Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published in Agron J 100:672-680 (2008)
DOI: 10.2134/agronj2007.0240
© 2008 American Society of Agronomy
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 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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gwathmey, C. O.
Right arrow Articles by Larson, J. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Gwathmey, C. O.
Right arrow Articles by Larson, J. A.
Agricola
Right arrow Articles by Gwathmey, C. O.
Right arrow Articles by Larson, J. A.
Related Collections
Right arrow Weed Management
Right arrow Other Crop Management
Right arrow Other Cropping Systems
Right arrow Cotton
Right arrow Crop Systems

COTTON

Solid and Skip-Row Spacings for Irrigated and Nonirrigated Upland Cotton

C. Owen Gwathmeya,*, Lawrence E. Steckela and James A. Larsonb

a Dep. of Plant Sci., Univ. of Tennessee, 605 Airways Blvd., Jackson, TN 38301
b Dep. of Agric. Econ., Univ. of Tennessee, 2621 Morgan Circle Dr., Knoxville, TN 37996

* Corresponding author (cogwathmey{at}utk.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Producers of upland cotton (Gossypium hirsutum L.) are interested in alternative row spacings and planting patterns to improve productivity. We conducted factorial experiments for 3 yr in adjacent irrigated and nonirrigated fields at Milan, TN, of cotton grown in 25-, 76- and 102-cm rows, each planted in a solid and 2 x 1 skip-row pattern. Narrower rows and solid plantings tended to close canopy earlier and more completely, to suppress weed growth, and to mature earlier than in wider rows and skip-row patterns. Incomplete canopy closure in wider skip-row spacings resulted in less weed suppression, but weed growth in skipped rows diminished with narrower row spacing. Skip-row cotton matured later than solid planting, but this effect also diminished with narrower rows. Without irrigation, row spacing and configuration for maximum lint yield depended on growing conditions of the particular year. Under irrigation, however, solid planted 76-cm rows consistently yielded more than either 25- or 102-cm solid or skip-row cotton. With or without irrigation, lint yields were 13 to 15% lower in 102-cm skip-row cotton than in 102-cm solid planting, but yields did not differ between solid and skip-row in narrower rows. Across years and treatments, plant density was the most influential yield component in nonirrigated cotton, but plant density and bolls plant–1 had similar influence under irrigation. Cotton producers interested in skip-row planting should consider rows spaced 76-cm or less to minimize weed problems and yield loss.

Abbreviations: BG, Bollgard • DAP, days after planting • DP, Deltapine • GLM, General Linear Model • PAR, photosynthetically active radiation • RR, Roundup Ready • SAS, Statistical Analysis System • UNR, ultra-narrow row


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

Received for publication July 9, 2007.
   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
AS COSTS OF SEED and other planting inputs have increased in recent years, producers of upland cotton have re-evaluated various row spacings and planting patterns to improve productivity. One alternative to solid planting is a skip-row pattern, in which entire rows are not planted adjacent to planted rows (Bange et al., 2005). A full 2 x 1 pattern refers to two planted rows adjacent to a skipped row of the same width (Parvin et al., 1999). Skipping rows reduces the seed planted per hectare, so planting costs are reduced. Another system involves planting in ultra-narrow rows (UNR ≤ 38 cm wide), with no post-plant cultivation, and harvesting with a finger-stripper harvester (Vories et al., 2001). Although more seed are typically planted in UNR, higher planting costs may be offset by lower equipment ownership and other cost savings relative to conventional wide row cotton (Parvin et al., 2000).

The growth habit of cotton allows the plant to partially compensate for potential yield lost to skip-row planting, such that yield reduction is less than proportional to the reduction in planted area. In Alabama, for instance, cotton yields increased in skip-row patterns by 29 to 62% on planted area basis, but decreased 12 to 46% on field-area basis, depending on number and frequency of skipped rows (King et al., 1986). The least yield loss relative to solid cotton occurred in the 2 x 1 skip-row pattern spaced at 102 cm, which yielded 88% of solid planting. Several Mississippi Delta studies, summarized by Parvin et al. (1999), showed that cotton in a full 2 x 1 skip-row pattern yielded from 68 to 92% of solid planting. Across locations in Australia, Bange et al. (2005) found that yield differences between solid and 2 x 1 skip-row cotton appeared as yields increased > 1.6 bales ha–1 (363 kg lint ha–1). However, none of these studies indicated which yield components accounted for the observed compensation. King et al. (1986) found that skip-row cotton generally matured later than solid planting. In a brief report on part of the present study, Gwathmey and Steckel (2006) also found that yield compensation was accompanied by delayed maturity, which can lead to losses of yield and fiber quality in short-season environments.

Yield compensation in dryland skip-row cotton may be partly due to soil water conservation. In a Texas dryland environment, Hons and McMichael (1986) showed that soil water stored beneath adjacent skipped rows tends to distribute the seasonal water supply more favorably for crop development. They found that yield response to 1-m bedded skip-rows differed by year. With relatively normal rainfall, yields per field area were similar in solid and 2 x 1 skip, but lower in 2 x 2 skip-rows. In a relatively dry year, however, yields (on land-area basis) were lower, but similar in all three planting patterns. Seasonal water extraction was similar for the planted rows of all patterns and for the fallowed row of the 2 x 1 pattern, indicating that cotton could extract as much water from the skipped row as from planted rows. This attribute of skip-row production is a major reason for its adoption in dryland regions (Bange et al., 2005), and it could also be valuable for nonirrigated cotton grown in sub-humid regions, especially in dry years.

Most skip-row research has been conducted in row spacings ≥ 76 cm, but few published studies have evaluated possible interactions between row spacing and planting pattern in cotton. Jones (1998) reported that plants in 76-cm rows produced more lint yield on a land-area basis than those in 102-cm rows in Mississippi Delta soils. Plant size was larger in the full 2 x 1 skip-row planting pattern compared with solid planting, but solid planting produced higher lint yield on a land-area basis. However, Jones (1998) did not indicate whether responses to row spacing and pattern were independent or interactive. Buehring et al. (2006) studied several row spacings and patterns in northern Mississippi. Their report did not specifically address interactions between row spacing and pattern, but inferences can be drawn from their data. In all six location-yr of yield data that they reported, similar yields were obtained from 38-cm rows planted in a solid and 2 x 1 skip-row pattern. In 76- and 97-cm rows, however, they obtained similar yields from solid and 2 x 1 skip-rows in three cases, but they obtained higher yields from solid plantings in three other cases. Their results suggest the possibility that skip-row effects on yield may not be equivalent at all row spacings.

Ultra-narrow row cotton is typically planted in solid rows spaced < 38 cm, with plant populations often in excess of 24.7 plants m–2 (Jones, 2001; Delaney et al., 2002). High seeding rates were recommended in the 1990s mainly to facilitate harvesting with a finger-type stripper (Delaney et al., 2002). Therefore, many comparisons of UNR and wide-row cotton involved fairly large differences in plant density and inconsistent effects on yield. In Texas, Jost and Cothren (2000) compared four row spacings in an irrigated bottomland soil. Hand-harvest yields were not affected by row spacing in a relatively wet year; but in a drier year, cotton in the 19- and 38-cm rows averaged 29% higher yields than in 76- and 102-cm rows. They attributed the yield response to higher boll numbers m–2 and to avoidance of late-season drought stress by earlier boll set in UNR. In Louisiana, Boquet (2005) compared hand-harvest yields from 102- and 25-cm rows under irrigated and nonirrigated conditions. He obtained 24 and 55% higher yields in wide rows than in UNR under irrigated and nonirrigated conditions, respectively. In both studies, yield differences were attributed mainly to more bolls m–2 and larger bolls in wide rows. He also attributed part of the yield difference to planting wide rows on beds to avoid the ponding of water that occurred in the flat seedbeds used for UNR cotton, and to drought stress in the nonirrigated study. In Arkansas, Vories et al. (2001) compared nonirrigated cotton yields in finger-stripped 19-cm rows, with those in spindle-picked 97-cm rows. They obtained higher lint yields in UNR in only 1 of 3 yr. Year-to-year variation in yield response may have been associated with differences in plant density. Drilling of 52 seeds m–2 in UNR produced from 20 to 37 plants m–2, while planting 17 seeds m–2 in wide rows produced from 6.8 to 12.1 plants m–2 in different years. The lowest lint yield from UNR, relative to wide rows, was obtained in the year with the highest plant density and the lowest rainfall. Considered together, these studies suggest that the capacity of UNR cotton to produce high yields at a high plant density may depend considerably on adequate moisture supply for boll development.

One reason cited for planting in UNR is to produce an earlier maturing crop (e.g., Jones, 2001), but achievement of the earliness potential of UNR has been inconsistent. In Texas, Jost and Cothren (2001) evaluated earliness of maturity by successive hand harvests in four row spacings, with plant density treatments varying within the UNR spacings. They found maturity differences on a heavy clay soil, where UNR cotton at ≥ 19.8 plants m–2 produced a greater proportion of total yield than either UNR or wide-row cotton at ≤ 13 plants m–2 by 112 d after planting (DAP). By 116 DAP, at least 70% of the yield had been produced in 19-cm rows at ≥ 19.8 plants m–2, while less than 50% had been produced in 76- or 101-cm rows. On a lighter silty clay loam soil, however, yields and earliness were unaffected by row-spacing and density treatments. In Australia, Roche et al. (2003) compared earliness and yield of irrigated cotton in 25-cm rows (at 36 plants m–2) with conventional 1-m row cotton (at 12 plants m–2). They found no significant differences between UNR and conventionally spaced treatments in time needed to reach 60% maturity or in lint yield, despite earlier leaf area development and more light interception in UNR early in the season. They found a higher light extinction coefficient in UNR, suggesting that less light reached leaves lower in the canopy to support boll development at earlier fruiting sites. These findings are consistent with Heitholt (1994), who found that efficiency of photosynthetically active radiation (PAR) interception per unit leaf area was greater at lower plant population density. Heitholt (1995) studied flowering and boll retention in three row spacings over a range of plant densities. While he did not directly measure earliness of maturity, he reported that cotton in 51-cm rows produced 12 to 21% more flowers m–2 d–1 at early to mid-bloom than in 102-cm rows, but similar boll retention percentages regardless of plant density. He attributed the modest yield advantage of 51-cm rows to flowering rate. Considered together, these studies suggest that earliness and yield potential in UNR is associated with early boll set, which in turn depends on adequate light penetration to lower leaves in the canopy.

Earlier canopy closure and PAR interception in UNR cotton is generally considered beneficial for weed suppression (Molin et al., 2004). Several studies of weed competition in soybean (Glycine max L.) found that planting in narrow rows can result in greater season-long weed control than wide-row soybean, due to a faster canopy closure providing greater shading and weed suppression. Légère and Schreiber (1989) reported that reducing soybean row spacing from 76 to 25 cm reduced redroot pigweed (Amaranthus retroflexus L.) biomass by 20% and increased soybean yield by 18%. Wax and Pendleton (1968) reported that soybean yields increased by 10 and 20% as soybean row width was reduced from 102 cm to 76 and 25 cm, respectively. Steckel and Sprague (2004) reported that as soybean row spacing decreased from 76 to 38 cm, more PAR was intercepted by the soybeans, resulting in less biomass production by the pigweed species, common waterhemp [Amaranthus rudis (Moq) J.D. Sauer]. Steckel et al. (2003) showed that as PAR was reduced, pigweed mortality increased and pigweed biomass production decreased. Insolation of the soil surface may also affect seed germination of Amaranthus spp. Gallagher and Cardina (1998) reported that Amaranthus spp. needed as little as 3 µmol m–2 of red light in buried seed and 1000 µmol m–2 in unburied seed to initiate germination, depending on temperature. Thus, crop PAR interception may affect both the population density and growth of competing weeds. In skip-row cotton, however, reduced light interception by the crop in unplanted rows can lead to interference by large weeds in mid to late season. As weeds in the skipped rows grow, they increasingly utilize the resources required by cotton as the season progresses (Charles, 2002). Control of these weeds is important, but little research has focused on this weed population.

Objectives of this study were to evaluate (i) light interception relative to weed suppression, (ii) earliness of maturity, and (iii) yield and yield components of cotton grown in solid and 2 x 1 skip-row patterns at several row spacings under irrigated and nonirrigated conditions.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
A study of row spacing and configuration was conducted from 2003 through 2005 on adjacent irrigated and nonirrigated fields at the Research and Education Center at Milan, TN. Soil type was predominantly Grenada silt loam (fine-silty, mixed, active, thermic Oxyaquic Fraglossudalfs) in both fields. Before planting in 2003, flat seedbeds were prepared by conventional disc and harrow tillage. In subsequent years, no-tillage practices were used in both fields. Before planting each year, fertilizer was uniformly broadcast to both fields as recommended by Agricultural Extension Service (2001) for cotton production. In 2003, applied rates of N, P, and K were 102, 15, and 56 kg ha–1, respectively. In 2004 and 2005, N–P–K rates were 115–39–74 and 96–20–74 kg ha–1, respectively.

In each field, a factorial experiment was conducted with row spacing treatments of 102, 76, and 25 cm, each planted in a solid and 2 x 1 skip-row configuration. The 2 x 1 configuration refers to one unplanted row adjacent to two planted rows of the same width. All combinations of row spacing and configuration were replicated four times in the nonirrigated field and five times in the irrigated field in randomized complete blocks. Row-spacing treatments were established by planting 13.1 seed m–1 row in 102- and 76-inch rows with 4-row John Deere 7100 no-till planters, and 8.2 seed m–1 row in 25-cm rows with a 7-row Kinze no-till tandem planter. Tractors used with the 76- and 102-cm planters straddled two rows. The Kinze planter was drawn by a tractor with a 152-cm track width. In 76- and 102-cm row plots, solid planting was obtained by engaging all planter units, while skip-row treatments were established by disengaging the third planter unit in designated plots. In 25-cm rows, all plots were solid planted, and skip-row treatments were established by removing plants from every third row of designated plots by hand before emergence of the fifth true leaf. Individual plot lengths were 9.14 and 12.2 m in the irrigated and nonirrigated studies, respectively. Widths of 25-, 76-, and 102-cm row plots were 5.33, 6.10, and 8.13 m, respectively, in both studies.

Commercial seed lots of ‘Paymaster 1218 BG/RR’ were planted on 15 May 2003 and 10 May 2004, and ‘DP 444 BG/RR’ was planted on 10 May 2005. Each seed lot had been treated by the company with commercial fungicides (baytan: beta-(4-chloropenoxy)-alpha-(1,1,-dimethylethyl)-1H-1,2,4,-triazole-1-ethanol; thiram: (tetramethylthiuram disulfide); and either mefenoxam: (R,S)-2-[(2,6-dimethylphenyl)-methoxyacetylamino]-propionic acid methyl ester; or metalaxyl: N-(2,6-dimethylphenyl)-N-(methoxyacetyl)alanine methyl ester); and insecticides (chlorpyrifos: O,O-diethyl O-(3,5,6-trichloro-2-pyridyl)phosphorothioate; and thiamethoxam: 4H-1,3,5,-oxadiazin-4-imine, 3-[(2-chloro-5-thiazolyl)methyl]tetrahydro-5-methyl-N-nitro). Commercial fungicides were also applied in furrow at planting: 0.9 kg ha–1 PCNB (pentachloronitrobenzene) and 0.22 kg ha–1 5-ethoxy-3-(trichloromethyl)-1,3,4,-thiadiazole. A granular formulation was used in the JD7100 planters and an emulsifiable concentrate of the same fungicide was applied with the Kinze planter. Plant stands were counted in all plots between 31 and 36 DAP each year.

Supplemental irrigation was applied to the irrigated plots by overhead lateral-move sprinkler boom. Irrigation timing was determined using the University of Arkansas Irrigation Scheduler (Cahoon et al., 1990). A total of 120 mm water was applied in 11 irrigations between 55 and 127 DAP in 2003, 102 mm in 8 irrigations between 56 and 95 DAP in 2004, and 112 mm in 11 irrigations between 14 and 95 DAP in 2005. Daily rainfall and temperature data were collected at a standard weather station located 3 km from the plots. Cumulative rainfall, irrigation, and heat units from planting to harvest-aid application and harvest are summarized in Fig. 1 .


Figure 1
View larger version (42K):
[in this window]
[in a new window]

  		Fig. 1. Amounts of irrigation, rainfall, and heat-unit accumulation from planting to defoliation and harvest in each year of the cotton row spacing and configuration study, Milan, TN.

 
Between 30 and 50 d before planting each year, plot areas were sprayed with labeled rates of dicamba (diglycolamine salt of 3,6-dichloro-o-anisic acid) and either glyphosate (N-(phosphonomethyl)glycine) or paraquat (1,1'-dimethyl-4,4'-bipyridinium dichloride) in aqueous solution. No pre-emergence residual herbicide was applied in this study. In 2003, 0.96 kg ha–1 glyphosate in form of its isopropylamine salt was uniformly applied over the plot area with a high-clearance sprayer at 20 and 33 DAP. In 2004, 1.06 kg ha–1 glyphosate in form of its isopropylamine salt was applied at 22 DAP, and in 2005, the same herbicide was similarly applied at 17 and 31 DAP. For further control of graminaceous weeds, clethodim ((E)-2-[1-[(3-chloro-2-propenyl)oxy]imino]propyl]-5-[2-(ethylthio)propyl]3-hydroxy-2-cyclohexen-1-one) was applied with a high-clearance sprayer at rates of 140 g ha–1 at 49 DAP in 2003, 105 g ha–1 at 47 DAP in 2004, and 210 g ha–1 at 59 DAP in 2005. For further control of broadleaf weeds, 71 g ha–1 pyrithiobac (sodium 2-chloro-6-[(4,6-dimethoxy pryimidin-2-yl)thio]benzoate) was applied at 38 DAP in 2004, and 5 g ha–1 trifloxysulfuron (2-pyridinesulfonamide, N-[[4,6-dimethoxy-2-primidinyl)amino]carbonyl]-3-(2,2,2,-trifluoro-ethoxy)-piridin-2-sulfonamide sodium salt) was applied at 58 DAP in 2005.

Natural populations of Amaranthus spp. that emerged after the fourth true leaf of cotton served as a general indicator of the propensity for broadleaf weeds to germinate and grow under these treatments. All plots were rated at 92 DAP in 2004 and at 84 DAP in 2005 for suppression of pigweed (including Amaranthus retroflexus, A. hybridus, and A. palmeri). The rating scale ranged from 0 (no suppression) to 100 (complete suppression). Rating data were subject to arcsine transformation for statistical analysis. This study was not designed to measure the interference effects of weeds on cotton. Weeds that emerge with the crop are typically the most competitive, but in this study they were removed at 22 or 31 DAP with glypyhosate. The observed pigweed had emerged at 4 to 5 wk and were 20 to 30 cm in height at the time of rating. They were removed with an application of pyrithiobac after rating, so they did not become large enough to interfere with cotton yields. Researchers have reported that pigweed that emerges 4 to 5 wk after crop emergence (R1 indeterminate soybean) caused yield loss similar to a weed-free check (Steckel and Sprague, 2004; Bensch et al., 2003).

Cotton growth was regulated by multiple applications of mepiquat chloride (N, N-dimethylpiperidinium chloride) to all plots, following University of Tennessee Agricultural Extension Service (2001) guidelines. A total of 147 g ha–1 was applied between 47 and 92 DAP in 2003, 37 g ha–1 between 32 and 47 DAP in 2004, and 104 g ha–1 between 49 and 77 DAP in 2005. Insect pests were controlled by methods recommended by the Agricultural Extension Service (2001) for Bacillus thuringiensis cotton. All spray applications after planting were made with a self-propelled, high-clearance sprayer with a 2.04-m track width, operating in 102-cm rows but covering all plots with a 15.24-m boom.

Light interception by the cotton canopy was measured in each plot of the irrigated and nonirrigated tests with a pair of quantum PAR sensors. These data were collected within 2h of solar noon at 48 and 88 DAP in 2003, 43, and 88 DAP in 2004, and 41 and 83 DAP in 2005. A point sensor (model LI-190, LI-Cor, Lincoln, NE) was situated above the canopy of each plot, while a LI-191 line sensor was placed under the canopy. The midpoint of the 1-m line sensor was placed in line with the row or skipped row and the sensor was oriented such that its ends were located in the row middles. Each plot was subsampled three times. In skip-row plots, one of three subsamples was taken in a skipped row. Pairs of simultaneous photon flux data from both sensors were recorded by a LI-1000 data logger.

All open bolls in a flagged 1-m row segment of each plot were hand harvested at 125, 132, 139, and 160 DAP in 2003, and at 122, 129, 136, and 150 DAP in 2004. Numbers of bolls and plants in each row segment were also recorded. Seedcotton from each hand harvest was air dried and weighed to calculate boll weight and number per plant. Earliness of maturity was calculated as the percent of total yield produced by each of four harvests. Sequential hand harvests were not collected in 2005 because of lodging of plants caused by strong winds. However, all open bolls in a 1-m row segment of each plot were hand harvested at 171 DAP in 2005. Weights and numbers of bolls and plants in each row segment were recorded as in previous years.

A mixture of 315 g ha–1 tribufos (S,S,S,-tributyl phosphorotrithioate), 56 g ha–1 thidiazuron (N-phenyl-N'-1,2,3,-thidiazol-5-ylurea), 28 g ha–1 diuron (3-(3,4-dichlorophenyl)-1,1-dimethylurea), and 1.68 kg ha–1 ethephon ((2-chloroethyl)phosphonic acid) was applied to all plots each year. These harvest aids were uniformly applied at 141 DAP in 2003, 140 DAP in 2004, and 137 DAP in 2005. In 2003, 6.7 kg ha–1 sodium chlorate was also applied to desiccate plants in 25-cm rows only, at 155 DAP. In 2004 and 2005, 420 g ha–1 paraquat dichloride was similarly applied to desiccate plants in 25-cm rows only, at 149 and 153 DAP, respectively.

Designated interior rows of all plots were mechanically harvested at 162 DAP in 2003, 180 DAP in 2004, and 160 DAP in 2005. Cotton in 76- and 102-cm row plots was harvested with a 2-row John Deere 9930 spindle picker set on 102-cm rows. In 76-cm row plots, a row of plants was removed by hand the day before harvest to allow harvest of the adjacent row with a single header of the same 2-row picker. Cotton in 25-cm row plots was harvested with a John Deere 4750 stripper harvester equipped with a 3.94-m-wide finger-type header and a precleaner. Row spacing was thus confounded by harvest method in this study, because finger stripping is the only method currently available for mechanical harvesting of cotton in 25-cm rows.

Seedcotton from each plot was weighed at harvest, and a subsample was collected for ginning at the West Tennessee Research and Education Center. All seedcotton samples were uniformly air dried before ginning with a 20-saw gin assembly that included a stick machine, incline cleaners, and two lint cleaners. Lint yield was defined as the mass of lint harvested per unit ground area, calculated from seedcotton weight, gin turnout, and plot area harvested. All yields are reported on a field area basis.

Multiyear data from irrigated and nonirrigated tests were analyzed separately by the General Linear Model (GLM) and Mixed procedures of SAS, version 9.2. The GLM procedure was run with fixed effects only to determine the relative magnitude of main effects of treatments and year x treatment interactions. Year x treatment interactions were considered non-negligible (in the sense of Littell et al., 2006) in cases where P(F) < 0.05 and F values were higher than those of corresponding main effects. In these cases, the Mixed procedure was applied to individual year data with replications as random effects. Otherwise, the Mixed procedure was applied to multiyear data with years and replications as random effects. Least square means were separated by independent pairwise t tests, using the pdiff option at the 0.05 level of significance.

Agronomic yield components analyzed in this study were plants m–2, bolls plant–1, seedcotton weight boll–1, and the lint fraction of seedcotton. The lint fraction was estimated by gin turnout, as ginning is a standard component of mechanical production of cotton lint. The relative influence of these components was estimated by a simple method developed by Piepho (1995) and verified in cotton by Wu et al. (2004). The contribution of each component to variability of lint yield was calculated as the coefficient:

Formula
where ci = sum of variance and associated covariances for each ln-transformed yield component; {sigma}i2 = Var [ln (xi)] = the variance of the ith component on a logarithmic scale; {sigma}ij = Cov [ln (xi), ln (xj)] = the covariance between the ith and the jth components on a logarithmic scale. In this model, lint yield was represented by the arithmetic product of the four components mentioned above, which became additive on logarithmic transformation. Variances, covariances, and coefficients were calculated with Excel spreadsheet functions.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
For all but one response variable, the main effects of treatment were more significant than effects of the corresponding year x treatment interaction (Tables 1 and 2 ), so their treatment effects were analyzed across years. The one exception occurred in the nonirrigated study, where the year x row spacing interaction for lint yield was more significant than main effects of row spacing (Table 1). Most of this interaction was produced when 2005 yield data were combined with the 2003 and 2004 data, so 2005 yields were analyzed separately. There was no significant year x row spacing interaction for lint yield in 2003 and 2004, so yield data were analyzed across those 2 yr.


View this table:
[in this window]
[in a new window]

  		Table 1. Observed significance levels of F tests of responses to nonirrigated row spacings and configurations in 2003–2005, assuming all fixed effects. Significant year x treatment interactions with higher F ratios than those for main effects are shaded.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Observed significance levels of F tests of responses to irrigated row spacings and configurations in 2003–2005, assuming all fixed effects. There were no significant year x treatment interactions with higher F ratios than those for main effects.

 
Satisfactory plant stands were obtained in all 3 yr of the study in both nonirrigated and irrigated plots (Table 3 ). Stands averaged 84% of seeds planted in 25-cm rows and 78% of seeds planted in 76- and 102-cm rows. In the nonirrigated study, plant density ranged from 6.8 plants m–2 in 102-cm 2 x 1 skip-rows to 28 plants m–2 in 25-cm solid-planted plots. In the irrigated study, density ranged from 6.6 plants m–2 in 102-cm skip-rows to 26 plants m–2 in 25-cm solid-planted plots.


View this table:
[in this window]
[in a new window]

  		Table 3. Plant population, final plant height, canopy interception of photosynthetically active radiation (PAR), and pigweed suppression in nonirrigated and irrigated cotton as influenced by row spacing and configuration, 2003–2005.

 
Both row spacing and configuration affected final plant height, with reductions in height related to plant density (Table 3). These height reductions are consistent with earlier reports (Jost and Cothren, 2000; Vories et al., 2001; Boquet, 2005), although those studies only examined solid-planted cotton. Cotton planted in 2 x 1 skip-rows averaged 6 to 7 cm taller than in solid plantings, consistent with the lower plant densities in skip-row patterns. Canopy interception of PAR was also directly related to density during squaring (41–48 DAP) in both irrigated and nonirrigated fields. The skip-row pattern intercepted significantly less radiation than solid plantings in early season. By cutout (83–88 DAP), however, row spacing and pattern had an interactive effect on canopy PAR interception. There were no significant differences between solid and 2 x 1 skip-rows in 25-cm rows, while skip-row continued to intercept less radiation than solid plantings in the other row widths. With PAR interception > 90%, the 25-cm skip-row planting had essentially closed canopy by cutout in both irrigated and nonirrigated crops. Canopy of the 76-cm solid planting also reached 90% interception by cutout.

Suppression of pigweed growth by the cotton crop was affected interactively by row spacing and pattern (Table 3). As expected, weed suppression was generally related to canopy PAR interception. Under nonirrigated conditions, weed suppression was maintained in solid plantings of all row widths, but was reduced in skip-rows where PAR interception was < 75%. Under irrigated conditions, however, pigweed suppression was reduced in plots with <90% PAR interception, possibly due to additional flushes of pigweed emergence induced by irrigation events. In both the nonirrigated as well as irrigated tests, lower pigweed numbers as well as less overall pigweed growth were observed in the narrow 25-cm rows compared with the 102-cm row spacing. In addition, the solid row pattern provided better overall pigweed control than the skip-row pattern. Pigweeds became most prominent in the 102-cm skipped rows. The pigweed response to the row spacing and row configuration can be explained by the 25-cm and solid cotton row configuration capturing more PAR than the wider and skip-row configuration treatments (Table 3). These results are similar to row width studies in soybean where narrow rows reduced pigweed biomass through less germination, less overall growth, and higher pigweed mortality rates in low PAR environments (Légère and Schreiber, 1989; Steckel et al., 2003; Steckel and Sprague, 2004). Because the pigweed in the study germinated much later than the cotton and was removed after 4 to 5 wk of growth, pigweed interference with lint yield was negligible. Cotton growers who adopt skip-row cotton production systems can manage late weeds either by adapting their directed sprayers to the row configuration or by planting transgenic varieties that tolerate late-season applications of glyphosate.

Under nonirrigated conditions in 2003 and 2004, lint yield was not significantly affected by main effects of row width or configuration, although there were interactive effects (Table 4 ). In 2005, yield generally decreased as row spacing increased. In all years, the 2 x 1 skip-row pattern reduced yields relative to solid planting only in 102-cm rows, where skip-row cotton yielded 85% of solid planting. The reason for a different yield response pattern in 2005 compared with 2003–2004 is unclear, but does not seem to be related to differences between years in rainfall or heat-unit accumulation, summarized in Fig. 1. It may be related to lodging of plants by high winds in 2005, which was more severe in wider rows and skip-row plots (data not shown). Under irrigation, however, yield responses were relatively consistent across years, with significant main effects and interactions (Table 5 ). Lint yields were higher in solid 76-cm rows than solid 25- or 102-cm rows under irrigation. Again, the skip-row pattern reduced yields only in the 102-cm rows, where skip-row yield was 87% of solid planting. The skip-row yield response differs from Hons and McMichael (1986), who found no significant differences in lint yields between solid and 2 x 1 skip-row cotton in 1-m rows in either a dry year or one with normal rainfall in Texas. Our results are similar to King et al. (1986), who reported a 12% reduction of seedcotton yield in 2 x 1 skip-rows, relative to solid planting in 102-cm rows in Alabama.


View this table:
[in this window]
[in a new window]

  		Table 4. Lint yield, gin turnout, and accumulation of seedcotton yield in nonirrigated cotton as influenced by row spacing and configuration, 2003–2005.

 

View this table:
[in this window]
[in a new window]

  		Table 5. Lint yield, gin turnout, and accumulation of seedcotton yield in irrigated cotton as influenced by row spacing and configuration, 2003–2005.

 
Under nonirrigated conditions, relative yield performance of UNR cotton differed considerably among years (Table 4). In 2003–2004, cotton in solid 25-cm rows yielded 11% less lint than in solid 102-cm rows. In 2005, yields were 28% higher in solid 25-cm rows than in solid 102-cm rows. However, yields apparently varied more among years in solid 102-cm rows than in solid 25-cm rows (Table 4). A similar pattern of yield variation between years can be seen in data reported by Jost and Cothren (2000). In contrast, data from the irrigated portion of our study indicated similar 3-yr lint yields in solid 25- and 102-cm rows (Table 5). This result differs from Boquet (2005) who obtained lower yields in 25-cm rows than in bedded 102-cm rows under irrigation. The current study showed consistently similar yields in 25-cm solid and skip-rows, and in 76-cm solid and skip-rows, although lint yields were higher in solid 76-cm than in solid 25-cm rows under irrigation. Our results in 25-cm rows are consistent with Buehring et al. (2006) who found similar yields in 38-cm solid and skip-rows. In 76-cm rows, however, Buehring et al. (2006) obtained higher yields in solid planting than in 2 x 1 skip-rows in three of six site-years reported.

Gin turnout of 25-cm row cotton was lower than that from 76- and 102-cm rows in both nonirrigated and irrigated studies (Tables 4 and 5). This difference in gin turnout was mainly attributed to harvest method, as the seedcotton from finger stripping often has more extraneous plant material than seedcotton from spindle picking (Vories et al., 2001). In studies where cotton was hand-picked, Jost and Cothren (2000) found no significant differences in lint percentage in samples from several row widths, although Jost and Cothren (2001) obtained slightly higher lint percentage from 38-cm rows than from 76 or 102 cm. Boquet (2005) also obtained slightly higher lint percentages in hand-picked samples from 25-cm than 102-cm rows under rainfed conditions, but obtained the opposite result with irrigation. The current results are consistent with the mechanically harvested studies of Vories et al. (2001).

Earliness of maturity was influenced by row width under irrigated and nonirrigated conditions and also by row configuration under irrigation. A greater proportion of seedcotton had accumulated by 130 DAP in 25- and 76-cm rows compared with 102-cm rows under nonirrigated conditions (Table 4). With irrigation, a greater proportion of seedcotton had accumulated by 137 DAP in solid 76- and 102-cm rows compared with corresponding skip-rows (Table 5). By contrast, row pattern did not influence earliness of cotton in 25-cm rows on any date in both irrigated and nonirrigated crops. Differences in yield accumulation due to treatment roughly paralleled those for PAR interception (Table 3) and are consistent with 1 yr of data reported by Jost and Cothren (2001). They found that > 19 plants m–2 in 19-cm rows produced a greater proportion of total yield by 112 to 122 DAP than ≤ 13 plants m–2 in 19-, 38-, 76-, or 102-cm rows on a heavy clay soil. They also reported higher leaf area index at 77 DAP with > 19 plants m–2 than with lower plant density treatments, which is consistent with the PAR interception differences in the current study. Results suggest that early canopy closure is conducive to early maturity of the crop, provided that plant density is not excessive for the particular environment. A delay of maturity increases the risk of encountering inclement weather as harvest is delayed, especially in short-season environments. Gwathmey and Steckel (2006) reported that cotton in 102-cm skip-rows required nine more days to produce 60% of lint yield than in 25-cm skip-rows in Tennessee. In a longer season environment, however, Roche et al. (2003) found no differences in maturity of cotton grown in 25- and 100-cm rows.

The number of bolls plant–1 varied inversely with plants m–2 in this study, as in several earlier comparisons of UNR and wide-row cotton (Jost and Cothren, 2001; Boquet, 2005; Vories and Glover, 2006). With narrower rows and solid plantings, the number of bolls plant–1decreased less than proportionally to the increase in plants m–2, generally producing more bolls m–2 than in wider rows and skip-row cotton (Table 6 ). However, bolls m–2 also appeared to vary inversely with seedcotton weight boll–1, compensating to some extent for bolls m–2. In solid-planted rows without irrigation, the number of bolls m–2 increased by 31% as row spacing decreased from 102 to 25 cm, unlike Boquet (2005) but similar to 1 yr of results of Jost and Cothren (2001). With irrigation, the number of bolls m–2 did not respond significantly to solid row spacing, but in skip-row configuration, bolls m–2 decreased as row width increased to 102 cm. In both irrigated and nonirrigated crops, boll weight decreased on the order of 10 to 15% as row spacing decreased from 102 to 25 cm in both solid and skip-row patterns. Vories and Glover (2006) found similar boll weights in 19- and 97-cm rows, but Jones (1998) reported larger bolls in 102-cm rows and 2 x 1 skip-rows compared with 76-cm rows and solid plantings, respectively.


View this table:
[in this window]
[in a new window]

  		Table 6. Boll number per plant, boll number per square meter ground area, and seedcotton weight per boll of nonirrigated and irrigated cotton as influenced by row spacing and configuration, 2003–2005.

 
The relative influence of lint yield components across years and treatments in the irrigated and nonirrigated studies are presented in Table 7 . Under nonirrigated conditions, about three-fourths of the variation in the lint yield product was attributed to plants m–2. The relatively large negative covariance between plants m–2 and bolls plant–1 indicates a compensatory interaction of these two yield components (Piepho, 1995). Thus bolls plant–1 exerted relatively little influence over lint yield after yield compensation was taken into account. Under irrigated conditions, however, plants m–2 and bolls plant–1 each accounted for about one-third of yield variation. The greater influence of bolls plant–1 under these conditions was apparently due to higher boll load with irrigation (Table 6). Results suggest that irrigated cotton may support a higher plant population with less compensatory adjustment of boll load than nonirrigated cotton. It also supports the hypothesis that the capacity to produce high yields at a high plant density depends considerably on adequate moisture supply for boll development.


View this table:
[in this window]
[in a new window]

  		Table 7. Variances, covariances, and relative influence (ci) of logarithmically transformed yield components in nonirrigated and irrigated studies of cotton row spacing and patterns, 2003–2005. Reported values = 100 x calculated values.

 

   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Cotton grown in narrower rows and solid plantings tended to close canopy earlier and more completely, suppress weed growth, and mature earlier than wider rows and a 2 x 1 skip-row pattern in this study. Incomplete canopy closure in wide 2 x 1 skip-row spacings resulted in less weed suppression, indicating a need for additional weed control in this configuration. However, weed growth in skipped rows tended to diminish with row spacing, suggesting an opportunity to gain some advantages of skip-row cotton without creating large weed problems in the field. Possible advantages of narrow skip-row planting may include lower planting costs, soil moisture conservation, and establishment of tram-lines for field equipment traffic, but these possibilities need further research. Later maturity with a 2 x 1 skip-row pattern tended to diminish with narrower row spacing. Time to crop readiness for harvest-aid application was 7 to 9 d later in 102-cm skip-rows than in 25-cm skip-rows, which is an important difference in short-season environments where risk of inclement weather tends to increase in late season. In nonirrigated conditions, row spacing and configuration for optimal lint yield depended on growing conditions of a particular year. Under irrigation, however, cotton grown in solid-planted 76-cm rows consistently yielded more than cotton in either 25-cm or 102-cm solid or skip-rows. With or without irrigation, however, lint yields were 13 to 15% lower in 102-cm skip-rows than in 102-cm solid planting, while yields did not differ between solid and skip-rows in narrower row widths. This finding suggests that some cost savings with skip-rows may be gained without yield sacrifice in rows spaced 76-cm or less, a finding that merits economic analysis. Future research should evaluate whether potential input cost savings offset any potential revenue losses with narrow skip-row planting.


   ACKNOWLEDGMENTS
 
This research was supported in part by a Cotton Incorporated State Support Project. We thank Pat Brawley, Ernest Merriweather, Carl Michaud, Clint Sharp, Kevin Willis, and the farm crew of the Research and Education Center at Milan for their technical assistance.

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


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




This article has been cited by other articles:


Home page
 Agron. J.Home page
J. A. Larson, C. O. Gwathmey, D. F. Mooney, L. E. Steckel, and R. K. Roberts
Does Skip-Row Planting Configuration Improve Cotton Net Return?
Agron. J., June 2, 2009; 101(4): 738 - 746.
[Abstract] [Full Text] [PDF]

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 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 HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gwathmey, C. O.
Right arrow Articles by Larson, J. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Gwathmey, C. O.
Right arrow Articles by Larson, J. A.
Agricola
Right arrow Articles by Gwathmey, C. O.
Right arrow Articles by Larson, J. A.
Related Collections
Right arrow Weed Management
Right arrow Other Crop Management
Right arrow Other Cropping Systems
Right arrow Cotton
Right arrow Crop Systems

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