Improved Bioherbicidal Efficacy of Bipolaris eleusines through Herbicide Addition on Weed Control in Paddy Rice

Bipolaris eleusines was mixed with herbicides to improve the control of barnyardgrass (Echinochloa crus-galli), a noxious weed in rice fields. The compatibility of B. eleusines with herbicides was evaluated for toxic effects on spore germination and mycelium growth in vitro tests, and varied effects were observed with different chemical products. Briefly, 25 g/L penoxsulam OD plus 10% bensulfuron-methyl WP were much more compatible with B. eleusines, and there was no inhibition of spore germination but the promotion of mycelium growth of B. eleusines at all treatment rates. Under greenhouse conditions, the coefficient of the specificity of B. eleusines conidial agent was determined as 3.91, closer to the herbicidal control of 2.89, showing it is highly specific between rice and barnyardgrass. Field experiments in 2011 and 2012 showed that B. eleusines conidial agent displayed good activity on barnyardgrass, monochoria [Monochoria vaginalis (Burm.f.) Presl. Ex Kunth.], and small-flower umbrella sedge (Cyperus difformis L.) and had no negative impact on the rice plant. It also reduced the loss of rice yield when compared with the non-treated control and could make this pathogen a conidial agent for commercial bioherbicidal development in the future.


Introduction
Barnyardgrass (Echinochloa crus-galli) is one of the most noxious weed species around the world [1][2][3]. It may cause severe losses in both yield and quality in rice production [4,5]. Chemical herbicides are generally the mainstay for weed control practices in many countries and are responsible for much of the unparalleled increased crop productivity [6,7]. However, the repeated, intensive, and indiscriminate use of the same herbicides has brought about ecological problems such as weed resistance [8][9][10], agro-ecosystem contamination, and deterioration [11,12]. In addition, the high costs involved in developing and registering chemical herbicides have prompted researchers to investigate alternative systems to control barnyardgrass and alleviate environmental and ecological concerns [6].
Microbial bioherbicide, with a different mode of action on weeds from that of chemical herbicide, is an emerging weed control strategy toward sustainable agriculture [13] and has the potential to replace or substitute some of the current chemical herbicides for the effective control of barnyardgrass [14]. To date, the most biologically effective alternatives to chemical weed control agents that have been extensively evaluated are plant pathogens, more specifically, plant pathogenic fungi such as mycoherbicides [6]. However, mycoherbicides are typically weed species-specific which causes it to be difficult to control multiple weeds at one time [15], and some pathogenic fungi spore types of products have suffered from poor efficacy or control in the field. One possible solution to the problem was the addition of surfactants and other adjuvants into a formulation [16]. Another one was to combine the spore biocontrol agent with chemical herbicide. Therefore, compatibility with herbicides used to control weeds is often studied [17][18][19][20][21]. There were synergistic responses to sub-lethal doses of 2,4-D (2,4-dichlorophenoxyaceticacid) followed 4, 8, or   And 10% cyhalofop-butyl EC, or a combination of 25 g/L penoxsulam OD plus 10% bensulfuron-methyl or pyrazosulfuron-ethyl WP, were much more compatible with B. eleusines and there was no inhibition of the spore germination of B. eleusines at all. However, 50% quinclorac WP, 10% bensulfuron-methyl WP, and 10% pyrazosulfuronethyl WP alone were noticeably more inhibitive, showing 50.6-61.3% inhibition after 12 h exposure at the recommended rates or half of the total.
Relatively, the addition of 10% bensulfuron-methyl, 10% pyrazosulfuron-ethyl WP, or 25 g/L penoxsulam OD plus 10% bensulfuron-methyl WP could promote the radial growth of B. eleusines mycelium at nearly all test rates (Table 1). Moreover, it is a dose-dependent increase both in the survival rate and the growth rate of B. eleusines in response to these herbicides, suggesting that these treatments were not significantly inhibitory at the high rates likely to be encountered in a foliar application. Other herbicide treatments caused no influence or significant inhibition to mycelium growth at the recommended rates or half rates. Tables 2 and 3 presented the weed control efficacy and crop safety of B. eleusines conidial control agent on barnyardgrass and three rice varieties after 4 weeks of treatment.  Rice symptoms of fungal infection were regularly observed 28 d after treatment. There was no visible damage to rice growth found. As shown in Table 2, the conidial agent of B. eleusines was safe at 28 DAT on three rice varieties, including japonica Ribenqing, indica 9311, and glutinous rice Guixiangsinuo P106. The fresh weight inhibition rate fluctuated between −5.35~0.38%, −6.02~−2.26%, and −5.15~0.24%, respectively, but all were lower than 10%. There was no significant difference compared to herbicide control. It indicated that their ED 10 was higher than 360 g a. i. ha −1 .

Weed Control of B. eleusines Combined Herbicides in Greenhouse
As shown in Table 3, the inhibition rate of the fresh weight of 3.75~180 g a. i. ha −1 B. eleusines conidial agent on barnyardgrass remained between 6.38~100%, while the herbicidal activity of 0.72~45 g a. i. ha −1 25 g/L penoxsulam OD stood at 2.88~100%. The result of the statistics regression (Dose-response analysis) indicated that the ED 90 values of the B. eleusines conidial agent and the control agent of 25 g/L penoxsulam OD were 92.06 g a. i. ha -1 and 23.33 g a. i. ha −1 , respectively. The coefficient of the specificity of the B. eleusines conidial agent was higher than 3.91 and closer to the herbicidal control of 2.89 and showed its high specificity between rice and barnyardgrass. It also suggested that the B. eleusines agent was safe for rice and highly efficient on barnyardgrass in the greenhouse.

Weed Control Efficacy of B. eleusines Conidial Agent with Herbicides under Field Conditions
Two-year field tests for the weed control efficacy and effect on the rice crop of B. eleusines conidial agent application were carried out at different experimental fields of Fuyang Experiment Station and the Shanghai Seed Breeding Center in 2011 and 2012, respectively.
There was no infection of rice plants by B. eleusines with eye observation and no adverse effect on the growth and development of rice plants in both experimental fields. Three weed species occurred to a substantial extent in density in the test field, i.e., barnyardgrass, 13.8 plants m −2 , small flower umbrella sedge (Cyperus difformis L.), 12.6 plants m −2 , and monochoria (Monochoria vaginalis), 7.0 plants m −2 in the Shanghai Seed Breeding Center. The occurrence of these weeds reduced rice yield (rough rice) by 15.2% in 2012 in the nontreated plot as compared with the rice yield of plots (5500.9 kg ha −1 ) where a commercial mixture of butachlor that is popular with farmers in China was treated as a control ( Table 4). The application of herbicide butachlor resulted in density reduction in barnyard grass by 68.0%, small-flower umbrella sedge by 95.1%, and monochoria by 63.5%. Under such a situation of weed occurrence and the weed control efficacy of this herbicide, the spray of the B. eleusines conidia at a concentration of 60 g a. i. ha −1 showed a reduction in the density of barnyardgrass by 93.9%, small-flower umbrella sedge by 99.3%, and monochoria by 83.8% (Table 4). The application of more B. eleusines conidia did not significantly result in more control of barnyardgrass and small-flower umbrella sedge, the overall weed control efficacy being superior to that of the herbicide application.  The same experiment result was obtained in field lots of the Fuyang Experiment Station. We also found that there was nearly no effect on the weeds in synergist-free B. eleusines formulation treatment compared to treatment with a synergistic agent (Table 5). It suggested that there was a positive synergetic effect to improving B. eleusines conidia formulation efficiency on weeds with herbicides of penoxsulam plus bensulfuron-methyl.

Discussion
Chemical herbicides are the most commonly utilized crop-protection compounds in the world and have played an important role in effectively and quickly repressing weed populations and significantly reducing labor intensity since it was invented. However, concerns have been gradually raised about environmental risks due to the long-term heavy use of herbicides. It could cause unsustainable ecological practices including resistant weeds' population increase and diversity decline in the rice paddy. Compared with it, bioherbicides could be more environmentally friendly and not easily lead to naturally herbicide-resistance weeds. Its disadvantage focused on the lower efficacy in the field. It might also need more time to experience fungal adsorption, colonization, germination, and growth at the host. Therefore, the weeding effect is slower than chemical herbicides. The integration of different action modes of pathogenic fungus and chemical herbicide could not only synergistically enhance the virulence of the bioherbicidal pathogens, but also significantly decrease the risk of occurring herbicide-resistance weeds. This strategy could synchronously meet the demands of government policies and consumer trends to reduce the use of synthetic pesticides and increase the use of nonchemical control methods [31]. These studies showed that a combination of B. eleusines with a lower rate of penoxsulam plus bensulfuron-methyl could effectively control barnyardgrass, monochoria, and smallflower umbrella sedge, and had no negative impact on the rice plant under field conditions. It suggested that this pathogen conidial agent had the potential to develop a commercial bioherbicide in the future.
The germination assay on agar provides an initial estimate of compatibility between herbicides and pathogens [32,33]. In this study, 50% quinclorac WP caused immediate losses in viability after 12 h of incubation and severe inhibition to the bioherbicide agent B. eleusines in the in vitro tests. However, 25 g/L penoxsulam OD and 10% bensulfuron methyl WP were not inhibitory to it. On the contrary, there was a dose-dependent increase both in the survival rate and the growth rate of B. eleusines in response to these two herbicides, so a tank mix of 25 g/L penoxsulam OD and 10% bensulfuron methyl WP with B. eleusines should be a better option to control weeds in the further greenhouse and field conditions because of the high compatibility.
The combinations of bioherbicides and synthetic herbicides can be synergistic in much research [33,34]. Some herbicides could lower the inherent defenses of the weed by inhibiting specific plant defense pathways such as pathogen-induced phytoalexin biosynthesis (phenyl-propanoid phytoalexins and steroid phytoalexins) and callose biosynthesis that stop pathogen attack, as well as many other defenses [15,35]. For example, Sharon found that a low concentration of Alternaria cassiae conidia elicited a hypersensitive response on the weed but not the crop after being sprayed with low levels of glyphosate together which suppressed weed phytoalexin biosynthesis [36]. Wymore et al. (1987) reported that coapplications of C. coccodes Wallr. and the herbicide thidiazuron to velvetleaf (Abutilon theophrasti Medic.) increased pathogen infection and weed control compared with either component applied alone [37]. Similarly, Peng et al. (2005) [33] also reported that coapplying propanil, quinclorac, or sethoxydim at a one-quarter rate with the pathogen at the sublethal dose of 2 × 10 7 spores/ml achieved complete control of green foxtail, because sethoxydim, with its ability to inhibit cell division, is highly effective on young and actively growing tissues [38]. In this study, there was a positive synergetic effect to improve B. eleusines conidia formulation efficiency on weeds with herbicides. When B. eleusines conidia were sprayed at a concentration of 60 g a. i. hm −2 combined with a sub-rate synergist, it showed a reduction in the density of barnyardgrass by 93.9%, small-flower umbrella sedge by 99.3%, and monochoria by 83.8% (Table 4). Briefly, 25 g/L penoxsulam OD and 10% bensulfuron methyl were used as synergetic herbicides with B. eleusines. It seemed that penoxsulam and bensulfuron methyl could lower weed defense responses, making weeds more susceptible to pathogen attack. Therefore, a lower dose of bioherbicide agent could effectively control weeds. However, the specific synergetic mechanism needs to be further studied.
Herbicide-resistance weeds resulting from pesticides abusage in the recent several decades have caused a great threat to rice production. It could lead to no herbicide available for some superweeds. Penoxsulam is a post-emergence triazolopyrimidine sulphonamide herbicide registered for weed control in rice crops [39,40] and bensulfuron-methyl is a representative sulfonylurea herbicide that controls broadleaf weeds [41]. Both of them could inhibit acetolactate synthase (ALS) activity in susceptible plant species, and appli-cation alone easily produced herbicide resistance in agricultural practices [42,43]. The integration application of different action modes of bioherbicide agent B. eleusines could potentially delay the development of herbicide resistance and greatly extend the shelf life of the herbicide.
Compared to the greenhouse test, there was a decline in weed control efficiency by the B. eleusines agent in the field application. Schnick and Boland (2004) also found that the dandelion response was reduced after treatment with sublethal doses of 2, 4-D followed later by P. herbarum inoculation under field conditions [22]. The environmental conditions or other factors play a role in efficacy after combining fungal pathogens and herbicides. Therefore, some UV protective agents and others need to be added to enhance the stability and efficacy of B. eleusines in the field application.
The meristems of many types of grass are protected by leaf sheaths [44]. It was observed that the young leaves of barnyardgrass tend to be more resistant to B. eleusines despite severe injuries on lower leaves, so barnyardgrass treated with B. eleusines alone could not be effectively killed and often restored 14 days after treatment. However, the efficacy of weed control is significantly improved by the addition of synergetic penoxsulam and bensulfuron methyl. This is similar to reports by [33]. They found that green foxtail treated with P. setariae alone often recovers from initial damages. However, after being combined with sethoxydim, with its ability to inhibit cell division, they became highly effective on young and actively growing tissues [38]. These synergetic herbicide effects might be complementary to the mode of action by bioherbicide agents on weeds.

Fungal Inoculum
Bipolars eleusines was isolated from severely diseased barnyardgrass and preserved at 4 • C in the Weed Lab at the China National Rice Research Institute (CNRRI), Hangzhou, China [30]. Conidia of B. eleusines were produced on solid media (rice flour 4%, soybean meal 1%, Na 3 PO 4 ·12H 2 O 0.2%, MgSO 4 ·7H 2 O 0.1%, distilled water 40%, and 100 mL perlite) in a pallet of 45 cm (length) × 35 cm (width) × 5 cm (height) at 28 • C for 2 weeks and collected by a Mycoharvester of the Solid State Fermentation (Nanjing Institute of Agricultural Mechanization, Ministry of Agriculture and Rural Affairs, Nanjing, China) for use in further experiments.
A conidial agent of B. eleusines was prepared with 1 g spore powder by the addition of 0.25 mL soybean oil, sodium carboxymethylcellulose (CMC) 0.49 g, SP-20 0.1 mL, synergist A (10% bensulfuron methyl WP) 6.56 mg, and synergist B (25 g/L penoxsulam OD) 26 µL. The conidial agent of B. eleusines without synergists A and B was prepared as a synergist-free control.

Plant Preparation
Barnyardgrass seeds, harvested at maturity from the Fuyang Experiment Station of China National Rice Research Institute, Hangzhou, China, were planted onto soil where herbicides had not been used and overlaid on a 0.5~1.0 cm fine sandy soil in plastic pots (7.5 cm diameter, 6.0 cm deep). Rice seeds (Japonica Ribenqing, Indica 9311, and Glutinous rice Guixiangsinuo P106, respectively) supplied by the Seed Bank of the China National Rice Research Institute were sown after soaking and sprouting. After sowing, the soil in each pot was moistened with tap water. Then, the pots were placed in a greenhouse maintained at a temperature between 20~35 • C and humidity greater than 50%. Two days after emergence, the seedlings were thinned to 11 plants per pot and grown for foliar treatment until the rice plants reached the 3~3.5-leaf stage and the barnyardgrass reached the 2.5-leaf stage.

Viability of B. eleusines Conidia with Herbicide Mixtures
Commercial formulations of herbicides used in this study were detailed in Table 6. Sterile centrifuge tubes were prepared with B. eleusines spores and Tween-20 (2.5% vol/vol) in water. Fresh stock suspensions of the herbicides were prepared before the experiment. The maximum use rates of the herbicides were determined by the manufacturer's label guidance and calculated assuming a 450 L ha −1 application volume. After a brief preincubation to disperse the spores, herbicide solutions were added from the stock solutions to yield the maximum labeled application rate, 1× (Table 6), or 0.5×, or 0.25×, or 0.1× (vol/vol). Each herbicide-rate treatment was hanging dropped onto one side of the bi-concavity slide, and the herbicide-free spore suspension was dropped onto another side as the control and then put into the sterilized Petri dish (Ø 90 mm) with absorbent filter paper to incubate moistly. After 12 h of incubation, the germination of the spore was observed microscopically, and we calculated the inhibition rate of spore germination with the following formula: Inhibition rate of spore germination (%) = (1 − Nx/Ny) * 100%. where Nx and Ny = data from the treated and non-treated control spore, respectively. Each treatment was replicated three times and the experiment was conducted twice.

Effect of Herbicide Mixtures on Mycelium Growth of B. eleusines
The chemical herbicides were diluted with sterilized water to the required concentration. A volume of 1 ml of each dilution was added into sterilized medium (2% glucose, 0.5% KNO 3 , 0.2% NaPO 3 12H 2 O, 0.1%MgSO 4 , 0.5% soybean powder, 1.7% agar, and water 99 mL) as herbicide media with the maximum labeled application rate of 1× (Table 6), 0.5×, 0.25×, 0.2×, or 0.1× (vol/vol), adequately mixing and putting in Petri dishes. The same volume of water was controlled. After solidifying, the mycelial discs (0.5-cm-diam agar plug) obtained from the margin of an actively growing colony of B. eleusines on PDA (Potato Dextrose Agar) plate culture media was inoculated on the center of the herbicide media and kept at 28 • C under constant darkness for 14 days. Each herbicide-rate treatment was conducted with three replications. The colony diameters of B. eleusines were determined after 14 days of treatment. The percentage of the growth inhibition rate was calculated by comparison to herbicide-free treatment.

Pot Experiment in Greenhouse
To investigate the effects on rice and barnyardgrass under different dosages of the B. eleusines conidial agent after foliar spraying treatment, a pot experiment was conducted in the greenhouse.
The conidial agent was sprayed using a spray tower 3WPSHZ-500 (Nanjing Institute of Agricultural Mechanization, Ministry of Agriculture and Rural Affairs) at 200 kPa air pressure and a distance of about 20 cm from the plants with 0.097 m 2 of the spray area. The B. eleusines conidial agent was sprayed at the rates of 3.75, 7.5, 15, 30, 60, 90, or 120 g a.i. ha −1 for weed control activity and 45, 90, 135, 180, 270, or 360 g a. i. ha −1 for rice safety evaluation, respectively. The pot was irrigated at 3 to 5 cm deep at 2 d per treatment. A commercial herbicide of penoxsulam (25 g/L a. i. by control) was applied at a rate of 0.72, 2.25, 5, 5.63, 6.75, 8.28, 22.5, or 45 g a. i. ha −1 for weed control activity and 2.25, 6.25, 11.25, 22.5, 45, or 67.5 g a. i. ha −1 for rice safety evaluation. Approximately 10 mL of conidial suspension was applied to each pot, resulting in slight runoff from the plant foliage. The control pots were sprayed with distilled water. The experiment was conducted in a completely randomized block design with four replicates for each treatment. The pots with inoculated plants were immediately covered with a plastic bag for 24 h in the greenhouse. All treatments were irrigated at 3 to 5 cm deep at 1 day per treatment.
The disease severity, the amount of viable rice and barnyardgrass, and the aboveground fresh weight of rice and barnyardgrass were recorded 4 weeks after treatment (WAT). The bioherbicideal efficacy or the inhibition rate of rice was calculated using the following formula: Reduction in fresh weight (%) = (1 − Nx/Ny) * 100%, where Nx and Ny = data from treated and non-treated control plants, respectively. The coefficient of specificity was calculated using the following formula: Coefficient of Specificity = ED 10 of crop/ED 90 of weed. In 2011, a conventional indica rice Yangdao 6 was cultivated as a single crop growing from May to September. In 2012, the conventional early maturing late japonica rice, "Wuyunjing 7", was cultivated from July to December. Three-week-old rice seedlings were transplanted with 25 cm by 20 cm spacing in plots measuring 5 m by 4 m for each. The field was drained 7 days after transplanting the rice seedlings, and the rice plants were sprayed with the conidial agent of B. eleusines. The conidial agent combined with herbicides was sprayed at the rates of 60, 90, 120, or 180 g a. i. hm −2 in a spray volume of 450 L ha −1 water. A synergist-free conidial agent was sprayed at the rates of 90 g a. i. hm −2 as a control in the Fuyang Experiment Station. The field was irrigated again at 3 to 5 cm deep at 2 days after treatment. A commercial herbicide of butachlor (60% a. i. by weight) that is popular with farmers for weed control in rice fields in Zhejiang and Shanghai, China was applied at a rate of 990 g a. i. of butachlor per ha, 5 days in Shanghai and 7 days in CNRRI after transplanting. The non-treated control was sprayed with water. Manual weeding was carried out 15 days after application. The experimental units were arranged in a randomized block design with three replications. The density and aboveground fresh weights of weeds per 0.25 m 2 were recorded at 4 WAT. The weed control efficacy, i.e., the reduction in density or weight of the weed, of the treatments was calculated as described previously. The whole plot was manually harvested, and the rice grain yield was recorded.

Data Analysis
All data analysis was performed using an SPSS 13.0 statistical package. The statistical comparisons of data for the percentage of the inhibition rate or reduction rate in Table 1,  Table 2, Table 4, and Table 5 were performed via analysis of variance, and the data in Table 3 through regression analysis. The treatment means in Tables 1 and 2, after uniformly normalizing to positive decimals, then square root, and arcsin transformation to normalize the variance, were compared by Tukey's multiple range tests at a 5% level of significance. The treatment means of the percentage of the reduction rate in Tables 4 and 5 were compared by Duncan's multiple range tests at a 5% level of significance after the same conversion. The results were back-transformed to the original ratings for presentation. The statistical analyses for the rice yield in Tables 4 and 5 were performed without transformation.

Conclusions
In brief, 25 g/L penoxsulam OD plus 10% bensulfuron-Methyl WP were compatible with B. eleusines in an in vitro test. The conidial agent of B. eleusines, by the addition of 25 g/L penoxsulam OD plus 10% bensulfuron-Methyl WP, was safe on three rice varieties, including japonica, indica, and glutinous rice, and showed high specificity between rice and barnyardgrass under greenhouse conditions. Field experiments showed that B. eleusines combined with chemical herbicides could improve the biological efficacy of the B. eleusines conidial agent. These results were critical for the commercial application of the B. eleusines conidial agent in the future.