Low 15N Natural Abundance in Shoot Tissue of Brachiaria humidicola Is an Indicator of Reduced N Losses Due to Biological Nitrification Inhibition (BNI)

The tropical forage grass Brachiaria humidicola (Bh) suppresses the activity of soil nitrifiers through biological nitrification inhibition (BNI). As a result, nitrate (${\text{NO}}_3^{-}$) formation and leaching are reduced which is also expected to tighten the soil nitrogen (N) cycle. However, the beneficial relationship between reduced ${\text{NO}}_3^{-}$ losses and enhanced N uptake due to BNI has not been experimentally demonstrated yet. Nitrification discriminates against the ¹⁵N isotope and leads to ¹⁵N depleted ${\text{NO}}_3^{-}$, but ¹⁵N enriched ${\text{NH}}_4^{+}$ in soils. Leaching of ¹⁵N depleted ${\text{NO}}_3^{-}$ enriches the residual N pool in the soil with ¹⁵N. We hypothesized that altered nitrification and ${\text{NO}}_3^{-}$ leaching due to diverging BNI magnitudes in contrasting Bh genotypes influence soil ¹⁵N natural abundance (δ¹⁵N), which in turn is reflected in distinct δ¹⁵N in Bh shoot biomass. Consequently, high BNI was expected to be reflected in low plant δ¹⁵N of Bh. It was our objective to investigate under controlled conditions the link between shoot value of δ¹⁵N in several Bh genotypes and leached ${\text{NO}}_3^{-}$ amounts and shoot N uptake. Additionally, plant ¹⁵N and N% was monitored among a wide range of Bh genotypes with contrasting BNI potentials in field plots for 3 years. We measured leaf δ¹⁵N of young leaves (regrown after cutback) of Bh and combined it with nitrification rates (NRs) of incubated soil to test whether there is a direct relationship between plant δ¹⁵N and BNI. Increased leached ${\text{NO}}_3^{-}$ was positively correlated with higher δ¹⁵N in Bh, whereas the correlation between shoot N uptake and shoot δ¹⁵N was inverse. Field cultivation of a wide range of Bh genotypes over 3 years decreased NRs in incubated soil, while shoot δ¹⁵N declined and shoot N% increased over time. Leaf δ¹⁵N of Bh genotypes correlated positively with NRs of incubated soil. It was concluded that decreasing plant δ¹⁵N of Bh genotypes over time reflects the long-term effect of BNI as linked to lower ${\text{NO}}_3^{-}$ formation and reduced ${\text{NO}}_3^{-}$ leaching. Accordingly, a low δ¹⁵N in Bh shoot tissue verified its potential as indicator of high BNI activity of Bh genotypes.

Introduction

Biological nitrification inhibition (BNI) by the tropical forage grass Brachiaria humidicola (Sylvester-Bradley et al., 1988; Subbarao et al., 2007) is an ecologically evolved trait to compete with nitrifying soil organisms for available ammonium (${\text{NH}}_4^{+}$). Certain root derived exudates (Coskun et al., 2017a), e.g., brachialactone, have been verified to block the activity of ammonia monooxygenase (AMO) and hydroxylamine oxidoreductase (HAO) pathways in nitrifiers such as Nitrosomonas europaea (Subbarao et al., 2009). By preventing the microbial transformation of less soil mobile ${\text{NH}}_4^{+}$ to more soil mobile nitrate (${\text{NO}}_3^{-}$), BNI is expected to reduce leaching of nitrogen (N) from ecosystems (Subbarao et al., 2009, 2013; Coskun et al., 2017b). Recent research has been undertaken to investigate the BNI effect by Bh on N dynamics, for example on the reduction of nitrous oxide (N₂O) emissions from soils (Byrnes et al., 2017), or on the influence of residual BNI effect on N uptake of subsequent crops (Karwat et al., 2017). However, indicators for reduced ${\text{NO}}_3^{-}$ leaching losses by effective BNI do not exist yet. The reduction of ${\text{NO}}_3^{-}$ leaching from agroecosystems due to BNI is one of the claimed central features in terms of BNI implementation (Subbarao et al., 2013; Coskun et al., 2017b).

Nitrification and N leaching are two main soil processes that lead to ¹⁵N:¹⁴N isotope fractionation (Mariotti et al., 1981). This results in a ¹⁵N enriched ${\text{NH}}_4^{+}$ and a ¹⁵N depleted ${\text{NO}}_3^{-}$ pool in soils (Delwiche and Steyn, 1970; Herman and Rundel, 1989; Jones and Dalal, 2017). Under high rainfall conditions, nitrification-derived ¹⁵N-depleted ${\text{NO}}_3^{-}$ is exposed to high leaching, whereby plants then would feed on remaining ¹⁵N enriched ${\text{NH}}_4^{+}$. In contrast, plants with effective BNI are expected to suppress microbial ${\text{NO}}_3^{-}$ formation. Consequently, less ¹⁵N depleted ${\text{NO}}_3^{-}$ would be lost from the system. Under both scenarios, Bh feeds mainly on ${\text{NH}}_4^{+}$, toward which it has a higher tolerance than other Brachiaria species (Castilla and Jackson, 1991; Rao et al., 1996). The ¹⁵N signature of soil mineral N can be reflected in a corresponding shoot ¹⁵N signature (Takebayashi et al., 2010). Previous field studies demonstrated that foliar δ¹⁵N increased in conjunction with increasing nitrification and N leaching by precipitation (Pardo et al., 2002; Stamatiadis et al., 2006; Huber et al., 2013; Yé et al., 2015). From these studies, the relationship between plant δ¹⁵N and soil nitrification and N losses was apparent, which indicates a possible strong relationship between BNI and the δ¹⁵N in plant tissue.

What remains unclear is if shoot δ¹⁵N of Bh genotypes is linked to plant induced BNI, and if this relates to reduced nitrification, thus reduced ${\text{NO}}_3^{-}$ leaching. We therefore hypothesized that (i) effective BNI is linked to reduced soil nitrification rates (NRs), enhanced N nutrition, reduced ${\text{NO}}_3^{-}$ leaching; and that (ii) this link is expressed in respective low δ¹⁵N in plant shoot tissue due to nutrition of Bh on a less naturally enriched N pool of ¹⁵N. In contrast, it was expected that, respectively, higher NRs and higher ${\text{NO}}_3^{-}$ leaching is expressed in higher δ¹⁵N of shoot biomass due to the uptake of ¹⁵N enriched soil N.

Materials and Methods

Experiment 1: Relationship Between δ¹⁵N of Bh, N Uptake, and ${\text{NO}}_3^{-}$ Leaching Under Controlled Conditions

A greenhouse study at the University of Hohenheim (UHOH), Stuttgart, Germany, was implemented as an α-design, i.e., a design with incomplete blocks, that are grouped into complete replicates. The trial contained four complete replicates with two blocks per replicate. The aim was to monitor the effect of N loss and N uptake after N fertilization on the δ¹⁵N in shoot biomass of different Bh genotypes. Experimental soil columns were manufactured from PVC-drainpipes (Ø 11 cm × 100 cm) in order to enable deep rooting of Bh and monitoring of ${\text{NO}}_3^{-}$ dislocation within the soil profile. A fossil tertiary clay loam (collected in Lich, Germany, 50°31′2.0”N, 8°50′55.9”E) with pH 5.7, 36% clay content, 0.25% total C, and 0.025% total N was used. The original soil was amended with sand (25 vol%) in order to improve drainage, and the resulting substrate was filled into the experimental pipes.

Five apomictic Bh hybrids (Bh08 selection: Bh08-1149, Bh08-700, Bh08-675, Bh08-696, Bh08-1253) and two germplasm accessions of Bh (CIAT 16888, CIAT 679 cv. Tully) were used as test genotypes (Rao et al., 2014). Bh stolons were first propagated from a Bh stock collection at UHOH and transferred to a turf-based culture substrate for 3 weeks for root establishment. Then, the young Bh plants were transplanted in August 2014 to the experimental pipes were exposed to supplementary artificial light (photosynthetically active radiation averaging 800 μmol m⁻² s⁻¹) for 12 h photoperiod. Day and night temperature regimes were adjusted to 25 and 20°C, respectively. Plants were raised with a basal fertilizer of N-P-K-S (analog to 30-60-150-35 kg ha⁻¹). After the establishment phase of 6 weeks, plants were cut to 2–3 cm height to facilitate regrowth. The experimental phase commenced with application of 150 kg N-${\text{NH}}_4^{+}$ ha⁻¹ as (NH₄)₂SO₄ (δ¹⁵N = −0.1) to stimulate the growth and activity of soil nitrifiers. Plants were watered every second day (with 100 ml H₂O), whereas the amount of water applied was doubled at 8, 19, and 28 days after N fertilization (DANF) in order to increase leaching of ${\text{NO}}_3^{-}$ derived from nitrification.

Rhizons (Soil Moisture Sampler, Ø 2.3 × 50 mm, porosity 0.1 μm, Eijkelkamp Agrisearch Equipment, Giesbeek, Netherlands) were installed horizontally into the pipes at 7.5 and 50 cm depth. This procedure enabled sampling of soil solution by suction pressure using a common medical syringe. Soil solution sampling (0, 3, 6, 7, 10, 14, and 20 DANF from topsoil and at 0, 18, and 27 DANF from 50 cm depth) was conducted at 3 h after irrigation with 100 ml of tap water to warrant sufficient soil moisture and time for establishing equilibrium for ${\text{NO}}_3^{-}$ level in the topsoil. A sample of 10 ml of soil solution was collected and frozen immediately. Concentration of ${\text{NO}}_3^{-}$ in soil solution samples was analyzed by complete reduction to ${\text{NO}}_2^{-}$ through hydrazine (in alkaline solution, with copper as catalyst) and subsequent reaction with sulfanilamide and Griess reagent (N-(1-Naphthyl)ethylenediamine) to form a pink compound measured photometrically at 550 nm (method DIN38405, ISO/DIS 13395; photometer: AutoAnalyzer 3, QuAAtro AQ2, SEAL Analytical Ltd., Southampton, United Kingdom).

Resin bags (ion-exchange resin in a fine nylon mesh) were installed at the bottom of the experimental pipe to trap leached ${\text{NO}}_3^{-}$. This allowed the quantification of cumulative ${\text{NO}}_3^{-}$-losses during the experiment. For this purpose, Resinex MX-11 (Jacobi Carbons GmbH, Frankfurt, Germany), a mixed anion/cation exchange resin with a maximum anion sorption capacity of 0.4 eq l⁻¹ was used. Each bag contained 200 ml resin amended with 200 ml of washed sand in order to slow down percolate passage through the coarse resin matrix. The anion traps were removed at the end of the experiment (i.e., at 42 DANF). The resin was thoroughly homogenized and an aliquot of 40 ml resin of each bag was further processed. Extraction of ${\text{NO}}_3^{-}$ was performed twice with 200 ml of 2 M KCL and extracts were subsequently analyzed photometrically for ${\text{NO}}_3^{-}$ content (see analysis of percolate samples). A pre-test on the extractability of Resinex MX-11 verified that a twofold extraction was sufficient to achieve extraction rates of 98% of trapped ${\text{NO}}_3^{-}$.

At 42 DANF, soil solution sampling showed ${\text{NO}}_3^{-}$ levels below the detection limit using a fast ${\text{NO}}_3^{-}$ test (<5 mg ${\text{NO}}_3^{-}$ L⁻¹). This indicated that the main effect of N fertilization should be reflected in terms of the δ¹⁵N and N uptake amount (mg N g plant dry matter⁻¹). The aboveground shoot biomass (cut 3 cm above soil surface) of all seven Bh genotypes was sampled subsequently and oven-dried (3 days 60°C). Dry matter was determined and ground aliquots of plant material were filled into tin capsules (HEKAtech GmbH, Wegberg, Germany).

¹⁵N and N% were measured for all plant samples from Experiments 1–3 at UHOH by using a Euro Elemental analyzer coupled to a Finnigan Delta continuous-flow isotope ratio mass spectrometer IRMS (Thermo Scientific, Bremen, Germany). The ¹⁵N natural abundance of the sample relative to the standard (atmospheric N₂) was expressed as: δ¹⁵N‰ = [(R_sample/R_standard) − 1] × 1000 (‰), where R represents the isotope ratio (¹⁵N/¹⁴N) and R_standard is ¹⁵N/¹⁴N for atmospheric N² that is 0.0036765 (δ¹⁵N‰ = 0) (Robinson, 2001).

Experiment 2: Differences in δ¹⁵N Leaf and Shoot of a Wide Range of Bh Genotypes Grown Under Field Conditions

A field trial was established in August 2013 with a range of Bh genotypes with contrasting BNI activity at Corpoica-La Libertad Research Center in the Piedmont region of Colombia (4°03′46” N, 73°27′47” W). The experimental field site was located at an altitude of 338 m above sea level with an annual mean temperature of 21.4°C and an average annual rainfall of 3,685 mm. The soil is classified as an Oxisol with a pH of 4.9, clay content of 42%, total N of 0.11%, and C/N ratio of 12.4. The trial included four out of the five Bh apomictic hybrids used in Experiment 1 and four Bh germplasm accessions. Each main plot had a size of 4 × 4 (16 m²) and received basal fertilization (kg ha⁻¹) in September 2013 of 69 N (urea, δ¹⁵N = 0.05), 25 P, 50 K, 50 Ca, 15 Mg, 10 S, 0.5 B, and 2.6 Zn. The second N (100 kg N ha⁻¹ as urea) fertilization (26 October 2015) was conducted after leaf sampling for Experiment 3 to stimulate both, BNI and soil nitrification.

For forage productivity and forage quality evaluation, grasses were cut every 6 weeks from October 2013 to November 2015. Aboveground shoot biomass samples of each plot were oven dried at 60°C. A sub-sample was ground and sent to UHOH for IRMS analysis (see above in Experiment 1).

Experiment 3: δ¹⁵N of Young Regrown Bh Leaves Linked to BNI Indicated by Soil Incubation

Two years after initiation of the field trial described for Experiment 2, a substantial BNI effect was expected based on preliminary tests (soil incubation, data not shown). The grass was cut on 5 October 2015 (end of the rainy season) and all plots received maintenance fertilization (kg ha⁻¹) of 40 P, 75 K, 110 Ca, 65 Mg, 19 S, 0.9 B, and 5.3 Zn. No N fertilizer was applied. It was intended that the plants take up soil mineral N and consequently reflect the δ¹⁵N of the soil mineral N. To avoid border effects during sampling, one sub-plot (1 m²) per main plot was randomly defined and marked. At 11 days after grass cutting (16 October), recently fully expended (second youngest) leaves from the regrown plants within the sub-plots were collected from the plots of three CIAT accessions as well as from Bh08 hybrids. Oven-dried and ground shoot samples were sent to UHOH for ¹⁵N and N% measurement (see above in Experiment 1).

Before cutting the grass in the field trial (Experiment 3) and adding the fertilizer in October 2015, topsoil samples (0–10 cm) were collected with an auger from eight randomly chosen points within each sub-plot. About 500 g of soil per sub-plot was thoroughly mixed, air-dried for 48 h, sieved (2 mm mesh size), and small stones as well as visible root material were removed. Representative sub-samples of 5 g of soil from each plot were filled in small glass tubes followed by application of 1.5 ml ammonium sulfate ((NH₄)₂SO₄) solution as substrate for soil nitrifiers. Tubes were sealed with parafilm that contained two holes for aeration and placed in a dark incubation chamber with constant 25°C and 60% air humidity. Soil ${\text{NO}}_3^{-}$ was extracted before starting the incubation (basal) and after 5 and 25 days (main active phase of nitrification) with 50 ml of 1 M KCl. ${\text{NO}}_3^{-}$ concentrations were determined as described above (Experiment 1). NRs were expressed as mg N-${\text{NO}}_3^{-}$ kg dry soil⁻¹ day⁻¹.

Statistical Analysis

Statistical analysis was performed using the SAS^® version 9.4 (SAS Institute Inc., Cary, NC, United States) with a mixed model approach. The assumptions of homogeneity of variance and normal distribution of residual errors were tested through the plots of studentized residuals vs. predicted value and quantile-quantile-plots, respectively, for all data sets. For the analysis of data from Experiment 2, the following model was set up: Genotype + Year + Rep + Genotype × Year. Block was set as random factor. Model-based least square means were used for data visualization with SigmaPlot for Windows version 12.0 (Copyright Systat Software, Inc.). The same software was also used for correlation analysis conducted for Experiments 1 and 3.

Results

Experiment 1 (Greenhouse at UHOH, Germany)

Plant δ¹⁵N at harvest (42 DANF) was found to be negatively correlated with plant N uptake amounts (p < 0.001) (Figure 1A). Furthermore, the relationship between plant δ¹⁵N and the amount of cumulative leached ${\text{NO}}_3^{-}$ (p = 0.018; R² = 0.16) was moderately positive (Figure 1B). In the case of low N uptake, plant δ¹⁵N was high, whereas plants with high N uptake were found to have lower plant δ¹⁵N (Figure 1). Increased leaching of ${\text{NO}}_3^{-}$-N lead to an increase of the δ¹⁵N of remaining N accumulating in the grass shoot biomass (Table 1). Furthermore, plant δ¹⁵N of all samples was relatively enriched compared to the δ¹⁵N of the applied ${\text{NH}}_4^{+}$ fertilizer (δ¹⁵N = −0.1). A linear regression analysis showed an inverse relationship (p < 0.001; R² = 0.37) between plant N uptake and leached N (Supplementary Material).

FIGURE 1

FIGURE 1. Linear regression of plant δ¹⁵N (‰) of Brachiaria humidicola (Bh) and plant N uptake (mg N pot⁻¹) (A) and cumulative leached ${\text{NO}}_3^{-}$ (mg N pot⁻¹) (B). The greenhouse study (Experiment 1) included seven Bh genotypes and was established in August 2014. Plants were sampled at 42 days after fertilization (6 weeks after transplanting to experimental pots) with (${\text{NH}}_4^{+}$)₂SO₄ (δ¹⁵N = –0.1‰). Plant N uptake and leached ${\text{NO}}_3^{-}$ are cumulative amounts determined at the date of harvest.

TABLE 1

TABLE 1. Plant δ¹⁵N (‰), plant N uptake (mg N pot⁻¹), and leached N (mg N pot⁻¹) at 42 days after N [(${\text{NH}}_4^{+}$)₂SO₄; δ¹⁵N = −0.1‰] fertilization of two Brachiaria humidicola (Bh) CIAT accessions and five Bh08 hybrids.

Experiment 2: Yearly δ¹⁵N Monitoring of Bh Genotypes in the Field (Field Trial La Libertad, Colombia)

The overall year effect showed an obvious trend of decreasing plant δ¹⁵N over the experimental seasons (p < 0.0001) (Figure 2). One month after experiment establishment and N fertilization (October 2013), most genotypes tended to a plant δ¹⁵N of around 7‰, except for the Bh08-1149 hybrid with a δ¹⁵N of almost 8‰. One year later (October 2014), δ¹⁵N of all genotypes had dropped below 5‰, however, a genotypic effect on δ¹⁵N abundance was absent (p = 0.13). At the last sampling (November 2015), δ¹⁵N of CIAT 26149 and CIAT 26146 had higher δ¹⁵N than the other three CIAT accessions and the two Bh08 hybrids (p = 0.02).

FIGURE 2

FIGURE 2. Plant δ¹⁵N (‰) of B. humidicola (Bh) of five contrasting (in terms of BNI) CIAT accessions and two hybrids (Bh08-population) with unknown BNI capacity sampled after the rainy season in the Colombian Llanos (Experiment 2). The field trail was established in August 2013. N fertilizer was applied as urea (δ¹⁵N = 0.05‰) in September 2013 (69 kg N ha⁻¹) and in October 2015 (100 kg N ha⁻¹). Error bars represent the standard error of the mean based on collected plants of three replicated completely randomized field plots. Means that share a common letter within the same year do not differ significantly at the a = 5% level.

To investigate an expected relationship between δ¹⁵N and N uptake by Bh, a regression analysis was conducted between the measured plant δ¹⁵N and the N concentration (N%) in the respective sampled Bh grass genotypes (Figure 3). A negative correlation (p < 0.001) was observed between plant δ¹⁵N and plant N%, indicating that the higher the N status of the plant is the lower the respective δ¹⁵N becomes.

FIGURE 3

FIGURE 3. Linear regression of plant δ¹⁵N (‰) and plant N concentration (%) of contrasting B. humidicola (Bh) genotypes (in terms of BNI) sampled after the rainy season in the Colombian Llanos (Experiment 2). The field trail was established in August 2013. N fertilizer was applied as urea (δ¹⁵N = 0.05‰) in September 2013 (69 kg N ha⁻¹) and in October 2015 (100 kg N ha⁻¹).

Experiment 3: Genotypic Leaf δ¹⁵N and Relation to Soil Nitrification (Field Trial La Libertad, Colombia)

Means of δ¹⁵N values of regrown leaves at 11 days after regrowth (October 2015) of the selected Bh genotypes were positively correlated with the observed NRs of the respective Bh genotypes (p = 0.007) (Figure 4). Leaf δ¹⁵N means were found to be different among genotypes (p = 0.001), whereas the genotype effect was not significant for NRs (p = 0.74) (Table 2). In more detail, CIAT 26146 had highest δ¹⁵N leaf signal and the corresponding incubated soil showed highest NR. CIAT 16888 and CIAT 679, with reported high BNI, had, compared to CIAT 26146, significantly lower ¹⁵N leaf signals and were among the genotypes tested those with lower NRs.

FIGURE 4

FIGURE 4. Linear regression of leaf δ¹⁵N (‰) of B. humidicola (Bh) and NRs (mg N-${\text{NO}}_3^{-}$ kg dry soil⁻¹) determined from incubated soil sampled from the respective field plots (Experiment 3). The field trial included Bh CIAT accessions with potential of high BNI (CIAT 16888), mid-high BNI (CIAT 679), and low BNI (CIAT 26146). Additionally, three hybrids (Bh08-population) with unknown BNI capacity were sampled. The field trail was established in August 2013 in the Colombian Llanos. N fertilizer was applied as urea (δ¹⁵N = 0.05‰) in September 2013 (69 kg N ha⁻¹). Plants were cut back at 5 October 2015 and topsoil (0–10 cm) samples were collected for incubation in the laboratory. Last fully developed leaves of Bh were sampled at 11 days after regrowth. Error bars represent the standard error of the mean based on collected leaf and soil samples of three replicated completely randomized field plots.

TABLE 2

TABLE 2. Nitrification rates (NRs) determined from incubated topsoil (0–20 cm, Oxisol, Colombian Llanos) sampled from field plots (October 2015) of three B. humidicola (Bh) CIAT accessions and three Bh hybrids (Experiment 3).

Discussion

Lower Plant δ¹⁵N Is Linked to Reduced ${\text{NO}}_3^{-}$ Losses and High Plant N Uptake

Our results of the greenhouse study confirmed the hypothesis that low plant δ¹⁵N is linked to enhanced N nutrition and reduced ${\text{NO}}_3^{-}$ leaching. Hence, the altered plant isotopic signals were indirectly linked to isotopic fractionation of ¹⁵N:¹⁴N between product (${\text{NO}}_3^{-}$) and substrate (${\text{NH}}_4^{+}$) during bacterial and archaeal nitrification (Delwiche and Steyn, 1970; Shearer et al., 1974). This led to subsequent leaching of the relatively ¹⁵N depleted ${\text{NO}}_3^{-}$. Our observations (Experiment 1) indicated that with increasing amounts of leached ${\text{NO}}_3^{-}$, the δ¹⁵N of the Bh grass increased (Figure 1B). These results suggest that losses of ¹⁵N depleted ${\text{NO}}_3^{-}$ resulted in relative enrichment of the plant available N. Similar observations of leaching processes leading to an ¹⁵N enrichment of the remaining soil ${\text{NH}}_4^{+}$ have been made by Pardo et al. (2007), Craine et al. (2009), and Stevenson et al. (2010). Furthermore, our observation of the positive relationship between increasing ¹⁵N:¹⁴N in vegetation due to increasing nitrification and N losses has also been described for leaves of forest trees (Pardo et al., 2002; Garten et al., 2008), cotton (Stamatiadis et al., 2006), mixed systems such as grass-heath-woodlands (Huber et al., 2013), as well as for comparative studies between perennial and annual grasses (Yé et al., 2015). Furthermore, evaluations at various sites demonstrated lower δ¹⁵N of ${\text{NO}}_3^{-}$ compared to δ¹⁵N of ${\text{NH}}_4^{+}$ in soil due to nitrification (Takebayashi et al., 2010). The yearly field evaluation revealed that ¹⁵N signals in Bh decreased with improved N status of the grass over time, indicating increased utilization of plant N as also observed in Central European grasslands by Kleinebecker et al. (2014). We suggest that in our field trial this was caused by decreasing ${\text{NO}}_3^{-}$ losses from the system due to expanded plant (e.g., root system) growth and development over the years.

Long-Term BNI Effect in the Field Expressed in Low Plant δ¹⁵N

The general trend of decreasing plant δ¹⁵N of Bh genotypes in the field study over the years indicated that N isotope fractionation and consequently the δ¹⁵N of the mineral N in the soil changed over time. Thus, our results confirmed that during the early stage after establishment of the Bh genotypes, there was no significant influence of BNI on microbial nitrification. This observation is consistent with other studies on BNI expression with young Bh plants (Miranda et al., 1994; Castoldi et al., 2013; O’Sullivan et al., 2016). Therefore, it is suggested that applied urea N (hydrolyzed to ${\text{NH}}_4^{+}$ within a few days) during transplanting had a strong stimulation effect on growth and activity of soil nitrifiers, and that the microbial formed ${\text{NO}}_3^{-}$ was leached rapidly in the first rainy season when the grass was still small. BNI was unlikely to be strongly expressed, since plants were less than 2 months old and a strong BNI effect, due to an accumulation of BNI substances in the soil, needs about 1 year of Bh establishment (Nuñez et al., 2018). The high δ¹⁵N of Bh plants observed in October 2013 could therefore reflect the strong ¹⁵N enrichment of the soil mineral N pool caused by substantial nitrification and leaching loss of ¹⁵N depleted ${\text{NO}}_3^{-}$ during the establishment of the trial (Nadelhoffer and Fry, 1994; Song et al., 2014). Plant δ¹⁵N after the second rainy season (October 2014) were lower than the first sampling. BNI ability has been shown to be promoted with developing root biomass leading to less nitrification over time (Subbarao et al., 2009). Reduced N losses during the second rainy season explained the lower plant δ¹⁵N caused by a lower ¹⁵N enrichment of soil mineral N pool. The general tendency of decreasing ¹⁵N natural abundance of Bh genotypes over the years was also visible in the third year. This indicated further reduction of ${\text{NO}}_3^{-}$ formation and loss due to increase of BNI. This was verified by other soil incubation studies (Arango et al., unpublished data) revealing low NRs during the second (2014) rainy season. In contrast, incubation of soil sampled during the third rainy season (2015) evidenced significantly lower average NRs (3.5 mg N-${\text{NO}}_3^{-}$ kg dry soil⁻¹ day⁻¹). However, since root systems expand over the years a general higher uptake of ${\text{NH}}_4^{+}$ is expected. This could, additionally to BNI substance release, increase the competiveness of Bh for ${\text{NH}}_4^{+}$ and indirectly reduce nitrifier activity.

Link Between High BNI and Low Leaf δ¹⁵N of Bh Genotypes

In our study, we linked leaf δ¹⁵N to BNI by Bh. It was evident that low nitrification in incubated soil taken from plots, where Bh was cultivated for more than 2 years, correlated with lower leaf δ¹⁵N. Exudation of brachialactone by Bh and other known nitrification inhibiting substances (Gopalakrishnan et al., 2009; Subbarao et al., 2009) are supposed to increase the relative ${\text{NH}}_4^{+}$-to-${\text{NO}}_3^{-}$ uptake, thereby reflecting primarily the δ¹⁵N signal of the plant available soil ${\text{NH}}_4^{+}$ pool (Kahmen et al., 2008). Furthermore, Bh root exudates have been shown to reduce N. europaea populations in soil (Gopalakrishnan et al., 2009). Our results thus support the hypothesis that high BNI (low NRs) results in lower leaf δ¹⁵N, as observed in our field study. Robinson (2001) suggested to measure whole plant δ¹⁵N when studies intend to indicate source δ¹⁵N in plant tissues to avoid uncertainties in isotopic discrimination during partitioning in the plant. In this respect, however, it has to be considered that intra-plant ¹⁵N discrimination (e.g., root-to-shoot) is generally small when N availability is low or when ${\text{NH}}_4^{+}$ is the primary mineral N form taken up by the plant (Evans, 2001).

Different Plant δ¹⁵N Among Contrasting Bh Genotypes in Terms of BNI

Biological nitrification inhibition differences among Bh accessions or hybrids have been revealed (Subbarao et al., 2007; Rao et al., 2014; Nuñez et al., 2018) and the effect of high BNI is expected to reduce ${\text{NO}}_3^{-}$ formation and leaching (Subbarao et al., 2009, 2013). But experimental evidence for the latter is lacking. We observed a strong genotypic effect on leaf δ¹⁵N in our field studies. For instance, Bh genotypes with low NRs in the field (CIAT 16888, CIAT 679, CIAT 26159) showed a strong BNI effect from their root exudates on N. europaea (Subbarao et al., 2007). At the end of the field study, the same genotypes showed lower δ¹⁵N than CIAT 26149 with known low BNI potential (Subbarao et al., 2007). Furthermore, the higher leaf δ¹⁵N of CIAT 26146 compared to CIAT 16888 and CIAT 679 fit to our hypotheses, and earlier BNI evaluations (Subbarao et al., 2009; Nuñez et al., 2018). However, apart from BNI, other factors could have altered shoot δ¹⁵N of Bh genotypes. Exemplary is the acknowledged symbiosis of mycorrhizae with plants (Evans, 2001). However, under low N availability the cycling of N through the fungus to the plant is rather negligible for plant δ¹⁵N (Högberg et al., 1999).

Conclusion

We studied the interlinkages of plant δ¹⁵N, BNI, microbial nitrification, N uptake, and N leaching losses under controlled as well as under field conditions based on a selection of contrasting Bh genotypes. Our main conclusion is that high BNI activity decreases plant δ¹⁵N of Bh. Thus, the ¹⁵N natural abundance of grass tissue might be linked to BNI activity in soil, suppressing the growth and activity of bacterial and archaeal nitrifiers. As a result, this led to enhanced ${\text{NH}}_4^{+}$ uptake by Bh and reduced ${\text{NO}}_3^{-}$ losses. This ecological concept is enhanced if: (i) BNI is expressed in soil due to long-term presence of Bh; (ii) there is a continuous substrate (${\text{NH}}_4^{+}$) supply (mineralization, fertilization) so that the source N for plant uptake is never converted completely (into ${\text{NO}}_3^{-}$); and (iii) a significant amount of ${\text{NO}}_3^{-}$ formed by nitrification is leached of the rooting zone of the plants. Since Bh can take up both N forms (Castilla and Jackson, 1991), the cumulative ¹⁵N shoot signal would be confounded in case of ${\text{NH}}_4^{+}$ and ${\text{NO}}_3^{-}$ uptake without nitrate loss. Furthermore, other microbial enzymatic reactions should not mask the discrimination process by nitrifiers, such as: (i) volatilization (driven by high soil pH, heat, not incorporated N fertilizer); (ii) denitrification (anaerobic conditions, high C availability, ${\text{NO}}_3^{-}$ substrate present). However, nitrate substrate left would even be higher enriched in ¹⁵N; and (iii) uptake of ¹⁵N depleted N derived by free living N fixing bacteria.

Our observations suggest that high BNI along with reduced microbial nitrification (one of the main reactions causing ¹⁵N:¹⁴N fractionation) and N leaching (enriching the remaining soil mineral N with ¹⁵N) are reflected in low δ¹⁵N leaf or shoot biomass signals under environments with high ${\text{NO}}_3^{-}$ leaching potential. We also suggest that the method described here can serve as an indicator of the extent of ${\text{NO}}_3^{-}$ leakiness for BNI field evaluations over the years, if combined with other BNI indicators like the abundance and activity of soil nitrifiers under the given conditions set out above.

Author Contributions

HK wrote the manuscript and had the overall task to modify it according to suggestions and corrections of the co-authors. Furthermore, HK measured all samples with the IRMS that have been used for this study. KE installed the greenhouse trial (Experiment 1) and conducted the sampling. JN assisted in sampling of plant material (Experiment 3) and delivered further plant material from other experiments for pre-tests (data not shown). IR was the leading senior scientist of the BMZ project at CIAT Colombia. FR contributed to the scientific interpretation and the concept of the study. JA was the leading young professional of the BMZ project at CIAT Colombia and enabled HK the access to the field trial of Experiments 2 and 3 for sampling. DM had the responsibility of taking samples of Experiment 2. AA assisted in the incubation study of Experiment 3. GC was the leading senior scientist of this study. All authors contributed to manuscript revision, read, and approved the submitted version.

Funding

The authors very much appreciate the financial support of the project “Climate-smart crop-livestock systems for smallholders in the tropics: Integration of new forage hybrids to intensify agriculture and to mitigate climate change through regulation of nitrification in soil” by the Federal Ministry for Economic Cooperation and Development (BMZ, Germany) under the GIZ Project Number 11.7860.7-001.00. In addition, this study was undertaken as part of the LivestockPlus project funded by CGIAR Research Program (CRP) on Climate Change, Agriculture and Food Security (CCAFS), which is a strategic partnership of CGIAR and Future Earth.

Conflict of Interest Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

We would like to thank Marc-André Sparke for his cooperation during the sampling period in the greenhouse and in the field trials in Colombia. For the support with the IRMS we are grateful to Elke Dachtler, Carolin Röhl, and Christian Brandt. We would also like to thank the Livestock CRP and the Ministry of Forestry and Fisheries of Japan (MAFF) for their support that allowed maintaining the field trials for this study. We thank all donors that globally support the work of the CRP programs through their contributions to the CGIAR system.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb.2018.02383/full#supplementary-material

References

Byrnes, R. C., Nùñez, J., Arenas, L., Rao, I., Trujillo, C., Alvarez, C., et al. (2017). Biological nitrification inhibition by Brachiaria grasses mitigates soil nitrous oxide emissions from bovine urine patches. Soil Biol. Biochem. 107, 156–163. doi: 10.1016/j.soilbio.2016.12.029

CrossRef Full Text | Google Scholar

Castilla, C. E., and Jackson, W. A. (1991). “Nitrogen uptake mechanism in two Brachiaria species: Y-315,” in Tropsoils Technical Report 1988-1989, ed. T. P. McBride (Raleigh, NC: Tropsoils Management Entity), 164–167.

Google Scholar

Castoldi, G., Reis, J. G., Pivetta, L. A., and Rosolem, C. A. (2013). Soil nitrogen dynamics after Brachiaria desiccation. Rev. Bras de Ciênc Solo 37, 1620–1627. doi: 10.1590/S0100-06832013000600018

CrossRef Full Text | Google Scholar

Coskun, D., Britto, D. T., Shi, W., and Kronzucker, H. J. (2017a). How plant root exudates shape the nitrogen cycle. Trends Plant Sci. 8, 661–673. doi: 10.1016/j.tplants.2017.05.004

PubMed Abstract | CrossRef Full Text | Google Scholar

Coskun, D., Britto, D. T., Shi, W., and Kronzucker, H. J. (2017b). Nitrogen transformations in modern agriculture and the role of biological nitrification inhibition. Nat. Plants 3:17074. doi: 10.1038/nplants.2017.74

PubMed Abstract | CrossRef Full Text | Google Scholar

Craine, J. M., Elmore, A. J., Aidar, M. P., Bustamante, M., Dawson, T. E., Hobbie, E. A., et al. (2009). Global patterns of foliar nitrogen isotopes and their relationships with climate, mycorrhizal fungi, foliar nutrient concentrations, and nitrogen availability. New Phytol. 183, 980–992. doi: 10.1111/j.1469-8137.2009.02917.x

PubMed Abstract | CrossRef Full Text | Google Scholar

Delwiche, C. C., and Steyn, P. L. (1970). Nitrogen isotope fractionation in soils and microbial reactions. Environ. Sci. Technol. 4, 929–935. doi: 10.1021/es60046a004

CrossRef Full Text | Google Scholar

Evans, R. D. (2001). Physiological mechanisms influencing plant nitrogen isotope composition. Trends Plant Sci. 6, 121–126. doi: 10.1016/S1360-1385(01)01889-1

PubMed Abstract | CrossRef Full Text | Google Scholar

Garten, C. T., Paul, J. R., Hanson, J., Todd, D. E., Bonnie, J. R., Lu, B., et al. (2008). “Natural 15N- and 13C-abundance as indicators of forest nitrogen status and soil carbon dynamics,” in Stable Isotopes in Ecology and Environmental Science, eds R. Michener and K. Lajtha (Hoboken, NJ: John Wiley & Sons).

Google Scholar

Gopalakrishnan, S., Watanabe, T., Pearse, S. J., Ito, O., Hossain, Z. A., and Subbarao, G. V. (2009). Biological nitrification inhibition by Brachiaria humidicola roots varies with soil type and inhibits nitrifying bacteria, but not other major soil microorganisms. Soil Sci. Plant Nutr. 55, 725–733. doi: 10.1111/j.1747-0765.2009.00398.x

CrossRef Full Text | Google Scholar

Herman, D. J., and Rundel, P. (1989). Nitrogen isotope fractionation in burned and unburned chaparral soils. Soil Sci. Soc. Am. J. 53, 1229–1236. doi: 10.2136/sssaj1989.03615995005300040040x

CrossRef Full Text | Google Scholar

Högberg, P., Högberg, M. N., Quist, M. E., Ekblad, A. L. F., and Näsholm, T. (1999). Nitrogen isotope fractionation during nitrogen uptake by ectomycorrhizal and non-mycorrhizal Pinus sylvestris. New Phytol. 142, 569–576. doi: 10.1046/j.1469-8137.1999.00404.x

CrossRef Full Text | Google Scholar

Huber, E., Bell, T. L., and Adams, M. A. (2013). Combustion influences on natural abundance nitrogen isotope ratio in soil and plants following a wildfire in a sub-alpine ecosystem. Acta Oecol. 173, 1063–1074. doi: 10.1007/s00442-013-2665-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Jones, A. R., and Dalal, R. C. (2017). Enrichment of natural 15N abundance during soil N losses under 20 years of continuous cereal cropping. Sci. Tot. Environ. 574, 282–287. doi: 10.1016/j.scitotenv.2016.08.192

PubMed Abstract | CrossRef Full Text | Google Scholar

Kahmen, A., Wanek, W., and Buchmann, N. (2008). Foliar δ15N values characterize soil N cycling and reflect nitrate or ammonium preference of plants along a temperate grassland gradient. Acta Oecol. 156, 861–870. doi: 10.1007/s00442-008-1028-8

PubMed Abstract | CrossRef Full Text | Google Scholar

Karwat, H., Moreta, D., Arango, J., Núñez, J., Rao, I., Rincón, Á., et al. (2017). Residual effect of BNI by Brachiaria humidicola pasture on nitrogen recovery and grain yield of subsequent maize. Plant Soil 420, 389–406. doi: 10.1007/s11104-017-3381-z

CrossRef Full Text | Google Scholar

Kleinebecker, T., Hölzel, N., Prati, D., Schmitt, B., Fischer, M., and Klaus, V. H. (2014). Evidence from the real world: 15N natural abundances reveal enhanced nitrogen use at high plant diversity in Central European grasslands. J. Ecol. 102, 456–465. doi: 10.1111/1365-2745.12202

CrossRef Full Text | Google Scholar

Mariotti, A., Germon, J., Hubert, P., Kaiser, P., Letolle, R., Tardieux, A., et al. (1981). Experimental determination of nitrogen kinetic isotope fractionation: some principles; illustration for the denitrification and nitrification processes. Plant Soil 62, 413–430. doi: 10.1007/BF02374138

CrossRef Full Text | Google Scholar

Miranda, C. H. B., Cadisch, G., Urquiaga, S., Boddey, R. M., and Giller, K. E. (1994). Mineral nitrogen in an oxisol from the Brazilian cerrados in the presence of Brachiaria spp. Eur. J. Agron. 3, 333–337. doi: 10.1016/S1161-0301(14)80163-9

CrossRef Full Text | Google Scholar

Nadelhoffer, K. J., and Fry, B. (1994). “Nitrogen isotope studies in forest ecosystems,” in Stable Isotopes in Ecology, eds K. Lajtha and R. Michener (Oxford: Blackwell), 22–44.

Google Scholar

Nuñez, J., Arevalo, A., Karwat, H., Egenolf, K., Miles, J., Ngonidzashe, C., et al. (2018). Biological nitrification inhibition activity in a soil-grown biparental population of the forage grass Brachiaria humidicola. Plant Soil 426, 401–411. doi: 10.1007/s11104-018-3626-5

CrossRef Full Text | Google Scholar

O’Sullivan, C. A., Fillery, I. R., Roper, M. M., and Richards, R. A. (2016). Identification of several wheat landraces with biological nitrification inhibition capacity. Plant Soil 404, 61–74. doi: 10.1007/s11104-016-2822-4

CrossRef Full Text | Google Scholar

Pardo, L. H., Hemond, H., Montoya, J., Fahey, T., and Siccama, T. (2002). Response of the natural abundance of 15N in forest soils and foliage to high nitrate loss following clear-cutting. Can. J. For. Res. 32, 1126–1136. doi: 10.1139/x02-041

CrossRef Full Text | Google Scholar

Pardo, L. H., McNulty, S. G., Boggs, J. L., and Duke, S. (2007). Regional patterns in foliar 15N across a gradient of nitrogen deposition in the northeastern US. Environ. Pollut. 149, 293–302. doi: 10.1016/j.envpol.2007.05.030

PubMed Abstract | CrossRef Full Text | Google Scholar

Rao, I. M., Ishitani, M., Miles, J., Peters, M., Tohme, J., Arango, J., et al. (2014). Climate-smart crop-livestock systems for smallholders in the tropics: integration of new forage hybrids to intensify agriculture and to mitigate climate change through regulation of nitrification in soil. Trop. Grassl.-Forrajes Trop. 2, 130–132. doi: 10.17138/TGFT(2)130-132

CrossRef Full Text | Google Scholar

Rao, I. M., Kerridge, P. C., and Macedo, M. C. M. (1996). “Nutritional requirements of Brachiaria and adaption to acid soils,” in Brachiaria: Biology, Agronomy, and Improvement, eds J. W. Miles, B. L. Maass, and C. B. do Valle (Cali, DC: CIAT Publication; no. 259), 53–71.

Google Scholar

Robinson, D. (2001). δ15N as an integrator of the nitrogen cycle. Trend Ecol. Evol. 16, 153–162. doi: 10.1016/S0169-5347(00)02098-X

CrossRef Full Text | Google Scholar

Shearer, G. B., Kohl, D. H., and Commoner, B. (1974). The precision of determinations of the natural abundance of nitrogen-15 in soils, fertilizers, and shelf chemicals. Soil Sci. 118, 08–316. doi: 10.1097/00010694-197411000-00005

CrossRef Full Text | Google Scholar

Song, G., Zhao, X., Wang, S. Q., and Xing, G. X. (2014). Nitrogen isotopic fractionation related to nitrification capacity in agricultural soils. Pedosphere 24, 186–195. doi: 10.1016/S1002-0160(14)60005-9

CrossRef Full Text | Google Scholar

Stamatiadis, S., Christofides, C., Tsadilas, C., Samaras, V., and Schepers, J. S. (2006). Natural abundance of foliar 15N as an early indicator of nitrogen deficiency in fertilized cotton. J. Plant Nutr. 29, 113–125. doi: 10.1080/01904160500416505

CrossRef Full Text | Google Scholar

Stevenson, B. A., Parfitt, R. L., Schipper, L. A., Baisden, W. T., and Mudge, P. (2010). Relationship between soil δ15N, C/N and N losses across land uses in New Zealand. Agric. Ecosyst. Environ. 139, 736–741. doi: 10.1016/j.agee.2010.10.020

CrossRef Full Text | Google Scholar

Subbarao, G. V., Nakahara, K., Hurtado, M. D. P., Ono, H., Moreta, D., Salcedo, A. F., et al. (2009). Evidence for biological nitrification inhibition in Brachiaria pastures. Proc. Natl. Acad. Sci. U.S.A. 106, 17302–17307. doi: 10.1073/pnas.0903694106

PubMed Abstract | CrossRef Full Text | Google Scholar

Subbarao, G. V., Rondon, M., Ito, O., Ishikawa, T., Rao, I., Nakahara, K., et al. (2007). Biological nitrification inhibition (BNI) - is it a widespread phenomenon? Plant Soil 294, 5–18. doi: 10.1007/s11104-006-9159-3

CrossRef Full Text | Google Scholar

Subbarao, G. V., Sahrawat, K. L., Nakahara, K., Rao, I. M., Ishitani, M., Hash, C. T., et al. (2013). A paradigm shift towards low-nitrifying production systems: the role of biological nitrification inhibition (BNI). Anal. Bot. 112, 297–316. doi: 10.1093/aob/mcs230

PubMed Abstract | CrossRef Full Text | Google Scholar

Sylvester-Bradley, R., Mosquera, D., and Mendez, J. (1988). Inhibition of nitrate accumulation in tropical grassland soils: effect of nitrogen fertilization and soil disturbance. J. Soil Sci. 39, 407–416. doi: 10.1111/j.1365-2389.1988.tb01226.x

CrossRef Full Text | Google Scholar

Takebayashi, Y., Koba, K., Sasaki, Y., Fang, Y., and Yoh, M. (2010). The natural abundance of 15N in plant and soil-available N indicates a shift of main plant N resources to NO3- from NH4 + along the N leaching gradient. Rapid Commun. Mass Spectrom. 24, 1001–1008. doi: 10.1002/rcm.4469

PubMed Abstract | CrossRef Full Text | Google Scholar

Yé, L., Abbadie, L., Bardoux, G., Lata, J., Nacro, H. B., and Masse, D. (2015). Contrasting impacts of grass species on nitrogen cycling in a grazed Sudanian savanna. Acta Oecol. 63, 8–15. doi: 10.1016/j.actao.2015.01.002

CrossRef Full Text | Google Scholar