Miscanthus sacchariflorus (Maxim.) Hack., a rhizomatous C4 grass species, also called as silver banner grass has received considerable attention as a potential bioenergy crop because of its high biomass production [1]. M. sacchariflorus has a broad geographic range (up to 50° N in Eastern Russia) [2–4]. It has also been investigated as a feedstock, for ensilage, large-scale paper production [5–7] and higher bioremediation capacity of heavy metals [7]. The vigorous rhizomes of M. sacchariflorus are useful in preventing soil erosion [8]. Some ornamental accessions of M. sacchariflorus are sold in nurseries [9] and the inflorescence is used as a winter bouquet [10]. Its annual biomass production is higher than any other capable perennial bioenergy crop [7]. It is fairly tolerant to a wide range of environmental conditions and propagates in varied water and nitrogen contents with minimal input [5,11–13]. Miscanthus sinensis Andersson (Poaceae), a C4 perennial bioenergy crop and also called Chinese silver grass is native to Asia, primarily in China, Korea and Japan, and grows in a wide range of environments and altitudes (from sea level to 2500 m) [14,15]. They originate in a region ranging as far north as Sakhalin and as far south as the Indochinese peninsula [4]. It is suitable for thatching, animal feeds and the production of compost, sold as an ornamental plant in nurseries, and as raw material for industrial use for the production of paper pulp [15].

Both species differ morphologically, phylogenetically, and ecologically [14,15]. Compared to M. sacchariflorus, the accessions of M. sinensis are not an aggressive competitor at the early growing stages due to a weak rhizome spread [12,16]. They occasionally grow in the same area forming a sympatric population from ~29° N to 43° N [3], and are considered a promising renewable biomass resource [14,17]. M. sinensis usually has a caespitose form, prefers aerobic soils and grows mostly in hills, and seminatural grasslands on sloppy lands [18] and is typically diploid with a monoploid genome size of about 2.5–2.8 pg. M. sinensis is an early colonizer [15,19] with great cold tolerance and yield in cold regions [15,20,21], whereas M. sacchariflorus, which has a rhizomatous habit, prefers to grow under wetter soil conditions, in riparian environments, on river banks, mostly on mesic, viscid, sandy soil formed by floodings, and dominates fertile lowland habitats on alluvial plains [14,22,23]. M. sacchariflorus can be diploid or tetraploid with a monoploid genome size of about 2.1–2.3 pg [22,24,25]. The population of M. sacchariflorus from China and Korea are generally diploid [26,27], whereas its accessions from Japan are predominantly tetraploid [28]. M. sacchariflorus flowers less readily than M. sinensis [11,29,30] and produce a lower number of stems per plant (average 26 stems during 3rd year of cultivation) than M. sinensis (average 49 stems) [31]. Climate change is posing a serious threat to the global environment. Fossil fuels emit a huge amount of polluting gases contributing to an increased greenhouse effect, triggering environmental pollution and climate change [32]. Due to the fast depletion of fossil reserves and the rapid growth of the world population, the demand for bioenergy is expected to increase by 10 fold by 2050 in both energy and transport sectors [33,34]. Therefore, the use of renewable energy (bioenergy) is inevitable and must be carried out in a sustainable way to reduce the environmental impacts [35]. Miscanthus is characterized by a rapid growth, and thus a high productivity, that can be grown with a low input of nutrients and water consumption, and may confer a high capacity to sequester carbon in the soil that enables to mitigate the global warming by reducing greenhouse gas emissions [35].

Weeds are a major threat to the growth of bioenergy crops as they constantly compete with them for nutrients, space, sunlight, and moisture. Some important broadleaf weeds and grass weeds that commonly grow in Miscanthus fields contribute to the loss of biomass yield of the crop [36]. The growth rate of Miscanthus, at the beginning of the establishment year, is slow and sensitive to weed interference [36], which reduces its ability to compete with weeds for growth. Previous studies on Miscanthus revealed difficulties in its establishment due to prolonged and various degrees of dormancy, cold stress, and water deficiency during growth [36]. These plants take considerable time to fully develop the canopy and fill the wide gap between seedlings [36]. The bare land between the seedlings resulting from low densities allows weeds to grow and compete with the young shoots of Miscanthus. In the absence of weed control, approximately a 97% reduction in M. sacchariflorus biomass yield was reported under field conditions by grass weeds such as Echinochloa crus-galli, Digitaria sanguinalis, Setaria viridis, and broadleaf weeds including Stellaria aquatic, Amaranthus lividus, Galinsoga ciliate, Calystegia sepium, Chenopodium album, Acalypha australis, Viola mandshurica, Rorippa palustris, Ipomoea hederacea and Trifolium repens [36]. The higher rate of competition of Miscanthus plants with weeds was observed during the first year of growth of Miscanthus in the field, and it was reduced in the subsequent years [37–41]. As the rate of growth of these weeds is higher than that of Miscanthus, it has a detrimental effect on the growth of the crop due to the deprivation of nutrients, space, light, and moisture. In the absence of a proper weed control system, young shoots and canopy die, show poor growth, or produce less biomass. Therefore, controlling or suppressing weed growth at or before the critical point of weed growth not only minimizes yield loss but also protects plants from diseases. Moreover, several previous studies indicated an improvement in the biomass of bioenergy crops by using pre-emergence or post-emergence herbicides [42,43].

Previous work on weed control on these perennial crops has mainly focused on Miscanthus × giganteus, which is widely cultivated in the UK, European countries and USA for biomass production [1,44]. Several pre-emergence and post-emergence herbicides have been tested for use in the early growth of Miscanthus × giganteus [1,40,41,44]. The herbicides such as pendimethalin, S-metolachlor, and isoxaflutole, applied at planting, and POST herbicides, such as 2,4-D (1060 g ai ha^–1), bromoxynil (840 g ai ha^–1), and dicamba (560 g ai ha^–1) were found to be effective on controlling the weeds and safety for Miscanthus × giganteus [44]. Other herbicides such as atrazine were effective in controlling a wide range of broadleaves weeds, whereas, acetochlor was less effective in controlling the emerged weeds (68%) in a Miscanthus × giganteus field [45]. Recently, interest in M. sinensis and M. sacchariflorus has been increased for use as biomass energy crops in Europe, North America, and South Korea for productivity trials [44,46,47]. However, very few field studies have been conducted on weed control systems for M. sacchariflorus using pre-and post-emergence herbicides [36]. Also, no study has comparatively evaluated herbicide-based weed management systems for M. sinensis and M. sacchariflorus under field conditions. This research hypothesized that (1) all the tested herbicides will reduce the emergence of weeds, sustain toxic effects and lowers the survivability of the distinct weed community of a M. sinensis and M. sacchariflorus field, (2) mulches pretreated with pre-emergence herbicides would impact early Miscanthus growth and development but have no adverse effect on yield, and (3) mulches pretreated with the pre-emergence and post-emergence herbicides would provide effective weed control by reducing the density and biomass (FW and DW) of weeds. Therefore, the main objective of the present study was to determine the effectiveness of specific pre-and post-emergence herbicides to control weeds. In addition, we investigated the safety of pre-and post-emergence applications of herbicides on M. sinensis and M. sacchariflorus under field conditions, and also evaluated the phytotoxicity of herbicides on M. sinensis and M. sacchariflorus.

The bioefficacy of pre-and post-emergence herbicides was evaluated in the M. sinensis and M. sacchariflorus fields at the Kangwon National University, South Korea, (37°56009.96″ N; 127°46055.21″ E; Fig. 1) at an average altitude of 100 m for two years (during the years 2018, 2019). Two different plots were used to study the effect of herbicides on the two species of Miscanthus and weeds. Meteorological data were obtained from the database of the Korea Meteorological Administration. The average temperature of the cultivated field ranged from −9.9°C (January) to 29.8°C (August). Soil samples in each plot were collected in April 2018 after plot construction by randomly selecting three sampling points. Briefly, 10 g of soil samples (10–20 cm) were gathered using a sterilized spatula. The collected soil samples were placed in individual whirl-pack bags and immediately transported to laboratory on ice at 4°C until use. Before the analysis, soil samples were air-dried, milled and sieved using a 2 mm sieve. The pH value and the characteristics of the soil samples were determined in the Department of Biological Environment, Kangwon National University, Chuncheon 24341, South Korea by following the methods described by Kim et al. [48].

The soil of the experimental field had a sandy loam texture. The M. sinensis and M. sacchariflorus experimental field was irrigated during the initial year of establishment on a regular basis. The research field was located in a temperate monsoon climate, with a wet and humid summer. In the first set of experiments, the harvested rhizomes of M. sinensis and M. sacchariflorus were planted 8 cm deep in the field (100 m × 100 m in size) in the year 2015. The same M. sinensis and M. sacchariflorus accessions were maintained at the experimental field until 2017 and transferred to the plots for herbicides treatment. There were six herbicide treatments and a control (without herbicide treatments) in each plot. Each experimental plot consisted of six rows of 10 m in length, spaced 1 m apart. Each row contained eight plants (in a 10 m length row), with an inter-plant distance of 1.25 m. The treatments consisted of two pre-emergence herbicides and four post-emergence herbicides (applied in recommended standard doses). Pre-emergence herbicides (alachlor and napropamide) were sprayed a week after planting M. sinensis and M. sacchariflorus in the field. Post-emergence herbicides (nicosulfuron, dicamba, bentazon, and glufosinate ammonium) were sprayed 30 days after planting the Miscanthus spp. (Table 1). None treated Miscanthus spp. in the areas were considered as the control for evaluating injury. The field experiments were a randomized block design, each with three replications. All the herbicides were applied within two days of planting Miscanthus plants. Both pre- and post-emergence herbicides were applied using a pressurized backpack sprayer equipped with a flat fan nozzle fitted to a knapsack stainless steel sprayer (MT-009, Taizhou, Zhejiang, China) with the capacity of 18 L at 1.0 MPa. The amount of herbicide sprayed on the Miscanthus spp. was based on the recommended rate. The fresh weight (FW) and dry weight (DW) of the weeds were measured at 30 DAT, after the post-emergence herbicide treatment. Weed density and measurement of FW and DW of weed samples were taken from three randomly selected spots with the help of a quadrat (0.5 m²). After 30 days, all weeds were collected and classified into broadleaf and grass weeds. Different weed species (both broadleaf species and grasses) were counted, and the FW of each weed plant was measured. To measure the DW, the collected weeds were dried at 90°C in an oven for 24 h. Weed densities in both the herbicide-treated and control plots were recorded 30 days after planting of Miscanthus spp. in the field.

Proper control of weeds in bioenergy crop fields is necessary as the infestation of weeds causes competition for available nutrients, water, light, and space, which may cause the inhibition of the growth of bioenergy crops, reducing their biomass [49,50]. These challenges can be addressed by the appropriate application of registered herbicides for weed control [51]. In this study, Miscanthus grown in the control plots without weeding resulted in poor growth and less biomass production compared to weed management by herbicides treatments. This is because the application of herbicides reduced the density of broadleaf and grass weeds in the field, which reduced competition and caused the surplus flow of nutrients to the growing Miscanthus plants. In the present study, comparatively, the biomass of both the M. sinensis and M. sacchariflorus was reduced by the treated herbicides during the first year of weed management (2018), and then in the second year (2019). These results corroborate the report of Everman et al. [52], who observed an increase in the biomass of M. giganteus after the treatment of herbicides in field experiments. In the present study, treatment with pre-emergence herbicides showed a significant variation in the impact on weeds and achieved good control over the broadleaf and grass weeds.

In this study, treatment with the pre-emergence herbicides showed that recommended dose of alachlor when compared to napropamide was more effective in controlling both broadleaf and grass weed species that grew in the M. sinensis and M. sacchariflorus fields. A significant decrease in weed density, FW, and DW was observed in the plot treated with alachlor. Higher efficiency of weed growth control was also reported for other crops treated with alachlor [53]. It has been reported that alachlor causes a reduction in plant growth by inhibiting cell division and cell enlargement [49]. Furthermore, Hemanth Kumar et al. [49] observed nuclear lesions and chromosomal abnormalities in alachlor-treated weeds. Moreover, they observed that the treatment of alachlor caused the inhibition of the activity of elongase and geranylgeranyl pyrophosphate (GGPP) enzymes and inhibited cell division, leading to the ultimate death of emerging weeds. In this study, treatment with post-emergence herbicides effectively reduced the density, FW, and DW of dominant broadleaf and grass weed species grown in the M. sinensis and M. sacchariflorus fields. The efficacy of these herbicides for the inhibition of the broadleaf and grass weeds was higher in 2019 than in 2018 in Miscanthus growing field.

The treatment of Pre-emergence herbicides, partially controls the density, FW, and DW of D. ciliaris, C. album L., and O. biennis L. whereas the treatment of post-emergence herbicides such as Bentazon, dicamba, Glufosinate ammonium completely suppressed these weeds species in Miscanthus field. According to Somerville et al. [54], a shorter length of soil residual activity of pre-emergence herbicides does not provide sufficient weed control when used alone. Moreover, others argued that due to the continuous use of different chemical herbicides in the cultivation land, many weeds have evolved resistance against different herbicides and its mode of action [55]. Other than chemical herbicides, genetic factors, biological characteristics of weed species, agronomic practices and characteristics of herbicides also play important role the evalution of and spread of herbicides resistance in weeds species [55]. In the present study, presence of D. ciliaris, C. album L., and O. biennis L. in the pre-emergence herbicides treated Miscanthus field could be due to insufficient weed control, rather than herbicides resistance of weeds. Others believed that weeds escapes are due to their ability to germinate for a long time during the growing season [56], indicating that the timing of herbicides applications is the key to its efficacy. Based on the present study, the combination of both pre-and post-emergence herbicides is necessary to obtain a good efficacy in total control of weeds in Miscanthus fields.

Among the four post-emergence herbicides, treatment with nicosulfuron inhibited the growth of most of the dominant weeds without affecting the biomass of the growing Miscanthus plants, indicating that the application of nicosulfuron, in particular, is necessary to control the late emergence or escape weeds. Moreover, in a similar study, nicosulfuron has been reported to control weed species without adversely affecting the growth of bioenergy crops [57–59]. In the present study, treatment with nicosulfuron partially inhibited the growth and density of grass weeds D. ciliaris (Retz.) and S. viridis L., which is in line with an earlier report by Zhang et al. [60], who observed nicosulfuron tolerance in D. sanguinalis. It has been reported that nicosulfuron herbicides are responsible for inhibiting acetolactate synthase (ALS), which is located in the photosynthetic tissues (chloroplast and plastids) of plants and is more susceptible in young foliar tissues [61]. In contrast, the application of dicamba and glufosinate ammonium resulted in severe visual injuries and caused various degrees of necrosis, stunting of growth, leaf distortion, twisting of leaves, and reduction in the FW and DW of the weeds in both M. sinensis and M. sacchariflorus fields. In the present study, broadleaf weeds P. asiatica L. and E. canadensis appeared in the glufosinate ammonium-treated Miscanthus plot. In contrast, the treatment with glufosinate ammonium completely inhibited the growth and density of grass weeds. The lower susceptibility of grass weeds to glufosinate ammonium was attributed to the presence of multiple meristematic zones in a single plant, which makes grasses generally possess a higher tendency to survive than broadleaf plants when exposed to glufosinate ammonium [62].

Recently, the application of herbicides has increased in weed management programs due to the scarcity of labour and the higher efficiency of herbicides in controlling their growth. Several studies have reported an increase in the FW of newly established bioenergy crop fields. However, herbicides may degrade the growth of bioenergy crops and cause severe losses in biomass yield. Therefore, it is important to assess the phytotoxicity of herbicides on bioenergy crops. Application of pre- and post-emergence herbicides at the recommended dose caused various degrees of visual injury in the two species of Miscanthus. Treatment with the recommended dose of bentazon resulted in lower phytotoxicity to the growing M. sinensis. However, treatment with bentazon at the same dose caused higher visual injury to the M. sacchariflorus plants. Application of glufosinate ammonium on plants either damaged plant tissues or killed them completely due to disruption in the biochemical and physiological processes of the cells. Both M. sinensis and M. sacchariflorus were susceptible to dicamba and glufosinate ammonium, resulting in a significant reduction in FW and DW compared to the control plants. In the present study, the phytotoxicity of dicamba and glufosinate ammonium was characterized by severe leaf chlorosis in the midrib region. Eventually, the plant showed rapid recovery after 30 DAT. Chlorophyll content of M. sinensis and M. sacchariflorus was significantly reduced by treatment with glufosinate ammonium, bentazon, dicamba, and alachlor compared to the chlorophyll content of plants in the control plot.

In this study, the application of dicamba and glufosinate ammonium in the Miscanthus field showed higher efficacy in reducing the weed density, FW, and DW of both broadleaf and grass weeds. However, application of these herbicides resulted in severe intoxication in both species of Miscanthus, leading to stunting of the plant and necrosis in the foliar part of the plants, which contributed to the reduced FW and DW. Glufosinate ammonium targets glutamine synthetase (the second most abundant protein in plant leaves [63–66] and is responsible for the production of the amino acid glutamine and ammonia detoxification [52]). Inhibition of glutamine synthetase causes decreased glutamine production and rapid accumulation of ammonia in plant tissues [63,67,68] and toxic accumulation of glyoxylate, which inhibits RuBP-carboxylate and carbon dioxide fixation [69]. It has been further observed that an interruption of photorespiration results from a deficiency of intermediates of the Calvin cycle [70,71]. Accumulation of higher levels of ammonia causes acute toxicity to plants, causing severe chlorosis of leaves, suppression of growth, and eventually plant death [72]. According to Takano and Dayan [73], glufosinate ammonium disrupts both photorespiration and the light reactions of photosynthesis, which causes photoreduction of molecular oxygen, which generates reactive oxygen species. Moreover, it can also inhibit carbon assimilation and plant growth [63,74]. In the present study, the application of dicamba and glufosinate ammonium not only suppressed the dominant broadleaf and grass weeds but also reduced the growth of Miscanthus plants. Therefore, before applying these herbicides, all of these factors should be considered. In the present study, M. sinensis showed higher tolerance to nicosulfuron and napropamide. In particular, treatment with nicosulfuron resulted in higher FW and DW of Miscanthus. Moreover, Miscanthus treatment with nicosulfuron caused higher chlorophyll content in both species of Miscanthus, which could contribute to the increased biomass in the plant. According to Barroso et al. [75], some plant species can absorb nicosulfuron in non-toxic compounds and can reduce phytotoxicity. Moreover, in the present study, M. sinensis was less susceptible to bentazon. However, treatment with bentazon resulted in reduced FW and DW, which is possibly due to the insufficient injuries to the broadleaf weeds such as Erigeron canadensis, Artemisia princeps, and Plantago asiatica L., which causes the re-growth of weeds in the M. sinensis field and competition for space, light, water, and nutrition. In the present study, treatment with bentazon resulted in lower chlorophyll content, higher visual injuries, and reduced FW and DW of Miscanthus species. Bentazon, a photosystem II (PSII) electron flow inhibitor, causes changes in the chloroplast ultrastructure, photosynthetic pigment ratios, and levels of chlorophyll proteins in the photosynthetic apparatus of plants [76]. Therefore, in the present study, it is possible that the reduction in photosynthetic pigments could reduce the FW and DW of Miscanthus species. In similar research, Hassannejad et al. [77] observed a lower photosynthetic performance index (PIABS) in Xanthium strumarium L. when treated with bentazon as compared to nicosulfuron.

In the present study, none of the herbicides alone treatments were found effective in controlling all the types of weeds. It was observed that the application of pre-emergence herbicides such as alachlor and napropamide at 951 and 1120 g.a.i.ha⁻¹, respectively, caused less visual injuries and biomass loss in M. sinensis and M. sacchariflorus, while the effect was inadequate to control invasive weeds such as biennis L., D. ciliaris (Retz.) and O. biennis L. The highest weed control was achieved by the treatment of glufosinate ammonium and dicamba at 400 and 482 400 g.a.i/ha⁻¹, respectively, which also showed severe phytotoxicity with a considerable amount of loss in the Miscanthus biomass, but these were often temporary. Treatment with nicosulfuron (40 g.a.i/ha⁻¹) showed good control of the broadleaf and grass weeds at the recommended dose with lower phytotoxicity and caused least visual injuries to the M. sinensis and M. sacchariflorus. The present study can be relevant for weed management in Miscanthus fields with multiple pre- and post-emergence herbicide options.