Resilient pastures for low rainfall mixed farms — crop and system benefits provided by legumes

Take home messages

• Mean wheat yield after medic pasture was increased by 1.11t/ha (+44%) and grain protein by 0.7%, compared to a continuous cereal rotation, across four sites.
• A simulation study based on a medium-low rainfall site found that the inclusion of legumes in the rotation can contribute 14—70kg N/ha per year to the nitrogen (N) bank depending on the intensity at which they are grown.
• Despite the upfront establishment costs of a legume phase, gains in subsequent cereal crop yields alone can provide substantial return on investment.
• Medics were most consistent for production and regeneration on neutral/alkaline sandy loam soils, but were less amendable to on-farm seed harvesting than other legumes.
• Common vetch is a good option where a sown legume ley of one year duration is preferred, and a regenerating pasture legume seedbank is not a priority.
• Two new medics and an arrowleaf clover are being developed for commercial release.

Background

The Dryland Legume Pasture Systems (DLPS) project is evaluating a range of annual pasture legumes on mixed farms in the low to medium rainfall zone (Ballard et al. 2020). The project aims to:

• provide a critical assessment of the regional performance of existing and potential pasture cultivars
• quantify the benefits provided by pasture legumes to cropping systems.

This paper reports on the findings from four trials targeting neutral and alkaline sandy loams receiving 275 to 400mm rainfall (Table 1). Sites were cropped with wheat in 2020, following i) sown vetch or ii) one year of sown pasture or iii) two years of pasture that had been sown in 2018 and regenerated in 2019. Regenerated pasture treatments were at Lameroo and Minnipa sites only. Legume break effects on wheat grain yield, protein and available soil N are reported, and performance of the different legume species are briefly discussed.

Legume production, benefits to the cereal crop and subsequent legume regeneration have been measured to understand adaptation of different legume species to the environment and farming system such that growers can be confident in their performance and benefits.

Results and discussion

Crop benefits

The mean medic pasture break effect was substantial, producing an additional 1.11t/ha of grain yield and increasing grain protein by 0.7%, compared to the continuous cereal treatment (Table 1).

At Lameroo, a larger break effect was measured after a second year regenerating medic pasture in 2019 (that is, 2 years of pasture), although the one year medic pasture sown in 2019 still resulted in 57% additional wheat yield. One instance of reduced wheat yield after medic pasture occurred at Minnipa.

Table 1: Wheat grain yield (kg/ha) and protein content (%) in 2020; following cereal, vetch and medic pasture treatments in 2019.

Differences in soil mineral N were measured (Table 2). The mean increase in available N was 42kg/ha and 35kg/ha after vetch and medic respectively. The soil N increase was greatest (81kg N/ha) at Piangil. There was a positive relationship between available N and grain yield at Lameroo and Piangil.

Table 2: 2019 legume and cereal biomass and 2020 soil mineral N preceding the cereal crop.

Legume biomass measured 16 to 20 Sept. (Lameroo, Waikerie, Piangil), 3 Oct (Minnipa).

The economic value of crop benefits from investing in legume phase

Using the Lameroo trial site results (see Tables 1 and 2), a return on investment (discounted cashflow) approach was used to evaluate the benefits of various legume phases to subsequent crops. This includes a 2-year 2018 sown medic (Seraph); serradella (Margurita); Trigonella and rose clover phase with regenerating pasture in 2019; and a single year sown serradella, medic, field peas and brown manure vetch compared to a continuous cereal treatment (wheat-barley-wheat-triticale). Assumed costs (based on PIRSA 2021 Gross Margin Guide) included pasture establishment ($145/ha), pasture management ($58/ha), and the opportunity cost (foregone profit) from not growing cereal in the legume years.

In each case, the value of benefits to subsequent cereal yield alone exceeded the costs of the legume phase (Figure 1). Largely due to very high legume break effects on wheat yield 2020, the net present value of including a legume was over $200/ha for the field pea, brown manure vetch and 2-year medic and serradella phases. It is important to note that the trials were not grazed, so no grazing value was included, nor was the ongoing value of the pasture legume seedbanks that were established. The effect of higher protein on grain price was also not included.

Figure 1. Return on investment from one-year (2019) and two-year (2018-19) legume phases ($/ha cumulative discounted cashflow) relative to continuous cereal sequence at Lameroo 2018-19 based on 2020 and 2021 cereal yield.

Despite the costs of establishment, gains in subsequent crop yields alone can provide substantial return on investment. Major value is also possible as the pasture species is then able to be generated through grazing and the legume seedbank in future pasture phases. The potential to take a long-term approach to legume nitrogen value is explored in the next section.

Long term systems modelling

An emerging approach to determining the amount of nitrogen to supply to crops has been coined the “N bank” strategy (Smith et al., 2019; Meier et al. 2021). The approach aims to maintain a level of N in the soil that will not limit cereal production (Meier et al. 2021), though it requires further field validation particularly in medium and low rainfall environments (Farrell et al. 2021; Meier et al. 2021). Through simulation modelling, we aimed to quantify the contribution of legumes in the rotation to the N bank.

APSIM was used to investigate the long-term (1991-2020) N contribution of grain and brown manure legumes (field pea) to the farming system. Soil and meteorological parameters were based on the Lameroo site. Nitrogen fertiliser was only applied to wheat crops in the rotation. All wheat crops received top-up nitrogen fertiliser at sowing to 40kg N/ha if soil mineral N in the surface 1m of soil was less than 40kg N/ha. In N bank scenarios, wheat crops received top-up nitrogen to the amount of the N bank at 65 days after sowing. At Lameroo, the optimal N bank for grain yield was determined as 120kg N/ha using methods of Meier et al. (2021). Grain legume and brown manure legumes were then grown in sequence at increasing intensity to determine the contribution of nitrogen to the N bank and the subsequent cereal crop yield (Table 3).

Table 3: Simulated annual nitrogen fertiliser applied to wheat to maintain the target N bank of 120kg/ha and wheat yield when grown in sequences with different legume intensities. Simulations were run from 1991-2020 and phased such that a grain yield was obtained for each year.

Current and future pasture legume options

Cereal yield response was positive for all legumes, in the range of +26 to +51% (Table 4). Vetch was the most productive (mean 1,851kg/ha, range 320 to 2,460kg/ha, 128% compared with Seraph) and readily harvestable legume (Table 4). Vetch remains the best option on neutral alkaline soils where a sown legume ley of one year duration is preferred. Amongst the pasture legume options, Seraph medic was most productive (mean 1414kg/ha, range 620 to 1,800kg/ha) and the only treatment to consistently regenerate after the cereal crop (mean 489 plants/m2, range 277 to 1,279 plants/m2). Whilst differences between the pasture legumes on cereal yield were modest, greater effect is expected longer term as differences in legume regeneration and production accrue.

Table 4: Cereal yield response as per cent increase compared to cereal on cereal treatment across four sites: Lameroo, Waikerie, Minnipa and Piangil. Production before wheat and regeneration after wheat (per cent of medic treatment) of five pasture legumes and common vetch on neutral and alkaline sandy loam soils. Seed harvestability (mean two sites) shown as kg/ha and per cent recovered of total seed.

Although Margurita French serradella and Trigonella APG5045 were readily harvestable (providing seed for resowing) compared to Seraph medic, they were less productive (Table 4). Margurita serradella regenerated at useful numbers at Lameroo (296 plants/m2) but not elsewhere. Trigonella failed to regenerate after crop at three of the four trials.

Annual medics

The performance of Seraph strand medic (formerly known as PM-250) confirms its suitability for neutral and alkaline sandy loams receiving 275 to 400mm rainfall (Ballard et al. 2020). Along with other contemporary medic cultivars, it has hard-seed levels that allow it to persist and regenerate after crop and soft enough to allow consecutive pastures to be grown. At Minnipa and Waikerie, sown and naturalised medics were the only legumes to regenerate at useful numbers after crop. The strand medics are best suited to sandy loam soils and barrel medics for loam-clay loams. Burr medics are best suited to loam-clay soils and tolerate lower soil pH (minimum 4.8 CaCl2).

Two spineless burr medic cohorts: one tolerant of boron (B), another tolerant to red legged earth mites (RLEM) are being evaluated. Two lines have been identified with high agronomic performance and RLEM tolerance. Five B tolerant lines have also been shortlisted and it is expected a line will be identified autumn 2022 for cultivar release.

Disc medics are well adapted to deep alkaline sandy soils. Historically, the cultivars Tornafield and Toreador were sold but no cultivar is now commercially available. Disc medics have performed well on sandy sites in DLPS trials. A cohort of disc and strand medics has been developed with increased ability to form effective symbiotic relationships with rhizobia strains that occupy Mallee soils. Early field evaluation is promising.

The DLPS project has found that medic pods can be summer sown to establish medic pastures (refer Flohr et. al, this issue), but the strategy is limited by the ability to harvest sufficient pod (Table 4). In subsequent work, where the medic sward was desiccated early to reduce pod drop, yields of 1,000kg/ha were achieved at one site but, remained around 100kg/ha at a second site. Whilst we have shown that it is possible to harvest medic pods, further work is needed to improve its reliability.

Hard-seeded French serradella (cultivars Margurita, Frano)

French serradella is widely grown on acidic sandy soils in WA and on some acidic soils in NSW. Pods are readily harvested (610kg/ha, refer Table 4) and break up into small segments which can be resown. They have an unusual seed softening process whereby light inhibits seed softening. This allows them to be sown at 20—30kg pods/ha in February at ~1cm depth, soften during autumn and establish with opening rain. In SA and Victoria, French serradella grew poorly on alkaline soils but has occasionally performed very well on deep mildly acidic to neutral sands. Serradella flowers later than medics and therefore may benefit from late rains. Conversely seed set may be low in dry springs. Frano, which was released in 2021, flowers earlier than Margurita and has improved early vigour.

Bladder Clover (cultivar Bartolo)

Developed in WA, bladder clover seeds can be harvested where canopy height is adequate and is suitable for February sowing at 20kg hard seed/ha at 0.5 to 1cm depth. This species maintains pasture quality longer, which can decrease the amount of feeding over the dry months. On some neutral pH soils, bladder clover has performed well, particularly in wetter years. At Piangil and Lameroo, pods were unable to be harvested by cereal header due to insufficient canopy height. Bladder clover has high levels of hard-seed in the first autumn and so the establishment year needs to be followed by a crop to allow hard seed to soften for germination in the following year. A new upright hay-type cultivar is being developed in WA.

Rose clover (cultivar SARDI Rose)

Developed for the upper mid-north of South Australia, it has been a middle of the road performer on Mallee soils. Production has been moderate and N-fixation lower. It has been inconsistent in its regeneration after crop. No further cultivar development is warranted.

Biserrula (cultivar Casbah)

Widely grown in WA and NSW, seed can be harvested and is suitable for summer sowing in NSW but not WA.. However, this species has not performed well on Mallee soils in SA or Victoria. It has high levels of hard-seed in the first autumn and so, the establishment year needs to be followed by a crop to allow these to soften for germination in the following year. Biserrula can cause temporary photosensitisation in grazing sheep and affected animals need to be removed

Trigonella (no cultivar currently available)

Trigonella balansae is closely related to annual medic. It is a species of interest because it can hold its pods and about 35% of seed can be harvested with a conventional grain harvester and used to reduce pasture establishment costs. In historic work, APG5045 was identified as having the best agronomic performance, but its hard-seed levels are too low for use as a ley legume pasture and was recommended as a phase pasture option. This hard-seed deficiency was confirmed in this study (18% regeneration after crop, Table 4). Two rounds of selection for increased hard-seed have been completed and the selections have grown well in the field. By end of autumn 2022, hard-seed studies and regeneration counts will be completed, following a 2021 wheat crop. Data will be reviewed to identify a suitable cultivar.

Trigonella is a new species for agriculture and before releasing a cultivar, it needs to pass a grazing study which considers animal performance, animal health and meat quality. This work is led by CSIRO in WA and needs to be completed before a decision is made about cultivar release.

Arrowleaf clover (cultivar Cefalu)

A minor species in low rainfall regions, it has shown promise in NSW and recent SA trials. If late spring rainfall occurs, it can produce green feed late in the season, which is valuable in finishing lambs, and seeds can potentially be harvested with grain harvester. The earliest flowering cultivar is Cefalu. Another line with increased winter and spring dry matter and earlier flowering has been developed. Thinner stems have also been selected, for the benefit of livestock production - modelling in the DLPS project has shown that relatively small differences in nutritive value can provide large benefits to livestock. Field evaluation in 2021 has shown that the new line had about 30% increased dry matter compared to Cefalu throughout the year. Hard-seed studies will be completed late autumn 2022 and a decision made on the suitability of the new line for cultivar release.

Conclusions

Substantial increases in cereal yield and grain protein were measured after medic pasture or vetch, compared to continuous cereal. Despite the costs of establishment, gains in subsequent crop yields alone can provide substantial return on investment, and is even greater when accounting for seedbank set up and grazing value. On neutral/alkaline sandy loam soils in the low rainfall regions, vetch remains the best option where a sown legume ley of one year duration is preferred. Where a self-regenerating pasture is preferred, annual medics provide the best option. Improved cultivars of burr medic (boron, RLEM tolerance), disc medic (improved N-fixation) and arrowleaf clover are being developed.

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 authors would like to thank them for their continued support. Similarly, the authors thank the South Australian Grains Industry Trust. We thank the Pocock, Schmidt, Hayden and Munro families for hosting trials at Lameroo, Waikerie in SA, and at Piangil and Speed in Victoria. The Dryland Legumes Pasture Systems project is supported by funding from the Australian Government Department of Agriculture, Water and Environment as part of its Rural R&D for Profit program, the Grains Research and Development Corporation, 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, NSW Department of Primary Industries and Charles Sturt University, as well as grower groups.

References

Ballard R, Peck D, Flohr B, Llewellyn R, Hill J, Scholz N, Tomney F, Gunn J, Richter I, McBeath T, Davoren B, Shoobridge W, Hackney B, Moodie M, Howie J (2020) New pasture opportunities to boost productivity of mixed farms in low/medium rainfall areas. Proceedings GRDC Grains Research Update, Adelaide, February 2020, pp.41-51.

Farrell M, Vadakattu G, Macdonald L (2021) Addressing the rundown of nitrogen and soil organic carbon. Adelaide Grains Research Update, February 2021

Flohr B, McBeath T, Ouzman J, Davoren B, Shoobridge W, Rebetzke G, Ballard R, Peck D, Llewellyn R, Kirkegaard J, Stummer B (2022) Adaptive sowing strategies to overcome a shifting seasonal break. Adelaide Grains Research Update, February 2022, in-press

Meier EA, Hunt JR, Hochman Z (2021) Evaluation of nitrogen bank, a soil nitrogen management strategy for sustainably closing wheat yield gaps. Field Crops Research 261, 108017.

Smith CJ, Hunt JR, Wang E, Macdonald BCT, Xing H, Denmead OT, Zeglin S, Zhao Z (2019) Using fertiliser to maintain soil inorganic nitrogen can increase dryland wheat yield with little environmental cost. Agriculture Ecosystems and Environment 286, 106644.

Contact details

David Peck
SARDI
PO Box 397, Adelaide SA 5001
08 8429 0475
david.peck@sa.gov.au

Bonnie Flohr
CSIRO
Locked Bag 2, Glen Osmond SA 5064
0475 982 678
bonnie.flohr@csiro.au
@BonnieFlohr