Managing herbicide-resistant weeds imposes a major constraint on broadacre crop production in Australia. Common weeds such as annual ryegrass, brome grass, wild oats, and wild turnip and radish exhibit varying levels of resistance to a range of herbicides with differing modes of action.
Alternatives to chemical use are being investigated in order to reduce reliance on herbicides for effective control. One such practice involves collecting seed in chaff and straw at harvest and ensuring that the seed is either destroyed during the harvesting process (using a new method of grinding, thereby destroying the seed) or collection of the harvested material, including the weed seeds, behind the header and carting it out of the paddock (either as bales or in bins).
Seed collection during harvest will work only if most of the plants remain in an upright position so the seed can be collected by the harvester; if the seed is shed before harvest this technique will be ineffective in reducing the weed population.
A survey of weed seed shedding and retention was conducted during the 2009 harvest period in 17 paddocks in south-eastern Australia (in South Australia, Victoria and southern New South Wales). The paddocks were cropped to either wheat or barley and were chosen in view of the presence of one of the weeds just listed.
In the case of annual ryegrass, seed shedding appeared to increase during the harvesting period, especially in response to hot weather during the middle of the harvest period.
Wild oats will shed seed during the harvesting period, and it is likely that hot weather during harvest speeds up the process.
Brome grass seed shedding was highly variable between the five sites where it was monitored, and no common trends could be discerned.
The amount of shedding of wild turnip seed was very variable between the two sites at which this species was monitored. The consensus seemed to be that turnip and radish pods at harvest are quite brittle and open very easily on contact—which would suggest that windy conditions or a header moving through the crop could cause sufficient pods to open before the header was able to capture all the seed and some would fall to the ground in front of the header.
Within-paddock variability in weed numbers influenced the results, and taking three replicates to assess seed shedding at the time of monitoring appeared to be insufficient. We recommend that if a similar study is conducted in future the use of permanent monitoring sites (quadrats) be established in paddocks: this would help overcome the spatial variability in weed seed numbers encountered in the present study.
It appears that, for ryegrass, wild oats and turnip, seed collection at harvest can be a technique for reducing the seed bank and hence the population of weeds in the following crop, although this does entail early and timely harvesting. The results for brome grass were too variable to allow any conclusions to be drawn about the likelihood of success in collecting a high percentage of seed at harvest.
To view the full ‘seed retention in weeds’ report from the Rural Industries Research and Development Corporation click here.
