application (2000) suggest that for both treatments, the remaining 15N was part of pools with similar mineralization characteristics.

Smaller contributions of fertilizer–15N cycled through soils and roots to total plant N uptake on Wasco sandy loam compared with Panoche clay loam and in the low-N vs. the high-N treatments were expected. Total cotton N uptake on Wasco sandy loam was greater than on Panoche clay loam (P < 0.001; Table 4) while amounts of fertilizer-N applied were almost the same (Table 5), thus a smaller percentage of N derived from residual fertilizer was not surprising. Similarly, plants had to rely much more on soil than fertilizer-N in the low-N treatments than the high N treatments since only minimal amounts of fertilizer-N had to be applied in the N-56 treatments to achieve the target rate.

Labeled Nitrogen Remaining in Belowground Pools
The total fertilizer–15N remaining in the soil at the end of the 2000 growing season was calculated by adding remaining amounts in the microplots amended with labeled aboveground residues to those determined in the microplots with labeled soil and roots. Total fertilizer–15N remaining in the soil at the end of the third season (2000) after fertilizer application did not differ between the two locations and averaged 40% (Table 5).

Although averages ranged from 48 to 86%, total 15N recovered from aboveground residue in the soil profiles did not differ between Acala locations and N treatments (Table 2). Two growing seasons after application of aboveground residue, analyses showed that virtually all of the 15N recovered from the labeled residue in the soil was found in the top 0.3-m and only traces were found at depths > 0.3 m (Fig. 3). Consistent with the results from Acala, 1 yr after residue application in Pima, nearly all labeled N was recovered in the top 0.3 m of the soil (Fig. 4).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 3. Acala fertilizer–15N recovery in microplots labeled either in the soil and root component or amended with labeled aboveground cotton residue. In microplots with labeled belowground components, fertilizer was applied in spring 1998, labeled cotton residue was exchanged with nonlabeled residue at the end of the 1998 season, and recovery in the soil was determined after harvest in 1999 and 2000. In microplots amended with 15N-labeled aboveground cotton residue in the fall of 1998, recovery was determined after harvest in 2000. Bars indicate standard errors of four replicates.

 


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 4. Fertilizer–15N recovery in Panoche clay loam of microplots labeled either in the soil and root component or the aboveground residue component. Pima cotton was fertilized in spring, and aboveground residue exchanges were made in the fall of 1999. Soil samples were collected after Pima harvest in 2000. Bars indicate standard errors of three replicates.

 
Analogous to the microplots amended with labeled residues, total fertilizer–15N recovery in the soil of microplots with labeled belowground component was not different between Panoche clay loam and Wasco sandy loam. However, while the biggest portion of labeled N was still recovered in the surface 0.3-m of soil, farther downward movement of 15N in the microplots where the 15N-fertilizer was originally applied was observed (Fig. 3). Interestingly, the proportions of 15N in lower soil horizons had not increased from 1999 to 2000. The recovery of large percentages of 15N in the topsoil and small recoveries in the second and third crop suggests that much of the N applied as fertilizer and residue was stabilized into more recalcitrant soil N pools. Traces of 15N in soil horizons near the maximum rooting depth observed for cotton in this study within 2 yr of fertilizer application signal that a potential for leaching losses may exist under the production conditions examined in this study. Particularly when excess amounts of residual mineral-N are present in the soil, leaching losses could be possible.


   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Three seasons after 15N-fertilizer application to Acala, between 37 and 46% of the 15N was recovered in soil and crop residue fractions. The cumulative recovery in seedcotton (N leaving the field) during three seasons varied between 32.2 and 37.1% of the 15N-fertilizer applied, while proportions unaccounted for ranged from 22.3 to 30.6% of the applied 15N-fertilizer. Observed recoveries of fertilizer–15N by the second crop after application averaged 5.8% for Acala and 2.9% for Pima. In the third crop after application, Acala recoveries averaged 2.6%. For the N-168 treatments of Acala, this translates into a recovery of originally applied fertilizer of {approx}8 kg N ha–1 in the second crop and 3.3 kg N ha–1 in the third crop after application. Recovery of residue 15N by Acala averaged 4.5% (3.6 kg N ha–1) in the first and 2.5% (2 kg N ha–1) in the second season after application. Nitrogen derived from crop residue incorporated in the previous two cotton growing seasons varied between 3.6 and 5.5%. Even in the N-168 treatments, the estimate for cotton N uptake from soil pools (not including previous year's residue N and current year's fertilizer-N) was >50%.

Virtually all the residue 15N recovered in the soil was found in the surface 0.3 m. Although most of the 15N-fertilizer applied in the first cotton season was recovered in the zone of highest rooting density (unpublished data, 1999), some 15N had moved into soil layers with low rooting density by the end of the second season. Approximately 37% of the fertilizer-N applied in the first year was recovered in the top 0.6 m of the soil profile even after three cropping seasons. When considered in combination with the fact that relatively low amounts of residual fertilizer-N were recovered in the second and third crop following fertilizer application, this suggests that much of the original fertilizer 15N was stabilized into more recalcitrant soil N fractions. Incorporation into more stable organic matter pools may have been favored by the high C/N ratio of aboveground cotton residue. Recovery of N from more stable organic matter pools should occur with only minor losses since seasonal mineralization patterns are likely to coincide with those of plant N uptake.

Under the production conditions examined, direct contributions of cotton residue-N and residual fertilizer-N to the N assimilation of subsequent cotton crops is of limited agronomic importance.


   ACKNOWLEDGMENTS
 
The authors would like to thank Wayne and Doug Wisecarver and the staff of the West Side Research and Extension Center and the University of California Cooperative Extension at the Kings County office for their cooperation, and Robert L. Nichols of Cotton Incorporated for his support.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
This research was supported in part by grants from Cotton Incorporated, the California Department of Food and Agriculture Fertilizer Research and Education Program, and the California Crop Improvement Association.

Received for publication December 18, 2003.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fritschi, F. B.
Right arrow Articles by Hutmacher, R. B.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Fritschi, F. B.
Right arrow Articles by Hutmacher, R. B.
Agricola
Right arrow Articles by Fritschi, F. B.
Right arrow Articles by Hutmacher, R. B.
Related Collections
Right arrow Nutrients
Right arrow Other Crop Management
Right arrow Other Cropping Systems
Right arrow Cotton
Right arrow Nutrient Cycling

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH