Previous research has shown that flumioxazin, a herbicide being developed as a postemergence-directed spray (PDS) in cotton, has the potential to injure cotton less than 30 cm tall if the herbicide contacts green stem tissue by rain splash or misapplication. In response to this concern, five-leaf cotton plants with chlorophyllous stems and older cotton, 16-leaf cotton plants, with bark on the lower stem were treated with a PDS containing flumioxazin plus crop oil concentrate (COC) or nonionic surfactant (NIS). Stems of treated plants and untreated plants at the respective growth stage were cross-sectioned and then magnified and photographed using bright-field microscopy techniques. More visible injury consisting of necrosis and desiccation was evident in younger cotton. Also, there was a decrease in treated-stem diameter and an increase in visible injury with COC compared with NIS in younger cotton. The effects of plant growth stage and harvest time on absorption, translocation, and metabolism of 14C-flumioxazin in cotton were also investigated. Total 14C absorbed at 72 h after treatment (HAT) was 77, 76, and 94% of applied at 4-, 8-, and 12-leaf growth stages, respectively. Cotton at the 12-leaf stage absorbed more 14C within 48 HAT than was absorbed by four- or eight-leaf cotton by 72 HAT. A majority (31 to 57%) of applied 14C remained in the treated stem for all growth stages and harvest times. Treated cotton stems at all growth stages and harvest times contained higher concentrations (Bq g−1) of 14C than any other tissues. Flumioxazin metabolites made up less than 5% of the radioactivity found in the treated stem. Because of the undetectable levels of metabolites in other tissues when flumioxazin was applied PDS, flumioxazin was foliar applied to determine whether flumioxazin transported to the leaves may have been metabolized. In foliar-treated cotton, flumioxazin metabolites in the treated leaf of four-leaf cotton totaled 4% of the recovered 14C 72 HAT. Flumioxazin metabolites in the treated leaf of 12-leaf cotton totaled 35% of the recovered 14C 48 HAT. These data suggest that differential absorption, translocation, and metabolism at various growth stages, as well as the development of a bark layer, are the bases for differential tolerances of cotton to flumioxazin applied PDS.

Populations of wild radish were collected from two fields in the northern Western Australian wheatbelt, where typical herbicide-use patterns had been practiced for the previous 17 seasons within an intensive crop production program. The herbicide resistance status of these populations clearly established that there was multiple-herbicide resistance across many herbicides from at least four modes of action. One population exhibited multiple-herbicide resistance to the phytoene desaturase (PDS)–inhibiting herbicide diflufenican (3.0-fold), the auxin analog herbicide 2,4-D (2.2-fold), and the photosystem II–inhibiting herbicides metribuzin and atrazine. Another population was found to be multiply resistant to the acetolactate synthase–inhibiting herbicides, the PDS-inhibiting herbicide diflufenican (2.5-fold), and the auxin analog herbicide 2,4-D amine (2.4-fold). Therefore, each population has developed multiple-herbicide resistance across several modes of action. The multiple resistance status of these wild radish populations developed from conventional herbicide usage in intensive cropping rotations, indicating a dramatic challenge for the future control of wild radish.

Herbicides currently registered for use near water have been ineffective for control of perennial pepperweed. Previous research has demonstrated that mowing followed by an application of glyphosate at 3.33 kg ae ha−1 to resprouting tissue can enhance the control of perennial pepperweed. The objectives of this study were to determine the mechanism(s) responsible for the enhanced effectiveness of glyphosate in combination with mowing. Mowing plants altered the leaf area distribution within the canopy. In mowed areas, the majority of leaf area was in the basal third of the canopy, whereas the bulk of the leaf area was in the top third of the canopy in unmowed plots. This change in plant architecture affected the deposition pattern of the spray solution. Unmowed plants retained 49 to 98% and 42 to 83% of a dye solution within the middle and top thirds of the canopy at the Colusa and Woodland sites, respectively, with only 1.9 to 6.0% dye deposited on the basal third of the canopy at both sites. In contrast, mowed plants had 18 to 34% and 26 to 70% of the dye retained in the basal third of the canopy at the Colusa and Woodland sites, respectively. Greenhouse studies showed that 14C-glyphosate applied to basal leaves of mowed plants translocated significantly more to belowground tissue. Unmowed plants accumulated 0.37% of the applied 14C-glyphosate in belowground tissue 48 h after labeling. In contrast, mowed plants accumulated 6.7% 14C-glyphosate in the belowground tissue. In field studies, estimates of basipetal seasonal translocation rates using total nonstructural carbohydrate pools of roots indicate that mowing did not change the translocation rate. However, the delay in application timing to allow plants to resprout appeared to synchronize applications with maximal translocation of carbohydrates to belowground structures. We hypothesize that the change in the canopy structure of perennial pepperweed after mowing results in fewer aboveground sinks and greater deposition of herbicide to basal leaves where it can preferentially be translocated to the root system. Furthermore, the delay between mowing and resprouting synchronized maximal belowground translocation rates with herbicide application timing. These factors all appear to be involved in the observed enhanced control of perennial pepperweed when combining mowing and glyphosate.

Experiments were initiated to determine the amount of time required for postemergence herbicides to render yellow nutsedge physiologically noncompetitive. The rate of net CO2 assimilation (A N) was chosen as the response variable to describe competitiveness. Specifically, the time required after herbicide treatment for A N to drop to 50% of that of the untreated control (A N50) was determined. A N50 values for halosulfuron, imazapic, glyphosate, and MSMA were 1.6, 2.1, 3.2, and 3.3 d, respectively. An A N50 value was not calculated for bentazon because A N rapidly decreased below 50% but recovered to > 50% by 9 d after treatment (DAT). Stomatal conductance to water vapor (g s) declined similarly with A N over time for halosulfuron, imazapic, and glyphosate treatments. However, g s of MSMA-treated plants was near 95% of the untreated control, whereas A N declined to 35% 11 DAT. At 11 DAT, all aboveground biomass was removed, and plants were returned to the greenhouse, and regrowth was determined after an additional 14 d. Yellow nutsedge regrowth for halosulfuron, imazapic, glyphosate, and MSMA was below 5% of the untreated control and was not statistically different. However, regrowth from bentazon was 44% of the control. Therefore, bentazon was the least effective herbicide tested, whereas halosulfuron and imazapic were most effective for yellow nutsedge control.

A field microplot experiment was conducted in 1996 and 1997 to determine the influence of root-knot nematodes on intra- and interspecific interactions between chile pepper (chile) and spurred anoda and between chile and yellow or purple nutsedge (or both) using a substitution design. An additional objective was to determine the influence of London rocket, a winter annual and host plant for root-knot nematodes, on the inter- and intraspecific interactions between chile and spurred anoda. Twelve plant combinations were planted into paired 76-cm-diam microplots at a density of 24 plants per microplot each year. Each pair of microplots had one root-knot nematode–infested and one uninfested plot. One randomly selected plant pair or triplet from each plot was destructively sampled in June, July, August, and September each year. Data included leaf area, plant dry weights (leaf, stem, root or root plus rhizome, chile fruit, and nutsedge tuber), and nematode egg production from the belowground biomass of the different plant species within a 2,355-cm3 sampled soil volume. Chile hosted the highest population of root-knot nematodes, followed by spurred anoda, purple nutsedge, and yellow nutsedge. Few root-knot nematode eggs were recovered from London rocket before incorporation into the microplots each spring. Root-knot nematode populations were higher in 1997 and, as a result, more interactions between nematodes and plant combinations were observed for chile. Spurred anoda and root-knot nematodes reduced chile shoot and root weights to levels not significantly different from zero in 1997. Yellow and purple nutsedge shoots, except for those from the original tuber, were removed throughout the season, and these species interfered less with chile. Spurred anoda was not affected by interspecific interference. Few interactions were observed between the spurred anoda plant combinations and root-knot nematodes. In June 1996, low populations of root-knot nematodes (< 4,000 eggs per gram of root) stimulated spurred anoda growth, but higher populations in June 1997 (> 30,000 eggs per gram of root) reduced spurred anoda growth. Prior presence of London rocket had little consistent influence on spurred anoda or chile. Yellow and purple nutsedge growth variables were affected by interference from chile and the other nutsedge species. Tuber number and tuber weight were higher when plants were infected with root-knot nematodes, particularly early in the season. The enhanced tuber production may increase early-season interference from nutsedge species under production conditions. The results suggest that the presence of root-knot nematodes influences plant interference, but the effect is species specific. Annual plants are affected differently compared with perennial nutsedges, possibly because of the continuous association between the perennials and the parasite.

Laboratory studies were conducted to evaluate variation in germination response of eight annual bluegrass ecotypes (‘Augusta 4’, ‘Augusta 8’, ‘Augusta 14’, ‘Augusta 17’, ‘Auburn’, ‘Birmingham’, ‘Columbia’, and ‘Purchased’) to photoperiod, temperature, and fenarimol, a fungicide–herbicide used for preemergence annual bluegrass. Seed collected from greenhouse-grown plants and stored for > 2 mo were evaluated under 18 environments (three day and night temperatures by six day and night durations). There was a significant ecotype by environment interaction affecting annual bluegrass germination. High temperature markedly restricted germination, with only the Birmingham ecotype exceeding 20% germination at day and night temperatures of 39 and 29 C, respectively. Maximum germination of all ecotypes was observed at a day and night temperature of 19 and 10 C, respectively. Maximum germination for a specific photoperiod was not consistent across ecotypes; however, all ecotypes germinated to some degree in complete darkness, which indicates that maintaining a dense turf canopy to eliminate annual bluegrass germination may not be completely effective. Ecotypes did not differ with respect to root length response to fenarimol but did vary with respect to shoot length response. Purchased and Columbia shoot growth were the most tolerant to increasing fenarimol concentrations. This information will be used to develop improved management strategies for annual bluegrass.

Biennial wormwood is native to North America and has become an important weed problem in soybean and dry bean fields of North Dakota, South Dakota, and Minnesota in the United States and in the prairie provinces of Canada. Intersimple sequence repeat (ISSR) markers were used to study the genetic diversity among six biennial wormwood and one annual wormwood populations. Deoxyribonucleic acid (DNA) sequences from internal transcribed spacer (ITS1 and ITS2) regions of ribosomal DNA and morphological diversity among the biennial and annual wormwood populations were also studied. High levels of genetic diversity were evident with Nei's gene diversity statistic (h) = 0.40 for biennial wormwood and h = 0.36 for annual wormwood. Total diversity of six biennial wormwood populations was H T = 0.40, and 22% of this diversity was among populations (G ST = 0.22). Estimated gene flow among biennial wormwood populations was low (Nm = 0.9), and high levels of differentiation may be due in part to low levels of genetic exchange among biennial wormwood populations. Although biennial wormwood behaves more like an annual than a biennial, the ISSR, ITS, and morphological studies show that the two species are dissimilar.

Studies were conducted to calibrate and validate a mathematical model previously developed to predict common lambsquarters seedling emergence at different corn seedbed preparation times. The model was calibrated for different types of soil by adjusting the base temperature of common lambsquarters seedling emergence to the soil texture. A relationship was established with the sand mineral fraction of the soil and was integrated into the model. The calibrated model provided a good fit of the field data and was accurate in predicting cumulative weed emergence in different soil types. The validation was done using data collected independently at a site located 80 km from the original experimental area. There were no differences between observed and predicted values. The accuracy of the model is very satisfactory because the emergence of common lambsquarters populations was accurately predicted at the 95% probability level. This model is one of the first to take into consideration seedbed preparation time and soil texture. This common lambsquarters emergence model could be adapted to model other weed species whose emergence is limited by low spring temperature.

Common waterhemp, giant foxtail, and velvetleaf seed germination in response to temperature was studied with a two-way thermogradient plate. Seeds were maintained under dark and wet conditions at 4 C for 12 wk, and velvetleaf seeds were scarified before the experiments were conducted. The seeds were germinated at 25 different temperature treatments. Minimum and optimum temperatures for velvetleaf germination were approximately 8 and 24 C, respectively. Temperature alternation did not affect the germination of this species. The minimum germination temperature was 10 C for common waterhemp and 14 C for giant foxtail. The optimum germination of giant foxtail occurred at approximately 24 C, but common waterhemp optimum germination was variable depending on temperature alternation. Increased amplitude of the diurnal temperature alternation increased percent germination of these two species, and this was more evident at lower temperatures. In the case of common waterhemp, the temperature required to reach specific germination percentages was reduced by increasing the amplitude of the temperature alternation.

Leafy spurge is a herbaceous perennial weed that reproduces asexually through adventitious vegetative buds and sexually by seeds. Seeds can remain viable in the soil for up to 8 yr. The objectives of this research were to determine whether the seed coat and endosperm restrict germination and to assess whether afterripening treatments affect germinability. Germination of nonafterripened intact seeds was 30% after 28 d of incubation. Afterripening seeds for 12 wk under warm moist conditions provided a twofold stimulation to 58% germination after 21 d compared with the control and three other afterripening treatments. Afterripening under warm moist conditions for an additional 12 wk provided nearly complete germination within 7 to 21 d, but seeds harvested in year 2000 also responded in the same way to cool moist conditions. The removal of seed coat (dehulled) and removal of endosperm surrounding the embryonic axis in dehulled seeds led to rapid germination, approaching 80 to 90% in 5 d. Fructose did not enhance germination of dehulled seeds or embryonic axes. Thus, leafy spurge displays coat-imposed seed dormancy that can be overcome by afterripening under moist conditions.

Greenhouse studies were conducted to assess the intensity of smooth pigweed and common purslane aboveground interference (AI) and belowground interference (BI) with lettuce and to determine primary mechanisms of interference of each species as affected by P fertility rates. Lettuce was transplanted in mixtures with either smooth pigweed or common purslane according to four partitioning regimes: no interference, full interference, BI, and AI. Soil used was low in P for optimum lettuce yields, therefore P was added at rates of 0, 0.4, and 0.8 grams of P per liter of soil. Shoot and root biomass and plant height were measured for each species, as well as P tissue content. The data obtained indicated that smooth pigweed interfered with lettuce primarily through light interception by its taller canopy. A secondary mechanism of interference was the absorption of P from the soil through luxury consumption, increasing the P tissue content without enhancing smooth pigweed biomass accumulation. In contrast, common purslane competed aggressively with lettuce for P. Because the weed grew taller than lettuce, light interception was a secondary interference factor.

An induced mutation of the common wheat (2n = 6x = 42, AABBDD genomes) cultivar ‘Fidel’ has been shown to provide resistance to the imidazolinone class of herbicides. This class of herbicide gives broad-spectrum weed control including the weedy relative of wheat, jointed goatgrass (2n = 4x = 28, CCDD genomes). Because wheat and jointed goatgrass share a common genome, genes present on the D genome may transfer between the two species as a result of natural hybridization and selective pressures. Our objectives were to determine which genome of common wheat contained the herbicide resistance gene in the mutated Fidel and to genetically map its position. We investigated the chromosomal location of this gene using both durum (2n = 4x = 28, AABB genomes) and common wheat (6x) backgrounds. From crosses of durum wheat genotypes as the recurrent parent with mutated Fidel (cv. 9804, resistant), only BC1 plants containing chromosome 6D (inherited from cv. 9804) were resistant to applications of labeled rates of imazamox, an imidazolinone herbicide. No other D-genome chromosome was absolutely associated with herbicide resistance. To confirm this chromosomal location and genetically map the position of this gene, two populations of F3 families from the cross of cv. 9804 to the common wheat cultivars ‘Cashup’ and ‘Madsen’ were screened for reaction to imazamox, followed by genetic mapping with microsatellite markers. Two linked microsatellite markers were associated with the resistance trait, and one of them, Xgdm127, was located to chromosome 6D using aneuploid stocks, confirming the location of this gene on 6D. These results indicate that this resistance gene is in the genome that common wheat shares with jointed goatgrass. Therefore, imidazolinone-resistant wheat will need to be carefully managed to minimize the occurrence and spread of resistant jointed goatgrass, whether such plants arise because of hybridization with resistant common wheat or by independent mutation, a frequent occurrence with this herbicide class.

The fungus Pleospora papaveracea is a potential biocontrol agent for opium poppy. The objective of this study was to characterize the growth and production of propagules of P. papaveracea on various substrates and determine their infectivity on opium poppy. Pleospora papaveracea was grown on agar media containing wheat bran, corn cobs, soy fiber, cottonseed meal, rice flour, cornstarch, pectin, dextrin, or molasses, all with the addition of brewer's yeast (BY). Maximum radial growth of P. papaveracea occurred on molasses, soy fiber, and wheat bran media. Pleospora papaveracea produced chlamydospores on dextrin–BY and cornstarch–BY only. Pleospora papaveracea growth in liquid media with 1% (wt/v) dextrin, cornstarch, soy fiber, or wheat bran resulted in the production of greater than 106 colony-forming units (cfu) ml−1 within 3 to 5 d of incubation. Pleospora papaveracea produced less than 105 chlamydospores ml−1 after 10 d of incubation in wheat bran–BY and soy fiber–BY liquid media compared with the production of greater than 105 chlamydospores ml−1 after 5 d of incubation in dextrin–BY or cornstarch–BY liquid media. Fewer cfu were produced by P. papaveracea in 0.25% dextrin or 0.25 and 0.50% soy fiber liquid media than with 1 or 2% substrate. Greater than 107 chlamydospores g−1 dry weight and 108 cfu g−1 dry weight of P. papaveracea were produced in dextrin–BY liquid media in a commercial bench-top fermentor. After air drying biomass for 6 d, propagules of P. papaveracea remained infective on opium poppy. Mycelia and chlamydospores of P. papaveracea grew and formed appressoria during the infection process. Air-dried biomass, when rehydrated in 0.001% Tween 20, caused necrosis within 48 h after application to detached opium poppy leaves. At least 94% of the propagules from air-dried biomass that germinated and infected detached opium poppy leaves were of mycelial origin.

The effects of the fungal protein Nep1 and Pseudomonas syringae pv. tagetis (Pst) applied separately or in combination on Canada thistle, common ragweed, and common dandelion were examined in growth chamber experiments. Experiments examined five treatments: (1) untreated control, (2) Silwet L-77 (0.3%, v/v) control, (3) Nep1 (5 μg ml−1) plus Silwet L-77 (0.3%, v/v), (4) Pst (109 colony-forming units [cfu] ml−1) plus Silwet L-77 (0.3%, v/v), and (5) Pst (109 cfu ml−1) and Nep1 (5 μg ml−1) plus Silwet L-77 (0.3%, v/v). Foliar treatments were applied at 28, 26, and 21 d after planting for Canada thistle, common dandelion, and common ragweed, respectively. For all three species, foliar application of Nep1 alone or in combination with Pst caused rapid desiccation and necrosis of leaves, with the greatest effect on recent, fully expanded (RFE) leaves. Within 4 to 8 h after treatment (HAT), 60 to 80% of RFE leaves of all three species were necrotic. Measured 72 HAT, Pst populations in Canada thistle leaves treated with Nep1 plus Pst were approximately 105 cfu cm−2 compared with 107 cfu cm−2 for leaves treated with Pst alone. Measured 2 wk after treatment, foliar application of Nep1 reduced shoot dry weight of the three weeds by 30 to 41%. Treatment with Pst reduced shoot growth of common ragweed, Canada thistle, and common dandelion by 82, 31, and 41%, respectively. The large suppression of common ragweed shoot growth caused by Pst treatment was associated with a high percentage (60%) of leaf area exhibiting chlorosis. Treatment with Pst plus Nep1 did not result in significant decreases in shoot dry weight for Canada thistle and common dandelion compared with either treatment alone. For common ragweed, shoot growth reduction caused by applying Pst and Nep1 together was not greater than that caused by Pst alone.

One hundred and thirty-three fungal isolates, pathogenic to green foxtail, were evaluated for weed control potential under controlled conditions. To determine weed control efficacy, these pathogens were applied as spore or mycelial suspensions at approximately 105 propagules ml−1 to green foxtail at the three-leaf stage. One week after inoculation, most isolates caused only minor injury to the plants, but 15 isolates caused 50 to 100% disease. Among the most efficacious isolates, only those of Pyricularia setariae exhibited strong host specificity to the target weed, revealing no significant pathogenicity on 28 other plant species tested, including many important crops such as wheat, barley, and oat. On green foxtail leaves, conidia of this fungus germinated readily at 14, 20, and 26 C, but the process of germination and appressorial formation was more rapid at the higher temperatures. The fungus applied at the concentration of 105 spores ml−1 reduced weed fresh weight by 34% 7 d after the treatment when compared with controls, whereas a concentration of 107 spores ml−1 reduced fresh weight by 87%. This efficacy was comparable with that of the herbicide sethoxydim. When applied to the weed at the one- to four-leaf stages, the fungus reduced green foxtail fresh weight by more than 80%. Efficacy was slightly lower on plants at the five-leaf stage or older. On the green foxtail biotype resistant to the herbicide sethoxydim, P. setariae caused 80% fresh weight reduction compared with untreated controls, as opposed to 17% achieved with the herbicide. At 20 C, the fungus required a minimum of 6-h dew period to initiate infection, but a 10-h dew period was needed to cause severe damage to green foxtail.

WeedSOFT® is a decision support system that was developed to help farmers and consultants in Nebraska with the selection of optimal weed management strategies. WeedSOFT® evolved from HERB, a bioeconomic model for soybean that was developed in North Carolina. The program is composed of four independent modules, namely, ADVISOR, EnviroFX, MapVIEW, and WeedVIEW. ADVISOR helps the user select a treatment based on maximum yield or maximum net gain. EnviroFX and MapVIEW provide environmentally relevant herbicide information and county soil maps that indicate vulnerability to groundwater contamination. WeedVIEW is a visual library of color images and line drawings of 46 common weed species. Over 500 farmers and consultants in Nebraska and adjacent states use WeedSOFT®. As a result of the current regionalization effort, the user base is expected to increase rapidly during the next 2 or 3 yr. This article explains the algorithms implemented in the current version of WeedSOFT®.

Reliable mechanical weed control requires knowledge of the achievable levels of weed control and crop damage when using certain implements in specific conditions. Quantitative methods that use weed, crop, soil, and cultivator characteristics to predict weed control and crop damage need to be developed. To that end, the relative susceptibility of weeds and crop plants to mechanical weeding and the selective ability of cultivators need to be quantified separately. The method presented in this study uses measured plant anchorage forces to quantify crop and weed sensitivity to being uprooted by a weed harrow and predicts the relationship between weed and crop uprooting by mechanical weeding. Uprooting and anchorage force of young perennial ryegrass and garden cress plants were measured in laboratory harrowing experiments on sandy soil. A nonlinear equation was introduced to describe the relationship between weed uprooting and crop uprooting. The parameters representing the selective potential of the actual crop–weed condition (K pot) and the implement selective ability (K cult) did not depend on crop uprooting. The relationship between potential weed and crop uprooting that could theoretically be obtained by a perfectly selective implement (i.e., pulling each plant with equal force) was calculated from plant anchorage force distributions measured before harrowing. The observed uprooting percentages achieved by harrowing were lower than the potential uprooting percentages. With K cult accounting for imperfect weeder selective ability, prediction accuracy was satisfactory. Field validation is required to confirm whether this method improves comparison and prediction of weeding performance of different weeding implements in different crop–weed situations.

The development of competitive cropping systems could minimize the negative effects of wild oat competition on cereal grain yield, and in the process, help augment herbicide use. A 3-yr field experiment was conducted at Kalispell, MT, to investigate the effects of spring wheat seed size and seeding rate on wheat spike production, biomass, and grain yield under a range of wild oat densities. Wheat plant density, spikes, biomass, and yield all increased as seed size and seeding rates increased. Averaged across all other factors, the use of higher seeding rates and larger seed sizes improved yields by 12 and 18%, respectively. Accordingly, grain yield was more highly correlated with seed size than with seeding rate effects. However, the combined use of both tactics resulted in a more competitive cropping system, improving grain yields by 30%. Seeding rate effects were related to spike production, whereas seed size effects were related to biomass production. As such, plants derived from large seed appear to have greater vigor and are able to acquire a larger share of plant growth factors relative to plants derived from small seed.

Competition between wet-seeded rice and barnyardgrass under two distinct environments was analyzed using a two-parameter response–surface model at the International Rice Research Institute in the Philippines. The findings showed that this model could predict crop yield loss due to weed competition over a wide range of crop and weed densities. The low-tillering, new plant–type cultivar was a weaker competitor and had a higher yield loss than high-tillering cultivar ‘IR72’ and a hybrid. Increasing the crop density could reduce yield loss due to weed competition. This effect was greater for the new plant type than for IR72 and the hybrid when barnyardgrass density was low. In contrast, this effect was less for the new plant type than for IR72 and the hybrid when the weed density was high. Competitiveness of the three rice cultivars was also affected by season. Crop yield loss was higher in the wet season than in the dry season.

Under certain conditions, application of glyphosate to glyphosate-resistant (GR) cotton can lead to fruit shedding and yield reductions. Field studies were conducted at the Texas Agricultural Experiment Station using GR cotton, cv. ‘DeltaPine 5690RR’, to determine if application method and timing affect cotton fruit retention. Glyphosate at 1.12 kg ai ha−1 was precisely postdirected (PD), postdirected with 25% foliage coverage (PDFC), or applied over the top (OT) at the 8- or 18-leaf stage after an initial topical application of 1.12 kg ha−1 glyphosate at the four-leaf stage. In one of the years of this study, 8 PD, 18 PDFC, and 18 OT reduced yield. In 1999 and 2000, 8 PDFC and 8 OT applications of glyphosate caused yield loss, mainly due to lower mean boll weight. Glyphosate applied topically at the eight-leaf stage also affected the Position 1 boll retention throughout the plant in both years. Glyphosate contact with leaves and stems should be avoided when applying glyphosate after the four-leaf stage to prevent possible yield loss.