Inclusion of Leguminous Green Manures Enhances Crop Biomass, Nutrient Uptake, Soil Phosphorus Dynamics and Bioavailability

Abstract

ABSTRACT

Introduction

In agroecosystems, phosphorus (P) applications over a long time have accumulated in soil as legacy P. This environmental challenge can be an agronomic opportunity as soil legacy P could be recovered in cropping systems using practices such as green manuring. We hypothesised that, at moderate soil available P levels, plant‐soil interactions under green manures can mobilise soil legacy P and promote cereal crop P uptake and growth.

Methods

Alongside a fallow treatment, three green manure treatments that included two legume treatments (narrow‐leaf lupin [Lupinus angustifolius], pea [Pisum sativum L.]) and one cereal treatment (wheat [Triticum aestivum] and barley [Hordeum vulgare]) were rotated with the main crops of wheat and barley in two phases on a pumice soil (27 mg kg⁻¹ Olsen P) in a microcosm experiment. Plant roots and shoots and end‐of‐experiment soil samples were collected for analysis.

Results

Over two crop rotations, inclusion of narrow‐leaf lupin and pea green manures significantly increased main crop biomass (27%–35%) and P uptake (15%–29%) relative to control, while the cereal green manure decreased the following crop's yield (−13%) and P uptake (−19%). Relative to fallow, microbial biomass P and soil organic P pools increased under all green manures yet total inorganic P decreased under leguminous green manures. This depletion (35 mg P kg⁻¹) under narrow‐leaf lupin was equivalent to ~47 kg P ha⁻¹. Phosphatase enzyme activities relevant to P cycling increased particularly under leguminous green manure treatments.

Conclusions

Leguminous green manures such as narrow‐leaf lupin could mobilise soil P to crops in field conditions, suggesting that drawdown of soil legacy P while sustaining crop yield can be tenable.

Introduction

In many contexts for crop production, P fertiliser is applied to compensate not only for expected P removal by crops but also for P complexed by soil (Coelho et al. 2020). Consequently, continued P applications exceeding P crop requirements (Condron 2004; Zhu, Li, and Whelan 2018) have accumulated significant amounts of surplus P in soils. This build-up of anthropogenic P inputs over time, called ‘legacy P’ (Condron et al. 2013; Zhu, Li, and Whelan 2018), degrades water quality for long periods afterwards (Sharpley et al. 2013). Recycling this legacy P would not only reduce eutrophication impacts but provide an agronomic opportunity in some systems, potentially obviating P fertiliser for up to several decades (Doydora et al. 2020). Since much of soil legacy P may be stored in less labile forms of inorganic and organic P, practices are needed to mobilise this P for crop uptake. Soil amendments are one option (Chassé and Ohno 2016; Edwards et al. 2016), but, more practically, conservation management practices could mobilise soil legacy P while providing additional soil conservation benefits. Notably, including a green manure crop in the off-season (Chavarría et al. 2016), also termed cover cropping, has strong potential to improve cash crop P uptake while minimising P inputs.

Green manures can provide several ecosystem services in cropping systems, including the protection of soil from erosion, interception of excess nitrate before it leaches, and the buildup of soil organic matter (Basche and DeLonge 2017; Blanco-Canqui et al. 2022; Wittwer and van der Heijden 2020). An additional yet understudied benefit of green manures is their ability to mobilise soil P to supply the following cash crop. Green manures can access soil P through a range of mechanisms, primarily in the rhizosphere, such as by promoting P desorption, chelating cations with root organic anions, and mineralising organic P via phosphatase activity (Hallama et al. 2019; Nguyen, McDowell, and Condron 2024; Richardson et al. 2009b). For example, Nguyen, McDowell and Condron (2024) found that Lupinus angustifolius (narrow-leaf lupin) depleted moderately labile inorganic P and stable inorganic P, and those depletions were correlated with malate exudation rates of five green manure crop species. Moreover, the authors reported that Pisum sativum (pea) mineralised moderately labile organic P. These effects on mobilising soil P may also carry over to the following cash crop, improving the crop's ability to acquire soil P (Hansen et al. 2022; Maltais-Landry and Frossard 2015; Mat Hassan, Hasbullah, and Marschner 2013). As a further benefit, depending on the green manure species, the crop, and soil P status, the green manure residues can supply as much as 20%–40% of crop P uptake (Hansen et al. 2022; Maltais-Landry and Frossard 2015). However, P mobilisation and crop P uptake under green manures varies greatly among green manure plant species and soil P status (Hansen et al. 2022; Soltangheisi et al. 2018), which may inhibit the adoption of green manures.

As green manure species vary in their root traits, growth, biomass composition, and more, so too do they vary in their ability to acquire P from less labile soil P pools. Currently, studies are inconsistent as to which green manure species or plant types improve soil P cycling and P mobilisation for the cash crop. Two general types of green manure species in particular, those in the Fabaceae (legumes) or Poaceae (e.g., cereals) families, generally produce high shoot biomass P (~7 to 10 kg P ha⁻¹) and improve crop yields (Hallama et al. 2019), albeit dependent on context. For example, Mat Hassan, Hasbullah and Marschner (2013) reported the greatest wheat (Triticum aestivum) P uptake following incorporation of white lupin (Lupinus albus) residue, while effects for soil P mobility were mixed when residue was removed for leguminous (including Cicer arietinum and Vicia faba) or wheat green manures. However, Maltais-Landry, Scow and Brennan (2014) found that cereals had larger effects on rates of P cycling than did legumes due to greater biomass production, although greater carbon (C) to P ratios in the residue could encourage microbial P immobilisation rather than crop P uptake. Furthermore, rhizosphere properties may not always relate to P acquisition by the cash crop. Maltais-Landry (2015) observed in a greenhouse experiment that, compared to cereal species (Avena sativa, wheat), legumes (e.g., Vicia faba, Vicia benghalensis) promoted more rhizosphere enzyme activity and the release of organic anions, leading to greater green manure P uptake. Another positive example is that white lupin released significant amounts of citrate under P deficient soils for an improved P uptake, while organic anions were not exuded or showed limited release under wheat (Richardson et al. 2009a). In contrast, however, for two different Californian soils, Maltais-Landry, Scow and Brennan (2014) measured similar changes in rhizosphere properties under legumes (e.g., Pisum sativum, Vicia dasycarpa, Vicia faba) yet greater P uptake rates for cereals (e.g., Avena sativa, Secale cereale) and a weed/pea mixture (including volunteer winter wheat). Further work should compare green manure species not only in agronomic terms but also in context of changes to soil P.

The benefit of green manures for mobilising soil P to the cash crop correlates with soil P status, with typically greater relative effects for soils with low P availability (Hallama et al. 2019). Part of the variation in soil P mobilisation under various green manures likely relates to soil P status, as soil P distributed in pools of varying lability interacts with microbial activity under green manures. In soils with significant organic P, increased phosphatase and other enzyme activities under green manures can improve the mobility of this pool (Hallama et al. 2019; Hallama et al. 2021). For example, Randhawa et al. (2005) reported that, in a soil with low available P (7 mg Olsen P kg⁻¹), mineralisation rates of soil organic P after green manure amendment were fivefold compared to unamended soil. Soil P complexed with reactive surfaces (e.g., metal oxides) may be mobilised through the release of organic anions under green manures (Richardson et al. 2009b); across five green manure species in two soils, malate exudation correlated with depletion of moderately labile and stable inorganic P pools (Nguyen, McDowell, and Condron 2024). The soil microbial biomass itself is also an important P pool, which several green manure species often increase (Hallama et al. 2019; Piotrowska-Długosz and Wilczewski 2020), thus increasing net soil P transfer into a pool of high turnover before the cash crop's growing season (Seeling and Zasoski 1993). These mechanisms may explain the benefits of green manures in soils with low available P (e.g., Olsen P < 10 mg kg⁻¹) (Hansen et al. 2022; Mat Hassan, Hasbullah, and Marschner 2013; Rick et al. 2011). At the same time, green manures may improve crop production on soils with moderate to high P availability, where legacy soil P reserves need to be drawn down to avoid environmental P losses (McDowell et al. 2020).

In this study, we assessed the ability of one cereal and two legume green manures to mobilise soil P for cereal crops (wheat and barley) in a rotation system on a volcanic soil with moderate P availability (i.e., at an optimal plant available soil P test concentration). Our main hypothesis was that green manures can mobilise soil legacy P (including less labile forms) at moderate available P levels and return labile P to soil for the next crop during incubation. We also hypothesised that phosphatase enzyme activities and microbial P would increase under green manures compared with an unplanted fallow.

Materials and Methods

Soil

A pumice soil (immature Orthic Pumice (NZ Classification) (Hewitt 2010): Typic Udivitrands (USDA); Vitric Andosol (World Reference Base)) was collected from the top 10 cm at a permanent grazed grassland in central North Island, New Zealand (Table 1). Orthic Pumice soils typically have low clay content (< 10%) yet moderate to high P retention (here, 46% anion storage capacity). The site received moderate inputs of P fertiliser (15–20 kg P ha⁻¹ year⁻¹) over many years. The soil was air dried and sieved < 4 mm.

Table 1 Chemical properties determined for the pumice soil.

Analysis	Units	Value
pHH2O	—	5.8
Total C	%	4.6
Total N	%	0.28
C:N	g g−1	16.4
CEC	cmolc kg−1	15
Total P	mg kg−1	645
Olsen P	mg kg−1	27
Anion Storage Capacity	%	46

Experimental Design

The experiment utilised a glasshouse under controlled conditions (10°C–25°C) over 12 months. Temperature was maintained through either heating or ventilation. Polyethylene pots were filled with 420 g of soil (pre-incubated). Using eight replicates per treatment in a completely randomised design, pots were randomly assigned either a fallow treatment or one of the following three green manure treatments (Table S1): narrow-leaf lupin (Lupinus angustifolius), pea (Pisum sativum), or cereal (wheat [Triticum aestivum] for the first rotation and barley [Hordeum vulgare] for the second rotation). All pots started in a green manure phase (40 d, before flowering), followed by a wheat crop (60 d), then a second green manure phase (40 d), then a barley crop (60 d). The main crops were only grown to the maximum vegetative stage (initiating reproduction stage) as soil P uptake reaches its maximum near this point. Fresh green manures after harvest were incorporated into soil and incubated for 30 d before sowing main crops, allowing time for partial decomposition and stability of the soil microbes. Seeds were pre-germinated (narrow-leaf lupin and pea, 4 d; wheat and barley, 2 d) in petri dishes lined with wet filter papers. Green manures and main crops were sown at the same density of 4 plants per pot. Both main crops received basal nutrients of 30 mg N, 15 mg S, 15 mg Mg, 2 mg Zn, 0.2 mg Mo and 0.3 mg B per kg soil. Before the second green manure phase, 3.5 g CaCO₃ kg⁻¹ soil was applied, followed by 1 week incubation, to maintain soil pH and supply more calcium. The soil moisture of all treatments including fallow was maintained at 70% field capacity (Randhawa et al. 2005) throughout the experiment by weighing pots and watering daily.

Green manures were harvested for determining fresh biomass and incorporation into soil. Before harvest, the soil was allowed to dry to 40% field capacity to minimise the soil adhering to roots while avoiding the wilting point. After measurements, shoots and roots were chopped into 0.5 cm long pieces, incorporated into the soil, and mixed thoroughly. The mixture of soil and plant biomass were transferred back to the pots, and incubated at 70% field capacity for 30 d to allow for decomposition. Root hairs of narrow-leaf lupin and pea were visually assessed by microscope.

Sampling and Analysis

Plant shoots were cut at the soil surface. Roots were separated from soil and washed. Crop shoot and root samples were dried at 65°C for 48 h for dry matter biomass determination, ground to < 1 mm and analysed for total N by the Dumas combustion method, and for total P by ICP-OES following digestion with nitric acid-hydrogen peroxide (Anderson 1996; Metson 1972).

After the wheat crop, moist soil samples (kept cool, 4°C) were collected for acid and alkaline phosphatase activities (Tabatabai 1994) and microbial biomass P (McLaughlin, Alston, and Martin 1986) within 7 d. Subsamples of air-dried soil were used for P fractionation, total carbon (C), and total nitrogen (N) analyses. Total C and N were determined by using an Elementar Vario Max CN Elemental Analyser. Soil P fractionation followed the sequential extraction scheme developed by Boitt et al. (2018b) based on Chen et al. (2000) and Condron and Newman (2011). This involved sequential extraction with 1 M NH₄Cl (soluble inorganic P), 0.5 M NaHCO₃ at pH 8.5 (labile P), 0.1 M NaOH (moderately-labile P), 1 M HCl (acid-soluble inorganic P), 0.1 M NaOH (stable P), followed by digestion of the residue with H₂SO₄ and H₂O₂ (residual P). Inorganic P in the extracts of 0.5 M NaHCO₃ and 0.1 M NaOH was determined by colourimetry via He and Honeycutt (2005), while inorganic P in the extracts of 1 M NH₄Cl and 1 M HCl, and total P in alkaline extracts (after ammonium persulphate digestion [Eisenreich, Bannerman, and Armstrong 1975]) were determined by the molybdenum-blue method (Murphy and Riley 1962). We refer to the difference between total and inorganic P in the NaHCO₃ and both NaOH extracts as organic P. Soil pH was measured at a ratio of 1 g air-dried soil to 2.5 mL deionised water after a 30 min equilibration.

Statistical Analysis

Uptake and plant biomass related variables throughout are expressed primarily as mass per pot and, when relevant, on an areal basis assuming 15 cm soil depth. Statistical analysis utilised R statistical software (version 4.3.1, R Core Team, 2023) assuming α = 0.05. Plotting and data manipulation were supported by the ‘tidyverse’ packages (Wickham et al. 2019). Differences between multiple green manure treatments were estimated by one-way analysis of variance (ANOVA), facilitated by the ‘multcomp’ package (Hothorn, Bretz, and Westfall 2008), including green manure biomass; crop yields, P and N uptake; soil C, N, C:N ratios, P fractions, microbial biomass P, and phosphatase enzyme activities (Tukey's HSD). Residual diagnostics ensured that statistical assumptions were met (e.g., with QQ-plots). Correlations between variables were estimated by Pearson's test and further assessed with linear regression.

Results

Green manure shoot and root biomass differed between the three green manure treatments (Table 2). Over both green manure phases, narrow-leaf lupin produced the greatest shoot and root biomass (34 and 38 g pot⁻¹), followed by pea (17 and 31 g pot⁻¹), and cereal (9 and 14 g pot⁻¹). Crop biomass (Figure 1) in the cereal green manure treatment was lower than in the fallow (decreases of 33% barley roots, 32% wheat shoots, 24% wheat roots), while legume green manures (narrow-leaf lupin and pea) increased biomass of barley (30%–42% shoots, 30%– 48% roots) and wheat (31%–41% shoots, 18%–19% roots). Total cumulative barley and wheat biomass were lower (13%) under cereal green manure compared with fallow, while increases occurred under legumes (35% for narrow-leaf lupin, 27% for pea).

Table 2 Mean fresh matter yields for the first and second green manures and their sum.

Treatment	Green manure 1	Green manure 2	Sum of green manure phases
Shoot	Root	Shoot	Root	Shoot	Root
g pot−1
Narrow-leaf lupin	19.85 ± 0.69 a	22.95 ± 1.34 a	14.01 ± 0.28 a	15.99 ± 0.78 a	33.86 ± 0.70 a	38.94 ± 1.77 a
Pea	6.69 ± 0.46 b	18.17 ± 1.35 b	9.98 ± 0.48 b	12.87 ± 0.99 b	16.67 ± 0.78 b	31.04 ± 1.84 b
Cereal	2.10 ± 0.07 c	6.97 ± ± 0.37 c	6.52 ± 0.35 c	6.75 ± 0.35 c	8.62 ± 0.38 c	13.72 ± 0.45 c

[IMAGE OMITTED. SEE PDF]

The cereal green manure decreased barley root P uptake by 28% compared with fallow (Figure 2), while there were increases in P uptake under both legume green manure treatments (shoots 14%–33%, roots 21%–37%). For wheat, the cereal green manure reduced root and shoot P uptake by 18% and 32%, respectively, compared with fallow, while legume green manures increased crop P uptake by 15%–33% and 9%–11% for shoots and roots, respectively. Overall crop P uptake decreased by 19% under the cereal green manure, but increased by 29% and 15% under the narrow-leaf lupin and pea green manures, respectively. Compared with the pattern of P uptake, N uptake by crops following green manures was similar (Figure 3). Inclusion of a cereal green manure decreased total N uptake by 25%, while the narrow-leaf lupin and pea green manures increased total N uptake by 63% and 21%, respectively, compared with fallow.

[IMAGE OMITTED. SEE PDF]

[IMAGE OMITTED. SEE PDF]

For soil P fractions (Figure S1), no differences were observed between fallow and green manure treatments for soil soluble inorganic P, acid-soluble inorganic P, and residual P (Table 3). Relative to fallow, all three green manures increased microbial biomass P and labile organic P but decreased labile inorganic P, although the magnitude of the respective changes was small (2–4 mg P kg⁻¹). Moderately-labile inorganic P (the largest pool in this soil) decreased under legume green manures (12–33 mg P kg⁻¹) compared with fallow but increased under cereal green manure (13 mg P kg⁻¹). Moderately labile organic P increased under pea (25 mg P kg⁻¹) and cereal (56 mg P kg⁻¹) green manures. Only narrow-leaf lupin caused a small decrease in stable inorganic P (1 mg P kg⁻¹), while stable organic P increased under the cereal green manure (10 mg P kg⁻¹) compared with fallow. The narrow-leaf lupin green manure decreased total inorganic P (35 mg P kg⁻¹) compared with fallow, while cereal green manure increased this pool by 15 mg P kg⁻¹. Total organic P increased by 71 and 36 mg P kg⁻¹ under cereal and pea, respectively, compared with fallow.

Table 3 Mean P fractions and microbial P (mg P kg⁻¹) for the soil following wheat crop.

P fractions	Fallow	Cereal	Narrow-leaf lupin	Pea
Microbial biomass P	7.25 ± 0.28 b	9.60 ± 0.49 a	10.90 ± 0.32 a	10.00 ± 0.28 a
Soluble inorganic P	0.40 ± 0.03 ab	0.41 ± 0 ± 0.02 ab	0.38 ± 0.01 b	0.46 ± 0.01 a
Labile inorganic P	27.84 ± 0.43 a	26.22 ± 0.13 b	24.36 ± 0.14 c	26.30 ± 0.38 b
Labile organic P	22.75 ± 0.59 c	28.15 ± 0.43 ab	26.62 ± 0.76 b	29.90 ± 0.47 a
Moderately labile inorganic P	233.92 ± 1.56 b	246.57 ± 2.81 a	200.89 ± 2.10d	222.02 ± 2.30 c
Moderately labile organic P	201.97 ± 2.43 c	258.34 ± 4.02 a	208.96 ± 3.76 c	227.34 ± 3.98 b
Acid soluble inorganic P	82.59 ± 1.18 a	85.51 ± 1.04 a	85.47 ± 2.83 a	87.04 ± 1.54 a
Stable inorganic P	15.18 ± 0.14 a	15.95 ± 0.40 a	13.83 ± 0.30 b	15.95 ± 0.35 a
Stable organic P	25.02 ± 0.59 b	34.79 ± 1.04 a	24.90 ± 1.34 b	29.24 ± 1.53 b
Residual P	40.99 ± 1.23 a	42.50 ± 1.90 a	41.16 ± 0.95 a	42.47 ± 1.11 a
Total inorganic P	359.93 ± 1.76 b	374.65 ± 2.82 a	324.93 ± 4.39 c	351.75 ± 2.90 b
Total organic P	249.75 ± 3.18 c	321.28 ± 4.36 a	260.48 ± 5.10 c	286.49 ± 5.16 b

Soil phosphatase activities following the wheat crop were greater for acid rather than alkaline phosphatases (Figure 4). Acid phosphatase activity was greater for the narrow-leaf lupin and pea (5.24 and 5.02 mmol p-nitrophenol kg⁻¹) green manure treatments compared with fallow (3.82 mmol p-nitrophenol kg⁻¹). Similarly, alkaline phosphatase activity was greater for the narrow-leaf lupin, pea, and cereal green manures (1.93, 1.60 and 1.73 mmol p-nitrophenol kg⁻¹, respectively), compared with fallow (1.41 mmol p-nitrophenol kg⁻¹). Likewise, alkaline phosphatase activities mirrored trends in soil pH, where pH was greatest for narrow-leaf lupin (6.56), followed by pea (6.48), fallow (6.42), then cereal (6.40).

[IMAGE OMITTED. SEE PDF]

At the end of the experiment, soil C was lower under narrow-leaf lupin and pea compared with cereal, while soil N was lower under fallow relative to the green manures except for pea (Table 4). The soil C:N ratio was lower for legume green manure treatments compared with fallow and cereal green manure, as expected due to N fixation by the legume green manures via symbiotic rhizobia increasing soil N concentration (Table 4) and, in turn, promoting C mineralisation.

Table 4 Soil total carbon (C), nitrogen (N), and mass C:N ratio following wheat crop across the green manure treatments.

Treatment	C %	N %	C:N g g−1
Fallow	4.03 ± 0.08 ab	0.25 ± 0.01 b	15.97 ± 0.65 a
Cereal	4.41 ± 0.11 a	0.28 ± 0.00 a	15.78 ± 0.22 a
Narrow-leaf lupin	3.85 ± 0.11 b	0.29 ± 0.01 a	13.33 ± 0.18 b
Pea	3.82 ± 0.08 b	0.27 ± 0.01 ab	14.27 ± 0.23 b

Correlation analysis showed that P uptake and moderately labile inorganic P were negatively correlated (Figure 5), but no other P fractions were correlated with P uptake. Moderately labile inorganic P concentration was also negatively correlated with pH. Microbial biomass P was positively correlated with labile organic P, acid phosphatase and alkaline phosphatase enzyme activities, while organic P pools were correlated with one another (Figure S2).

[IMAGE OMITTED. SEE PDF]

Discussion

Leguminous Green Manures Outperform Cereals in Mobilising Soil Phosphorus

In this soil with moderate P availability, legume green manures (pea, narrow-leaf lupin) increased crop biomass and P uptake, while the cereal (wheat, barley) green manure decreased these parameters. Of the legume green manures we studied here, narrow-leaf lupin, followed by pea, showed the greatest capacity to mobilise soil P, thereby returning more labile P from their biomass decomposition to soil for following crops: 3.7 mg P pot⁻¹ (~9 kg P ha⁻¹) was delivered by narrow-leaf lupin to the subsequent crops after two rotations. This P delivery is consistent with the field studies reviewed in Hallama et al. (2019), where P that cycled to cover crop biomass varied from 1 to 30 kg P ha⁻¹. In contrast, the cereal green manure treatments led to lower crop biomass, less P uptake by crop roots and shoots, as well as less N uptake. However, we recognise that cereal green manure species, such as those in the Poaceae family, may perform better in other contexts, such as in Maltais-Landry (2015), producing up to ~10 kg P ha⁻¹ in shoot P content.

Regardless of specific climatic and soil conditions, legume green manure species have traits that improve their ability to mobilise soil P. In this study, legume green manures likely delivered more P to the subsequent crop due to the green manure's biomass stoichiometry and root characteristics. Leguminous green manure biomass typically has lower C per P or N, which promotes decomposition and therefore more available P and N for the following crops (Alamgir et al. 2012; Espinosa, Sale, and Tang 2017; Hallama et al. 2019). Indeed, we observed a strong correlation across all treatments between P uptake and N uptake by succeeding crops. The leguminous green manure treatments also produced more than double the root biomass of the cereal green manure, thus providing an important organic P source directly in the soil environment. Assuming that rhizosphere volume generally scales with root biomass, greater green manure root biomass likely provides more opportunities for soil P mobilisation through a variety of mechanisms discussed below.

Green Manure Rhizosphere Properties Likely Improve the Cycling of Soil Phosphorus

In legacy P contexts, soil P mobilisation under green manures may be an important strategy for producing crops without significant P inputs. To unlock such soil P reserves, green manure species must be adept at mobilising multiple pools of soil P. In this study on a soil with moderate P availability, green manures, particularly legumes, mobilised agronomically significant amounts of soil P across several pools. Relative to fallow, the legume green manures depleted the moderately labile inorganic P pool the most, followed by the labile inorganic P pool. The cereal green manure treatment tended to have greater concentrations than the fallow in the inorganic P pools. Narrow-leaf lupin depleted total inorganic P the most, even depleting some of the stable inorganic P pool. Assuming 10–15 cm soil depth, the narrow-leaf lupin treatment depleted total inorganic P equivalent to 32–47 kg P ha⁻¹, an amount comparable to or even greater than typical P fertiliser applications (MacDonald et al. 2011; Zou, Zhang, and Davidson 2022). This liberated P likely supported crop biomass production while also redistributing among other soil organic P pools. However, all green manures increased total organic P relative to fallow, likely due to rhizodeposition and biomass decomposition (Damon et al. 2014) with narrow-leaf lupin having the smallest increase (11 mg P kg⁻¹). In this soil, the net increases in organic P pools may be inconsequential as crop P uptake correlated well with depletion of the moderately labile inorganic P pool. Indeed, the moderately labile inorganic P pool was the largest fraction in this soil, much like for comparable New Zealand soils (Boitt et al. 2018a; Tian et al. 2017). Relatedly, all green manures increased the soil microbial biomass P pool (3–4 mg P kg⁻¹), which is beneficial: the turnover of microbial biomass P, and thus exchange with available inorganic P pools, can be as large as ~17 kg P ha⁻¹ (Achat et al. 2010; Frossard et al. 2000; Turner et al. 2003).

The mechanisms for soil P mobility under green manures are important for determining how this conservation practice may generalise to settings with different soil properties, climates, cash crops, and green manure species. This study suggests that, for soils with moderate to high P availability, well-selected green manures can benefit crops through the profound impacts on the chemical and biological cycling of soil P. Relative to fallow, the cereal green manure decreased soil pH while the leguminous green manures increased soil pH. As soil pH decreases from circumneutral in non-calcareous soils, P solubility (largely controlled by Fe/Al compounds) also decreases (Goldberg and Sposito 1984; McDowell and Sharpley 2003). This dynamic may explain the good correlations between final soil pH, depletion in moderately labile inorganic P, and crop P uptake. When returned to soil, green manure biomass, especially that of legumes, may restore soil pH during decomposition by supplying alkalinity (Vanzolini et al. 2017; Yan, Schubert, and Mengel 1996).

As for effects on biological P cycling, the legume green manures increased both acid and alkaline phosphatase enzyme activities. The total increases here, 1.39–1.94 mmol kg⁻¹ h⁻¹ under legume green manures, were even greater than that observed by Hallama et al. (2021) and Chavarría et al. (2016), and may also contribute to the greater crop P uptake by mobilising some organic P pools (Clarholm, Skyllberg, and Rosling 2015). Another major biological effect of green manures on soil P cycling is likely due to their rhizodeposits. On pumice and volcanic ash soils in New Zealand, Nguyen, McDowell and Condron (2024) found that narrow-leaf lupin mobilised more moderately labile inorganic P and stable inorganic P in the rhizosphere than did pea. The depletion of these P pools positively correlated to malate exudation rates across several green manure species, with narrow-leaf lupin in particular releasing relatively more malate (Nguyen, McDowell, and Condron 2024). Indeed, narrow-leaf lupin has longer root hairs (> 1 mm) than for example, pea, as we visually observed (Figure 6), which supports greater root areas and thus, potentially, more soil P mobilisation (Chen et al. 2002; Jungk 1987). Additionally, narrow-leaf lupin may release more organic anions such as citrate and malate (Egle, Römer, and Keller 2003), thus mobilising more of the less-labile soil P than either pea or cereal green manures (Nuruzzaman et al. 2005). In our study, narrow-leaf lupin produced the greatest biomass, followed by pea and cereal, thus supplying more available carbon for microbial activity and therefore P cycling.

[IMAGE OMITTED. SEE PDF]

Improvements in crop P fertility under green manures may be more pronounced for soils with lower P availability (Hallama et al. 2019). However, mobilising P in soils with moderate to high P availability, as was demonstrated here via leguminous green manures, is important to the sustainability of agricultural P. Globally, significant areas of cropland currently contain more available P (e.g., as Olsen P) than what is agronomically optimal (McDowell, Pletnyakov, and Haygarth 2024). These areas are often emblematic of legacy P (Sharpley et al. 2013), where strategies are needed to improve accessibility of soil P to relieve P input requirements and simultaneously lower risk of P pollution (Rowe et al. 2016). We argue that green manures are a viable conservation practice for mobilising labile as well as less labile soil P, effectively lowering the threshold needed for optimum P fertility. This benefit for sustainable P cycling in croplands complements the other benefits of green manures, or cover-cropping, such as biological N fixation (for legumes), protection from erosion, and improvements in soil organic matter and structure (Basche and DeLonge 2017; Hanrahan et al. 2021; Wittwer et al. 2017).

Conclusions

In this study, the inclusion of narrow-leaf lupin or pea as green manure crops increased cereal crop biomass, P uptake, and N uptake over two rotations compared with a fallow treatment as well as a cereal green manure treatment. This was attributed to a combination of factors including enhanced mobilisation and acquisition of moderately labile forms of soil P and increased biological cycling of P in soil. We conclude that the inclusion of a legume green manure may be able to sustain the yield of following crops while supplanting the need to fertilise in the short term. This effect should be investigated in the field over longer experimental duration to observe whether it sustains and how soil P cycling may shift in the long-term. Green manures may be a beneficial practice for drawing down legacy P for soils with moderate P availability, thus potentially providing water quality benefits while capitalising on agronomic opportunity.

Author Contributions

P. V. Nguyen: conceptualisation, data curation, formal analysis, investigation, methodology, validation, project administration, software, writing–original draft, writing–review and editing. L. M. Condron: conceptualisation, project administration, supervision, methodology, funding acquisition, resources, writing–review and editing. Z. P. Simpson: formal analysis, methodology, software, visualisation, writing–original draft, writing–review and editing. R. W. McDowell: conceptualisation, supervision, writing–review and editing.

Acknowledgements

This work was supported by Lincoln University. PVN was supported by a PhD scholarship from Ministry of Agriculture and Rural Development of the Socialist Republic of Viet Nam. RWM was supported by the Our land and Water National Science Challenge, funded by the New Zealand Ministry of Business, Innovation and Employment (contract C10X1507). Open access publishing facilitated by Lincoln University, as part of the Wiley - Lincoln University agreement via the Council of Australian University Librarians.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Achat, D. L., C. Morel, M. R. Bakker, et al. 2010. “Assessing Turnover of Microbial Biomass Phosphorus: Combination of an Isotopic Dilution Method With a Mass Balance Model.” Soil Biology and Biochemistry 42: 2231–2240. [DOI: https://dx.doi.org/10.1016/j.soilbio.2010.08.023].

Zou, T., X. Zhang, and E. A. Davidson. 2022. “Global Trends of Cropland Phosphorus Use and Sustainability Challenges.” Nature 611: 81–87.