The mechanisms of herbicide resistance

GRDC code: UA00124

Author

Christopher Preston, School of Agriculture, Food & Wine, University of Adelaide

Take home message

For most herbicide modes of action there is more than one resistance mechanism that can provide resistance and within each target site, there are a number of amino acid modifications that provide resistance. This means that resistance mechanisms can vary widely between populations; although, some patterns are common. While some broad predictions can be made, a herbicide test is the only sure way of knowing which alternative herbicide will be effective on a resistant population.

What are we selecting for and why?

Mechanisms of herbicide resistance – what are we selecting for and why?

Herbicide resistance has become a major concern for grain producers across Australia. Currently there are 41 weed species in Australia with resistance to one or more of 11 herbicide modes of action. Management of herbicide resistance usually relies on an understanding of the biology of the weed species and the herbicides that are still effective for control. At times the mechanisms of herbicide resistance that are common can provide information about potential cross-resistance to other herbicides and rules of thumb about herbicide management.

Every time a herbicide is used to kill weeds or stop their seed production it selects for resistance. How strongly herbicides select for resistance depends on the proportion of weeds that are controlled by each application and how frequently applications of the same mode of action are made. Herbicide applications that control a smaller proportion of weeds will select for both weak and strong resistance mechanisms, whereas applications that control a high proportion of weeds tend to select for stronger mechanisms only. However, any mechanism that is present in the weed population and can allow the weed to survive has the potential to be selected by herbicide use.

Generally mechanisms of herbicide resistance are divided into target site and non-target site mechanisms. However, within each category there are distinct resistance mechanisms.

Target site resistance mechanisms

Target site mechanisms involve a change to the protein that binds the herbicide resulting in a lack of inhibition of the biochemical pathway. 

The most obvious of these is where a mutation within the target protein reduces or eliminates binding of the herbicide. This is the classic target site mutation and is typically seen as providing virtual immunity to the herbicide. However, that is not always the case and it is possible to have weak target site mutations as well as strong target site mutations. Target site mutations arise due to single point mutations in the underlying DNA that change an amino acid in the protein. This changed amino acid may either remove a bond required for binding the herbicide or change the shape of the binding pocket.

Target site mutations are common in weeds with resistance to Group A, Group B and triazine Group C herbicides, but also occur with resistance to Group D and Group M herbicides. With target site resistance it is common to get cross-resistance to other herbicides of the same herbicide mode of action.

For most target sites there is more than one possible mutation that will provide resistance to herbicides. In many cases, different mutations give different levels of resistance and patterns of cross-resistance. For example, resistance selected by sulfonylurea herbicides may not result in resistance to the imidazolinone herbicides. Our data suggests that cross-resistance to imidazolinone herbicides occurs about 30% of the time with broadleaf weeds and 50% of the time with grass weeds. However, this is variable between species. 

There are 8 different amino acids within the ALS protein where mutations are known to result in resistance to herbicides. Of these 4 give strong resistance to sulfonylurea herbicides and 6 give strong resistance to imidazolinone herbicides. Therefore, only some of the mutations provide resistance to both groups of herbicides. The reasons for this are that the different groups of herbicides bind differently in the binding pocket, so different mutations affect only one or both types of herbicides (Figure 1).

Figure 1.  Different mutations within a target site can result in different patterns of resistance. A target has a binding site where two chemically different herbicides, H1 and H2, can bind (A). The herbicides bind to different parts of the binding site. A mutation within the target site (B) may stop binding of one herbicide, but not the other. A different mutation elsewhere within the target site (C) may stop both herbicides from binding.

How the different mutations within a target site affect the binding of herbicides. Text description follows image.

Figure 1 text description: Target A has no mutation to the target site so both herbicide 1 and herbicide 2 is able to bind to the target. Target B has a mutation to one side of the target site which prevents herbicide 1 from binding to the target site but does not affect herbicide 2 from binding. Target C has a mutation to the target site which prevents both herbicide 1 and herbicide 2 from binding to the target site.

A different situation has occurred with Group A herbicides. For ACCase there are 7 amino acids within the protein where mutations are known to provide resistance to herbicides. Most of these provide resistance to the fop herbicides, but only 3 provide any resistance to clethodim, with only 1 giving high level resistance on its own. Therefore, most target site mutations selected by fop herbicides can be controlled by clethodim. This has allowed growers in southern Australia to exploit fop herbicides first to control ryegrass and after those herbicides had failed, to use clethodim to control ryegrass. Once clethodim started to fail, higher rates were used as there was only one mutation that provided high resistance to clethodim.

The other type of target site resistance is where there are many more copies of the target site than would normally be present. In this type of resistance, the extra target sites act like a sponge soaking up the herbicide. This mechanism has so far only been seen with glyphosate resistant weeds. If this type of mechanism was to occur for another herbicide target site it would be expected to provide resistance to every herbicide in that mode of action.

Non-target site resistance mechanisms

Non-target site resistance mechanisms allow plants to survive application of the herbicide by not allowing sufficient herbicide to reach the target site. The weed may be initially affected by the herbicide application, but will survive and set seed.

The most common example of non-target site resistance is due to increased herbicide detoxification. With this resistance mechanism, there is more rapid breakdown of the herbicide inside the plant and less of the active herbicide reaches the target site to kill the plant. With this mechanism, the species has to start with some ability to metabolise the herbicide, but that becomes greatly enhanced in the resistant individuals. For this reason, enhanced metabolism is typically observed in herbicides that can be used selectively in the crop; such as Groups A, B, C, D and I.

The exact nature of the mutation that results in resistance due to enhanced herbicide degradation has not been identified. However, evidence points to the increased activity of several enzymes, rather than a single enzyme. One outcome of this type of mechanism is that it frequently leads to cross-resistance to herbicides of different modes of action. This seriously complicates resistance management with herbicides. Cross-resistance patterns tend to be highly variable and unpredictable, suggesting there are numerous types of enhanced herbicide detoxification occurring.

A variant on herbicide detoxification is reduced activation of the herbicide. Several herbicides are applied as pro-herbicides and rely on the plant metabolising them to the active compound. If the plant fails to do this, the herbicide will not work. This mechanism has been observed with triallate resistance in Canada, but has not been found in Australia.

A second non-target site mechanism involves changes to the translocation of herbicides within the plant. In this mechanism, the herbicide becomes trapped in the leaf tips and reduced amounts are present in the meristem and other parts of the plant. Where the herbicide has to be present in the growing tissue to kill the plant, reducing translocation will reduce the concentration of herbicide at the target site in these key tissues. This type of mechanism is common in weeds resistant to Group L and Group M herbicides, but has also been seen in weeds resistant to Group A herbicides.

There are several mechanisms whereby a plant can reduce the translocation of herbicides. The main mechanism seems to be through pumping the herbicide into the cell vacuole. As this involves specific transporters for the herbicide, resistance usually occurs to a single herbicide only. The exception to this is resistance to paraquat where cross resistance to diquat always occurs.

An alternative type of reduced translocation is where the herbicide is trapped in tissues that are then shed from the plant. This ‘rapid necrosis’ resistance resembles the plant response to pathogen attack, but on a massive scale where the whole leaves rapidly die and fall off taking the herbicide with them. This type of resistance has been observed for glyphosate resistance elsewhere, but not in Australia.

Two other types of non-target site resistance are theoretically possible, but have not been well documented. Reducing the absorption of the herbicide into the plant will reduce the concentration of herbicide at the target site. Such a mechanism is only likely to be effective for herbicides absorbed only or primarily through leaf tissue. The other type of mechanism is where the plant is able to avoid the detrimental effect of the herbicide action, usually through increased capacity to deal with oxygen radicals. This has been proposed as a mechanism of paraquat resistance for example, but is only a practical mechanism if the plant also has the ability to rapidly remove the herbicide from the target site.

What types of herbicide resistance are being selected and why?

We should expect that every mechanism of resistance that is present in a population will be selected by herbicide use. In practice what tends to happen is that the strongest resistance mechanism present becomes dominant. The strongest mechanism will have greater fitness under selection and so individuals carrying it will contribute more to the seed bank. In most broadleaf weeds target-site resistance to the sulfonylurea herbicides is found most commonly because this typically provides 100 fold resistance to the herbicide. The types of resistance mechanisms known for the different herbicide modes of action in Australia are listed in Table 1.


 Mode of action
Target site mutation
Target site duplication
Increased detoxification
Reduced translocation
Unknown
Table 1. Herbicide resistance mechanisms that have been identified for different modes of action in Australia
A
Yes

Yes
Yes

B
Yes

Yes

 
C
Yes

Yes


D
Yes

Yes


F




Yes
I


Yes?


J




Yes
L



Yes

M
Yes
Yes

Yes

Q




Yes
Z


Yes?


Due to variations in the strength of the resistance mechanisms between herbicides of the same mode of action, different herbicide selection can lead to different outcomes. A good example of this is that in the 1990s, diclofop-methyl (Hoegrass®) was the main selecting agent used against wild oats. Most of the resistant populations had target site resistance across the fop herbicides with some resistance to the dim herbicides. Later fenoxaprop-p-ethyl (Wildcat®) and clodinafop-propargyl (Topik®) became the most commonly used wild oat herbicides and a different type of resistance mechanism became common. This was a non-target resistance that gave cross-resistance to flamprop-methyl (Mataven®), but these populations often had susceptibility to other fop herbicides.

The patterns can also be different between weed species. Most of the resistance to Group B herbicides in annual ryegrass is target site resistance, generally because of the strong resistance provided by this mechanism to chlorsulfuron (Glean®) and triasulfuron (Logran®). In addition, at least 50% of the resistant populations have cross-resistance to imidazolinone herbicides. However, in brome grass, which has been selected mainly by iodosulfuron (Atlantis®) and pyroxsulam (Crusader®), most of the Group B resistance is low-level, non-target site resistance with little or no cross-resistance to imidazolinone herbicides.

Continued selection of the resistant populations with herbicides will result in stronger herbicide resistance, usually occurring through the stacking of resistance mechanisms. For example, when annual ryegrass became resistant to clethodim (Select®) at 250 mL ha-1 it was discovered that most populations could be controlled by 500 mL ha-1. Then populations became resistant to the higher rate by picking up extra target site mutations. When glyphosate resistant annual ryegrass first occurred in vineyards, some growers increased the rates of glyphosate in an attempt to control the ryegrass. What they selected for was annual ryegrass with two different mechanisms of resistance, a target site mutation and reduced herbicide translocation, which was much more resistant to glyphosate.

Weed species also become resistant to multiple modes of action by accumulating herbicide resistance mechanisms. This usually occurs through the sequential application of different modes of action, but can also occur with herbicide rotations and the use of herbicide mixtures. Multiple resistance occurs more readily in outcrossing weed species, but can also occur in self-pollinated weed species. The worst case of multiple resistance is a ryegrass population with resistance to Group A, B, C, L and M herbicides. This population has a combination of target site and non-target site resistance mechanisms.

The role of herbicide testing

While it is possible to make some general predictions of the most common resistance mechanisms that will occur from specific types of herbicide selection, the diversity of resistance mechanism present and the variations in herbicide history mean that it is difficult to predict the resistance mechanism in any one population. This is where herbicide testing becomes useful, not so much to determine whether the population is resistant, but to identify herbicides that will still work. Frequently a population in one field will respond differently to alternative herbicides to a population in an adjacent field. This is because a different mix of resistance mechanisms has been selected. Therefore, a test conducted on a population from one field may not be a good predictor of what will happen in the adjacent field.

Contact details

Christopher Preston
University of Adelaide
Ph: 08 8313 7237
Email: christopher.preston@adelaide.edu.au

GRDC Project Code: UA00124,