Introduction

Wheat (Triticum aestivun L.) is one of the three highest production cereals in the world along with corn and rice, and it is used to produce food products and balanced diets in livestock. In 2016, in Mexico, the harvested area of wheat grain was 612, 866 ha with a production of 3, 606, 144 t. The states with the largest area sown that year were Sonora, Baja California, Sinaloa, Guanajuato and Michoacán (^{SIAP, 2016}).

Wheat production is affected by the interference of weeds, which are one of the main limitations that reduce yields per surface unit (^{Ross and Lembi, 2009}). ^{Cruz et al. (2004)} showed that when weeds competed with wheat throughout the cycle, yield decreased by 90 %. Within the crop there is a great variety of weeds that differ according to the region of the country, but there are some very aggressive in most of the producing areas of this cereal.

In the state of Guanajuato, the most important weeds in the wheat crop are: of narrow leaf, Avena fatua L., Phalaris minor Retz., P. paradoxa L. and Echinochloa spp; and of broad leaf, Bidens spp., Rhaphanus raphanistrum L., Brassica campestris L., Helianthus annuus L., Brassica nigra (L.) W.G.J. Koch, Chenopodium album L. Bosc ex Moq., Sonchus oleraceus L., and Malva parviflora L. Echinochloa crus-galli occurs mainly in the rainy season and can cause total losses in yield if no efficient control is carried out.

Due to the type of wheat planting, it is difficult to apply an alternative control method to the use of herbicides (^{Tafoya et al., 2009}; ^{Tafoya and Carrillo, 2009}). Chemical control is practical, efficient and relatively inexpensive, but its repeated use leads to a selection of resistant weed populations (^{Powles and Yu, 2010}). The development of resistant populations occurs through the selection pressure imposed by the frequent use of one or more herbicides with the same action mode (^{Christoffers, 1999}, ^{Fischer, 2013}), the biological characteristics of the weed, the specificity of the herbicide, the diversity of the resistance genes involved and herbicide efficacy (^{Valverde et al., 2000}; ^{Cerdeira and Duke, 2006}).

^{Heap (2016)} points out that in the world there are 47 weed species resistant to inhibitor herbicides of the enzyme acetyl coenzyme A carboxylase (ACCase) and 159 resistant to herbicides inhibitors of the enzyme acetolactate synthase (ALS), in both cases species of the genus Echinochloa are included. Of these, 19 biotypes have developed resistance to herbicides that inhibit ALS and 17 to herbicides that inhibit ACCase. Echinochloa crus-galli is considered the sixth weed most resistant to herbicides after Lolium rigidum Gaud., Amaranthus palmeri S. Watson, Conyza canadensis (L.) Cronquist, Avena fatua L. and A. tuberculatus (Moq.) J.D. Sauer (Heap, 2014).

^{Mallory-Smith et al. (1990)} and ^{Devine (1997)} mention that some herbicides, such as ACCase inhibitors, can select resistant biotypes in one to five generations. This limited number of generations is due to the high specificity of action site, the high frequency of mutations in the nuclear gene that encodes the enzyme and the possibility that different semi-dominant mutations alter the site of coupling of the herbicide in the enzyme (^{Gressel, 2002}, ^{Tranel and Wright, 2002}, ^{Délye, 2005}). The rapid increase in resistance to ALS inhibitor herbicides is attributed in part to the high frequency of natural mutation in the center of action of the enzyme (^{Devine and Preston, 2000}).

In the state of Guanajuato the control of E. crus-galli has been deficient (visual appreciation of the first author), which is attributed to the resistance development of this species to the herbicides of traditional use in the region, however there is no scientific evidence. Knowledge of the presence of resistance is necessary to determine the actions in the weed control, which range from the use of herbicides with different modes of action, to alternative control methods (mechanical, manual, biological and legal). Therefore, the objective of the present research was to determine the biological efficacy of the herbicides commonly used in wheat crop and the resistance presence to these herbicides in E. crus-galli collected in the state of Guanajuato. The hypothesis was that there are biotypes of E. crus-galli that have developed resistance to the herbicides used in Guanajuato for the control of narrow leaf weeds in cereals.

Materials and Methods

During the spring of 2015, in the state of Guanajuato, we collected samples of five biotypes of E. crus-galli in commercial wheat plots in which herbicides were applied, and this species was not controlled (Table 1). The biotype considered susceptible was collected in an area surrounding the initial section of an irrigation water distribution channel, where we were certain that no herbicides were used before. Then, we separated the seeds (approximately 200 g per collection site) from the rest of the infrutescence by using stainless steel sieves and stored them in paper bags under laboratory conditions until they were used.

We placed the seeds in a germination chamber (APT.line^® KBWF E5.2, Binder, Tuttlingen, Germany), and exposed them at 38 °C for 24 h. After removing them from the chamber, we placed them in 10 cm diameter Petri dishes on filter paper (Whatman No. 1), and in each Petri dish we applied 20 mL of 0.3 % p/v KNO₃ solution.

We put the boxes again in the germination chamber at 30 °C for 24 h, under constant light and 80 % relative humidity; then we washed the seeds with distilled water to remove the remains of the KNO₃ solution and kept them in the germination chamber, under the same conditions for 6 d. When the seeds germinated and the seedlings reached 2 cm in height, we transplanted them in plastic pots with a capacity of 500 mL filled with a mixture of field soil and peat moss substrate (Promix flex^®) in a ratio of 7:3.

Results and Discussion

Biological efficacy

In the biological efficacy phase of the four herbicides, we observed significant statistical differences due to the effect of the herbicides in the control of E. crus-galli biotypes at 30 DDA. ACCase-inhibiting herbicides (clodinafop propargyl and pinoxaden) showed good control of all the E. crus-galli biotypes examined in this evaluation and in the two previous ones (data not shown). The control was on average 99 %, with the exception of biotype III, where the control was 95 % (Table 4 and Figure 1). These results were decisive to rule out the suspicion of resistance to these herbicides.

Table 4 Control percentages (30 DDA) of the E. crus-galli biotypes of the state of Guanajuato for the evaluated herbicides. 

Tratamiento	Biotipo
	Sus	I	II	III	IV	V
Flucarbazone sódico	75.00 b	60.00 c	42.75 b	25.50 b	34.12 b	55.12 b †
Clodinafop propargil	99.50 a	99.50 a	99.50 a	94.43 a	95.50 a	99.50 a
Pinoxaden	99.50 a	99.50 a	99.50 a	95.83 a	99.50 a	99.50 a
Meso.+ Iodos¶	79.37 b	79.37 b	42.75 b	25.50 b	34.12 b	25.50 c
Testigo absoluto	0.50 c	0.50 d	0.50 c	0.50 c	0.50 c	0.50 d

^† Means with different letter in a column are statistically different (Tukey; p ( 0.05).

^¶ Mesosulfuron methyl+iodosulfuron methyl

Figure 1 Effect at 30 d after of the application of ACCase inhibitor herbicides (clodinafop propargil and pinoxaden) on the biotypes of E. crus-galli, collected in Guanajuato. 

The ALS inhibitor herbicides (sodium flucarbazone and mesosulfuron methyl+iodosulfuron methyl) provided an average control of 77 % in the susceptible biotype (Table 4).

This percentage of control is considered regular, according to the European scale in use. However, with the exception of biotype I, in the other biotypes its effectiveness was low (Figure 2), so we could confirm the suspicion of resistance of these biotypes to the aforementioned herbicides.

Figure 2 Effect at 30 d after of the application of the herbicides inhibitors of ALS (Meso. + Iod. = Mesosulfuronmetil+iodosulfuron-methyl and sodium flucarbazone) in the biotypes of E. crus-galli, collected in Guanajuato. 

The determinations of dry biomass confirmed the results obtained from the final visual evaluation. The biotypes treated with pinoxaden and clodinafop propargil had the lowest dry biomass, that is, they were the treatments that exhibited the greatest control.

In contrast, the dry biomass recorded for flucarbazone and the combination of mesosulfuron methyl+iodosulfuron methyl was high in the biotypes suspected of having resistance (Figure 3).

Figure 3 Dry biomass of plants from the biotypes of E. crus-galli per experimental unit 30 d after the application of the herbicides inhibitors of ALS and ACCase. Sus = susceptible 

The first author's observations about the biological efficacy of ACCase-inhibitor herbicides can be attributed to poorly performed applications when not using the appropriate equipment, particularly the flat fan tips recommended for the application of herbicides. In addition, bad equipment calibration, which can lead to the application of inadequate doses (^{Salas, 2001}).

Conclusions

The clodinafop propargil and pinoxaden herbicides showed good biological effectiveness on E. crus-galli biotypes from Guanajuato; the herbicides flucarbazone and mesosulfuron methyl+iodosulfuron methyl showed an acceptable effectiveness on the susceptible biotype and low on the biotypes with suspected resistance. In two of the six biotypes studied, we confirmed the E. crus-galli resistance to the herbicide mesosulfuron methyl + iodosulfuron methyl.