Comparing cotton fiber quality from conventional and round module harvesting methods

Abstract

The introduction of the John Deere 7760 spindle harvester to the Australian cotton industry in 2008, with on board module building capacity producing round modules, has led to the rapid uptake of this technology by growers due to a reduction in labor requirements. There have, however, been anecdotal reports from cotton classing facilities and growers that the quality of cotton harvested by the John Deere 7760 is different and more variable compared to cotton harvested by the conventional spindle basket harvesting and separate module building method. The aim of this research was to compare the average fiber quality and the variability of quality between bales of cotton produced by these two harvesting methods. Four fields located in the southern and central cotton growing areas of Australia were planted with two popular upland varieties. Alternate groups of rows across each field were harvested using either the round module or conventional basket harvesting method, with harvested seed cotton ginned at the same gin. Fiber samples were assessed for quality attributes typically used to class and value cotton in Australia. There was no significant difference of average results between the two harvest methods for High Volume Instrument (HVI) determined upper half mean length and bundle strength. However, HVI micronaire was marginally yet significantly lower and HVI reflectance higher for fiber from the round module system, which was attributed to the round harvesters being able to harvest more fiber (including immature fiber from the top of plants) and less trash. The small difference in reflectance, however, did not translate into a practical difference, with there being no significant difference measured between the two harvesting systems for visually determined color and leaf grade. The normalized variability (% coefficient of variation) of fiber quality between bales was greater for the round module system, which was attributed to less blending during the sequential ginning of round modules in contrast to the vertical building of equivalent multiple (round module) layers of seed cotton in larger conventional modules that undergo more blending when fed longitudinally into the gin. While the round modules in this research were harvested and ginned according to industry standards and in sequence, less than favorable in-field conditions and out-of-sequence ginning would more than likely compound the variation in fiber quality between bales. The impact of these variables on fiber quality would have significant practical implications, which will require further investigation.

Keywords

fiber ginning harvesting cotton quality testing properties

Australian cotton has earned a reputation amongst international spinners for its high fiber quality, low contamination and good spinning ability, and is mainly used in spinning mills in South East Asia to produce fine to medium count combed ring and compact spinning yarns.¹ The cotton grown in Australia is mainly upland cotton (Gossypium hirsutum L.) of which 90% is grown under either partial or full irrigation, with the rest grown as dryland or rain-fed cotton. All cotton grown in Australia is mechanically harvested utilizing once-over harvesting, with the bulk of the crop harvested by selective-type harvesters that use rotating tapered, barbed spindles (spindle harvesters) to remove seed cotton from opened bolls into the machine.

Cotton harvesting represents the largest single cost item in cotton production and is the largest capital investment other than land.² The cost of cotton production in Australia is one of the highest in the world at almost three times the world average. High yields and high-quality cotton fiber ensure that the industry has remained competitive. The cost of production is a critical issue for a cotton grower and since harvesting on average contributes about 14% to the total cost of production, it is no surprise that there is a focus in Australia on making efficiency gains in this area.³

Traditionally seed cotton was harvested by conventional spindle harvesters with a basket system that would either dump the harvested seed cotton directly into the module builder (Figure 1(a)) or use boll buggies (a tractor-drawn bin) to transport the seed cotton from the harvester to the module builder, allowing the harvester to continue operating. In Australia, conventional modules containing compressed seed cotton for transport to the cotton gin are about 2.4 m wide, 12 m long and 3 m high and weigh 12,000–16,000 kg, producing about 24 (227 kg) bales of ginned fiber. Typically harvesters with basket systems require up to four pieces of support equipment (tractor-drawn boll buggies as well as module builders) along with workers to operate the equipment. Two (four-row) harvesters in operation with all this other equipment will generally require a crew of 8–10 workers and this represents a significant cost and safety risk.⁴

Figure 1.

Conventional basket harvester with separate module building operation (a), and simultaneous harvesting and module building operation by the John Deere 7760 harvester (b).

The release of alternative harvesters by Case IH (Racine, WI) and John Deere (JD; Moline, IL) with on board module building capacity offer significant opportunity to reduce the amount of equipment and the number of operators required for the harvesting operation. The Case IH 625 Module Express six-row harvester, which can operate at a speed of 6.4 km h⁻¹, produces half-size conventional modules 2.4 m wide, 4.6 m long and 2.4 m high and can weigh 2000–5500 kg, producing about 6–7.5 bales. The JD 7760 six-row harvester, which can operate at a speed of 6.8 km h⁻¹, has been described as a hybrid of a cotton harvester and an oversized round hay baler, produces round modules that are covered with an engineered polyethylene film that both protects the seed cotton and provides compressive force to maintain the module density. These modules can have a diameter of 2.44 m and a width of 2.39 m and, depending on moisture content, can weigh 2000–2600 kg, producing about four (227 kg) bales^3,5,6 (Figure 1(b)).

There are currently no Case IH 625 Module Express harvesters in Australia; however, since 2008, the uptake of the JD 7760 round module harvester in Australia has been rapid. It is estimated that in the 2010/2011 cotton season there were approximately 80 round module machines that harvested approximately 44% of the 4.2 million bale crop, while in the 2011/2012 season there were over 200 machines that harvested approximately 75% of the 5.4 million bale crop. This is the largest percentage of any crop harvested by these machines worldwide. While round module harvesters have a greater initial capital cost and consume expensive plastic wrapping, the Australian cotton industry has embraced these harvesters because they can harvest cotton continuously when conditions permit, which makes it very efficient, and dispenses with the requirement of sourcing reliable seasonal workers to undertake laborious module building.^2,5

Despite the advantages of the JD 7760, some concerns have been raised regarding soil compaction and the potential effect on yield of subsequent crops.⁷ Furthermore, there have been some suggestions and anecdotal reports from growers and classers that the quality of the cotton fiber harvested by the JD 7760 harvester is different and potentially inferior to the traditional basket harvesting and separate module building method. In particular, concerns have been raised in regards to differences or greater variability in the color and trash content of fiber harvested by the JD 7760. Some explanations for these hypothesized differences include that there is limited in-field blending with round module harvesting due to a greater number of round modules being produced per area harvested, which may be handled differently (i.e. not in sequence) at the gin. Similarly, the degree of seed cotton blending during module building and subsequent opening at the gin may be different for the two harvesting and module building techniques.

In addition, the JD 7760 harvester is a more powerful machine with more powerful doffers, and with different air-flow dynamics delivering a greater volume of air across the machine heads, compared to its non-module building counterpart. This allows the JD 7760 to harvest cotton with a higher moisture content, and therefore it can start earlier in the morning and harvest longer into the night when higher moisture levels (dew) are present. The typical recommendation is that the surface moisture content of seed cotton should not exceed 12% (generally measured via a handheld moisture meter) during harvesting.^4,6,8 Cotton that is too moist during harvest can be prone to degradation in the module, which can adversely affect quality, color and reflectance, and will also lead to elevated temperatures that may accelerate fiber degradation and cause module fires.⁹

Some growers and harvester operators have also speculated that although some trash can be dispersed in the accumulator basket prior to the delivery of the harvested cotton to the module building apparatus of the JD 7760, the traditional harvesters appear to disperse more trash out of the collecting basket. The different power, air-flow dynamics and direct module building mechanics of the JD 7760 may also interact adversely with seed cotton to negatively impact other quality attributes, including the shortening of fibers and the generation of entanglements (neps).

Figure 2 shows the variability in the appearance of cotton classing samples from bales of cotton (n > 20) that were produced from one gin run of six round modules. While it is recommended that modules, including large conventional modules, be ginned in a similar sequence to how they were produced in the field; practical logistics during harvesting, transport and storage, and the selection of modules for ginning, are all factors that may cause such variability. While such variability may be normal in many respects, including differences in quality and appearance attributes for bales from large conventional modules, unfounded reports from industry stakeholders suggest that the round module harvesting system results in more variable classing results and that the quality of cotton from the round module system is inferior or different to the conventional harvesting and module building technology.

Figure 2.

Classing samples (each approximately 100 g in weight, 120 mm wide and 220 mm long) in a classing facility conditioning tray. The majority of samples represent approximately 24 bales that were produced by a single gin run of six round modules. Note the variation in color of the samples.

Little research has been undertaken to rigorously compare the impacts on fiber quality of the round module harvesters with the conventional basket harvesters under similar harvesting conditions. A survey of fiber quality from bales produced by these different harvesting and module building technologies used in the USA in 2008 found that there were differences in quality (micronaire, strength, length and color) as measured by the High Volume Instrument (HVI).^10,11 However it was difficult to draw any definitive conclusions from that study as the modules used were not of the same variety and were not produced under similar growth and harvest conditions. In another study,¹² the quality of 2000 bales produced from two varieties grown on subplots, with rows of the plots harvested alternatively by a JD 7760 and a JD basket system harvester, were compared. Fiber quality data from HVI and Advanced Fiber Information System (AFIS) instruments indicated that the fibers from bales produced from round modules were stronger, longer and more uniform in length, with fewer short fiber and neps, compared to bales produced from the conventional harvesting and module construction method. This also resulted in rotor yarns that were stronger, with better elongation and toughness. Unfortunately, only minimal information was provided in the conference abstract published, which makes it impossible to draw any definitive conclusions from that study on why the differences in quality occur.

This study is a more in-depth version of that reported by van der Sluijs,¹³ and aims to specifically investigate the average quality and the consistency or variability of the quality of fiber harvested via the JD 7760 round module harvester compared with fiber harvested via the traditional basket harvesting and separate module building method. The research was undertaken using popular high yielding (>2000 kg ha⁻¹ fiber) commercial varieties common in the Australian system, and attempted to control for potentially confounding factors such as variety, location, moisture during harvesting and ginning. The results of these experiments will help assess and quantify the impact of harvesting method on fiber quality.

Materials and methods

Four experiments were conducted to compare the fiber quality of cotton harvested via both the conventional basket harvesting with separate module construction method, and the JD 7760 round module harvesting machine. Experiments were undertaken during the 2011/2012 growing season (planted in 2011; defoliated, harvested and ginned in 2012) in two cotton growing areas in New South Wales, Australia. Two experiments were conducted in fields at Boomi (28°44’S 149 °35’E) in the McIntyre Valley (central region), designated fields ‘A’ and ‘B’, and two experiments at Hillston (33°29’07’S 145 °31’58’E) in the Lachlan Valley (southern region), designated fields ‘C’ and ‘D’ (Figure 3).

Figure 3.

Map (shaded portion enlarged) showing the main cotton growing regions of Australia. Experimental fields A and B were at Boomi (•) and fields C and D were at Hillston (•).

A summary of respective field operations employed on each of the fields is presented in Table 1. The cotton varieties used for the experiments were Sicot 71BRF and Sicot 74BRF,^14,15 the two most popular upland varieties grown in Australia at the time that this work was undertaken. Fields A and B were planted on the 17 and 18 October 2011, respectively, while fields C and D were planted on 8 October 2011. All four fields were subjected to standard management practices for irrigated upland cotton in Australia.

Table 1.

Details of the location, variety used, size, planting date, harvest aid application dates, harvest dates and ginning dates for each field used to conduct experiments

Field	Location	Variety	Field size (ha)	Planting date	1st Harvest aid date	2nd Harvest aid date	Harvesting date	Ginning date
A	Boomi	Sicot 74BRF	50.5	17 Oct	20 Mar	12 Apr	17 & 18 Apr	8 May
B	Boomi	Sicot 71BRF	63	18 Oct	29 Mar	9 Apr	19 & 20 Apr	8 & 9 May
C	Hillston	Sicot 74BRF	42	8 Oct	15 Apr	11 May	22 & 23 May	16 Aug
D	Hillston	Sicot 74BRF	22	8 Oct	25 Apr	11 May	23 & 24 May	17 Aug

Fields A and B were first subjected to harvest aids by air with a mixture of leaf defoliant (0.1 L ha⁻¹ Dropp® liquid from Bayer Crop Science), boll opener (0.5 L ha⁻¹ Prep® from Bayer Crop Science) and defoliant aid spray (l L ha⁻¹ of D-C-Tron® from Caltex). They were again sprayed by air with a mixture of leaf defoliant (0.1 L ha⁻¹ Dropp®) and defoliant aid spray (1 L ha⁻¹ D-C-Tron®). Fields C and D were first treated by air with a mixture of leaf defoliant (0.18 L ha⁻¹ Dropp® Ultramax from Bayer Crop Science), a boll opener (0. 8 L ha⁻¹ Prep® from Bayer Crop Science) and 0.5 L ha⁻¹ Canopy® oil from Caltex. They were sprayed again by air with a mixture of leaf defoliant (0.12 L ha⁻¹ Dropp® Ultramax) and boll opener (2.8 L ha⁻¹ Prep®).

The harvesting of each field occurred over a two-day period. Harvesting took place during the day and seed cotton moisture was continually monitored via a handheld Delmhorst C-2000 cotton moisture meter with cup-like electrode (Delmhorst Instruments Co., Towaco, NJ) to ensure that harvested cotton did not have a surface moisture level greater than the recommended level of 12%. Groups of rows of cotton were alternately harvested by either the round or conventional module harvesting technique, with harvesting starting at one end of the field and sequentially occurring across each field. These groups of rows harvested by both methods, or ‘treatments’, each being the full length of the field (row length) were designated into blocks. Each block represented the total number of rows harvested by either method (multiple treatment pairs), and subsequently the groups of modules constructed for the area of the field that was required to build one conventional module; that is, the area of a field required to build one conventional module and multiple associated round modules was designated as a single block. This blocking (or module treatment pairs) factor was used during statistical testing as a way of taking into account any spatial variation across each field. The number of blocks was dependent on the field size, with fields A, B, C and D having 8, 11, 10 and 5 blocks for each field, respectively.

Grower-owned and operated harvesting machinery was employed to harvest fields at Boomi, while contractor-owned and operated harvesting machinery was used to harvest fields at Hillston. All machine harvesters employed during the harvesting of experimental fields were maintained and operated via normal industry practice and manufacturers’ recommendations. Harvesters were operated at a ground speed of 6.4 km h⁻¹. For the harvesting of fields A and B, a JD 7760 harvesting machine was used to harvest seed cotton from the first six rows, then skipped four rows, and then harvested the next six rows until the field was completed. One JD 9967 four-row basket harvester was used to harvest the remaining rows, with resulting seed cotton being dumped directly into a module builder (Figure 1(a)). For fields C and D, a JD 7760 harvesting machine harvested seed cotton from the first 12 rows then skipped 12 rows and harvested the next 12 rows until the field was completed. Two four-row basket harvesters were used to harvest the remaining cotton, a Case 2555 machine and a Case 2155 machine. Seed cotton harvested by the Case machines was dumped into tractor-drawn boll buggies that were then unloaded into the module builders.

For all fields, round modules were dropped in the field (Figure 4(a)) and picked up by a mast-type tractor-mounted implement that holds the module with the axis parallel to the tractor rear axle, and were then staged together in the sequence that they were harvested. The modules were placed to allow easy access for the equipment and trucks, on a smooth, even and firm compact surface that allows water to drain away. Round modules were staged in a ‘sausage’ (end to end) method with a gap between modules to facilitate water runoff (Figure 4(b)).

Figure 4.

The harvesting of an experimental field showing a deposited round module (a), and a harvested field showing staged round modules and a conventionally built module (b).

All modules from fields were ginned in the sequence that they were produced. Modules produced from fields A and B were ginned under standard commercial conditions at Brighann Ginning in Moree, NSW. This gin is a modern Lummus (Savannah, GA) super-high-capacity saw gin equipped with four 170 gin stands, producing 60 bales per hour. Fields C and D were ginned under standard commercial conditions at the Australian Cotton Ginning Company in Hillston, NSW. This gin is a modern Continental (Prattville, AL) super-high-capacity saw gin equipped with three 181 gin stands, producing 38 bales per hour. The percentage of the weight of usable fiber per the weight of un-ginned seed cotton (gin turnout) was calculated by the commercial ginning operators at a commercial scale using module and ginned bale weights. Table 2 summarizes the details of modules and ginned bales of fiber produced from each field.

Table 2.

Breakdown of type, number and total weight of modules, and number of bales of fiber produced and the gin turnout (% fiber per un-ginned seed cotton) for each field

Field	Module type	Number of modules	Total weight of modules (kg)	Number of 227 kg bales of ginned fiber	Gin turnout (%)
A	Conventional	8	106,860	197	43.2
A	Round	66	161,640	291	43.9
B	Conventional	11	141,420	237	40.0
B	Round	85	208,280	349	40.4
C	Conventional	10	129,480	240	41.9
C	Round	58	130,540	241	41.7
D	Conventional	5	65,720	126	43.6
D	Round	30	68,220	130	43.5

Classing samples were collected at the gin after bale formation. Two classing samples, from two opposite sides of each bale, were collected per bale. Fiber samples from each bale were subjected to manual visual classing and HVI (Uster Technologies Inc, Knoxville, TN) testing at Australian Classing Services (Wee Waa, NSW).^16,17 For visual classing, color grading describes the degree of reflectance and yellowness of samples with a 21-grade or Strict Middling cotton appearing brighter and less yellow than a 31-grade or Middling cotton. Leaf grade is a visual estimate of the amount of cotton plant leaf particles in fiber on a scale of 1 (low) to 7 (high).^18,19 HVI testing determined yellowness (+b), Rd (% reflectance), fiber upper half mean length (mm), bundle strength (g tex⁻¹) and micronaire (a combined measure of fiber fineness and maturity). The above-mentioned quality attributes (excluding HVI Rd and +b) are used by merchants in Australia to value and trade cotton bales. In addition, a limited sample set (recovered from field B) was subjected to testing via the AFIS (Uster Technologies Inc, Knoxville, TN) to measure total neps, fiber neps, seed coat neps (total neps = fiber neps + seed coat neps), trash, dust and percent visible foreign matter.²⁰ The same field B samples that were subjected to AFIS testing were also tested for the determination of maturity ratio via the use of the CSIRO Cottonscan instrument²¹ and Lord’s calculation.²²

Average fiber quality was calculated from multiple bales for each experimental unit (bales from groups of either conventional or round harvested rows designated to each block; average number of bales per experimental unit = 27). To gauge the consistency or variability of HVI fiber quality between bales for either harvesting method, the coefficient of variation (CV%) (CV% = (standard deviation/mean) × 100) was calculated for each experimental unit (group of multiple bales, n = 27). A single value (either bale average or bale CV%) for each experimental unit was used for statistical testing. To test for statistical differences between the two harvesting methods for visual classing and HVI results, restricted maximum likelihood (REML) modeling was employed using Genstat 14 (Lawes Agricultural Trust, IACR Rothamsted, UK), with the fixed model designated as the harvesting method, and the random model designated as the block nested under field (field/block). REML modeling enabled the collective assessment of data from all four fields as an unbalanced design.²³ Analysis of variance was employed to assess AFIS and maturity ratio data for field B.

Results and discussion

According to grower feedback, production output from all fields was normal with no adverse disease or other management issues identified. Varieties performed as expected for gin turnout, with minimal differences in gin turnout being detected between the harvesting method treatments (Table 2).²⁴

For fiber quality results for each field for each harvest method, fiber was visually classified on average as color 21 Strict Middling at either 2 or 3 leaf grade, which is equivalent or better than the Australian base grade of 31-3. Base grade refers to the grade of cotton that is used by cotton merchants as a basis for contracts, premiums and discounts. For average HVI results, upper half mean length ranged between 30 and 31 mm, bundle strength ranged from 28 to 33 g tex⁻¹ and micronaire ranged from 3.7 to 4.7 (Table 3). These results were all within the expected ranges for the varieties used.²⁴

Table 3.

Mean and standard deviation (s.d.) results values (n = blocks) for visual classing color and leaf grade, and High Volume Instrument (HVI) determined yellowness (+b), % reflectance (Rd), upper half mean length, bundle strength and micronaire for cotton fiber from all experimental fields harvested either by conventional or round module harvesting methods. For visual classing results, grades are also reported to the nearest values of practical significance in parenthesis. Results and mean values for restricted maximum likelihood (REML) statistical analysis are presented indicating the degree of significance between the two harvest treatments (n.s. = not significant)

Field	Module type	Block number	Color grade	s.d.	Leaf grade	s.d.	+b	s.d.	Rd %	s.d.	Length mm	s.d.	Strength g tex⁻¹	s.d.	Micronaire	s.d.
A	Conventional	8	21.0 (21)	0.0	2.1 (2)	0.3	7.7	0.1	83.2	0.3	31.0	0.2	33.2	0.3	4.68	0.06
A	Round		21.0 (21)	0.0	2.0 (2)	0.0	7.6	0.1	84.0	0.3	30.9	0.2	31.1	0.8	4.69	0.02
B	Conventional	11	22.9 (21)	4.0	2.5 (3)	0.5	7.6	0.2	82.6	0.7	30.6	0.2	30.1	1.3	4.46	0.08
B	Round		21.8 (21)	2.5	2.4 (2)	0.5	7.8	0.2	83.4	0.7	30.4	0.2	29.8	0.4	4.27	0.10
C	Conventional	10	23.0 (21)	4.2	3.0 (3)	0.0	6.8	0.3	82.0	0.4	30.3	0.3	28.3	0.3	3.73	0.08
C	Round		21.0 (21)	0.0	3.0 (3)	0.0	6.6	0.3	82.2	0.6	30.5	0.4	29.6	0.6	3.72	0.07
D	Conventional	5	21.0 (21)	0.0	3.0 (3)	0.0	6.6	0.2	82.5	0.6	30.5	0.2	30.2	0.4	3.92	0.08
D	Round		21.0 (21)	0.0	3.0 (3)	0.0	6.5	0.2	83.4	0.4	30.3	0.3	29.5	0.3	3.84	0.08
REML statistical analysis
	Conventional		22.2 (21)		2.6 (3)		7.2		82.6		30.6		30.4		4.20
	Round		21.2 (21)		2.6 (3)		7.1		83.2		30.6		30.1		4.13
	P-value		0.14 (n.s.)		0.09 (n.s.)		0.61 (n.s.)		P < 0.001		0.70 (n.s.)		0.34 (n.s.)		P < 0.01

There was no significant difference between the two harvesting methods for either manually determined or HVI average cotton fiber color (Table 3). Prior research has shown that fiber moisture at harvesting will influence cotton color.⁹ All harvesting was conducted on cotton fields with a surface moisture content not exceeding the industry recommendation, and as such it is clear that for the storage times employed for this work, the type of harvesting and module building technology had no detrimental effect on ginned cotton fiber color. However, the variability (CV%) of the color (+b) of cotton fiber between bales was significantly greater for the round module method (Table 4). This was also the case for the other HVI fiber quality attributes, and is likely due to less blending of seed cotton during round module building and opening at the gin. In contrast, when larger conventional modules are built, multiple layers of equivalent (round module) amounts of seed cotton are packed vertically. These modules are opened at the gin starting at one short side or ‘end’ and then fed in longitudinally, which enables the simultaneous and more comprehensive blending of the vertically packed seed cotton. While the round modules in this research were harvested and ginned according to industry standards and in sequence, variable conditions in the field and out-of-sequence ginning would likely compound the variation in fiber quality between bales. The impact of these variables on fiber quality would have significant practical implications, which should be further investigated.

Table 4.

Mean coefficient of variation (%) values (n = blocks) indicating the variability of fiber quality between bales for High Volume Instrument (HVI) determined yellowness (+b), % reflectance (Rd), upper half mean length, bundle strength and micronaire for cotton fiber from all experimental fields harvested either by conventional or round module harvesting methods. Results and mean values for restricted maximum likelihood (REML) statistical analysis are presented indicating the degree of significance between the two harvest treatments

Field	Module type	Block number	+b	Rd	Length	Strength	Micronaire
A	Conventional	8	0.99	0.38	0.59	1.35	1.04
A	Round		1.39	0.35	0.77	1.83	1.15
B	Conventional	11	1.03	0.38	0.65	1.20	1.18
B	Round		2.16	0.58	0.99	1.63	1.85
C	Conventional	10	2.06	0.57	0.69	1.20	0.97
C	Round		3.47	0.99	1.39	2.72	2.17
D	Conventional	5	1.30	0.65	0.75	1.92	0.99
D	Round		3.17	0.91	1.55	2.99	3.08
REML statistical analysis
	Conventional		1.37	0.49	0.68	1.42	1.09
	Round		2.52	0.71	1.16	2.27	2.00
	P-value		P < 0.001	0.001	P < 0.001	P < 0.001	P < 0.001

There was no statistical difference between harvesting methods for average visually determined leaf grade, although HVI reflectance results showed that round module cotton was slightly yet significantly higher in reflectance compared to conventional module cotton (Table 3), which indicates that round module cotton is lower in trash. Indeed, any notion that the conventional system allows more trash to disperse freely and thus benefitting trash levels in ginned fiber appears not to be the case. The round harvesters appear to be more efficient at harvesting cotton relative to unwanted trash, although practically this small gain may never be captured via visual classing. In addition, it must be acknowledged that while Case harvesters have spindle drums mounted on either side of the (row) head, JD harvesters have both spindle drums in-line on one side of the row. While there is no scientific evidence of significant differences between both systems, it is conceded that there may be some confounding of effects, since fields C and D used Case conventional harvesters. The consistent differences in reflectance between harvesting methods for all four fields suggests that there was minimal impact of any confounding effects.

For HVI fiber micronaire, there was a significant difference between the two harvesting methods, with cotton fiber from round modules generally having lower micronaire (Table 3). This difference was relatively small on average across experiments and similar to the tolerance error of the instrument of 0.1²⁵ and considering the micronaire range that cotton fiber typically can be without incurring any price discount penalties (the G5 micronaire range 3.5–4.9).²⁶ Nonetheless, the result was statistically valid, and may be due to the sensitivity that the air-flow micronaire instrument has to other properties of a cotton sample that can affect air resistance. This was demonstrated by Sui et al.,²⁷ who showed that cotton that had been machine harvested, pre-cleaned and then hand ginned had lower micronaire compared to the same cotton without pre-cleaning; these authors concluded that generally increased mechanical harvesting and processing of seed cotton resulted in reduced micronaire because of more immature fiber removal from the plant relative to hand harvesting, and because of differences in trash size and amount affecting the reading. Certainly HVI reflectance results indicated lower trash in round module cotton, and AFIS results from field B showed that dust, trash and percent visible foreign matter all tended to be lower (marginally significant P < 0.1) for the round module system (Table 5). The sensitivity of HVI reflectance and AFIS measurements to small changes in dust and foreign matter support the small yet significant difference in fiber micronaire between the two harvesting systems. Cheng and Cheng²⁸ also reported a weak yet significant relationship between AFIS percent visible foreign matter and micronaire. While the round module harvesters may be more effective at harvesting cotton fiber relative to unwanted dust and trash, this also includes, like the hand versus machine effect reported by Sui et al.,²⁷ a harvesting action that more effectively removes smaller less mature fruit locules from the top and outer branches of the plant; such immature fiber would lower micronaire. Indeed, cotton fiber maturity results (only available for field B) showed that round module cotton had a significantly lower fiber maturity ratio compared to conventional module cotton (Table 5).

Table 5.

Mean and standard deviation (s.d.) values for Advanced Fiber Information System (AFIS) neps, trash, dust and percent visible foreign matter (VFM), and for calculated maturity ratio (MR) for cotton fiber from bales from field B harvested by either the conventional or round module harvesting methods. Relevant probability values from analysis of variance (ANOVA) testing indicate the degree of significance between the two harvest treatments (n.s. = not significant)

Module type	Total neps g⁻¹	s.d.	Fiber neps g⁻¹	s.d.	Seed coat neps g⁻¹	s.d.	Trash count g⁻¹	s.d.	Dust count g⁻¹	s.d.	VFM %	s.d.	MR	s.d.
Conventional	276	11	254	11	22	2	45	13	244	61	0.93	0.28	0.86	0.02
Round	270	16	250	16	20	2	41	10	215	35	0.81	0.19	0.81	0.04
P-value	0.41 (n.s.)		0.50 (n.s.)		0.76 (n.s.)		0.08		0.06		0.06		<0.05

There was no significant difference between conventional and round module harvesting methods for HVI length and strength (Table 3). This was in contradiction to the findings of Thibodeaux et al.,¹² who reported that cotton from round modules was longer and stronger than cotton from conventional modules; full experimental details of this earlier study were not published so it is difficult to establish the reasons for differences. While the air-flow dynamics of the round module building harvesters are acknowledged to be different, the mechanical spindle harvesting action, conducted for the experiments reported herein at the same ground speed for both technologies, was not different enough to cause any measureable effect on fiber length. Certainly it was also not anticipated that a difference in fiber strength could be affected by the harvesting and module building method, because strength is primarily a varietal trait; any physical action or treatment that might negatively impact the cellulose make up of, and therefore the strength of, fibers (e.g. excessive heat) was not observed.

For AFIS neps for field B, there was no significant difference between the two harvesting methods for either total neps, fiber neps or seed coat neps (Table 5). The mechanical manipulation of cotton fiber is the cause of fiber entanglements, with the most intense manipulation, such as post gin stand saw lint cleaning, having the greatest impact. Machine harvesting is typically seen to be less intensive because fibers are not opened and disturbed to the same degree as they are during ginning. For example, Sui et al.²⁷ showed that compared to hand harvesting and manual seed removal (hand ginning), machine harvesting with some pre-cleaning of seed cotton that was hand ginned contributed to an increase of 96 neps g⁻¹ of fiber, while machine harvesting with full machine ginning and a standard post gin stand lint cleaning passage contributed to an increase of 284 neps g⁻¹. Similarly, Bange and Long²⁹ showed that machine harvesting with a spindle picker compared to hand harvesting, which was processed similarly, produced an average of 53 neps g⁻¹ and this did not vary with fiber micronaire. Therefore, the harvesting action of either of the assessed methods would not have been different enough to have contributed to a measurable difference in the level of neps in commercially ginned fiber, which infers that there is little chance that the round module harvesting method is responsible for an adverse neps problem in cotton fiber.

Conclusion

The average quality and the variability of quality between bales of cotton, produced from the traditional machine spindle harvesting and separate module building method, was compared with cotton produced via the all-in-one machine spindle harvester and on board round module builder. There was no significant difference between harvest methods for average HVI length and strength. Micronaire was overall lower, and HVI reflectance was higher, for round modules, which was attributed to the round harvesters being able to more efficiently harvest more fiber relative to unwanted trash. The difference in HVI reflectance was practically too small to be detected via visually determined color and leaf grade. While the round harvesters may be able to more effectively harvest more cotton inclusive of less mature unopened fruit, and thus lowering average micronaire, a good growth regulator and harvest aid management would minimize or negate any adverse effects. The variability of fiber quality between bales was higher for the round module system, which was attributed to a lesser blending effect when round modules are sequentially ginned. In comparison, large conventional modules are built with multiple (round module) amounts of vertically packed seed cotton, which undergo more blending as these modules are fed into the gin longitudinally. The practical implications of capturing this greater round module bale variability was beyond the scope of this research, although growers and ginners can now consider taking appropriate action to negate any potential negative effects. For example, the way in which the practice of module averaging^18,30 is undertaken, and the frequency of its use, might need to be reviewed in light of these results. Further, merchants and spinners may find value in bale fiber quality data that more accurately represents in-field variation. This should enable more precise bale management during storage and mill lay-down preparation, and allow more control in targeting specific, more consistent, end products.

Footnotes

Funding

This work was supported by the Australian Cotton Research and Development Corporation and CSIRO.

Acknowledgments

The authors acknowledge the generous cooperation of Greg Morris from Boomi and Malcolm Pritchard from Hillston for facilitating experiments on their farms. Thanks also goes to Brighann Ginning and the Australian Cotton Ginning Company for ginning experimental cotton, to Australian Classing Services for fiber quality testing and to Susan Miller and Glenda Howarth for technical assistance.

References

van der Sluijs MHJ and Johnson PD. Determination of the perceptions and needs of mills that purchase and process Australian cotton. CMSE, Geelong, 2011, p.128.

Willcutt MH, Buschermohle MJ, Barnes EM, et al. In field time in motion comparisons of conventional, John Deere 7760 and Case 625 module express cotton pickers. In: beltwide cotton conference, San Antonio,TX, 2009, pp.462–476.

Chaudhry

. Cost of production of raw cotton, Washington, DC: ICAC, 2010, pp. 121–121.

van der Sluijs

MHJ

. Best management practice for harvesting, Geelong: CMSE, 2013, pp. 24–24.

Martin SW and Valco TD. Economic comparisons of on-board module builder harvest methods. In: beltwide cotton conference, Nashville,TN, 2008.

Willcutt

Buschermohle

Huitink

. The spindle-type cotton harvester, Cary, NC: Cotton Incorporated, 2010, pp. 42–42.

Braunack

Price

Hodgson

. The effect of picker traffic on soil compaction: a preliminary survey. Aus Cotton Grower 2011; 32: 12–16.

Lalor

Willcutt

Curley

. Seed cotton storage and handling. In: Cotton ginners handbook, agricultural handbook number 503, Washington DC: United States Department of Agriculture, Agricultural Research Service, 1994, pp. 16–25.

Curley RG, Roberts B, Kerby T, et al. Effect of seed cotton moisture level and storage time on the quality of lint in stored modules. In: beltwide cotton conference, National Cotton Council, Dallas, TX, 4–8 January 1987, pp.504–510.

10.

Byler RK, Willcutt MH, Buschermohle MJ, et al. The effects of three module types on cotton ginning and fiber quality. In: ASABE annual international meeting, Reno, NV, 21–24 June 2009.

11.

Byler RK, Willcutt MH, Buschermohle MJ, et al. The effects of three module types on cotton ginning and fiber quality. In: beltwide cotton conference, New Orleans, LA, 2010, pp.645–656.

12.

Thibodeaux DP, Insley RE and Senter H. Studying the impact of harvesting with round bale modules on cotton quality. In: belwide cotton conference, San Antonio, TX, 2009, p.1453.

13.

van der Sluijs MHJ. Comparison of fiber quality harvested with round and conventional modules. In: beltwide cotton conferences, National Cotton Council, San Antonio, TX, January 2013, pp.722–729.

14.

Stiller

. Sicot 71BRF. Plant Varieties J 2008; 21: 194–197.

15.

Stiller

. Sicot 74BRF. Plant Varieties J 2010; 23: 143–146.

16.

ASTM International. Standard test method for measurement of physical properties of raw cotton by cotton classification instruments D5867-12. In: Annual book of ASTM standards – section seven textiles vol 07.02, West Conshohocken, PA: ASTM International, 2012, pp. 388–392.

17.

ASTM International. Standard practice for lighting cotton classing rooms for color grading D1684-07. In: Annual book of ASTM standards – section seven textiles vol 07.01, West Conshohocken, PA: ASTM International, 2012, pp. 396–399.

18.

Cotton Incorporated. The classification of cotton, Cary: Cotton Incorporated, 2013.

19.

USDA. Cotton classification understanding the data. Washington DC: United States Department of Agriculture, Agricultural Marketing Services, Cotton Program, 2005.

20.

ASTM International. Standard test method for neps in cotton fibers (AFIS-N instrument) D5866-05. Annual book of ASTM standards – section seven textiles vol 07.02, West Conshohocken, PA: ASTM International, 2012, pp. 385–387.

21.

Abbott

Higgerson

Long

. An instrument for determining the average fiber linear density (fineness) of cotton lint samples. Text Res J 2010; 80: 822–833.

22.

Montalvo

. Relationships between micronaire, fineness, and maturity. Part I – Fundamentals. J Cotton Sci 2005; 9: 81–88.

23.

Piepho

H-P

Richter

Spilke

. Statistical aspects of on-farm experimentation. Crop Pasture Sci 2011; 62: 721–735.

24.

CSD. Cotton variety guide 2012, Wee Waa: Cotton Seed Distributors, 2012.

25.

USDA. Guidelines for HVI testing. Tennessee: United States Department of Agriculture, Agricultural Marketing Services, Cotton Program, 2001.

26.

van der Sluijs MHJ. Classing, the Australian cotton production manual. Cotton Research and Development Corporation, 2013, pp.126–130.

27.

Sui

Thomasson

Byler

. Effect of machine-fiber interaction on cotton fiber quality and foreign-matter particle attachment to fiber. J Cotton Sci 2010; 14: 145–153.

28.

Cheng

KPS

Cheng

YSJ

. Relationship between appearance and physical properties of raw cotton. Text Res J 2003; 73: 245–250.

29.

Bange

Long

. Impact of harvest aid timing and machine spindle harvesting on neps in upland cotton. Text Res J 2013; 83: 651–658.

30.

Earnst DW. ‘Module averaging – a new perspective’, proceed. Beltwide Cotton Conferences, National Cotton Council, Nashville TN, 6–10 January 2003, pp.677–684.