| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Lousiana State Univ. Agric. Cent., P.O. Box 438, Saint Joseph, LA 71366
b Texas A&M Univ., Dep. of Soil and Crop Sciences, College Station, TX 77843-2474
c Lousiana State Univ., Dep. of Experimental Statistics, 161 Agric. Administration Building, Baton Rouge, LA 70803-5606
* Corresponding author (eclawson{at}agcenter.lsu.edu)
Received for publication February 1, 2005.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: CR, conventional-row • HVI, high volume instrument • UNR, ultra-narrow row
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Row spacings in CR cotton production (including "narrow row" systems) range from 76 to 102 cm. Although some research has demonstrated lint yield advantages for UNR when compared with CR cotton, the consistency of the response has varied among experiments. A reliable yield advantage for reduced row spacings was found by Hawkins and Peacock (1973), who compared two plant populations across row spacings ranging from 25.4 to 101.6 cm. Averaged over three tests, greater hand-picked seedcotton yield was found in the 25.4-cm than in the wider spacings. Inconsistent responses were reported by Jost and Cothren (2000, 2001), who found instances of greater lint yield for UNR than for CR cotton in 1 yr, but no differences among treatments in the other in each of two studies. Similarly, Rinehardt et al. (2004) found an instance of greater lint yield in UNR relative to two N rates in CR cotton, but only at one location. At the other, no lint yield differences were present between the systems regardless of N.
In several studies, equivalent lint yields between the systems were found in the majority of years or in data combined over years. Nichols et al. (2003) and Vories et al. (2001) each found yield advantages for UNR relative to CR cotton in 1 of 3 yr, but no differences in the others. When the data was averaged over all years, Vories et al. (2001) showed similar lint yields and no significant differences among row spacings. In a 3-yr study, Baker (1976) showed no significant lint yield differences among CR, UNR, and high population twin-row cotton. When compared under some conditions, UNR may have lower yield potential than CR cotton. Over 4 yr, in irrigated and nonirrigated tests, Boquet (2005) found that flat-planted UNR cotton was lower yielding than CR checks planted on raised beds. The variation in the results of previous research suggests that the yield potential of UNR cotton may be dependent on environmental conditions and cultural practices.
The yield potential of cotton is strongly influenced by N fertility, and experiments have been conducted to determine optimal N fertilizer rates for CR cotton in virtually all cotton-producing areas. The responses of UNR cotton to N fertilizer can be compared with those of CR cotton by the inclusion of both systems in N fertilizer experiments. Studies examining these production systems at varying N rates include those of Koli and Morrill (1976), Rinehardt et al. (2004), and Marois et al. (2004). Lint yield showed little or no response to applied N in the latter two studies. In the case of Koli and Morrill (1976), lint yields were generally decreased when N was applied. Boquet (2005) compared UNR cotton fertilized at a series of N rates with CR cotton checks that received the optimal N rate for the local area. Although lint yield did not respond to N in the UNR cotton, N rates below the optimal for CR cotton were not included. Further research is needed to fully establish N responses of lint yield in UNR relative to CR cotton.
In addition to lint yield, variations in N fertilizer rate may be reflected in plant architecture, which has important implications for UNR production. Large plants generally have high numbers of fruiting sites and may be prone to fruit abscission from sites low in the canopy. Because these effects can promote delays in boll set and crop maturity, reduced plant size is preferable in UNR cotton. Limited plant size may also help to optimize the harvest performance of finger-type strippers. This is important for reasons that include greater foreign matter and reduced lint turnout in finger-stripped UNR relative to CR cotton, as shown by Anthony et al. (1999). The additional foreign matter may result in fiber quality problems if it is deemed necessary to subject UNR cotton to extra steps in ginning. Under these circumstances, McAlister (2001) found that finger-stripped UNR cotton was lower in high volume instrument (HVI) length uniformity index and higher in AFIS neps than CR cotton.
The prevalence of finger-type stripper use for UNR cotton harvest has heightened the emphasis on fiber quality in this production system. However, equipment has recently been developed to allow 38-cm rows to be spindle-picked (Deere and Co., 2004). This harvest method may alter traditional gin turnout and fiber quality relationships between UNR and CR cotton, and future advances may do so as well. Although harvest equipment influences on fiber quality may interact with N fertilizer rate, they depend on the machinery used. Intrinsic effects of N on the quality of UNR cotton may also be important in the determination of optimal N fertilizer rate. Data from hand-picked cotton is valuable in identifying these effects without influence from harvest machinery type.
The primary objective of this study was to determine if UNR N requirements differ from those of CR cotton based on lint yield, fiber quality, and plant architecture. A secondary objective was to compare the yield potential of UNR and CR systems.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
Production practices applicable to the local area were utilized as much as possible. Mepiquat chloride (N,N-dimethylpiperidinium chloride) was applied uniformly across all treatments. Applications of 24.5 g a.i. ha–1 were made on 16 June 2000; 21 June and 4 July 2001; and 12 June and 12 July 2002. Prior to the final application in 2002, an additional mepiquat chloride application was washed off by rain. Insect pests were controlled as guided by scouting, with pest pressure considerably higher in 2001 than in other years. Boll weevil (Anthonomus grandis grandis Boheman) eradication, initiated in the area in 2001 (Texas Boll Weevil Eradication Foundation, 2005), became instrumental in maintaining low insect populations in 2002. The fields were irrigated by an overhead linear-move sprinkler system.
Climatic conditions are listed in Table 3. During 2000, high temperatures were experienced in the latter part of the growing season. In 2001, frequent rain occurred during harvest. The year 2002 was characterized by moderate temperatures and plentiful, well distributed rainfall.
|
Seedcotton was collected by repeated hand harvest of a 1.52 by 4.57 m area in each plot. For each year, seedcotton weight totaled over all harvests was multiplied by lint percentage to obtain lint yield. Lint percentage was obtained by mixing the seedcotton from all harvests in each plot in each year, pulling a sample from this composited harvest, and ginning it on a laboratory gin. Part of the resulting lint was sent to the International Textile Center in Lubbock, TX, for HVI quality analysis. Due to time constraints, the final harvest was not included in the ginout–fiber quality sample for 2002. However, averaged across replications, the included harvests represented at least 93% of total seedcotton yield for each treatment. Fiber quality data is not available for 2000.
For analysis of boll weight, bolls with <1 g of seedcotton were considered unharvestable and excluded, as were damaged bolls or bolls originating on damaged branches or peduncles. Bolls on plants having terminal damage were excluded; however, certain exceptions to this rule were made in 2001. In that year, when a branch had clearly taken over as the main stem following terminal damage, bolls were assigned a main stem node on that basis and included in the analysis. Bolls were also included in 2001 if found on branches below the site of terminal damage occurring high on the plant. Analysis was performed on a whole-plant basis and treatments were also compared within a subset comprised of first position bolls at Nodes 6 to 10.
Data analysis was performed in SAS using PROC MIXED at the 0.05 level of significance (SAS Institute, 2002). Analysis is also included for some variables at the 0.10 level of significance. Least squares means are reported, separated by the Tukey-Kramer method. Data was combined over the 3 yr of the study for all variables other than fiber quality parameters, which are combined over 2001 and 2002. Years were considered a random effect. Significance of effects of N rate, row spacing, and N rate x row spacing interaction on each reported parameter are given in Table 4.
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Excessive plant height was not an issue regardless of row spacing or N. For the highest N rate, plant heights were 58.1, 62.3, and 71.1 cm in the 19-, 38-, and 76-cm row spacings, respectively. These heights do not suggest plant sizes great enough to cause problems with finger-type stripper harvest. However, there may be a need to avoid excessive N fertilization in UNR cotton under conditions that are conducive to rank growth.
Because each fruiting or vegetative branch arises at a main stem node, the increase in main stem nodes plant–1 with greater N suggests a plant structure capable of supporting a larger boll load. Although this increase was small in magnitude, such changes may reduce the likelihood of early crop maturity in UNR cotton. The reduction in nodes plant–1 associated with the UNR spacings may be important in maintaining appropriate plant architecture when elevated N rates are applied in a growth promoting environment.
Individual Boll Weight
Row spacing and N treatments have potential to affect boll distribution among fruiting sites. This may influence individual boll weight when averaged over the whole plant, since boll weight has been found to vary by fruiting branch position and by main stem node (Jenkins et al., 1990; Bednarz et al., 2000). Weight of first position bolls at Nodes 6 through 10 (subset individual boll weight) is reported to show treatment effects on individual boll weight with minimal influence of changes in boll distribution. Whole-plant individual boll weight is discussed to provide a more general insight into yield results.
Subset individual boll weight was decreased by reduction in row spacing (Table 6). Relative to CR, earlier canopy closure (Jost and Cothren, 2000) and faster development of leaf area index (Jost and Cothren, 2001) have been documented for UNR cotton. These factors have potential to reduce light penetration into the vicinity of lower first position bolls, and may have contributed to reductions in subset individual boll weight. Earlier canopy closure was visually apparent with reduced row spacing in our study. A reduction in the ability of the entire plant to provide photosynthate may have also been important in subset individual boll weight decreases, because plants were smaller under the increased competition of UNR spacings.
|
= 0.10, the 19- and 76-cm spacings differed significantly (Tables 4 and 6). Whole-plant individual boll weight in 19 cm was 95% of that in 76-cm rows. When expressed by us in a similar manner, previously reported data also show limited individual boll weight reductions for UNR relative to CR cotton. Averaged over planting date and variety, Galanopoulou-Sendouka et al. (1980) report significant boll weight differences in 1 of 2 study years. For that year, the 25 cm was 92% of the 100-cm row spacing mean. Using 25-boll samples, Baker (1976) found that boll weight reductions from high population-reduced row spacing treatments were significant. However, these reduced means were within 90% of a CR check. In contrast, instances of larger boll weight differences among equidistant plant spacings are evident in the data reported by Fowler and Ray (1977). Subset individual boll weight was significantly increased by N, which may indicate greater ability of plants to provide photosynthate to these bolls (Table 6). In CR cotton, Boquet et al. (1994) found that weight of first position bolls on main stem Nodes 5 through 10 was optimized at 28 to 56 kg N ha–1, and was decreased by N rates above 56 kg N ha–1. They suggested that changes in canopy environment with increasing N may have reduced localized carbohydrate availability and thereby decreased individual boll weight in this subgroup. In our study, subset individual boll weight generally increased through the highest rate of N. Although greater plant height may have increased shading near subset bolls in high N treatments, N appears to have been more limiting than localized boll environment under our study conditions.
Whole-plant individual boll weight also increased with N (Table 6). Although the range in means was similar to that of subset individual boll weight, significant differences did not exist among the highest three N rates. As was the case for row spacing effects, a relatively narrow range of boll weights was found among N treatments, with the 0 equal to 88% of the 151 kg N ha–1 mean. In a greenhouse study, Jackson and Gerik (1990) found that boll weight increased significantly as biweekly N application increased from 0 to 144 mmoles. For their reported data, the 0 N means were 73 and 74% respectively, of the 144 mmole N means in the 2 study years.
Effects of row spacings and N were more pronounced on subset than on whole-plant individual boll weight. This may be due to treatment effects on boll distribution or nonsubset individual boll weight, both of which are included in the whole-plant means. Whole-plant individual boll weight was changed to a greater degree by N rates than by row spacings. This suggests that increases in N may have potential in maintaining boll weight in UNR cotton. Although small, the observed row spacing and N effects on whole-plant individual boll weight have potential to affect lint yield. However, the final effect on lint yield is also dependent on responses of other yield components, including boll numbers and lint percentage.
Lint Percentage
Lint percentage was significantly increased with each reduction in row spacing (Table 7). When independent of harvesting equipment effects, UNR lint percentage has often been similar to or higher than that of CR cotton. Jost and Cothren (2000) found no significant differences in lint percentage between hand-picked UNR and CR cotton. Likewise, Baker (1976) found no significant differences in lint percentage from boll samples among high population-reduced row spacing systems and CR cotton. Our results show increases in lint percentage for UNR cotton. This agrees with Galanopoulou-Sendouka et al. (1980), who under hand-picked conditions found UNR to be significantly higher in lint percentage than CR cotton in 1 of 2 yr. Jost and Cothren (2001) found that a high population 19-cm spacing and a low and a medium population 38-cm spacing each had significantly higher lint percentage than 76- and 101-cm CR cotton. In that study, the increased lint percentage was attributed to decreased seed size in the reduced row spacings. Data on seed size was not obtained in our study, but if such an effect is typical for UNR systems it may provide an explanation for the observed lint percentage increases.
|
Although UNR increased lint percentage, the magnitude of the differences was limited. Nevertheless, these results confirm that UNR is not intrinsically lower in lint percentage than CR cotton. If advances in technology improve UNR harvesting efficiency, producers may realize comparable gin turnouts for the two systems. Lint percentage did not change in response to N rate, and for this reason was not influential in the N responses of lint yield.
Fiber Quality
Micronaire was significantly higher in 38- than in 19-cm spacing, but neither differed from the 76-cm rows (Table 8). The maximum difference between means was <0.2 micronaire units. The lack of significant differences between UNR and CR spacings is in accord with previous hand-picked studies, which report that micronaire differences among these production systems were uncommon or inconsistent. Galanopoulou-Sendouka et al. (1980) found an instance of slightly higher micronaire in UNR than in CR cotton, but only in 1 of 2 yr. In studies involving UNR and CR cotton, Jost and Cothren (2000, 2001) found no differences in micronaire among treatments. Baker (1976) found that two UNR and three high-population twin-row treatments did not differ in micronaire from a 91.4-cm CR check, although the micronaire of a remaining UNR treatment was reduced.
|
Nitrogen effects on fiber quality were minimal. A small but significant reduction in micronaire was observed with higher N (Table 8). On average, individual bolls in these treatments may have been subject to a slight reduction in carbohydrate availability. Potential reasons for this include shading of boll-supplying leaves and high numbers of bolls plant–1 (Hake et al., 1996), both of which were likely in the high N treatments. High yield, which may be related to greater numbers of bolls plant–1, is also associated with reduced levels of micronaire (Hake et al., 1996). The small difference in micronaire was probably a consequence of low yields and reduced canopy development under the N-deficient low N treatments. It does not represent an expected response under all circumstances. For example, under conditions in which yields did not respond to increased N, Boquet (2005) found no effect of N on micronaire in UNR cotton.
Fiber length, length uniformity, and strength did not vary with N rate. Similar results were found by Boquet (2005) in both irrigated and nonirrigated UNR cotton. In that study, under irrigation, N effects on these parameters were limited to minimal differences in fiber length. Without irrigation, fiber quality was not affected by N. In our study, an N by row spacing interaction on fiber length was present at the 0.10 level of significance (Table 4). The range in fiber length means of individual N and row spacing combinations was <0.006 cm, suggesting that the interaction was of limited importance.
Fiber quality was similar across row spacings and N rates. The most notable difference was a minor reduction of micronaire with increased N. However, the interaction of N and row spacing on micronaire was not significant (Table 4), suggesting a similar response to N for each row spacing. Over the range of N rates tested, there appears to be little variation in fiber quality of hand-picked UNR and CR cotton.
Lint Yield
Lint yields were very similar for the three row spacings, with no differences significant (Fig. 1
). Yield component data provide a measure of insight into this result. Whole-plant individual boll weight was decreased slightly by UNR treatments, but increases in lint percentage in the reduced row spacings provided some compensation for the boll weight reduction. Unmeasured factors such as bolls plant–1 and bolls ha–1 may also have varied among row spacings to prevent yield differences.
|
Flat planting was utilized for all row spacings. This eliminated the advantages that raised beds, which are typical in CR cotton, may offer in terms of stand establishment and drainage. The use of a bedded system may have prevented the initial stand failures that occurred in 2000 and 2001. However, once the stand was established, the lack of raised beds did not appear to decrease yields. In some environments, the use of raised beds may be more important. After finding greater yields for CR cotton on raised beds than for flat-planted UNR cotton, Boquet (2005) suggested that a bedded system for UNR cotton might be beneficial in increasing its yield potential in his area.
The multiple hand harvests ensured that virtually all seedcotton was removed from the plants. This method also limited the time that bolls remained open before harvest. This reduced the potential for seedcotton to fall to the ground under inclement weather, although losses of this type were observed following the lengthy rains that delayed the second harvest in 2001. Under these harvest methods, the data represent yield potential of the row spacings, and do not account for differences in harvest machinery between UNR and CR cotton.
Careful consideration of the effects of harvest methods, planting practices, irrigation availability, and other sources of variability is important in forming expectations about yields of UNR relative to CR cotton. Under the conditions of our study, lint yield potential was unaffected by row spacing.
Lint yield was increased by N fertilizer, with differences decreasing in magnitude with each additional increment of N (Fig. 2 ). Lint percentage was unaffected by N fertilizer rate, suggesting that increased whole-plant individual boll weight under N fertilization contributed directly to increased yields.
|
In assessing N fertilizer requirements of UNR cotton, a primary factor for consideration is yield. Interaction between N and row spacing on lint yield was virtually absent (Table 4). This indicates a comparable response of UNR and CR cotton across the N rates tested. The N requirements of UNR cotton do not appear to be lower than those of CR cotton. The similar response of lint yield to N suggests the possibility of equivalent N requirements for the two systems. This appears likely in light of the results of Boquet (2005). Under both irrigated and nonirrigated conditions, UNR cotton lint yield did not respond to a series of N rates at or above the current recommended rate for CR cotton, leading to the suggestion that UNR did not require N rates above the CR cotton optimum. However, in our study, it cannot be certain that the optimal N rate was reached, and further research may be needed to compare UNR and CR cotton responses to N rates above those that are optimal for CR systems.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Variation in the response of UNR and CR cotton to N may have implications for UNR N fertilizer recommendations. However, interaction between N and row spacing was essentially absent for all parameters reported, showing a comparable response to the applied N rates for both systems. Plant height and main stem nodes plant–1 were greater under increased N. However, study conditions were not conducive to excessive growth, and the need to consider these parameters in N fertilization practices for UNR cotton was minimal. Fiber quality, measured on hand-picked cotton, was largely unaffected by N and row spacing treatments and does not appear to be important in UNR N fertilizer recommendations. Interactions of machine harvest method with N rate on the fiber quality of UNR cotton were not addressed and may be an important topic of future research.
Lint yield was increased with each additional increment of N fertilizer, although increases were smaller in magnitude for higher N rates. The high yield levels and diminishing lint yield response to N suggest that the greatest N rate in this study was approaching optimal. Lack of interaction was especially pronounced on lint yield. Therefore, over the range of N rates applied, each row spacing responded similarly to N. It appears that N fertilizer recommendations for CR should not be reduced for UNR cotton. Further investigation may be needed to verify that differential response among UNR and CR cotton does not occur at elevated N rates.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. P. Viator, C. O. Gwathmey, J. T. Cothren, J. T. Reed, E. D. Vories, R. C. Nuti, K. L. Edmisten, and R. Wells Influence of Ultranarrow Row and Conventional Row Cotton on the Last Effective Boll Population Agron. J., August 11, 2008; 100(5): 1327 - 1331. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. L. Clawson, J. T. Cothren, D. C. Blouin, and J. L. Satterwhite Timing of Maturity in Ultra-Narrow and Conventional Row Cotton as Affected by Nitrogen Fertilizer Rate Agron. J., February 29, 2008; 100(2): 421 - 431. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
