New pasture opportunities to boost productivity of mixed farms in low/medium rainfall areas

Take home messages

• A critical assessment of the regional performance of existing and new pasture legumes over two years has shown that annual medics continue to provide the best pasture option for neutral/alkaline sandy soils in the Mallee. Common vetch is an alternative option where a sown legume ley of one-year duration is preferred.
• PM-250 strand medic will be released in 2021. For the first time, it combines resistance to the foliar fungal pathogen, powdery mildew, and tolerance to sulfonylurea herbicide residues.
• Aim to maximise pasture legume seed set in the establishment year. Legumes with very high hard seed levels (>90%) are best cropped in the year following establishment.
• Differences in legume production and N2-fixation have been measured and impacts on wheat production will be measured at multiple sites in 2020.
• Alternative pasture establishment methods (for example; summer sowing) are viable in the Mallee, however, are not suitable for all legume species. Further investigation is needed to define the conditions where summer sowing and twin sowing practices are reliable.

Background

A project, informally known as the Dryland Legume Pasture Systems (DLPS) project, is evaluating a diverse range of annual pasture legumes on mixed farms in the low to medium rainfall zone (<450mm). The DLPS project aims to:

• Provide a critical assessment of the regional performance of existing and new pasture lines.
• Determine if pasture legumes can be established more efficiently.
• Quantify the benefits provided by pasture legumes to crops and livestock.

Legumes close to commercial release, including strand medic line, PM-250, existing legumes not widely utilised in south-eastern Australia (for example; serradella, bladder clover and biserrula) as well as undomesticated legumes (for example; Trigonella and Astragalus spp.) are being compared with traditionally grown medics and vetch. Commercial legume species options are shown in Table 1. Legume production, N2 fixation, nutritive value and ability to regenerate after cropping phases is being measured to understand different legume species adaptation to soil type, so that growers can be confident in their performance and benefits for the crops that follow.

A significant obstacle to the adoption of new pastures legumes is the high cost of pasture seed and difficulty in establishment, particularly in low to medium rainfall areas. A feature of some legumes under investigation is their aerial seeded habit and retention of seed, allowing seed to be grower harvested and re-sown with standard cropping equipment. The project is examining the potential of different pasture legume species to be established more efficiently.

Table 1. Annual pasture legumes. Cultivars (release date and key traits) and indicative adaptation.

Results and discussion

A new strand medic

A new strand medic (Medicago littoralis) cultivar is scheduled for release in 2021. Currently known as PM-250, it provides a significant advantage over the cultivar Angel, which it will replace. PM-250 combines for the first time, resistance to the foliar fungal pathogen, powdery mildew (Erysiphe trifolii), and tolerance to sulfonylurea (SU) herbicide residues. It is suited to neutral and alkaline sandy loams receiving 275 to 400mm rainfall.

The commercialisation of PM-250 is based on the assessment of its performance at 10 field sites. Across these sites, 34 assessments of dry matter (DM) production and 12 assessments of seed yield were completed. Overall, PM-250 produced 16% more DM than Angel medic and similar high seed yields (Figure 1). Production increases are likely to be greatest, but not limited to, where powdery mildew occurs. Increases of up to 49% and reduced levels of the phytoestrogen, coumestrol have been measured in the presence of powdery mildew. PM-250 is being further assessed in the DLPS project described below.

Figure 1. Relative mean herbage production (% site maximum) and seed yield (kg/ha) of PM-250 and Angel strand medics across ten field sites. Includes 34 assessments of dry matter production and 12 seed yield assessments. Bars above columns indicate standard error.

Casting the net to identify the next opportunity

On 15 June 2018, 30 annual pasture legumes (12 medics, 10 clovers, two serradellas, two lotus, two trigonella, biserrula and astragalus) and two vetches were established at the earliest opportunity after late opening rains in a small plot trial at Lameroo, SA. A similar trial sown at Minnipa SA (27 June 2018) contained an extra vetch, but only seven clovers (Table 2). The Lameroo trial was located on a sandy soil, on the lower-mid dune (pHCa 5.8). The Minnipa trial was located on a uniform area of sandy loam (pHCa 7.8). Seed was inoculated with the appropriate rhizobia strain and sown at 5, 7.5, 10 or 40kg/ha germinable seed for the small, small-medium, medium and large seeded legumes, respectively. Plots were un-grazed and managed to maximise seed set. In 2019, the legume plots were allowed to regenerate. Plant DM production (2018 and 2019), seed set (2018) and plant regeneration (2019) were measured.

Growing season rainfall was 48% in 2018 and 71% in 2019 of the long-term average at Lameroo (269mm), and 62% in 2018 and 89% in 2019 of the long-term average at Minnipa (242mm).

Performance of commercial legume species

Production in 2018 was limited to less than 1,500kg/ha by seasonal conditions. Even so, differences in the production and seed set of the commercially available legumes were measured (Figure 2). Vetch was most productive (1,098kg DM/ha), followed by barrel medic (820kg DM/ha) and strand medic (688kg DM/ha). Barrel medic was the most productive pasture species at Minnipa, consistent with the recommendation for use on alkaline loam soils. Legumes developed for acidic sands in WA (bladder clover, serradella and biserrula) were less productive.

Seed set of the commercial legume species generally exceeded 200kg/ha, with the exception Margurita serradella (129 and 47kg/ha) and vetch at Minnipa (55kg/ha) (Figure 2). The later flowering time of the French serradella likely contributed to its low seed production.

Figure 2. Dry matter production, seed set and regeneration of strand medic (multiple cultivars), barrel medic (multiple cultivars), bladder clover (cv. Bartolo), rose clover (cv. SARDI Rose), French serradella (cv. Margurita), biserrula (cv. Casbah) and vetch (cv. Studenicia) at Lameroo and Minnipa, SA. Numbers in parentheses accompanying the legume name (e.g. 62, 58) indicate % performance relative to the best legume entry at Lameroo and Minnipa, respectively. Bars above columns indicate standard error.

There were large differences in legume regeneration in 2019. Strand and barrel medics regenerated adequately (>200plants/m2) at both sites, as did rose clover at Lameroo. This provided some flexibility to extend the pasture phase into a second year and consolidate the seed bank (Figure 2). Although biserrula produced a reasonable seed yield in 2018, it regenerated at <20plants/m2 in 2019. This is due to its high hard seed level (Table 1) and is consistent with the recommendation that this legume be cropped the year following its establishment, to enable some breakdown of hard seed. Vetch, which has been selected to have <5% hard seed to prevent it becoming an in-crop weed, did not regenerate.

DM production of the commercial legumes in 2019 was generally consistent with the results for 2018. The annual medics (developed for alkaline soils) generally produced most winter DM. Rose clover performed better on the sandy loam soil at Lameroo. The WA bred legumes produced less DM, the result of poor regeneration (for example; Casbah) and sub-optimal adaptation to soil type.

Performance of other pasture legume species, cultivars and lines

Ranked performance of all legumes sown at Lameroo and Minnipa is shown in Table 2.

In 2018 when growing season rainfall was less than 200mm, vetches and barrel medics were consistently the most productive species. In the absence of powdery mildew, PM-250 strand medic ranked 11th, achieving about 65% of the best legume lines, namely Studenica vetch at Lameroo and Caliph barrel medic at Minnipa. Rose clover and astragalus were the most productive alternative species, even though astragalus is known to have been constrained by poor nodulation.

In 2019, strand medics (Herald, Harbinger, Jaguar, PM-250 and Pildappa) and the strand medic hybrid (Toreador) occupied six of the top ten ranked positions. They established and grew well at both sites, regenerating at >250plants/m2 and producing more than 1,100kg/ha biomass. Sultan SU was the best barrel medic (rank 9th). Caliph and Cheetah (best two pasture legumes in 2018), performed less well in 2019, falling to ranks 17 and 20. The best alternative legumes were the early flowering selection of trigonella, burr medic with putative boron tolerance, rose clover and two lotus species. These legumes performed best on the loam soil at Minnipa. Astragalus fell to rank 27 in 2019, due to high hard seed levels.

Legume performance in other environments

A sub-set of the legumes in Table 2 is being tested in other low rainfall environments.

On a neutral (pHCaCl 7.4) sandy soil in Piangil, Victoria, the production of several legumes established in 2019 exceeded 4,000kg/ha, more than double that measured at the SA sites. Even so, relative legume production at Piangil was significantly correlated (n=19, P<0.01, R2 = 0.57) with production in the establishment year (2018) at Minnipa (Table 2). Studenica vetch (4,880kg/ha) and the barrel medics (Caliph, Sultan SU and Cheetah) were most productive (≥3 500kg/ha). Margurita, Santorini serradellas and biserrula produced less than 1 ,000kg/ha DM at Piangil.

In NSW, legume performance has been different on the acidic red loams at Kikoira (pHCaCl 4.9) and Condobolin (pHCaCl 5.1). In trials established in 2018, biserrula was the outstanding species across both sites. It was the only legume to survive extreme drought conditions at Condobolin. Biserrula produced more than 120kg/ha seed at Kikoira with approximately one-third of that produced prior to the end of October. Other species including Margurita, Santorini serradella and arrowleaf clover also produced useful quantities of herbage (around 1,200kg/ha) under severe drought at Kikoira but had not commenced reproductive growth by late October. Whilst they managed some seed set after 53mm rain in November, had this not occurred, these later maturing species may have failed to produce seed. Both Casbah biserrula and Lotus ornithopodioides regenerated well in 2019.

Table 2. Ranked performance by mean of measures and sites, of 33 legumes in 2018 (establishment year) and 2019 (regenerating year). Parentheses show % performance relative to the best legume at Lameroo and Minnipa respectively.

Pastures in rotations

A cropping systems experiment at Lameroo is evaluating the duration of pasture benefits and pasture regeneration after cropping, using a range of legume species grown for two years, (PM-250 medic, Margurita serradella, SARDI rose clover and Trigonella balansae), or one year (PM-250 medic, and Margurita serradella). Crop benefits will be measured in 2020 after the one or two-year pasture phase, when the pasture systems will be compared against three control treatments; vetch-cereal, pea-cereal and continuous cereal. Similar experiments (not reported here) are being undertaken at Piangil in Victoria, and Harden and Uranquinty in NSW.

Growing season rainfall at Lameroo in 2018 was 140mm. In 2018 pastures were established primarily to set seed for regeneration in 2019. Seed set was adequate for each species and was estimated to range between 190-320kg/ha. PM-250 medic produced the greatest DM up until late September (1.8t/ha, Table 3), however late rains in October/November (33mm) may have supported some further growth and seed set of the later flowering species, particularly serradella.

After the first season, soil mineral nitrogen was the parameter that varied most. Measured in early 2019 it reflected N fixed by the pasture species in 2018 (Table 3), medic>rose clover>trigonella>serradella>wheat. Some serradella plants were pale yellow and because nodulation in adjacent plots was observed to be less than ideal, we speculate that sub-optimal nodulation was probably limiting in the system experiment. While there were significant differences in nutritive values of metabolisable energy (ME), digestibility and crude protein, they were not large.

Table 3. Summary of N2-fixation and biomass production, metabolisable energy (ME), digestibility and crude protein at peak biomass from 2018 sampling, and soil mineral nitrogen and moisture from 0-100cm from soil cores taken in May 2019 at Lameroo, SA.

In 2019, regenerating pasture treatments had higher plant establishment than plots sown in autumn, namely PM-250 and Margurita. PM-250 density in the regenerating plots was five times (232 versus 38plants/m2) and Margurita density seven times (373 versus 47plants/m2) levels in the sown plots. Rose clover and trigonella regenerated at 304 and 151plants/m2, respectively. These differences affected production (Figure 3).

Growing season rainfall (April to October) in 2019 at Lameroo was 205mm. All treatments produced most DM in mid-October when all species were podding (Figure 3). There was no significant difference at the mid-October cut between autumn sown medic, regenerating medic, regenerating rose clover and regenerating serradella. The extent to which lower production of trigonella and autumn sown serradella effect crop production, will be measured in 2020.

Figure 3. Biomass in Lameroo 2019 for pasture species either sown on 14 May 2019 (○) or regeneration from seed set in 2018 (●). Solid vertical line in the Medic figure is LSD (5%) at each biomass measurement, and markers are the date that first flowers were observed in autumn sown treatments (∆) and regenerating treatments (×).

Pasture establishment in the Mallee

Alternative pasture establishment methods were evaluated at Waikerie, SA, and Piangil, Vic, in 2019 using a range of annual pasture legumes, including some not commonly grown in the Mallee region. Indicative sowing rates are shown in Table 4. Establishment methods evaluated were:

*             Twin-sown, where ‘hard’ pasture seed/pod was sown with wheat seed in 2018 for pasture establishment in 2019.

*             Summer-sown (February), where ‘hard’ seed/pod was sown in summer and softens to establish on the autumn break.

*             Autumn-sown (control treatment), where ‘soft’ germinable seed is sown on the break of the season.

In 2019 at Waikerie, the seasonal break occurred on 9 May with 20mm rainfall. Rainfall prior to 9 May was 22mm. In Piangil, the seasonal break occurred on 2 May with 19mm rainfall, and rainfall prior to 2 May of 17mm. At both sites, all establishment treatments emerged within two weeks of each other. Sowing method had a significant effect on plant density at both sites (Figures 4A and 4B). The targeted population for sown pastures is typically 150-200plants/m2.

Table 4. Indicative rates of sown pod or seed (kg/ha) and equivalent amount (kg/ha) of viable hard seed sown in twin- and summer-sown treatments; and rate of germinable seed (kg/ha) in the autumn sown treatment.

Seedling establishment

At Waikerie, mean plant density across all legumes were; autumn-sown 132plants/m2, twin-sown 64plants/m2 and summer-sown 159plants m2 (Figure 4). In Piangil, mean plant density across all legumes were; autumn-sown 73plants/m2, twin-sown 42plants/m2 and summer-sown 60plants/m2. An observation relevant to the lower establishment in twin-sown plots, is that seed may have been buried too deep as a result of collapse of furrows and sand movement over the 2018/19 summer period.

At both sites, serradella had the highest establishment for all twin- and summer-sown treatments compared to other species but established best when summer-sown. Medic densities were greatest when autumn sown.

Figure 4. Plant establishment resulting from different establishment methods at A) Waikerie on 25 June 2019 and B) Piangil on 5 June 2019.

Production

Treatment differences in dry matter production were measured at Waikerie, despite production being limited by low rainfall (Figure 5). Production was greatest for summer and autumn-sown PM-250 medic. Although serradella and rose clover produced more DM when summer-sown, their overall production was lower, suggesting they are less well adapted to Mallee soils. Dry matter was lowest in the twin-sown treatment, consistent with lower plant numbers.

Figure 5: Biomass production in 2019 at Waikerie in the establishment treatments; autumn-sowing (●), twin-sowing (○) and summer-sowing (■).

At Piangil, twin-sown treatments performed better than at Waikerie (Figure 6). Higher plant density did not necessarily result in higher biomass production. For example, there was higher plant density in summer-sown serradella, but twin-sown treatments produced more biomass. Medic produced similar biomass in the autumn- and twin-sown treatments. Production of trigonella and gland clover was generally low, indicating they are less adapted to the soil type.

Figure 6. Biomass production in 2019 at Piangil in the establishment treatments; autumn-sowing (●), twin-sowing (○) and summer-sowing (■).

Results from 2019 indicate that twin and summer-sowing may be a viable establishment method for the Mallee region, however it is not suitable for all legume species. In both environments, Margurita serradella gained most advantage from the alternative establishment methods. Results for PM-250 medic were inconsistent, with twin-sowing inferior at Waikerie and summer-sowing inferior at Piangil. Given that all treatments emerged on similar dates, and there was very little summer rainfall in 2019, further exploration of the methods are required under a range of growing seasons such that risks and/or benefits associated with earlier seasonal or false breaks can be evaluated.

Weed management

Weed control is an important consideration with twin and summer-sowing. At Waikerie there were significantly greater numbers of broad leaf weeds in the twin and summer-sown plots, compared to autumn-sown plots. Weed DM for the treatments was; twin-sowing 500kg/ha, summer-sowing 440kg/ha and autumn-sowing 360kg/ha (P<.001). Autumn-sown plots received a knock-down spray at sowing, while twin and summer-sown plots did not. Twin and summer-sowing methods should only be considered for paddocks with low weed levels.

Seasonal analysis

To understand the likely suitability of summer and twin-sowing in other low rainfall environments, historic climate records (1970 to 2018) were analysed to reveal 25th to 75th percentiles of when the seasonal break occurred. Using the APSIM model (version 7.10) and historic weather records, the approach of Unkovich (2010) was used to estimate the mean break of a season, that is, when over a seven-day period, accumulated rainfall exceeds accumulated pan evaporation. An additional rule was added, which was that soil temperature should be below 20°C. Figure 7 shows ‘box and whisker’ plots for six locations, and the probability of a break occurring on 25 April.

The analysis revealed that Lameroo and Condobolin have the earliest median break, and higher probability of a break occurring before 25 April, while Minnipa and Waikerie typically have the latest seasonal break. In environments with a greater probability of an early seasonal break, summer-sowing will likely be more beneficial — soil conditions are warmer, and a longer growing season can be exploited more often. In environments where the seasonal break is often later, there is greater risk of seed losses or burial, rhizobia death and exposure to pathogens. Establishment following autumn, summer and twin-sowing methods will also be measured in Lameroo in 2020.

Figure 7. ‘Box and whisker’ plots showing 25th to 75th percentiles of when the autumn break occurred in historic data set 1970-2018, using Unkovich (2010), and the probability of the seasonal break occurring on 25 April.

Conclusion

Pasture legume production, regeneration and persistence is determined by multiple factors (Nichols et al. 2012), including adaptation to soil type (texture and pH), capacity to set seed (early flowering desirable in low rainfall areas) and hard seed levels that allow regeneration and persistence through the cropping sequence.

On neutral/alkaline soils in the low rainfall regions, annual medics continue to provide the best option where a self-regenerating pasture is preferred. The SA trials reported in this paper, reiterate strand and disc medics as the best pasture legume choice for the lighter sands and barrel medics for the heavier loams in the Mallee. PM-250 strand medic is scheduled for release in 2021 and has demonstrated a production benefit of 16% over the cultivar Angel which it will replace. In addition, larger benefits are expected where powdery mildew and herbicide residues are present. Cohorts of disc, strand and burr medic have been developed and are being assessed by the DLPS project.

Legumes developed for WA soils and farming systems (biserrula, serradella and bladder clover) have so far performed less well on Mallee soils in SA but have performed well on other soil types. Specifically, biserrula has grown and regenerated well on acidic red sands in NSW. Pasture legume species other than medics have on occasion shown promise in the Mallee but have neither been outstanding or consistent. If trialling the ‘alternative’ species, it suggested that small areas are initially sown. Common vetch may be a better option where a sown legume ley of one year is preferred, because of its ability to provide early production and options for late weed control. A new vetch cultivar (Studenica), scheduled for release in 2021, has performed well in the DLPS trials.

The aim in the establishment year of legume pastures should be to maximise seed set, and if done well the resultant seed bank (25 times what is sown) will support pasture regeneration for many years. Alternative establishment methods have demonstrated potential in the Mallee but are not suitable for all legume species. Margurita serradella gained greatest advantage from the alternative establishment methods. Results for PM-250 medic were inconsistent but showed some promise and are worthy of further investigation given their potential to provide growers with greater sowing flexibility and reduce seed costs. Differences in N2-fixation by the different legumes have been measured. The impact of this and other pasture impacts on wheat production will be measured in 2020.

The studies reported in this paper have focussed on legume monocultures. Legume mixtures such as medic and vetch in the establishment year may be useful to achieving more consistent production through the season and across variable soils.

Acknowledgements

The research reported in this paper is made possible by significant contributions of growers through both trial cooperation and the support of GRDC and the South Australian Grains Industry Trust.

The Dryland Legumes Pasture Systems project is supported by funding from the Australian Government Department of Agriculture as part of its Rural R&D for Profit program, the GRDC, Meat and Livestock Australia and Australian Wool Innovation. The research partners include the South Australian Research and Development Institute, Murdoch University, the Commonwealth Scientific and Industrial Research Organisation, the WA Department of Primary Industries and Regional Development, and Charles Sturt University, as well as grower groups.

We thank the Pocock and Schmidt families for hosting trials at Lameroo and Waikerie, SA.

Useful resources

Wurst, M., Pasture legumes for temperate farming systems. The ute guide. (2004) (Primary Industries and Resources South Australia and Grains Research and Development Corporation).

Pasture picker. Pastures Australia.

Legumes for temperate pastures. EverGraze.

References

Flohr, B.M., Hunt, J.R., Kirkegaard, J.A. and Evans, J.R. (2017) Water and temperature stress define the optimal flowering period for wheat in south-eastern Australia. Field Crops Research 209, 108-119.

Nichols P. G. H., Revell C. K., Humphries A. W., Howie J. H., Hall E. J., Sandral G. A., Ghamkhar K. and Harris C. A. (2012) Temperate pasture legumes in Australia—their history, current use, and future prospects. Crop and Pasture Science 63, 691-725.

Puckridge, D.W. and French, R.J. (1983) The annual legume pasture in cereal ley farming systems of southern Australia- a review. Agriculture Ecosystems & Environment 9, 229-267.

Revell, C.K., Ewing, M.A. and Nutt, B.J. (2012) Breeding and farming system opportunities for pasture legumes facing increasing climate variability in the south-west of Western Australia. Crop and Pasture Science 63, 840-847.

Unkovich, M. (2010) A simple, self-adjusting rule for identifying seasonal breaks for crop models. In ‘Proceedings of the 15th ASA Conference - Food Security from Sustainable Agriculture’. (Lincoln, New Zealand).

Contact details

Ross Ballard
SARDI Soil Biology and Diagnostics
GPO Box 397, Adelaide SA 5001
+61 8 8429 2217
ross.ballard@sa.gov.au