Managing wild radish and brome grass seedbanks

Cameron Taylor and Amy Smith (BCG)  BCG logo with white background

Take home messages

  • A two-spray approach of Clearfield® rotation and premium wild radish herbicide products such as Velocity®, Jaguar®, Precept® + Lexone® and Flight® proved to be the best at controlling wild radish and brome grass at the Pira site, over three years.
  • Short-term pain led to long-term gain: after three years, the more expensive herbicide regime proved to be at least equally or more profitable than the other strategies.
  • Effective products can lower the seedbank to manageable levels in three years however once a resistant population is selected, the population can explode quickly reducing yield and profitability.
  • It is important to consider the whole system and not just control one weed at the expense of a blowout of another


Wild radish populations are currently in a state of flux, with more incidences of resistance to herbicide Groups B, F and I emerging across North West Victoria (Craig 2012). Much of this has been attributed to an over reliance on Group B herbicides to control grass weeds such as brome and ryegrass. The consistent use of MCPA ester and amine formulations (Group I) and Group B (SU and SAs) is resulting in the effectiveness of these products diminishing. Subsequently, the selection pressure being placed on these herbicides is increasing.

To manage herbicide resistant wild radish and to avoid its spread efficiently, it is necessary to understand its lifecycle. This will allow growers to identify and implement methods that decrease the radish seed bank in a problem paddock. The weed seed bank is defined as the mature seeds that exist in the soil. Adopting practices that help reduce the number of viable weed seeds present in the seed bank (predominantly by preventing new plants from setting seed) improves the productivity of the paddock. It also decreases the need for costly herbicides, delaying the onset of further herbicide resistance.

A radish plant can produce up to 5000 seeds which, due to the hard-seeded nature of the pods, can stay in the top soil for up to eight years without germinating. The plants are highly competitive and when present in large numbers can greatly affect crop yields. This is the reason such importance has been placed on controlling 100 per cent of wild radish every year.

A paddock at Pira was selected to host BCG’s 2014 wild radish research trials after a survey conducted in 2013 determined that it contained wild radish populations with Group B resistance and suspected tolerance to Groups F and I as seen in paddock 6 (Figure 1). In 2013, a survey of the paddock showed that approximately 40% of plants were 100% resistant to Eclipse® as well as 100% resistance in other Group B herbicides including Logran® and Intervix® (Craig 2012).

bcg 2017 f1

Figure 1. Wild radish survey results. The points on the graph represent the variation of the surveyed paddocks across North-West Victoria in 2013. The treatments with varied results could have been due to poor herbicide efficacy or resistance. Paddock 6 is the site at Pira that came back with 100% confirmed resistance to Group B.

After multiple years of collecting data on herbicide efficacy of wild radish, a treatment list was developed based on that collected knowledge of herbicide type, rates and timing. In 2013, BCG completed a herbicide by timing trial that showed a good example of why resistance could be confused with application timing (Figure 2) (Taylor 2013). Jaguar for example, is a good chemical when the weeds are small, but when the weeds start growing in size the herbicide efficacy drops quickly due to being a contact product and the possibility of being shaded from the crop (Taylor 2014).

bcg 2017 f2

Figure 2. GRDC emerging weeds wild radish timing trial in 2013. The trial had three spray timings at 2 leaf radish, 6-8 leaf radish and large radish plants. The EWRC scores are comparing the different treatments and spray application timing scored at 28 days after application.

The trial discussed in this article is designed to compare the impact of a range of crop sequences and herbicide regimes on wild radish numbers and growth habits, and on subsequent crop performance and profitability. Along with wild radish, the Pira site also had a low population of brome grass. Given the increasing difficulty in managing this weed it was decided to also investigate how the different systems impacted on brome grass numbers throughout the duration of the trial.

Some of the herbicides used in this trial are not registered for use in certain crops, and were tested for experimental purposes only. Always read the label and adhere to directions when using herbicides.


To measure the effectiveness of herbicides used in Clearfield® and non-Clearfield® cropping rotations to control wild radish and brome grass in the Mallee.

Trial details

Location:                                Pira

Soil type:                                Deep sandy loam without sub-soil constraints

Paddock history:                  2013 grazed fallow

Treatments:                           Refer to Table 2

Target plant density:           130 plants/m²

Seeding equipment:            Knife points, press wheels, 30cm row spacing

Replicates:                             Four

Pests, disease and fertiliser were managed according to best management practice.

bcg 2017 t1

Table 1. Trial details by year.


A replicated field trial was sown using a complete randomised block trial design. The treatments are outlined in Table 2 below. Assessments included wild radish and brome counts before and after knockdowns and treatment sprays, and soil cores for seedbank assessments. Seeds in soil cores were allowed to germinate in planting trays for several months in 2014, 2015 and 2016. Trays were first refrigerated for several days (to account for brome germination temperature requirements) then placed outdoors and watered regularly. Soil was disturbed several times to increase germination. In 2017, radish seed was germinated by piercing the seed coat and placing in damp paper towel until no more seeds germinated. Brome was germinated in planting trays as described previously.

bcg 2017 t2

Table 2. Treatment descriptions for the trial at Pira.

Results and Interpretation

Wild radish

Herbicide efficacy

At the end of each season the number of wild radish plants were counted in each plot. The number of plants that germinated in 2014 was very high, and tested the ability of the herbicides to combat the weed pressure. Treatments 2 and 4 – two spray approach with premium products – controlled more than 99 per cent of the plants in all seasons. Treatment 1 (Tigrex®) and 3 (Midas®) had surviving plants that set seed at the end of the 2014 and 2016 seasons (Figure 3). The only year in which these treatments did not set seed was 2015, where many weeds died before setting seed because of drought.

In 2016, the major difference between treatments 1 and 3 was the number of seeds being set by the number of surviving plants. In year one of the trial these two treatments set the same amount of seed (800 seeds/m2). By the end of the trial Midas was having limited effect on the surviving radish population having more plants/m2 and those plants producing significantly more seed.

bcg 2017 f3

Figure 3. Start of 2014 and end of season wild radish plants counts per metre squared for each year of the trial for each treatment (see Table 2). Letters denote significant difference between treatments for that year. End of season counts show number of plants that survived the treatments each year to set seed. Start 2014 p=0.788 (NS), cv%=45, LSD=58.75. End 2014 p=0.035, cv%=111, LSD=5.9 . 2015 p=0.133 (NS), cv%=155, LSD 0.04. 2016 p=0.01, cv%=122, LSD=8.1.


In this trial the initial seed bank had 214 plants/m2 across the trial site. From the summer rain in 2014, 31 per cent of the seed bank germinated, providing an excellent opportunity to decrease the seed bank before the growing season began. This doesn’t happen every year: it is important to take these opportunities as they arise.

A double knock with glyphosate, followed by paraquat, is a good strategy to control any large radish plants that may have survived the initial glyphosate spray. From there, subsequent weed counts indicated that another 30 per cent of the seed bank germinated in crop; a total of 61 per cent for the season.

A two-spray approach (treatment 2 and 4) was found to have the greatest impact on seed bank reduction over three years by preventing any plants surviving to set seed (Figure 3, Figure 4). The seedbank for these treatments dropping by a mean of 57 per cent (Figure 3: treatments 2 and 4). The products used in these treatments, involved using a range of herbicide groups over the three years to minimise plant survival and potential resistance. Products with multiple modes of action come at a premium price however the benefit is clear for a long dormancy weed like radish; by 2017 the final seedbank had dropped by approximately 88 per cent since the first year.

A single spray approach with either Tigrex or Midas (treatment 1 and 3) was effective in the first year of the trial, with the seedbank decreasing by 83 per cent (Figure 3) (though it must be noted that the reduction in the first year can be largely attributed to two good knockdowns prior to sowing). A proportion of the population, approximately 7 radish/m2 survived the first year and so the population began to increase in the following years (Figure 4).

This increase occurred steadily under ‘farmer practice’ (treatment 1) conditions and more rapidly under ‘CLF farmer practice’ (treatment 3). Clearfield products are effective at reducing radish numbers initially, however resistance can build quickly. In both one-spray approach systems, a proportion of radish survived to set seed the following year.

bcg 2017 f4

Figure 4. Mean wild radish/m2 seedbank taken from soil cores at the beginning of each year shown for each treatment (see Table 2). Statistical analysis for Treatment by Year p=0.011, cv%=16.5, LSD not provided as data was transformed for analysis.

The large difference observed between treatment 1 and 3 was due to the amount of seed being set from each plant. While some of the wild radish plants survived the Tigrex (Group FI) herbicide spray in treatment 1 the wild radish plants were severely stunted and produced significantly less weed seed numbers than treatment 3. It was clearly evident that the resistant population had been selected in treatment 3 with plants surviving the Midas herbicide (group BI) and going on to produce many seeds increasing the resistant population quickly (Figure 3). This is supported by the 2013 resistance tests which found plants which survived a group B spray had 100% group B resistance.

Brome grass

Managing the balance for broadleaf and grass control can be tricky in the rotation. With brome grass monitoring throughout the trial, keeping Clearfield herbicides in the rotation was key to keeping the seedbank low, as seen in treatment 3 and 4 (Figure 5).

Having very limited brome control by only relying on pre-emergent chemistry like in treatment 1, can lead to very large seed set in one year. A slight increase in the seedbank in year one of the trial has set up enough seed to cause an explosion in seed set in 2016 when an above average rainfall year occurred (Figure 5).

Treatment 2 had a large increase in brome in 2017 from seed set in 2016 but not to the level of treatment 1. The addition of metribuzin herbicide in this treatment, which has a suppressive effect on brome, could have contributed to this decrease in seed set (Figure 5).

Brome grass populations can be close to eradicated after three seasons as the seed bank does not have the longevity of wild radish. This may allow a move away from the reliance on Group B herbicides after three years. Alternatives to this could be using a pulse or canola crop to mix herbicide chemistries or to have a brown manure crop (Taylor 2014).

bcg 2017 f5

Figure 5. Mean brome/m2 taken from soil cores at the beginning of each year shown for each treatment (see Table 2). Treatment by Year p<0.001, cv%=63.5, LSD not provided as data was transformed for analysis

Commercial practice

It is vital to know the resistant status of wild radish before choosing herbicide options as selection pressure in wild radish can explode very quickly due to its ability to set a large amount of seed. A two-spray approach with premium products can drive wild radish down over three seasons. Brome grass numbers can be kept very low while controlling wild radish numbers with good spray strategies. It is important to consider the whole system and not just control one weed at the expense of a blowout of another.

In low rainfall environments, it may be thought that low cost (risk averse) production systems are the most sustainable for long-term profitability. This trial has shown over three years that taking a ‘no survivors’ attitude to weed control can decrease seedbank to manageable levels and open rotation options in the future.

When levels are low enough in paddock situations, spot spraying and hand weeding may even become an option to lower the cost of production. The key principals of weed control cannot be forgotten – spray when the weeds are small and follow up for any survivors. Every wild radish plant that survives could mean dealing with the issue for another six to eight years due to the longevity of the seedbank.

One thing that was not investigated in this trial was harvest weed seed management techniques. Narrow windrow burning and now newer techniques such as integrated harvest weed seed destructors on headers can provide another alternative to weed seed control program. One of the weaknesses of this system is that the weeds must be captured in the front of the header for effective control. Through observations at harvest time, wild radish and brome grass start to shed their seeds before the harvester arrives in the paddock. Although the harvester would collect most of the weed seeds, the ability to shed seeds would maintain a low seedbank, but not eradicate the problem completely.

Site specific management could be key to lowering your costs of production. Targeting your paddock weed populations with weed mapping and the applying the right chemicals and management to the relevant areas of the paddock may be a way forward, in keeping costs as low as possible.

On-farm profitability

Often with weeds it is a case of spending more to make more. At the time, premium products might seem like a large expense but quite often standard products will not cut it and resistance has the potential to begin and build up, causing a serious issue which could take years to overcome. Where there were high weeds numbers in the 2016 season – as seen in treatments 1, 2 and 3 – yield was affected (Table 3).

bcg 2017 t3

Table 3. Yield and quality of wheat in 2016 season. The yields were analysed using a 90% confidence interval due to the variable nature of weed populations.

In 2014 the wheat yield was not significant between treatments, averaging 1.8t/ha. The weed density at this stage had no effect on yield. In 2015, a year of severe drought, there was no significant difference between the barley yields, which averaged 0.7t/ha across the site. These factors have all contributed to income after three years of production (Table 4).

After three years, the more expensive herbicide regime has proven to be at least equally or more profitable than the other strategies. Treatment 4, which controlled brome grass and wild radish every year made $326.54/ha while the farmer practice produced $270.24/ha over the trial (Table 4). This was mainly due to the downgrading of the sample at harvest time in 2016 due the high amount of weed seeds in the sample (Table 3).

bcg 2017 t4

Table 4. Profit of the different herbicide strategies after three years. Each year’s costs have been derived from the Rural Solutions Farm Gross Margin Guide + herbicide cost. Prices for grain used for individual years in association with its respective BCG results compendiums.


Taylor C., 2013, ‘Fighting the war against weeds: wild radish control in wheat’, BCG 2013 Season Research Results, pp. 154 – 160.

Taylor C., 2014, ‘Wild radish control’ BCG 2014 Season Research Results, pp. 129 – 139.

Craig S., 2012, ‘Short-term pain for long-term gain: controlling wild radish in wheat” BCG 2012 Season Research Results, pp. 94 – 98.


This research was funded through the GRDC emerging weed issues project ‘Improving IWM practice in the Southern Region’ (UA00105) and BCG members funds through their membership




About BCG

Birchip Cropping Group Inc. (BCG) is a not-for-profit agricultural research and extension organisation led by farmers in the Victorian Wimmera and Mallee.
This entry was posted in No category. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s