1 MANAGING NITROGEN LOSS FROM GRAZED PASTURE

Intensive agricultural land use can negatively impact freshwater quality in its receiving environment (e.g., Cartwright et al., 1991). Inorganic nitrogen (N) is a potentially transformational element in receiving aquatic, marine and atmospheric environments, and therefore a vast body of research has focussed on reducing N emissions to these ecosystems, typically via leaching and gaseous transport processes. Within in-situ grazed forage systems, early research identified the ruminant urine patch, an area with a high N load (>70 g N m⁻²), as the predominant source of N leaching (Ball & Ryden, 1984; Di & Cameron, 2002). Up to 90% of the N originating from urine patches can be leached to groundwater (Ledgard et al., 2009). At present, interest in the N load of urine patches is also increasing in the search for options to mitigate emission of nitrous oxide, a potent greenhouse gas (De Klein et al., 2001).

One way to reduce N leaching from grazed forage systems is to limit the N inputs from fertilizer and supplementary feeds, reducing the overall amount of urine N returned to the soil. Several dairy farmlet-scale comparisons in the Waikato, New Zealand, have shown that lower N inputs decreased N leaching (Ledgard et al., 1997; Ledgard et al., 2006; Shepherd et al., 2017; Table 1). However, milk production ha⁻¹ was also lower and, depending on prices of milk and inputs, profitability was also reduced due to the relatively low cost of N fertilizer and high dry matter (DM) growth response rate (extra kg DM per kg N applied) achieved in this region (Clark et al., 2019).

Studies in drystock systems have been fewer, partly because the lower levels of N fertilizer and supplementary feed typically used in these systems limit the ability to manipulate these inputs. Leaching losses from sheep-grazed hill country systems were shown to be linearly related to N fertilizer inputs (Hoogendoorn et al., 2017) and other research has demonstrated lower leaching associated with sheep and deer relative to cattle (Hoogendoorn et al., 2011; Maheswaran et al., 2023).

Applied research to reduce environmental impacts in New Zealand dairy systems turned its focus to selecting forages that impact the N cycle of grazed dairy pastures without reducing production and profitability (Chapman et al., 2014). This approach has the further advantage of minimizing the need for substantive system change and associated capital investment. New Zealand lowland pastures are predominantly perennial ryegrass (Lolium perenne L.; PR) and white clover (Trifolium repens L.; WC), a high-quality forage mixture shown to achieve pasture and milk production at moderate N fertilizer rates similar to perennial ryegrass monocultures with high N fertilizer rates (Collins et al., 2014). However, quality comes with a challenge – mixtures of perennial ryegrass and white clover often have a higher N content than required for cow maintenance and milk production, resulting in a large proportion of N intake being excreted in urine (Pacheco & Waghorn, 2008).

Studies in dairy systems during the period 2005–2012 found that more diverse perennial pastures containing ryegrass, white clover and the herbs chicory (Cichorium intybus L.) and narrow-leaved plantain (Plantago lanceolata L.; plantain or PL hereafter) could maintain milk yield while reducing the urine N concentration and urinary N excretion (Totty et al., 2013; Woodward et al., 2012). A modelling exercise indicated that these differences could substantially reduce N leaching, e.g., by 19% if half the farm area consisted of this grass-clover-herb mixture (Beukes et al., 2014). It appeared that a small number of species of different functional groups could be a promising solution to manage N in grazed pastures (Pembleton et al., 2015; Vibart et al., 2016).

Plantain was suggested as a key species to achieve N leaching reductions via several of these pathways (Judson et al., 2018a): specifically reduced urine N concentration; lower partitioning of excreted N to urine; a nitrification inhibiting effect from urine from animals fed PL; and a nitrification inhibiting effect from plant material and root exudates. These aspects of PL are explored in detail in this paper.

2 THE USE OF PLANTAIN IN NEW ZEALAND GRASSLANDS

Plantain is considered a short-lived perennial plant, persisting for 2 to 4 years under common grazing practices (Stewart & Judson, 2019). It was first commercialized in New Zealand in the late 1990s with the release of two cultivars, ‘Grasslands Lancelot’ and ‘Ceres Tonic’ (Rumball et al., 1997; Stewart, 1996). Naturalized PL is largely a winter-dormant, prostrate weed colonizing poor and lower fertility soils throughout the world (Cavers et al., 1980). These two cultivars aimed to target differing grazing contexts based on their growth habit; the more prostrate Lancelot was aimed at close-grazed sheep systems (Rumball et al., 1997), while the more erect Ceres Tonic was more suited to cattle systems.

Early research was mainly with short-term pure crops (Fraser & Rowarth, 1996) and the results of lamb growth trials with cv. Lancelot, a winter-dormant cultivar, created a stigma of a poor-quality forage that persevered in the industry. Early commercial use of PL was as a minor component, sown at low rates (1–2 kg ha⁻¹) in diverse pasture mixtures, with a seed market volume that grew from c. 20 to 80 t year⁻¹ nationally between 1996 and 2005. During the period 2005–2010, PL was researched and promoted as a major component in “herb mixtures” that included clovers and sometimes chicory (Cranston et al., 2015). The effectiveness of this for finishing young stock (Moorhead et al., 2002) and increasing ewe liveweights (Judson et al., 2009) led to an increase in seed sales of the Ceres Tonic cultivar and an overall increase in market volume to c. 110 t year⁻¹. From 2010 to 2015 pure crop use grew rapidly, to see market volume peak at over 300 t year⁻¹. However, poor sector understanding of its niche and pest issues in warmer climates led to a market retrenchment to c. 200 t year⁻¹ by 2015. The most recent development is the inclusion of PL in perennial ryegrass-based swards in the dairy sector (Edwards et al., 2015), as a result of two major national research programmes: “Forages for Reduced Nitrate Leaching” (https://www.dairynz.co.nz) and “Greener Pastures” (https://www.agricom.co.nz). Since 2015, other proprietary cultivars have entered the market with other proprietary cultivars, such that market volume is currently c. 250 t year⁻¹. These cultivars range from late-heading, winter-hardy types to upright, winter-active forage plants which are significantly more productive than wild types.

The cultivars in use establish rapidly and are adapted to a wide range of soil types and climatic conditions. They are tolerant of a wide range of insect pests and diseases, although the PL moth (Scopula rubraria and Epyaxa rosearia) and the red-legged earth mite (Halotydeus destructor) can devastate pastures.

Plantain is highly palatable to grazing animals and capable of supporting excellent animal performance across many farming systems. Compared with PR, PL has been shown to have a lower DM content, lower structural fibre and higher non-structural fibre contents, similar crude protein content but less soluble and degradable N, and a similar digestibility (Minnée et al., 2019). It also has higher contents of trace elements than perennial ryegrass, when sufficiently available in the soil (Moorhead et al., 2002). In non-irrigated systems, compared with PR-WC pastures, the benefits of PL pastures during lactation for ewes include improved intake, reduced faecal parasite egg output (Judson et al., 2009), increased milk production (Kenyon et al., 2010), and better lamb liveweight gains (Hutton et al., 2011;Kemp et al., 2013; Moorhead et al., 2002).

These properties of PL, as well as the prevalence of various secondary compounds, are causing the differences in N cycling in grazed pastures described in the next sections, compared with New Zealand's standard PR-WC mixtures.

3 URINARY N DILUTION

Nitrogen concentration in urine from ruminant livestock decreases by 13%–75% as PL inclusion in the animals' diet increases (Box et al., 2017; Minnée et al., 2017; Minnée et al., 2020; Navarrete et al., 2022; Navarrete et al., 2023; Nkomboni, 2017). In a meta-analyses of studies, Nguyen et al. (2022b) described a significant negative association between the content of PL in the diet of cattle and urine N concentration (relative change of urine N concentration (%): y = −0.0645 × PL content (g kg⁻¹ DM) + 0.61; R² = 0.69, p < .001), mainly due to an increase in urine volume (relative change of daily urine volume (%): y = 0.0659 × PL content (g kg⁻¹ DM) – 12.73; R² = 0.64, p < .001) and to a lesser extent reduced urinary N excretion (relative change of UN excretion (%): y = −0.0297 × PL content (g kg⁻¹ DM) – 5.86; R² = 0.22, p = .036).

The greater urine output is associated with more frequent urination events per animal per day (Mangwe et al., 2019; Marshall et al., 2022a; Minnée et al., 2020). The dilution effect of PL has also been observed in sheep (O'Connell et al., 2016) and red deer (Cervus elaphus L., Beck et al., 2020).

The increase in urine output is suggested to be a result of diuresis, driven by the excessive consumption of water due to the low DM% in the herbage. Plantain herbage can contain up to 30% more water relative to ryegrass (Minnée et al., 2019) resulting in total water intakes of 10–20 L cow⁻¹ d⁻¹ greater than cattle offered ryegrass-clover diets. The increased water intakes were largely from water consumed in the herbage, as the animals were also observed to reduce consumption of drinking water as PL content in the diet increased (Al-marashdeh et al., 2019; Dodd et al., 2019a; Minnée et al., 2020). This observation is important for livestock farmers who provide mineral or medications through drinking water, as with this method livestock consuming PL may not receive the desired dose.

It has been suggested that increased mineral concentrations in PL herbage could also contribute to diuresis, or that the PL bioactive aucubin has diuretic properties (Deaker et al., 1994). The latter is supported by a study that compared urine output from sheep fed different cultivars of PL that varied in bioactive concentration (Judson et al., 2018b). Further research is required to quantify the contribution of each mechanism towards diuresis.

4 ANIMAL N PARTITIONING

The diets of livestock in forage-based pastoral systems often contain levels of N that exceed animal requirements (Pacheco & Waghorn, 2008). In addition, the efficiency with which ruminants utilize dietary N is low, with lactating dairy cows utilizing 13%–31% of N consumed (Castillo et al., 2000). Because ruminants cannot store excess N, any dietary N not utilized for production will be excreted, mainly in urine (Kebreab et al., 2001). While urinary N excretion (g cow⁻¹ d⁻¹) is closely related to N intake, there are certain dietary characteristics that can affect the utilization of N, and concurrently the amount of N that is partitioned to urine. These characteristics include soluble carbohydrate content, N digestibility, and bioactive compounds that reduce rumen N metabolism. Plantain herbage possesses all three characteristics, and differences in partitioning of dietary N between livestock fed diets with or without PL were initially reported by Cheng et al. (2017), and later confirmed by Minnée et al. (2020) and Marshall et al. (2022a).

Soluble carbohydrates: Soluble carbohydrates (SC, i.e., sugars) provide a key source of energy for rumen microbiota. When there is a ready supply of SC in the rumen, dietary N can be incorporated into microbial protein by microbial protein synthesis, which is available for use by the animal for production or growth. Conversely, when SC availability limits microbial protein synthesis, N accumulates in the rumen as ammonia. It is then absorbed into the bloodstream and processed into urea in the liver and excreted in urine. Temperate pastures generally contain high concentrations of N and low concentrations of SC, and this imbalance results in poor utilization of dietary N for ruminant products, with high amounts of N excreted in urine (Belanche et al., 2013). In a review of studies with perennial ryegrass, Edwards et al. (2007) showed that increasing SC concentration in herbage reduced urinary N excretion.

Plantain herbage consistently contains greater SC content than ryegrass (on average 299 vs. 211 g kg⁻¹ DM) with similar N content (Minnée et al., 2019). Evaluation of the production performance of cattle and sheep showed improved milk production and/or live-weight gain per unit of dietary N intake when fed diets which included PL, relative to ryegrass-clover diets (Box et al., 2017; Mangwe et al., 2019; Minnée et al., 2017; Moorhead et al., 2002), indicating improved dietary N utilization.

To confirm a N partitioning effect of PL, cows were housed in metabolisms stalls and offered ryegrass-clover diets with graded PL substitution (0%, 15%, 30% and 45% of the diet on a DM basis; Minnée et al., 2020). Nitrogen intake was similar across the diets, but SC intake and the amount of dietary N partitioned to milk production increased with increasing PL content of the diet. Up to a 6% increase in N utilization was observed, through an additional 23 g N d⁻¹ partitioned to milk, and a 30% increase in milk solids (milk fat + protein) yield from cows fed 45% of the diet as PL relative to cows fed ryegrass-clover-only diets. Further, the amount of dietary N partitioned to urine decreased from 50 to 39% with increasing PL substitution, equating to 66 g less N day⁻¹ excreted to urine N from cows fed diets of 45% PL. Similarly, Cheng et al. (2017) reported that dairy heifers fed PL partitioned 34% less dietary N to urine compared with heifers offered a ryegrass-clover diet.

Some of the reduction in N partitioned to urine is found to be associated with increased N excreted to faeces from cows fed PL and this is expanded upon below.

Protein degradability: Much of the protein in forages is degradable in the rumen (75%–90%; Waghorn & Clark, 2004). Degradable protein can be further defined into soluble protein (SP) which is rapidly degraded, and insoluble degradable protein which is more slowly degraded in the rumen. The proportions of rumen degradable and SP in the forage influence the rate of ammonia accumulation in the rumen, and in turn, depending on energy supply, influence N excretion. Dietary protein that is not degradable in the rumen (rumen undegradable protein, RUP), largely is protein associated with plant cell walls and is difficult for rumen microbes to access and thus is either digested in the small intestine (digestible RUP) or is excreted in faeces. Feeds with lesser protein degradability can promote greater partitioning of dietary N to milk and faeces, and less to urine through less rapid ammonia production in the rumen and greater bypass of N from the rumen to the intestines (Kebreab et al., 2001).

In general, PL and PR herbage contain a similar amount of total N, but PL has less SP (on average 12 vs. 38% of total N) and more RUP (on average 64% vs. 31% of total N) (Minnée et al., 2019) than ryegrass. Minnée et al. (2020) reported a strong positive relationship between PL content in the diet and partitioning of N to faeces (R² = 0.8, p < .001) from lactating dairy cows, with no negative effect on animal production. In agreement, the study by Marshall et al. (2021) showed that lactating dairy cows fed diets of 100% PL partitioned 30% more dietary N to faeces than cows fed ryegrass diets.

Bioactive compounds: Plantain contains various bioactive compounds such as iridoid glycosides (aucubin, catalpol) and phenylpropanoid glycoside (acteoside) (Box et al., 2019), and an in vitro study (Navarrete et al., 2016) showed that ammonia production was reduced by up to 40% when these compounds were added to incubating rumen fluid. The authors suggest this may have been associated with an antimicrobial effect of the bioactive compounds reducing dietary N degradation, or the provision of an alternative energetic substrate. These results require further investigation to fully elucidate the contribution of these three bioactive compounds for altering N partitioning in ruminants. Condensed tannins are another group of secondary compounds known to influence N partitioning through reducing ruminal degradation of dietary protein resulting in increased flow of protein to the small intestine (McMahon et al., 2000). Condensed tannins are noted as not being present in PL (Sanna et al., 2022) or at low to moderate levels (Hamacher et al., 2021; Jackson et al., 1996; Stewart, 1996), though the direct effects of tannins in forage PL appear to be little studied. It is possible, however, that tannins in PL are associated with the low rumen ammonia concentrations measured in livestock fed diets including PL (Minnée et al., 2017).

5 SOIL N PROCESSING

Urine from sheep and cattle fed PL appears to contain yet unidentified compounds that delay soil nitrification (Judson et al., 2018a, 2018b; Judson et al., 2019; Peterson et al., 2022). Inhibiting nitrification in the soil retains N in the less-leachable ammonium form and extends the window for plant N uptake, giving soil microbes more time to accumulate their favoured N source and reducing the amount of potentially leachable nitrate and/or emitted nitrous oxide.

Soil microcosms treated with urine from sheep grazing pure swards of PL showed reduced nitrate production for up to a month compared with those treated with urine from sheep grazing ryegrass, and delayed nitrate production in the first 10–12 days of the incubation compared with a treatment with no urine added (Figure 2; Peterson et al., 2022). The degree of inhibition appears to be related to sufficient concentrations of the necessary compounds being excreted in the urine and so is dependent on a number of factors: (1) length of time grazing the forage, (2) the percentage of PL in the diet, and (3) the cultivar of PL consumed.

While PL has been shown to induce diuresis in sheep after only a 24 h grazing period (O'Connell et al., 2016), the inhibitory effect in microcosms treated with urine collected from sheep after 48 h of grazing on pure PL stands was not obvious, and only became apparent after at least 7 days of grazing PL (Peterson et al., 2022; Podolyan et al., 2020). Increasing the proportion of PL in the diet of sheep participating in a controlled feeding experiment from 25% to 50% resulted in not only a 30% decrease in total urine N, but a 20% decrease in nitrate produced in soil microcosms after 21 days (Peterson et al., 2023a). Finally, significant differences in nitrate produced in soil microcosms treated with urine from sheep grazing different PL cultivars have been observed, particularly in the early phase of the incubation (Fraser et al., 2018; Peterson et al., 2023b; Figure 2). These differences are thought to be related to the concentrations of bioactive compounds in the leaf material at the time of consumption, which are known to vary with cultivar and season (Box & Judson, 2018). The compounds responsible for this urine-derived inhibition, along with the mechanism by which they delay nitrification are yet to be confirmed. Nonetheless, there is evidence to suggest that these compounds directly inhibit the enzymes that convert ammonium to nitrite, without deleterious effect on the soil microbiome as a whole (Peterson et al., 2023b).

There is evidence to suggest that compounds exuded from the roots of PL may also act as inhibitors of nitrification in soil. Rauber et al. (2008) observed a decline in N mineralisation in field plots when potatoes were under-sown with PL, while soil mesocosms dominated by PL had significantly reduced nitrate concentration, mineralisation, and nitrification rates compared with mesocosms dominated by Anthoxanthum odoratum L. or Lotus corniculatus L. (Massaccesi et al., 2015). Indeed, root exudates collected from PL grown in hydroponics show cultivar-dependent inhibition in a bioassay using a pure culture of a nitrifying bacterium, and correlations with root exudate chemistry (as determined with untargeted metabolomic profiling) have been made (Figure 3; Peterson et al., 2023b).

Lysimeter and field trials have been critical in helping to understand the direct effects of PL root exudates on soil N processes and estimating any contribution of inhibitory compounds excreted in the urine. Woods et al. (2017) reported a 45% reduction in N leaching when urine from dairy cows grazing PR-WC swards was applied to lysimeters in IR-WC-PL pasture (42% PL on a DM basis), compared with lysimeters in PR-WC. Further, when urine derived from dairy cows grazing swards containing PL was applied to lysimeters with IR-WC-PL, the reduction was 89% compared with when urine derived from cows grazing PR-WC swards was applied to PR-WC swards. However, this may be attributed in part to the lower concentration of urine-N in the IR-WC-PL urine (applied at 508 kg N ha⁻¹), compared with the PR-WC urine (applied at 664 kg N ha⁻¹) and greater uptake of N due to higher growth rate of Italian ryegrass over the cooler season. Carlton et al. (2019) demonstrated a 74%–82% reduction in N leaching from lysimeters with swards containing 20%–30% PL compared with PR-WC swards, when standardized fresh urine from cows fed PR-WC was applied. In similar work, Welten et al. (2019) observed reductions in N leaching of 15% when urine collected from dairy cows grazing a PR-WC pasture was deposited on lysimeters containing PL monoculture in summer (relative to ryegrass); this increased to 50% when the urine was deposited in winter. Talbot et al. (2021) also examined the timing of urine deposition (from cows grazing PR-WC) on lysimeters with swards containing 25%–35% PL and found reductions in N leached of 14%–24% compared with lysimeters under PR-WC, irrespective of timing.

The potential for PL in both the diet and in the sward to reduce nitrous oxide emissions that originate from the urine patch has also been assessed. The direct addition of PL leaf extract or aucubin (a bioactive compound found in PL thought to be implicated in nitrification inhibition) to the urine of cattle grazing PR-WC, resulted in decreased urine nitrous oxide emissions of 50% and 70%, respectively when this urine was applied to a PR-WC sward (Gardiner et al., 2018). A follow-up experiment, however, found that nitrous oxide emissions were not significantly reduced when urine containing aucubin was applied at a lower N loading rate, although evidence for short-term nitrification inhibition was observed (Gardiner et al., 2020). Application of urine from dairy cows grazing PR-WC at the same rate to monoculture plots of PL and perennial ryegrass resulted in a 28% reduction in nitrous oxide emissions from the PL plots compared with the ryegrass, suggesting that these observed difference in emissions were solely plant-induced (Luo et al., 2018).

Pijlman et al. (2020) examined the effect of root exudation in the absence of added urine and found almost 40% lower potential nitrification in mesocosms under monoculture of PL versus a monoculture of ryegrass. In the field, the presence of PL resulted in cumulative nitrous oxide fluxes that were 39% lower compared with the monoculture of perennial ryegrass. Simon et al. (2019) evaluated whether any effects on nitrous oxide emissions could be attributed to a ‘urine’ or a ‘plant’ effect when PL was incorporated into the sward and found that reductions in the nitrous oxide emission factor (nitrous oxide-N emitted as % of N applied) could only be attributed to a direct ‘plant’ effect, possibly due to nitrification inhibition via root exudation.

There are a number of factors that may explain the inconsistencies in N loss reduction reported when PL is incorporated into the pasture sward. Contrasting soil types will have vastly different mineral N concentrations, organic carbon content, microbiomes, soil texture and water-holding capacities – all of which could be expected to alter soil N processes. Soil nitrification inhibition has been shown to be dependent on the soil type in which the plant is grown (Gopalakrishnan et al., 2009; Illarze et al., 2021; Ipinmoroti et al., 2008; Subbarao et al., 2012), likely due to differences in the abundance and diversity of ammonium-oxidizing bacteria (AOB) and archaea (AOA) in these soils (Clark et al., 2021). Carlton et al. (2019) report that the AOB abundance was lower in soils under PL compared with PR-WC after urine treatment and they attribute lower nitrate leaching losses in the PL soil to reduced nitrification rates as a result of this lesser abundance.

The inherent structure of a soil may also influence the extent of N losses. A free-draining soil may allow rapid leaching and so a shorter residence time of N in the root zone could reduce the potential for root exudates to alter N transformations. Talbot et al. (2021) suggests that the smaller reduction in N leaching observed in their lysimeter study may be due to the shallow stony soil used compared with the deep silt loam and fine sandy loam soils used by Welten et al. (2019) and Carlton et al. (2019), respectively.

Differences in root architecture, biomass and growth rate may also influence N leaching or N emissions by altering soil hydraulic properties and/or mineral N uptake (Dunbabin et al., 2003; Gregory, 2006). There is evidence that efficiencies in reducing N leaching under some plants may be not only due to the exudation of nitrification inhibitors from the roots, but also to retardation in solute flow associated with a fine root system and a more complex pore network (Galdos et al., 2020). While there is evidence for exudation of inhibitory compounds, PL is thought to have a coarse root system that consists of both a tap root and fibrous roots; having other species present in the sward to maximize the uptake of soil mineral N or modify the soil structure sufficiently may be more useful than maximizing the amount of PL in the sward.

6 PADDOCK SCALE N LEACHING

Paddock scale investigations provide confidence that the PL attributes able to reduce N leaching in grazed pastures, as described above, are scalable. Navarrete et al. (2018) measured 85%–90% lower N leaching losses in the first year from hydrologically isolated and grazed paddocks containing PL, compared with PR-WC swards. In their study, the total N leaching from grazed pastures was determined using paddocks individually isolated with a mole-pipe drain system from which all drainage water can be collected and analysed for nitrate concentration (Bowler, 1980). This methodology for measuring N leaching from grazed pasture provides a useful “proof of concept” of the combined effects of PL, rather than estimating this from combining the various effects on farm system components in a computer model.

The evaluation of N leaching losses, using this methodology, from incorporating PL in PR-WC pastures, in a dairy system for multiple years, can provide farmers evidence and confidence that PL works for reducing N leaching from their farming systems. Results showed that pastures with 30 to 50% of PL (and otherwise PR-WC) resulted in N leaching reduction of up to 60% in year 1, and by up to 46% during year 2 when compared with PR-WC (Table 2).

The grazing of PL-PR-WC pastures by lactating cows over 2 years demonstrated the PL ability to reduce nitrate leaching. Lower urine N concentration, reduced urine N excretion and reduced nitrification, as described in the sections above, all contribute to more opportunity for plant uptake of urine N and a lower surplus subject to being lost via N leaching (Li et al., 2012).

7 EFFECTS OF FEEDING PLANTAIN ON MILK AND MEAT COMPOSITION

When feeding larger proportions of PL to production livestock, it needs to be clear whether there could be any detrimental effects on product quality. The nutritional composition of PL does not suggest that feeding it to dairy cows will have a substantial impact on milk composition, relative to feeding PR-WC pastures. However, Minnée et al. (2020) fed PL at 3 levels up to 45% of a pasture diet in late lactation and showed a decline in bulk milk fat content with increasing PL “dose”, in association with a 15%–20% increase in milk solids yield. Box et al. (2017) fed PL at levels up to 100% of the diet and demonstrated similar effects on bulk milk fat content and milk yield in late lactation.

A meta-analysis of trials comparing the wider impacts of PL in dairy cattle diets, including effects on milk composition (major components), has been published recently (Nguyen et al., 2022b). The data presented show a small reduction in milk fat content of 0.13 g/100 g milk in early lactation and 0.29 g/100 g milk in late lactation. The meta-analysis covered data from 12 studies (eleven from NZ and one, Pembleton et al., 2016, from Tasmania) where PL had been fed to dairy cattle. Plantain feeding maintained, and sometimes enhanced, milk and milk protein production. The reduction in milk fat content was an inconsistent response and milk yield responses to PL feeding were confined to late lactation. There was no impact of PL feeding on milk protein content apart from one report of a small increase which was associated with an increase in milk protein yield (Nkomboni et al., 2021). Marshall et al. (2022b) also reported changes in multiple minor milk components (mostly intermediary metabolites) from cows fed PL vs. PR herbage.

The cause of the reduction in milk fat content with PL diets, when it occurs, is currently unknown. In part, the inconsistency of this response may have been because some individual experiments were underpowered. Evaluation of volatile fatty acids in the rumen of deer during PL feeding showed a reduction in the acetate: propionate ratio (Swainson & Hoskin, 2006). Similar changes were reported in the rumen of dairy cows by Marshall et al. (2021). Increased propionate is consistent with a reduction in the supply of lipogenic precursors to the mammary gland for milk fat synthesis.

Changes in the fatty acid composition of milk fat, in response to PL feeding, have been reported (Mangwe et al., 2018). Data for fatty acids greater than C14 chain length were presented alongside odd-number carbon chain and branched chain fatty acids. Both C18:2 (linoleic acid) and C18:3 (α-linolenic) content were close to double in milk fat from cows fed PL relative to PR-WC, yet the content and intake of α-linolenic acid in the PL herbage were similar (within 10%) to that of PR-WC. Linoleic acid intake was >40% higher on the PL diet relative to PR-WC. These results suggest that this fatty acid and the other major polyunsaturated fatty acid in PL lipid, linoleic acid, must be relatively protected from biohydrogenation in the rumen compared to the same fatty acid in PR-WC. There were also small changes in branch-chain and odd-number carbon length fatty acids in milk fat reported by Mangwe et al. (2018). Similar changes in milk fatty acids in response to feeding PL herbage have been reported by Marshall et al. (2022b). Feeding PR-red clover-PL silage, with the latter at 35% inclusion in the stack, also resulted in a marked increase in the C18:3 fatty acid content of milk fat (>5-fold relative to total mixed ration, TMR, Ineichen et al., 2019). Effects on C18:2 were much smaller. Marshall et al. (2022b) also studied phytochemical content and noted multiple profile changes in response to feeding PL compared with PR, suggesting the potential for altering the nutraceutical value of milk.

The effect of PL feeding on meat quality in grazing ruminants has not received the same research attention as dairy. In lambs there was no consistent effect of PL feeding on the sensory characteristics of meat, but some perceived improvement in nutritional quality through increased content of polyunsaturated fatty acids (Campbell et al., 2011). Similar data on beef meat composition when fed PL as a major part of the diet are not available.

8 IMPLEMENTATION CHALLENGES

The inclusion of PL in the forage base on farms should be very attractive and manageable where farmers are already familiar with establishing and managing ley and mixed-species pastures to improve forage supply and quality for livestock. For example, pasture renewal rates on New Zealand dairy farms range between 6% and 8% of farm area on an annual basis (Kerr et al., 2015), with rates likely to be lower on drystock farms. In addition, there is now a 25+ year experience base with PL across the New Zealand pastoral sector to draw on.

9 CONCLUSIONS

The results of lysimeter studies demonstrated that including PL in pasture reduced N leaching by 14%–89% when cow urine was applied. Results of paddock-scale measurements confirm significant reductions in N leaching: 20%–90% when PL comprised 30%–40% of the DM of PL-PR-WC grazed pastures.

No negative effects on milk production and composition have been found when feeding PL and indeed some improvements in milk fatty acid profiles have been observed. On the contrary, pasture mixtures with PL often outyield PR-WC pastures in summer and autumn in warm, dry regions. However, challenges with implementation remain. Significant differences between PL cultivars were found for their secondary compounds content and associated effect on soil nitrification rates, and this needs to be better understood to ensure this pathway contributes to N loss reductions. Furthermore, management options need to be clarified to ensure successful establishment and maintenance of PL in productive, grazed pastures, so that meaningful proportions of PL in pasture and the grazing animal's diet can be achieved and maintained.