Tillage impacts of weed seed burial and subsequent management

Take home message

• Weed seed ecology (emergence and persistence) need to be taken into consideration when considering the potential role of tillage as a weed management tactic.
• In a zero till system, effective control of seedlings will rapidly deplete the weed seed bank of most of our key weeds. This is especially the case for feathertop Rhodes grass, windmill grass, fleabane and sowthistle.
• Herbicide resistance threatens our ability to effectively control emerged weeds.
• Applying cultivation may reduce weed emergence compared with a zero tillage system, especially in a wetter season. However, in a dry season, tillage may increase emergence.
• Seed burial is likely to increase the persistence of weed seeds.
• A second pass of cultivation can result in subsequent emergences of weeds as a result of seeds being brought closer to the soil surface.
• New approaches are being investigated to apply tillage in a targeted manner to minimise soil disturbance and weed seed burial.

Background

The widespread adoption of no-till farming has resulted in reliance on herbicides for the control of weeds. As a result of this reliance, the weed spectrum in the northern grain region has changed and become difficult to control. Herbicide resistance has become a common problem and weeds have shifted toward surface germinating species such as common sowthistle, fleabane and feathertop Rhodes grass. Continued reliance on herbicides will result in further cases of herbicide resistance and proliferation of these weeds. 

The change in weed spectrum has forced industry to investigate alternative approaches for weed management, including judicious or targeted use of tillage. The challenge with the reintroduction of tillage is to retain the benefits gained through zero tillage (improved soil structure, reduced erosion and improved soil water conservation) while addressing the weed management issues mentioned above.

Tillage can be applied to target mature, uncontrolled plants or to manipulate the seedbank and thereby improve weed management. This paper primarily addresses the use of tillage to manipulate the weed seed bank.

The importance of weed seed bank ecology

Each weed species has a different seed bank ecology, including depth from which it can emerge and duration seed persist in the soil, as affected by burial depth. These ecological factors are important to understand when considering the potential role and consequences of using tillage.

Emergence

In a zero tillage system, where soil disturbance is largely removed, most weed seeds will remain in the soil surface layer (0-2cm). Germination of some of our most common weeds is favoured from these layers and helps to explain their prevalence in zero tillage systems. For example, feathertop Rhodes grass seeds emerge mostly from the top 2cm of soil with a greatly reduced emergence from 5cm (Table 1). Over 12 months, 43% of seed buried near the surface germinated, compared with 5% at 5 cm and 0% at 10cm depths.

Similarly, sowthistle will mostly emerge from the top 1cm of soil, with limited emergence from a depth of 2cm (Figure 1). Fleabane will only emerge from the top 1cm of soil, with no emergence at or below 2cm.

Other species, such as liverseed grass and barnyard grass, are able to emerge from deeper in the soil profile (Table 1). While awnless barnyard grass still prefers to emerge from a depth of 2cm, it is also capable of emerging from a depth of 5cm. Liverseed grass prefers to be buried to 5cm for optimal emergence and can even emerge from a depth of 10cm (Table 1). As a result, liverseed grass is less common in zero tillage systems.

Table 1. Cumulative emergence (% of viable seed sown) of summer grasses at 24 months following burial at different depths (cm)

Depth of burial (cm)	Feathertop Rhodes grass	Barnyard grass	Liverseed grass
0-2	43*	27	36
5	5*	5	74
10	0	0.8	16

*emergences ceased after 12 months in all studies conducted on this grass

Figure 1. Cumulative emergence of common sowthistle as % of viable seed buried on the soil surface and at different depths (cm)

Persistence

Weed seeds left on or near the soil surface generally have a short life as fluctuating temperature and moisture reduces viability quickly. Generally, seed burial, for example via tillage, will promote persistence of seeds, and generally speaking, the deeper the seed is buried, the longer it will persist.

A clear example of this is barnyard grass and liverseed grass (Figure 2). For both species, seeds only remain viable for a short time in the soil surface layers, but persistence increases with depth of seed burial. Only 1 – 2% of seed remains viable after two years buried at a soil depth of 0 to 2cm, in contrast with approximately 20% remaining after two years buried at the depth of 10cm.

Depth and duration of seed burial also affect the persistence of feathertop Rhodes grass and windmill grass (Figure 2). A recent pot experiment on the eastern Darling Downs showed that after three months of burial, a large number (>70%) of FTR seed persisted at 10cm burial depth (Figure 2). There were a significantly lower number of viable seeds persisting at the 0cm depth. Burial depth had less impact on the persistence of WMG with both depth treatments having <20% viable seed remaining.

There were no further emergences of feathertop Rhodes grass from soil exhumed at 12 and 18 months showing this weed to be short lived in our southern Queensland environment. No emergences were recorded for windmill grass in soil exhumed at 18 months.

Figure 2. Persistence as assessed through emergence of seed from exhumed soil (% of viable seeds) of feathertop Rhodes grass (FTR), windmill grass (WMG), awnless barnyard grass and liverseed grass  in response to different burial depths (cm) and durations of burial (months).

Even though small-seeded, both fleabane and sowthistle seed can persist at depth for more than three years. A pot study on the Darling Downs showed that after 3 years of burial 1, 10 and 8% of viable fleabane seed remained at depths of 0-2, 5 and 10cm respectively (Figure 3). For sowthistle, after 30 months of burial, over 10% of seed remained viable when buried at 5 or 10cm compared with only 1% at 1cm burial depth (Figure 4).

Figure 3. Persistence of flaxleaf fleabane buried at different depths and for different times (years).

 

Figure 4. Persistence of common sowthistle seed buried at different depths for either 8 or 30 months.

Tillage effects

The Queensland Department of Agriculture and Fisheries (DAF) have conducted a series of four field experiments since 2011 to investigate the effect of tillage on seed burial and emergence of key weed species.

The first three experiments were established at the DAF Hermitage Research Facility, and the fourth was established at the DAF Wellcamp Research Station. At field experiments one to three, five tillage treatments were imposed with different levels of soil disturbance and inversion;

• Zero tillage (ZT)
• Harrows
• Gyral
• Offset discs
• One-way discs.

At the fourth field experiment, we explored the impact of a second tillage pass with a gyral after one-way disc and offset disc treatment, on the subsequent emergence of weeds.

Seeds of awnless barnyard grass, feathertop Rhodes grass, windmill grass, liverseed grass, common sowthistle and flaxleaf fleabane were sown on the soil surface prior to tillage application. Also, small coloured beads, to represent weed seeds, were included so we could track soil inversion via soil coring and bead recovery.

Weed emergence was counted in each treatment for up to 18 months. We were unable to get emergence of every species in each experiment therefore some of the data presented here excludes certain species.

Glass bead (seed) burial

The burial of glass beads was quite consistent across experiments. As such we are only presenting the data from the fourth experiment (Figure 5) as it also explores the impact of the second pass with the gyral.

Figure 5. Burial of glass beads (cm) under different types of tillage as assessed through bead recovery from soil cores. Lettering is based upon LSD’s of the transformed means. Means with the same letter within each burial depth are not significantly different at the 5% significance level.

The application of different tillage treatments affected the burial of glass beads (Figure 6). Generally, the zero tillage treatment had a larger proportion of glass beads in the top 2cm of soil and as tillage intensity increased, this proportion decreased.

Analysis showed there was a significant difference between tillage type for seed burial (P<0.001) at 0-2, 2-5 and 5-10cm soil depths but not for the 10-20cm depth.

For the 0-2cm depth, the offset disc treatment with and without the second gyral pass and the one-way disc treatment all significantly reduced the number of glass beads at this depth compared to zero tillage. Of note, the one-way disc double pass treatment had significantly more beads at this depth than the single one-way disc treatment showing that the second pass of the gyral returned more beads back into this layer.

For the 2-5cm depth, the zero tillage treatment had significantly fewer glass beads than all other treatments except for the offset/gyral double pass treatment. There was no significant difference between the single and double pass treatments for the offset and one-way disc treatments, showing that the second pass did not significantly alter the number of glass beads at this depth.

For the 5-10cm depth the zero tillage treatment had significantly fewer beads than all other treatments. For this depth there was a significant difference between the offset disc and the offset disc/gyral double pass treatments with a significantly greater bead count for the double pass treatment. This result shows that the double pass moved more beads into this layer.

Weed emergence

Given the difference in glass bead burial under the different tillage treatments and the impact of seed burial on emergence, it is not unexpected that we found tillage to have a significant impact on weed emergence. However, the effect of tillage on weed emergence was different across field experiments.

In this paper, we firstly report on the first field experiment to show the potential, favourable impact of tillage on reducing weed emergence (Figure 6). We will then provide an example, using sowthistle, to demonstrate how season can impact on weed emergence under different tillage treatments.

In the first experiment, cumulative weed emergence density in zero tillage treatments was 233 (fleabane), 149 (feathertop Rhodes grass), 433 (windmill grass), 267 (barnyard grass), 380 (sowthistle) and 72 (liverseed grass) plants/m² across a 3m² assessment area (to account for horizontal seed movement).

Most forms of tillage greatly reduced the emergence of all weed species (Figure 6) with the greatest reduction evident in the small-seeded species fleabane. Generally, as the intensity of tillage increased, the emergence of weeds decreased. The greatest reduction in emergence was generally under a one-way disc, which caused large amounts of soil to be inverted and seed burial.

Figure 6. Emergence of key northern region weed species, as a % of emergence in zero tillage treatment, under different types of tillage

The effect of tillage on weed emergence was not consistent across experiments. For example, for sowthistle, off set discs and one-way discs reduced emergence compared with zero tillage in all four field experiments (Figure 7). However, in field experiment 2 both the harrow and gyral treatments increased seedling emergence and in experiment 3 the gyral treatment increased emergence.

This difference can be explained by considering the season (temperature and rainfall) and depth of seed burial. At the start of experiments 2 and 3, there was hot and dry weather, resulting in minimal emergence from zero tillage treatments and a rapid depletion of seed on the soil surface. In these experiments, seed buried by the harrow and gyral treatments were preserved at a depth from which they could later emerge once a favourable environment was present (sufficient moisture and suitable temperature). The emergence from the offset disc and one-way disc treatments was always less than in zero tillage treatments as the seed was buried too deep for emergence.

In experiments 1 and 4, there was a wet start to the experiment, resulting in a large flush of sowthistle emergence from zero tilled treatments. In these experiments, a portion of seed was buried below the depth of emergence in all tilled treatments and therefore a reduction in emergence was measured.

Figure 7. Impact of different forms of tillage on the emergence of common sowthistle as % of emergence in zero tilled plots

Impact of second pass

The second pass with the gyral altered the distribution of glass beads in the soil profile (Figure 5). This in turn resulted in further emergences of all species, but not in all treatments (Figure 8). The emergence of barnyard grass was greatest of all species and while there was an increase in emergence following the gyral double pass after the one-way disc, the same effect was not evident for the gyral double pass after the offset disc. For sowthistle, feathertop Rhodes grass and windmill grass there was a small increase in emergence following the application of the gyral double pass.

Figure 8. Emergence (/m²) of key weed northern region weed species after the second pass gyral treatment had been applied following different forms of tillage

New tillage approaches

In a desire to find the balance between retaining the benefits of zero tillage and using tillage to improve weed control, current research is investigating and evaluating the use of robotics and targeted tillage in combination with weed detector technology. Both of these innovative approaches are being investigated to target low density weed populations. Thereby, stopping weed seed set and spread of herbicide resistance, whilst causing minimal soil disturbance.

There are several research groups investigating the role of robotics in the management of weeds. Whilst all are still in the development stages, there has been some very positive progress made. The robots are able to detect weeds, with some systems able to distinguish between weed types (grass vs broad leaf). One research group are developing a system that can apply different weed management tactics, depending upon the weed type present. For example, if it detects a grass weed, it may apply tillage. If the robot detects a broad leaf weed, it may apply herbicide.

Targeted tillage research is developing a hydraulically driven, rapid response tine system for the conduct of spot tillage operations. This approach relies upon a tractor operator but will use technologies already on the market (weed detection, tines).

Management implications

Effective weed management is reliant on an integrated approach. As with any other weed management tactic, the positives and negatives of tillage need to be considered. There have been many positive gains through the adoption of zero till and reduced tillage systems. However, the negative result of herbicide resistance cannot be ignored. We have demonstrated that tillage can have a positive impact in improving weed control. However, tillage across the whole paddock on a regular basis is likely to undo the positive gains achieved through zero tillage. Targeted tillage aimed at disturbing less of the soil and thereby reducing weed seed burial is the optimal approach if tillage is reintroduced in our current farming systems.

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.

Contact details

Michael Widderick
Queensland Department of Agriculture and Fisheries
Leslie Research Facility
13 Holberton Street, Toowoomba
Ph: 07 46398856
Fx: 07 46398888
Email: michael.widderick@daf.qld.gov.au