Multispecies forages in the Australian dairy feedbase: is there a biological business case?

Introduction

Globally, it is estimated that 1 billion hectares, 7% of the total land area of Earth, is devoted to dairy production (FAO 2016). Most of this land area is grassland, used either for grazing or as a source of conserved forage. A need to intensify milk production to meet global demand, combined with plentiful and inexpensive access to inorganic nitrogen (N) fertiliser, has resulted in a preference to sow dairy grassland with monoculture species that display a linear yield response to inorganic N fertiliser application (Oenema et al. 2014). The species of choice in most temperate Australian dairy systems is perennial ryegrass (Lolium perenne; PRG), a species that has benefitted from significant genetic improvement in targeted plant breeding programs to optimise its productivity and nutritive value for grazing livestock (Lee et al. 2014). However, in recent years a combination of fluctuating and extreme prices of inorganic N fertilisers (Humphreys et al. 2012), combined with restrictions on N fertiliser rates per hectare per annum in certain countries (e.g. Water Framework Directive in the EU), and societal pressure to reduce environmental impacts (Martínez-Dalmau et al. 2021), have combined to create a driver for farmers to consider the reintroduction of functional diversity to their pastures. The most common example of this is the addition of clover to PRG pastures creating a binary mixture (Humphreys et al. 2017). In the late 1990s, research suggested that further increases in diversity were beneficial (Hector et al. 1999), leading to the concept of targeted multispecies swards becoming a topic of scientific interest in the 2000s until the present day. Hector et al. (1999) conducted a foundational study involving eight field sites across Europe that showed an overall correlation between an increase of functional groups and increased pasture productivity in the absence of confounding effects of N fertilisation. Further research into multispecies swards grew in more recent years from this finding that diversity can increase yield, partly or entirely replacing the need for N fertilisation, thus creating a cheaper and more self-reliant way of intensifying grassland productivity (Döring et al. 2013; Finn et al. 2013). Such swards, in the modern context, are defined in this review as temporary pastures (4–6 years) designed to incorporate several species from multiple functional groups in harmony, such that different species fill different ecological niches resulting in complementarity effects. The functional groups most often represented are perennial grasses, legumes, and herbs (forbs), comprising species that have nutritional value for feeding livestock; however, brassicas, cereals, or annual pasture species may also be featured. Multispecies swards have also been known by other names in the literature, including ‘mixed swards’, ‘herbal leys’, or ‘diverse forages’. The purported benefits of modern multispecies swards are threefold, namely, productive (Jaramillo et al. 2021), resilient (Lüscher et al. 2022), and environmentally beneficial (Huyghe et al. 2012). Productivity as a defining feature is key because it excludes sward types from this review that, although diverse, are not sown with the aim of producing high yields, such as, for example, meadowlands, permanent pasture, and agri-environmental mixtures comprising wildflowers.

This review identifies both opportunities and challenges related to the incorporation of multispecies swards into temperate Australian dairy systems and builds on historical reviews of this topic (Pembleton et al. 2015) by incorporating more recent international studies. Dairy farming in Australia is largely concentrated in southern temperate regions where cattle graze all year round due to relatively mild winters. Systems are divided between irrigated and rain-fed, or can be a mixture of both, depending on the available water sources in each dairying region. A current lack of Australian government policy controlling inorganic N application rates has resulted in monoculture PRG remaining the most common pasture sward within temperate dairy farming systems, unlike many other similarly temperate dairying nations that routinely combine legumes and grasses (e.g. New Zealand, Ireland). A 2019 national industry survey (Dairy Australia 2019) reported that 67% of respondents had sown PRG, and 65% had sown an annual or short-term ryegrass (e.g. Italian ryegrass), in the past 5 years. Only 12% reported sowing clover and just 5% had sown a mixed sward in the past 5 years, indicating little use of binary mixtures with clover, or more diverse swards in general. However, the same economic pressures (e.g. N fertiliser price volatility and water scarcity in drought years) that have driven interest in pasture diversity overseas, combined with the possibility of greater restrictions around N applications in Australia in the future (Rawnsley et al. 2019), are now fuelling rising interest in multispecies pastures among Australian dairy producers and thus a review is timely to re-assess their potential.

Soil–plant interactions under multispecies swards

One of the first considerations influencing the uptake of multispecies swards in southern Australian dairying regions is whether the soils found in these regions can support multispecies sward growth, since they are typically either moderately or very acidic, such as in the State of Victoria where over 50% of all Australian milk is produced (Dairy Australia 2019; Fig. 1). A recent UK study contrasted multispecies sward productivity on two soil types, pH 5.4 and 7.4 (measured in water), where soil was removed from the paddock and placed in pots for a growth study (Darch et al. 2022). Darch et al. (2022) observed that a mixture of 15 perennial species (six grasses, five legumes and four herbs) yielded both greater total dry matter (DM) per pot and greater DM per functional group per pot in all three functional groups using the pH 5.4 soil than the pH 7.4 soil. Further examination of soil properties by Darch et al. (2022) showed a greater concentration of soil organic carbon, and soil N (as total oxidisable N plus ammonium N) in the pH 5.4 soil, which may have contributed to this outcome. The relative success of the more acidic soil type in the study of Darch et al. (2022) is promising for the application of multispecies swards in Australia. The functional group most likely to suffer in acidic soils is legumes. Determining the pre- and post-establishment fertiliser and liming requirements for a multispecies sward can be more complicated than with a monoculture where the needs of only one plant needs to be accounted for (Bolland and Russell 2010). Hill et al. (2005) demonstrated that phosphorus (P) requirement can differ significantly among common Australian pasture species, including legumes and grasses, and that a deficiency in P provision can alter a sward botanical composition to favour plants better adapted to nutrient stress. Furthermore, as plant communities become more complex, a greater range of micronutrients should also be considered in addition to macronutrients. For example, the rhizobia-driven N-fixing ability of certain legumes has been shown to be inhibited in boron-deficient soils (Hamilton et al. 2015). Another study conducted in southern Australian pastures also identified copper, molybdenum and zinc as being important to legume productivity (Brennan et al. 2019).

Fig. 1.

A map of Victoria, Australia, showing the pH of topsoil (0–10 cm) in areas identified to have pasture as the land use, according to the ‘VLUIS’ land registry database (source: https://vro.agriculture.vic.gov.au/dpi/vro/vrosite.nsf/pages/vrohome).

To achieve the purported efficiency benefits of multispecies swards, legumes must be utilised to fix atmospheric N in the soil, resulting in a reduced need for expensive inorganic N fertiliser application (Suter et al. 2015). Fixed N can be transferred from the legumes to the non-legumes in the mixture through various pathways, including decomposition of legume root tissue, root exudates comprising soluble N compounds, and via mycorrhizal systems (Thilakarathna et al. 2016). This transfer of N from legumes to grasses and herbs within multispecies swards is an example of niche complementarity. Previous studies have measured N derived from symbiotic sources in legume-containing mixtures in excess of 200 kg N/ha.annum, a rate that completely negates the need for additional inorganic N application (Nyfeler et al. 2011). However, there is a risk that practitioners will continue to apply inappropriately high inorganic N fertiliser rates to multispecies swards with high concentrations of fixed N in the soil, which would lead to N loss to the environment as leached nitrate and as gaseous compounds, namely, nitrous oxide, ammonia, dinitrogen, and nitric oxide (Bracken et al. 2020; Bracken et al. 2022). Nitrous oxide is a potent greenhouse gas (GHG) and therefore any environmental benefits of multispecies swards could be negated should they lead to increased GHG emissions. In an Irish study, Bracken et al. (2020) first demonstrated that a mixture comprising 75% grass and 25% legume resulted in a cumulative release of ~50 g N₂O-N/ha in the 2 months post-fertilisation with 40 kg N/ha, whereas a mixture comprising 25% grass and 75% legume resulted in ~150 g N₂O-N/ha released when both swards were fertilised at the same rate. However, in a follow-up study by Bracken et al. (2022), where the study was repeated but with reducing fertiliser rates matched to sward legume content, PRG monocultures fertilised at 250 kg N/ha.annum produced a 71% increase in N₂O emissions relative to a legume-heavy mixture receiving no N fertiliser. Since Bracken et al. (2022) reported no significant difference in the herbage yield achieved from these two treatments, the mixture resulted in the least GHG emission for the same biomass yield. A different study by Cummins et al. (2021) agreed with this finding and further demonstrated that N₂O reduction through multispecies sward use was greater when the functional group proportions of the mixture were even and lesser in legume-dominant mixtures. Bracken et al. (2020, 2022) also found that the herb ribwort plantain (Plantago lanceolota) seemed to partially inhibit nitrification processes in soil, with small beneficial effects on N losses. However, the finding was not consistent with other studies (Podolyan et al. 2020). Further research is warranted to assist producers in correctly matching N fertilisation levels to multispecies sward compositions where legume concentrations can vary considerably from season to season and year to year.

Another purported benefit of introducing diverse communities of species is that key soil characteristics (other than plant available N) can be improved to provide positive long-term impacts on soil function. Such characteristics include soil structure, water-holding capacity, and soil carbon (C) concentration. A German study (the Jena Experiment) that aimed to reintroduce permanent diverse grasslands into land that was previously used for arable cropping showed a strong correlation between species richness and C sequestered after 9 years of the experiment (Lange et al. 2015). However, conflicting evidence has arisen on whether temporary, intensive multispecies leys can have the same beneficial soil effects. A UK study investigating three mixtures of increasing species richness compared with a monoculture PRG control showed no discernible evidence for increased soil C after 2 years under mixed swards, in comparison to the control sward (Shepperd et al. 2020). This was consistent with a US study by Skinner et al. (2006) that measured change in soil C under three different mixtures of varying diversity over 3 years and found no change or even decreases in soil C. However, in a longer-term study, conducted by the same US researchers, in which soil changes were evaluated after 9 years under a two- or five-species mixture, the more diverse mixture was shown to have more than three times the soil C accumulation rate of the binary mixture (1.80 vs 0.50 Mg C/ha.annum; Skinner and Dell 2016). Results pointed to functional diversity needing to be maintained for many years before a significant soil C concentration change can be observed. This is also likely to be associated with other factors such as previous paddock history, soil type, grazing practices, and water availability. Equally important is the assessment of C fractions, given the different contributions to soil functionality and C retention. There is evidence to show that multispecies swards result in greater root mass and rooting depth than do grass species alone (McNally et al. 2015; Eisenhauer et al. 2017), particularly compared with PRG due to its particularly shallow rooting nature. Deeper rooting structures coupled with a faster turnover of root material can explain the mechanism by which multispecies swards may increase soil organic C stocks over time. McNally et al. (2015) demonstrated that a moderately diverse pasture not only had deeper rooting system than a two-species mixture, but also had twice the annual root turnover (5411 vs 2672 kg/ha). Deeper rooting systems are also associated with improved soil structure and ability to adapt to and withstand drought (Hoekstra et al. 2015). With future climate conditions predicted to result in more extreme weather events, the ability to withstand drought is an especially important pasture characteristic for Australian dairy farms.

Many of the interactions discussed above are governed by soil microbial populations, including the degree of N fixation achieved (Unkovich 2012), carbon sequestration (Lange et al. 2015), and the amount of denitrification that occurs in the soil (des Roseaux et al. 2022). However, it has only been in recent years that analytical capability to describe microbial populations, particularly in the rhizosphere, has evolved and findings on factors that affect the soil microbiome have started to emerge. Gaining a greater mechanistic understanding of this area in relation to multispecies swards is a promising topic for further research and may help explain conflicting results on soil C sequestration under multispecies swards.

Species selection and seed mixture design

The agronomy of multispecies swards begins in the establishment phase when practitioners must initially decide on a seed mixture for the new sward. Even at this early stage, already a complex array of factors must be considered, including the total number of species (species richness), the specific species to be included, the proportional seed rate of each species within the mixture (determined by the desired botanical composition for the new sward), and finally, the seed rate per hectare at which to sow the seed mixture. A further complication is dealing with any weeds that arise, especially in Australian soil where the seedbank of weeds can be prolific, and with limited chemical weed control options due to the diversity of the sown species. Multiple authors have remarked that species richness alone does not guarantee productivity in multispecies swards, rather, objective species choice from the start is crucial for eventual success (Huyghe et al. 2012; Chapagain et al. 2020; Jordon et al. 2022). Notably, the number of studies focussing on how exactly to achieve the optimum seed mixture is limited. There are a greater number of studies that have compared effects of species richness on productivity (Hector et al. 1999; Grace et al. 2019a; Barker et al. 2021; Komainda et al. 2022) than the number of studies that have focussed on species choice (Ozcan et al. 2015) or functional group proportion (Finn et al. 2013; Grace et al. 2018a). Still fewer studies report differential effects of establishment methods for multispecies swards (e.g. seed rate at sowing, soil preparation or sowing method).

Fig. 2 shows a summary of the species included within seven previous studies. These previous studies all concerned the effect of species choice on subsequent botanical composition and sward performance. The seven studies were all conducted in Ireland, the UK, or Europe; however, many of the species utilised can be commonly purchased as commercial seed worldwide, including in Australia (except for certain herb species). The most popular species within key functional groups were PRG, cocksfoot (Dactylis glomerata) and timothy (Phleum pratense) for the grasses; white clover (Trifolium repens), red clover (Trifolium pratense) and birdsfoot trefoil (Lotus corniculatus) for the legumes; and ribwort plantain (Plantago lanceolota), yarrow (Achillea millefolium), chicory (Cichorium intybus), and burnet (Sanguisorba minor) for the herbs. A commonality among the studies in Fig. 2 is that they, mostly, use species that have undergone some cultivar testing and development as opposed to the use of native species when designing mixtures. Aside from PRG and certain clovers that have been extensively selectively bred, many species in Fig. 2 still represent untapped genetic potential that could be further exploited to increase their survivability and productivity within specified multispecies swards. Fig. 2 also highlights a gap in European research around the use of annual functional groups such as brassicas and cereals that can be sown together with grasses, legumes, or herbs, or even oversown into existing perennial swards annually to make short-term multispecies mixtures. Given the harsh summer conditions in parts of Australia, the use of annuals to formulate multispecies swards may be an attractive local option, providing that farmers are happy to reseed large areas of their farms each year.

Fig. 2.

A table of seven studies concerning multispecies swards and the species chosen to form treatment mixtures in those studies (left) and a graphical summary of how often each species featured in treatment mixtures across those studies (right).

Regarding species richness, several studies agree that greater diversity increases potential complementarity effects (Jordon et al. 2022). Early studies focussed on proof-of-concept that moderate levels of diversity promoted greater productivity than monocultures. The terms ‘overyielding’ or ‘transgressive overyielding’ refer to a mixture either outperforming the weighted average of its constituent species sown in monoculture or outperforming the most productive constituent species sown in monoculture respectively. In the 3-year study of Finn et al. (2013), in which the same plot study design comparing four-species mixtures with accompanying monoculture plots of their constituent species was replicated at 31 sites across Europe, 97% of mixtures showed overyielding and 60% showed transgressive overyielding. Further to this, there are also numerous examples in the literature of multispecies swards receiving low (<100 kg N/ha.year) or no N fertiliser equalling or exceeding the yields of monoculture PRG swards receiving high (>200 kg N/ha.year) rates of applied N (Collins et al. 2014; Barker et al. 2021; Baker et al. 2023). More recent studies have sought to ascertain the optimum level of diversity, considering mixtures with greater richness of species; for example, Barker et al. (2021) observed superior yield from 12- to 17-species mixtures than from six-species mixtures.

Finn et al. (2013) also reported a correlation between a greater magnitude of overyielding effects and greater evenness in the functional group proportions within the mixtures. However, it has been demonstrated that the species proportions within mixtures vary dynamically over time (Barker et al. 2021), likely due to environmental conditions, varying establishment and persistence profiles of the different species, and cyclic N fluxes linked to legume growth patterns. While certain management interventions could be implemented to influence species proportions (for example, the withholding of inorganic N fertiliser to favour legumes over grasses, or oversowing seed of a certain missing species), it is not clear whether taking such interventions are worthwhile as opposed to working with the sward that has naturally developed over time.

Management of multispecies swards as a feed source

Most temperate Australian dairy farming systems are characterised by all-year-round grazing, with reliance on forage conservation, as silage, to manage pasture accumulation excess to animal requirements. This silage can be offered during the summer months to ensure supply of forage to the animals in adequate quantity and quality during challenging drought periods. However, studies that have compared the effects of grazing or cutting on multispecies sward botanical composition and productivity often note negative outcomes associated with grazing (Collins et al. 2014; Grace et al. 2019b). Grace et al. (2019b) observed an average yield reduction of 1.9 t DM/ha.annum in grazed multispecies swards versus cut multispecies swards, while there was no significant difference attributed to defoliation method in a PRG control. Collins et al. (2014) also observed a yield penalty in the region of 4 t DM/ha.annum associated with grazing versus cutting of a multispecies sward at one experimental location in Wales, but not at two others in Europe, and concluded that the response to grazing depended on sward composition as well as the grazing animal species (ewes at the Wales site, and dairy or beef at the European sites). A contrasting result was obtained in a recent German study, where the average 2-year productivity of an eight-species mixture was 2.7 t DM/ha.annum greater under a grazing regime than a cutting regime (Loges et al. 2020). Success under grazing is likely to depend on a combination of initial species choice, management (e.g. applied fertiliser/natural fertilisation through defecation), and grazing practices (e.g. defoliation frequency) determining whether additional species remain productive and persistent over time (Pembleton et al. 2015). This is especially the case for legumes in species-diverse swards that have a greater requirement for phosphorus and sulfur fertilisers than N and may require adapted grazing strategies suited to the legume species and cultivar and the grazing ruminant species to maintain the legume component in the sward (Frame and Newbould 1986; Black et al. 2009). Herb species, excluding chicory and plantain, may also be fragile and poorly suited to persistence under grazing due to effects of trampling or hard grazing. Since continuation of diversity-driven overyielding effects over a period of several years requires that diversity is maintained over time (Nyfeler et al. 2009), a loss of diversity under grazing may be the cause of the poorer performance of grazed multispecies swards in the afore-mentioned studies, where grazed swards yielded less than cut multispecies swards. In a grazing study, Sanderson et al. (2005) highlighted that continual re-establishment of the chicory and legume components in the pasture was necessary to retain the yield benefits of their complex mixtures. Australian graziers should be aware that multispecies swards may suffer some yield depression when grazed, as opposed to cut, if species loss occurs due to trampling or hard-grazing and further work into species choice for robustness and persistence under grazing is warranted to minimise this effect. There is also a need for research works that continue to monitor pastures over a longer period of up to 6 years, since few studies continue to observe swards, particularly under grazing, until the end of their productive lifetimes to measure persistence.

One purported advantage of multispecies swards in a grazing rotation is their year-round productivity and ability to produce biomass during challenging ‘shoulder-seasons’ (summer and winter) when PRG growth rates tend to be minimal due to unfavourable climatic conditions. In a grazing study in USA, Sanderson et al. (2005) concluded that planting a mixture of grasses, legumes, and chicory benefits herbage production during dry years and reduces weed invasion for a few years after planting. However, Sanderson et al. (2005) also observed variation within the response of different mixtures to dry conditions as one mixture produced less average herbage yield for the grazing seasons than did other mixtures (4800 vs 7600 kg DM/ha). When rain was plentiful there were no differences in herbage yield among the mixtures, all averaging 9800 kg DM/ha. This agrees with a 3-year UK study by Barker et al. (2021), in which it was reported that multispecies swards receiving no N fertiliser yielded less than N-fertilised perennial rye grass (PRG) in the first year when rain was plentiful, had equal yield to PRG in the second year when there was summer drought, and exceeded the yield of PRG in year 3 when there was a second summer drought. Furthermore, in the USA, Skinner et al. (2004) found that complex 5-species mixtures yielded more herbage than a simple 2-species mixture under dry conditions, highlighting that the vigorous growth of chicory in the complex mixture accounted for most of the yield increase, likely due to its deep-rooted growth habit allowing access to moisture from lower in the soil profile. These studies underlined the possible drought tolerance of multispecies swards and their potential to aid farmers to continue grazing without needing to offer additional conserved forage or purchased supplements through ‘shoulder-seasons’, and particularly in summer drought periods.

A key aspect of any grazed pasture is its nutritive characteristics in terms of macronutrients such as crude and metabolisable protein, digestibility and metabolisable energy, fibrous components, and sugars. The concentration profile of macronutrients in the forage has direct effects on eventual milk yield and milk solid concentration in the grazing dairy cow as well as pasture palatability, which further affects forage intake, utilisation and selective grazing behaviours. The selective breeding of PRG has led to the development of cultivars that are highly digestible, with high crude protein concentrations, providing grazing occurs at an ideal stage of maturity, making it an ideal nutrient-dense feed. In contrast, when PRG plants encounter environmental challenges such as prolonged high temperature and drought, typical of the Australian summer, their nutritive value has been demonstrated to decline (Rogers et al. 2022). However, it should be noted that there is a lack of sufficient studies that examine the effect of challenging climatic conditions on this contrast. Studies have reported nutritive values for multispecies swards in comparison to either PRG alone (Grace et al. 2018b; Barker et al. 2021) or in simple combinations with clover (Jing et al. 2017; Loges et al. 2020; Hearn et al. 2022). Loges et al. (2020) found that an eight-species mixture had, on average, marginally lower metabolisable energy (11.0 vs 11.1 MJ/kg DM) and crude protein concentration (21.3 vs 22.5% DM) pre-grazing than did a PRG and white clover combination over 2 years. However, these differences, while significant, were not considered large enough to affect the productivity of grazing animals and both swards were considered to produce high-quality forage. Loges et al. (2020) also observed that, when grazed at an optimal maturity stage, cattle preferred the more diverse multispecies sward leading to greater forage utilisation (pre-grazing minus post-grazing mass) of 9.03 t DM/ha.annum utilised for the diverse treatments versus 8.41 t DM/ha.annum utilised for the ryegrass–clover treatment. Similarly, Jing et al. (2017) noted a small reduction in both in vitro organic-matter digestibility and crude protein concentration of cut herbage in a 12-species mixture, compared with a three-species mixture of grass and two clovers; however, the reduction was small and inconsistent over cuts and years. Hearn et al. (2022) furthered this work by considering seasonal variation of nutritive value in both binary and complex mixtures (up to five species) and found there to be significant effects of both timepoint and pasture type but no interaction between these two for organic-matter digestibility. This lack of interaction was because all pasture types were found to have lower digestibility in summer than during the rest of the year. Furthermore, Grace et al. (2018b) compared the nutritive characteristics of a PRG monoculture and a six- and nine-species mixtures under grazing in spring, summer and autumn, and found no significant differences apart from a significant increase in crude protein concentration in the six-species mixture relative to the PRG monoculture in spring. From these studies, the evidence suggests that multispecies swards are likely to be either equally, or slightly less, nutritious than the conventional monoculture grasslands typically found currently in temperate Australian dairy farms. One explanation for reductions in nutritive quality versus PRG pasture can include certain species becoming mature and unpalatable, such as, for example, if chicory enters the reproductive stage. This scenario can also lead to undesirable selective grazing behaviours and was noted by UK farmers as a key drawback of multispecies swards when tried to implement in practice (Jordon et al. 2023). However, if DM yield increases are seen under multispecies swards, then it is possible that multispecies swards could still result in greater yield of nutrients per hectare despite having similar or lower nutrient concentration profiles than does traditional pasture. Indeed, evidence from a modelling study conducted in New Zealand has suggested that the nutrient profile of diverse pastures might better match N and energy requirements of cattle than that of traditional pasture types, leading to less excess N wastage in urine (Beukes et al. 2014). This effect combined with a lower DM concentration of diverse swards promoting dilute urine was suggested by Beukes et al. (2014) to lead to N leaching reductions of 11–19%, depending on the proportion of farm sown to diverse swards in the modelled scenarios. However, further work to definitively show how this effect translates to actual farming conditions is needed.

Production, health, and environmental footprint of grazing dairy cattle

Different research groups have demonstrated that diversifying the grazing sward from grass alone, or in combination with a legume, by inclusion of one or more herb species can modulate milk production and composition, but results from the literature are not consistent (Chapman et al. 2008; Bryant et al. 2017; Dodd et al. 2019). Two recent meta-analyses, including studies conducted predominantly in New Zealand and Australia, have reported that inclusion of chicory and/or ribwort plantain (McCarthy et al. 2020) or plantain alone (Nguyen et al. 2022) to grass or grass plus legume swards increased yields of milk and milk protein (1.20 and 1.02 kg/day, and 0.05 and 0.02 kg/day respectively), with milk fat yield being increased in the analysis including multiple herbs (0.04 kg/day) but not with plantain alone, compared with cows grazing grass or grass plus legume swards. It is important to note that production results varied depending on the stage of lactation of the cow. This is likely to be associated with differences in the stage of growth of different species (Grace et al. 2018b) and nutritive characteristics of the sward in different seasons, which can result in differences in DM digestibility or DM intake (DMI) of more diverse swards compared with less diverse swards (McCarthy et al. 2020; Nguyen et al. 2022). Future research efforts should consider interactions among the stage of lactation of the cow, seasonal nutritive characteristics of the sward, and type and amount of supplementary feed as potential drivers of cow production.

Sward nutritional characteristics and composition are important factors modulating the ruminant gastrointestinal environment, which in turn influences cow production and health. Studies with small ruminants have reported anthelmintic effects of more diverse pastures compared with less diverse pastures (Marley et al. 2003; Grace et al. 2019a). While the mechanisms involved in decreasing the gastrointestinal parasite load of animals grazing more diverse pastures is not fully elucidated, it is hypothesised that plant secondary metabolites may play a role (Min and Hart 2003). However, there is a paucity of data on the potential of diverse swards in reducing parasite load in dairy cows. It is also suggested that diversifying the grazing sward could help mitigate some of the nutritional imbalances or antinutritional factors that can occur in cows grazing monoculture pastures. Enriquez-Hidalgo et al. (2014) reported that cows grazing PRG plus white clover, as opposed to PRG alone in autumn increased mean ruminal pH (6.16 vs 6.00), and decreased time below ruminal pH of 6 (357 vs 704 min/22 h) and concentration of D-lactic acid (4.7 vs 6.3 mmol/L), suggesting a reduced risk for ruminal acidosis in cows grazing a PRG plus white clover; however, another study reported that cows offered fresh-cut forb species (plantain or chicory) in the afternoon and PRG plus white clover both morning and afternoon had a lower ruminal pH in the hours post-forb feeding than did cows offered only PRG plus white clover at each feeding, despite lower daily DMI by cows in the forb treatments (Minneé et al. 2017). Additional studies have reported that ingestive and ruminating behaviour, DM digestibility and ruminal microbial communities are dependent on sward type (Gregorini et al. 2013; McCarthy et al. 2020; Smith et al. 2020). Therefore, the combination of species included in the sward requires further consideration before nutritional recommendations can be made to mitigate animal health risks associated with digestive disturbances.

An area that has gained attention in recent years is the potential for diversifying the grazing sward to mitigate the environmental footprint of dairy farming. The N-fixing properties of legume forage species and their role in reducing N₂O emissions as well as the need for inorganic N application is well understood and has been discussed earlier in this review (Jensen et al. 2012; Cummins et al. 2021). In addition, inclusion of forb species and legumes in the sward can contribute to improvements in N-use efficiency and lowering N leaching to the environment relative to PRG–clover pastures (Beukes et al. 2014; Rawnsley et al. 2019). For example, Woods et al. (2018) reported that after application of urine from cows grazing an Italian ryegrass–plantain–white clover sward or a PRG–white clover back to each respective sward, urine N leaching was reduced by 88.9% in the more diverse sward compared with the least diverse sward. Other studies have reported on the ruminal methane-mitigation potential of multispecies swards for dairy cows associated with plant secondary metabolites (e.g. tannins; Loza et al. 2021a) and nutritive characteristics of the sward (Jonker et al. 2019; Loza et al. 2021b), but results have been inconclusive. However, the relative contribution of multispecies swards to mitigating undesirable environmental outcomes is likely to be determined by the type and density of species in the sward and environmental conditions (e.g. soil type, temperature, rain patterns, etc.). Further, evaluations at a system level are required to account for soil, plant and animal interactions across seasons that ultimately determine the productivity, environmental footprint, and profitability of dairy farms (Unkovich 2012; Bryant et al. 2020).

Conclusions

The present review has sought to identify research gaps, challenges and opportunities associated with multispecies swards and their potential to enhance the Australian dairy feedbase. International interest in this topic has intensified in the past 5 years, with several new studies being published covering a variety of aspects of multispecies swards from soil to plant to animal. These studies have highlighted the potential benefits of multispecies swards such as equal or even increased productivity in low N-input multispecies swards compared against high N-input PRG monocultures. The literature also suggests that multispecies swards can be persistent through summer drought, resilient to grazing if managed appropriately, and a source of high-quality feed for ruminants. Through the ability to reduce inorganic-N inputs, multispecies swards have been shown to reduce nitrous oxide emissions and increase N-use efficiency of systems, while also having positive effects on soil fertility in the longer term (~10 years). Although only a few studies have specifically tested milk productivity and animal-health effects of grazing ruminants consuming multispecies swards, initial results are encouraging. However, studies have often reported that mixtures did not exhibit these potential benefits consistently, perhaps due to different seed-mixture designs, eventual botanical composition, environment effects, or interactions with management systems. Identified challenges associated with multispecies swards centred around the complexity of options they offer. From seed-mixture design, the eventual botanical composition, functional group choice and proportion, species richness, and ongoing management, no two multispecies swards will ever be identical. There remains uncertainty around what is the best way to approach the above-mentioned aspects of multispecies sward implementation for best effect and it can be expected that there will not be a ‘one size fits all’ solution. There were examples in the literature of studies where multispecies swards required early renovation to rectify species loss, were less nutritious than a PRG control, or did not yield as well as expected under grazing, showing that greater understanding of the mechanisms that drive successful multispecies swards is needed. The potential benefits and identified challenges are summarised in Fig. 3.

Fig. 3.

A summary of potential benefits and identified challenges concerning the adoption of multispecies swards in farming systems.

Topics where further Australian-specific research would be beneficial include (1) a greater understanding on how to design the species mixture according to the target use-case and environment, (2) longer-term studies (5+ years) to understand effects on soil function, persistency of species and the sward as a whole, with a focus on rotational grazing scenarios, (3) optimal sward management and how this interacts with changes to botanical composition of the sward to guarantee that a multispecies sward reaches its full production potential, (4) a greater focus on animal studies to understand the effect of consuming different multispecies swards on production and health responses of lactating cows as well as a feed source to support dairy youngstock, and dry dairy cows, and (5) the potential of multispecies swards to contribute to mitigation of methane and N₂O emissions. Due to an overwhelming number of possible multispecies sward designs, advanced modelling methods may be useful to test a greater number of options than can subsequently be measured in physical studies, and to extrapolate the impact of findings. In conclusion, although there is still much to understand about the mechanisms governing what makes a multispecies sward successful, the mounting evidence that diverse pastures can be environmentally positive, resilient to future climate scenarios, and adequately productive and nutritious, compared against intensive monoculture grasslands, suggests that they will have a positive impact if adopted by Australian dairy farmers. The greatest challenge to their adoption will likely be in providing optimal advice to producers, as they depart from traditional methods, to help them navigate the potential complexities of this alternative approach.

Data availability

The data that support this study will be shared upon reasonable request to the corresponding author.

Conflicts of interest

The authors declare no conflicts of interest.

Declaration of funding

This research did not receive any specific funding.