HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Hutmacher, R. B. Articles by Delgado, R. Search for Related Content PubMed Articles by Hutmacher, R. B. Articles by Delgado, R. Agricola Articles by Hutmacher, R. B. Articles by Delgado, R. Related Collections Best Management Practices Cotton Nutrient Management Nitrogen

PRODUCTION PAPER

Response of Recent Acala Cotton Varieties to Variable Nitrogen Rates in the San Joaquin Valley of California

^a Univ. of California Coop. Ext., Shafter, CA, and Univ. of California, Davis, CA
^b Dep. of Agron. and Range Sci., 1 Shields Ave., Univ. of California, Davis, CA 95616
^c Univ. of California, Madera, Kings, Merced, Tulare, Fresno, Kern, and Glenn Counties, respectively
^d Univ. of California, Shafter, CA
^e Dep. of Agron. and Range Sci., Univ. of California, Davis, CA (currently USDA, Stoneville, MS)
^f Cotton Inc., Cary, NC
^g Univ. of California, Shafter, CA

^* Corresponding author (rltravis{at}ucdavis.edu).

Received for publication December 17, 2002.

	ABSTRACT

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

 
Nitrogen fertilizer is routinely applied to crops grown in rotation with upland cotton (Gossypium hirsutum L.) in the San Joaquin Valley (SJV) of California. However, increasing N fertilizer costs, the potential overuse of N resulting in excessive vegetative growth and harvest delays, increasing pest pressure, and concern for nitrate contamination of groundwater support a reassessment of current N fertilization practices. The primary goal of this research was to provide information that would assist SJV cotton growers in updating and improving N management practices. Plot site selection included two university field stations and six on-farm locations representing all SJV cotton-producing counties. Nitrogen treatments of 56 to 224 kg N ha^–1 were over a 5-yr period. Cotton lint yield responded positively to increasing N applications in only 41% (16 out of 39) of the test sites. Yield response to fertilizer N was related to residual soil N in the upper 0.6 m of soil as follows: below 70 kg ha^–1 residual NO₃–N, 9 of 17 sites responded positively to increasing applied N; at 70 to 125 kg ha^–1, 5 of 11 sites responded; and at greater than 125 kg ha^–1, only 2 of 11 sites responded. Changes in soil NO₃–N levels from postplanting to postharvest were generally larger within the upper 1.2 m of soil than at lower depths. However, net increases in soil NO₃–N also occurred in the 1.2- to 2.4-m range at sites prone to leaching.

Abbreviations: REC, Research and Extension Center • SJV, San Joaquin Valley • WS REC, West Side Research and Extension Center

	INTRODUCTION

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

 
NITROGEN IS GENERALLY considered a yield-limiting factor in upland cotton production systems. In California cotton systems, N fertilization is focused on optimizing lint yield, while avoiding excessive applications that can make defoliation more difficult and delay harvest, yet maintaining historically excellent fiber quality. For the most part, growers have been successful in meeting yield and quality goals using present N management practices.

Total annual N use in California for all crops has increased in recent years and is currently estimated at 450000 Mg (USDA Econ. and Stat. Syst., 2001). Average N fertilizer application rates for California upland cotton increased from about 120 kg ha^–1 in the late 1970s to about 200 kg ha^–1 by the mid 1990s (Fritschi et al., 2003). Current management strategies for SJV cotton production call for N applications ranging from 171 to 228 kg ha^–1 (Weir et al., 1996); however, actual rates will depend upon soil type and previous cropping history. In some cases, annual N application for cotton may exceed 228 kg ha^–1 (Weir et al., 1996). Assuming annual California cotton production has averaged 400000 ha over the past decade, cotton would account for nearly 18% of all agricultural N used in California during that time. Increases in applied N in the SJV above levels prevailing 20 to 30 yr ago are a result of a combination of factors. For one, the higher yields associated with the newer varieties suggest greater need for N, yet there is little evidence to support this contention. In addition, use of the growth regulator mepiquat chloride has altered irrigation and N fertilizer management practices in many cotton production areas of the state. Since mepiquat chloride can be used to control excessive growth (Kerby et al., 1996), there is less need for restricted water and N applications.

Most of the newer Acala varieties are more determinate in growth habit than the older varieties. Plant populations typically used are higher, and length of the growing period has been reduced. This raises the question of whether current N applications accurately reflect crop needs (Marsh et al., 2000). One possible scenario is that N is being applied in excessive amounts. If so, this could have a negative impact on the ecosystem (Franco and Cady, 1997; Burow et al., 1998). If the rotation sequence lacks sufficient deep-rooted crops to intercept applied and residual N, the movement of NO₃–N through the soil profile can result in groundwater contamination that may lead to human health problems (Comly, 1987; Kross et al., 1993). The relatively low cost of N fertilizer sources over the past decade has increased the likelihood for overuse of N and resulting groundwater contamination. Growers often consider making higher N fertilizer applications, even in excess of crop N requirements, as cheap insurance against N limitations to yield. More recently, increases in energy costs, which constitute a large part of N fertilizer production costs, have been passed onto growers as increases in fertilizer cost, making it doubly important to reevaluate N recommendations.

Development of an effective N management system for cotton will require close attention to plant needs while balancing those needs against environmental concerns. Nitrogen deficiency reduces vegetative and reproductive growth and induces premature senescence, potentially reducing yields (McConnell et al., 1993; Gerik et al., 1994). Likewise, excess N can negatively impact cotton growth and development (Hutchinson et al., 1995; McConnell et al., 1995). Higher-than-desired N levels during bloom/early boll fill can promote vegetative development at the expense of fruit retention (Gaylor et al., 1983; Mullins and Burmester, 1990; Boquet and Breitenbeck, 2000). Excess mid- and late-season N can delay progress toward cutout, delay the timing of defoliation, and increase the cost of defoliation (Roberts et al., 1996). In addition, excess N can increase problems with late-season silverleaf whitefly (Bemisia argentifolii) and aphid (Aphis gossypii) infestations that can complicate defoliation and reduce lint quality (Roberts et al., 1996; Cisneros and Godfrey, 2001; Bi et al., 2001). The use of mepiquat chloride further complicates this issue.

Although there has been considerable research on cotton N use, the literature does not provide a clear assessment of plant need. Bassett et al. (1970) in California and Halevy (1976) and Halevy and Kramer (1986) in Israel reported that total N accumulation in aboveground plant tissue in older varieties ranged from 10 to 12 kg N per 100 kg lint under moderate yields of 1200 to 1500 kg lint ha^–1. Lutrick et al. (1986), following a 3-yr demonstration, suggested that the "sufficient-N zone" established in prior years was higher than desirable and should be redefined to reduce excess N applications. Later, Unruh and Silvertooth (1996) in Arizona indicated an average of 12 kg N per 100 kg lint, with aboveground plant accumulation of 205 kg ha^–1. Mullins and Burmester (1990) concluded that newer non-Acala southeastern USA cotton varieties take up an average of 18 kg N per100 kg lint. Although they indicated that this high ratio is typical of southeastern varieties, this is not relevant to California. Lint yields of 1600 to 1900 kg lint ha^–1 are common in the SJV. This would result in uptake estimated at 290 to 340 kg N ha^–1. If fertilizer rates were adjusted to supply that amount of N on an annual basis in California, current evidence suggests that significant N leaching losses would occur.

Although the newer California upland varieties typically produce high yields, there is limited information on N uptake in these systems. Hutmacher et al. (1994)(1995), using drip irrigation with recent Acala varieties Germain ‘GC-510’ and CPCSD ‘Maxxa’, reported total N accumulation in aboveground plant tissue close to 10 kg N per 100 kg lint. These values were closer to those reported by Bassett et al. (1970) for the older California varieties than to recent averages from the southeastern USA. Thus, in the large plants and high-yield conditions in much of SJV, the average N uptake requirement per unit yield may be lower than in other parts of the Cotton Belt. To achieve sustainable high yields while limiting leaching of NO₃–N below the root zone, more information is needed regarding optimal amounts of N fertilizer required to achieve acceptable yields.

In the absence of clearly defined N requirements for cotton, SJV recommendations continue to be based upon earlier studies (Weir et al., 1996). While current recommendations may reflect plant needs, they do not consider the potential contribution of other N sources. It is becoming increasingly clear that a substantial amount of N can be present in the soil, either as inorganic N remaining from previous years or in organic forms (Marsh et al., 2000; Fritschi et al., 2003) that will at least partially satisfy this requirement. Thus, a first approximation of N needs should consider that contribution. This has been hindered by difficulties in accurately assessing soil N reserves. Previous efforts in California (Weir et al., 1996) have attempted to relate soil NO₃–N levels in the upper region of the profile to yield responses, but no widely accepted recommendations have been developed. The problem is that the variable release of NO^–₃, NH⁺₄, and organic N also impacts growth and yield.

This communication summarizes the results of a 5-yr study that assessed the N management practices for upland cotton production in the SJV of California. Large-scale plots were established at seven or eight locations per year, with at least one site in each of the six cotton-producing counties in the SJV. Year-by-year yields as influenced by N application, residual N in the upper 0.6 m of the soil profile, and climatic conditions are reported here. The results will show that excessive N is often applied in SJV crop rotations that include cotton. Methods of assessing crop N needs are discussed.

	MATERIALS AND METHODS

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

 
This study was predicated upon the requirement that the research be done at sites and under conditions that were typical of cotton production in the SJV of California. This led to several unusual experimental circumstances. First, there was no attempt to select research sites based on spring residual soil N levels, nor did site selection criteria specify specific soil NO₃–N levels or ranges. Second, in deference to requests from grower/cooperators, a zero-N treatment was omitted. The plots were relatively large and represented a significant investment for the growers, and they indicated that they did not want to risk a major yield loss with a zero-N treatment. Third, with the introduction of several new varieties of commercial importance in California over the past few years, it was not possible to require grower/cooperators to use the same variety in each location every year. However, only Acala upland varieties were planted in this study. Finally, since the trials were primarily in grower fields, where rotation is a critical component of the cropping system, it was not possible to maintain the plots on the same sites every year at all locations. Repeat sites were used where possible.

Plot Size and Locations
The experiments were conducted on the University of California West Side Research and Extension Center (WS REC) in western Fresno County, Shafter Research and Extension Center (REC) in north-central Kern County, and at on-farm sites in Fresno, Kern, Kings, Madera, Merced, and Tulare Counties. The University research center plots were four rows wide by approximately 90 m long. Grower plots were six or eight rows wide and ranged in length from 200 m to more than 450 m (the full length of the irrigation runs). Row widths ranged from 0.76 to 1.02 m. Four replications were used in a randomized complete block design in all cases. The WS REC plots were in the same location for 1996 and 1997 but were moved to an adjacent field for 1998 through 2000. The Shafter REC plots were moved to a new site in 1999 due to crop rotation requirements of the research center. Trials were repeated on other sites as permitted by grower rotation schedules. The 1998 to 2000 Kings County and the 1997 to 1999 Tulare County trials were maintained on the same sites. In addition, there were seven trials located on the same site for two consecutive years. The remaining 11 trials were on single-year sites. Table 1 summarizes the rotation sequences and soil types for all experimental sites.

Nitrogen Treatments
All N fertilizer treatments were applied after planting but before the first in-season irrigation. Fertilizer applications were typically made in late April through late May, depending upon site and year. The first within-season irrigation was applied in mid- to late May to early June. Fertilizer N rates were 56, 112, 168, or 224 kg ha^–1 after correction for soil residual N in the upper 0.6 m of the soil profile. When residual soil NO₃–N exceeded 56 kg N ha^–1, that amount was set as the baseline treatment (identified as 56 kg ha^–1). The remaining treatments were established by applying an additional 56, 112, and 168 kg N ha^–1, producing treatments that will hereafter be referred to as the 112, 168, and 224 kg ha^–1, respectively. Only in 1996, three supplemental treatments were established by applying an additional 56 kg N ha^–1 to all but the 224 kg ha^–1 treatment before the second irrigation. These treatments were omitted in subsequent years due to the lack of a yield response in 1996 and grower requests for simplified fertilizer applications. Liquid urea solutions were applied as the fertilizer N source at most sites. The solutions were dripped through tubes behind or through a soil shank pulled to a depth of 15 to 25 cm adjacent to every plant row at a distance of 20 to 35 cm from the row. Various implements were used to close the shanked soil opening following application. Granular dry urea was applied through a hollow shank at a depth of 12 to 20 cm at the Shafter REC site in 1996, the Merced County site in 1997, and the WS REC site in 1998 and 1999. Anhydrous NH₃ was used at the Kern County site in 1997. Field application equipment was calibrated in nonplot test rows before test plot applications in all cases.

Soil Sampling
Samples were collected from all plots within the first 9 to 17 d after planting for residual soil NO₃–N analyses. Six locations per field replication in the low-N treatment were sampled at each site. Samples were taken in 0.3-m increments to a depth of 0.6 m with a hand auger, composited within each replication, and air-dried at 35 to 40°C for NO₃–N analyses. Deep sampling was done two times per year at each site to evaluate NO₃–N uptake and movement within the soil: (i) within 11 to 27 d after planting across all sites and years but before N fertilization in all cases and (ii) within 20 d after harvest. Hand- or power-auger–collected samples (4.45-cm-diam. tube) were taken on the edge of the beds within 7.5 to 15 cm of the planted row. Subsamples ranging from 40 to 50 g in weight were taken from each primary sample to determine gravimetric water content using the method of Gardner (1965). Four locations within each plot were sampled in the first, second, and third years of the study while three locations per plot were sampled in Years 4 and 5. Replicate plots for each treatment were sampled at each location in 0.3-m increments to a depth of 1.2 m and then in 0.6-m increments to 2.4 m. After 1996, it was determined that average levels and patterns of soil NO₃–N across treatments were not significantly different when using three vs. four field replications. As a result, deep soil sampling (depths greater than 0.6 m) at postplanting and postharvest was not done in the fourth field replication in subsequent years. The soil samples were refrigerated (2–4°C) until subsampled or analyzed. Separate subsamples were used for each analysis class.

Soil and Water Analyses
Soil NO₃–N was determined on 2 M KCl extracts (Keeney and Nelson, 1982). Total micro-Kjeldahl N (Bremner and Mulvaney, 1982) and NH₄–N (Keeney and Nelson, 1982) were assayed on a limited number of samples from select field locations in 1997 and 1999. These data were used to compare relative levels of total N and NH₄–N with NO₃–N. Separate subsamples were prepared by compositing samples from three to four sample locations for each depth. Those were air-dried at 35 to 45°C and analyzed for total extractable PO₄–P by 2% acetic acid extraction (Olsen and Sommers, 1982) and ammonium acetate–exchangeable K (Cassman et al., 1990; Thomas, 1982). Nitrate contributed by irrigation water was estimated from NO₃–N analyses and volume of water per application (see "Cultural Practices" section). Three replicate 50-mL water samples were collected in Nalgene bottles from midstream of the irrigation water source. Samples were collected at each site during one irrigation per month in June, July, and August each year. Water delivery was allowed to run for a minimum of 15 min before sample collection. If there was more than one irrigation per month at any site, no additional samples were collected. This was assumed to represent an average NO₃–N concentration for all other irrigations that month unless a change in water source occurred. In that case, additional samples were collected and analyzed. Water NO₃–N was measured using an ion-selective electrode (Keeney and Nelson, 1982).

Groundwater Depth
Shallow groundwater may contribute N to the cropping system in some situations. Shallow groundwater (above a depth of 2.4 m) was evident at postplanting and postharvest soil sampling times at two Fresno County sites (1996 and 1997–1998) and one Merced County site (1999–2000). Soil sampling was limited to a depth of 2.4 m, so if shallow groundwater existed at lower depths at other sites, it was not detected. Groundwater depths were taken as static water table depth 1 to 3 d after drilling a 4.45-cm hole 2.4 m in below the bed surface. Average postplanting groundwater depths at the Fresno County sites in early May were 1.9 m (1996), 2.15 m (1997), and 1.6 m (1998). In late October and early November, average postharvest groundwater depths at the Fresno County sites were 2.25 m (1996), 1.85 m (1997), and below 2.4 m (1998). Postplanting depths to static groundwater averaged 1.5 m (1999) and 1.65 m (2000) at the Merced County site while postharvest averages were 2.25 m (1999) and 2.1 m (2000).

Bulk Density and Moisture Release Determinations
Bulk density was determined at the same time as the spring deep soil sampling. Three replications of the 56 kg ha^–1 treatment at each field site were selected. Samples were collected in 0.3-cm increments to a depth of 0.9 m from two locations per plot. Bulk density was determined by a double-cylinder core method (Blake, 1965). Deep samples (0.9–2.1 m) were taken in 0.3-m increments to avoid soil compression. Samples were dried at 105°C and bulk density calculated. During the first year at each site, soil subsamples collected along with bulk density samples were used to determine moisture release characteristics at two levels of soil water tension (0.03 and 1.5 MPa) using a pressure plate method (Klute, 1965).

Supplemental Fertilization
Supplemental P and K fertilizer applications were made at three sites over the 5-yr period. Potassium was added when postplanting soil samples indicated exchangeable-K levels of less than 130 mg K kg^–1 in the 0.13- to 0.40-m depth (Miller et al., 1997). Phosphorous was applied when the soil level averaged less than 10 mg PO₄–P kg^–1 in the upper 0.45 m of the profile (Olsen and Sommers, 1982). The 1998 Shafter REC site (116 mg K kg^–1) and 1999 Kern County (127 mg K kg^–1) site received supplemental K fertilizer as K₂SO₄ at rates of 225 and 260 kg K₂0 ha^–1, respectively, banded at 0.20- to 0.25-m depth down the furrow center. The 1999 WS REC (9 mg PO₄–P kg^–1) site received 125 kg ha^–1 of triple superphosphate also as a banded application. No foliar N or K applications were made at any of the trial locations.

Root Weight Distribution
Soil samples for estimating root weight were collected from three field replicates at two sites per replication. Sampling was done in 1997 through 2000 at two sites (WS REC and Shafter REC) and at select other sites in 1997 and 1999. The hand auger samples (4.45-cm diam.) were taken in 0.15-m increments to a depth was 1.8 m at a distance of 7.5 to 15 cm away from and perpendicular to the plant row. Samples were transported in an insulated cool box and refrigerated at 4 to 6°C until analyzed. Soil samples were placed in 2.5-L containers, and adequate water was added to fully immerse the soil. After soaking 12 to 24 h, samples were poured through a 0.2-cm sieve. Roots were physically separated from remaining soil particles based on visual observation and root identification, dried at 40 to 45°C, and weighed.

Cultural Practices
Dairy waste consisting of uncomposted cow manure or cow manure plus dairy lagoon waste was spread after harvest and before listing beds before the cotton crops at the Tulare County experiment sites in 1996, 1997, and 1999 and at the Madera County sites in 1996, 1999, and 2000. Estimated total N content of those applications averaged 41 kg N ha^–1 (ranging from 27–58 kg N ha^–1). These values were obtained by determining N content of three replicate samples taken after application in each field and using grower estimates of average waste application rates. Since a minimum of at least 2 mo, and as much as 5 mo, elapsed between application of dairy waste and the postplanting soil sampling, some of the organic N would have mineralized and should be reflected in the postplanting soil NO₃–N measurements.

Additional cultural practices were similar across locations. All fields were irrigated before planting to replenish soil water extracted by prior crops and not replenished by winter rainfall. Furrow irrigation was used at all locations. Water delivery was by siphon tubes, gated pipe, or ditch supply methods. Since soil texture, rooting depth, and therefore available soil water-holding capacity differed across sites, the total number of irrigations also differed. In-season water applications between the first and final irrigations ranged from 3 to 11. Field sites requiring only three irrigations were usually deep, finer-textured soils with higher water-holding capacity and potential for deep (1.2–1.8 m) rooting depths. Estimated within-season irrigation water application (not including preplant irrigation) ranged from as low as 650 mm (some sites in a low evapotranspiration year, 1998) to more than 900 mm in sandy loam soils in warm years such as 1996. Estimates of irrigation applications were made using irrigation run times and average furrow flow rates.

The plant growth regulator mepiquat chloride was applied near first bloom (generally at 15 to 16 main-stem nodes) at all but 1 (Shafter REC in 1996) of the 39 sites over the 5-yr trial. Growth regulator treatment decisions and application rates were determined using plant height, main-stem node number, and square retention counted on the first five fruiting branches. Only the first fruiting positions closest to the main stem were counted (Kerby et al., 1996). Although growth regulator application rates were different across sites and years, they were made at uniform rates across all N treatments to avoid introduction of an additional variable.

Plant Mapping and Yield
Plant vigor was monitored at most, but not all, treatment sites. Plant height, main-stem node counts, and fruit retention (as square or boll retention) (Kerby et al., 1996) were taken at approximately 2 to 3 wk before first bloom, within 1 wk before or after first bloom, 2 to 3 wk after first bloom, and at vegetative cutout. Nodes above white flower counts (Kerby et al., 1996) were added after first bloom. Fruit retention was calculated as percentage retention of flower buds (squares) and bolls on first fruiting position sites on the lowest five fruiting branches and uppermost five fruiting branches present at each measurement time.

Seed cotton was mechanically harvested with commercial-type spindle pickers. Seed cotton was weighed in the field, and 2.5-kg subsamples were taken for moisture content, percentage lint, and fiber quality measurements. The seed cotton was ginned at the University of California Shafter REC research gin and the lint, seed, and trash contents determined. Lint and seed yields were calculated and adjusted for moisture content. Although fiber quality data will not be discussed in this paper, subsamples from each field sample were sent for high-volume instrumentation (HVI) analyses.

	RESULTS AND DISCUSSION

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

 
Initial Soil Nitrate Nitrogen Concentration
Soil NO₃–N concentrations in the upper 0.6 m of the soil profile at planting time varied widely among sites and years. Nitrate N ranged from 77 to 251 kg ha^–1 in 1996, 48 to 158 in 1997, 40 to 114 in 1998, 40 to 271 in 1999, and 39 to 161 in 2000 (Table 2). Generally, sites with relatively low residual soil NO₃–N followed either cotton or wheat (Triticum aestivum L.) while moderate-to-high residual soil NO₃–N levels were typical following field corn (Zea mays L.) (for silage or grain), processing tomato (Lycopersicon esculentum Mill.), or forage alfalfa (Medicago sativa L.) (Table 1). This underscores the importance of considering cropping history, as it relates to previous fertilizer applications, when projecting N needs for cotton. Other factors to consider are the impact of farm locations on rotation crop options and N fertilization, shallow groundwater containing soluble N, and the availability of organic amendments containing N (such as dairy lagoon waste or manure).

The timing of and amount of irrigation water applied may impact the downward movement of NO₃ in the soil profile; however, this issue was not thoroughly addressed in this study. Data collection on seasonal changes in soil water content was limited. In addition, the volume of water per irrigation was estimated from irrigation run times and average flow rates. The latter is less accurate than direct metering of applied water. However, some insight may be gained by comparing estimated applied water with soil water storage capacity. The latter was determined using soil moisture release data and gravimetric analyses of water content. Gravimetric analyses were done at selected times during the growing season (after planting but before the first irrigation, 2 to 17 d before first irrigation and within 10 d after harvest). The potential for early-season water movement through the soil profile can be evaluated by comparing after-planting gravimetric soil water content (in this case, to a depth of 2.4 m) with similar measurements taken before the first within-season irrigation (typically late May to early June) to a depth of 1.5 m. Measurement of soil water content at 30 out of the 39 site-year locations before the first irrigation (2 to 17 d before the first irrigation) indicated that there was adequate capacity to store estimated applied water from that irrigation in the upper 1.5 m of the profile. This suggests that there was limited potential for downward flow (below about 1.5 m) of water and NO₃–N early in the season at those sites. It is recognized that care must be exercised in interpreting these data since this approach does not consider the potential effects of preferential flow paths whereby downward movement of water could bypass some soil pores where water could be stored. At the other nine site-year combinations (Merced and Fresno Counties in 1996 and 1998, Kings County in 1996, Madera County in 1998, and Shafter REC in 1996, 1998, and 2000), soils were either lower in water storage capacity (coarser-textured soils) or had higher initial water contents. Thus, they were unable to store as much of the water from the first irrigation.

Soil water depletion from the upper 1.2 m at postharvest sampling exceeded 75% of the calculated water capacity (between 0.3 and 1.5 MPa soil water tension) in the upper 1.2 m of the soil profile at 27 of the 39 sites across the 5 yr. At the same 27 sites, soil water depletion from the 1.2- to 2.4-m zone ranged from 26 to 57%. Postharvest soil water contents in both the upper 1.2-m and 1.2- to 2.4-m zones of the soil profile were higher at the other 12 site-year combinations (Fresno and Madera Counties in 1996 and 1998; Merced County in 1998, 1999, and 2000; Tulare County in 1998; Kern County in 1998; Kings County in 1996 and 1999; and Shafter REC in 1998), particularly in the 1.2- to 2.4-m zone. Thus, in the latter situation, there would be a higher probability of water and solute leaching losses when irrigation amounts exceeded plant uptake and soil water storage capacity. These data will be discussed more fully in a subsequent paper addressing soil NO₃ dynamics.

Other Forms of Soil Nitrogen
Caution must be used when basing estimates of soil N availability simply on NO₃–N measurements. Clearly, there are other forms of N present in the soil, the relative mix of which may be impacted by changing soil water status, temperature, and biological factors. To partially address this issue, total Kjeldahl N and NH₄–N were determined for comparison purposes in a more limited number of field tests (postplanting and postharvest sampling, upper 1.2-m soil profile, 1997 and 1999; data not shown). Ammonium nitrogen varied widely across soils, test sites, and years, but values were relatively low (generally 5 to <20% of soil NO₃–N levels) and not well correlated with total Kjeldahl N or NO₃–N. Treatment and location differences in Kjeldahl-N analyses were in reasonable agreement (positive correlation, r² = 0.71) with relative NO₃–N values at sites where specific comparisons were made. However, Kjeldahl analyses are considerably more expensive and time-consuming and are not commonly used for soil testing in the SJV.

Impact of Plant Uptake on Soil Nitrogen Concentration
Plant uptake can also impact soil NO₃ concentrations. Postharvest root mass (root fresh weight) was determined in soil cores from the upper 1.8 m at 15 sites over the 39 site-year combinations. Samples were collected between 35 and 57 d after first bloom in three replications of the 168 kg N ha^–1 treatment at each location (Table 4). Since the samples were collected from 4.45-cm soil cores taken at a distance of 7.5 to 15 cm perpendicular from the plant rows, they represent a limited soil volume and provide only a relative approximation of root mass. More than 92% of the root mass occurred in the upper 1.2 m of the soil profile at the Tulare County, Madera County, and Shafter REC location sites. Deeper rooting was apparent at the Kings County, Merced County, Kern County, and WS REC locations where between 9 and 21% of measured root mass occurred in the 1.2- to 1.8-m depth (Table 4). Deeper rooting depths should provide more opportunity for plants to utilize soluble nutrients moving into lower parts of the soil profile.

As noted above, there was considerable variation in soil NO₃–N levels in the upper 0.6 m of the soil profile near planting time. Preplant (or immediate postplant) sampling of the upper 0.6 m of the soil profile provided a reasonable compromise between minimizing sample number for ease of collection while maintaining costs at a level acceptable by growers and consultants. When all yield responses from all years and sites were regressed against residual NO₃–N in the upper 0.6 m of soil plus applied N (Fig. 1), it was evident, as expected, that many factors in addition to N impact cotton lint yields across sites and years. Analyses of the combined data suggests that under present production practices, about 1500 kg lint ha^–1 is a practically achievable goal and that about 200 to 220 kg available N ha^–1 is required for such a yield from combined soil NO₃–N and applied-N sources. The analysis further suggests that a quantity of N equal to the spring residual NO₃–N level could be deducted from the quantity of applied N; however, the measure is imprecise in that it only accounts for about one-third of the total variation in yields (r² = 0.328). Conversion of per-hectare yields to relative yields, by expressing each treatment's yield as a percentage of yield in the lowest N treatment at each site, did not significantly reduce the large degree of scatter in these data (data not shown). This analysis suggests that while residual soil NO₃–N measurements alone may be imprecise as an indicator of fertilizer N needs, they may have use as one of several criteria in evaluations of crop N needs and likely responses to fertilizer N.

View larger version (27K): [in this window] [in a new window]   Fig. 1. Lint yield as a function of residual soil NO3–N (top 0.6 m of soil profile) plus applied N for all sites and years from 1996 to 2000. Curve fit is second-order polynomial. r2 = 0.328.

Other Crop Management and Environmental Factors
The partitioning of nutrients and photosynthate to developing fruit has a large effect on the amount and duration of vegetative growth in cotton (Guinn, 1982; Gerik et al., 1994). Low temperatures during early vegetative growth or very high temperatures and/or water stress during flowering or fruiting can negatively impact yield potential through direct fruit losses. In addition, the duration of active growth, eventual size of plants, and the amount of N taken up and removed from the soil–plant system by harvesting can affect N availability and total crop N demand. In general, higher boll loads and higher seed cotton yields will require and remove more N from the soil system than will lower boll loads and yields (Bassett et al., 1970; Halevy, 1976; Guinn, 1982; Gerik et al., 1994).

The data set grouping of yield by residual NO₃–N (Table 12) was used with midbloom fruit retention data to test the hypothesis that fields most likely to respond positively to N applications would be those with lower residual soil NO₃–N and some indications of moderate-to-high yield potential. Fruit retention data during midbloom (12–17 d after first bloom) for the 56 and 168 kg N ha^–1 treatments for 36 of the 39 site-year combinations are shown in Table 13. Fruit retention on the bottom five fruiting branches and square (flower bud) retention on the top five fruiting branches are shown as an indicator of the degree to which early bolls were retained and midseason squares were retained rather than aborted. The midbloom period was evaluated since prior studies in the SJV (Weir et al., 1996) indicated that this is about 7 d before the time that growers should consider the last opportunity to apply water-run or foliar N applications.

	SUMMARY AND CONCLUSIONS

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

 
The results clearly show that under current management strategies, N applications in some California crop rotations that include cotton can be in considerable excess of that needed for a positive yield response. It is also clear that under less-than-optimum growing conditions, and in the presence of adequate-to-high residual soil N, cotton can be grown with little applied N. The problem is that neither of these situations can be predicted with complete confidence. Historically, the approach has been to apply relatively high N rates. Growers trying to maximize yields and financial returns during difficult economic times are reluctant to reduce relatively inexpensive fertilizer applications and risk yield losses due to N deficiencies. As a result, the average N application rate is 180 to 200 kg N ha^–1, with higher rates not uncommon under high yield potential conditions.

A meaningful reduction in N use in SJV cotton production systems will require that growers develop an integrative approach to estimating crop needs. Information on cropping history and associated N use, residual spring soil NO₃–N level, N supplied by irrigation water, early-season heat unit accumulation, and in-season measurements of crop N status and fruit retention, when taken together, should allow an improved assessment of crop N needs. Although agronomically it makes sense for farm managers to measure soil N status, adjust N application rates, and monitor in-season plant N status, economically the perception may be that the cost of management time and analytical services might not represent a savings compared with following the historical N fertilizer guidelines. However, increased concern for potential negative impacts of excess N applications on both the crop and the environment and real cost savings incurred by reducing N applications should make comprehensive management approaches more attractive.

	ACKNOWLEDGMENTS

 
The assistance and generosity of growers/cooperators in this research is gratefully acknowledged. Financial support for this project was provided by the Cotton Incorporated California State Support Committee of Cotton Incorporated in combination with national funds from the Cotton Incorporated Agricultural Research budget. Funding for related projects on cotton N nutrition that partially supported this project came from the California Department of Food and Agriculture, Fertilizer Research and Education program, the University of California Cooperative Extension, and California Crop Improvement Association.

	NOTES

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

 
This research was supported in part by cooperative research agreements with the California State Support Committee of Cotton Incorporated.

	REFERENCES

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

 

• Bassett, D.M., W.D. Anderson, and C.H.E. Werkhoven. 1970. Dry matter production and nutrient uptake in irrigated cotton (Gossypium hirsutum L.). Agron. J. 62:299–303.[Abstract/Free Full Text]
• Bassett, D.M., and A.J. MacKenzie. 1983. Plant analysis as a guide to cotton fertilization. p. 16–17. In H.M. Reisenauer (ed.) Soil and plant tissue testing in California. Bull. 1879. Univ. of California, Div. of Agric. and Nat. Resour., Oakland.
• Bi, J.L., G. Ballmer, D.L. Hendrix, T.J. Henneberry, and N.C. Toscano. 2001. Effect of cotton nitrogen fertilization on Bemisia argentifolii populations and honeydew production. Entomol. Exp. Appl. 99:25–36.
• Blake, G.R. 1965. Bulk density. p. 374–390. In C.A. Black, D.D. Evans, L.E. Ensminger, J.L. White, F.E. Clark, and R.C. Dinauer (ed.) Methods of soil analysis. Part 1. Physical and mineralogical properties. Agron. Monogr. 9. ASA, Madison, WI.
• Boquet, D.J., and G.A. Breitenbeck. 2000. Nitrogen rate effect on partitioning of nitrogen and dry matter by cotton. Crop Sci. 40:1685–1693.[Abstract/Free Full Text]
• Bremner, J.M., and C.S. Mulvaney. 1982. Nitrogen—total. p. 595–624. In A.L. Page, R.H. Miller, and D.R. Keeney (ed.) Methods of soil analysis. Part 2. Chemical and microbiological properties. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Burow, K.R., S.V. Stork, and N.M. Dubrovsky. 1998. Nitrate and pesticides in ground water in the eastern San Juaquin Valley, California: Occurrence and trends. Water Resour. Investigations Rep. 98-4040. USGS, Sacramento, CA.
• Cassman, K.G., D.C. Bryant, and B.A. Roberts. 1990. Comparison of soil test methods for predicting cotton response to soil and fertilizer potassium on potassium fixing soils. Commun. Soil Sci. Plant Anal. 21:1727–1743.
• Cisneros, J.J., and L.D. Godfrey. 2001. Midseason pest status of cotton aphid (Homoptera: Aphididae) in California cotton: Is nitrogen a key factor? Environ. Entomol. 30:501–510.
• Comly, H.H. 1987. Cyanosis in infants caused by nitrates in well water. JAMA 257:2788–2792.[Medline]
• Franco, J., and C.W. Cady. 1997. Preventing nitrate groundwater contamination in California: A nonregulatory approach. J. Prod. Agric. 10:52–57.
• Fritschi, F.B., B.A. Roberts, R.L. Travis, D.W. Rains, and R.B. Hutmacher. 2003. Response of irrigated Acala and Pima cotton to nitrogen fertilization: Growth, dry matter partitioning, and yield. Agron. J. 95:133–146.[Abstract/Free Full Text]
• Gardner, W.H. 1965. Water content. p. 82–127. In C.A. Black (ed.) Methods of soil analysis. Part 1. Physical and mineralogical properties. Agron. Monogr. 9. ASA, Madison, WI.
• Gaylor, M.J., G.A. Buchanan, F.R. Gilliland, and R.L. Davis. 1983. Interactions among a herbicide program, nitrogen fertilization, tarnished plant bugs, and planting dates for yield and maturity of cotton. Agron. J. 903–907.
• Gerik, T.J., B.S. Jackson, C.O. Stockle, and W.D. Rosenthal. 1994. Plant nitrogen status and boll load of cotton. Agron. J. 86:514–518.[Abstract/Free Full Text]
• Guinn, G. 1982. Causes of square and boll shedding in cotton. USDA Tech. Bull. 1672. USDA, Beltsville, MD.
• Halevy, J. 1976. Growth rate and nutrient uptake of two cotton cultivars grown under irrigation. Agron. J. 68:701–705.[Abstract/Free Full Text]
• Halevy, J., and O. Kramer. 1986. Nitrogen fertilizer management of cotton grown under drip irrigation in a grumusol. Irrig. Sci. 7:63–72.
• Hutchinson, R.L., G.A. Breitenback, and R.A. Brown, and W.J Thomas. 1995. Winter cover crop effects on nitrogen fertilization requirements of no-till and conventional tillage cotton. p. 73–76. In M.R. McClellend et al. (ed.) Conservation tillage systems for cotton: A review of research and demonstration results from across the cotton belt. Arkansas Agric. Exp. Stn. Fayetteville, AR.
• Hutmacher, R.B., M.P. Keeley, C.J. Phene, K.R. Davis, J.E. Ayars, M.S. Peters, S.S. Vail, J. Covarrubias, and A. Nevarez. 1999. Cotton root systems in a clay loam soil: Effects of growth stage, irrigation and nitrogen treatments. p. 637–640. In Proc. Beltwide Cotton Conf., Orlando, FL. 6–10 Jan. 1999. Natl. Cotton Counc., Memphis, TN.
• Hutmacher, R.B., C.J. Phene, K.R. Davis, T. Pflaum, M.S. Peters, and S.S. Vail. 1994. Acala and Pima cotton water relations and nutrient management under subsurface drip irrigation. p. 1355–1359. In Proc. Beltwide Cotton Conf., San Diego, CA. 5–8 Jan. 1994. Natl. Cotton Counc., Memphis, TN.
• Hutmacher, R.B., C.J. Phene, K.R. Davis, S.S. Vail, T. Pflaum, M.S. Peters, C.A. Hawk, T.A. Kerby, and M.P. Keeley. 1995. Nitrogen uptake of Acala and Pima cotton under high-yield, drip irrigated conditions. p. 1295–1300. In Proc. Beltwide Cotton Conf., San Antonio, TX. 4–7 Jan. 1995. Natl. Cotton Counc., Memphis, TN.
• Keeney, D.R., and D.W. Nelson. 1982. Nitrogen—inorganic forms. p. 643–698. In A.L. Page (ed.) Methods of soil analysis. Part 2. Chemical and microbiological properties. 2nd ed. Agron. Monogr. 9. ASA, Madison, WI.
• Kerby, T.A., B.L Weir, and M.P. Keeley. 1996. The uses of PIX. p. 294–304. In S.J. Hake et al. (ed.) Cotton production manual. Publ. 3352. Univ. of California Div. of Agric. and Nat. Resour., Oakland.
• Klute, A. 1965. Water capacity. p. 273–278. In C.A. Black (ed.) Methods of soil analysis. Part 1. Physical and mineralogical properties. Agron. Monogr. 9. ASA, Madison, WI.
• Kross, B.C., G.R. Hallberg, R. Bruner, K. Cherryholmes, and K.J. Johnson. 1993. The nitrate contamination of private well water in Iowa. Am. J. Public Health 83:270–272.[Abstract/Free Full Text]
• Lutrick, M.C., H.A. Peacock, and J.A. Cornell. 1986. Nitrate monitoring for cotton lint production on a Typic Paleudult. Agron. J. 78:1041–1046.[Abstract/Free Full Text]
• Marsh, B.H., R.B. Hutmacher, B.R. Roberts, R.L. Travis, and D.W. Rains. 2000. Why develop new nitrogen guidelines for California cotton? p. 1385. In Proc. Beltwide Cotton Conf., San Antonio, TX. 3–7 Jan. 2000. Natl. Cotton Counc., Memphis, TN.
• McConnell, J.S., W.H. Baker, D.M. Miller, B.S. Frizzell, and J.J. Varval. 1993. Nitrogen fertilization of cotton cultivars of differing maturity. Agron. J. 85:1151–1156.[Abstract/Free Full Text]
• McConnell, J.S., R.E. Glover, E.D. Vories, W.H. Baker, B.S. Frizzell, and F.M. Bourland. 1995. Nitrogen fertilization and plant development of cotton as determined by nodes above white flower. J. Plant Nutr. 18:1027–1036.
• Miller, R.O., B.L. Weir, R.N. Vargas, S.D. Wright, R.L. Travis, B.A. Roberts, D.W. Rains, D. Munk, D.J. Munier, and M. Keeley. 1997. Cotton potassium fertility guidelines for the San Joaquin Valley of California. Publ. 21562. Univ. of California Div. of Agric. and Nat. Resour., Oakland.
• Mullins, G.L., and C.H. Burmester. 1990. Dry matter, nitrogen, phosphorus and potassium accumulation by four cotton varieties. Agron. J. 82:729–736.[Abstract/Free Full Text]
• Olsen, S.R., and L.E. Sommers. 1982. Phosphorus. p. 403–430. In A.L. Page (ed.) Methods of soil analysis. Part 2. Chemical and microbiological properties. Agron. Monogr. 9. ASA, Madison, WI.
• Roberts, B.A., R.G. Curley, T.A. Kerby, S.D. Wright, and W.D. Mayfield. 1996. Defoliation, harvest and ginning. p. 305–323. In S.J. Hake, T.A. Kerby, and K.D. Hake (ed.) Cotton production manual. Publ. 3352. Univ. of California Div. of Agric. and Nat. Resour., Oakland.
• Thomas, G.W. 1982. Exchangeable cations. p. 159–165. In A.L. Page (ed.) Methods of soil analysis. Part 2. Chemical and microbiological properties. Agron. Monogr. 9. ASA, Madison, WI.
• Unruh, B.L., and J.L. Silvertooth. 1996. Comparisons between an Upland and a Pima cotton cultivar: II. Nutrient uptake and partitioning. Agron. J. 88:589–595.[Abstract/Free Full Text]
• USDA Economics and Statistics System. 2001. Economics and statistics systems database [Online]. Available at http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/#field (verified 3 Nov. 2003). Mann Library, Cornell Univ., Ithaca, NY.
• Weir, B.L., T.A. Kerby, K.D. Hake, B.A. Roberts, and L.J. Zelinski. 1996. Cotton fertility. p. 210–227. In S.J. Hake, T.A. Kerby, and K.D. Hake (ed.) Cotton production manual. Publ. 3352. Univ. of California Div. of Agric. and Nat. Resour., Oakland.
• Zelinski, L.J., and B.L. Weir. 1986. Soil nitrogen tests and yields for San Joaquin Valley cotton. p. 26–29. In California cotton progress report. Univ. of California, Div. of Agric. and Nat. Resour., Oakland.

This article has been cited by other articles: