Mechelle Owen, Pippa Michael and Stephen Powles
Australian Herbicide Resistance Initiative
- WA farmers have significant weed seed contamination in their crop seed
- WA farmers are sowing herbicide-resistant weed seed into their cropping paddocks
- Foreign seed contamination may
- add to weed burden,
- introduce unwanted species,
- introduce herbicide resistant biotypes
- Crop hygiene (harvest, seeding, transporting) is important to minimize spread of weeds
Growers regularly employ in-crop management to reduce weed seed set thus minimising the number of seeds returning to the seed bank. While many growers are making a conscious effort to minimize weed seed set during the growing season they may be introducing seed back into the seed bank at seeding.
Across Australia is it common practice for farmers to grow and conserve their own crop seed rather than purchase it. However this crop seed can be contaminated with weed seeds, and if seed cleaning operations are only partially effective, then this weed seed may be sown into paddocks with crop seed in the following season. Minimising the introduction of weeds into the farming system through sowing of clean crop seed is an important component of best farming practice. As well as adding to the weed burden already present, foreign seed contamination may also introduce unwanted species and herbicide resistant biotypes into farming systems with potentially serious consequences. Herbicide resistant annual ryegrass (Lolium rigidum L.), wild oats (Avena fatua L.), and wild radish (Raphanus raphanistrum L.) populations are widespread throughout the WA grain belt (Owen et al. 2007, Walsh et al. 2007, Owen and Powles 2009) and several studies have also identified high levels of weed seed contamination in crop seed in the seed box at seeding time (Powles and Cawthray 1999, Moerkerk 2002, Niknam 2002).
Recent research conducted by WAHRI ,(Pippa Michaels, Muresk), aimed to determine the extent of weed seed contamination present in crop seed sown by growers in WA and to understand the effect of seed source (i.e. farmer retained, certified seed) and seed cleaning techniques on the degree of contamination. As part of this study, weed seeds infesting the crop seed were found to be herbicide resistant; therefore many growers are actually sowing resistant weed seed into their cropping paddocks.
During 2007/8 the study was conducted to quantify the extent of weed seed contamination in crop seed used for sowing. Farmers were asked a series of questions about seed cleaning methods, whether crop seed was cleaned prior to sowing and the source of their crop seed. A total of 183 grain samples (~10kg), of which half were wheat and half an alternative crop (i.e. barley, lupins, pea, canola and oats), were provided by 78 farmers from across the WA grain-belt (see sidebar).
Crop seed samples were cleaned by hand and the total weed seed contamination was determined. Weed seeds collected from the crop seed samples were then screened for herbicide resistance status using the most common herbicides used for the control of the particular weed species. Seedlings were sprayed at the 2-3 leaf stage and assessed for mortality 21 days after treatment.
Level of grain contamination
Of the 78 farms surveyed, the majority of crop seed samples (97%) were cleaned before crop sowing. Nearly all (95%) farmers surveyed retained their own grain for crop seed, with only 5% purchasing external seed. Despite most farmers growing their own crop seed, the majority (69%) of crop seed was cleaned by external contract seed cleaners (either on farm contractors or seed cleaners in town). Wheat accounted for over half of the samples, followed by barley and lupins (in more northern regions).
In total, 74% of grain samples collected in this survey had some level of weed seed contamination even though 97% of farmers stated that they had cleaned their grain. Contamination levels were highly variable between samples with an average of 8.7 weed seeds per kg of crop seed (Table 1). The main contaminant weed was annual ryegrass (4.6. seeds/kg), which occurred in over half of all samples, followed by wild radish (2.0 seeds/kg), brome grass (Bromus sp) (1.0 seeds/kg) and wild oats (<1 w/kg), which were all found in approximately one third of samples. A number of other weed seed contaminates were found in the grain sample but at low levels (Table 1). Figure 1 shows that, as expected, annual ryegrass was the most frequent contaminant, followed by wild radish, brome and wild oat. Volunteer cereal and legume grains accounted for 15-18% of foreign material. There were at least 13 different weed species contaminating the grain samples surveyed. Most growers expected some level of contamination; however, the level of contamination expected was much lower than what was found.
Table 1 Average weed seed number infesting 1 kg of crop seed for all samples
Figure 1 Seed contamination as a proportion of total samples (183)
The type of cleaning method employed had a significant effect of the level of weed seed contamination. There were differences in the level of contamination between cereal and pulse crops, with cereals having slightly higher levels of contamination for annual ryegrass, and were more frequently contaminated. Samples that were not cleaned had much higher levels of contamination than the cleaned crop seed samples. Samples that were only cleaned with air had much higher levels of contamination that those that were cleaned with a gravity table, and other methods such as a rotary screen and sieves, although sieves appeared more effective for cleaning wild radish.
Herbicide resistance status of weed seed contaminating crop seed
Weed seeds found contaminating the crop seed sample were tested for herbicide resistance with Group A and B herbicides. The majority of annual ryegrass seeds found in the crop seed were resistant to Group A – diclofop-methyl (84%) and Group B – sulfometuron (91%) (Table 2). Wild radish showed resistance to the group B herbicide chlorsulfuron. Wild oat showed resistance to the Group A herbicide diclofop-methyl, however, no resistance was found in brome grass to any Group A herbicide tested. Brome grass populations were not treated with Group B herbicides. The majority of wild radish resistant populations came from the more northern regions, while ryegrass resistance was spread across the whole state. This result is similar to that found in recent herbicide resistance surveys in WA (Owen et al. 2007; Walsh et al. 2007). All of the ryegrass populations resistant to the in-crop herbicides were controlled by the knockdown herbicide glyphosate (Table 2).
Table 2 The percentage resistance of weed populations contaminating crop seed displaying resistance to each herbicide for all samples
Implications of weed seed contamination
- Yield Loss in crop year
- Increased weed seed in seed bank
- Spread of herbicide resistant weeds
- Weed seed contaminants in grain exported to other countries e.g. Canada & Japan and if herbicide resistant could be a threat to export markets
On-farm management solutions
There are several ways to reduce the level of contamination in grain retained for crop seed. One of the most important factors to take into consideration when choosing paddocks to harvest seed from is to consider the weed burden in that paddock, higher levels of weeds will result in a higher level of contamination in the seed.
Cleaning methods used are also an important factor to consider, results showed there is a large difference in some of the cleaning methods. The trend was for gravity tables to be the most effective means, closely followed by rotary screens, sieves and a combination of methods, although on-farm hygiene, paddock selection and crop type may also influence contamination levels. Uncleaned crop seed had 22x more contamination that cleaned crop seed.
- Pick best paddock for crop seed (choose in advance and ensure good weed control in previous year – stop seed set
- On farm hygiene is very important (clean silos, truck, seeding & harvesting equipment) to stop carry over seed burden and spreading of weed seed and herbicide resistance
- Sow weed free crop seed (avoid adding weed seed to the seed bank unnecessarily)
- Delay seeding to allow adequate weed control before planting
- Consider crop type and variety – some crops are more competitive and allow greater choice of herbicide options for in-crop weed control
- Determine weed seed contamination levels prior to sowing – allow time for seed cleaning
- Clean crop seed – grading, sieves, gravity table
- Seed grading is never totally effective so consider other factors which may contribute to contamination
- Widespread herbicide resistance means weeds surviving the crop phase and setting seed are likely to be resistant
- Contaminated seed may introduce herbicide resistance into clean paddocks
Take home message
- 26% of samples were ‘weed free’– it is possible!
- Check crop seed before sowing
- Use 1kg sample to check level of contamination
- 10 seeds/kg ~ 1 plant/metre2
- 1 ryegrass plant produces about 500 seeds with is equivalent to 10kg/ha seeding rate of weed seed!
- Weeds increase the costs and reduce yield
- Farmers need to be vigilant to ensure the productivity of their farming systems (hygiene, management) and continued quality of end products
Owen, M.J., Walsh, M.J, Llewellyn, R.S. and Powles, S.B. (2007) Widespread occurrence of multiple herbicide resistance in Western Australian annual ryegrass (Lolium rigidum) populations. Australian Journal of Agricultural Research, 58 (7), 711-718.
Walsh, M.J., Owen, M.J. and Powles, S.B. (2007) Frequency and distribution of herbicide resistance in Raphanus raphanistrum populations randomly collected across the Western Australian wheatbelt. Weed Research, 47 (6), 542-550.
Moerkerk, M.R. (2001) Seed box survey of field crops in Victoria during 1996 and 1997. In ’ Proceedings of the 13th Australian Weeds Conference; threats now and forever’, Perth, Australia. Pp 55-58.
Sharon Roya Niknam, Michael Moerkerk and Roger Cousens (2001) Weed seed contamination in cereal and pulse crops. In ’ Proceedings of the 13th Australian Weeds Conference weeds; threats now and forever’, Perth, Australia. Pg 59-62