The Journal of Cotton Science 11:242–251 (2007) 242 http://journal.cotton.org, © The Cotton Foundation 2007 BREEDING AND GENETICES Variation in Marginal Bract Trichome Density in Upland Cotton F. M. Bourland* and J. M. Hornbeck ABSTRACT ike many other plant species, cotton (Gossypium Lhirsutum L.) plants have hair-like protrusions Elimination or reduction of trichomes around or trichomes on the surfaces of several plant parts the margins of bracts should improve the clean- (Bradow and Wartelle, 1998; Lee, 1984; Oosterhuis ing efficiency of cotton (Gossypium hirsutum L.) and Jernstedt, 1999). Most research on cotton lint. The objective of the study was to establish pubescence has focused on abaxial leaf and stem sampling protocols for measuring marginal bract trichomes. Cotton cultivars have been developed that trichome density by examining variation over exhibit leaf and stem pubescence levels ranging from fruiting positions of the cotton plant, sampling glabrous to very hairy or pilose. As summarized by time, cultivars, and environments. To determine Jenkins and Wilson (1996), increased cotton plant variation of bract and leaf trichome density, pubescence on leaves and stems may be associated plants of six contrasting cultivars in two environ- with increased resistance to some cotton insect pests, ments were evaluated. To determine variation in but increased susceptibility to others. The glabrous bract trichome density over time, bracts of three trait generally had neutral or positive effects on yield contrasting cultivars were sampled from three and fiber quality, while the pilose characteristic often canopy sites on three dates from two years. To had negative effects. In some studies, relationships determine variation in bract trichome density over with yield and fiber traits were different with specific cultivars and environments, bracts were evaluated genes affecting pubescence. Also, low pubescence for multiple cultivars at multiple sites over four levels on cotton have been associated with improved years. Bract trichome density tended to increase seedcotton cleaning efficiency and low foreign matter as leaf trichome density increased and to decline content in harvested lint, which resulted in higher leaf with canopy age. Significant interactions involving grades in ginned cotton (Novick et al., 1991). sympodial position indicated that bracts should be Pubescence on cotton bracts has received little sampled from the same position. The first-position attention. Bracts are modified leaves surrounding the is suggested. Interactions involving nodes became flower buds and bolls of the cotton plant. Morey et non-significant by dropping the highest and lowest al. (1976) found that bracts are a major contributor nodes. Bract trichome density declined with older to leaf trash in harvested cotton. This seems reason- canopy sites and later sample dates, but variation able, since bracts are in closer proximity than leaves among cultivars was relatively consistent over to the cotton fibers on the plant, and most leaves are locations and years. Although cultivars varied removed from the plant prior to harvest if defoliation significantly, none of the cultivars had glabrous is successful. Bract tissue has also been implicated bracts. A significant cultivar by location interaction as a causative agent in byssinosis, a lung disease of in one year became non-significant by dropping a cotton mill workers (Ayer, 1971). highly stressed location. These data indicate that Methods with the potential to reduce lint bract samples should be collected from full-sized, contamination by bract tissue have included the mid-canopy, first-position bolls soon after flower- development of cotton lines having smaller bracts ing ceases and that bract trichome density can be (Bowman and Jones, 1982, 1983; McDaniel, 1996; adequately characterized by sampling bracts at Milam et al., 1975), deciduous or caducous bracts one test site. (Muramoto et al., 1987), and withering bracts (Knight, 1952). Reducing bract size or lessening their persistence may have negative effects on the F. M. Bourland, Northeast Research & Extension Center, physiology of the plant, since the role of bracts University of Arkansas, P.O. Box 48, Keiser, AR 72351; J. M. relative to leaves increases under stress conditions, Hornbeck, Lon Mann Cotton Research Station, University of Arkansas, P.O. Box 789, Marianna, AR 72360 such as drought (Wullschelger et al., 1990; Zhao *Corresponding author: [email protected] and Oosterhuis, 2000). BOURLAND AND HORNBECK: Variation IN MARGINAL BRACT TRICHOME DENSITY 243 Among Upland cotton genotypes, both glabrous square centimeter. Each branch of stellate trichomes and pubescent genotypes, as well as glaborous and was counted as an individual trichome. Maximum pubescent stem genotypes, have trichomes subten- bract length was measured from base of the bract to ding from the margin of bracts (Fig. 1). Intuitively, the tip of the center tooth. these marginal bract trichomes might play an impor- Whole plant study. A study was established tant role in the cleaning efficiency of ginned cotton, in 1996 to survey the variation in bract trichome since they would likely increase the adherence of density over the surface of the plant. Six cultivars, bract tissue to cotton lint. Only one report on bract selected for their variation in leaf pubescence, trichomes was found, but it addressed trichomes on were planted at two contrasting locations. Plots adaxial bract surfaces rather than marginal bract were four rows by 12 m long on 1-m centers and trichomes (Dimitropoulou et al., 1980). arranged in a RCB design with two replications. The two locations were Fayetteville in the Boston Mountains of Northwest Arkansas and Keiser in Mississippi River Delta of Northeast Arkansas. Soils were a Captina silt loam (fine-silty, siliceous, active, mesic Typic Fragiudults) at Fayetteville and a Sharkey clay (very-fine, smectitic, thermic Chromic Epiaquerts) at Keiser. The six cultivars and their expected leaf pubescence (based on rating scale described by Bourland et al., 2003a) included Stoneville 474 (very hairy) (Stoneville Pedigreed Seed Co.; Memphis, TN), Paymaster H1330 (hairy), Paymaster H1244 (light hair) (Paymaster Cotton- seed; Stuttgart, AR), SureGrow 501 (light hair) Figure 1. Trichomes subtending from margin of cotton (Sure-Grow Seed, Inc.; Leland, MS), Paymaster bracts. HS200 (glabrous leaf) (Paymaster Cottonseed), and The objective of this study was to establish a Deltapine 50 (glabrous leaf) (Delta Pine and Land sampling protocol for measuring marginal bract Co.; Scott, MS). In addition to having glabrous trichome density by examining variation over fruit- leaves, HS200 also displays glabrous stems, while ing positions of the cotton plant, time, cultivars and the other cultivars have hairy stems. environments. In late August, after termination of flowering but prior to defoliation, leaf pubescence of plants was ex- Materials AND METHODS amined and six representative plants were taken from each plot and transported to the laboratory. Abaxial Three separate studies were conducted to address trichomes were counted on all leaves associated with the objective. A conventional cotton production first, second, and third sympodial positions on two system with furrow-irrigation was used in each test, plants per plot. To increase the probability of sampling except those designed as non-irrigated sites, and cul- bolls at every position, bracts from each retained boll tural practices were standard for each individual re- on all six plants per plot were sampled. Marginal gion. To facilitate counting of abaxial leaf trichomes bract trichomes were counted and bract length was and marginal bract trichomes, a hole-punch (6-mm determined for one bract per boll for each first, second, diameter) was made in an index card and placed and third sympodial position boll. Measurements over the plant tissue. Using a viewing microscope, were recorded by main-stem node number from the trichomes exposed through the hole were counted. base of the plant and by sympodial position from the Marginal bract trichome density was determined by main-stem. Data were analyzed by SAS using PROC counting marginal trichomes on two representative GLM (SAS Institute, Inc.; Cary, NC). marginal areas of the center tooth of each bract, Temporal bract trichome study. The effect of then converting to number per centimeter. Abaxial sampling time was evaluated in a study conducted at leaf trichome density was determined by counting the Lon Mann Cotton Research Station near Marian- trichomes for a representative abaxial, mid-vein area na, AR, on a Calloway silt loam soil (fine-silty, mixed, of each sampled leaf, then converting to number per active, thermic Fraglossudalfs). The test included JOURNAL OF COTTON SCIENCE, Volume 11, Issue 4, 2007 244 selected cultivars in the 2001 and 2002 Arkansas In 1999 and 2000, all cultivars in the tests were Cotton Variety Test at the Marianna irrigated location. sampled. In 2001 and 2002, only cultivars common Plots were two rows, 12 to 15 m long, on 1-m centers to the 2000 test were examined. Samples were taken and were arranged in RCB designs with four repli- from 10 random plants per plot in two replications. cations. Cultivars (and expected leaf pubescence) On each plant, a full-size, mid-canopy leaf was visu- examined included SureGrow 215BR (glabrous ally rated for pubescence (as described by Bourland leaf) (Delta Pine and Land Co.), PayMaster 1218BR et al., 2003a) by observing