Published online 19 March 2008
Published in Crop Sci 48:601-605 (2008)
© 2008 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
Genetic Gain in Yield Potential of Upland Cotton under Varying Plant Densities
Brian M. Schwartz* and
C.W. Smith
Dep. of Soil and Crop Sciences, Texas A&M Univ., College Station, TX 77843-2474
* Corresponding author (aggiegtr{at}ufl.edu).
 |
ABSTRACT
|
|---|
Genetic gain studies have been used to evaluate the historical improvement of different traits, to provide insights into magnitudes of gain possible in future cultivars, and to defend the role of genetics during periods of stagnant or decreasing yield trends. This study was conducted over a 2-yr period and included nine current or obsolete cotton (Gossypium hirsutum L.) cultivars grown in five plant densities to evaluate genetic gain with varying levels of interplant competition. The rates of genetic gain for lint yield were highest in the commercial, 1 by 0.3 m, and 1 by 1 m plant spacing treatments with slopes of 8.7, 8.2, and 7.1 kg ha–1 yr–1, respectively. Slopes were reduced in the 2 by 2 m and 3 by 3 m spacing treatments with gains of 3.6 and 1.5 kg ha–1 yr–1, respectively, implying that for lint yield, genetic gains have been made for tolerance to interplant competition and not only yield potential per se.
|
INTRODUCTION
|
|---|
ESTIMATES FOR THE RATE OF genetic gain for yield in cotton (Gossypium hirsutum L.) over at least a 40-yr time frame have ranged from 3.7 to 10.2 kg ha–1 yr–1 (Bayles et al., 2005; Bassett and Hyer, 1985; Bridge et al., 1971; Culp and Green, 1992; Hoskinson and Stewart, 1977; Meredith, 2002; Wells and Meredith, 1984c). Such calculations are always contingent on the time period represented by the cultivars, the cultivars or germplasm lines selected to represent eras, and the environment in which they were tested (i.e., there are no absolute genetic gain values). Bridge (1990) suggested that resistance or tolerance to biotic stress such as disease and insect pests and abiotic stresses such as drought and heat have influenced genetic gain, and Duvick and Cassman (1999) demonstrated that tolerance to interplant competition also played a role in gain from selection in maize (Zea mays L.). Generally, the evaluations of yield gains in cotton have suggested that modern cultivars support more bolls per unit production area, implying also that they support more bolls per plant, and consistently generate higher lint percentages. Other underlying components of increased yield potential have been attributed to smaller bolls, smaller seed, earlier maturity, and higher micronaire values (Bridge and Meredith, 1983; Bridge et al., 1971; Culp and Green, 1992; Hoskinson and Stewart, 1977; Meredith et al., 1997; Moser and Percy, 1999; Turner et al., 1976). However, some of the genetic gain in "stripper" type cottons cultivars adapted to the Plains in Oklahoma was attributed to larger bolls and seeds (Bayles et al., 2005). In a series of publications, Wells and Meredith (1984a, 1984b, 1984c) reported on the physiological differences between obsolete and modern cultivars, such as an earlier, more complete transition from vegetative to reproductive dry matter partitioning that resulted in a greater amount of boll development occurring when leaf area and mass are at a maximum. Recent cultivars also partitioned more dry matter into reproductive structures without increasing total dry matter. They generally produce a greater number of smaller bolls with a higher lint percentage before the shedding of fruit that is common later in the season.
Research involving variable planting densities in upland cotton have not included densities designed to eliminate interplant competition altogether. The majority of past studies have concentrated on identifying the optimum number of plants per hectare for increasing yield in commercial situations (Bridge et al., 1973; Fowler and Ray, 1977; Hawkins and Peacock, 1970; Smith et al., 1979). Duvick and Cassman (1999) demonstrated that modern maize hybrids only outperform older ones when grown at higher plant densities or under stressful conditions, but in a stress-free isolation environment, the genetic yield potential of older and newer hybrids were comparable. Results indicating increased productivity in modern cotton cultivars are limited to experiments comparing modern and obsolete cultivars planted at densities common to then current production practices (Bayles et al., 2005; Bridge and Meredith, 1983; Culp and Green, 1992).
Research reported herein was initiated to determine if the rates of genetic gain for yield and selected yield components in a set of obsolete and modern cultivars varied with plant densities.
|
MATERIALS AND METHODS
|
|---|
Nine cotton cultivars, one modern and eight obsolete, were planted into five discrete plant density treatments. Each cultivar was selected for evaluation on the basis of adaptation to growing conditions in College Station, TX, during their respective eras of production and the proximity of their year of development or release to the beginning of each decade, from 1900 through 2000.
In 2003 and 2004, the genotypes were grown in five plant densities designed to evaluate varying levels of interplant competition. Plant density treatments were single plant culture with plants spaced 3 by 3 m, 2 by 2 m, 1 by 1 m, 1 by 0.3 m, and 1 m by 0.07 to 0.1 m, hereafter referred to as the commercial density. Plots within plant densities were single rows, 12 m long in a randomized complete block design with four replications. All other cultural practices were standard for College Station, TX, including furrow irrigation. Genotypes were compared within and among each density treatment to estimate genetic change for: lint yield, lint percentage, boll size, seeds per boll, and seed index.
Seed for all genotypes in the 3 by 3 m, 2 by 2 m, 1 by 1 m, and 1 by 0.3 m density treatments were hand planted, three seeds per hill, and thinned to one plant per hill 2 wk after emergence. Seed for all genotypes in the commercial population were machine planted with a plot planter. The experiments were monitored weekly and weeds removed as needed. Five mature bolls from the middle of the plant were taken randomly from each plant in the 3 by 3 m, 2 by 2 m, 1 by 1 m, and 1 by 0.3 m densities before application of defoliant. Fifty mature bolls were taken from the middle of the fruiting zone in the commercial density treatment before harvest. Defoliant was applied to each density treatment only when the latest maturing cultivar reached the visually estimated 60% open boll stage to ensure that each plant reached its genetic potential. All plants in the 3 by 3 m (five plants) and 2 by 2 m (seven plants) density treatments were hand harvested. In the 1 by 1 m density, only those plants grown under equal levels of interplant competition throughout the season, a maximum of 11 plants, were hand harvested. Five plants grown under equal levels of interplant competition were hand harvested randomly in the 1 by 0.3 m density treatment. Plots in the commercial density were harvested with a one-row picker modified for plot harvest with grab samples taken from each plot for determination of gin turnout.
Seed cotton from each individual plant from the 3 by 3 m, 2 by 2 m, 1 by 1 m, and 1 by 0.3 m density treatments was ginned on an eight-saw laboratory gin. Lint yield per plant was recorded and lint percentage calculated. Boll sample weights were added back to the individual plant weights and lint yield per plant was converted to lint yield per hectare for ease of comparative purposes. Grab samples from the commercial density plots were ginned on an eight-saw laboratory gin and lint percentage calculated to determine lint yield per hectare. Several yield components were measured or calculated using the five boll samples for each plant in all plant densities except the commercial density where the 50-boll samples were used. Yield components determined in this study were boll size as the weight of seedcotton per boll, seeds per boll, and seed index as the weight of 100 fuzzy seed.
Plot means were used for the analyses of variance and regression reported herein. Bartlett's test for homogeneity of variances was performed for each trait, testing the equality of variance between cultivars in each plant density treatment. Variances among each trait's rate of genetic change slopes over the five plant density treatments were tested also. When conditions of homogeneity of variances were not met, the following data transformations were applied to the dataset and tested by Bartlett's: x2, x3, x–1, x–2, x–3, x1/2, x–1/2, log(x), ln(x), or ex. Transformed data sets were used in the analysis of variance if their variances were homogeneous. If no transformation succeeded, the untransformed data were used in the analyses of variance and regression analyses.
An analysis of variance was performed on each of the measured traits, in each density treatment. The General Linear Models (GLM) procedure in SAS software (SAS Institute, 2007) assuming a mixed model with genotypes and years considered as fixed variables and replications considered as a random variable was chosen for the analysis.
Linear regression analyses were performed for each trait to obtain the corresponding rate of genetic change slope in each replication for all five plant density treatments. An analysis of variance was performed on the rates of genetic change slopes of each of the measured traits, across all density treatments. Data were analyzed by PROC GLM (SAS Institute, 2007) assuming a mixed model with density and years considered as fixed variables and replications considered as random.
|
RESULTS AND DISCUSSION
|
|---|
Lint Yield
The four genotypes released between 1905 and 1930 produced less lint per hectare than the five cultivars released between 1941 and 2002 across all plant densities evaluated (Table 1
). ‘Deltapine 491’, released in 2002, yielded more lint per hectare than any other cultivar tested regardless of year of release. Average yields of these cultivars indicate a general trend toward improved yield potential during the 20th century, regardless of plant density.
View this table:
[in this window]
[in a new window]
|
Table 1. Mean lint yield of nine obsolete and modern upland cotton cultivars grown in five plant density treatments at College Station, TX, in 2003 and 2004.
|
|
Rates of genetic change for lint yield varied across plant densities and no density x year interaction was noted (Table 2
). The rates of genetic change were highest in the commercial, 1 by 0.3 m, and 1 by 1 m density treatments with regression slopes indicating genetic gains of 8.7, 8.2, and 7.1 kg ha–1 yr–1, respectively, across the 20th century (Table 3
). In the 2 by 2 m and 3 by 3 m spaced populations, the rates of change (i.e., slopes) were each smaller but not equal to zero, with gains of only 3.6 and 1.5 kg ha–1 yr–1, respectively. Higher rates of genetic change under conditions of greater interplant competition are a result of increased lint yields in the more recently released cultivars, particularly Deltapine 491. This implies that for this trait, genetic gains have been made for tolerance to interplant competition and not only yield potential per plant per se. These results, while not as distinct, are similar to those of Duvick and Cassman (1999) who reported that modern maize hybrids only out-performed obsolete hybrids at higher plant densities. Tollenaar and Wu (1999) and Tollenaar et al. (1997) suggested that newer maize hybrids reached their per unit land area yield potential through genetic improvements in stress tolerance, the same pathway suggested by Bridge (1990). While increased yields in maize across the 20th century may be associated with increased commercial planting densities, this is not the case with upland cotton since recommended plant spacing within the drill has been 8 to 10 cm (equal to the commercial density in this study) throughout the last half of the century with research data supporting this spacing before 1950 (Christidis and Harrison, 1955; Hawkins and Peacock, 1970). Possible explanations for the higher rates of gain at the higher plant densities in this study could include improved tolerance to shading, and thus an increased capacity to retain fruit when grown at commercial plant densities.
View this table:
[in this window]
[in a new window]
|
Table 2. Mean squares for rates of genetic change for lint yield, lint percentage, boll size, seeds per boll, and seed index of upland cotton grown in five plant densities at College Station, TX, in 2003 and 2004.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Mean rates of genetic change for lint yield, lint percentage, boll size, seeds per boll, and seed index of upland cotton grown in five plant densities at College Station, TX, in 2003 and 2004.
|
|
Yield Components
Breeders generally make decisions for cultivar release based on the competitiveness of an individual genotype relative to lint yield and fiber quality, with secondary consideration of individual components of yield that make up final total lint yield. Table 2 shows that rates of genetic change for lint percentage varied across plant densities and no density x year interaction was noted. In addition, rates of genetic change for boll size, seeds per boll, and seed index did not differ across plant densities, but density x year interactions were significant for seeds per boll and seed index. Overall, interpretation of the rates of genetic change for yield components was more difficult than for lint yield (Table 3). The rates of change for lint percentage were highest in the 1 by 0.3 m and 1 by 1 m treatments, each having slopes of 0.11% yr–1. While differences in mean values of other yield components (Tables 4
and 5
) in these cultivars were detected, regression analyses did not suggest any trends in rates of genetic change relative to plant density treatments. Data presented in Table 4 suggest that lint percentage tended to increase across this set of cultivars with year of release regardless of plant density. Lint percentage of cultivars in this study that were released by 1930 ranged from 26 to 34% lint while those released in 1941 or later averaged >35% across the plant density treatments studied. However, ‘Half & Half’, which was released in 1910, averaged 37% lint and was not different than several cultivars released later in the century. Trends of increasing lint percentage have been reported throughout the literature and have been used in part to explain expanding lint yields (Bridge et al., 1971; Culp and Green, 1992; Hoskinson and Stewart, 1977; Miller and Rawlings, 1967; Miller et al., 1958; Moser and Percy, 1999; Wells and Meredith, 1984c). In contrast, Bridge and Meredith (1983) concluded that lint percentage had not changed in cultivars released after ‘Deltapine 14’ in 1941. The number of seeds per boll and seed index tended to decrease with year of cultivar release (Table 5). This decrease in seeds per boll and seed index probably contributed to the general decrease in boll size and increased lint percentage with year of release as shown in Table 4.
View this table:
[in this window]
[in a new window]
|
Table 4. Means for lint percentage and boll size of nine obsolete and modern upland cotton cultivars grown in five plant density treatments at College Station, TX, in 2003 and 2004.
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Means for seeds per boll and seed index of nine obsolete and modern upland cotton cultivars grown in five plant density treatments at College Station, TX, in 2003 and 2004.
|
|
These data support the hypothesis that increased yields in upland cotton are primarily influenced by enhanced retention of bolls with minor influence from lint percentage, which is impacted by seed size and number as proposed by Culp and Green (1992), Harrell and Culp (1976), Hoskinson and Stewart (1977), Miller et al. (1958), Moser and Percy (1999), Ramey (1972), Turner et al. (1976), and Wells and Meredith (1984c).
|
CONCLUSIONS
|
|---|
Estimates of genetic gain in lint yield in upland cotton are affected by plant density, with higher estimates for cultivars grown in densities of 1 by 1 m spacing or less, and lower estimates derived when plants are spaced 3 by 3 m. These data suggest breeders have made less progress in selecting for higher genetic yield potential in upland cotton than previously reported. However, these data do demonstrate that breeders have made gains in yield potential per plant as indicated by 1.5 and 3.6 kg ha–1 yr–1 gain estimates when plants were grown in single spaced culture at 3 by 3 m and 2 by 2 m, respectively. Contrary to lint yield, plant density treatments had little impact on estimates of genetic gain for the yield components lint percentage, boll size, seeds per boll, and seed index. Overall, this study provides support for additional research to compare the efficacy of selection in field experiments evaluated under different levels of interplant competition.
|
NOTES
|
|---|
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. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication January 25, 2007.
|
REFERENCES
|
|---|
- Bassett, D.M., and A.H. Hyer. 1985. Acala cotton in California: 60 years of varietal improvement. p. 76. In Proc. Beltwide Cotton Prod. Res. Conf., New Orleans, LA. 6–11 Jan. 1985. Natl. Cotton Council Am., Memphis, TN.
- Bayles, M.B., L.M. Verhalen, W.M. Johnson, and B.R. Barnes. 2005. Trends over time among cotton cultivars released by the Oklahoma Agricultural Experiment Station. Crop Sci.
45
:966–980.[Abstract/Free Full Text]
- Bridge, R.R. 1990. Improvements in cotton through genetics and breeding since 1975. p. 16–18. In Proc. Beltwide Cotton Prod. Conf., Las Vegas, NV. 10–11 Jan. 1990. Natl. Cotton Council Am., Memphis, TN.
- Bridge, R.R., and W.R. Meredith, Jr. 1983. Comparative performance of obsolete and current cotton cultivars. Crop Sci.
23
:949–952.[Abstract/Free Full Text]
- Bridge, R.R., W.R. Meredith, Jr., and J.F. Chism. 1971. Comparative performance of obsolete varieties and current varieties of upland cotton. Crop Sci.
11
:29–32.[Abstract/Free Full Text]
- Bridge, R.R., W.R. Meredith, Jr., and J.F. Chism. 1973. Influence of planting method and plant population on cotton (Gossypium hirsutum L.). Agron. J.
65
:104–109.[Abstract/Free Full Text]
- Christidis, B.G., and G.J. Harrison. 1955. Cotton growing problems. McGraw-Hill, New York.
- Culp, T.W., and C.C. Green. 1992. Performance of obsolete and current cultivars and Pee Dee germplasm lines of cotton. Crop Sci.
32
:35–41.[Abstract/Free Full Text]
- Duvick, D.N., and K.G. Cassman. 1999. Post-green revolution trends in yield potential of temperate maize in the north-central United States. Crop Sci.
39
:1622–1630.[Abstract/Free Full Text]
- Fowler, J.L., and L.L. Ray. 1977. Response of two cotton genotypes to five equidistant spacing patterns. Agron. J.
69
:733–738.[Abstract/Free Full Text]
- Harrell, D.C., and T.W. Culp. 1976. Effects of yield components on lint yield of upland cotton with high fiber strength. Crop Sci.
16
:205–208.[Abstract/Free Full Text]
- Hawkins, B.S., and H.A. Peacock. 1970. Yield response of upland cotton (Gossypium hirsutum L.) to several spacing arrangements. Agron. J.
62
:578–580.[Abstract/Free Full Text]
- Hoskinson, P.E., and J.M. Stewart. 1977. Field performance of two obsolete cotton cultivars. p. 78–79. In Proc. Beltwide Cotton Prod. Res. Conf., Atlanta, GA. 10–12 Jan. 1977 Natl. Cotton Council Am., Memphis, TN.
- Meredith, W.R., Jr. 2002. Factors that contribute to lack of genetic progress. In Proc. Beltwide Cotton Conf., Atlanta, GA. 8–12 Jan. 2002. Natl. Cotton Council Am., Memphis, TN.
- Meredith, W.R., Jr., J.J. Heitholt, W.T. Pettigrew, and S.T. Rayburn, Jr. 1997. Comparison of obsolete and modern cotton cultivars at two nitrogen levels. Crop Sci.
37
:1453–1457.[Abstract/Free Full Text]
- Miller, P.A., and J.O. Rawlings. 1967. Selection for increased lint yield and correlated responses in upland cotton, Gossypium hirsutum L. Crop Sci.
7
:637–640.[Abstract/Free Full Text]
- Miller, P.A., J.C. Williams, Jr., H.F. Robinson, and R.E. Comstock. 1958. Estimates of genotypic and environmental variances and covariances in upland cotton and their implications in selection. Agron. J.
50
:126–131.[Abstract/Free Full Text]
- Moser, H.S., and R.G. Percy. 1999. Genetic improvement of yield, yield components, and agronomic characteristics of pima cotton: 1949–1991. p. 488–489. In Proc. Beltwide Cotton Conf., Orlando, FL. 3–7 Jan. 1999. Natl. Cotton Council Am., Memphis, TN.
- Ramey, H.H., Jr. 1972. Yield response of six cultivars of upland cotton, Gossypium hirsutum L., in two cultural regimes. Crop Sci.
12
:353–354.[Abstract/Free Full Text]
- SAS Institute. 2007. The SAS system for Windows: Release 9.1.3. SAS Inst., Cary, NC.
- Smith, C.W., B.A. Waddle, and H.H. Ramey, Jr. 1979. Plant spacings with irrigated cotton. Agron. J.
71
:858–860.[Abstract/Free Full Text]
- Tollenaar, M., A. Aguilera, and S.P. Nissanka. 1997. Grain yield is reduced more by weed interference in an old than in a new maize hybrid. Agron. J.
89
:239–246.[Abstract/Free Full Text]
- Tollenaar, M., and J. Wu. 1999. Yield improvement in temperate maize is attributable to greater stress tolerance. Crop Sci.
39
:1597–1604.[Abstract/Free Full Text]
- Turner, J.H., H.H. Ramey, Jr., and S. Worley, Jr. 1976. Trends for yield and quality in cotton breeding since 1960. p. 107. In Proc. Beltwide Cotton Prod. Res. Conf., Las Vegas, NV. 5–7 Jan. 1976. Natl. Cotton Council Am., Memphis, TN.
- Wells, R., and W.R. Meredith, Jr. 1984a. Comparative growth of obsolete and modern cotton cultivars: I. Vegetative dry matter partitioning. Crop Sci.
24
:858–862.[Abstract/Free Full Text]
- Wells, R., and W.R. Meredith, Jr. 1984b. Comparative growth of obsolete and modern cotton cultivars: II. Reproductive dry matter partitioning. Crop Sci.
24
:863–868.[Abstract/Free Full Text]
- Wells, R., and W.R. Meredith, Jr. 1984c. Comparative growth of obsolete and modern cotton cultivars: III. Relationship of yield to observed growth characteristics. Crop Sci.
24
:868–872.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
B. M. Schwartz and C. W. Smith
Genetic Gain in Fiber Properties of Upland Cotton under Varying Plant Densities
Crop Sci.,
July 1, 2008;
48(4):
1321 - 1327.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
