Abstract
Neps are fiber entanglements created during the mechanical processing of cotton and are often associated with immature fibers. Even in small amounts neps can affect textile quality and cotton marketability. Machine harvesting, lower fiber linear density (fineness), and more immature bolls at harvest are factors that contribute to neps. However, it is not clear whether differences in fiber linear density or immature bolls at harvest combine with harvest method to substantially affect neps. The aim of this study was to compare machine spindle and hand-harvested cotton collected from four field studies with treatments that differed in percentage of immature bolls and fiber linear density at harvest (resulting from differences in harvest aid timing) and to test for statistical interactions. By systematically varying the timing of harvest aids to cease crop growth, removing fruiting branches, or both, differences in the percentage of immature bolls and fiber linear density were generated. In all studies spindle harvesting increased neps, but there were no significant statistical interactions between the harvest method with harvest aid timing or branch removal treatments. When all measurements of neps were combined across studies there was a multiple regression that explained the level of neps with the harvest method and fiber linear density (R2 = 0.66). These responses supported the individual season analyses, finding no statistical interaction of harvest method with either variable. Spindle harvesting increased neps by an average of 53 count/g compared to hand harvesting. Identifying reasons for differences in nep levels between cotton growing regions may assist in developing strategies to reduce neps.
Neps are fiber entanglements created during mechanical processing of cotton fiber and are often associated with dead or immature fibers. 1 Even in small amounts, neps are undesirable in cotton processing as they decrease mill processing efficiency and ruin the appearance of finished yarns and fabrics. 2 While cotton growers are rarely penalized for the presence of high levels of neps, it can, however, affect the reputation of a particular industry when cotton arrives at a spinning mill. A value of greater than 250 neps/g is considered high. 3
In a review by Van der Sluijs and Hunter, 4 many reasons are documented for differences in neps in cotton lint, yarn, and fabric resulting from changes in field production practices, cultivar choice, climatic conditions during crop growth, harvesting, and ginning. One factor that was identified as a reason for neps was the use of machine harvesting when compared to cotton harvested by hand. The review stated that neps could be increased by up to 30% prior to ginning using machine harvesting, but there was only one non-citable study reported to support this. There are few other documented investigations comparing increases in nep levels in lint cotton generated from spindle-harvested cotton compared to hand-harvested cotton.5,6
Research into the machine harvest effects on neps have included the impacts of harvest timing7–9 and harvester setup, 10 and comparisons between spindle and stripped harvested cotton. 11 Mangialardi et al. 7 and Bange et al. 12 also showed that the level of neps in spindle machine-harvested cotton was related to fiber maturity and linear density caused by differences in the maturity of bolls presented at harvest. Bange et al. 12 generated differences in boll maturity by systematically varying the timing of the harvest aids. Relationships between neps and fiber maturity, fiber linear density (fineness), or micronaire are well established4,12,13 and indicate that fewer neps are associated with fiber that is more mature as measured by the higher maturity ratio, higher fiber linear density (fineness), and higher micronaire. In all these relationships, nep levels were measured following the ginning process. Studies have also shown that increasing the mechanical handling of lint through the ginning process also leads to an increase in neps,8,12,14,15 and differences in neps could partly be attributed to fiber maturity at harvest.
No reports were identified that have specifically investigated or quantified the level of neps associated with both differences in harvest operation and changes in crop maturity at harvest. The objectives of this study were to:
Establish whether differences in crop maturity (as defined by the percentage of immature bolls) and changes in fiber linear density resulting from differences in harvest aid timing influence the level of nep generation following machine harvesting; contribute to the literature quantifying the impacts of machine spindle harvesting on nep generation.
Knowledge of these effects will help us to understand the reasons for higher neps in some cotton production regions and may assist in refining harvest management strategies to optimize fiber quality.
Materials and methods
Cultural details and treatments
Impact of harvest aid timing on fiber micronaire, linear density, and neps of spindle-harvested samples in Experiment 1 (2005/2006). Neps for corresponding hand-harvested samples are also presented. Interaction of harvest treatment and harvest aid timing was not significant (0.05 level) (DAS – days after sowing)
Significant at the 0.05 level.
Significant at the 0.01 level.
Least significant difference (LSD) calculated at the 0.05 level.
n.s. – no significant difference at the 0.05 level.
Impact of harvest aid timing on fiber micronaire, linear density, and neps of spindle-harvested samples in Experiment 2 (2006/2007). Neps for corresponding hand-harvested samples are also presented. Interaction of harvest treatment and harvest aid timing was not significant (0.05 level) (DAS – days after sowing)
Significant at the 0.05 level.
Significant at the 0.01 level.
Significant at the 0.001 level.
Least significant difference (LSD) calculated at the 0.05 level.
n.s. – no significant difference at the 0.05 level.
Impact of harvest aid timing on fiber micronaire, linear density, and neps of spindle-harvested samples in Experiment 3 (2008/2009). Neps for corresponding hand-harvested samples are also presented. No interactions between treatments were significant (0.05 level) (DAS – days after sowing)
Significant at the 0.01 level.
Significant at the 0.001 level.
Least significant difference (LSD) calculated at the 0.05 level.
n.s. – no significant difference at the 0.05 level.
Significant at the 0.1 level.
Impact of harvest aid timing on fiber micronaire, linear density, and neps of spindle-harvested samples in Experiment 4 (2009/2010). Neps for corresponding hand-harvested samples are also presented. No interactions between treatments were significant (0.05 level) (DAS – days after sowing)
Significant at the 0.05 level.
Significant at the 0.01 level.
Significant at the 0.001 level.
Least significant difference (LSD) calculated at the 0.05 level.
n.s. – no significant difference at the 0.05 level.
The application of harvest aids caused leaves to defoliate and bolls to open, resulting in differences in the number of open immature bolls at harvest. Each experiment used a randomized complete block design with four replications.
Plots (8 m × 10 m) contained eight rows spaced at 1 m. In the center two rows of each plot, a mixture of leaf defoliant and a boll opener was applied as a harvest aid. Harvest aids were sprayed with a calibrated CO2 pressurized boom with total swath width of 3.0 m using flat fan nozzles (110-01) at 200 k Pa delivering 100 l/ha of spray solution. The chemical and rates were 0.2 l/ha Dropp Liquid® (Bayer CropScience, active constituent 500 g/l thidiazuron); 3 l/ha Prep 720® (Bayer CropScience, active constituent 720 g/l ethephon); and 2 l/ha D-C Tron® (Caltex, active constituent 991 ml/l petroleum oil).
Upland cotton (Gossypium hirsutum L.) was sown with a commercial row crop planter on 15 October 2005 (Experiment 1), 20 October 2006 (Experiment 2), 16 October 2008 (Experiment 3), and on 15 October 2009 (Experiment 4). In Experiments 1 and 2, the CSIRO (Commonwealth Scientific and Industrial Research Organisation, Australia) cultivar Sicot 71BR containing Monsanto traits (Bollgard II® and Roundup Ready®) was used, while in Experiments 3 and 4, the CSIRO cultivar Sicot 71BRF 17 containing Monsanto traits (Bollgard II® and Roundup Ready Flex®) was used. Sicot 71BRF was the commercial replacement for Sicot 71BR and has a similar genetic background to Sicot 71BR. Both have normal leaf shape, medium-to-late maturity, and compact growth habit.
Each experiment was established and grown with full irrigation using non-limiting N and thorough insect control as described by Hearn and Fitt. 18 Nitrogen was applied as anhydrous ammonia (injected below and to the side of the plant line) four weeks before sowing at a rate of 200 kg N/ha. The rate of N was determined on the basis of a N replacement program that accounts for N use in previous cotton crops. 19
To determine lint yield and quality, the fourth row of each plot was harvested with a spindle harvester (John Deere 9910 modified for single row harvesting) and the seed cotton weighed. In the fifth row of the plot, a 0.5 m2 hand-harvested sample was also taken at the same time as the machine harvest. Seed cotton from both harvest methods for each plot was ginned separately using a 20 saw gin (Continental Eagle, Prattville, Alabama, USA) with a pre-cleaner; there are no lint cleaning facilities located at the ACRI. Machine and hand harvests occurred at 212 DAS (Experiment 1), 196 DAS (Experiment 2), 207 DAS (Experiment 3), and 211 DAS (Experiment 4). One sub-sample (>8 g) was taken from each plot for both machine and hand-harvested samples for fiber quality analyses.
Crop status at harvest aid application
To assess crop status when harvest aid treatments were applied, a number of measurements were taken on the control plots on each day of treatment application. On each day of treatment application, five plants were taken from 0.5 m2 within the control plots and all bolls (regardless of size or age) were removed from all plants to determine the percentage of immature bolls (immature bolls/total bolls × 100). Immature bolls were distinguished from mature bolls by cutting bolls perpendicular to their vertical axis and assessing the color of the seed coats within the bolls. A seed coat that was not dark was classified as immature. 20
Fiber quality analyses
Using part of the sub-sample for each plot of machine-harvested cotton, the micronaire (no units) of fiber samples was determined using a Zellweger Uster High Volume Instrument (HVI) 900 (Knoxville, TN). The reported HVI micronaire is an average of two measurements taken from each plot sub-sample. Each of these samples was then blended individually through one passage of a Shirley Development Laboratories ‘Shirley Analyser Mk2’, and tested for fiber linear density (µg/m or millitex (mtex = mg/km)), which is often referred to as fiber fineness, using the CSIRO Cottonscan™. 21 Plot measurements of fiber linear density were the mean of five measurements taken from the sample. Cottonscan was chosen to measure fiber linear density as it provides a direct measure of the total length of a known mass of the fiber snippets to calculate linear density. 21 Immature fiber typically has micronaire values <3.8, while fiber that is too mature and coarse has micronaire values >4.5, and linear density values >220 µg/m. 22
Nep analysis (including total nep count/g, seed coats nep count/g, fiber nep count/g) of ginned lint sub-samples from each plot was undertaken using the remaining part of the machine harvest sub-sampled and the sub-sample taken for each plot for the hand-picked harvest using the Advanced Fiber Information System (AFIS) PRO instrument (Uster Technologies, Switzerland). The AFIS instrument requires a 0.5 g sample prepared into a 25 cm long sliver, which is subsequently delivered through a series of rollers and card flats to align fibers and to isolate neps and trash. Individualized fibers and neps enter an electro-optical sensor with measured voltage/time wave forms being proportional to the amount of light scattered, which thus allows the type, number, and size of neps to be quantified. 4 Plots of AFIS PRO neps (count/g) are an average of five measurements of five individual sub-sub-samples.
Data analysis
All analyses of variance (ANOVAs) and multiple linear regression analyses were conducted using the Genstat 9 statistical package (Lawes Agricultural Trust, IACR, Rothamsted, UK). To assess the impact of machine harvesting on neps across all experiments, neps were related to fiber linear density using a quadratic response, and multiple regression analysis was used to account for differences in harvest method. Statistical differences resulting from the ANOVA were compared using Fisher’s least significant difference (LSD) method and, unless otherwise stated, statistical significant differences were compared using a better than 95% confidence interval (P < 0.05).
Results and discussion
Harvest aid timing, branch removal, and harvest method all influenced the level of total neps; there were no statistical interactions between treatments (Tables 1–4). When neps were separated into seed coat and fiber neps, as measured by the AFIS PRO, there were also no significant statistical interactions between treatments (Tables 1–4). Increases in total neps for machine-harvested cotton were a result of both significant increases in fiber and seed coat neps for this treatment. The only exception was fiber neps in Experiment 2, where there was no significant difference. Harvest aid timing affected changes in total neps through changes in fiber neps, with the exception of Experiment 3 where there was no significant difference between harvest aid timing treatments. Only in Experiment 4 did harvest aid timing affect seed coat neps. Branch removal increased fiber neps in Experiment 3, while in Experiment 4 branch removal lowered seed coat neps.
In general, across all experiments, earlier harvest aid timing treatments had more total neps and lower fiber linear density (fineness). This was not, however, always reflected in lower micronaire, with no significant differences measured in Experiment 3. Others have shown that earlier applications of harvest aids resulting in crop termination have lowered micronaire.12,23–25 Bange et al. 12 and Thibodeaux et al. 26 also found that total neps increased while fiber linear density (fineness) and maturity was lowered with earlier applications.
In Experiments 3 and 4, the effect of branch removal did not significantly affect micronaire, but lowered fiber linear density only in Experiment 3. Associated with the lower fiber linear density caused by branch removal in Experiment 3, neps were increased. There were, however, no differences in either fiber linear density or total neps caused by branch removal in Experiment 4. Lower fiber linear density in Experiment 3 is most likely associated with the delay in crop development associated with fruiting branch removal. 27 Faircloth et al. 16 showed that a delay in growth resulting from a fruiting gap (generated by fruiting branch removal) in some years lowered micronaire.
In all experiments, machine spindle-harvested cotton had higher levels of total neps (Tables 1–4; Figure 1) compared to hand-harvested cotton. Machine harvest increased neps (count/g) by 28 in Experiment 1, 25 in Experiment 2, 83 in Experiment 3, and 77 in Experiment 4. There was no statistical interaction of harvest method with either harvest aid timing or fruiting branch removal. Hughs et al.
5
have shown total neps to increase by 30 count/g in spindle-harvested cotton for Gossypium barbadense, while Sui et al.
6
had total neps increasing with machine harvest, but it was not statistically significant.
Average impact of harvesting method on total neps in (a) Experiment 1, (b) Experiment 2, (c) Experiment 3, and (d) Experiment 4. Machine-harvested cotton refers to spindle-harvested cotton. The vertical lines represent the least significant difference of the harvest method at P = 0.05.
Differences in the number of the neps between machine- and hand-harvested cotton among years may be associated with changes in machine setup, operation, and climatic conditions at the time of harvest.4,10,28 No specific details regarding machine setup, operation, or climatic conditions at the time of harvest were collected.
To generate an average effect of the spindle machine harvest and to further assess for any potential interaction of harvest method and harvest aid timing, total neps data collected across all seasons was regressed with fiber linear density. Bange et al.
12
has previously related the change in total neps to fiber linear density caused by differences with harvest aid timing. A multiple regression of total neps to fiber linear density (R2 = 0.66) accounting for harvest method was highly significant (P < 0.001) (Figure 2). Only the constant was significantly different between harvest methods, indicating that there was no interaction of fiber linear density with harvest method to generate a different response. Quadratic regressions also provided better fit to the data than a linear regression. A curvilinear response was also found by Bange et al.
12
On average, spindle machine harvesting caused 53 (count/g) more total neps across the range of differences in fiber linear density (approximately 170–210 µg/m) (Figure 2).
The relationship of fiber linear density (ld) and total neps for hand (open symbols) and machine-harvested cotton (closed symbols). Regressions are fitted to the combined data of all four experiments.
The outcomes of this study are supported by previous investigations by Bange et al., 12 where they related the percentage of immature bolls at the time of harvest aid application to an increase in total neps following ginning, and subsequent lint cleaning passages. Bange et al. 12 showed there was only a significant statistical interaction between the timing of harvest aid treatment and the number of lint cleaning passages to substantially increase total neps in very early harvest aid treatments (>80% immature bolls), and with samples that had low fiber linear density (<158.5 µg/m), which were not present in this study. Lint cleaning is known to be more aggressive than machine spindle harvesting in generating neps,6,8 and this comparison assists interpretation of these outcomes.
Conclusion
This study demonstrated and quantified the effect of machine spindle harvesting on increasing neps in upland cotton compared with hand harvesting. Despite differences in fiber linear density resulting from differences in harvest aid timing, at the time of harvest the machine spindle harvest method did not combine with harvest aid timing to further increase neps (total, seed coat, and fiber). Multiple regression analysis was able to show that, on average, machine spindle harvesting contributed 53 more total neps (count/g) than hand-harvested cotton. More research needs to be undertaken assessing the impact of machine spindle harvesting on genotypes that vary considerably in other fiber quality attributes. Additional knowledge of machine harvesting effects in combination with crop management helps to identify reasons for higher neps in some cotton production regions, and assist the development of harvest management strategies to optimize fiber quality.
Footnotes
Funding
This work was supported in part by the Cotton Research and Development Corporation of Australia and the Cotton Catchment Communities Co-operative Research Centre.
Acknowledgements
Thanks to Jane Caton, Darin Hodgson, Rebecca Giles, Sue Miller, and Geni Kozdra for assistance in the field and with AFIS measurements, Dr Stuart Gordon and Mr Rene van der Sluijs for helpful discussions about the results, and Cotton Seed Distributors for provision of the planting seed.