Herbicide resistant barnyardgrass in Iran and Turkey

Background: Inconsistent control of Barnyardgrass has been reported repeatedly by farmers in major rice growing areas of Iran and Turkey. Objective: Thus, a greenhouse study was conducted to investigate the resistance of Barnyardgrass to acetyl CoA carboxylase (cyhalofop-butyl) and acetolactate synthase (penoxsulam, bispyribac-sodium) inhibiting herbicides. Methods: The seeds were sown in pots in a greenhouse and after screening, were sprayed with various rates of cyhalofop-butyl, penoxsulam and bispyribac-sodium herbicides at 3-4 leaf stage. Four weeks later, the above ground biomass was cut, dried in an oven and weighted. The results were then analyzed in the R software (drc package) using a four or three parameter log-logistic function. All experiments were repeated twice. Results: While 30% of biotypes collected from Turkey were not controlled by ACCase inhibitors at twice the recommended rates, no resistance was observed in Iran’s samples. Only one biotype exhibited cross-resistant to penoxsulam and bispyribac-sodium in Iran, which was due to consecutive application of herbicides with modes of action similar to these herbicides. In contrast, several Turkish biotypes showed cross-resistance as well as multiple resistance to ACCase and ALS inhibitors. Conclusions: The rapid expansion of herbicide resistance in both countries indicate the necessity of adopting integrated weed management practices to hinder the further evolution of resistance in future.


INTRODUCTION
Rice has become the main crop produced in the eastern part of the world due to the key role it plays in dietary pattern of these regions. In 2013, around 493000 ha of rice were grown in northern parts of Iran, approximately 85% of Iranian rice production. In Turkey, 30000 ha of rice were grown in Samsun, Sinop and Corum, which was about 25% of total rice production in Turkey (FAO, 2018). Rice cultivation is quite different in these two countries. For example, in Iran transplanting and in Turkey direct seeding methods are widely practiced.
The genus of Echinochloa has 50 species (including subspecies and varieties) which are the Planta Daninha 2020;38:e020227592 -https://doi.org/10.1590/S0100-83582020380100060 2/8 most important weeds occurring in rice fields around the world (Andres et al., 2007). The broad ecological adaptation such as the ability to mimic rice, rapid germination, growth and abundant seed production make this species a successful weed in rice production system (Kaya Altop et al., 2014).
Echinochloa crus-galli possesses the C4 photosynthetic pathway, spreads and reproduces through seeds and is one of the world's most serious grass weeds (Rao et al., 2007). This weed can tolerate flooding or poor drainage, but is unable to bear serious drought (Bajwa et al., 2015). Yield losses caused by weeds may range from 15 to 42% depending on weed densities, weed species, seeding method and rice cultivar (Mennan et al., 2012). Bajwa et al. (2015) reported that E. crus-galli may reduce rice yield between 21 to 79%, depending on the management and cropping system. Weed management is a major issue for rice producers. Application of herbicide is the best option to control E. crus-galli in Turkey and Iran. Currently the usage of this controlling system has decreased in Turkey and some other countries due to the elevated extent of herbicide resistance (Mennan et al., 2011). Rice-specific herbicides such as propanil were introduced in last 26 years. Since then, the number of registered herbicides to control weeds in rice has increased in Turkey and Iran. Some of these herbicides are single chemicals, but during last 5 or 10 years products with combination of two to three chemicals became popular. Recently, also, 'one-shot' (achieving desired control with a single usage) single herbicides including sulfonylurea compounds such as Inhibition of acetyl CoA carboxylase (ACCase), acetolactate synthase ALS (acetohydroxyacid synthase AHAS) and few others have been the most common. According to Heap (2020), E. crus-galli is reported to be resistant to many herbicides including cyhalofop-butyl (ACCase inhibitors, A/1), bispyribacsodium and penoxsulam (ALS inhibitors, B/2), in various parts of the world, including Brazil, China, and Egypt, among others.
Due to consecutive application of herbicides common in rice cultivation, E. crus-galli has evolved resistance in northern provinces of Iran (Haghnama and Mennan, 2017). Thus, this study was conducted to examine E. crus-galli biotypes in terms of resistance to bispyribac-sodium, cyhalofop-butyl and penoxsulam, existence of multiple or cross resistance, assessment of resistance to herbicides commonly associated with rice production in Iran.

Seed source
Mature seeds of E. crus-galli biotypes were collected from different rice growing regions of Iran and Turkey. In 2013, 56 different samples were collected from three provinces of Turkey and from the fields with a long history of herbicide use where weed control problems had been reported. Seeds were collected from field after full maturity. In Iran, seeds of 128 E. crus-galli populations were gathered from three provinces. All seeds were cleaned and stored in a laboratory to be used in the experiments. Susceptible biotypes were gathered from areas of each region with no history of being sprayed with the studied herbicides. It must be noted that the mechanisms responsible for the resistance of these biotypes had been determined previously (Haghnama, 2016).

Screening assay
In order to rule out the populations falsely reported as resistant, a single-dose experiment has been carried out according to Moss (Moss et al., 1999) to estimate possible resistance within 184 biotypes. If the herbicides effectiveness at recommended dose was less than 80%, these samples were involved in the subsequent dose-response experiments.

Dose-response experiments
Based on the single-dose-assay results, the putative tolerant as well as susceptible biotypes were seeded into cell trays containing commercial potting mix. Plants were grown at average daily temperatures ranging from 24 to 30 o C and at a 16 h day length photoperiod and 80% relative humidity. After germination, seedlings were transferred into 0.3 L square plastic pots filled with rice paddy field soil. Each pot was fertilized with ammonium sulphate (21% N, 24% S) at rates of 350 kg ha -1 in two split doses before and after transplanting and first tillering, respectively. When seedling established, each pot contained 4 equidistantly spaced uniform plants and herbicides were applied at the 3-4 true leaf stages.

Data analysis
Since no interaction was observed between the experimental runs and the herbicide treatments, the results of the two experimental runs were pooled.
Typically the data are best described with a fourparameter log-logistic curve (Ritz, 2010). y = in which y is biomass, C and D are lower and upper value of y, respectively, b is slope of the curve at ED50 point, which is the dose required to halve the biomass relative to D.
Since C parameter was not significant, the equation was altered to: y = Analysis of the dose-response curves was performed using R (version 2.15.2) with the add-on package drc (version 2.03.0). The multiple curves have been fitted simultaneously, as well as, each set quality of dose-response models has been differentiated with an analysis of variance through a lack-of-fit F-test (Seefeldt et al., 1995). The ED50 response level have been calculated from the model and graphical analysis of residuals has been used to estimate regression fits (Ritz, 2010).
Resistance index (RI) and Resistance-Susceptible (R/S) ratio were scored based on following formula (Ritz, 2010).

RI
Resistant biotypes of each province were evaluated with the susceptible biotypes collected from that same region.

RESULTS AND DISCUSSION
The biotypes which maintained 80% dry weight compared to control under recommended dose of herbicide were chosen as resistant ones and underwent dose-response assay. No cyhalofop-butyl resistance was observed among biotypes collected from Iran, and only one biotype showed resistance to penoxsulam. Bispyribac-sodium resistance was more common though. Nevertheless, resistance was more severe in Turkey, and several biotypes proved to have developed resistance to studied herbicides (Tables 1 and 2).
Results of dose-response assay (Figures 1, 2 and 3) showed that the sole biotype resistant to penoxsulam in Iran (GOL-59) had a resistance factor of 7.07. Herbicide rates leading to 50% reduction in plant dry weight were 23.63-85.41 g a.i. ha -1 . Thus, resistance factor of bispyribac-sodium resistant biotypes collected from Iran ranged from 2.21 to 8.00. Resistance factor and ED50 of Turkey biotypes which had evolved resistance to penoxsulam and bispyribacsodium were respectively 37.31-60.76 g a.i. ha -1 , 2.30-3.74 and 38.86-70.31 g a.i. ha -1 , 2.05-3.70 (Table 1). Biotypes resistant to cyhalofop-butyl herbicide collected from Turkey showed ED50s of 193.2-471.4 g a.i. ha -1 . Since the variation among susceptible biotypes collected from each province were relatively more than the ones associated with penoxsulam and bispyribac-sodium, the resistance factors were in some cases not in co-ordinance with ED50 values. Nonetheless, resistance factors ranged from 2.20 to 5.37 (Table 2). According to Table 3 to evolution of further weed resistance in the future. In case of multiple resistances to herbicides and due to the limited number of alternative herbicides, the cost of preventive operation would be significantly less than fighting against evolution of resistance. When weed control in first year is poor, farmers will incur great costs to dispel resistance, but if alternate affordable herbicide is available, resistance control cost would reduce (Peterson, 1999). Rubin (1997) defined cross-resistance as a case where a weed biotype is resistant to two or more herbicides acting at the same primary site of action Each data point represents the average of the two experiments.   and multiple resistances, when a biotype is resistant to two or more herbicides functioning at different sites. ALS inhibitors such as bensulfuron-methyl, cinosulfuron and ethoxysulfuron are used in some specified areas, but there is no history of bispyribacsodium application to rice fields in Iran. Apparently, bispyribac-sodium resistance outbreak in E. crus-galli in Golestan province, Iran may be considered as cross-resistance. Resistant genes exist in the plant before selection (the initial frequency of resistant genes). Hence, if the initial frequency of resistant gene is high, resistance development will be fast. Also, there are numerous reports of natural resistance in weeds to ALS inhibitors (Preston et al, 2006).
Results also showed that, in Iran, especially in Golestan province, a greater percentage of samples were resistant to both ALS inhibitors. Northern provinces of Sinop in Turkey possessed a less ratio of weeds resistant to both ALS and ACCase inhibitors than Samsun and Corum. Since cyhalofop herbicide has been used consecutively in turkey, resistance to this herbicide is within logic. This is also the case for lack of cyhalofop-resistance in Iran, as this herbicide has not been used in the country. Resistance to ALS inhibitors bispyribac sodium and penoxsulam in Iran may be justified with the fact that there is history of some ALS herbicides being continuously applied in crops cultivated in rotation with rice, which has eventually led to evolution of cross-resistance to the studied herbicides.

CONCLUSIONS
The rapid expansion of herbicide resistance in both countries indicate the necessity of adopting integrated weed management practices to hinder the further evolution of resistance in future.

CONTRIBUTIONS
KH: data curation, analysis, writing of the manuscript. HM: supervision, editing of the manuscript.