Cross-resistance to auxinic herbicides in dicamba-resistant Chenopodium album

ABSTRACT

The responses of two dicamba-resistant Chenopodium album (fathen) populations (L and M) were compared with the responses of two dicamba-susceptible populations (A and P) to the auxinic herbicides mecoprop, clopyralid, 2,4‐D and aminopyralid in a preliminary experiment. The dicamba-resistant fathen was cross-resistant to the pyridine carboxylic acid herbicides clopyralid and aminopyralid, but not the phenoxy acid herbicides, 2,4-D and mecoprop. The level of cross-resistance to aminopyralid and picloram (another pyridine carboxylic acid herbicide) was investigated in two separate dose-response experiments. The results of the first dose-response experiment showed populations L and M were 12 and 19 times more resistant to aminopyralid, respectively, than susceptible populations (A and P). The dicamba-resistant fathen populations were also shown to be resistant to picloram although the levels of resistance ranged from three- to 17-fold. These levels of cross-resistance for both herbicides were confirmed in the second dose-response experiment. The results of this study help with planning control strategies for this resistance problem.

KEYWORDS:

• Auxinic herbicides
• Chenopodium album
• cross-resistance
• dose-response
• forage crops
• herbicide-resistance

Introduction

Maize and forage brassica crops are extensively grown by livestock farmers in New Zealand to provide supplementary feed at times of the year when pasture growth is poor (Minnee et al. 2009). However, both maize and brassica crops are susceptible to weed competition during early establishment (Ghanizadeh et al. 2010, 2014; Dumbleton et al. 2012). Chenopodium album L. (fathen) is one of the most troublesome weeds associated with crops in temperate countries such as New Zealand, especially in maize (James et al. 2005) and brassicas (Dumbleton et al. 2012). Fathen was the first species known to have developed resistance to herbicides (mainly atrazine) in New Zealand, within maize crops (Rahman et al. 1983). The benzoic acid group herbicide dicamba was then suggested as a useful post-emergence herbicide to control atrazine-resistant fathen in affected maize crops (Rahman et al. 1983). However, dicamba-resistant fathen has now also been reported from these maize fields (James et al. 2005). Greenhouse trials have shown these populations of fathen have seven- to 19-fold levels of resistance to dicamba (Ghanizadeh et al. 2015).

Cross-resistance can be defined as the expression of a single mechanism of resistance that also confers resistance in a biotype to chemicals from the same or different classes (Beckie & Tardif 2012). The pattern of cross-resistance has been studied in a number of auxinic herbicide-resistant weeds and varied results have been reported. For example, Van Eerd et al. (2005) reported a cross-resistance to phenoxy acids (MCPA), benzoic acid (dicamba), and pyridine carboxylic acid (picloram, fluroxypyr and triclopyr) classes of auxinic herbicides in a quinclorac-resistant biotype of Galium spurium. Also, it was found that a dicamba-resistant biotype of Sinapis arvensis was cross-resistant to 2,4-D, mecoprop and picloram (Heap & Morrison 2002). However, Harrington (1989) did not note any cross-resistance to either dicamba or picloram in 2,4-D-resistant populations of Carduus nutans. The pattern of cross-resistance has also been investigated for dicamba-resistant Kochia scoparia by Cranston et al. (2001), which was found to be cross-resistant to dichlorprop, 2,4-D, mecoprop, MCPA and picloram.

As inconsistent patterns of cross-resistance have been recorded among weeds resistant to auxinic herbicides (reviewed by Beckie & Tardif [2012]), it was difficult to predict whether the cross-resistance found in K. scoparia could be extrapolated to the dicamba-resistant fathen in New Zealand, despite these weeds being closely related as both species are members of the Amaranthaceae family. Information on cross-resistance to herbicides helps with an understanding of the mechanism of resistance, and is also needed when developing strategies to control resistant populations. Therefore, the objective of this study was to evaluate the response of two dicamba-resistant populations of fathen to other auxinic herbicides.

Materials and methods

Plant materials

Dicamba-resistant (L and M) and dicamba-susceptible (A and P) populations of fathen which had previously been studied by Ghanizadeh et al. (2015) were used in this study. Plants of each population were grown in a glasshouse in hand-watered polythene planter bags (700 mL) filled with potting mix (50% bark, 30% fibre, 20% Pacific Pumice (7 mm) and slow-release fertiliser (Woodace) by sowing five seeds per bag. More than 80% of the seeds had germinated after 2 weeks for all populations, and were thinned to one plant per pot when plants were 4–5 cm tall. The plants were then left to establish before spraying with different rates of herbicides when they were 6 weeks old.

Preliminary experiment

The response of the four fathen populations to dicamba, mecoprop, clopyralid, 2,4-D and aminopyralid was assessed in a preliminary experiment that was conducted on 25 March 2015. In the preliminary experiment, all herbicides were applied when plants were 11–12 cm tall. All four fathen populations were treated with dicamba (Kamba 500, dimethylamine salt) at 0, 200, 400 and 800 g ae ha⁻¹, mecoprop-P (Duplosan KV, potassium salt of the optically active isomer) at 0, 450, 900 and 1800 g ae ha⁻¹, clopyralid (Versatill, amine salt) at 0, 150, 300 and 600 g ae ha⁻¹, 2,4-D (Baton, dimethylamine salt) at 0, 400, 800 and 1600 g ae ha⁻¹, and aminopyralid (Tordon Max, triisopropanolammonium salt) at 0, 22.5, 45 and 90 g ae ha⁻¹. Each herbicide was applied using a dual-nozzle laboratory track sprayer calibrated to deliver 210 L ha⁻¹ of herbicide solution at 200 kPa.

Plants were kept in a glasshouse after application with the daily maximum and minimum temperatures in the 2 weeks following application averaging 27.7 °C and 17.5 °C, respectively, and the average day length was 11.4 h. To evaluate the effect of the herbicides, plants were harvested from all pots at ground level 7 weeks after application. Then harvested material was oven-dried at 70 °C for 48 h and weighed. The effect of each herbicide treatment was calculated as a percentage of the dry weight of untreated plants for that population.

Dose-response experiments

To determine whether the four populations of fathen differed in susceptibility to picloram, and also to estimate the size of differences between the populations in their resistance to both aminopyralid and picloram, a dose-response experiment was conducted on 12 August 2015. Picloram was selected for this study as previous work had suggested that dicamba-resistant biotypes are always cross-resistant to picloram (Beckie & Tardif 2012). Plants of each population were established using the method described above and kept in a glasshouse. When plants were 10–11 cm tall, they were sprayed with aminopyralid at rates of 0, 11.25, 22.5, 45, 90, 180 and 360 g ae ha⁻¹ for susceptible populations (A and P) and 0, 45, 90, 180, 360, 720 and 1440 g ae ha⁻¹ for dicamba-resistant populations (L and M). Picloram (Spike, amine salt) was applied at rates of 0, 25, 50, 100, 200, 400 and 800 g ae ha⁻¹ for susceptible populations (A and P) and 0, 100, 200, 400, 800, 1600 and 3200 g ae ha⁻¹ for resistant populations (L and M). Both herbicides were applied using a dual-nozzle laboratory track sprayer calibrated to deliver 227 L ha⁻¹ of spray solution at 200 kPa. Plants were kept in a glasshouse with an average daily maximum and minimum temperature in the 2 weeks after application of 24.8 °C and 17.2 °C, respectively. Daylight was supplemented using four 500 W hydrogen gas lamps to maintain a 14 h day length to stop plants from flowering too rapidly. All plants were harvested at ground level 7 weeks after herbicide application and harvested materials were evaluated using the method described above. There were five replicate pots for each rate.

This dose-response experiment was repeated on 4 November 2015 using the same method. Daily maximum/minimum temperatures in the 14 d following application of herbicides in the second dose-response experiment averaged 26.6/17.8 °C (day/night), respectively, and the average day length was 14.1 h.

Statistical analyses

All three experiments were conducted using a randomised complete block design with five replicates of each rate. The basic assumptions for regression models were checked to ensure they were satisfied (Onofri et al. 2010). Dose-response curves for the populations were then determined for each of the latter two experiments by fitting the data to a three parameter logistic model:(1) where Y was plant biomass as a percentage of untreated control plants, d was the upper limit, x was the herbicide rate, GR₅₀ was the rate of herbicide causing 50% reduction in biomass, and b was the slope around GR₅₀. Data were fitted to the three parameter logistic model using the statistical software R (v.2.15.2) with its dose-response curve (drc) package (Knezevic et al. 2007). The ratios of GR₅₀ values for resistant and susceptible populations were calculated.

Results

Preliminary experiment

Figure 1 illustrates the effect of the five different herbicides applied at three rates on plant growth compared with untreated plants for the preliminary experiment. Plants of populations L and M were confirmed as being resistant to dicamba, shown by significant differences in biomass of these populations compared with the susceptible populations (A and P) following treatment with dicamba. The dry weight of the susceptible populations, A and P, decreased significantly with increasing dicamba rates, while both dicamba-resistant populations had average dry weights of more than 90% of untreated plants at all three dicamba application rates.

Figure 1. The reduction in plant shoot dry weight compared with untreated plants of four populations of Chenopodium album (A [dicamba-susceptible] from Waikato maize fields, P [dicamba-susceptible] from Palmerston North, L and M [dicamba-resistant] from Waikato maize fields), at 49 DAT with five different herbicides each at three rates. Vertical bars represent standard error of the mean.

Dicamba-resistant populations were severely affected by increasing dose of 2,4-D and mecoprop-P, with little difference occurring between the four populations to these herbicides (Figure 1). Both dicamba-resistant and susceptible plants showed the typical symptoms of auxinic herbicide damage when treated with 2,4-D and mecoprop-P, such as epinasty of the leaves, tissue swelling, bursting stems and twisting. At 800 g ae ha⁻¹of 2,4-D, and 900 g ae ha⁻¹ of mecoprop-P, a significant reduction in plant dry weight compared with untreated plants was recorded for both dicamba-resistant and susceptible populations, with more than 50% reduction in biomass compared with untreated plants for all four populations (Figure 1).

Both dicamba-resistant populations appeared to be cross-resistant to clopyralid and aminopyralid, although clopyralid at 300 g ae ha⁻¹ was less effective on dicamba-susceptible populations than aminopyralid at 45 g ae ha⁻¹ (Figure 1). At 300 g ae ha⁻¹ of clopyralid, population A and P (dicamba-susceptible) suffered only 10% and 30% reduction in plant dry weight, respectively, compared with untreated plants, whereas with 45 g ae ha⁻¹ of aminopyralid, both populations A and P had 85% and 65% reduction in biomass, respectively, compared with untreated plants (Figure 1). Increasing the clopyralid rate to 600 g ae ha⁻¹ resulted in a significant reduction in plant biomass compared with untreated plants for populations A and P; however, clopyralid was still less effective on dicamba-susceptible populations than aminopyralid.

Dose-response experiments

Aminopyralid

The herbicide rates calculated to cause a 50% reduction in biomass (GR₅₀) for the fathen populations in the first and second aminopyralid dose-response experiments are shown in Table 1. The dose-response curves for both experiments are shown in Figure 2A–B. Aminopyralid at 45 g ae ha⁻¹ resulted in significant reductions in biomass for both susceptible populations, with both less than 50% of untreated plants 7 weeks after application in the first experiment (Figure 2A). Very similar results were obtained for the second aminopyralid dose-response experiment (Figure 2B).

Figure 2. Fitted aminopyralid dose-response curves (on a logarithmic dose scale) for reduction in shoot dry weight for four populations of Chenopodium album (A [dicamba-susceptible] from Waikato maize fields, P [dicamba-susceptible] from Palmerston North, L and M [dicamba-resistant] from Waikato maize fields) in the A, first, and B, second dose-response experiments. Vertical bars represent ± standard error of the mean.

Table 1. Parameters (see footnote) estimated from the non-linear regression analysis of aminopyralid dose-response experiments for Chenopodium album populations at 49 days after treatment as illustrated in Figure 2A–B.

Population	d ± SE	b ± SE	GR50 ± SE	R/S (A) GR50 ratio	R/S (P) GR50 ratio	R^2
First dose-response experiment
A (susceptible)	103.9 ± 1.3	1.0 ± 0.1	47 ± 15	–	–	0.93
P (susceptible)	103.6 ± 1.9	1.0 ± 0.2	55 ± 18	–	–	0.90
L (resistant)	100.1 ± 2.7	1.4 ± 0.2	699 ± 53	14.8	12.7	0.95
M (resistant)	102.7 ± 3.4	1.6 ± 0.2	896 ± 79	19.1	16.3	0.90
Second dose-response experiment
A (susceptible)	102.6 ± 1.6	1.2 ± 0.2	44 ± 10	–	–	0.94
P (susceptible)	103.1 ± 1.8	1.2 ± 0.1	54 ± 16	–	–	0.98
L (resistant)	101.5 ± 2.4	1.2 ± 0.1	317 ± 46	7.2	5.8	0.95
M (resistant)	101.4 ± 1.5	1.4 ± 0.2	543 ± 85	12.3	10.0	0.95

d, the upper limit; b, the slope around the GR₅₀; SE, standard error; GR₅₀, the rate of herbicide (g ae ha⁻¹) required to reduce shoot fresh weight by 50%; R/S GR₅₀, resistant/susceptible GR50 ratio; R², coefficient of determination.

The GR₅₀ values for populations L and M were significantly higher than those for susceptible populations (A and P) in both aminopyralid dose-response experiments (Table 1). By comparing the GR₅₀ of population L against those of the susceptible populations for the first experiment, it was estimated that population L was 14.8 and 12.7 times more resistant to aminopyralid than susceptible populations A and P, respectively (Table 1). The level of resistance to aminopyralid for population M was slightly higher than that of population L, and was estimated to be 19.1 and 16.3 times more resistant to aminopyralid than susceptible populations A and P, respectively (Table 1). In the second dose-response experiment, however, the level of resistance to aminopyralid in both resistant populations was slightly lower than that of the first dose-response experiment and it was estimated that populations L and M had resistance levels to aminopyralid that ranged from five- to 12-fold (Table 1).

Picloram

Dose-response curves for picloram showed that dicamba-resistant populations of fathen (L and M) were also significantly less susceptible to picloram than populations A and P (Figure 3A–B). A significantly greater reduction in shoot dry weight of populations A and P compared with populations L and M at each rate of picloram (Figure 3A–B) resulted in higher GR₅₀ values for populations L and M in both dose-response experiments (Table 2).

Figure 3. Fitted picloram dose-response curves (on a logarithmic dose scale) for reduction in shoot dry weight for four populations of Chenopodium album (A [dicamba-susceptible] from Waikato maize fields, P [dicamba-susceptible] from Palmerston North, L and M [dicamba-resistant] from Waikato maize fields) in the A, first, and B, second dose-response experiments. Vertical bars represent ± standard error of the mean.

Table 2. Parameters (see footnote) estimated from the non-linear regression analysis of picloram dose-response experiments for Chenopodium album populations at 49 days after treatment as illustrated in Figure 3A–B.

Population	d ± SE	b ± SE	GR50 ± SE	R/S (A) GR50 ratio	R/S (P) GR50 ratio	R^2
First dose-response experiment
A (susceptible)	103.5 ± 2.2	1.0 ± 0.2	94 ± 17	–	–	0.92
P (susceptible)	102.5 ± 2.3	1.2 ± 0.1	199 ± 24	–	–	0.95
L (resistant)	101.6 ± 1.6	0.8 ± 0.1	627 ± 182	6.7	3.1	0.95
M (resistant)	102.3 ± 2.2	1.1 ± 0.2	1609 ± 254	17.1	8.1	0.90
Second dose-response experiment
A (susceptible)	101.9 ± 1.1	0.9 ± 0.1	75 ± 18	–	–	0.98
P (susceptible)	104.2 ± 2.2	1.8 ± 0.2	153 ± 12	–	–	0.96
L (resistant)	100.7 ± 2.7	1.6 ± 0.1	871 ± 54	11.6	5.7	0.98
M (resistant)	101.4 ± 1.8	2.0 ± 0.2	1407 ± 63	18.7	9.2	0.97

d, the upper limit; b, the slope around the GR₅₀; SE, standard error; GR₅₀, the rate of herbicide (g ae ha⁻¹) required to reduce shoot fresh weight by 50%; R/S GR₅₀, resistant/susceptible GR₅₀ ratio; R², coefficient of determination.

The GR₅₀ values of population L were 6.7 and 3.1 times greater, respectively, than those of populations A and P for the first experiment. In the second experiment, population L had 11.6 and 5.7 times higher GR₅₀ values, respectively, than populations A and P (Table 2). Likewise population M was estimated to have a resistance level to picloram of 17.1- and 8.1-fold, respectively, in the first dose-response experiment (Table 2). In the second experiment, the GR₅₀ values of populations M were 18.7 and 9.2 times higher, respectively, than those of populations A and P (Table 2).

Discussion

The pattern of cross-resistance was investigated in this study and the results showed that the dicamba-resistant fathen is cross-resistant to pyridine carboxylic acids but not phenoxy acids. This is in contrast to the results reported by Cranston et al. (2001) who noted that dicamba-resistant K. scoparia was cross-resistant to both carboxylic acid and phenoxy acid classes of auxinic herbicides. However, as occurred in our work, Fuerst et al. (1996) noted that picloram-resistant Centaurea solstitialis was cross-resistant to benzoic acids (dicamba) and pyridine carboxylic acids (clopyralid) but not phenoxy acids (2,4-D).

Clopyralid appeared to be not effective on fathen plants that were susceptible to dicamba, whereas aminopyralid was more effective, causing significant decreases in the biomass of dicamba-susceptible plants. Similarly, Blackshaw (1989) noted that that clopyralid only suppressed fathen growth slightly compared with other weed species. The physiological reason for this greater activity of aminopyralid compared with clopyralid in fathen is unknown.

This is the first report of herbicide cross-resistance in dicamba-resistant fathen and is particularly important because this species is a major weed in many agricultural and horticultural crops (Aper et al. 2010). The results of this study showed a cross-resistance to pyridine carboxylic acid type herbicides such as picloram and aminopyralid in dicamba-resistant fathen and this suggests that these herbicides could not be used where dicamba-resistant fathen is present. In New Zealand, there have been anecdotal reports from farmers regarding the lack of control of fathen in brassica crops where aminopyralid and clopyralid/picloram mixes have been used for a number of seasons in succession. Although it is unknown if the same pattern of cross-resistance as found in our work exists for fathen from brassica crops for aminopyralid and clopyralid/picloram mixes, it could presumably evolve in the same way as has occurred in maize crops.

In maize where this weed species was first found to be resistant to dicamba, there are a number of alternative herbicides that can be used to manage this resistant species, such as nicosulfuron and mesotrione (Rahman et al. 2008). Hence, the cross-resistance shown in our study to other two auxinic herbicides (picloram and aminopyralid) might not be a major issue for weed management in maize. However, this could become important in crops such as forage brassicas where few other herbicide options are available.

Disclosure statement

No potential conflict of interest was reported by the authors.