Effect of grass species and fungal endophyte on soil nitrification potential

Abstract

This experiment tested the impact of grass species and fungal endophyte status and strain on soil nitrification. Three C3 grass–endophyte combinations were tested: Festuca arundinacea–Neotyphodium coenophialum; F. pratensis–N. uncinatum; Lolium perenne–N. lolii, and included different strains of endophyte and endophyte-free grasses. Three C4 species— Brachiaria decumbens, Paspalum dilatatum and Pennisetum clandestinum—of unknown Acremonium endophyte status, were also tested. Plants were grown for 6 weeks without fertiliser in pots in a field soil (Mollic Psammaquent), then rhizosphere soil was sampled and incubated at 27 °C with and without (NH₄)₂SO₄. The –N production was measured weekly. The presence of plants significantly increased the nitrification rate compared with bare soil. There were significant differences (P=0.02) among plant species. In all cases, endophyte presence increased nitrification in soil samples (P<0.001) relative to endophyte-free within a species.

Keywords:

• nitrification
• endophyte
• grass
• ryegrass
• fescue

Introduction

Nitrification, the microbial oxidation of ammonia (NH₃) to nitrate (), is a key step in the nitrogen (N) cycle in soil, with major economic and ecological consequences. In the context of pastoral farming, nitrification is important to farm productivity and profitability as nitrate leaching and reduction of to gaseous N compounds can reduce the availability of N for plant growth. Leached is a water contaminant and nitrous oxide (N₂O) is a greenhouse gas with a long-term climate-forcing potential 310 times that of carbon dioxide (CO₂) on a mass for mass basis (IPCC 1997).

In New Zealand, there is a strong focus on mitigating the undesirable effects of nitrification in pastoral systems (Di et al. 2007; Ledgard et al. 2007). Best practice fertiliser management, nutrient budgeting and systems analysis are all important in designing and implementing effective mitigation (Monaghan et al. 2007). One promising mitigation option is the use of nitrification inhibitors. During the last decade, many experiments have been conducted to assess the effectiveness of synthetic chemicals such as dicyandiamide (DCD) (Di & Cameron 2008), nitrapyrin (Calderón et al. 2005) and dimethylppyrazole phosphate (Zerulla et al. 2001) to inhibit nitrification in laboratory and farm settings. These chemicals act by inhibiting the ammonia monoxygenase enzyme found in all ammonia-oxidising bacteria (AOB) that drive the conversion of to (McCarty 1999), the first step in nitrification. In New Zealand dairy pastures, treating the soil with DCD can effectively reduce leaching (Di et al. 2009a) and N₂O emissions (Di & Cameron 2008). However, the effectiveness of inhibition varies with season and soil type (Di et al. 2009b).

Plants have long been suspected as being modifiers of soil nitrification (Munro 1966). However, it has not been possible to specifically assess their effects on soil bacteria until recently, when a modified strain of nitrifying bacteria was developed as a precise bio-assay (Subbarao et al. 2006). This strain has been used to show, in both field and laboratory studies (Ishikawa et al. 2003; Subbarao et al. 2006, 2007a, 2009), that the presence per se of specific plant species will suppress nitrification, a phenomenon termed biological nitrification inhibition (BNI). This occurs by the constitutive and inducible release of compounds from the plant into the rhizosphere, which act via the ammonia mono-oxygenase and the hydroxylamine oxidoreductase pathway to inhibit AOB (Subbarao et al. 2009). Studies have shown that BNI activity varies among species and that the tropical (C4) grasses such as brachiaria (Brachiaria spp.) and sorghum (Sorghum bicolor) exhibit greater endogenous BNI activity than temperate (C3) grass species such as ryegrass (Lolium spp.) and fescue (Festuca spp.) (Subbarao et al. 2007a). In situ production of naturally occurring nitrification inhibitors by plant roots within the rhizosphere of growing plants has appeal as a low-cost supplement or alternative to currently used synthetic nitrification inhibitors in pastoral systems.

Novel endophytic fungal associations with temperate forage grasses are used in pastoral agriculture as these provide agronomic benefits including enhanced plant persistence as well as improved animal health and production (Easton 2007). The pasture grasses attracting the most research and commercial attention are perennial ryegrass (Lolium perenne), tall fescue (Festuca arundinacea) and meadow fescue (F. pratensis); each with its respective mutualistic Neotyphodium fungal endophyte symbiont. The fungal biomass of these grass–endophyte associations is restricted to the foliar portion of the plant, however they can influence both primary and secondary metabolism in their host plants (Rasmussen et al. 2007). Endophytes may also increase root biomass, synthesise unique chemical compounds and influence the adaptation and ecology of host plants (Schardl et al. 2004). These effects can differ depending on the combinations of grass species and endophyte strain (Ahlholm et al. 2002).

A link between soil N status and fungal endophyte mass has been shown by Rasmussen et al. (2007) who found that endophyte concentration in perennial ryegrass was reduced by 40% at higher soil N availability. Field experiments have shown different soil C/N ratios under endophyte-infected versus endophyte-free swards (Franzluebbers & Stuedemann 2005). Endophyte has also been shown to have an impact on insects, nematodes, ruminant mammals, other fungi (Liu et al. 2011) and plants (Schardl et al. 2004) and earthworms (Humphries et al. 2001). However, there is little known about their effects on soil bacterial populations (Omacini et al. 2004; Sayer et al. 2004).

While there is some information on the impact of grass species on nitrification (Subbarao et al. 2006, 2007a, 2009), the authors are not aware of any study that has considered endophyte infection status in temperate grass species. This study tested the effect on soil nitrification of a range of common temperate pasture grasses with and without associated Neotyphodium fungal endophyte, together with tropical species previously shown to have significant BNI activity.

Materials and methods

Plant and endophyte species

The temperate C3 grass species perennial ryegrass, tall fescue and meadow fescue were tested, each without fungal endophyte and with three novel fungal endophyte combinations depending on species. In total, there were 12 treatment combinations, but only 10 are presented in this paper (Table 1). Tall fescue was infected with strains of Neotyphodium coenophialum endophyte, meadow fescue with strains of N. uncinatum and perennial ryegrass with strains of N. lolii. Endophyte strains included commercially available and experimental types. In the case of perennial ryegrass, the host plants were clonal copies of a single endophyte-free plant genotype, with replicate tillers of that plant artificially infected with different endophyte strains (Simpson 2009). For tall fescue and meadow fescue, plant genotype differed between each endophyte treatment, as these were naturally occurring grass–endophyte associations and not formed by the process of artificial infection. In both tall fescue and meadow fescue, the genotypes came from different populations; in the case of tall fescue, the populations used came from different subspecies with differing genomic constitutions adapted to different climatic conditions (Hand et al. 2010). For comparison, the study also included two C4 grasses that have been shown to have BNI activity (Subbarao et al. 2007a): brachiaria (Brachiaria decumbens) and kikuyu (Pennisetum clandestinum). A further C4 species, paspalum (Paspalum dilatatum), which has not been previously tested for effects on nitrification but is a widespread grass in northern New Zealand pastures, was also included. These C4 species do not host fungal endophytes of the Neotyphodium/Epichloë genera but may be endophytically infected with Acremonium spp. fungi (brachiaria [Kelemu et al. 2001], kikuy: [McGee et al. 1991] and paspalum [Clay et al. 1985)]. Plants in this experiment were not examined for the presence of fungi other than Neotyphodium/Epichloë. For each of the C4 species tested, eight plants from the same population were used in the experiment.

Table 1  Grass species and Neotyphodium fungal endophyte strains used in an experiment to assess the effect of grass–endophyte combinations on nitrification potential of soil. All C4 grasses are free of Neotyphodium endophytes, but the status of Acremonium fungal endophyte infection was untested.

Grass species (common name)	Description	Neotyphodium endophyte (infection status and strain)	Photosynthetic pathway
Brachiaria decumbens (Signal grass)		Endophyte-free	C4
Paspalum dilatatum (paspalum)		Endophyte-free	C4
Pennisetum clandestinum (kikuyu)		Endophyte-free	C4
Lolium perenne (perennial ryegrass)	cv. ‘Grasslands Samson’ clone #1	AR1	C3
	cv. ‘Grasslands Samson’ clone #1	AR6	C3
	cv. ‘Grasslands Samson’ clone #1	AR37	C3
	cv. ‘Grasslands Samson’ clone #1	Endophyte-free	C3
Festuca arundinacea (tall fescue)	Parent plant A F. arundinacea subsp. atlantigena (octaploid)	AR501	C3
	Parent plant B F. arundinacea subsp. atlantigena (octaploid)	AR542	C3
	cv. ‘Kentucky 31’ F. arundinacea subsp. arundinacea (hexaploid) ‘Continental’ form	Endophyte-free	C3
Festuca pratensis (meadow fescue)	Parent plant 1	AR1002	C3
	Parent plant 2	AR1006	C3
	J110	Endophyte-free	C3

Plant growth experiment

The kikuyu and paspalum plants were grown from seeds on agar plates until they germinated, then transplanted to commercial potting mix (60% peat and 40% sand mixed with ‘Nutricoat’ commercial nutrient mixture). For the other species, tillers were clonally propagated into potting mix from established plants of each grass species. For all temperate species, tillers were tested for the presence or absence of endophyte by tissue-print immunoblot (Hahn et al. 2003; Simpson 2009) and only tillers with the correct endophyte infection status were included in the experiment. Twelve clonal tillers of each entry were planted into 1.5 l pots containing potting mix and grown in a ventilated greenhouse during autumn 2009 with watering as required by capillary mats.

After 8 weeks, each plant was removed from its pot, rinsed free of potting mix and transferred to 500 g of a Pukepuke Black Sand soil at 60% water holding capacity (WHC) in plastic bags inside 750 ml plastic pots. The plastic bags were used to prevent loss of N in the drainage water. All water for the plants throughout the experiment was reverse osmosis (RO) purified. Pukepuke Black Sand (Mollic Psammaquent) soils are characterised by a 0.25 m black loamy fine-sand top horizon underlain by greyish-brown, fine sand-textured horizons (Cowie & Hall 1965). The soil was collected from a sheep-grazed paddock near the Flock House experimental site on the coastal fringe of the Manawatu plains of New Zealand (40.27°S, 175.29°E), and was taken from under vegetation at depths of 0–180 mm within 200 mm of a fence rigged with an electric wire to minimise effects from ruminant animal excreta. At the time of field collection, the soil had a pH of 5.9, a total C content of 45.2 g/kg soil and a total N content of 3.7 g/kg soil. The soil was stored dry for 2 months then wet to 60% WHC and thoroughly mixed before weighing each pot.

Six pots were filled with soil, maintained at 60% WHC, and left without plants to assess the nitrification properties of soil alone. The transplanted grasses were grown for an acclimatisation period of 2 weeks, and were then trimmed to a height of 4 cm. Trimmings were removed from the pot and did not contribute to the N status of the soil samples. Eight clonal copies from each of the 15 entries were selected on the basis of size uniformity and used for the continuation of the experiment. Plants were left to grow for a further 4 weeks. During the plant acclimatisation and growth phases, soil moisture content was maintained at approximately 60% WHC by watering to weight once a week, supplemented by watering with a fixed volume of water every second day.

Soil nitrification assay

Soil was sampled after 4 weeks of plant regrowth. Each plant was removed from the pot and a 45 g sample of soil directly surrounding the root zone of each plant was collected, visible root fragments were removed and the soil was mixed thoroughly and split into nine sub-samples of 5 g each. Four of the nine sub-samples were amended with 3 mg of (NH₄)₂SO₄ in solution (1 ml RO water); four sub-samples received 1 ml of RO water only. These sub-samples were placed in small plastic bags containing air to allow circulation, sealed with a rubber band and incubated at 27 °C. The ninth sub-sample was analysed to quantify initial and concentrations at harvest using the method detailed later in the paper. The same protocol was used with soil sampled from six pots without plants.

At weekly intervals for 4 weeks, one (NH₄)₂SO₄ amended soil sub-sample and one RO water treated soil sub-sample for each replicate of each entry was removed from the incubator and analysed. Each sub-sample was extracted by shaking with 30 ml of 2 M KCl for 60 min using an end-over-end shaker running at approximately 30 rpm, adding phenylmecuric acetate to stop microbial activity and filtering using Whatman no. 42 filter paper. Each solution was then analysed using a FIAstar 5000 Analyser (Foss Tecator, Hoganas, Sweden) to quantify and concentrations.

Statistical analysis

The means of the net soil –N produced from the amended –N over the 4-week incubation cycle were statistically analysed with a linear mixed effects model using the statistical analysis package ‘R’ version 3 (Pinheiro et al. 2009).

Net soil nitrification rate was calculated using weekly net soil –N production and plotted with time (data not presented). Maximum weekly nitrification rate (achieved at different weeks for different species) was statistically analysed using a linear mixed effects model using R version 3 (Pinheiro et al. 2009).

Results

–N production from added NH₄–N increased over the course of the experiment due to nitrification in all soils (data not presented). The average rate of nitrification in bare soil was 79.3 µg –N/g dry soil/4 weeks (i.e. lower than the values recorded in the presence of plants, see Tables 2 and 3). There were significant (P < 0.05) effects on nitrification attributed to plant species and to endophyte infection status.

Soil –N total production over the 4-week period differed significantly among grass species (P = 0.02), as did maximum rates of production (P = 0.01). The lowest nitrification was associated with meadow fescue both in terms of the total amount and maximum rate of –N production over the 4 weeks (Table 2). Brachiaria was the next lowest for total –N production, but was not significantly different from other species with the exception of paspalum, which had the highest total.

Table 2  Mean total –N production and mean maximum potential nitrification rate from an –N-amended soil over a 4-week incubation interval. The week that the maximum nitrification was achieved is indicated in parentheses. The soils were sampled from the rhizosphere of different plant species after 6 weeks of growth. Means exclude endophyte infected entries. Least significant difference (LSD) (5%) is for all comparisons within the same column.

Grass species	Total soil –N production (µg/g dry soil)	Maximum potential nitrification rate (µg/g dry soil/week)
Brachiaria	99.7	42.6 (3)
Paspalum	123.1	43.9 (3)
Kikuyu	112.5	46.4 (3)
Perennial ryegrass	111.7	46.0 (3)
Tall fescue	108.9	49.1 (3)
Meadow fescue	95.1	28.9 (2)
LSD (5%)	16.3	9.9

Endophyte status significantly influenced nitrification, with mean soil –N production over 4 weeks for endophyte-free and endophyte-infected plants of 105.2 µg/g dry soil and 128.8 µg/g dry soil respectively(P < 0.001).The endophyte-related increase in nitrification was apparent in all plant–endophyte combinations. (Table 3).

Table 3  Mean total –N production and mean maximum soil nitrification rate from an –N-amended soil over a 4-week incubation interval. The soils were sampled from the rhizosphere of different plant species and species–endophyte strain combinations after 6 weeks of growth. The week that the maximum nitrification was achieved is indicated within the brackets. LSD (5%) is for all comparisons within same column.

Grass species	Endophyte	Total soil –N production (µg/g dry soil)	Maximum soil nitrification rate (µg/g dry soil/week)
Perennial ryegrass	AR1	132.2	48.7 (2)
Perennial ryegrass	AR6	134.6	49.2 (2)
Perennial ryegrass	AR37	126.0	48.9 (2)
Perennial ryegrass	Endophyte-free	111.7	46.0 (3)
Tall fescue	AR501	124.5	44.5 (2)
Tall fescue	AR542	132.7	52.9 (3)
Tall fescue	Endophyte-free	108.9	49.1 (3)
Meadow fescue	AR1002	124.4	52.1 (2)
Meadow fescue	AR1006	127.6	49.9 (2)
Meadow fescue	Endophyte-free	95.1	28.9 (2)
LSD (5%)		13.7	10.5

The mean maximum soil nitrification rate for endophyte-free and endophyte-infected plants was 41.1 and 49.4 µg –N/g dry soil/week respectively (P = 0.01). This result depended heavily on the meadow fescue data, with no difference in ryegrass and tall fescue (Table 3).

Discussion

This experiment showed plant- and endophyte-mediated differences in –N production from soil amended with ammonium over a 4-week incubation period. The initially uniform soil had different grass and grass–endophyte combinations grown for 6 weeks prior to the assay. This finding supports the recent overview of the capacity of plants to modify soil nitrification (Fillery 2007; Subbarao et al. 2009), although the finding of Subbarao et al. (2009) of significant inhibition of nitrification by some field-grown plants relative to bare soil plots was not borne out in the current data, which showed that bare soil treatment had a lower nitrification rate than plant-treated soils in pots. It should be noted that the field plot soil of Subbarao et al. (2009) was bare for 3 years, whereas the soil used in this experiment was bare for less than 3 months after being collected from an actively growing mixed species pasture.

Plants may influence soil nitrification capacity in several ways, including altering the availability of for nitrifiers and by the release of inhibitory compounds. Plants can compete directly for with nitrifiers, so any physiological and morphological differences among plants that affect the uptake of from soil (Craine et al. 2002) may have influenced our assay results by altering the number of nitrifiers in each soil sample prior to the beginning of the incubation. This experiment did not quantify differences in plant physiology or morphology. However, the C3 grasses have adapted to fill different niches as reflected in fundamental differences in physiology between C3 and C4 species. Plant species are also known to differ in their preference for or (Wiltshire 1973; Haynes & Goh 1978; Bolan et al. 2005) and the influence of endophyte status on plant physiology has been well documented (Schardl et al. 2004).

Plants may also alter availability indirectly through their influence on heterotrophic microbes. Carbon released in root exudates may stimulate heterotrophic microbial growth, increasing consumption by heterotrophs and decreasing the substrate for nitrifiers (Ross et al. 2001; Booth et al. 2005), thus reducing the abundance of AOB (Verhagen et al. 1992). The quantity and type of carboniferous compounds in root exudates differs among grass species (Lambers et al. 2009) and with endophyte status (Hadacek & Kraus 2002; Malinowski et al. 2004; Cecilia et al. 2010), therefore different levels of heterotrophic proliferation can be expected within the rhizosphere of different grass species, which may influence even short-term experiments such as the one described here.

A further way in which some plants alter nitrification is by the release of compounds from the plant into the rhizosphere that specifically inhibit the activity of AOB, the phenomenon termed BNI (Subbarao et al. 2007a; Subbarao et al. 2008; Subbarao et al. 2009). Recently, inducible inhibitory compounds have been identified in exudates of roots of the tropical grasses Brachiaria spp. and sorghum as causative of the BNI effect (Gopalakrishnan et al. 2007; Subbarao et al. 2009; Zakir et al. 2009). However, it should be pointed out that, compared with recently bare soil, there was increased nitrification in brachiaria-treated soil in our short-term experiment, implying more complex processes are operating than simply the release of compounds capable of inhibiting nitrification.

Even in the absence of induction (Subbarao et al. 2007b), the brachiaria-treated soil samples studied here exhibited relatively low nitrification potential compared with other species; this is consistent with BNI evidence published elsewhere with ammonia used for experimental induction of BNI activity in plants (Subbarao et al. 2007b). The species used in this study, B. decumbens, has been shown to be comparable to the more widely studied BNI species B. humidicola, with both exhibiting the highest inducible BNI activity per unit root biomass among a range of C3 and C4 pasture and crop plants (Subbarao et al. 2007a). Brachiaria spp. are now known to occasionally be infected with endophytic Acremonium fungi (Kelemu et al. 2001). As the authors only became aware of this information after the experiment was conducted, the Acremonium status of the brachiaria was not checked; further investigation is clearly warranted given the impact of Neotyphodium endophytes measured in this study and as a way of confidently interpreting this work compared with previous results for Brachiaria spp. (which was also untested for Acremonium fungi). In the work reported here, it is notable in the experiment that soil samples from the endophyte-free meadow fescue, a C3 species, matched the low B. decumbens level of nitrification inhibition. This may present opportunities for breeders of temperate forage grasses to utilise this trait in fescue and ryegrass breeding, given the ability to introgress genes and to generate a variety of stable inter-specific Festulolium cultivars (Kopecký et al. 2006).

The results of this study indicate that the infection status of grasses, with Neotyphodium fungal endophytes, has an effect on soil nitrification. However, all the C3 and C4 grasses tested in this experiment may host a range of other fungi (Clay et al. 1985; McGee et al. 1991; Kelemu et al. 2001) and the effect of their presence on soil nitrification was not quantified in this experiment. Similarly, plant health and growth were not specifically monitored or tested. While plants were selected for uniformity, it is possible that treatment effects were influenced by differential plant shoot or root growth in the experiment, or that endophyte-related enhanced plant growth may have occurred (Schardl et al. 2004).

In this experiment, perennial ryegrass was generated as clonal material (clone #1 from cultivar Samson) so that any endophyte-free and endophyte-infected differences could be attributed solely to the endophyte treatment rather than the effect of plant genotype (Easton et al. 2002). However, the authors did not have access to plant clonal material infected with an endophyte series for the two fescue species so, while the treatment effects have been demonstrated, our ability to attribute these to endophyte per se is limited.

There is a specific need to explore the ammonia induction effect on BNI in temperate grass–endophyte combinations to assess the longer term and field impact of these combinations and to address specific experimental design improvements to clarify the findings in this experiment.

Conclusions

The results of this study suggest that the presence of plants significantly increased the potential soil nitrification rate compared with bare soil. There were significant differences among plant species, and endophyte presence in perennial ryegrass, tall fescue and meadow fescue increased nitrification potential of soil samples. These findings add further levels of complexity to the factors already known to modify nitrification in pasture soils.

Acknowledgements

We would like to thank the New Zealand Foundation for Research, Science and Technology (FRST) for a Technology in Industry Fellowship (TIF) supporting CL. The authors are grateful to Mr Wayne Simpson of AgResearch Grasslands for supplying endophyte plant material.