Introduction
The use of cover crops has increased significantly across the United States within the last decade. The state of Indiana has been a leader in cover crop adoption, with a little over 650,000 ha planted to cover crops in 2022 in comparison with 498,000 ha in 2014 (ISDA 2023). Despite all the incentives from federal and state agencies as well as the crop production industry, this represents only 12.8% of Indiana’s cropland (USDA-NASS 2023). Cover crops have been recommended as one strategy to improve the physical, chemical, and biological properties of soil (Baumhardt et al. Reference Baumhardt, Stewart and Sainju2015; Chen et al. Reference Chen, Rejesus, Aglasan, Hagen and Salas2022; Du et al. Reference Du, Jian, Du and Stewart2022; Muhammad et al. Reference Muhammad, Wang, Sainju, Zhang, Zhao and Khan2021) and reduce ground and surface water contamination (Lacey and Armstrong Reference Lacey and Armstrong2015; Ruffatti et al. Reference Ruffatti, Roth, Lacey and Armstrong2019). Furthermore, some cover crop species such as cereal rye (Secale cereale L.) can suppress weed emergence and growth, representing a valuable integrated weed management tool (Hodgskiss et al. Reference Hodgskiss, Young, Armstrong and Johnson2020; Loux et al. Reference Loux, Dobbels, Bradley, Johnson, Young, Spaunhorst, Norsworthy, Palhano and Steckel2017; Petersen et al. Reference Petersen, Zimmer, Maia, Armstrong, Young and Johnson2023).
One of the most frequently documented benefits of cover crop use is the increase in the soil organic matter (SOM) content (Moore et al. Reference Moore, Wiedenhoeft, Kaspar and Cambardella2014; Poeplau and Don Reference Poeplau and Don2015; Villamil et al. Reference Villamil, Bollero, Darmody, Simmons and Bullock2006). SOM is the primary source of energy used by microorganisms to survive and multiply (Fontaine et al. Reference Fontaine, Mariotti and Abbadie2003; Gunina and Kuzyakov Reference Gunina and Kuzyakov2022). Fungi and bacteria are examples of microorganisms that are constantly producing enzymes and releasing some of them into the soil solution. These enzymes are categorized into indicators of overall microbial activity (e.g., dehydrogenase, intracellular) or specific to certain nutrient cycles (e.g., hydrolases, extracellular). Dehydrogenase (DHA) is classified within the oxidoreductases, the largest enzyme group, and is responsible for catalyzing redox reactions of organic compounds (e.g., pesticides) in the soil (Dixon et al. Reference Dixon, Webb and Thorne1979). The hydrolases are responsible for catalyzing the carbon (C), nitrogen (N), phosphorus (P), and sulfur (S) cycles in the soil. Within the hydrolases, β-glucosidase (BG) decomposes SOM, which will ultimately result in the production of glucose, an energy source for soil microbes (Deng and Tabatabai Reference Deng and Tabatabai1994).
Among the several plant species currently used as cover crops, cereal rye is the most commonly used and is known for producing large amounts of above- and belowground biomass as well as scavenging residual N left in the soil by the previous crop or applied as fertilizer in fall (Kaspar et al. Reference Kaspar, Jaynes, Parkin and Moorman2007; Kladivko et al. Reference Kladivko, Kaspar, Jaynes, Malone, Singer, Morin and Searchinger2014; Lacey and Armstrong Reference Lacey and Armstrong2015; Ruffatti et al. Reference Ruffatti, Roth, Lacey and Armstrong2019). Once terminated, the cereal rye residue is slowly mineralized and incorporated into the SOM. Although not known for producing large amounts of biomass, leguminous cover crop species such as crimson clover (Trifolium incarnatum L.) also contribute to the SOM pool, primarily by fixing N from the atmosphere (Smith et al. Reference Smith, Knight and Peterson1982), which is released into the soil upon degradation of the plant residue. More recently, the vast majority of the initiatives that promote the use of cover crops list soil health as one of the main benefits from this practice (Chami et al. Reference Chami, Niles, Parry, Mirsky, Ackroyd and Ryan2023; Myers et al. Reference Myers, Weber and Tellatin2019). Although the term “soil health” is subjective, increased microbial activity is a principal component documented in several studies (Adetunji et al. Reference Adetunji, Ncube, Meyer, Mulidzi and Lewu2020; Brennan and Acosta-Martinez Reference Brennan and Acosta-Martinez2019; Finney et al. Reference Finney, Buyer and Kaye2017; Kim et al. Reference Kim, Zabaloy, Guan and Villamil2020; Nevins et al. Reference Nevins, Nakatsu and Armstrong2018, Reference Nevins, Lacey and Armstrong2020). The increased microbial activity is, for the most part, related to the increases in the SOM content as a result of cover crop residue decomposition, moisture conservation, enhanced aggregate stability, and improved drainage (Dinesh et al. Reference Dinesh, Ghoshal Chaudhuri, Sheeja and Shiva2009; dos Santos Cordeiro et al. Reference Echer and Araujo2021; Hu et al. Reference Hu, Jiang, Thomas, Chen, Xie, Hu, Kong, Yang, Chen, Zhang and Shi2023; Mendes et al. Reference Mendes, Bandick, Dick and Bottomley1999).
In addition to mineralization of SOM, soil microorganisms can also utilize herbicides and other commonly used pesticides as alternative sources of C and N (Qiu et al. Reference Qiu, Pang, Zhou, Zhang, Feng and Sheng2009). In fact, most of the herbicide degradation that occurs in soils is promoted by microbes (Van Eerd et al. Reference Van Eerd, Hoagland, Zablotowicz and Hall2003). Currently, there is no consensus about the effect of herbicides on soil enzymes. While some studies have reported no effect (Niemi et al. Reference Niemi, Heiskanen, Ahtiainen, Rahkonen, Mäntykoski, Welling, Laitinen and Ruuttunen2009; Omar and Abdel-Sater Reference Omar and Abdel-Sater2001), others have described either negative (Du et al. Reference Du, Zhu, Zhu, Zhang, Li, Wang, Wang, Zhang and Cheng2018; Mukherjee et al. Reference Mukherjee, Tripathi, Mukherjee, Bhattacharyya and Chakrabarti2016) or positive impacts (Kucharski et al. Reference Kucharski, Tomkiel, Baćmaga, Borowik and Wyszkowska2016; Singh and Ghoshal Reference Singh and Ghoshal2013) of herbicides on BG activity. Unlike BG, DHA generally shows reduced activity in the presence of herbicides (Bennicelli et al. Reference Bennicelli, Szafranek-Nakonieczna, Wolińska, Stępniewska and Bogudzińska2009; Sebiomo et al. Reference Sebiomo, Ogundero and Bankole2010; Tomkiel et al. Reference Tomkiel, Baćmaga, Borowik, Kucharski and Wyszkowska2019). However, this reduction is temporary, and the activity increases as the population of microbes capable of degrading the herbicide increases (Cole Reference Cole1976; Robertson and Alexander Reference Robertson and Alexander1994; Sebiomo et al. Reference Sebiomo, Ogundero and Bankole2010; Tyagi et al. Reference Tyagi, Mandal, Kumar and Kumar2018). Research conducted by Weaver et al. (Reference Weaver, Krutz, Zablotowicz and Reddy2007) demonstrated that, even when applied at 3-fold label rates, glyphosate did not cause microbial community shifts in the soil. Furthermore, these authors concluded that the application of glyphosate under laboratory settings caused only small and transient (<7 d) effects on the soil microbial community. In general, the response of a soil enzyme to a given pesticide is practically unpredictable, because different pesticides can either increase, decrease, or result in no effect to the enzyme. In addition, the response of soil enzymes to the presence of pesticides also vary by pesticide rate and soil type (Schaffer Reference Schaffer1993).
Establishing cover crops as the sole weed management strategy rarely results in acceptable season-long weed control (Burgos and Talbert Reference Burgos and Talbert1996; Teasdale Reference Teasdale1993; Teasdale et al. Reference Teasdale, Pillai and Collins2005). Competition for light, water, and nutrients and the release of allelochemicals are the primary means of weed suppression during cover crop growth. Once terminated, cover crops can suppress weeds through the physical barrier created by the residue left on the soil surface. The more biomass that is accumulated by the cover crop, the greater amount of weed suppression is achieved (MacLaren et al. Reference MacLaren, Swanepoel, Bennett, Wright and Dehnen-Schmutz2019; Osipitan et al. Reference Osipitan, Dille, Assefa, Radicetti, Ayeni and Knezevic2019). However, one disadvantage of late cover crop termination (i.e., more biomass accumulation) is the interference with cash crop development. Corn (Zea mays L.), for instance, is very sensitive to late termination of cover crops, showing nutrient deficiencies and stunted growth due to nutrient (primarily N) immobilization during mineralization of the cover crop residue (Nevins et al. Reference Nevins, Lacey and Armstrong2020; Reed et al. Reference Reed, Karsten, Curran, Tooker and Duiker2019; Rosa et al. Reference Rosa, Creech, Elmore, Rudnick, Lindquist, Fudolig, Butts and Werle2021). Thus, in most cases, the recommendation is to terminate the cover crop at least 2 wk before corn planting (Acharya et al. Reference Acharya, Bakker, Moorman, Kaspar, Lenssen and Robertson2017). At this stage, biomass accumulation by the cover crop is usually not enough to physically suppress weed emergence throughout the entire growing season. Therefore, the inclusion of soil-residual herbicides at cover crop termination is essential to extend the period of weed control (Whalen et al. Reference Whalen, Shergill, Kinne, Bish and Bradley2020).
To provide adequate weed control, soil-residual herbicides require proper placement, movement into the weed germination zone of the soil, and length of residual control. The length of residual control is affected by some soil properties such as SOM and clay content and environmental conditions such as temperature and rainfall volume. Furthermore, the residual activity of herbicides is also influenced by the overall microbial activity of the soil (García-Delgado et al. Reference García-Delgado, Barba-Vicente, Marín-Benito, Mariano Igual, Sánchez-Martín and Sonia Rodríguez-Cruz2019). Therefore, the adoption of agronomic practices that have the potential to increase the activity of soil microorganisms, such as the use of cover crops, could also lead to an increased degradation of herbicides in the soil. However, to date, there is no evidence in the literature that supports this hypothesis.
Another way cover crops can influence the fate of soil-residual herbicides is by interception, which occurs at the time of application and is directly related to the amount of cover crop biomass present, with more biomass resulting in more interception (Nunes et al. Reference Nunes, Arneson, DeWerff, Ruark, Conley, Smith and Werle2023). Once intercepted, residual herbicides can only move onto the soil surface with rainfall or irrigation. This movement, or leaching, of the herbicide is affected by the volume of water that falls onto the cover crop and also by the chemical properties of the herbicide. The more rainfall or irrigation, the greater amount of herbicide will leach onto the soil (Khalil et al. Reference Khalil, Flower, Siddique and Ward2019). Herbicides with higher water solubility (e.g., mesotrione, cloransulam-methyl) have a tendency to be washed off of the plants more easily than those with lower solubility (e.g., S-metolachlor, trifluralin) (Khalil et al. Reference Khalil, Flower, Siddique and Ward2019). In addition to the water volume and herbicide solubility, the maturity of the cover crop also affects the herbicide leaching from the plants onto the soil, with older plants having a lower ratio of cellulose:lignin than younger plants (i.e., during plant residue decomposition, enzymes such as BG will break down cellulose molecules, thus exposing lignin molecules). Lignin is considered a recalcitrant cell wall component (Vanholme et al. Reference Vanholme, Demedts, Morreel, Ralph and Boerjan2010), and in some plant parts, can account for 60% to 80% of the secondary cell wall composition (Musha and Goring Reference Musha and Goring1975). Research conducted by Dao (Reference Dao1991) suggested that most of the herbicide binding onto the plant surface occurs at the exposed lignin sorption sites, while herbicide binding to cellulose is minimal. The lack of rainfall or irrigation after the application of a residual herbicide at cover crop termination results in reduced concentrations of these pesticides in the soil. Not only because the herbicide that was intercepted by the cover crop residue will not leach onto the soil but also because the herbicide that reaches the soil at the time of application will not be incorporated into the top layer of soil. Ultimately, lower concentrations of residual herbicides in the soil lead to a greater reliance on postemergence herbicides to achieve acceptable weed control during the critical weed-free period (Loux et al. Reference Loux, Dobbels, Johnson and Young2011), which goes against the principles of herbicide-resistance management.
Most of the research conducted thus far has focused on the weed control efficacy of residual herbicides when applied at cover crop termination. However, there are still confounding factors such as microbial degradation and interception by cover crop residue that could be negatively impacting the efficacy of these herbicides. Knowing how these two factors affect herbicide fate in cover-cropping systems is essential to further improve this practice. Therefore, the objectives of this research were (1) to evaluate the effect of cover crop use on soil microbial activity, (2) to measure the concentration of residual herbicides in the soil when applied at cover crop termination, (3) to correlate microbial activity and herbicide concentration in the soil, (4) to investigate herbicide interception by cover crops and further leaching onto the soil, and (5) to assess early- and late-season weed biomass as influenced by cover crop and herbicide treatments.
Materials and Methods
Field trials were established at Throckmorton (40.29°N, 86.90°W) and Pinney (41.44°N, 86.92°W) Purdue Agricultural Centers (TPAC and PPAC, respectively) near Lafayette and Wanatah, IN, respectively, in September of 2019. Both trials remained in the same location until October of 2023 when final data were collected. Fields for each trial were managed as a conventional corn–soybean [Glycine max (L.) Merr.] rotation before the initiation of the study, and were planted to soybeans during the 2019 growing season. The soil at TPAC consisted of a Drummer silty clay loam (22% sand, 53% silt, and 25% clay). The soil at PPAC consisted of a Tracy sandy loam (70% sand, 17% silt, and 13% clay). Soil samples were taken in March of each year and used to determine fertility parameters of the soil in each trial (Table 1). Soil from each site was tilled before first cover crop planting in September of 2019, using a rotary tiller at 10 cm deep and then managed as a transitional no-till system following a corn–soybean rotation during the subsequent years.
Table 1. Chemical properties and bulk density of the soil from each cover crop and fallow treatment at PPAC and TPAC at 0- to 10-cm depth for all 4 yr of the study. a
a Abbreviations: CEC, cation exchange capacity; OM, organic matter; PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
b Soil samples were taken in March of each year.
Treatments were arranged in a split-plot design and included two cover crop species, cereal rye and crimson clover, as well as a no cover crop control as main plots. The soil-residual herbicides tested were randomized in each main plot, replicated four times, and divided into no residual, medium residual, and heavy residual (residual load based on the number of herbicides applied) (Table 2) for a total of 36 experimental units. Plots were 3 m by 8 m in size. Cover crops were planted in the fall of each year using a no-till drill (John Deere 1590, John Deere, Moline, IL) at 19-cm row spacing, at seeding rates of 112 and 22.5 kg ha−1 of cereal rye (‘Hazlet’, Cisco Company, Indianapolis, IN) and crimson clover (‘Dixie’, Cisco Company), respectively (Table 3). In the spring of 2020 and 2022, herbicide-resistant corn (SmartStax™ DKC 62-52RIB, Bayer Crop Science, St Louis, MO) was planted at 86,450 seeds ha−1 in 76-cm row spacing (Table 3). In both years, starter fertilizer was applied at corn planting at 34 kg N ha−1 (19-17-00) and a side-dress application was made near the V6 growth stage at 200 kg N ha−1 (UAN 28-00-00). In the spring of 2021 and 2023, herbicide-resistant soybean (Enlist E3® P26T57E, Corteva, Johnston, IA) was planted at 350,000 seeds ha−1 in 38-cm row spacing (Table 3).
Table 2. Herbicide treatments and rates applied at cover crop termination. a
a Abbreviations: PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
b Herbicides were applied in tank mixture at 2 wk before cash crop planting each year.
Table 3. Cover crop planting and termination dates and cash crop planting dates at TPAC and PPAC from 2019 until 2023. a
a Abbreviations: PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
Herbicides were applied traveling at 4.8 km h−1 using a CO2-pressurized spray boom equipped with eight AIXR 110015 (TeeJet®, Spraying Systems, Wheaton, IL) nozzles spaced 38 cm apart and calibrated to deliver 140 L ha−1 and operating at 165 kPa. Glyphosate (Roundup PowerMax®, Bayer Crop Science) and glufosinate (Liberty® 280 SL, BASF, Research Triangle Park, NC) were applied in tank mix at 1,750 g ae ha−1 and 737 g ai ha−1, respectively, at 4 wk after corn or soybean planting to all plots. Non-ionic surfactant (Class Act® Ridion®, WinField Solutions, St Paul, MN) and ammonium sulfate (Amsol®, WinField Solutions) were added to all herbicide applications at 0.25 and 5% v/v, respectively (cover crop termination and postemergence application).
Cover crop and weed biomass were determined separately 1 d before spring termination using a 0.25-m2 quadrat that was randomly placed within the first 1 m (lengthwise) of each plot. All aboveground plant material inside the quadrat was harvested by cutting the plants at the base (1 cm above soil surface) with scissors and was placed in separate bags for cover crop and weed biomass samples. Bags were placed in a forced-air oven at 80 C for 96 h. Dry weights were recorded and converted to kg ha−1.
Weed biomass was also determined at 4 (from 2021 until 2023) and 18 wk after termination (WAT) (from 2020 until 2023). Two 0.25-m2 quadrats were randomly placed between the two center rows of the cash crop, one in the front and one in the back of each plot. All plant material within each quadrat was harvested by cutting the plants at the base (1 cm above soil surface) with scissors. Samples were placed in a forced-air oven at 80 C for 96 h. Dry weights were recorded and converted to kg ha−1. Starting in 2021, weed biomass was also determined following the same method at 4 wk after cash crop planting.
In 2020, soil samples were collected 5 d before cover crop termination and at 21, 28, 56, 84, and 112 d after termination (DAT) to determine soil enzyme activity and the concentrations of residual herbicides. For all subsequent years, soil samples were collected 5 d before cover crop termination and at 0, 10, 14, 28, 56, 84, and 112 DAT. Soil samples taken before cover crop termination were used to determine the base levels of soil microbial activity. Fourteen soil cores were collected at 0- to 5-cm depths for each plot, using a 2-cm-diameter probe. The cores were homogenized to form one composite sample per plot and refrigerated at 4 C until processing. The soil probe was cleaned with a 50% acetone solution between plots to avoid sample contamination. No more than 1 d after collection, soil samples were passed through a 2-mm sieve and thoroughly homogenized, and then an aliquot of approximately 60 g was placed in a 50-ml Falcon tube and stored at −20 C before analysis of herbicide concentration. A 50% acetone solution was used to clean the sieve between samples. The remainder of the soil sample was kept at 4 C and then used to measure the activity of BG and DHA. Sample storage time was kept constant before each enzyme assay across sampling events, sites, and years and was never more than 72 h after sampling. Soil moisture from each sample was determined before the enzyme assays from a 5-g subsample that was placed in a forced-air oven at 105 C for 48 h.
BG activity was measured according to the method described by Eivazi and Tabatabai (Reference Eivazi and Tabatabai1988), with adaptations. The method used in this study differs from the standard bench-scale method by reducing all chemicals and soil amount by 90%. All glassware was substituted by 2-ml microcentrifuge tubes, and the filtration step was substituted by centrifuge. Briefly, 0.1 g (±0.01) of soil was weighed (in triplicate plus one control sample) into 2-ml microcentrifuge tubes, and 20 µl of toluene (Fisher Scientific, Fair Lawn, NJ) was added to all samples. After 15 min, 100 µl of PNG solution (50 mM p-nitrophenyl-β-d-glucopyranoside; Acros Organics, Pittsburgh, PA) and a modified universal buffer solution (MUB; pH 6) were added to the samples, except for the control samples, which received only MUB. Samples were then thoroughly mixed and incubated at 37 C (±1 C) for 1 h. After incubation, 100 µl of calcium chloride (0.5 M; Fisher Scientific) and 400 µl of tris(hydroxymethyl)aminomethane buffer (THAM; 100 mM, pH 12) were added to all samples. PNG was added to the control samples only after the addition of THAM, which stops the enzymatic reaction. Eight blank samples, without soil, were included in the assay and treated like the experimental samples (same chemicals added and same incubation conditions). Tubes were then placed in a centrifuge and spun at 13,000 rpm for 8 min. Two hundred microliters of the supernatant was transferred to a 96-well microplate (costar flat-well medium binding polystyrene, cat no. 9017, Corning, Corning, NY). Absorbance of the supernatant was measured at 405 nm with a microplate reader (Multiskan Sky 96-well Microplate Reader, Thermo Scientific, Waltham, MA). The amount of p-nitrophenol (pNP 1 mM; Sigma Chemicals, St Louis, MO) released from each sample was calculated from a calibration curve that was prepared using a 10 mM pNP solution diluted in water to achieve concentrations equivalent to 0, 0.01, 0.02, 0.03, 0.04, and 0.05 mM of pNP. BG activity was then expressed in µmol pNP g−1 dry soil h−1.
DHA activity was measured according to the method described by Benefield et al. (Reference Benefield, Howard and Howard1977) and adapted by von Mersi (Reference von Mersi1996), von Mersi and Schinner (Reference von Mersi and Schinner1991), and Shaw and Burns (Reference Shaw and Burns2006). Additional adaptations were made to improve efficiency of the process and reduce the use of chemicals and laboratory supplies. This assay was conducted in the dark to avoid photodegradation of INT [2-(4-iodophenyl)-3-(4-nitrophenyl)-5-phenyl-2H-tetrazolium chloride; Sigma Chemicals]. Briefly, 0.1 g (±0.01) of soil was weighed (in triplicate plus one control sample) into 2-ml microcentrifuge tubes. Control samples were autoclaved at 120 C for 20 min. Then, 200 µl of INT solution (5 mg mL−1; Alfa Aesar, Ward Hill, MA) and 150 µl THAM buffer (1 M; pH 7) were added to all samples. Tubes were closed, shaken vigorously, and placed in the incubator for 2 h at 37 C (±1 C). Immediately after incubation, 1 ml of an extractant solution (1:1 N,N-dimethylformamide and ethanol) was added to each tube, which were then left at room temperature for another incubation of 1 h, being shaken (vortex mixer) every 20 min. Tubes were centrifuged at 13,000 rpm for 8 min after the last incubation. Then, 200 µl of supernatant was transferred to a 96-well plate (polypropylene, Eppendorf, Hamburg, Germany). Absorbance of the supernatant was measured at 464 nm with a microplate reader (Multiskan Sky 96-well Microplate Reader, Thermo Scientific, Waltham, MA). The amount of iodonitrotetrazolium formazan (INTF; Sigma Chemicals) released from each sample was calculated from a calibration curve prepared using INTF standard (100 mg INTF ml−1; dissolved in extractant solution) concentrations equivalent to 0, 25, 50, 100, 200, 300, and 500 μg of INTF in 6.75 ml of solution. DHA activity was expressed in μg INTF g−1 dry soil h−1.
The concentration of soil-residual herbicides in the soil was determined following the QuEChERS (Quick-Easy-Cheap-Effective-Rugged-Safe) extraction method (Anastassiades et al. Reference Anastassiades, Lehotay, Štajnbaher and Schenck2003), with modifications. Briefly, soil samples were thawed and a 3-g aliquot was weighed into 50-ml Falcon tubes. Then, 15 ml of deionized water and acetonitrile (Optima™ LC/MS grade with 1% formic acid; Fisher Scientific, Fair Lawn, NJ) and 10 µl of internal standards were added to each tube, which were then shaken (vortex mixer) for 1 min. After shaking, 6 g of MgSO4 and 1.5 g of NaOAc were added to each sample. Subsequently, tubes were placed in a Geno/Grinder 2010 (SPEX Sample Prep, Metuchen, NJ) and shaken at 1,000 rpm for 3 min and then centrifuged at 2,500 rpm for 10 min. Twelve milliliters of the supernatant was transferred into 15-ml dispersive solid-phase extraction tubes (part no. 5982-5158, Agilent Technologies, Palo Alto, CA) that were then shaken at 1,100 rpm for 5 min on the Geno/Grinder. Tubes were centrifuged at 4,000 rpm for 5 min, and the entire supernatant was transferred into a 15-ml centrifuge tube. Samples were dried in a SpeedVac (SPD1030, Thermo Scientific, San Jose, CA) overnight. Pellets were resuspended with 150 ml of 100% acetonitrile (Optima™ LC/MS grade, Fisher Scientific, Fair Lawn, NJ) and vortexed until fully dissolved. Tubes were placed in a centrifuge at 4,000 rpm for 5 min, and 130 µl of the supernatant was transferred into a 96-well microplate (polypropylene, Eppendorf, Hamburg, Germany) before ultra–high performance liquid chromatography (UHPLC) analysis. All samples were analyzed in an Agilent 1290 Infinity II UHPLC with a 6,470 triple quadrupole mass spectrometry and a EclipsePlus C18 RRHD 1.8 μm, 2.1 by 50 mm column (Agilent Technologies, Santa Clara, CA) at the Bindley Bioscience Center at Purdue University. Recoveries from fortified untreated soil samples indicated that recovery was 112%, 80%, 74%, 113%, and 70% for sulfentrazone, S-metolachlor, cloransulam-methyl, atrazine, and mesotrione, respectively.
All data were subjected to an ANOVA using the PROC GLIMMIX procedure in SAS (v. 9.4, SAS Institute, Cary, NC). There was a significant treatment by year interaction for the early- and late-season weed biomass, residual herbicide concentration, and enzyme activities. Therefore, results are presented separately by year. The interaction between cover crop and herbicide treatments for early- and late-season weed biomass was nonsignificant, therefore, data were combined over cover crop and fallow treatments within each year. Assumptions of normality and homogeneity of variance were evaluated by visual assessment of residual plots. Data were log or square-root transformed when needed. However, original mean values are presented. Means were separated using Tukey’s honest significant difference (HSD) (enzyme activity and herbicide concentration data) or Fisher’s protected LSD (early- and late-season weed biomass data) (α = 0.05). Pearson’s correlation coefficients (PROC CORR procedure in SAS) were used to identify significant (α = 0.05) relationships between soil microbial activity and concentration of residual herbicides in the soil.
Results and Discussion
Cover Crop and Weed Biomass at Termination
Cereal rye and crimson clover biomass accumulation was, on average, greater at TPAC than at PPAC (Figure 1). Although the cover crop planting and termination dates were similar between sites, the overall soil fertility parameters (Table 1) from TPAC were better than those from PPAC, which resulted in greater cover crop growth. Furthermore, average monthly temperatures during the cover crop growing season were slightly higher at TPAC (data not shown). The average cereal rye biomass accumulation at TPAC and PPAC, across all years, was 4,210 and 2,623 kg ha−1, respectively. The average crimson clover biomass accumulated in that same period was 1,342 and 1,099 kg ha−1 at TPAC and PPAC, respectively. In general, crimson clover biomass accumulation was fairly low in both sites. According to the Midwest cover crop decision tool, the cutoff date for crimson clover planting in Indiana is September 18 (MCCC 2024). With the exception of the first cover crop planting in 2019, all other planting dates were between the first and second week of October. In the fall of 2022, the combination of a late planting date with exceptionally dry weather conditions (Figure 2) resulted in the germination of very few crimson clover plants in the fall. Hodgskiss et al. (Reference Hodgskiss, Young, Armstrong and Johnson2020) conducted field trials at TPAC in the 2 yr before this research and reported up to 1,476 and 3,709 kg ha−1 of crimson clover and cereal rye biomass, respectively. Also, those authors did not have success in establishing crimson clover in 1 of the 2 yr of their study.
Figure 1. Cover crop and weed biomass 1 d before cover crop termination. Weed biomass was collected in all 4 yr in the fallow plots and whenever there were weeds present in the cover crop plots. Abbreviations: CR, cereal rye; CC, crimson clover; FL, fallow; PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
Figure 2. Thirty-year average and actual precititation for each month that data were collected over the 4 yr of study period at PPAC (Pinney Purdue Agricultural Center, Wanatah, IN) and TPAC (Throckmorton Purdue Agricultural Center, Lafayette, IN).
Fallow plots were not kept weed-free during the study. At the time of cover crop termination, the predominant weed species present in the fallow plots and plots with low crimson clover stands were: common chickweed [Stellaria media (L.) Vill.], henbit (Lamium amplexicaule L.), purple deadnettle (Lamium purpureum L.), and shepherd’s purse [Capsella bursa-pastoris (L.) Medik.]. Cereal rye plots were, for the most part, free of winter-annual weed species at the time of spring termination.
Early-Season Weed Control
The primary weed species present in the trial areas were: giant ragweed (Ambrosia trifida L.; TPAC only), waterhemp [Amaranthus tuberculatus (Moq.) Sauer], redroot pigweed (Amaranthus retroflexus L.), common lambsquarters (Chenopodium album L.), horseweed [Conyza canadensis (L.) Cronquist; syn.: Erigeron canadensis L.; TPAC only), giant foxtail (Setaria faberi Herrm.), yellow foxtail [Setaria pumila (Poir.) Roem. & Schult.], barnyardgrass [Echinochloa crus-galli (L.) P. Beauv.], and fall panicum (Panicum dichotomiflorum Michx.). The overall weed density was much higher at TPAC compared with PPAC. The experimental area at PPAC was used for commercial grain production for several years before the beginning of the study, while the field at TPAC was maintained as a weed science research area for at least 10 yr before the study initiation.
No cover crop by herbicide interactions were observed for early-season weed biomass, while the main effect of herbicide was significant. Therefore, data were pooled over cover crop and fallow treatments within each year (Table 4). The use of residual herbicide at cover crop termination consistently reduced weed biomass at 4 weeks after planting (WAP) relative to the termination without residual herbicides. Between 2021 and 2023, the application of residual herbicides resulted in an average of 88% and 81% reduction in weed biomass at 4 WAP at PPAC and TPAC, respectively. At PPAC, the heavy residual herbicide program increased weed control relative to the medium herbicide program only in 2023. In 2022 and 2023 at TPAC, the inclusion of a third residual herbicide in the tank mixture applied at cover crop termination improved weed control by an average of 18% compared with the termination with two residual herbicides. This improved weed control observed at TPAC was likely the result of adding either cloransulam-methyl (2023) or mesotrione (2022) to the tank mixture, which resulted in greater control of A. trifida (data not shown). Investigating the control of A. trifida following the application of several residual herbicides in combination or not with mesotrione, Westrich et al. (Reference Westrich, Johnson and Young2024) reported up to 84% reduction in A. trifida biomass as result of mesotrione application (alone or in tank mixture with the other herbicides evaluated). Similarly, other authors have observed up to 94% control of acetolactate synthase inhibitor (ALS)-susceptible A. trifida at 4 wk after the application of cloransulam-methyl (Follings et al. Reference Follings, Soltani, Robinson, Tardif, Lawton and Sikkema2013). In this present study, the use of cereal rye and crimson clover cover crops resulted in weed control similar to that in the fallow treatment at 4 WAP in both locations and all years of data collection.
Table 4. Weed biomass at 4 wk after cash crop planting, from 2021 until 2023.
a Abbreviations: PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
b Herbicide treatments: heavy, three residual herbicides; medium, two residual herbicides; no, no residual herbicides.
c Data were log transformed. However, original mean values are presented. Numbers followed by the same letter within year and location are not significantly different according to Fisher’s protected LSD (P < 0.05).
Late-Season Weed Control
Late-season biomass was estimated in all 4 yr of the study. However, in 2021 and 2023, when soybean was planted as cash crop, all plots from both locations were weed-free at 18 WAP. The primary weed species present in the trial areas in 2020 and 2022 were S. media (predominant species), common purslane (Portulaca oleracea L.), A. tuberculatus, A. retroflexus, C. album, S. faberi, S. pumila, E. crus-galli, and P. dichotomiflorum.
No cover crop by herbicide interactions were observed for late-season weed biomass, while the main effect of herbicide was significant. Therefore, data were pooled over cover crop and fallow treatments within each year (Table 5). For the 18 WAP evaluation timing, the use of residual herbicides at cover crop termination did not result in greater weed control relative to the no residual program. One exception was at PPAC in 2020, when the application of three residual herbicides improved weed control at 18 WAP by 74% in comparison to the application of glyphosate plus glufosinate only. In the same location and year, the use of two residual herbicides at cover crop termination provided similar weed control as the no residual program. The average weed biomass for the whole trial area was 119 and 676 kg ha−1 in 2020 and 4 and 36 kg ha−1 in 2022 at PPAC and TPAC, respectively, at 18 WAP. Although crop yield was not determined in this study, other researchers have observed corn yield losses ranging from 11% to 74% in the presence of 2,220 to 5,900 kg ha−1 of A. tuberculatus biomass (Steckel and Sprague Reference Steckel and Sprague2004). Fausey et al. (Reference Fausey, Kells, Swinton and Renner1997) investigated the impact of increasing S. faberi densities in corn yield and reported losses ranging from 14% to ∼50% when densities were increased from 10 to 100 S. faberi plants m−1 of corn row. Similar to what was observed at 4 WAP, the use of cereal rye and crimson clover cover crops did not provide greater weed control than the fallow treatment later in the growing season at PPAC and TPAC between 2020 and 2023.
Table 5. Weed biomass at 18 wk after cash crop planting (WAP) in 2020 and 2022.
a Abbreviations: PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
b Herbicide treatments: heavy, three residual herbicides; medium, two residual herbicides; no, no residual herbicides.
c All plots were weed-free at 18 WAP at PPAC and TPAC, in 2021 and 2023. Therefore, only data from 2020 and 2022 are shown. Data were log transformed. However, original mean values are presented. Numbers followed by the same letter within year and location are not significantly different according to Fisher’s protected LSD (P < 0.05).
Soil Enzyme Activities
BG
The effect of cover crop use on BG activity is shown in Figure 3, and the detailed means separation for each site and year can be found Supplementary Tables S1 and S2 (available in the Supplementary Material). No cover crop by herbicide interactions were observed for BG activity, while the main effect of cover crop was significant. Therefore, data were pooled over herbicide treatments within each year and sample timing (Figure 3). In general, BG activity was 38% higher at TPAC compared with PPAC. This result was expected, because the average SOM content from TPAC (2.7%) was 1.7-fold greater than that from PPAC (1.6%) (Table 1). BG has a critical role in organic matter cycling in soils and is therefore responsive to changes in SOM content (Bandick and Dick Reference Bandick and Dick1999; Debosz et al. Reference Debosz, Rasmussen and Pedersen1999; Eivazi and Tabatabai Reference Eivazi and Tabatabai1988; Monreal and Bergstrom Reference Monreal and Bergstrom2000; Sinsabaugh et al. Reference Sinsabaugh, Lauber, Weintraub, Ahmed, Allison, Crenshaw, Contosta, Cusack, Frey, Gallo, Gartner, Hobbie, Holland, Keeler and Powers2008; Turner et al. Reference Turner, Hopkins, Haygarth and Ostle2002). BG activities measured in our field studies are comparable to those reported in previous studies with similar SOM contents. Bandick and Dick (Reference Bandick and Dick1999) measured BG activities ranging from 0.3 to 1.5 µmol pNP g−1 h−1 in soils with SOM varying from 2.3% to 3.8%. Similarly, Eivazi and Tabatabai (Reference Eivazi and Tabatabai1988), who developed the assay to measure BG activity in soils, reported activities from 0.07 to 2.12 µmol pNP g−1 h−1 in soils with SOM ranging from 0.8% to 9.4%.
Figure 3. β-glucosidase (BG) activity in the soil (0- to 5-cm depth) from 5 d before until 112 d after cover crop termination from 2020 until 2023. Data points represent mean ± SE of four replications. PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
The use of crimson clover as cover crop for four growing seasons resulted in increased BG activity relative to the fallow in an average of 34% and 32% of the soil sample timings at TPAC and PPAC, respectively. However, significant amounts of crimson clover biomass were achieved only in the first year of the study, with an average of 3,145 kg ha−1 of biomass across TPAC and PPAC. For the subsequent years, the average crimson clover biomass for both sites was 580 kg ha−1. Therefore, data from this research are not enough to provide meaningful conclusions regarding the effect of crimson clover on BG activity. Nevertheless, other researchers have found increased BG activity as result of crimson clover use as cover crop when average biomass accumulation reached 5,972 kg ha−1 (Tyler Reference Tyler2020). On the other hand, when cereal rye was used as cover crop, we observed increased BG activity relative to the fallow in an average of 32% and 86% of the soil sample timings at TPAC and PPAC, respectively, across 4 yr of data collection (Supplementary Tables S1 and S2). When treatment differences occurred, the use of cereal rye resulted in an average of 14% and 27% increase in BG activity relative to the fallow at TPAC and PPAC, respectively. These results are consistent with those of previous research by Eivazi et al. (Reference Eivazi, Pinero, Dolan-Timpe and Doggett2024) that evaluated the effect of cereal rye management strategies on soil enzyme activities and reported up to 39% greater BG activity in plots with cereal rye relative to the no cover crop control. With respect to the frequency of increased BG activity with cereal rye, other researchers have found results contrary to our study. Tyler (Reference Tyler2020) assessed the effect of cereal rye (average biomass: 3,840 kg ha−1) on BG activity in soils with SOM contents above 4.9%. These authors found consistent increases in BG activity in the soil of plots with cereal rye in comparison to the no cover crop control for all sample timings during 3 yr. However, our results showed that the consistent increase in BG activity with cereal rye was achieved only in our low SOM site (1.6%; PPAC) and after 2 yr of cover crop adoption. In the moderate SOM site (2.7%; TPAC), cereal rye use resulted in increased BG activity relative to the fallow in no more than five out of eight sample timings during 4 yr of data collection. We suggest that the consistent BG activity increase as result of cereal rye use at PPAC is due to the low SOM background levels from that site. In other words, the C:N input (not excluding other elements present in the biomass but in lower concentrations) from cereal rye use was enough to result in significant differences in BG activity between the cereal rye and fallow treatments in a short period of time only where the starting point for the SOM content was low. Whereas at TPAC, where the SOM levels were higher (relative to PPAC) to begin with, the C:N input from cereal rye was not enough to result in significant differences between those treatments. Nevertheless, significant differences in BG activity could also happen in moderate to high SOM sites, but most likely only after several years of cereal rye use.
DHA
Effect of cover crop use on DHA activity is shown in Figure 4, and the detailed means separation for each site and year can be found Supplementary Tables S1 and S2. No cover crop by herbicide interactions were observed for DHA activity, while the main effect of cover crop was significant. Therefore, data were pooled over herbicide treatments within each year and sample timing (Figure 4). The average activity of DHA was 29% higher at TPAC (3.8 µg INTF g−1 h−1) compared with PPAC (2.7 µg INTF g−1 h−1). We also attribute this difference in enzyme activity to the greater SOM from TPAC. More SOM means more substrate to support microbial growth and, therefore, more enzyme activity (Yuan and Yue Reference Yuan and Yue2012). With respect to the effect of cover crop use on enzyme activity, the pattern of DHA and BG activities were similar for each site, with cover crop use resulting in more instances of increased activity relative to the fallow at PPAC than at TPAC. For example, at TPAC, out of the 30 soil sample timings within the 4 yr of the study, the use of cereal rye increased DHA activity relative to the fallow in only seven timings. At PPAC that number increased to 25 sample timings. After 4 yr of cover crop use at PPAC, the average DHA activity in the soil of plots with cereal rye (3.3 µg INTF g−1 h−1) was 42% higher than the activity measured in the soil of fallow plots (1.9 µg INTF g−1 h−1), whereas at TPAC, the increase in DHA activity was only 5%. Results from PPAC are in agreement with those observed by Eivazi et al. (Reference Eivazi, Pinero, Dolan-Timpe and Doggett2024), who reported a 47% increase in DHA activity in the presence of cereal rye in comparison to the no cover crop control.
Figure 4. Dehydrogenase (DHA) activity in the soil (0- to 5-cm depth) from 5 d before until 112 d after cover crop termination from 2020 until 2023. Data points represent mean ± SE of four replications. PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
Correlation between Herbicide Concentration and Enzyme Activities
The Pearson’s correlation analysis in Table 6 shows that BG was not strongly correlated with herbicide concentration in the soil. Conversely, DHA was strongly correlated with the concentration of sulfentrazone, S-metolachlor, and cloransulam-methyl in 2021 and atrazine and S-metolachlor in 2022. In 2021, there was a clear trend of DHA activity increase from 0 to 112 DAT (Figure 4), while the concentrations of the herbicides used that year were decreasing through the growing season. Conversely, a positive and strong correlation was observed in 2022 between DHA activity and atrazine and S-metolachlor concentrations in the soil, when both enzyme activity and herbicide concentrations declined through most of the growing season. In that year in particular, the cover crop termination at TPAC was followed by frequent rainfall events within 14 DAT (Figure 2; Supplementary Figure S1; month of May), which resulted in a sharp decline in the concentration of the herbicides in the soil soon after application. However, an extended period of drought between June and July (Figure 2; Supplementary Figure S1) impacted DHA activity that only begun to recover in August of that year. Therefore, the correlations between DHA and the concentrations of atrazine and S-metolachlor in the soil are explained by environmental factors that affected each variable separately, rather than the direct effect of the herbicides in the enzyme activity. In agreement with some previous reports (Cole Reference Cole1976; Davies and Greaves Reference Davies and Greaves1981; Dennis et al. Reference Dennis, Kukulies, Forstner, Plisson, Eaglesham and Pattison2023; Tomkiel et al. Reference Tomkiel, Baćmaga, Wyszkowska, Kucharski and Borowik2015; Tyler Reference Tyler2022), results from this research suggested that none of the five residual herbicides used impacted BG or DHA activity when applied at cover crop termination and within the recommended label rates.
Table 6. Pearson’s correlation coefficients between herbicides concentrations in the soil and enzyme activities, from 2021 until 2023. a , b
a Abbreviations: PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN).
b Correlation coefficients were nonsignificant (ns) or significant at *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.
c n = 168 for sulfentrazone, S-metolachlor, and atrazine; n = 84 for cloransulam-methyl and mesotrione.
Concentration of Residual Herbicides in the Soil
The effect of cover crop on the concentrations of sulfentrazone, S-metolachlor, cloransulam-methyl, atrazine, and mesotrione is shown in Figures 5–7, and detailed means separation are shown in Supplementary Tables S3–S5. The use of cereal rye as cover crop reduced the concentration of sulfentrazone in the soil relative to the fallow only at 0 DAT at PPAC in 2021 and 10 DAT at TPAC in 2023. For all other sample timings within 2021 and 2023, the concentration of sulfentrazone was similar for all treatments. As mentioned previously, the fallow plots were not kept weed-free during our study. Thus, the presence of weeds in those plots at 0 DAT resulted in the interception of residual herbicides, similar to what happened in plots with cereal rye. The concentration of cloransulam-methyl in the soil of plots with cereal rye was lower than fallow plots only at 10 and 14 DAT at TPAC and PPAC, respectively, in 2023. No significant effect of cover crop use on cloransulam-methyl concentration was identified for all other sample timings in 2021 and 2023. Among all herbicides evaluated, S-metolachlor was the only one that had reduced concentrations in the soil of plots with cereal rye relative to the fallow in 20 out of the 28 sample timings between 2021 and 2023, with most of the reductions observed at PPAC (up to 5 out of 7 sample timings). When differences occurred, the concentration of S-metolachlor was, on average, 34% lower in plots with cereal rye in comparison with the concentration measured in fallow plots. Cereal rye biomass intercepted 39% more atrazine than the weeds present in the fallow plots at PPAC in 2021. However, for all other sample timings, the concentration of atrazine in plots with cereal rye was either similar or higher than the concentration from fallow plots. In addition to the interception by cover crop or weed biomass, the herbicides tested in this study were also likely intercepted by the crop residue (e.g., corn and/or soybean stubble) left on the soil surface after the previous year’s harvest. Even though this factor was not taken into consideration in this study, we must recognize that the crop residue lying on the surface does affect residual herbicide fate through interception and retention (Banks and Robinson Reference Banks and Robinson1986; Bauman and Ross Reference Bauman and Ross1983; Dao Reference Dao1991; Ghadiri et al. Reference Ghadiri, Shea and Wicks1984; Reddy et al. Reference Reddy, Locke, Wagner, Zablotowicz, Gaston and Smeda1995). Furthermore, aged crop residue tends to adsorb more herbicide than fresh residue, giving that there are more sorption sites (lignin) exposed upon degradation of the residue (Dao Reference Dao1991).
Figure 5. Sulfentrazone, S-metolachlor, and cloransulam-methyl concentration in the soil (0- to 5-cm depth) and daily rainfall amounts from 0 until 112 d after cover crop termination at PPAC (Pinney Purdue Agricultural Center, Wanatah, IN) and TPAC (Throckmorton Purdue Agricultural Center, Lafayette, IN). in 2021. Data points represent mean ± SE of four replications.
Figure 6. Atrazine, S-metolachlor, and mesotrione concentration in the soil (0- to 5-cm depth) and daily rainfall amounts from 0 until 112 d after cover crop termination at PPAC (Pinney Purdue Agricultural Center, Wanatah, IN) and TPAC (Throckmorton Purdue Agricultural Center, Lafayette, IN) in 2022. Data points represent mean ± SE of four replications.
Figure 7. Sulfentrazone, S-metolachlor, and cloransulam-methyl concentration in the soil (0- to 5-cm depth) and daily rainfall amounts from 0zero until 112 d after cover crop termination at PPAC (Pinney Purdue Agricultural Center, Wanatah, IN) and TPAC (Throckmorton Purdue Agricultural Center, Lafayette, IN) in 2023. Data points represent mean ± SE of four replications.
In general, although the use of cereal rye resulted in increased BG and DHA activities, we did not identify any specific trends that would suggest an increase in the degradation of sulfentrazone, S-metolachlor, cloransulam-methyl, atrazine, or mesotrione. Similarly, Reddy et al. (Reference Reddy, Locke, Wagner, Zablotowicz, Gaston and Smeda1995) investigated the effect of no-till and conventional till on the activity of soil enzymes and degradation of chlorimuron. No-till adoption resulted in increased enzyme activity compared with conventional till, but only minimal changes in the pattern of chlorimuron degradation were detected (Reddy et al. Reference Reddy, Locke, Wagner, Zablotowicz, Gaston and Smeda1995). On the other hand, previous studies have demonstrated that the use of perennial ryegrass (Lolium multiflorum Lam.) as cover crop in no-till or conventional till systems can increase the soil microbial activity and lead to enhanced degradation of fluometuron in comparison to the same systems without ryegrass (Zablotowicz et al. Reference Zablotowicz, Locke and Gaston2007).
Herbicide Interception by Cereal Rye Residue
Herbicide interception was calculated as the percent reduction from the expected concentration of the residual herbicide in the soil (considering complete incorporation of the herbicide applied into into the top 0 to 5 cm of soil) based on the actual concentration measured at 0 DAT (Table 7). The use of cereal rye as cover crop did not affect residual herbicide fate by increasing the activities of BG and DHA in the soil (Figures 3 and 4). However, cereal rye biomass did intercept substantial amounts of the herbicides applied at termination, leading to significant reductions in the initial concentration of all herbicides in the soil. The average amount of herbicide interception across the 3 yr of data collection and two locations was 77%, with S-metolachlor being the herbicide with the least amount of interception (55%) and mesotrione the herbicide with the highest amount (91%). Although the interception of herbicides by cereal rye biomass results in lower concentrations in the soil, the early-season weed biomass data (Table 4) from this research showed that even under reduced initial concentrations in the soil, the use of residual herbicides at cover crop termination improved weed control relative to the treatment without residual herbicides. This improved weed control was in part due to rainfall events that occurred on the days following the application (Figures 5–7), thus washing off the herbicide from the biomass and incorporating it into the soil. The total rainfall volume for the 7 d following the herbicide application between 2021 and 2023 and across the two locations was, on average, 37 mm (Supplementary Figure S1). Previous studies have demonstrated that 50 mm of rainfall was enough to move 90% of the atrazine initially intercepted by wheat (Triticum aestivum L.) straw into the soil, thus increasing the concentration of the herbicide in the soil more than 2-fold (Ghadiri et al. Reference Ghadiri, Shea and Wicks1984). In the absence of rainfall, low concentrations of residual herbicides in the soil have the potential to increase the selection pressure for herbicide-resistant weed biotypes (Busi et al. Reference Busi, Gaines, Walsh and Powles2012; Neve and Powles Reference Neve and Powles2005), which has been the main issue in weed science for decades. For example, a multiple-resistant rigid ryegrass (Lolium rigidum Gaudin) population that was subjected for three generations to recurrent low doses of pyroxasulfone had more than 30% survival rate at 240 g ai ha−1 (2.4-fold the label rate) (Busi et al. Reference Busi, Gaines, Walsh and Powles2012). In that same study, Busi et al. (Reference Busi, Gaines, Walsh and Powles2012) suggested that only full label rates of pyroxasulfone would be enough to provide adequate weed control.
Table 7. Expected and actual concentrations of residual herbicides in the soil (0- to 5-cm depth) and interception by cereal rye at the time of cover crop termination from 2021 until 2023 at PPAC and TPAC. a
a Abbreviations: DAT, days after cover crop termination; PPAC, Pinney Purdue Agricultural Center (Wanatah, IN); TPAC, Throckmorton Purdue Agricultural Center (Lafayette, IN); UHPLC ultra–high performance liquid chromatography.
b Expected concentration of the herbicide in the soil (i.e., 100% of the applied herbicide is incorporated into the top 5 cm of soil). Equivalent to the herbicide application rate in ng ha−1 divided by the soil weight in g ha−1 (0- to 5-cm depth).
c Herbicide concentration measured in the UHPLC at 0 DAT.
d Percentage reduction from the expected concentration to the actual herbicide concentration measured at 0 DAT.
Our study demonstrates that cover crops occasionally increase activity of BG and DHA during the growing season, in comparison to the fallow control. Even in times when there was an increase in the activity of these two enzymes, increased degradation of any of the residual herbicides was not observed. Furthermore, results from TPAC in 2021 showed a strong correlation between increased DHA activity and decreased concentrations of sulfentrazone, S-metolachlor, and cloransulam-methyl, suggesting that the herbicides were being used by that enzyme as a substrate to sustain population growth. Our results also indicated that the initial concentrations of all residual herbicides in the soil were significantly reduced due to interception by cereal rye biomass. However, our results showed significant reductions in early-season weed biomass when residual herbicides were included in the tank mixture relative to the application of glyphosate plus glufosinate. The inclusion of two or three residual herbicides in the tank mixture applied at cover crop termination resulted in similar early-season weed biomass reductions in 2 and 1 out of 3 yr at PPAC and TPAC, respectively. In this study, for the most part, the application of residual herbicide at cover crop termination did not improve the late-season weed control relative to the no residual herbicide program. Season-long weed control was achieved only in 2021 and 2023, most likely due to the narrower row spacings adopted for soybean planting (38 cm). The use of cereal rye or crimson clover as cover crops did not improve the control of summer annual weed species relative to the fallow treatment in any of the locations and years of the study. Overall, results from this study suggest that soil-residual herbicides can and should be included in the tank mixture applied at cover crop termination without risks of increased degradation. Furthermore, even with significant interception by the cover crop biomass, soil-residual herbicides still reduced weed biomass early in the season. The combination of cover crops and soil-residual herbicides is, therefore, one excellent alternative to improve weed control following an integrated management approach. Future research could investigate the adoption of more comprehensive herbicide programs that include split applications of residual herbicides along with postemergence applications to achieve season-long weed control.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/wsc.2025.13
Funding statement
This research was partially funded by Indiana Corn Marketing Council.
Competing Interests
The author declare no conflicts of interest.
Open access
