Introduction
Vegetable production in Georgia contributes $1.2 billion to the farm gate value (Stubbs Reference Stubbs2020). Cole crops and leafy greens contribute more than $119 million annually to the Georgia farm gate value and are grown on 37,160 ha (Stubbs Reference Stubbs2020). Georgia ranks first in collard production and fourteenth in broccoli production in the United States (USDA-NASS 2022a, 2022b). As production of these crops has increased across the Southeast, the method of planting has shifted from being seeded to mostly transplanted; approximately 80% of broccoli and 70% of collard plants are currently transplanted (Coolong et al. Reference Coolong, Sparks and Dutta2016, Reference Coolong, Kelley, MacDonald and Adams2017). The shift toward transplanting is due to significantly reduced germination associated with high fall soil temperatures (Coolong et al. Reference Coolong, Sparks and Dutta2016). Transplanting also improves uniformity and shortens the production season thus facilitating a greater potential for higher market prices (Orzolek Reference Orzolek1991).
In addition to transplanting, beginning the season with a weed-free field is crucial for profitability. Deemed the critical weed-free period, beginning at planting and continuing for several weeks, this period is when young plants are most susceptible to competition and interference from weeds, resulting in lower yield (Zimdahl Reference Zimdahl2018). Shadbolt and Holm (Reference Shadbolt and Holm1956) reported that even a low density of weeds, 15% of a normal stand, was sufficient to reduce yield in carrot, onion, and beet crops by 30%, 25%, and 22%, respectively, when weeds were allowed to grow alongside the crop for 4 wk after planting. They also determined that managing weeds from planting through 4 wk after planting was sufficient to mitigate yield losses from weed competition. Additionally, a study conducted on cabbage and tomato transplants found the critical weed-free period to be from planting through 3 to 5 wk after planting, respectively (Weaver Reference Weaver1984). By planting into weed-free fields, vegetable crops can establish and grow competitively. However, when weeds are present at planting, it is more difficult to manage these weeds in-crop, often resulting in reduced yields and quality.
Tillage, glyphosate, and paraquat are effective tools available to help eliminate weeds prior to planting cole crops in Georgia (Coolong et al. Reference Coolong, Kelley, MacDonald and Adams2017; Culpepper and Randell Reference Culpepper and Randell2022). Of these options, tillage to prepare a planting bed followed by an application of glyphosate to control weeds emerging between land preparation and planting are the standard choice to ensure fields are free of weeds at planting. Growers apply glyphosate because of its broad-spectrum activity on many of the common and troublesome weeds including yellow and purple nutsedge (Cyperus esculentus L. and C. rotundus L., respectively), wild radish (Raphanus raphanistrum L.), pink purslane (Portulaca pilosa L.), and cutleaf evening primrose (Oenothera laciniata Hill) (Culpepper et al. Reference Culpepper, Carlson and York2005; Culpepper and Randell Reference Culpepper and Randell2022; Van Wychen Reference Van Wychen2019; Webster et al. Reference Webster, Grey, Davis and Culpepper2008) and because, at least historically, the herbicide was thought to pose no threat to the succeeding crop. In fact, glyphosate’s 47-d soil half-life typically does not offer residual activity because of its propensity to bind to soil (Koc =24,000 mL g−1; Kd = 324 to 600 mL g−1) (Shaner et al. Reference Shaner, Jachetta, Senseman, Burke, Hanson, Jugulam, Tan, Reynolds, Strek, McAllister, Green, Glenn, Turner and Pawlak2014).
However, literature nearly three decades old explains that residual herbicidal activity of glyphosate can occur, although it is quite rare (Cornish Reference Cornish1992; Eberbach and Douglas Reference Eberbach and Douglas1983; Salazar and Appleby Reference Salazar and Appleby1982). Additionally, recent research has confirmed the residual activity of glyphosate can damage fruiting vegetable and cucurbit crops with subsequent yield losses on sandy soils with low organic matter when crops are transplanted. Goodman et al. (Reference Goodman, Randell, Hand, Vance and Culpepper2019) reported watermelons transplanted into bareground systems were injured 30% with a 16% yield loss when 2.5 kg ha−1 glyphosate was applied 1 d before planting. That study also reported cucumber injury of 14% to 52% when glyphosate was applied at 2.5 or 3.7 kg ha−1 at 1, 4, or 7 d prior to transplanting; yield loss was again prevalent with most treatments. Randell et al. (Reference Randell, Vance, Hand, Wright and Culpepper2022) found similar results in squash with 11% to 56% injury when glyphosate was applied at 1.5, 3.1, or 4.6 kg ha−1 1, 4, and 7 d prior to transplanting with yield loss ranging from 21% to 41%. Studies reported by Goodman et al. (Reference Goodman, Randell, Hand, Vance and Culpepper2019) and Randell et al. (Reference Randell, Vance, Hand, Wright and Culpepper2022) were conducted on soils with sand content of 88% to 90% with 0.5% to 0.9% organic matter, which likely influenced glyphosate residual activity. Other researchers have also noted that glyphosate is not readily adsorbed by sandy soils, thus soils high in sand content have higher potential for glyphosate to remain in the soil solution where it can be taken up by plants (Salazar and Appleby Reference Salazar and Appleby1982; Sprankle et al. Reference Sprankle, Meggitt and Penner1975).
With recent research confirming that the residual activity of glyphosate can damage fruiting vegetable and cucurbit crops on sandy soils with low levels of organic matter, research is needed to determine the vulnerability of transplanted broccoli and collard to glyphosate applied preplant in similar production systems and soils. Additionally, if glyphosate applied preplant poses a risk to broccoli and collard, practical mitigation measures must be demonstrated.
Materials and Methods
Site Selection and Trial Establishment
Field studies were initiated October 24, 2019, and October 12, 2020, at the University of Georgia Ponder Research Farm near Ty Ty, Georgia (31.507°N, 83.657°W) to evaluate broccoli and collard responses to glyphosate applied preplant. Soil at this location was a Tifton loamy sand consisting of 92% sand, 6% silt, 2% clay, and 0.58% organic matter, pH 6.5; typical for the production region (Coolong et al. Reference Coolong, Kelley, MacDonald and Adams2017). This experiment consisted of a three-factor factorial treatment arrangement in a randomized complete block design with four replications. The first factor, glyphosate rate, consisted of three levels: 0 kg ae ha−1, 2.5 kg ha−1, and 5.0 kg ha−1 (Roundup PowerMAX® II; Bayer CropScience LP, St. Louis, MO). The 2.5 kg ha−1 rate is the highest rate allowed in cole crops and is suggested for preplant control of nutsedge species (Anonymous 2020). The 5 kg ha−1 rate was included to evaluate crop safety at a 2× rate. The second factor in the treatment factorial consisted of two levels of tillage: roto-tilling soil to a depth of 5 cm after glyphosate application (prior to planting) or no tillage. The third factor also consisted of two levels of irrigation: sprinkler irrigation of 0.6 cm after glyphosate application (before planting) or no irrigation. Trials were maintained weed-free by selecting a site with low weed populations coupled with a preplant application of trifluralin (Treflan 4L; Loveland Products Inc., Greely, CO) at 35 g ai ha−1 and oxyfluorfen (Goal® 2XL; Corteva Agriscience, Indianapolis, IN) at 105 g ai ha−1. Irrigation was implemented following this application, prior to planting. Weed escapes were removed by hand prior to reaching 4 cm in height or diameter.
The land was conventionally prepared using a disk harrow (International Harvester, Chicago, IL), followed by ripping and bedding (Kelley Manufacturing Co., Tifton, GA), and rototilling (Maletti, Modena, Italy), leaving a smooth, flat surface for planting prior to trial initiation. After the study area was prepared, glyphosate treatments requiring irrigation were applied. Three hours later, 0.6 cm of sprinkler irrigation was implemented across the entire study area. Overhead sprinkler irrigation was applied at a rate of 0.25 cm per 10 min. One to two hours after irrigation, the remaining glyphosate treatments were applied. Tillage commenced 3 h later. Tillage treatments were implemented by identifying all plots requiring tillage, driving a tractor equipped with a rototiller down each row, and lowering the rototiller in the alley prior to each identified plot, lightly tilling the plot, then raising the rototiller in the alley at the end of the plot. The rototiller was allowed to turn for several seconds following each plot to remove any treated soil prior to continuing. All glyphosate treatments were applied with a CO2-pressurized backpack sprayer at 276 kPa and equipped with TTI 110015 nozzles (TeeJet Technologies, Glendale Heights, IL) calibrated to deliver 140 L ha−1.
Eight hours after treatment initiation, transplant holes were made using a tractor-mounted twin-row transplant hole-punch wheel (Kennco Manufacturing, Inc., Ruskin, FL) with broccoli (cv. ‘Emerald Crown’) and collard (cv. ‘Top Bunch’) immediately transplanted by hand, a standard production practice in this region (Coolong et al. Reference Coolong, Sparks and Dutta2016, Reference Coolong, Kelley, MacDonald and Adams2017). Plot dimensions were 1.8 m wide and 7.6 m long with one row of broccoli and one row of collard planted using a between-row spacing of 38 cm and an in-row spacing of 31 cm. Broccoli and collard fertility, irrigation, and insect and disease management followed University of Georgia recommendations (Coolong et al. Reference Coolong, Sparks and Dutta2016, Reference Coolong, Kelley, MacDonald and Adams2017).
Data Collection and Analysis
Visible estimates of crop injury were obtained weekly beginning 1 wk after planting (WAP) and continuing through harvest, with the greatest level of injury recorded 4 WAP. Injury was evaluated on a scale of 0% to 100%, with 0% being no injury (chlorosis, necrosis, stunting) and 100% being crop death. Plant growth was measured by recording diameter, at the widest point, of 10 plants per plot weekly beginning 2 WAP and continuing until harvest for each crop. Fresh weight and root biomass for each crop was collected 4 WAP from seven plants per plot. Plants were lifted from the soil after irrigation with a shovel ensuring roots were not lost during the root and shoot harvesting process. The root ball was placed into a paper bag and taken to the greenhouse for 4 wk of drying to facilitate soil separation from the roots naturally, followed by weighing. Aboveground shoots were placed in their own bag and weighed immediately. Yield data were collected at plant maturity from 10 plants per plot, with broccoli harvested up to 15 times over 7 wk and collard three times over 3 mo. For each plot at each harvest, broccoli marketable heads and collard marketable leaves were counted and weighed (USDA-AMS 2008a, 2008b)
Maximum injury, plant diameter, shoot and root biomass, broccoli head numbers and weights, and collard leaf numbers and weights were determined relative to the control plots having the same irrigation and tillage treatments but no glyphosate. Data were analyzed using the MIXED procedure in JMP Pro 15 software (SAS Institute, Cary, NC). Replication was nested within year and was treated as a random effect, while glyphosate rate, irrigation, and tillage were fixed effects. The two-way and three-way interactions between variables were also evaluated. An ANOVA was conducted, and means were separated using Tukey’s honestly significant difference test with a significance level of α = 0.05.
Results and Discussion
Broccoli
Visible injury estimates were influenced by glyphosate rate, tillage, and irrigation at 4 WAP (Table 1). Glyphosate at 2.5 and 5.0 kg ha−1 caused 43% and 79% injury, respectively, when neither tillage nor irrigation followed application (Table 2). The level of injury observed was severe and was likely influenced by the transplant wheel hole-punch method, which is standard for current production in the region. When punching the hole, soil from the surface is pushed downward, placing herbicide-treated soil throughout the plant hole, thereby increasing direct contact of the herbicide with the moist transplant root ball. When implementing irrigation as a method to mitigate the residual activity of glyphosate, visible injury was reduced 28% to 48%; however, injury observed remained unacceptable at 15% with the low rate of glyphosate and 31% with the higher rate (Table 2). Tillage was more effective than irrigation in mitigating injury, with less than 7% injury observed with either rate of glyphosate if tillage was implemented. Tillage plus irrigation was no more effective than tillage alone. These results reflect those observed by Randell et al. (Reference Randell, Vance, Hand, Wright and Culpepper2022), when tillage following glyphosate applications of 3.1 and 6.2 kg ha−1 was sufficient to eliminate injury to squash, while irrigation was less beneficial. Additionally, Cornish (Reference Cornish1992) suggested that mixing glyphosate with soil would mitigate its effects on sensitive crops due to increased adsorption.
Table 2. Broccoli injury, plant diameter, and marketable head weights as influenced by glyphosate rate, tillage, and irrigation.a, b, c
a Injury, plant diameter, shoot biomass, and root biomass measurements taken 4 wk after planting when maximum injury was observed. Broccoli marketable heads harvested up to 15 times.
b Data are combined over 2019 and 2020. Means were separated using Tukey’s honestly significant difference test with a significance level of α = 0.05. Means followed by the same letter within a column are not significantly different.
c All variables except injury are presented as a percent of the nontreated control (no herbicide) for each irrigation and tillage treatment. Averages for each variable in the control are as follows: plant diameter, 31.0 cm; shoot biomass, 34.7 g; root biomass, 8.2 g, broccoli marketable head number, 9.8; and broccoli marketable head weight, 3.1 kg.
Plant diameters 4 WAP were similarly influenced by glyphosate rate, tillage, and irrigation (Table 1). Without mitigating damage with tillage or irrigation, glyphosate reduced plant diameter measurements by 33% to 56% with 2.5 to 5.0 kg ha−1 of glyphosate compared to the control plants. The addition of irrigation resulted in healthier plants, but plant diameters remained 12% to 16% smaller than that of the control. Tillage eliminated the negative impacts from glyphosate applied preplant, regardless of rate, with plant diameters being 102% to 103% of the control.
Aboveground fresh weight biomass was reduced 81% by both rates of glyphosate at 4 WAP (Table 2). Irrigation resulted in a 45% increase in biomass weight when applying glyphosate at 2.5 kg ha−1, but benefits were not significant for the higher rate. Tillage was once again more effective than irrigation, with biomass weights similar to that of the control when tilling after applying either rate of glyphosate. Root biomass was influenced less than shoot biomass; glyphosate at 5.0 kg ha−1 without tillage or irrigation was the only treatment to significantly reduce dry weight biomass (33%). Eberbach and Douglas (Reference Eberbach and Douglas1983) reported a 6.5% reduction in root growth of subterranean clover following an application of glyphosate at 2 µg g−1 made to soil with 82% sand when plants were grown in a greenhouse (equivalent to a 1× field application of glyphosate in this study).
Glyphosate at both 2.5 and 5.0 kg−1 reduced total head weights by 35% and 63%, respectively, when no mitigation was implemented (Table 2). Previous research indicated that glyphosate uptake through roots can reduce vegetable yields (Semidey and Almodóvar Reference Semidey and Almodóvar1987). The number of broccoli heads produced was reduced by glyphosate only at 5.0 kg ha−1 (56%), indicating that the impact from glyphosate, especially at the lower rate, was predominately through reducing head size (data not shown). Using irrigation, tillage, or irrigation plus tillage to mitigate the residual activity of glyphosate were all successful with head weights harvested from these treatments similar to that of the nontreated control. The ability of the broccoli to recover from injury levels observed with these treatments during early season is feasible, especially in production systems providing ample irrigation and fertilizer (Bhowmik and McGlew Reference Bhowmik and McGlew1986).
Collard
Collard visible injury estimates were influenced by glyphosate rate, tillage, and irrigation (Table 3). Collard transplants were more sensitive to glyphosate applied preplant than broccoli transplants, with injury ranging from 71% with glyphosate at 2.5 kg ha−1 to 93% with glyphosate at 5.0 kg ha−1 (Tables 2 and 4). Similar to the result observed in broccoli, irrigation did lessen injury observed in collard, but damage remained severe, ranging from 30% to 63% (Table 4). In contrast to broccoli, tillage did not reduce crop damage to a level below 10% for all treatments. Even with tillage, collard injury of 17% was observed with the high rate of glyphosate applied preplant.
Table 4. Collard injury, plant diameter, and shoot biomass influenced by glyphosate rate, tillage, and irrigation.a, b, c
a Injury, plant diameter and shoot biomass measurements taken 4 wk after planting when maximum injury was observed.
b Data are combined over 2019 and 2020. Means were separated using Tukey’s honestly significant difference test with a significance level of α = 0.05. Means followed by the same letter within a column are not significantly different.
c All variables except injury are presented as a percent of the nontreated control (no herbicide) for each irrigation and tillage treatment. Averages for each variable in the control are as follows: plant diameter, 28.1 cm and shoot biomass, 33.6 g.
By 4 WAP, plant diameter was reduced 48% to 70% by glyphosate (Table 4). Similar to injury, irrigation alone did not mitigate the impact of either glyphosate rate. Tillage was again more effective than irrigation, but a 13% reduction in plant diameter remained when the high rate of glyphosate was followed by tillage. In fact, only tillage or tillage plus irrigation following the low rate of glyphosate produced growth diameter parameters similar to that of the control.
Aboveground fresh weight of collard was reduced 83% to 95% at 4 WAP (Table 4). When irrigation was implemented, shoot biomass was still reduced 49% to 79%; tillage more effectively mitigated glyphosate damage with biomass levels similar to that of the control. Collard root biomass was influenced by glyphosate rate and tillage (Table 3). When combined over irrigation options, glyphosate reduced root biomass 25% to 34%, whereas implementing tillage following either rate of glyphosate alleviated root biomass losses (Table 5).
a Root biomass measurement was taken 4 wk after planting when maximum injury was observed. Collard marketable leaves harvested three times.
b Data are combined over 2019 and 2020 and averaged over irrigation treatment. Means were separated using Tukey’s honestly significant difference test with a significance level of α = 0.05. Means followed by the same letter within a column are not significantly different.
c All variables are presented as a percent of the nontreated control (no herbicide) for each irrigation and tillage treatment. Averages for each variable in the control are as follows: root biomass, 4.7 g; shoot biomass, 34.7 g; root biomass, 8.2 g, collard marketable leaf number, 249; and collard marketable leaf weight, 4.4 kg.
Collard leaf number harvested and relative weights were influenced only by glyphosate rate and tillage, with leaf number reduced by 46% (data not shown) and weights reduced by 50% when glyphosate was applied at 2.5 kg ha−1 (Table 5). When glyphosate was applied at 5.0 kg ha−1, leaf number and weight were reduced by 70% to 71%. Leaf number and weight remained 26% and 31% lower, respectively, than the control even when implementing tillage after applying the higher rate of glyphosate. Leaf number and weight response by collard indicate crop safety when tillage follows a glyphosate application of only 2.5 kg ha−1.
In conclusion, these results confirm that care must be taken when applying glyphosate prior to transplanting broccoli and collard. There is evidence that a pretransplant glyphosate application can result in significant injury, plant growth reductions, and yield loss in these crops. The residual damage of glyphosate to these crops is likely influenced by soil type, planting method, irrigation scheduling, herbicide rate, tillage between application and planting, and interval between application and planting. In this study, overhead irrigation effectively reduced the impact from the residual activity of glyphosate on these crops. However, irrigation of 0.6 cm was not adequate enough to alleviate glyphosate damage. Rainfall or irrigation of at least 1.3 cm and waiting 7 d to plant is recommended to mitigate injury if tillage is not implemented (Culpepper and Randell Reference Culpepper and Randell2022). Tillage to a depth of 5 cm was more effective than irrigation and can be used to mitigate injury when applying glyphosate at rates equal ≤2.5 kg ha−1 for these crops.
Acknowledgments
We thank Tim Richards for his technical support and the student workers for their hard work during the course of this study. We also thank the Georgia Fruit and Vegetable Growers Association for funding this research. No conflicts of interest have been declared.
Open access
