Options for managing barley grass in crop

Take home messages

• Reducing row spacing from 30cm to 18cm lowered grass weed dry matter substantially, and increased grain yield of wheat by more than 0.5t/ha in two consecutive years at Minnipa.
• Increasing wheat seeding rate also lowered late grassy weed dry matter and barley grass weed seed set, and increased grain yield at Minnipa.
• Barley has consistently outperformed wheat in its ability to compete with barley grass, particularly when sown at high seeding rates.
• Row orientation of east-west or north-south sowing showed variable yield results in 2015 and 2016.
• At Appila, in the Upper North, increasing seeding rate of barley in the presence of barley grass increased crop yields and reduced weed seed production. This applies particularly to competitive barley varieties such as Fathom, but also to less competitive varieties such as Hindmarsh.
• Light interception was very sensitive to sowing direction, and not having ‘weeds’ resulted in higher light within the canopy in the north-south direction compared to east-west.

Background

Controlling barley grass (Hordeum sp) in the Upper Eyre Peninsula and the Upper North low rainfall farming systems is becoming a major focus for growers, due to the intensity of cereal cropping, development of herbicide resistance and the changing ecology of the weeds, such as increasing frequency of delayed emergence of barley grass in populations.

There are effective but costly options for grass weed control using pre-emergent and post emergent herbicides. However, for long-term sustainability, a range of management techniques besides reliance on herbicides is required to address the issue. Potential non-chemical options for managing barley grass in a crop addressed by this research project include increasing crop competition by reducing row spacing or increasing sowing rate, and changing row direction. The Australian Herbicide Resistance Initiative (ARHI) based at University of Western Australia has shown an increase in grain yield with wheat and barley sown in an east–west (E-W) orientation over crops sown in a north-south (N-S) orientation due to a decrease in ryegrass competition. This effect is due to lower light interception by the weed due to the crop row orientation resulting in a decrease in weed seed (Borger, 2015).

An important requirement is to find practices which both maximise crop yield in the presence of background grass populations as well as suppress weed seed carryover. This research was funded as part of the GRDC ‘Overdependence on Agrochemicals’ project, which aims to find ways to reduce dependence on agrochemicals in our current farming systems.

Method

Minnipa Agricultural Centre - row spacing and seeding rate trial

Replicated trials were undertaken in both 2015 and 2016 at Minnipa Agricultural Centre (MAC) with Mace wheat sown at three seeding rates (targeting 60, 120 or 240 plants/m²) on two different row spacings of 18cm (7”) and 30cm (12”) with two different seeding boots, a narrow row Harrington point and an Atom-Jet spread row (ribbon) seeding boot with press wheels.

The Minnipa Agricultural Centre soil is an alkaline red sandy loam, with a pH (CaCl₂) of 7.8 and adequate Colwell P and soil mineral N. A base fertiliser rate of 60kg/ha of 18:20:0:0 was applied to all treatments in both seasons. For more detailed information on the trial management refer to Eyre Peninsula Farming Systems Summary articles (EPARF website).

Table 1. Rainfall and sowing dates.

Trial measurements taken during the seasons included soil moisture, PreDicta B root disease test, soil nutrition, weed establishment, weed seedbank germination, crop and weed establishment, crop and weed biomass (early and late), light interception in crop rows (using AccuPAR PAR/LAI ceptometer), grain yield and quality.

Grass weed seed set in 2016 was calculated using the total panicle length and number of panicles/m² of individual plots. Weed seeds per panicle were counted from selected treatments and a regression was used to calculate weed seed set per plot. Data were analysed using Analysis of Variance in GENSTAT version 16.

Results

Minnipa Agricultural Centre - row spacing and seeding rate trial

There were no statistically significant interactions for row spacing and seeding rate in either 2015 or 2016 seasons, so results are presented for the individual factors only.

Table 2. Grass weed density and canopy measurements taken in seeding rate and row spacing trial sown with Mace wheat at Minnipa, 2015.

Table 3. Grass weed density and canopy measurements taken in seeding rate and row spacing trial sown with Mace wheat at Minnipa, 2016.

Table 4. Wheat growth, yield and grain quality measurements taken in seeding rate and row spacing trial sown with Mace wheat at Minnipa, 2015 and 2016.

This research targeted barley grass weeds but there were also other grass weeds present. The average late barely grass population at the Minnipa trial was 11 plants/m² in 2015 and 15 plants/m² in 2016.

Seeding rate increased the number of wheat plants/m² however no rate achieved the targeted plant densities despite high moisture at seeding 2016, and 2015 was a drier seeding. The 18cm row spacing resulted in higher plant densities than the 30cm row spacing (Table 4), but the seeding system boots had no impact on plant numbers (data not presented). There were no differences in early weed numbers for row spacing or seeding rates (Table 2), however barley grass germination has been shown to be later in cropped paddocks.

Early crop dry matter was greater in the 30cm row spacing than in the 18cm, and this trend carried through to late dry matter. Seeding rate progressively increased dry matter early in the season but the effect had largely disappeared by late season dry matter cuts (Table 3).

Total late grass weed dry matter was lower in the higher seeding rate. The 18cm row spacing also had lower late grass weed dry matter compared to the 30cm row spacing (Table 3).

The late barely grass and ryegrass weed seed set followed similar trends to the grassy weed dry matter. Barley grass seed production was lower with narrower 18cm row spacing compared to 30cm (Table 3). There was no difference in ryegrass numbers or weed seed set with the narrow row spacing as ryegrass density was similar. The increase in seeding rate and plant density also decreased barley and ryegrass weed seed set (Table 3).

Grain yield increased with seeding rate (Table 4). The 18cm row spacing also out-yielded the 30cm row spacing for the second season, by 0.88t/ha in 2015 (but large differences in plant establishment between 18cm and 30cm row spacings) and 0.48t/ha in 2016, but again there were no differences between the two seeding systems (data not presented).

Method

Minnipa Agricultural Centre - row orientation trial

Replicated plot trials were sown in 2015 and 2016 with two row orientations; N-S and E-W. Treatments within row orientations included two row spacings, 18cm (7”) and 30cm (12”), sown with two different seeding boots (a Harrington knife point and an Atom-Jet spread row ribbon seeding boot). The plots were direct drilled with press wheels. Oats were spread as a surrogate weed at 70 plants/m², through hoses at the front of the seeder during the seeder pass in 2015 and before the seeder pass in 2016.

In 2015 the trial was sown into 2014 standing stubble in each orientation and additional ‘control’ plots were sown near each trial block but in the opposite row orientation to that in each block. In 2016 the trial was sown into a pasture paddock, not stubble, with two row orientations of E-W and N-S.

Trial measurements taken during the season included soil moisture, PreDictaB root disease test, soil nutrition, weed establishment, ‘weed’ germination, crop and weed establishment, crop and weed biomass (early and late), light interception in crop rows (using AccuPAR PAR/LAI ceptometer), grain yield and quality.

Results

Minnipa Agricultural Centre - row orientation trial

Using oats as a surrogate grass weed resulted in an even weed pressure across the large area of the trial which was unlikely to be achieved by only relying on the background grass weed levels. Using oat ‘weeds’ gives a relative indication of the outcome that would have been achieved with other grass weeds such as ryegrass and barley grass at high populations in the system. The oat-only treatment (no wheat sown) resulted in 72 plants/m², achieving the targeted plant density for weed pressure, unlike 2015 when the weed pressure was only 26 plants/m².

In 2015 E-W row orientation increased yield over N-S by 0.15t/ha in a season with a 2.7t/ha yield average in the trial. The trial results showed a decline in yield due to weed competition, but no direct effect on weed competition due to row direction. Sowing in an E-W direction may give a slight yield increase in higher yielding seasons but no differences were detected in weed seed set (data not shown).

In 2016 there were no interactions between row spacing, seed rate or seeding system in terms of the effect on weeds. There was no difference in crop establishment due to row direction with the average being 112 plants/m². There was a difference in plant numbers between the row spacing treatments, with 120 wheat plants/m² established in the 18cm row spacing treatment and 105 plants/m² in the 30cm row spacing (Table 5). The type of seeding point or the addition of weeds had no impact on wheat establishment.

There were no differences in late crop dry matter due to sowing direction or seeding systems in the absence of weeds (Table 5). The late crop dry matter was greater in the narrow row spacing than in the wider row spacing (Table 5).

In 2016 there was no detectable difference in wheat yield due to sowing direction in the absence of weeds (Table 5). The narrow row spacing resulted in higher yields compared to wider row spacing (Table 5), but with no significant differences in grain quality.

There was a significant difference in grain yield due to ‘weeds’ in the system with an average wheat grain yield decrease of 0.7t/ha. The ‘oat’ weed seed set averaged 0.23t/ha and there was no effect on weed seed set due to sowing direction or row spacing in 2016 (data not presented).

In both seasons the light interception measured as leaf area index (LAI) showed greater shading in the E-W sowing direction compared to N-S. The narrow 18cm row spacing also showed greater shading due to canopy cover compared to the 30cm row spacing (data not presented).

Table 5. Mace wheat growth, yield and grain quality with different sowing direction, row spacing and seeding systems at Minnipa 2016.

*LSD not available due to lack of replication (>8 required for statistical comparison)

** in absence of weeds

- Analysed data not provided

The gross margin analysis of yields of the different systems at Minnipa over two seasons show the change in row spacing had the greatest financial gain (Table 6) however many other issues must be considered to implement this within a farming system such as stubble management, timeliness and speed of seeding, herbicide efficacy, machinery changes or upgrades.

Seeding rate increases are easy to implement in a farming system, however plant counts should be undertaken to ensure adequate plant populations are being achieved to maximise returns.

Row orientation may have a small gain in good seasons however the length of paddock runs, hills and prevailing wind direction will impact on the decision to implement this change in the farming systems.

Table 6. Grain yield and gross margins with different row spacing, seeding rate and sowing direction at Minnipa in 2015 and 2016 ($193/t at Port Lincoln on 1 December).

*Non-significant

Method

Upper North – Seeding rate, crop type and variety

In 2015 and 2016 seasons, one wheat variety (Scepter) and two barley varieties (Fathom, a vigorous, competitive variety and Hindmarsh which is considered less competitive) were sown with three treatments for each variety - this involved two seeding rates (60 and 120kg/ha) and a further treatment which aimed at best practice weed control (high seeding rate of 120kg/ha plus appropriate chemical weed control of Sakura^® @ 118g/ha on wheat and TriflurX^® @ 2.5L/ha on barley). The crop was established using 72kg/ha 18:20:0:0 fertiliser with 70kg/ha urea banded below the seed with 30cm row spacings.

Results

Upper North – Seeding rate, crop type and variety

The average barley grass population at the trial site was 25 plants/m² in 2015 and 107 plants/m² in 2016.

In the presence of high grass weed numbers, wheat performed poorly compared to both barley varieties. Wheat showed grass carryover of 2 to 3 times that of barley. As in 2015, doubling the wheat seeding rate provided no benefit. The 2016 wheat yield data is questionable, given the level of frost impact, but also supports the fact that Scepter wheat performed quite poorly as a competitor to barley grass, when compared with barley.

Comparison of species and variety impact on weed infestation and seed set at different seeding rates

At the higher seeding rate of 120kg/ha, weed measurements taken at anthesis showed that both barley varieties had reduced grass weed panicles to well under half that observed in the wheat plots (Table 7). At the low seeding rate, this reduction in grass seed carry-over was still evident, but not to the same extent (Table 8).

Table 7. Species and variety impact on weed infestation at 120kg/ha seeding rate.

Table 8. Species and variety impact on weed infestation at 60kg/ha seeding rate.

Discussion

The 18cm row spacing achieved higher plant numbers than the 30cm row spacing with the same seeding rate, but the seeding system (ribbon or narrow boots) had no significant impact on crop numbers.

There were no differences in early weed numbers due to row spacing or seeding rates. At Minnipa the paddock barley grass populations have been shown to have a later germination pattern in the paddock compared to areas which are not cropped. The total late grass dry matter declined with the higher seeding rate, and also declined with narrower row spacing. Late grassy weed dry matter was 65% lower, and barley grass weed seed set was 57% lower with a higher seeding rate. The late barley grass showed similar trends, with decreasing weed seed set in the narrow row spacing, and also the higher seeding rate.

In the 2016 season the 18cm row spacing again yielded higher (+0.48t/ha) than the 30cm system. In 2015 the higher seeding rates also resulted in higher grain yield, but larger differences were recorded in plant numbers, but no grain quality differences were present in either season. Previous research from WA showed there is no difference in yield due to row spacing in crops that yield less than 0.5t/ha, but in crops with yields greater than 3.0t/ha there is a yield penalty with wider row spacing. The yield penalty experienced by the wheat crops (with yields between 2.7 – 3.4t/ha) was an 8% decrease in yield for every 9cm increase in row spacing (GRDC, 2011).

A more recent review of row spacing of winter crops in broad scale agriculture in southern Australia, by Scott et al. in 2013, suggests the direct effect on yield of adopting wider rows (reduced yield at greater than 18cm) has often been overlooked, due to the relative ease of stubble management in wider rows. At yields of 2.0t/ha widening row spacing from 18cm to 36cm reduced yield by 1860kg/ha (Scott, 2013). This review also noted crops sown on wider rows are less competitive with weeds, mainly ryegrass.

Research into using crop competition for weed control in barley and wheat in 2015 at Hart showed varying the seeding rates (increasing from 100 to 300 plants/m²) reduced the yield loss due to weed competition (Goss, 2015). This research also showed there were differences in wheat and barley varieties’ ability to compete with grass weeds, and it also found no difference between normal or spreader seeding boots (Goss, 2015). Spreader boots were used to try to reduce the row spacing (by spreading the seed) and increase grass weed competition, however this effect did not occur at Minnipa in the last two seasons.

Row orientation research at Minnipa has shown in poor seasons N-S sowing increased yield due to shading effects, however in good seasons E-W orientation had yield benefits. The presence of ‘weeds’ reduced grain yield but there was no differences in weed seed biomass or seed set due to row orientation. Light interception showed greater shading in the E-W sowing direction compared to N-S. The narrow 18cm row spacing also showed greater shading due to canopy cover compared to the 30cm row spacing.

Research in the Upper North of SA showed barley sown at higher seeding rates was more effective than wheat at reducing barley grass seed set, particularly with more vigorous varieties such as Fathom, compared to less vigorous varieties such as Hindmarsh (Mudge, EPFS Summary 2016). This research demonstrated that increasing the seeding rate of barley in situations where barley grass is not controllable by herbicides, can have substantial benefits, both in terms of yield and reducing weed seed carryover. Wheat would not be a preferred option in such circumstances and increasing seeding rate of wheat in high grass pressure (104 barley grass plants/m²) is unlikely to provide any benefit.

At Minnipa achieving 166 plants/m² instead of 87 plants /m² (targeted rate was district practice rate of 120 plants/m²) reduced barley grass seed set by 57% and ryegrass by 36%. Sowing to achieve a district practice seeding rate of 60kg/ha (actually 108 plants/m² establishment) at 18cm spacing instead of 30cm led to a 44% decrease in barley grass seed production. Overall the reduction in barley grass numbers demonstrates that using crop competition (either by using a narrow 18cm row spacing or by increasing plant density) is a potentially effective non-chemical methods to reduce barley grass and ryegrass numbers in current farming systems. Using narrow row spacings of 18cm with wheat crops with an expected yield greater than 2t/ha has been shown to have a yield advantage in this environment.

Conclusion

• Reducing row spacing to 18cm from 30cm resulted in 42% lower grass weed dry matter and increased grain yield of wheat by more than 0.5t/ha in 2015 and 2016 at Minnipa, in trials with greater than 2t/ha wheat yields.
• Late grass weed dry matter was 65% lower, and barley grass weed seed set was 57% lower with a higher seeding rate, and increasing seeding rate also increased grain yield at Minnipa in 2015 and 2016.
• There may be small yield benefits as a result of E-W sowing in average to good seasons over N-S sowing (with no difference in weed seed set), however paddock orientation, hills and wind direction will be influencing factors in implementation.
• Increasing the seeding rate of barley in the presence of barley grass can provide substantial benefits to both yield and levels of weed seed carryover. This applies particularly to competitive varieties such as Fathom but also to less competitive varieties such as Hindmarsh.
• In contrast, doubling the seeding rate of wheat in 30cm row spacing had no beneficial effect on yield or weed carryover at Appila in the Upper North.
• Doubling the district practice seeding rate in barley substantially reduced the competitive effect of barley grass to the stage where crop yields were similar to those of control plots where herbicide was applied.
• During the trials, barley consistently outperformed wheat in its ability to compete with barley grass, particularly when sown at high seeding rates.

References

Cook, A. & Richter, I. (2017). Seeding rate by row spacing for barley grass management.  In B. Gontar (Ed.), Eyre Peninsula Farming Systems Summary 2016 (pp. 105-107). Port Lincoln: The Printing Press.

Cook, A., Richter, I., Dyson, C. & Wilhelm, N. (2017). Row orientation, seeding systems and weed competition. In B. Gontar (Ed.), Eyre Peninsula Farming Systems Summary 2016 (pp. 108-111). Port Lincoln: The Printing Press.

Mudge, B. (2017) Overdependence on agrochemicals – UNFS barley grass trial In B. Gontar (Ed.), Eyre Peninsula Farming Systems Summary 2016 (pp. 122-126). Port Lincoln: The Printing Press.

Cook, A., Shepperd, W., Richter, I., & Wilhelm, N. (2016). Seeding rate by row spacing for barley grass management In B. Gontar (Ed.), Eyre Peninsula Farming Systems Summary 2015 (pp. 159-162). Port Lincoln: The Printing Press.

Cook, A., Wilhelm, N., Shepperd, W., & Richter, I. (2016). Row orientation and weed competition. In B. Gontar (Ed.), Eyre Peninsula Farming Systems Summary 2015 (pp. 163-165). Port Lincoln: The Printing Press.

Mudge, B. (2016) Overdependence on agrochemicals – UNFS barley grass trial In B. Gontar (Ed.), Eyre Peninsula Farming Systems Summary 2015 (pp. 166-170). Port Lincoln: The Printing Press.

GRDC Factsheet, Crop placement and Row Spacing, Jan. 2011.

Scott, B.,Martin P., Riethmuller G.Row spacing of winter crops in broad scale agriculture in southern Australia (2013). Monograph No. 3, Graham Centre.

Goss, S., and Wheeler, R. (2015) Using crop competition for weed control in barley and wheat, GRDC Update Papers, Adelaide, 2015.

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.

Thank you also to Sue Budarick, Tegan Watts and Katrina Brands for conducting the weed counts.

Contact details

Amanda Cook
SARDI, Minnipa Agricultural Centre, PO Box 31, Minnipa, SA 5654
0427 270 154
Amanda.Cook@sa.gov.au