Evaluation of Preemergence and Postemergence Herbicide Programs on Weed Control and Weed Seed Suppression in Mississippi Peanut ( Arachis hypogea

: Weed control and reducing weed seed deposition to the soil seedbank is a challenging issues for Mississippi peanut producers. Research was established during 2017 and 2018 at the Delta Research and Extension Center in Stoneville, Mississippi, to evaluate herbicide programs for weed control and reducing weed seed production in Mississippi peanut production. Treatments were combinations of acetochlor, clethodim, ﬂumioxazin, lactofen, paraquat, and S -metolachlor with their respective adjuvants if needed. Treatments were applied PRE, two to three weeks after emergence (EPOST), and / or four to ﬁve weeks after emergence (MPOST). All treatments included a PRE application followed by (fb) application of EPOST and / or MPOST application. Flumioxazin PRE fb lactofen plus clethodim MPOST provided greater than or equal to ( ≥ ) 88% control of barnyardgrass, hemp sesbania, Palmer amaranth, pitted morningglory, and prickly sida. Additionally, this treatment reduced total weed seed production 88% compared to the nontreated control. Flumioxazin PRE fb lactofen plus clethodim EPOST fb acetochlor MPOST provided similar weed control and peanut yield as ﬂumioxazin PRE fb lactofen plus clethodim MPOST. This treatment reduced total weed seed production 93%. Treatments containing PRE, EPOST, and MPOST herbicide applications provided the best season-long control of weeds and weed seed suppression in Mississippi peanut.


Introduction
Weed control is essential for maximizing peanut yield [1,2]. Competitive nature and allelopathic interactions are major characteristics of weeds that influence crop yield [1]. Research has suggested a relationship between crops and weeds resulting in peanut yield loss from weed interference [3][4][5]. The relationship between crops and weeds can be explained by the critical period for weed control (CPWC) [6]. The CPWC is the period of time during the crop growth cycle in which weeds must be controlled to prevent peanut yield loss ≥5% [2,6,7]. Studies have reported the first three to five weeks of the growing season without weed competition will allow maximum crop yields if environmental conditions are favorable [8]. Studies have reported peanuts kept weed free six weeks after planting had no yield loss when weeds emerged later [3].
To effectively control weeds, producers should develop integrated weed management practices that minimize environmental impacts without forfeiting profitability [9]. In commercial crop production, weed control is built on the use of herbicides [10]. Producers prefer to use a single-pass herbicide program to achieve season-long weed control, which reduces labor costs and time [11,12]. However, in the southeastern U.S., weed management relies heavily on preplant-incorporated (PPI)-, PRE-, and POST-applied herbicides [13]. Preplant-incorporated herbicides are assimilated into the soil prior to crop planting with mechanical incorporation [14]. Preemergence herbicides help to reduce yield loss from weed interference, allow time for crop establishment, and mitigate selection pressure to POST herbicides [15]. Postemergence herbicides are most effective when the application rate, application timing, and weed size at application are in compliance with the herbicide label [10].
The biology and competitive nature of weeds is critical for producers to understand when developing a weed management strategy. Producers that implement effective weed management strategies to control competitive weeds during the CPWC can potentially maximize crop yield [3,6,7]. Using multiple herbicide MOA can mitigate potential for crop yield loss and also reduce the potential for herbicide-resistant weeds to evolve [10].
To overcome yearly weed infestations, control of emerged seedlings and depletion of the soil seedbank is essential [16]. A soil seedbank is a reserve for viable seeds contained in the soil profile or soil surface in a given area [17]. Mature weeds will release seeds to the soil surface, many of which accumulate in the soil seedbank [16]. However, if weed seed production is prevented, weed soil seedbanks can be depleted [16]. In agricultural fields, the soil seedbank provides long-term persistence of weed seed that survive less than 1 year [16,18,19]. Weed seed in the soil seedbank may not germinate for several years due to dormancy traits [20]. Understanding seed-producing characteristics of weed species is critical for developing methods that mitigate the addition of weed seed to the soil [21].
Following a 4-yr fallow soil with intensive tillage, research reported that only 10% of the original velvetleaf [Abutilon throphrasti (Medik)] seed was still present in the soil [22]. Additional studies observed reported viable weed seeds were reduced in the soil by 95% across five sites in Nebraska following a 5-yr period of controlling weeds exclusively with herbicides [23]. Cultivated soils accommodate large amounts of weed seed that interfere with crop production, which forces long-term weed management strategies to focus on controlling emerged seedlings [16]. Weed seed replenishing the soil seedbank will increase weed populations in successive growing seasons; therefore, an increase in herbicide applications are needed to control those weed populations [24]. However, a system of weed management that completely controls aboveground weeds for a long period of time will not completely eliminate weeds from that area [25]. Therefore, weed management can only be successful when control methods are focused on manipulating the weed seedbank [26].
Controlling competitive weeds during early season peanut growth can reduce yield loss potential and mitigate weed seed additions to the seedbank. It is imperative for peanut producers to reduce weed seed dispersion to the soil seedbank. Allowing weeds to mature and produce seed may increase weed populations in successive growing seasons. Depletion of the soil seedbank by controlling weeds prior to producing seed can potentially reduce weed populations and reduce the amount of herbicide applications needed in successive growing seasons. Currently, herbicides are the most widely used weed control option for Mississippi peanut producers. Based on previous research and current field observations, multiple herbicide applications are needed to control weeds in Mississippi peanut. Increasing the number of herbicide applications may result in greater weed control and weed seed suppression. Therefore, a study was established at the Delta Research and Extension Center to evaluate various herbicide programs for controlling common weed species that interfere with peanut in Mississippi. The purpose of this study was to identify the best herbicide program for weed control and reducing weed seed production in a Mississippi peanut production system.

Field Preparation and Study Design
A field study was conducted at the Delta Research and Extension Center in Stoneville, MS, USA in 2017 and 2018 (33 • 26 37.1 N, 90 • 54 29.88 W) to evaluate PRE and POST herbicide programs in Mississippi peanut. Soil series at each site was a Commerce silty clay loam (fine-silty, mixed, superactive, nonacid, thermic Fluvaquentic Endoaquepts) with a pH of 7.4 and an organic matter content of approximately 1.5%. The 2017 and 2018 sites were fallow the previous year. Therefore, both sites had a variety of weed species that were allowed to mature and replenish the soil seedbank naturally. However, barnyardgrass, hemp sesbania, Palmer amaranth, pitted morningglory, and prickly sida seed were broadcasted both years to achieve uniform weed populations.
In both siteyears, fields were disked and beds were established. Paraquat (Gramoxone SL 2.0, herbicide, 841 g ai ha −1 , Syngenta Crop Protection, Greensboro, NC, USA) was applied immediately after planting to control existing weeds. In both siteyears, 'Georgia-06G' (University of Georgia-Coastal Plain Experiment Station, Tifton, GA, USA) was planted to a depth of 5 cm at a seeding rate of 258,300 seeds ha −1 using a small-plot planter (John Deere MaxEmerge Plus 1700, Deere and Company, Moline, IL, USA). Peanut were planted on 23 May and 5 May in 2017 and 2018, respectively. Plot size was 4 × 6 m and consisted of four rows spaced 101.6 cm apart. All four rows received herbicide treatments, but only rows two and three were harvested.
The study was designed as a randomized complete block with four replications. Treatments were different combinations of herbicides which included acetochlor, clethodim, flumioxazin, lactofen, paraquat, and s-metolachlor (Table 1). Applications were made using a CO 2 -pressurized backpack sprayer equipped with turbo induction nozzles (Turbo TeeJet Induction 110015 nozzle, TeeJet Technologies, Springfield, IL, USA) set to deliver 140 L ha −1 at 206 kPa. All treatments received a PRE application and a sequential application at 2 to 3 weeks after emergence (EPOST), 4 to 5 weeks after emergence (MPOST), or both EPOST and MPOST. A nontreated control was included.

Data Collection
Visible estimates of peanut injury and weed control were recorded 21 d after PRE (DA PRE), 7 d after EPOST (DA-EPOST), and 7 and 35 d after MPOST (DA-MPOST) on a scale of 0 to 100% where 0 indicated no visible effect and 100 indicated complete plant death. Prior to each application timing, weed density (m −2 ) and height (cm) were recorded for barnyardgrass, hemp sesbania, Palmer amaranth, pitted morningglory, and prickly sida ( Table 2). Prior to peanut harvest, all weeds present in two 1-m 2 quadrats in each plot were harvested by hand, placed in paper bags, dried for 8 weeks at ≥32 • C, and then weighed. All weed species excluding Palmer amaranth were hand-threshed to remove seeds, and total seed weight for each species was recorded. Additionally, for each weed species except Palmer amaranth, five samples consisting of 100 seeds each were weighed to determine average seed weight for each species. The average seed weight for each species was then utilized to estimate the total seed m −2 in each plot. Total seed m −2 was calculated for barnyardgrass, hemp sesbania, pitted morningglory, and prickly sida with the following equation: Total number of seeds m −2 = (Total seed weight for sample × 100)/(Weight of 100 seeds) Palmer amaranth seed were removed by plant grinder (Thomas Model 4 Wiley Mill, Thomas Scientific, Swedesboro, NJ, USA) to determine total seed weight. Five 0.5-g samples of Palmer amaranth seed were counted to determine the average number of seed in each sample. The average seed number was then used to calculate the amount of seed produced m −2 in each plot. Average seed number for Palmer amaranth was calculated using the following equation: Total number of Palmer amaranth seed m −2 = (Number of seed in 0.5 g × Total weight of seed m −2 )/0.5 (2) To determine effectiveness of each herbicide program in reducing overall weed seed production, the following equation was used: Percent weed seed reduction following each herbicide program = [1 -(Total weed seed following each herbicide program/Total weed seed in nontreated control)] (3) Following weed sample collection, peanuts were unearthed by a digger. The peanut digger consisted of a blade underneath a conveyer belt to cut peanut taproot and overturn plants to be left on the soil surface. Peanut remained on the soil surface for 2 weeks before harvest. Rows two and three in each plot were unearthed and harvested to record peanut yield. In 2017, peanuts were harvested on November 20; however, inclement weather prevented peanut harvest in 2018.

Statistical Analysis
Square roots of visible estimates of peanut injury and weed control were arcsine transformed. The transformation did not improve homogeneity of variance; therefore, nontransformed data were used in analysis. Nontreated data were excluded when injury and weed control were analyzed. However, nontreated data were included in analysis for weed above-ground dry weight, weed seed production, and peanut yield. Data were pooled across siteyears and subjected to ANOVA using the PROC GLIMMIX procedure in SAS 9.4 (SAS Institute Inc., Cary, NC, USA) with year and replication (nested within year) as random effect parameters [27]. Least square means were calculated and mean separation (p ≤ 0.05) were produced in SAS, which converts mean separation output to letter groupings [28].

Results and Discussion
No peanut injury was observed following PRE applications. However, peanut injury was significant (p < 0.0001) at 7 DA-EPOST and 7 DA-MPOST applications. Peanut injury was observed when EPOST and MPOST treatments contained paraquat, lactofen, or lactofen plus clethodim. At 7 DA-EPOST, peanut injury was greatest from treatments containing paraquat EPOST (8 to 10%). All remaining treatments causing peanut injury 7 DA-EPOST contained lactofen with injury from 2 to 6%. Peanut injury 7 DA-MPOST was greatest following paraquat EPOST fb lactofen or lactofen plus clethodim MPOST with injury from 3 to 4%.

Barnyardgrass Control
Barnyardgrass control was significant at every evaluation. At 21 DA-PRE (p = 0.0012), a difference among treatments was detected for barnyardgrass control (Table 3). S-metolachlor provided the greatest level of control (92 to 95%) compared with flumioxazin and acetochlor (76 to 82%). The greatest difference among PRE applications was flumioxazin and s-metolachlor which provided 82 and 95% barnyardgrass control, respectively. At 21 DA-PRE, s-metolachlor provided greater control of barnyardgrass than flumioxazin and acetochlor.

Prickly Sida Control
Prickly sida control was significant 21 DA-PRE, 7 DA-EPOST, and 35 DA-MPOST. At 21 DA-PRE (p < 0.0001), all treatments provided ≥93% control of prickly sida ( Table 7). All treatments 7 DA-EPOST (p = 0.0192) provided ≥98% control of prickly sida. At 35 DA-MPOST (p = 0.0039), all treatments provided ≥94% control of prickly sida.   flumioxazin PRE fb acetochlor EPOST fb lactofen + clethodim MPOST 100 a 100 a 100 a 99 ab a All data pooled over two siteyears. Means within a column followed by the same letter are not different at p ≤ 0.05. b EPOST treatments were applied 2 to 3 weeks after emergence. MPOST treatments were applied four to five weeks after emergence. c All paraquat treatments included NIS at 0.25% v/v. d All lactofen and lactofen plus clethodim treatments included COC at 1% v/v.

Peanut Yield
Peanut were not harvested in 2018 due to inclement weather. However, there was a peanut yield difference among treatments in 2017 (p < 0.0001). Peanut yield among treatments ranging from 2680 to 3130 kg ha −1 were not different, but were different from treatments resulting in peanut yield of 982 to 2087 kg ha −1 (data not shown). Weed interference resulted in an 88% reduction in peanut yield when the nontreated control (368 kg ha −1 ) was compared to flumioxazin PRE fb lactofen plus clethodim MPOST (3126 kg ha −1 ).

Pitted Morningglory and Hemp Sesbania: Above-Ground Dry Weight and Seed Production
Prior to peanut harvest, weed above ground dry weight (AGDW) was significant for hemp sesbania (p < 0.0001) and pitted morningglory (p = 0.0009). Differences in seed production were detected for barnyardgrass (p = 0.0002), hemp sesbania (p < 0.0001), Palmer amaranth (p = 0.0443), and pitted morningglory (p = 0.031). Hemp sesbania and pitted morningglory were the only two species with differences for both AGDW and seed production ( Table 8). Hemp sesbania AGDW was 153 g and produced 1942 seeds m −2 for the nontreated control. Flumioxazin PRE fb paraquat EPOST fb acetochlor MPOST provided the least reduction of AGDW, which resulted in 86.12 g m −2 . However, S-metolachlor PRE fb paraquat EPOST was the worst treatment in terms of seed production which resulted in 1046 seeds m −2 deposited to the soil seedbank. Ten treatments provided comparable reduction in hemp sesbania seed production ranging from 0 to 207 seeds m −2 .
For pitted morningglory, all treatments reduced AGDW from 0 to 23 g m −2 (Table 8). Pitted morningglory following s-metolachlor PRE fb lactofen plus clethodim MPOST resulted in AGDW of 23 g m −2 which was different from treatments resulting in 0 to 8 g m −2 of AGDW. S-metolachlor PRE fb paraquat EPOST resulted in comparable reduction of pitted morningglory AGDW (15 g m −2 ) as s-metolachlor PRE fb lactofen plus clethodim MPOST. All treatments resulted in pitted morningglory seed production ranging from 0 to 138 seeds m −2 . Acetochlor PRE fb lactofen EPOST (138 seeds m −2 ) was different from all treatments that resulted in pitted morningglory seed production ranging from 0 to 55 seeds m −2 . The only comparable treatment to acetochlor PRE fb lactofen EPOST was flumioxazin PRE fb paraquat EPOST fb acetochlor MPOST, which resulted in pitted morningglory producing 63 seeds m −2 . Table 8. Above-ground dry weight of hemp sesbania and pitted morningglory m −2 , as well as seed production for barnyardgrass, hemp sesbania, Palmer amaranth, and pitted morningglory m −2 following PRE, EPOST b , and MPOST b herbicide programs prior to peanut harvest in Stoneville, MS, averaged over 2017 and 2018 a .

Herbicide Programs
Above

Barnyardgrass Seed Production
Barnyardgrass AGDW was not affected by the treatments imposed in this study (data now shown). However, treatments were different in terms of reducing seed production ( Table 8). All treatments resulted in barnyardgrass seed production from 2239 to 13,696 seeds m −2 . Barnyardgrass seed production following flumioxazin PRE fb lactofen plus clethodim MPOST resulted in the greatest suppression of seed production (2239 seeds m −2 ). Flumioxazin PRE fb paraquat EPOST resulted in the greatest level of barnyardgrass seed production at 13,696 seeds m −2 .

Palmer Amaranth Seed Production
Palmer amaranth AGDW was not affected by the treatments imposed in this study (data not shown). Palmer amaranth produced 58,196 seeds m −2 in the nontreated control ( Table 8). All treatments provided comparable reduction of Palmer amaranth seed production ranging from 538 to 30,420 seeds m −2 . Flumioxazin PRE fb lactofen plus clethodim EPOST fb acetochlor MPOST, which resulted in 538 Palmer amaranth seeds m −2 , was the only treatment different from s-metolachlor PRE fb lactofen plus clethodim MPOST, which resulted in 30,420 seeds m −2 . Flumioxazin PRE fb lactofen plus clethodim EPOST fb acetochlor MPOST reduced Palmer amaranth seed ≥99% compared with the nontreated control.

Prickly Sida Seed Production
Prickly sida AGDW and seed production were not affected by treatments in this study (data not shown). However, several treatments effectively reduced or eliminated seed production of prickly sida. In the nontreated control, prickly sida produced 4621 seeds m −2 . All treatments reduced prickly sida seed production from 0 to 3263 seeds m −2 .

Summary of Results
The current research demonstrated that herbicide programs provide different levels of control, AGDW, and seed production of barnyardgrass, hemp sesbania, Palmer amaranth, pitted morningglory, and prickly sida in Mississippi peanut.
At 35 DA-MPOST, 11 treatments provided comparable control of hemp sesbania ranging from 89 to 100% control (Table 4). At 35 DA-MPOST, s-metolachlor PRE fb paraquat EPOST provided 60% control of hemp sesbania, which was the lowest level of control. Pitted morningglory control 35 DA-MPOST ranged from 72 to 100% following all treatments (Table 6). Ten treatments provided comparable control of pitted morningglory ranging from 92 to 100% at 35 DA-MPOST. All treatments provided 94 to 100% control of prickly sida 35 DA-MPOST.
The treatments in this two-year study resulted in differences for control, AGDW, and weed seed production among all weed species. Previous research reported that weeds developed herbicide resistance by responding to overreliance of a single pest control method [10,29,30]. Additionally, research stated that multiple modes of action applied simultaneously or sequentially can reduce the potential for weeds to develop herbicide resistance [10]. Therefore, weeds exposed to treatments such as flumioxazin PRE fb lactofen plus clethodim MPOST may have the potential to develop resistance due to flumioxazin and lactofen being WSSA group 14 herbicides (Table 1). However, treatments such as flumioxazin PRE fb paraquat EPOST fb lactofen plus clethodim or flumioxazin PRE fb lactofen plus clethodim EPOST fb acetochlor MPOST may help to reduce the risk for weeds to develop herbicide resistance due to an additional herbicide with a different mode of action included in both treatments.
Several treatments in this study resulted in some weed species producing zero seed (Table 8). Even if treatments in this study provided weed control ≥95%, weeds that matured and produced seed may have replenished the soil seedbank. Several treatments reduced weed seed production, but did not completely eliminate the possibility of increasing weed populations in successive growing seasons [25]. If weed seed replenishes the soil seedbank, an increase in herbicide applications may be needed in sequential growing seasons to control them [24]. To effectively control weeds and deplete the soil seedbank, weed management strategies must implement methods for reducing the addition of weed seed to the soil seedbank [21]. Therefore, weed management strategies cannot be focused on aboveground vegetation alone, but must implement practices, such as tillage, to reduce the soil seedbank [22,26].

Conclusions
Among all weed species, the most consistent control was provided by flumioxazin PRE fb lactofen plus clethodim MPOST, flumioxazin PRE fb lactofen plus clethodim EPOST fb acetochlor MPOST, and flumioxazin PRE fb paraquat EPOST fb lactofen plus clethodim MPOST, which provided 88, 88, and 93% control 35 DA-MPOST. All treatments reduced hemp sesbania AGDW as compared to the nontreated control. However, pitted morningglory following s-metolachlor PRE fb lactofen plus clethodim MPOST had greater AGDW than the nontreated control and was different from all other treatments except s-metolachlor PRE fb paraquat EPOST. Treatments with PRE and MPOST or PRE, EPOST, and MPOST applications resulted in 82 to 93% reduction in weed seed production. Therefore, weed control programs must include a PRE and MPOST application to reduce weed seed production greater than PRE fb EPOST programs.