Residual Herbicides and Cover Crops Interactions for Soybean Weed Control

Residual herbicides and cover crops are important tools inside an integrated weed management program. The straw produced in crop rotation can interacts with herbicides. The aim of this study was to evaluate the interaction of diclosulam, sulfentrazone, imazethapyr, flumioxazin, s-metolachlor and pyroxasulfone with black oat, cereal rye, common vetch and oilseed radish cover crops and it´s reflection on weed control and soybean production. Were evaluated the biomass production of cover crops and its influence on the soybean population. A phytosociological survey of the weed community was carried out, further evaluating the control provided by the herbicides and its effects on soybean productivity. Diclosulam was the more efficient herbicide tested, reducing both weed density and biomass (68% and 89%, respectively) compared to the fallow, independently of cover crop straw. The best control levels for the population identified were provided by the combination of the herbicides diclosulam with black oat, radish or fallow. We observed that herbicide efficacy in this case was more related with control spectrum than with herbicide-straw interation. This research demonstrates that the integration of cover crops and residual herbicides is efficient in the suppression and control of weeds in the soybean crop in the no-tillage system.


Introduction
Soybean (Glycine max (L.) Merr.) is a important oilseed for human and animal nutrition, with 37.03 million hectares being cultivated in Brazil with production of up to 126 million tons of grain, produced in more than 58% of the country's arable area (Companhia Nacional de Abastecimento [CONAB], 2020).
Is currently cultivated in the no-tillage system (NTS) which has as fundamental precepts the permanent vegetation cover and the seeding on the plant biomass of the predecessor crop. The success of the no-tillage system in tropical and subtropical environments is due to the cultivation of cover species combined with the use of the herbicide glyphosate for its desiccation (Day et al., 2020;Kan et al., 2020) The advent of transgenic RR (RoundUp Ready) soybean made it possible to control weeds also in crop post-emergence. However, the repeated use of the herbicide ended up selecting resistant weeds, which led to an increase of 149.14% in the use of herbicides between 2007 and 2014 in Brazil (Agrofit, 2020), in addition to the increase in environmental problems such as the drift of herbicides to non-target areas (Vieira et al., 2020).
The worldwide demand for the production of food without specific patterns of chemical residues and the recent cases of limitation and prohibition of the use of glyphosate in countries in Europe and Latin America, also place Brazil at the center of the discussion on reducing the use of pesticides in food production on a global scale (Miyazaki, Bauer-Panskus, Bøhn, Reichenbecher, & Then, 2019). accumulation of the weed community in the autumn and winter period and in summer crops as weeds and crops compete for light, water and nutrients (Akobundu et al., 2000;Blackshaw et al., 2001;Brennan & Smith, 2005;Grimmer & Masiunas, 2004;Peachey et al., 2004;Stivers-Young, 1998). In addition, cover crops are responsible for the interception and absorption of light that acts on the germination of weed seeds and, in some cases, their suppression by allelopathy (Teasdale & Daughtry, 1993).
Rotating mechanisms of action by using residual herbicides can also reduce glyphosate use within a management program, as well as new cases of resistance (Heap & Duke, 2018). When applied at the time of seeding, pre-emergent herbicides promote a prolonged control of weeds in the early stages of the crop, increasing the period prior to interference (PPI) (Rizzardi, Rockenbach, & Schneider, 2020).
Studies show that the use of pre-emergence herbicides in conjunction with cover crop cultivation within a weed management program can be effective in reducing interference in early summer (Cornelius & Bradley, 2017;Wiggins, Hayes, & Steckel, 2016).
The cover straw cultivated in the off season promotes soil shading and contributes to the reduction of weed species germination, on the other hand, this physical barrier can also reduce the efficiency of herbicides, since they have solubility characteristics that depend on a certain amount of water to break the straw and reach the ground (Da Silva et al., 2020). Higher straw densities can mitigate herbicide transposition, increasing the possibility of degradation (Fornarolli, Rodrigues, Lima, & Valério, 1998;Reddy et al., 1995).
Under these conditions, research is needed that point to the development of sustainable and economically viable production systems, focusing on the search for integrative solutions for different weed control methods, from the establishment of cover crops in winter to the end of the cycle of summer culture. The objective of this work was to evaluate the efficacy of weed control, soybean selectivity from residual herbicides interacting without and with cover crops straw.
The accumulation of precipitation during the experiment was 1250 mm, with low precipitation accumulated in the summer months. After fertilization with mineral N, a precipitation of 20 mm was observed, that contributed to dissolve the fertilizer. A precipitation of 18 mm was also observed five days after the summer sowing, which contributed to a good germination and emergence of the soybean, also favoring the action of pre-emergent herbicides applied on the day of sowing. Figure 1 shows precipitation and temperature data during the conduct of the experiment. Four species of winter cover crops were established in this area on 25/06/2020: black oat (Avena strigosa S.), cereal rye (Secale cereal L.), common vetch (Vicia sativa L.), oilseed radish (Raphanus sativus var. Oleiferus) plus a control treatment (fallow without cultivation). Sowing densities were: 66 kg ha -1 for black oat, 84 kg ha -1 for cereal rye, 96 kg ha -1 for common vetch and 20 kg ha -1 for oilseed radish, defined according to the technical manual of plants cover (Calegari, 2016). Seeding was carried out by broadcast, followed by light harrowing. The seeds used in the experiment were purchased commercially.
During the winter cover cycle, nitrogen fertilization (150 kg of nitrogen per hectare in form of urea 45%) was carried out for each treatment as recommended in the fertilizer and liming guidelines for the states of Rio Grande do Sul and Santa Catarina (Sousa & Ermani, 2016). After a period of 110 days from sowing, the covers were desiccated with 1,860 g ae ha -1 of glyphosate (Glizmax Prime, 480 g ea L -1 , Dow AgroSciences, São Paulo, Brazil).
Applications were carried out using a backpack spray pressurized by CO 2 at a constant pressure of 3 kPa, equipped with an one-meter application bar equipped with two tips and AIXR110015 nozzles (TeeJet Technologies, Wheaton, IL), regulated to deliver 200 L ha -1 of solution.The sown soybean cultivar was DM 53i54, which has an indeterminate growth habit, average cycle of 125 days in the state of Paraná and belongs to maturity group 5.4, showing stability and high yield potential in a subtropical climate. The sowing density was 350,000 plants per hectare with 0.45 m spacing between rows.

Data Collection and Analysis
In order to measure the accumulation of dry mass provided by the cover crops during the winter and the amount of straw at the time of sowing, the plant cover was collected before desiccation with the aid of a 0.062 m² frame, in four replications, totaling 0.25 m² sampling. Samples were dried by heating in an oven with forced air circulation at 60 °C for 72 hours. For soybean, were evaluated in the two central lines of the plot the number of plants per linear meter at 28 days after emergence and the visual injury at 14 and 28 days after sowing (DAS), according to the European Weed Research Council (EWRC, 1964) rating scale (0 to 100%). At 42 days from 0% t aid of a 0.2 phytosocio horizontal treatments The soybe meters, tot average pr yield per h 0.05). In c software (R

Cover
The bioma radish at th ton ha -1 ), Overall, ce Figure      For the visual control of herbicides, there was an interaction between the analyzed factors and a significant difference between the averages of the treatments. Several herbicides analyzed showed very low control, such as s-metolachlor (5.6%), pyroxasulfone (3%) and flumioxazin (13.51%), including that observed in fallow, demonstrating that this behavior is linked to control inefficiency for the observed species and not the interference of the cover crop in the movement of the herbicide to its target. The herbicide diclosulam promoted, regardless of the cover crop, that is, the levels of dry mass residue present at the time of application, greater control of weeds when compared to the others (Table 1). Imazethapyr, showed similar results to diclosulam, in coverings with black oat and oilseed radish. Note. Averages in the same column followed by distinct capital letters and in the row followed by differ lowercase letters differ by Tukey test (p ≤ 0.05).
Cover crops did not modify the observed efficacy of diclosulam, flumioxazin, and pyroxasulfone. Black oat improved the observed efficacy of sulfentrazone and imazetaphyr. Cereal rye cover improved the observed efficacy of imazethapyr and s-metolachlor. Common vetch cover improved the observed efficacy of imazetaphyr.
The herbicide flumioxazin (enzyme protopor-phyrinogen oxidase inhibitor, PROTOX) was not efficient in controlling the weed population, however, it was superior when compared to the efficacy of the herbicides s-metolachlor and pyroxasulfone.

Soybean Yield
For soybean, symptoms of phytotoxicity can occur because residual herbicides application. Low symptoms were noted 14 days after applications (less than 5%), for sulfentrazone. After 28 days, no further damage to crop development was observed (data not shown). For average soybean yield, a significant interaction of the effect of herbicides with the different winter coverings was observed.
In the unfolding of the herbicide factor, the treatment without winter covering with the applications of imazethapyr and diclosulam increased soybean yield. Flumioxazin and sulfentrazone applications had no effect. The application of diclosulam and imazetaphyr don't showed differences from a fallow condition (Table 2). In black oat, the crop that most reduced the presence and development of weeds, the treatments with diclosulam and imazethapyr had higher yields. The same was observed for cereal rye and common vetch. For oilseed radish only treatment with imazethapyr reduced weed development. Note. Averages in the same row followed by distinct capital letters and in the column followed by differ lowercase letters differ by Tukey test (p ≤ 0.05).
Comparing the chemical and cultural method, the yield obtained by using cover crops as the only method of weed suppression was not sufficient to guarantee higher yields. The summation effect of chemical control can be observed when comparing the isolated effect of black oat crop (1820 kg ha -1 ) with the effect of application of herbicides in fallow.Values ranged from 2094 kg ha -1 to 3556 kg ha -1 (treatments with sulfentrazone and diclosulam, respectively). The worst situation observed was in fallow without application of residual herbicide, reaching soybean yield of 1679 kg ha -1 .

Cover Crop Biomass
The establishment of winter cover crops is essential in maintaining the no-tillage system, enhancing soil healthy in addition to reducing weed emergence in the early stages of the crop. According to (Nunes et al., 2006), the ideal minimum amount of biomass for ground cover in the no-tillage system should be 6.0 ton ha -1 near what was observed for black oat. On the other hand, excess of straw can difficult sowing, because straw cutting and seeds/fertilizers depositions are affected (Trogello, José Modolo, Scarsi, & Dallacort, 2013).
Moderated cereal rye densities, as 67 kg ha -1 , do not affect the soybean stand, however, seeding densities varying between 100-135 kg ha -1 , can cause a reduction in the soybean population up to 4% (Essman et al., 2020;Schramski, Sprague, & Renner, 2021). Our results support these observations, since the sowing density of cereal rye used in this experiment was 84 kg ha -1 , with no differences in soybean emergence compared to the no cover crops treatment.
According to Modolo et al. (2020), the exacerbated accumulation of biomass can negatively affect the number of plants per hectare, reducing the successor crop's productivity. In this sense, actions such as the reduction of nitrogen fertilization of the covers and early desiccation can be interesting management alternatives to avoid a substantial increase in biomass in years with greater precipitation.

4.2Weed Suppression
The production systems are much diversified and vary according to their location, altitude, soil and relief. The same happens in the establishment of weed communities. Although there was no significant interaction of the factors, both the cover crops and the herbicides decreased the density and dry mass of weeds growing with soybean. The use of residual herbicides has been one of the main methods of annual weed control, and is one of the main recommendations for the management of herbicide resistance (Somerville, Powles, Walsh, & Renton, 2017).
Previous research shows that the application of residual herbicides in combination with cover crops residues can provide farmers adequate weed control at the end of the season (Cornelius & Bradley, 2017;Wiggins et al., 2016), because the use of coverings during winter reduces the incidence of light on the soil, reducing the temperature range, reducing the germination rate of some weed species (Teasdale & Mohler, 1993). In addition to promoting a change in edaphoclimatic conditions, biomass residues have the ability to suppress weed growth by releasing allelopathic compounds that act as growth inhibitors in the weed rooting zone (Burgos, Talbert, & Mattice, 1999;Olofsdotter, Jensen, & Courtois, 2002).
According to Brennan and Smith (2005), the amount of cover crop biomass determines the level of plant suppression, as greater amounts of residues provide greater suppression. However, the suppression exclusively by the cultivation of winter covers varies from year to year, because the production of dry mass is influenced by jas.ccsenet.org Journal of Agricultural Science Vol. 14, No. 9; particular conditions of each environment, such as fertility, average rainfall and average temperature during its cycle and weed community, basically composed by soil seedbank (Schramski et al., 2021).
Even with the average dry mass reduction of the herbicides diclosulam and imazethapyr being higher than the average of the effect of all the coverings, we cannot say that the herbicides were more efficient than the coverings for weed suppression in this case, because coverages such as common vetch and oilseed radish caused higher density reduction and dry mass when compared to the herbicide sulfentrazone, for example. In this case, soybean yield and economic return should be considered.

Weed Control
For visual control, there was a significant interaction between herbicide and cover factors, showing synergism of some combinations, demonstrating the importance of adopting weed management programs that include control methods both in winter and in summer. The same result was observed by Whalen et al. (2020), where they concluded that the use of pre-emergence herbicides in conjunction with cover crops is more efficient in controlling weeds when compared to management only with herbicides in the summer. Wiggins et al. (2016) found that rye, hairy vetch (Vicia villosa), clover (Trifolium) or winter wheat (Triticum aestivum) in combination with a residual herbicide resulted in 87% control of pigweed (Amaranthus albus), with a control of 65% being observed where cover crops did not receive added residual herbicide effect. The main characteristic of residual herbicides is their prolonged action on the soil seed bank, reducing germination and seedling emergence. After its application, the residual herbicide control efficiency decreases over time, because in contact with the environment, the molecule can be subject to photodegradation, volatilization, chemical and biological degradation and sorption processes (Silva et al., 2007). Schramski et al. (2021), observed 99% control of horseweed (Conyza spp.) when using the mixed herbicides glyphosate + 2,4-D + flumioxazin + metribuzin, regardless of the presence of plant cover, however, even with the efficiency of control by herbicides, the author emphasizes the need for a crop cover in order to reduce horseweed biomass and obtain better control with the application of a post-emergence herbicide only.
The herbicide action dynamics can change due to precipitation conditions, however, there was no observation of the negative influence of cover on the chemical control of weeds, since there were no water limitations after the application of the herbicides in this work. An accumulated precipitation of 18 mm was observed up to five days after application (Figure 7), enough to reduce the possibilities of adsorption of molecules on the straw and facilitate the concentration of the herbicide in the solution (Clark et al., 2019;Da Silva et al., 2020).
The inferior control performances of the herbicides s-metolachlor, pyroxasulfone and sulfentrazone in this case, occurred due to the predominance of wild radish plants in the experiment, as these herbicides act mainly in the control of monocotyledonous and are not recommended for the two species with highest relative importance identified (Agrofit, 2020).
It is worth mentioning that there are already cases of wild radish resistance to ALS-inhibiting herbicides in Brazil and multiple resistance in Australia (Heap, 2021), in this case, the better performance of ALS inhibitors herbicides could be compromised by the presence of biotypes resistant to this mechanism of action.

Soybean Yield
For average soybean yield, a significant interaction of the effect of herbicides with the different winter coverings was observed. Such behavior was also observed by Schramski et al. (2020), when soybean yield varied between 52-145% higher only using residual herbicides, and between 19-75% higher only using cover crops.
The yield obtained in the treatment characterized by the absence of weeds and residual herbicides confirms that the volume of biomass produced by cover crops did not change yield, even though, previously observed, affected soybean seed emergence. Such a difference in the plant stand is a factor that can be compensated for, as soybean plants can modify their individual development to compensate for empty row spaces (Moore, 1991;Pires et al., 1998;Thomas et al., 1998).
Cover species such as sorghum (Sorghum bicolor L.) cause soybean suppression by releasing allelopathic compounds with yield effects (Denadai, De Mello, Chioderoli, & De Niro Gazola, 2016), however, for Almeida (1991), allelopathic compounds released by cereal rye, oilseed radish and oat (Avena sativa) roots do not interfere in the germination percentage of soybean seeds. Bortolini and Fortes (2005), studying allelopathy caused by cover crops, concluded that root exudates from black oat and hairy vetch plants do not reduce soybean seed germination.
Comparing the chemical and cultural method, the yield obtained by using cover crops as the only method of weed suppression was not sufficient to guarantee higher yields. The summation effect of chemical control can be observed when comparing the isolated effect of black oat crop (1820 kg ha -1 ) with the effect of application of herbicides in fallow. Values ranged from 2094 kg ha -1 to 3556 kg ha -1 (treatments with sulfentrazone and diclosulam, respectively). The worst situation observed was in fallow without application of residual herbicide, reaching soybean yield of 1679 kg ha -1 . This research confirms the results of other studies that have shown the ability of winter cover crops to reduce weed density and dry mass, but reinforces that residual herbicides are still essential in the management program (Cornelius & Bradley, 2017;Essman et al., 2020).
The effect of cover crops plus straw interactions is an important management alternative that should be increasingly explored due to the use of more than a single control method. The interaction of the herbicide with the straw and the soil is a complex process, each herbicide has particularities regarding its physical/chemical factors, and the variation in coverage makes it even more difficult to study their relationships. More studies should be carried out considering the herbicide dynamics in different weed communities.