Impact of chaff lining and chaff tramlining on survival of weed seeds

Take home messages

  • There is no evidence that weed seeds deposited in chaff lines and chaff tramlines will decay more rapidly than if deposited outside of concentrated chaff
  • A proportion of winter weed seeds buried in tramlines and chaff lines were still capable of germination in the following autumn and could compete with winter crops
  • Tramlines and chaff lines should be monitored and emergence controlled if necessary
  • The key reason to use chaff lining or chaff tramlining is to concentrate weed seeds into a narrow area for targeted, more efficient control of weed emergence

Background

Herbicide resistance is a major concern for northern region crop production due to the increasing frequency of resistance in key weeds. Non-herbicide weed management alternatives are needed to delay the spread and onset of further herbicide resistance (Walsh et al., 2013). One such alternative is harvest weed seed control (HWSC). Harvest weed seed control (HWSC) refers to a suite of management practices - all of which target the seed of weeds present at harvest and borne at harvest height (typically around 15cm above ground height, depending on header set-up).

Current HWSC systems include narrow windrow burning, chaff tramlining/chaff lining, chaff carts, bale direct and seed destruction (Walsh et al. 2013). Chaff lining and chaff tramlining have potential for wide-spread adoption in northern Australia owing to their low cost and ease-of-implementation relative to some other HWSC practices. Chaff tramlining is the practice of concentrating the weed seed bearing chaff material on dedicated tramlines in controlled traffic farming (CTF) systems, typically using a chaff deck to deposit chaff into 2 lines (one per wheel track). Chaff lining is a similar concept, where the chaff material is concentrated using a chute into a single narrow row between stubble rows and directly behind the harvester (i.e. not specifically onto the tramlines).

Aims

Anecdotal reports suggest weed seeds decay rapidly when deposited in chaff tramlines/chaff lines, but so far this remains scientifically unsubstantiated. The experiments described in this study were designed to:

  • Investigate weed seed survival over time under barley chaff in tramlines created using Draper and Shelbourne header fronts, compared to weed seeds outside of tramlines.
  • Evaluate the influence of different types of chaff on weed seed survival in chaff lines, compared to weed seeds outside of chaff lines.

Method

Field experiments were set up at Wagga Wagga in NSW and Irongate in southern Queensland. For the experiment in Wagga Wagga, seeds of annual ryegrass (ARG), brome grass, and wild oats were placed in bags under chaff lines of wheat, barley, and canola. For the control treatment, bags were placed next to the chaff line. For each crop type there were 4 replicates for each weed species. Bags of weed seeds were put out in December 2017 and removed in April 2018.  Seeds recovered from the bags were used in a germination test lasting 21 days.  The number of germinated and filled (i.e. embryo present) seeds was used to calculate weed seed survival as a percentage of the seed initially placed into each bag (100 seeds per species per bag).

For the study at Irongate, near Mount Tyson in Queensland’s eastern Darling Downs, two field experiments were set up following the 2017 winter crop harvest. Chaff was directed onto tramlines during barley harvest using Draper (conventional) and Shelbourne (stripper) fronts. Bags containing seeds of annual ryegrass, wild oats, turnip weed, and common sowthistle were placed under the chaff tramlines. For the control treatment, bags containing weed seeds were placed outside of chaff tramlines, in a nearby part of the paddock. For each trial (Shelbourne and Draper) and each sampling time there were 3 or 6 replications (for the control and tramline treatments, respectively) for each of the four weed species tested. The bags containing weed seeds were originally placed in the field in November 2017 and were removed on two occasions; in April 2018 and September 2018. Seeds recovered from the bags were used in a germination test lasting 28 days, where the number of germinated and filled (i.e. embryo present) seeds was used to calculate weed seed survival as a percentage of the seed initially placed into each bag. For common sowthistle, germination alone was used to calculate seed survival. The total seed initially placed in each bag was 133/bag (ARG), 106/bag (wild oats), 200/bag (turnip weed), and 250/bag (sowthistle). The number of seeds used for each species differed to account for variation in initial seed viability among the species, so that there would be 100 viable seeds per packet.

Results and discussion

Results from the barley tramlines at Wagga Wagga showed significant differences in seed survival between each of the 3 weed species after 5 months in the field (annual ryegrass 95%, Brome 84%, wild oats 23%, l.s.d.=7.1%), but there was no significant difference between the control and the chaff line treatments. In canola (Figure 1), there was a significant interaction between weed species and treatment (i.e. chaff line and control). Brome had significantly higher seed survival under the chaff lines compared to the control treatment (seed bags not under the chaff line). The reverse occurred in annual ryegrass, where seed survival was significantly higher in the control treatment. Seed survival of wild oats did not differ significantly between the chaff line and control.

This is a column graph showing the rate of survival of weed seeds placed under a canola chaff line compared to a non-chaff line treatment. Brome had significantly higher seed survival under the chaff lines compared to the control treatment (seed bags not under the chaff line).

Figure 1. Survival of weed seeds (% of initial viable seed) placed under a canola chaff line compared with the control treatment (non-chaff line) at Wagga Wagga, NSW. The l.s.d is indicated using the error bars. Predictions with the same letter are not significantly different

In wheat (Figure 2), there was again a significant interaction between species and location (i.e. near or under the chaff line). Again, there was significantly higher seed survival for brome in the chaff line treatment compared with the control treatment, but there was no significant difference between chaff line and control for the other 2 species.

This is a column graph showing the rate of survival of weed seeds placed under a wheat chaff line compared to a non-chaff line treatment. Again, there was significantly higher seed survival for brome in the chaff line treatment compared with the control treatment, but there was no significant difference between chaff line and control for the other 2 species.

Figure 2. Survival of weed seeds (% of initial viable seed) placed under a wheat chaff line compared with the control treatment (non-chaff line) at Wagga Wagga, NSW. The l.s.d is indicated using the error bars. Predictions with the same letter are not significantly different.

The data from the Draper tramlining study at Irongate (Figure 3) showed a significantly higher average survival for seeds placed under the tramlines (18.4%) compared to the control (8.9%). There was a significant 2-way interaction between species and time, because the change in seed survivorship over time differed among species. In an analysis of species x time (combined for chaff lining and control treatments), seed survival of turnip weed did not change significantly between sampling times (7% April 2018 and 6% September 2018), but there was a significant decline in seed survival for annual ryegrass (29% April 2018 and 6% September 2018) and common sowthistle (7% April 2018 and 0.8% September 2018), and a highly significant decline in survival of wild oats (87% April 2018 and 0.8% September 2018).

The data from the Shelbourne tramlining study at Irongate (Figure 4) had a significant 3-way interaction between species and time. When analysed separately for each species, seed survival declined significantly over time in annual ryegrass (37% April 2018 and 11% September 2018) and common sowthistle (19% April 2018 and 3% September 2018), and in wild oats the decline approached significance (Fprob=0.063), but there was little remaining wild oat seed at both time periods (6% April 2018 and 0.3% September 2018). Seed bags containing wild oats appeared to have been predated by mice. For turnip weed, there was a 2-way interaction of tramline treatment and time. Seed survival declined dramatically for turnip weed in the control treatment (39% April 2018 and 7% September 2018), but there was no significant difference in the tramlining treatment (17% April 2018 and 22% September 2018). The slight but non-significant increase in seed survival in the September sampling is likely the result of reduced germination in the first sampling period, potentially due to seed dormancy. The suggestion is that decay of turnip weed is more rapid in a non-tramline environment and could have contributed to the overall lower seed survival across weed species in the control environment compared with Draper tramlines.

This is a column graph showing survival of weed seeds over time in tramline and non-tramline treatments in a barley trial at Irongate, Qld

Figure 3. Survival of weed seeds over time in tramline and control (non-tramline) treatments in the Draper barley trial at Irongate in QLD. The error bars represent the back-transformed standard errors of the means. There was a significant main effect of treatment and interaction between species and time (these effects were fitted to obtain the predictions).

This is a column graph showing the survival of weed seeds over time in tramline and non-tramline treatments in the barley trial at Irongate, Qld

Figure 4. Survival of weed seeds over time in tramline and control (non-tramline) treatments in the Shelbourne barley trial at Irongate in QLD. The error bars represent the back-transformed standard errors of the means. There was a significant interaction between treatment, species and time.

The above results indicate that weed seed survival under chaff lines and chaff lines is influenced by both chaff type and weed species. The differences between the Draper and the Shelbourne study in barley indicate that chaff volume could also play a role, with stripper fronts like Shelbourne known to have less chaff output (Broster et al., 2018). However, the lack of consistent trends in the data point towards the involvement of other factors beyond those measured or controlled for in these trials. It is clear from the scientific literature that factors influencing weed seed persistence are many and varied. Some weeds have transient seed banks, with seeds that germinate within a year of initial dispersal, while others can form persistent seed banks with seeds that remain in the soil for longer than a year. Losses of seed from the seed bank result from genetically controlled physiological responses to environmental cues including light, temperature, water, oxygen tension and chemical stimulants; as well as interaction with animals and pathogens leading to death (Simpson et al., 2012).

Conclusions

It is evident that weed seeds of key northern species can remain viable in chaff lines and chaff tramlines until the following autumn, which is the key period of recruitment for winter species. Weeds that emerge at this time have the potential to compete with winter crops.

It is important to monitor chaff lines and chaff tramlines, so that weed seedlings can be controlled at an early growth stage, while they are relatively easier to control and before they substantially impact the crop.

The key benefit of using chaff lining and chaff tramlining is not to promote decay of weed seeds, but to concentrate weed seeds captured during harvest into one or two lines per header pass, rather than spreading the weeds throughout the paddock. The advantages are that weed seedlings are easily visible in chaff lines and chaff tramlines, and emerging seedlings can be monitored and controlled in a targeted manner (e.g. using high labelled rates and a shielded sprayer) with potential gains in weed management efficacy.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the author would like to thank them for their continued support.

Our sincere appreciation to the farmers who have hosted trials on their properties. Your commitment to furthering this type of research is critical in our war against weeds.

References

Broster J., Rayner A, Ruttledge A (2018). Impact of stripper fronts and chaff lining on harvest weed seed control.

Simpson RL, Leck MA, Parker VT (2012) Seed banks: General concepts and methodological issues, Ecology of soil seed banks. Elsevier, pp. 3-7.

Walsh MJ, Newman P, Powles SB (2013) Targeting weed seeds in-crop: A new weed control paradigm for global agriculture. Weed Technol. 27:431-436

Contact details

Annie Ruttledge
Research Scientist, Weeds Management
Queensland Department of Agriculture and Fisheries
Ph: 07 4529 1247
Email: annemieke.ruttledge@daf.qld.gov.au

GRDC code: US00084 (innovative crop weed control for northern region cropping systems)

GRDC Project code: US00084