Nodulated N2-fixing Casuarina cunninghamiana is the sink for net N transfer from non-N2-fixing Eucalyptus maculata via an ectomycorrhizal fungus Pisolithus sp. using 15NH4+ or 15NO3 supplied as ammonium nitrate

Authors

  • Xinhua He,

    Corresponding author
    1. Department of Botany, University of Queensland, Brisbane, Queensland 4072, Australia;
    2. Department of Land, Air and Water Resources, University of California, Davis, CA 95616, USA;
    3. College of Life Sciences, Yunnan Normal University, Kunming, Yunnan 650092, China
    • Author for correspondence: Xinhua He Tel: +1 530 752 4131 Fax: +1 530 752 1552 Email: huahe@ucdavis.edu

    Search for more papers by this author
  • Christa Critchley,

    1. Department of Botany, University of Queensland, Brisbane, Queensland 4072, Australia;
    Search for more papers by this author
  • Hock Ng,

    1. Department of Botany, University of Queensland, Brisbane, Queensland 4072, Australia;
    Search for more papers by this author
  • Caroline Bledsoe

    1. Department of Land, Air and Water Resources, University of California, Davis, CA 95616, USA;
    Search for more papers by this author

Summary

  • • To determine the effects of nitrogen source on rates of net N transfer between plants connected by a common mycorrhizal network, we measured transfer of N supplied as 15NH414NO3 or 14NH415NO3 in three Casuarina/Eucalyptus treatments interconnected by a Pisolithus sp. The treatments were nonnodulated nonmycorrhizal/nonmycorrhizal; nonnodulated mycorrhizal/mycorrhizal; and nodulated mycorrhizal/mycorrhizal.
  • • Mycorrhization was 67% in Eucalyptus and 36% in Casuarina. N2 fixation supplied 38% of the N in Casuarina. Biomass, N and 15N contents were lowest in nonmycorrhizal plants and greatest in plants in the nodulated/mycorrhizal treatment.
  • • Nitrogen transfer was enhanced by mycorrhization and by nodulation, and was greater when N was supplied as 15NH4+ than 15NO3. Nitrogen transfer rates were lowest in the nonmycorrhizal treatment for either 15N source, and greatest in the nodulated, mycorrhizal treatment. Transfer was greater to Casuarina than to Eucalyptus and where ammonium rather than nitrate was the N source.
  • • Irrespective of 15N source and of whether Casuarina or Eucalyptus was the N sink, net N transfer was low and was similar in both nonnodulated treatments. However, when Casuarina was the N sink in the nodulated, mycorrhizal treatment, net N transfer was much greater with 15NH4+ than with 15NO3. High N demand by Casuarina resulted in greater net N transfer from the less N-demanding Eucalyptus. Net transfer of N from a non-N2-fixing to an N2-fixing plant may reflect the very high N demand of N2-fixing species.

Introduction

In nonagricultural soils, ammonium (NH4+) is the dominant form of nitrogen. In most agricultural soils, nitrate (NO3) is the dominant form of N, as NH4+ is not readily available because of rapid nitrification (Marschner, 1995; Forde & Clarkson, 1999; Brady & Weil, 2002). Most agricultural crops grow better on NO3 (Haynes & Goh, 1978; Malhi et al., 1988; Oaks, 1992; Marschner, 1995) while most forest trees grow better on NH4+ (Haynes & Goh, 1978; Oaks, 1992; Kronzucker et al., 1997; Hall & Matson, 1999). Most plant species grow best on a mixture of NH4+ and NO3 (Marschner, 1995), while some others, especially those confined to nutrient-impoverished soils, use either NH4+ or NO3 (Atkin, 1996). Like plants, mycorrhizal fungi use either NH4+ or NO3 (Alexander, 1983, 1989; Hogberg & Alexander, 1995; Turnbull et al., 1995; Michelsen et al., 1996; Hawkins et al., 2000; Martin & Plassard, 2001), although under axenic conditions most fungi grow better on NH4+ than NO3 (Finlay et al., 1992; Keller, 1996; Baar et al., 1997). When plants are mycorrhizal, the preference for using an inorganic N source for NH4+ or NO3 in both plants and fungi must be considered, but there is little evidence in the literature on this topic. This study considers effects of ammonium and nitrate on net N transfer between plants connected by mycorrhizal hyphae.

A single plant can form mycorrhizas with many fungi in intra- or interspecific combinations. Hyphae of a single fungal species can interconnect many plants. As a consequence, a common mycorrhizal network (CMN) can form within and between plant roots (Newman, 1988). CMNs provide pathways for movement of nutrients such as carbon (Simard et al., 2002) and N (He et al., 2003), or even of genetic material (Giovannetti et al., 2004). Transfer of NH4+ or NO3 between N2-fixing and non-N2-fixing plants is mediated by arbuscular mycorrhizal (AM) CMNs (Bethlenfalvay et al., 1991; Frey & Schuepp, 1993; Johansen & Jensen, 1996). Transfer of NH4+ is mediated by ectomycorrhizal (EM) CMNs (Arnebrant et al., 1993; Ekblad & Huss-Danell, 1995). Nitrogen gradients between N-rich donors and N-limited receivers may be a driving force for unidirectional N transfer via CMNs (Bethlenfalvay et al., 1991; Frey & Schuepp, 1993). However, net N transfer was not determined in most studies. One study did report low levels of net N transfer (0.3%) from barley to pea connected by an AM fungus, Glomus intraradices (Johansen & Jensen, 1996).

Two tree species, N2-fixing (with the actinomycete Frankia) Casuarina and non-N2-fixing Eucalyptus, often grow together in Australia. They are predominantly EM but can form AM associations as well (Attiwill & Adams, 1996). Eucalyptus can obtain N from adjacent nodulated Casuarina and other N2-fixing plants (Subbarao & Rodriguez-Barrueco, 1995; Attiwill & Adams, 1996). We previously reported two-way or net N transfer of 15NH4+ and 15NO3, when supplied separately, between 6-month-old Casuarina and Eucalyptus seedlings linked by the ectomycorrhizal fungus Pisolithus sp. Nitrogen transfer was affected by mycorrhizal fungi and/or nodulation (He et al., 2004). As NH4+ and NO3 have different chemical mobility in soil, and their uptake by plants and fungi may be different, it is important to compare effects of NH4+ and NO3 on mycorrhizal mediation of N transfer between plants. In that study (15NH4)2SO4 or K15NO3 was provided separately, so the effects of each N form on uptake and transfer of the other N form were not possible in two individual experiments. In order to compare the differences in N transfer, in the present study we used 14NH414NO3 as the N source to fertilize Casuarina and Eucalyptus seedlings for 11 months. We then applied 15N as 15NH414NO3 or 14NH415NO3 for 1 month. Thus this study differed from our previous one in that we determined the effects of each N form on uptake and transfer of the other N form. In addition, the 12-month-old seedlings tested were older and larger than in the previous study. In this study we asked four questions. (1) What are the net N transfer rates of 15NH4+ and 15NO3? (2) Do mycorrhization and nodulation affect rates and direction of 15NH4+ and 15NO3 transfer? (3) Which plant species is the stronger sink for N? (4) Does 15NH4+ affect uptake and transfer of 15NO3 and vice versa?

Materials and Methods

Experimental design

The experimental system, growth media and plant growth conditions were the same as those used by He et al. (2004). There were three treatments with either Casuarina or Eucalyptus as donor (Table 1). First, as a control, nonnodulated nonmycorrhizal Casuarina/nonmycorrhizal Eucalyptus (NodMyc/Myc) seedlings were paired to monitor any N transfer via the soil pathway, as there were no hyphal connections between roots. Second, nonnodulated mycorrhizal Casuarina/mycorrhizal Eucalyptus (NodMyc+/Myc+) seedlings were paired to study mycorrhizal effects on N transfer, as there was no nodulation in Casuarina. Third, nodulated mycorrhizal Casuarina/mycorrhizal Eucalyptus (Nod+Myc+/Myc+) seedlings were paired to study effects of mycorrhizas plus N2 fixation on N transfer. To minimize mass flow and diffusion through soil pathways, two perforated Perspex plates (0.25 cm wide) were inserted in the middle of a 5 l plastic box (300 × 12 × 150 cm) so that they formed a 5 mm air gap. This divided the box into two chambers. On one side of the plates there was 37 µm nylon mesh that allowed only hyphal connections. To further minimize water movement through the nylon mesh between the two chambers, high water-holding-capacity crystals [0.5% (w/w), RainSaver, Hortex Australia Pty Ltd, NSW] were included in the potting mix. Each m3 potting mix contained 8.3 kg stock fertilizer mix (1.5 kg cattle blood and bone powders, 0.3 kg K2SO4, 1.5 kg superphosphate, 2.5 kg dolomite, 1.5 kg hydrated lime, 1.0 kg gypsum). Each chamber contained 3 kg potting mix (sand : peat = 2 : 1 v/v, pH 6.5).

Table 1. Experimental design for two-way nitrogen transfer between Casuarina and Eucalyptus seedlings grown in two-chambered pots
Pair nameN donorN receiverTreatment status
MycNodCode
  1. 15N was supplied to growth medium of donor plant for 4 wk. The mycorrhizal fungus was Pisolithus sp. and the N2-fixing symbiont was Frankia. C, Casuarina; E, Eucalyptus; Myc or M, mycorrhizal status; Nod or F, nodulation status.

Nonnodulated Casuarina Eucalyptus CM–F– → EM–
 nonmycorrhizal pairs Eucalyptus Casuarina EM– → CM–F–
Nonnodulated Casuarina Eucalyptus +CM+F– → EM+
 mycorrhizal pairs Eucalyptus Casuarina +EM+ → CM+F–
Nodulated Casuarina Eucalyptus ++CM+F+ → EM+
 mycorrhizal pairs Eucalyptus Casuarina ++EM+ → CM+F+

Production of aseptic ectomycorrhizal seedlings

Sterilized seeds of Casuarina cunninghamiana Miq. and Eucalyptus maculata Hook were germinated on agar media modified from Ahmad & Hellebust (1991), containing 14NH414NO3 as the N source. For mycorrhization and/or nodulation, germinated seeds were incubated on agar plates with or without Pisolithus sp. and/or Frankia strains in a growth chamber (23/18°C, day/night) (Brundrett et al., 1996; He et al., 2004). As controls, boiled fungal cultures and/or nodule suspensions were added to the Myc and/or Nod treatment. Mycorrhizas and/or nodules formed after 6 wk. Seedlings were then transplanted into experimental pots and watered weekly (5–10 ml, 2.0 mmol 14NH414NO3 as needed) for 11 months (September 1998 to July 1999). Plants were grown in a temperature-controlled glass house (St Lucia campus, University of Queensland). To prevent transformation of NH4+ to NO3, N-serve, a nitrification inhibitor, was added [2-chloro-6-(trichloromethyl) pyridine, Sigma-Aldrich Pty Ltd, Castle Hills, NSW]. N-serve was dissolved in ethanol, diluted in distilled water and mixed with potting mix (5 mg kg−1 potting mix). When plants were 11 months old, 5 mg 15N (15NH414NO3 or 14NH415NO3, ≈98.0 at %15N) was added to the growth media of the donor side each week for 4 wk (July–August 1999). No 15N was supplied to the receivers.

Nitrogen assays, N2 fixation and nitrogen-transfer calculations

Fresh Casuarina needles were collected before 15N labelling to estimate N2 fixation using the 15N natural abundance technique (Boddey et al., 2000; Unkovich et al., 2001). Fresh root samples were preserved in 50% ethanol–water solution for determination of mycorrhizal colonization (Brundrett et al., 1996). Shoots, roots and nodules were harvested after 4 wk 15N labelling. Dried plant matter was ground and analysed for total N, at (atom) %15N and 15N (automated 15N/13C analyser–mass spectrometer, Europa Scientific, Crewe, UK) at the University of Queensland, Australia. N2 fixation was expressed as specific nodule activity (µg tissue N mg−1 d. wt nodule) and as percentage plant N derived from fixation (%Nfixed) (Warembourg, 1993). Nitrogen transfer was expressed as percentage N transfer; percentage N in the N receiver derived from transfer (%NDFT); and amount of N transferred (mg per plant) (Johansen & Jensen, 1996, also see He et al., 2004). Net N transfer was calculated as the difference between N transferred from Casuarina to Eucalyptus and from Eucalyptus to Casuarina.

Statistical analyses

Each treatment had nine replicates for a total of 27 paired samples. Data (n = 9) were arc-sine transformed if zero values existed, then analysed by anova procedures for N donor and N receiver, respectively. Differences in means were compared (Tukey's honestly significant difference method) and considered significant at P≤ 0.05 or 0.01 (Sokal & Rohlf, 1995).

Results

Biological nitrogen fixation in Casuarina plants

There were no statistically significant differences in N2 fixation between Nod+Myc and Nod+Myc+Casuarina seedlings (as donors or receivers), and data were combined (n = 36). Nodule biomass was 4.6 ± 0.4 g d. wt per plant, plant percentage N was 1.86 ± 0.09, and δ15N values were −0.21‰ ± 0.22. We also grew nodulated Casuarina without N fertilization in order to measure the basic N2 fixation level. Non-N-fertilized Casuarina seedlings had significantly lower δ15N values (−6.44‰ ± 0.78). For NH4NO3 grown Casuarina seedlings, N2 fixation supplied 38% ± 2.6 of their total N requirement, and their specific nodule activity was 18.0 ± 0.2 µg N mg−1 d. wt nodule.

Mycorrhizal formation

Mycorrhizal colonization of Eucalyptus was twice as much as in Casuarina. For both plant species, colonization was not affected by N source (Table 2). In both mycorrhizal (NodMyc+/Myc+ and Nod+Myc+/Myc+) treatments, colonization was similar in Casuarina (3638%) and in Eucalyptus (64%−69%). As expected, colonization was very low (<4%) in NodMyc/Myc treatment, where no fungal mycelia were observed in the growth media. In both mycorrhizal and nonmycorrhizal treatments, roots did not cross the nylon mesh.

Table 2. Mycorrhizal colonization and percentage nitrogen of Casuarina and Eucalyptus seedlings grown in two-chambered pots
TreatmentLabelled 15N formColonization (%)N (%)
MycNod Casuarina Eucalyptus Casuarina Eucalyptus
  1. All plants were fertilized with 14NH414NO3 for 11 months, then donor plants were treated with 15NH414NO3 or 14NH415NO3 for 1 month. As donor species did not affect mycorrhizal colonization and percentage N, data for donor and receiver seedlings were combined (means ± SE, n = 18). For each parameter values followed by different letters designate significant differences (P ≤ 0.05) between treatments for a given 15N form and a given species within columns (a, b, c), and between species for a given N form and a given treatment within rows (x, y, z), respectively.

  2. Myc, mycorrhizal status; Nod, nodulation status.

15N-ammonium 4 ± 1(b, x) 3 ± 1(b, x)0.79 ± 0.05(c, x)0.59 ± 0.03(b, y)
   15N-nitrate 4 ± 1(b, x) 4 ± 1(b, x)0.83 ± 0.08(c, x)0.65 ± 0.05(b, y)
+ 15N-ammonium36 ± 4(a, y)64 ± 5(a, x)1.14 ± 0.09(b, x)0.62 ± 0.02(b, y)
   15N-nitrate36 ± 4(a, y)65 ± 6(a, x)1.11 ± 0.07(b, x)0.68 ± 0.03(b, y)
++ 15N-ammonium37 ± 5(a, y)69 ± 7(a, x)1.48 ± 0.09(a, x)0.82 ± 0.05(a, y)
15N-nitrate38 ± 6(a, y)68 ± 5(a, x)1.42 ± 0.06(a, x)0.82 ± 0.07(a, y)

Biomass production

For each plant species, as biomass of N donors or N receivers was similar, and as application of 15NH4+ or 15NO3 for 4 wk did not affect growth, biomass data from donors and receivers were combined (Fig. 1). For each 15N source, biomass was least in the NodMyc/Myc; greater in the NodMyc+/Myc+ and greatest in the Nod+Myc+/Myc+ treatment. Casuarina had twofold greater biomass production than Eucalyptus for all three treatments.

Figure 1.

Effects of mycorrhizas (M), nodulation (F) and nitrogen source on biomass of Casuarina (C, lined bars) and Eucalyptus (E, open bars) seedlings (means ± SE, n = 18). Donor and receiver biomass data were combined. Plants were fertilized with 14NH414NO3 for 11 months and 15N (N donor only) for 1 month before harvest. Values followed by different letters designate significant differences (P = 0.05) between treatments for a given 15N form and a given species (a, b, c); and between 15N forms for a given species and a given treatment (x, y, z), respectively.

Plant percentage nitrogen and nitrogen content

Plant percentage N was significantly greater for Casuarina than for Eucalyptus in all treatments. For each plant species, plant percentage N was similar when supplied from either N source (Table 2). For Casuarina seedlings, plant percentage N was least in NodMyc; greater in NodMyc+ and greatest in the Nod+Myc+ treatment. In contrast, for Eucalyptus seedlings, plant percentage N in MycEucalyptus was similar to that in Myc+Eucalyptus, when the Myc+Eucalyptus was partnered with NodMyc+Casuarina, but was significantly lower than in Myc+ eucalypts partnered with Nod+Myc+Casuarina.

As either donors or receivers in a given treatment, total plant N content in Casuarina or Eucalyptus was not affected by application of 15NH4+ or 15NO3 for 4 wk (Fig. 2). For a given N source for Casuarina or Eucalyptus, N content had the following pattern: NodMyc/Myc < NodMyc+/Myc+ (1.7–2.4 times) < Nod+Myc+/Myc+ (3.6–3.9 times). Casuarina, as either donor or receiver, had approximately two to four times greater N content than Eucalyptus, independent of N source, but not independent of mycorrhizal or nodulation treatment. Biomass and N content were positively correlated in both Casuarina and Eucalyptus (Fig. 3, r2 = 0.82–0.93), and were least in NodMyc/Myc; greater in NodMyc+/Myc+ and greatest in the Nod+Myc+/Myc+ treatment.

Figure 2.

Effects of mycorrhizas (M), nodulation (F), nitrogen source, and identity of N donor or N receiver on N content of Casuarina (C) and Eucalyptus (E) seedlings (lined bars, donors; open bars, receivers). Plants were fertilized with 14NH414NO3 for 11 months and with 15N (N donor only) for 1 month before harvest. Values (means ± SE, n = 9) followed by different letters designate significant differences (P = 0.05) between treatments for a given 15N form and a given species (a, b, c); between 15N forms for a given species and a given treatment (x, y, z); and between donors and receivers for a given species, a given 15N form and a given treatment (α, β, γ), respectively.

Figure 3.

Relationship between biomass and nitrogen content of Casuarina (circles) and Eucalyptus (triangles) seedlings. 15N donors are indicated by solid symbols; 15N receivers by open symbols. Plants were fertilized with 14NH414NO3 for 11 months and 15N (N donor only) for 1 month before harvest. Data are means ± SE (n = 9). Regressions are shown for Casuarina (solid lines) and Eucalyptus (dashed lines). C, Casuarina; E, Eucalyptus; M, mycorrhizal status; F, nodulation status.

15N at % excess and 15N content

Roots had greater 15N at % excess than shoots for both plant species, whether Casuarina or Eucalyptus was the donor or the receiver, and whether 15NH4+ or 15NO3 was the 15N source (Table 3). As the donor supplied with either 15NH4+ or 15NO3, Casuarina had lower shoot 15N at % excess, but greater root 15N at % excess than Eucalyptus. As receivers supplied with either 15NH4+ or 15NO3, Eucalyptus had greater shoot and root 15N at % excess than Casuarina. Distribution patterns of 15N at % excess between treatments were similar in shoots and roots. For both Casuarina and Eucalyptus donors, shoot and root distribution patterns for total 15N at % excess decreased significantly as follows: NodMyc/Myc > NodMyc+/Myc+ > Nod+Myc+/Myc+ treatment. The opposite was true for both Casuarina and Eucalyptus as receivers. Total 15N at % excess increased significantly as follows: NodMyc/Myc < NodMyc+/Myc+ < Nod+Myc+/Myc+ treatments. As expected, with a given 15N source and a given treatment, total 15N at % excess was greater in donors (Casuarina or Eucalyptus) than in receivers (Eucalyptus or Casuarina). Surprisingly, total 15N at % excess was greater in Casuarina donors than in Eucalyptus donors, but was lower in Casuarina receivers than in Eucalyptus receivers in all three treatments. For both Casuarina and Eucalyptus as donors in all three treatments, and for Casuarina and Eucalyptus receivers in the NodMyc/Myc treatment, total 15N at % excess was significantly greater when 15NO3 was the N source (Table 3). For both Casuarina and Eucalyptus as receivers, 15N at % excess was similar in the NodMyc+/Myc+ and the Nod+Myc+/Myc+ treatment, regardless of N source.

Table 3. 15N at % excess of Casuarina and Eucalyptus seedlings grown in two-chambered pots
Codes 15N source 15N donor 15N (at % excess)* 15N receiver 15N (at % excess)
ShootsRootsTotal (weighted)ShootsRootsTotal (weighted)
  1. All plants were fertilized with 14NH414NO3 for 11 months, then donor plants were treated with 15NH414NO3 or 14NH415NO3 for 1 month. Data are means ± SE (n = 9). For each parameter values followed by different letters designate significant differences (P ≤ 0.05) between 15N forms for a given species and a given treatment within columns (a, b, c), and between treatments for a given 15N form and a given species within columns (x, y, z), respectively.

  2. C, Casuarina; E, Eucalyptus; M, mycorrhizal status; F, nodulation status.

  3. *Values of total 15N at % excess calculated by WAE = [AE(a) × TN(a) + AE(b) × TN(b) +… AE(n) × TN(n)]/TN(a + b +…n), where WAE is weighted 15N at % excess, AE is 15N at % excess in different plant parts represented as a, b, … , n, and TN is total N in parts a, b, … , n (Warembourg, 1993).

CM–F– 15NH4+ Casuarina 0.148 ± 0.025(b, x)5.117 ± 0.109(b, x)1.556 ± 0.183(b, x) Eucalyptus 0.004 ± 0.001(a, y)0.027 ± 0.002(b, z)0.010 ± 0.001(b, z)
→ EM– 15NO3 Casuarina 0.245 ± 0.006(a, x)8.325 ± 0.177(a, x)2.742 ± 0.079(a, x) Eucalyptus 0.007 ± 0.001(a, y)0.229 ± 0.032(a, z)0.060 ± 0.007(a, z)
EM– 15NH4+ Eucalyptus 0.544 ± 0.029(a, x)1.970 ± 0.135(b, x)0.942 ± 0.025(b, x) Casuarina 0.002 ± 0(a, y)0.014 ± 0.002(b, z)0.005 ± 0.001(b, z)
CM–F– 15NO3 Eucalyptus 0.580 ± 0.074(a, x)2.739 ± 0.083(a, x)1.167 ± 0.068(a, x) Casuarina 0.001 ± 0(a, z)0.032 ± 0.003(a, z)0.010 ± 0.001(a, z)
CM+F– 15NH4+ Casuarina 0.079 ± 0.005(b, y)2.806 ± 0.077(b, y)0.739 ± 0.057(b, y) Eucalyptus 0.006 ± 0(a, y)0.258 ± 0.005(a, y)0.082 ± 0.002(a, y)
→ EM+ 15NO3 Casuarina 0.148 ± 0.003(a, y)5.786 ± 0.041(a, y)1.436 ± 0.091(a, y) Eucalyptus 0.010 ± 0.001(a, y)0.307 ± 0.017(a, y)0.096 ± 0.002(a, y)
EM+ 15NH4+ Eucalyptus 0.170 ± 0.013(b, y)1.406 ± 0.026(b, y)0.534 ± 0.012(b, y) Casuarina 0.003 ± 0(a, y)0.082 ± 0.011(a, y)0.025 ± 0.001(a, y)
CM+F– 15NO3 Eucalyptus 0.333 ± 0.012(a, y)2.287 ± 0.048(a, y)0.899 ± 0.017(a, y) Casuarina 0.008 ± 0.001(a, y)0.078 ± 0.007(a, y)0.022 ± 0.001(a, y)
CM+F+ 15NH4+ Casuarina 0.048 ± 0.001(b, z)1.868 ± 0.057(b, z)0.466 ± 0.042(b, z) Eucalyptus 0.014 ± 0.001(b, x)0.561 ± 0.006(b, x)0.194 ± 0.018(a, x)
→ EM+ 15NO3 Casuarina 0.116 ± 0.005(a, z)4.195 ± 0.043(a, z)1.070 ± 0.062(a, z) Eucalyptus 0.031 ± 0.003(a, x)0.626 ± 0.022(a, x)0.212 ± 0.011(a, x)
EM+ 15NH4+ Eucalyptus 0.081 ± 0.006(b, z)0.950 ± 0.027(b, z)0.315 ± 0.010(b, z) Casuarina 0.008 ± 0.001(b, x)0.413 ± 0.012(a, x)0.096 ± 0.009(a, x)
CM+F+ 15NO3 Eucalyptus 0.121 ± 0.013(a, z)1.386 ± 0.074(a, z)0.458 ± 0.022(a, z) Casuarina 0.025 ± 0.005(a, x)0.301 ± 0.016(b, x)0.087 ± 0.008(a, x)

Roots had greater 15N content than shoots for both plant species, regardless of N source or species of the donor (Table 4). Casuarina had greater shoot and root 15N content than Eucalyptus, except in the NodMyc/Myc treatment. As receivers, Eucalyptus generally had greater 15N content than Casuarina. As donors, both Casuarina and Eucalyptus had significantly greater total (shoot plus root) 15N content when supplied with 15NO3 than with 15NH4+ for all three treatments. For Casuarina donors, whether 15NH4+ or 15NO3 was applied, total 15N content was low in NodMyc; greater in NodMyc+ and greatest in Nod+Myc+ treatment. Total 15N content in Eucalyptus donors was decreased from Myc treatment to Myc+ treatment when 15NH4+ was used. Total 15N content in Myc+Eucalyptus donors was lowest when Nod+Myc+Casuarina was the partner and when 15NO3 was applied. In addition, Casuarina donors had greater total 15N content than Eucalyptus donors in all three treatments, while only Nod+Myc+Casuarina receivers had greater total 15N content than Myc+Eucalyptus receivers. For both Casuarina and Eucalyptus receivers, regardless of the 15N source, total 15N content increased as follows: NodMyc/Myc < NodMyc+/Myc+ < Nod+Myc+/Myc+ treatments. As receivers, Casuarina and Eucalyptus had greater total 15N content only when supplied with 15NO3 in the NodMyc/Myc treatment. Casuarina and Eucalyptus receivers had greater total 15N content in the NodMyc+/Myc+ and in the Nod+Myc+/Myc+ treatment when 15NH4+ was supplied.

Table 4. Total 15N content of Casuarina and Eucalyptus seedlings grown in two-chambered pots
Codes 15N source 15N donor 15N (mg per plant) 15N receiver 15N (mg per plant)
ShootsRootsTotal (weighted)ShootsRootsTotal (weighted)
  1. All plants were fertilized with 14NH414NO3 for 11 months, then donor plants were treated with 15NH414NO3 or 14NH415NO3 for 1 month.

  2. Data are means ± SE (n = 9). For each parameter values followed by different letters designate significant differences (P ≤ 0.05) between 15N forms for a given species and a given treatment within columns (a, b, c), and between treatments for a given 15N form and a given species within columns (x, y, z), respectively.

  3. C, Casuarina; E, Eucalyptus; M, mycorrhizal status; F, nodulation status.

CM–F– 15NH4+ Casuarina 0.374 ± 0.040(b, z) 5.204 ± 0.362(b, y) 5.578 ± 0.402(b, y) Eucalyptus 0.005 ± 0.001(a, z)0.011 ± 0.002(b, z)0.016 ± 0.002(b, z)
→ EM– 15NO3 Casuarina 0.608 ± 0.044(a, z) 9.216 ± 0.501(a, z) 9.824 ± 0.545(a, z) Eucalyptus 0.009 ± 0(a, z)0.090 ± 0.007(a, z)0.099 ± 0.007(a, z)
EM– → 15NH4+ Eucalyptus 0.796 ± 0.102(b, x) 1.108 ± 0.049(b, xy) 1.904 ± 0.146(b, x) Casuarina 0.004 ± 0.001(a, z)0.013 ± 0.002(b, z)0.017 ± 0.002(b, z)
CM–F– 15NO3 Eucalyptus 0.910 ± 0.109(a, x) 1.611 ± 0.046(a, y) 2.520 ± 0.139(a, x) Casuarina 0.003 ± 0.001(a, z)0.030 ± 0.004(a, z)0.033 ± 0.005(a, z)
CM+F–  15NH4+ Casuarina 0.465 ± 0.053(b, y) 5.484 ± 0.473(b, y) 5.949 ± 0.500(b, y) Eucalyptus 0.012 ± 0.001(a, y)0.270 ± 0.015(a, y)0.282 ± 0.016(a, y)
→ EM+ 15NO3 Casuarina 0.933 ± 0.010(a, y)10.838 ± 0.900(a, y)11.771 ± 0.834(a, y) Eucalyptus 0.016 ± 0(a, y)0.226 ± 0.008(b, y)0.242 ± 0.008(b, y)
EM+ → 15NH4+ Eucalyptus 0.372 ± 0.048(b, y) 1.204 ± 0.078(b, x) 1.576 ± 0.118(b, y) Casuarina 0.008 ± 0.001(b, y)0.119 ± 0.013(a, y)0.127 ± 0.014(a, y)
CM+F– 15NO3 Eucalyptus 0.656 ± 0.030(a, y) 2.107 ± 0.107(a, x) 2.763 ± 0.135(a, x) Casuarina 0.022 ± 0.004(a, y)0.074 ± 0.007(b, y)0.096 ± 0.005(b, y)
CM+F+  15NH4+ Casuarina 0.541 ± 0.022(b, x) 6.854 ± 0.025(b, x) 7.395 ± 0.751(b, x) Eucalyptus 0.055 ± 0.006(a, x)0.766 ± 0.174(a, x)0.821 ± 0.175(a, x)
→ EM+ 15NO3 Casuarina 1.252 ± 0.149(a, x)13.272 ± 0.717(a, x)14.524 ± 1.807(a, x) Eucalyptus 0.052 ± 0.003(a, x)0.553 ± 0.083(b, x)0.605 ± 0.086(b, x)
EM+ → 15NH4+ Eucalyptus 0.166 ± 0.020(b, z) 1.051 ± 0.112(b, y) 1.217 ± 0.128(b, z) Casuarina 0.053 ± 0.011(b, x)0.922 ± 0.100(a, x)0.975 ± 0.104(a, x)
CM+F+ 15NO3 Eucalyptus 0.349 ± 0.078(a, z) 1.753 ± 0.244(a, y) 2.102 ± 0.211(a, y) Casuarina 0.170 ± 0.016(a, x)0.511 ± 0.029(b, x)0.681 ± 0.030(b, x)

Nitrogen transfer between paired Casuarina/Eucalyptus seedlings

Data on N transfer were calculated from 15N at % excess (Table 3) and 15N content (Table 4) of donor and receiver plants (Johansen & Jensen, 1996; see also He et al., 2004). Nitrogen transfer was expressed in three ways: percentage N transfer (Table 5); percentage N in the receiver derived from transfer (%NDFT) (Table 5); and the total amount of N transferred (mg per plant) (Fig. 4). Two-way N transfer occurred in both nonmycorrhizal and mycorrhizal treatments, regardless of 15N source. In the nonmycorrhizal (NodMyc/Myc) treatment, percentage N transfer was very low (<1.3, Table 5). In the nonnodulated mycorrhizal (NodMyc+/Myc+) treatment, percentage N transfer was 2.2–3.3% greater when 15NH4+ was supplied, and 5.0–11% greater when 15NO3 was supplied. In the nodulated mycorrhizal (Nod+Myc+/Myc+) treatment, percentage N transfer was much greater and ranged from 5.0 to 10% or from 24 to 39% when either 15NH4+ or 15NO3 was applied. Mycorrhizal treatments had greater percentage N transfer when 15NH4+, rather than 15NO3, was supplied. Percentage N transfer was greater from Eucalyptus to Casuarina than from Casuarina to Eucalyptus in all three treatments, whether 15NH4+ or 15NO3 was applied.

Table 5. N transfer (%) and NDFT (%) in Casuarina/Eucalyptus pairs grown in two-chambered pots
Treatments 15N sourceDirection of transfer C → ERatio 15NH4+/15NO3Direction of transfer E → CRatio 15NH4+/15NO3Ratio C → E/E → C
  1. Plants were fertilized with 14NH414NO3 for 11 months, and then with 15NH414NO3 or 14NH415NO3 (N donor only) for 1 month before harvest.

  2. Data are means ± SE (n = 9). For each parameter values followed by different letters designate significant differences (P ≤ 0.05) between 15N forms for a given pair and a given treatment within columns (a, b, c), between treatments for a given 15N form and a given pair within columns (x, y, z), and between directions of transfer in pairs for a given N form and a given treatment within rows (α, β, γ), respectively. C, Casuarina; E, Eucalyptus; M, mycorrhizal status; F, nodulation status.

% N transfer
CM–F– ↔ EM– 15N-ammonium 0.3 ± 0(b, z, α)0.3 0.9 ± 0.2(b, z, α)0.73.0
15N-nitrate 1.0 ± 0.1(a, z, α)  1.3 ± 0.2(a, z, α) 1.3
CM+F– ↔ EM+ 15N-ammonium 3.3 ± 0.9(a, y, β)1.810.5 ± 2.6(a, y, α)2.12.3
15N-nitrate 2.2 ± 0.2(b, y, β)  5.1 ± 1.2(b, y, α) 2.0
CM+F+ ↔ EM+ 15N-ammonium10.1 ± 2.6(a, x, β)1.939.1 ± 5.5(a, x, α)1.73.8
  15N-nitrate 5.3 ± 1.3(b, x, β) 23.6 ± 5.9(b, x, a) 4.5
% NDFT (N in the receiver derived from transfer)
CM–F– ↔ EM– 15N-ammonium 1.6 ± 0.1(b, z, α)0.6 0.5 ± 0.1(a, z, β)0.70.3
15N-nitrate 2.7 ± 0.2(a, z, α)  0.7 ± 0(a, z, β) 0.3
CM+F– ↔ EM+ 15N-ammonium 6.0 ± 0.5(a, y, α)1.0 4.3 ± 0.9(a, y, β)1.20.7
15N-nitrate 6.3 ± 0.2(a, y, α)  3.6 ± 0.1(a, y, β) 0.6
CM+F+ ↔ EM+ 15N-ammonium32.1 ± 7.9(a, x, α)1.129.1 ± 6.4(a, x, α)1.00.9
15N-nitrate28.3 ± 1.5(a, x, α) 30.2 ± 1.5(a, x, α) 1.1
Figure 4.

Nitrogen transfer (a,b) and net N transfer (c,d) in Casuarina (C)/Eucalyptus (E) pairs as affected by mycorrhizas (M), nodulation (F), N source, and species of N donor or N receiver. Plants were fertilized with 14NH414NO3 for 11 months and with 15N (N donor only) for 1 month before harvest. Values (means ± SE, n = 9 or 18) followed by different letters designate significant differences (P = 0.05) between treatments for a given 15N form and a given pair (a, b, c); between 15N forms for a given pair and a given treatment (x, y, z); and between directions of transfer in pairs for a given 15N form and a given treatment (α, β, γ), respectively.

Nitrogen transfer was also calculated as %NDFT, an estimate of transferred N relative to total N in the receiver. Among treatments, %NDFT was: NodMyc/Myc (0.5–2.7%) < NodMyc+/Myc+ (3.6–6.3%) < Nod+Myc+/Myc+ (28–32%) treatment, when either 15NH4+ or 15NO3 was supplied (Table 5). %NDFT was similar in nonmycorrhizal and mycorrhizal treatments, independent of 15N source. Regardless of N source, there was greater %NDFT from Casuarina to Eucalyptus than from Eucalyptus to Casuarina in all three treatments, except in the Nod+Myc+/Myc+ treatment with 15NO3 as the N source.

We also calculated the amount of N transferred between Casuarina/Eucalyptus pairs (Fig. 4a,b). Regardless of N source, the amount of N transferred was low (1–4 mg per plant) in the NodMyc/Myc treatment, then increased (17–44 mg per plant) in the NodMyc+/Myc+ treatment, and was much greater (120–420 mg per plant) in the Nod+Myc+/Myc+ treatment. The amount of N transferred was greater in mycorrhizal (NodMyc+/Myc+, 42–44 vs 17–19 mg per plant) and (Nod+Myc+/Myc+, 238–420 vs 120–250 mg per plant) treatments when 15NH4+ was supplied, compared with 15NO3. In both nonmycorrhizal (NodMyc/Myc) and mycorrhizal (NodMyc+/Myc+) treatments, the amount of N transferred was similar for either species as donor. However, in the nodulated mycorrhizal (Nod+Myc+/Myc+) treatment, a substantially greater amount of N was transferred from Eucalyptus to Casuarina than from Casuarina to Eucalyptus.

Net nitrogen transfer in Casuarina/Eucalyptus pairs

Our experiments were designed to enable us to calculate net N transfer in paired Casuarina/Eucalyptus seedlings (Fig. 4c,d). As expected, the amounts of net N transfer were quite low (0.7–0.8 mg per plant) in the nonnodulated nonmycorrhizal (NodMyc/Myc) treatment. Net N transfer was greater (1.8–1.9 mg per plant) in the nonnodulated mycorrhizal (NodMyc+/Myc+) treatment, and much greater (127–185 mg per plant) in the nodulated mycorrhizal (Nod+Myc+/Myc+) treatment. In both NodMyc/Myc and NodMyc+/Myc+ treatments, amounts of net transfer were similar regardless of whether 15NH4+ or 15NO3 was supplied. In contrast, in the Nod+Myc+/Myc+ treatment, amounts of net transfer were greater when 15NH4+, rather than 15NO3, was supplied. In both NodMyc/Myc and NodMyc+/Myc+ treatments, Casuarina was the sink plant for N acquisition when NH4+ was supplied, while Eucalyptus was the N sink plant when 15NO3 was supplied. Unexpectedly, Casuarina rather than Eucalyptus was the N sink plant in the Nod+Myc+/Myc+ treatment, when either 15NH4+ or 15NO3 was applied.

Discussion

Biological nitrogen fixation

In this study, N2 fixation rate was 38% for nodulated 11-month-old C. cunninghamiana grown with 14NH414NO3. This rate was greater (9%) than that in our previous experiment with 5-month-old C. cunninghamiana grown on (14NH4)2SO4 or K14NO3 (He et al., 2004). Our N2 fixation data were comparable with data reported for Casuarina by Mariotti et al. (1992); Parrota et al. (1996) and Subbarao & Rodriguez-Barrueco (1995), but were different from those of Reddell & Bowen (1985) and Reddell et al. (1988, 1997). We used needles to analyse 15N natural abundance and to calculate the percentage of plant N supplied by N2 fixation (Boddey et al., 2000; Unkovich et al., 2001). Reddell and coworkers calculated the amount of N in the nodule that was derived from fixation as mg N fixed per mg nodule d. wt.

Mycorrhizal colonization and formation of common mycorrhizal networks

Mycorrhizal colonization was similar for 12-month-old (this study) and 6-month-old Casuarina or Eucalyptus seedlings (He et al., 2004). Other reports of EM colonization were less: 10–40% in Casuarina (Gardner, 1986; Theodorou & Reddell, 1991; Subbarao & Rodriguez-Barrueco, 1995) and 10–60% in Eucalyptus (Burgess et al., 1993; May & Simpson, 1997; Reddell et al., 1999; Chen et al., 2000). Aseptic conditions and suitable glasshouse environments may have favoured colonization in our studies. Nodulation of Casuarina did not increase mycorrhizal colonization, and was consistent with results of Gardner (1986) and Subbarao & Rodriguez-Barrueco (1995). Similar to our previous report (He et al., 2004), we observed living fungal hyphae growing on and through the nylon mesh in the two-chambered pots, indicating that EM networks formed between Casuarina and Eucalyptus. No mycorrhizal hyphae were observed in the NodMyc/Myc treatment.

Effects of mycorrhizal colonization and nodulation on plant growth

Biomass of both plant species in Casuarina/Eucalyptus pairs was greater when plants were mycorrhizal, and then greatest when Casuarina was both mycorrhizal and nodulated. Compared with the NodMycCasuarina, N content was twice as high in the NodMyc+ and four times as high in the Nod+Myc+Casuarina. Compared with the MycEucalyptus, N content of Eucalyptus increased 1.5 times in the NodMyc+/Myc+ and twice in the Nod+Myc+/Myc+ treatment. Our data agree with previous reports that the main effect of mycorrhization on plant N status was in N2-fixing plants (Hayman, 1986; Vassilev et al., 2001). Common host genes could potentially enhance the formation of mycorrhizal associations, and are required for both nodulation and mycorrhizal formation in N2-fixing plants (Albrecht et al., 1999; Anéet al., 2004) for their high N-demanding requirements (McKey, 1994; Sprent, 1994; Sprent, 2001). About twice as much N was transferred from Myc+Eucalyptus to Nod+Myc+Casuarinas than the reverse, resulting in increased growth by Casuarina caused by interspecific N transfer. Our data support the hypothesis that N2-fixing plants may have evolved in an N-rich environment (McKey, 1994; Stewart & Schmidt, 1999), and are consistent with the observation that a non-N2-fixing plant (Baccharis pilularis) facilitates the establishment of an N2-fixing plant (Lupinus arboreus) (Rudgers & Maron, 2003). Furthermore, regardless of whether Casuarina or Eucalyptus was the donor or the receiver, and whether 15NH4+ or 15NO3 was the 15N source, seedlings in the mycorrhizal treatments were taller, and had larger root systems than those in the nonmycorrhizal treatment (He, 2002). Similar results were obtained by others (Smith & Read, 1997; Jonsson et al., 2001). For both donors and receivers, both biomass and N content in Casuarina and Eucalyptus were similar, suggesting N transfer through CMNs from donors to receivers. This indicates that CMNs can provide direct conduits for nutrient exchange between plants, and that CMNs are effective in supplying growth-supporting amounts of nutrients to a receiver plant. The increased growth of recipients suggests that they benefited physiologically from mycorrhizal-mediated nutrient transfer between plants.

Two-way nitrogen transfer between plants and its potential ecological significance

Roots had greater 15N at % excess and 15N content than shoots for both plant species, whether Casuarina or Eucalyptus was the donor or the receiver, and whether 15NH4+ or 15NO3 was the 15N source. Both 15N at % excess and 15N content were greater in the 12-month-old seedlings than in the 6-month-old ones (He et al., 2004). Our results demonstrated two-way N transfers, generated from the 15N content of both donor and receiver plants, between 12-month-old paired Casuarina and Eucalyptus seedlings supplied with either 15NH4+14NO3 or 14NH4+15NO3. In the nonnodulated nonmycorrhizal (NodMyc/Myc) treatment, amounts of N transfer were less than 2% of the total plant N. This low N transfer may have occurred from diffusion along the interface between mesh and perforated Perspex barriers, or perhaps from leaf absorption of volatilized N-containing compounds (Frank et al., 2004). In the nonnodulated mycorrhizal (NodMyc+/Myc+) treatment, amounts of N transfer were 16% of total plant N. This N transfer was enhanced by mycorrhizal interconnections. In the nodulated mycorrhizal (Nod+Myc+/Myc+) treatment, much greater amounts of N were transferred, up to 30% of the total plant N. Thus mycorrhizas and Frankia together had the greatest positive effect on N flux between nodulated mycorrhizal Casuarina and mycorrhizal Eucalyptus. Furthermore, significantly greater amounts of N were transferred from non-N2-fixing mycorrhizal (NodMyc+) Eucalyptus to nodulated mycorrhizal (Nod+Myc+) Casuarina than the reverse. These data were consistent with our previous study with 6-month-old Casuarina/Eucalyptus pairs. Thus our two-way N transfer results only partially support the hypothesis of Bethlenfalvay et al. (1991) and Frey & Schuepp (1993) that one-way N flows from legume donors to nonlegume receivers.

Nitrogen transfer between legumes and nonlegumes can take place by indirect or direct routes (McNeill & Wood, 1990). Indirect N transfer occurs through recapture of N-containing materials from rhizodeposition (Stern, 1993; Dubach & Russelle, 1994; Chalk, 1996; Paynel et al., 2001; Walker et al., 2003), although rhizodeposition is regulated by interactions between plants and mycorrhizas (Jones et al., 2004). Direct N transfer occurs when nutrients are obtained through mycorrhizal hyphae where CMNs connect plants (Newman, 1988; Frey & Schuepp, 1993; He et al., 2003). To eliminate transfer through indirect pathways, we used fine mesh screen and air gaps to exclude root contact and water movement, and to allow passage of hyphae. Similar techniques have been used previously in other studies to demonstrate one-way N transfer via CMNs (Bethlenfalvay et al., 1991; Frey & Schuepp, 1992, 1993; Arnebrant et al., 1993; Johansen & Jensen, 1996).

We investigated two-way or net N transfer as the difference from Casuarina to Eucalyptus and from Eucalyptus to Casuarina. Net N transfer was greater in mycorrhizal treatments when supplied with 15NH4+ than with 15NO3. Independent of N source, net transfer was very low (<1 mg per plant) in the nonnodulated nonmycorrhizal treatment. Mycorrhizas doubled this amount (≈2 mg per plant) in the nonnodulated mycorrhizal treatment, but the greatest increase (127–185 mg per plant) was in the combination of mycorrhizas and nodulation treatment (Fig. 4c,d). Amounts of net N transfer from 12-month-old pairs supplied with 15NH414NO3 or 14NH415NO3 was greater than that from 6-month-old pairs supplied with (15NH4)2SO4 or K15NO3 (He et al., 2004). In the present study, for both the NodMyc/Myc and the NodMyc+/Myc+ treatment, 12-month-old Casuarina was the N sink plant when 15NH4+ was supplied, while 12-month-old Eucalyptus was the N sink plant when 15NO3 was supplied. However, for both NodMyc/Myc and NodMyc+/Myc+ treatments in our previous study, either the 6-month-old Casuarina or Eucalyptus was the N-sink plant, independent of 15NH4+ or 15NO3 (He et al., 2004). Both studies demonstrated a much greater net flow of N from mycorrhizal Eucalyptus to nodulated mycorrhizal Casuarina, suggesting that Casuarina was the stronger N sink in the Nod+Myc+/Myc+ treatment. Our data are consistent with the hypothesis proposed by McKey (1994) that N fixation by legumes may have evolved to help fuel an N-rich lifestyle. We found only one other study that reported <1% N transfer from non-N2-fixing barley to N2-fixing pea connected by an AM fungus, G. intraradicesJohansen & Jensen (1996). Other two-way N-transfer studies did not report effects of mycorrhizas, and N transfer was insignificant from grass to legume (Brophy et al., 1987; Tomm et al., 1994; Hogh-Jensen & Schjoerring, 2000) or from rice to peanut (Shen & Chu, 2004).

We assume that the following scenario may exist in nature for plants that have different N requirements (McKey, 1994; Sprent, 1994; Sprent, 2001). Differences in net N transfer with different nodulation/mycorrhizal combinations could have important ecological implications both for nutrient cycling and for the structure and function of agricultural or natural plant communities, particularly with respect to those plants that can potentially construct a ‘wood-wide web’ or common mycorrhizal network to transfer nutrients underground. Our data support the hypothesis that mycorrhizal colonization can result in resource equalization or sharing (Read, 1997; Perry, 1998), thus reducing dominance of aggressive species and promoting coexistence and biodiversity (Read, 1997; van der Heijden et al., 1998). Benefits of N nutrition to non-N2-fixing plants by neighbouring N2-fixing plants have been reported in both agriculture (Stern, 1993; Chalk, 1996; Anon., 1998) and forests (Binkley, 1992; O’Connell & Grove, 1996; Binkley et al., 2003). However, non-N2-fixing plants may also benefit N2-fixing plants by increasing germination, survival and growth; such facilitation could benefit other members in a plant community (Rudgers & Maron, 2003). From an evolutionary perspective, plants may evolve with increased competitive ability for resource requirements, and may respond differently to a more nutrient-rich condition. For instance, an N2-fixing plant may not exploit its capacity to fix atmospheric N2, but may use soil available N (McKey, 1994; Marschner, 1995), especially early in establishment, and thus become relatively N-poor (receivers). In the meantime, plants adapted to extremely nutrient-low environments, especially those with an evolving capacity to live in nutrient-impoverished forests and woodlands (Attiwill & Leeper, 1987; Stewart & Schmidt, 1999), such as many native Australian plants, may become relatively N-rich (donors). If such provisional N-rich and N-poor plants become neighbours, they may alter their donor and receiver behaviour and N-transfer may switch from the low-N-demand non-N2-fixing plant to the high-N-demand N2-fixing plant in order to satisfy the latter's higher N requirements.

Nitrogen transfer between plants: ammonium vs nitrate

Patterns of 15N distribution and N transfer between plants in this study were similar to patterns from our previous study using single (15NH4)2SO4 or K15NO3 (He et al., 2004). As NH4+ and NO3 have different chemical mobility in soil, and their uptake by plants and fungi may be different, it is important to compare effects of NH4+ and NO3 on mycorrhizal mediation of N transfer between plants. In this study we were able to compare differences in N transfer by fertilizing plants with the same N-containing fertilizer (14NH414NO3) but monitoring 15N movement between plants by labelling distinctive 15N labels (15NH4+14NO3 and 14NH4+15NO3).

In this study our data showed that more N was transferred when plants were supplied with the cation 15NH4+ than with the anion 15NO3, although NO3 is more mobile in plants and in soil (Fig. 4a,b). This effect of inorganic form of 15N source on differences in amount of N transfer supplied with 15NH4+ vs 15NO3 was modest (−2.5 less to Eucalyptus vs −1.0 mg per plant less to Casuarina) in the nonnodulated nonmycorrhizal/nonmycorrhizal (NodMyc/Myc) treatment; greater (+23 more to Eucalyptus vs +27 mg per plant more to Casuarina) in the nonnodulated mycorrhizal/mycorrhizal (NodMyc+/Myc+) treatment; and substantially greater (+118 more to Eucalyptus vs +175 mg per plant more to Casuarina) in the nodulated mycorrhizal/mycorrhizal (Nod+Myc+/Myc+) treatment.

Was the applied 15NH4+ transformed into 15NO3? We were concerned that the concentration of NO3 might increase because of rapid nitrification of NH4+ by microbes (Paul & Clark, 1996; Brady & Weil, 2002). To prevent nitrification in the soil, a nitrification inhibitor, N-serve, was added (5 mg kg−1 media). N-serve retards the first step in nitrification of NH4+ to hydroxylamine by Nitrosomonas sp. (Slangen & Kerkhoff, 1984). We did not observe any toxic effects of N-serve on plant growth or mycorrhizal development in this study, or in our earlier study (He et al., 2004). Our observations agree with other studies. Chambers et al. (1980) showed that there was little or no effect of N-serve on mycorrhizal development under high concentration of 15 mg kg−1 soil. In addition, N-serve had little effect on other soil microorganisms when 10 mg kg−1 soil was used (Laskowski et al., 1975). We were also concerned that the concentration of NH4+ and NO3 might be changed because of leaking out of the experimental system. Our experimental pots were closed, so there was no NH4+ or NO3 leaching. Thus 15NH4+ or 15NO3 concentrations should be maintained at the same level, whether 15NH4+ or 15NO3 was the 15N source.

Why was more N transferred when 15NH4+ was supplied, although NO3 is more mobile? We suggest that the following three factors could be responsible. First, the cation NH4+ is bound to negatively charged sites of clay lattices in soil, reducing mobility and leading to reduced availability, so that diffusion to the root surface becomes the rate-limiting step (Marschner, 1995; Tinker & Nye, 2000; Brady & Weil, 2002). Therefore plants may require more extensive root systems to access NH4+ (Raven et al., 1992; Engels et al., 2000), such as mycorrhizal roots (Marschner & Dell, 1994; Smith & Read, 1997) or cluster roots in both Casuarina (Reddell et, 1986; Diem et al., 2000; Neumann & Martinoia, 2002) and Eucalyptus species (Grierson & Adams, 2000; Neumann & Martinoia, 2002). Both mycorrhizas and cluster roots can enhance nutrient acquisition by increasing root surface area and by physiological adaptations (Marschner, 1995; Diem et al., 2000; Neumann & Martinoia, 2002). However, cluster roots, which are adapted to habitats of extremely low soil fertility as an effective nutrient-uptake mechanism, were not observed in either Casuarina or Eucalyptus in our experiments, as suitable N-nutrition and glasshouse growth conditions might not favour their formation. In contrast, the anion NO3, which is not bound to negatively charged soil particles, is more mobile in the soil, and mass flow to the root surface becomes the rate-limiting step (Marschner, 1995; Tinker & Nye, 2000; Brady & Weil, 2002). Theoretically, NO3 in soils could be extracted by a nonmycorrhizal root system assuming high fine-root density. Thus mycorrhizas would add little advantage to NO3 uptake (Tinker & Nye, 2000). However, NH4+ is often the dominant inorganic N form in Australian forest soils (Adams & Attiwill, 1986; Attiwill & Leeper, 1987; Connell et al., 1995). Both Eucalyptus (Shedley et al., 1995; Barros & Novais, 1996) and Casuarina (Torrey, 1983; Subbarao & Rodriguez-Barrueco, 1995) have mycorrhizal association and inherently prefer NH4+ rather than NO3. Therefore it is possible that N uptake and translocation could be promoted by mycorrhizal colonization in soils where NH4+ is the main N form. Our data from both mycorrhizal and the nodulated mycorrhizal treatments support this hypothesis, because a greater amount of N was transferred to Casuarina or to Eucalyptus when 15NH4+, rather than 15NO3, was the 15N source.

Second, increased ammonium transfer may be related to differences in uptake and assimilation of 15NH4+ and 15NO3 by mycorrhizal fungi. Mycorrhizas play important roles in N nutrition of trees by increased uptake at the root surface and by extensive mycelial growth and soil exploration (Smith & Read, 1997). Most EM fungi can assimilate NH4+ (Jongbloed et al., 1991; Martin & Plassard, 2001), but only some can reduce NO3 (Sarjala, 1990; Scheromm et al., 1990; Martin & Plassard, 2001). EM fungal mycelia are abundant in humus and litter layers in soils where NH4+ concentrations are greater than NO3 (Alexander, 1983, 1989; Read, 1992). Thus hyphae might be effective absorbing structures for NH4+ in ectomycorrhizal roots. EM associations may enhance NH4+ uptake (Bledsoe & Zasoski, 1983; Littke et al., 1984; Rygiewicz et al., 1984; Javelle et al., 1999; Plassard et al., 2000). Disruption of external mycelia can decrease NH4+ uptake (Javelle et al., 1999). However, the influence of EM associations on NO3 uptake is controversial. NO3 uptake can be either increased (Bledsoe & Zasoski, 1983; Gobert & Plassard, 2002; Plassard et al., 2002) or reduced (Kreuzwieser et al., 2000; Plassard et al., 2002). In general, NH4+ uptake was greater than NO3 uptake by ectomycorrhizas (Bowen & Smith, 1981; Littke et al., 1984; Kreuzwieser et al., 2000). Uptake rates for NH4+ were greater than for NO3 for the EM fungus Hebeloma crustuliniforme (Littke et al., 1984). Lower NO3 uptake rates by mycorrhizal seedlings indicate that mycorrhizas may play a less important role in accessing more mobile NO3. However, nothing is currently known about its N-uptake ability for the Pisolithus species used in our experiment. We only knew that all the 37 isolates from three unidentified Australian Pisolithus species utilized a range of amino acids, and most of them showed a preference for NH4+ rather than NO3 (Anderson et al., 1999).

In our study, high water-holding-capacity crystals were added to the potting mix in order to minimize diffusion and mass flow. In this way, roots could obtain water while the growth media remained relatively dry. Nutrient uptake is inhibited in dryer soils as decreased soil water leads to a decrease in rates of mass flow and diffusion, and consequently to reduced rates of nutrient transport to root surfaces (Marschner, 1995; Tinker & Nye, 2000; Brady & Weil, 2002). External mycelia might be of great importance under drier conditions because they occupy a larger soil volume for greater access to nutrients (Davies et al., 1992; Smith & Read, 1997). Mycorrhizal hyphae may thus play a greater role in transport of less mobile NH4+ than in transport of more mobile NO3 to plant roots, except when the mobility of NO3 from soil to root surface is reduced in water-deficient soil (Tobar et al., 1994; Wu et al., 1999). Nevertheless, information about the water-use capacity of the mycorrhizal fungus Pisolithus sp. is needed.

The third factor may be associated with differences in energy required to metabolize NH4+ and NO3. Uptake and assimilation of inorganic NH4+ or NO3 by mycorrhizal fungi or by plants is a prerequisite to transfer of assimilated N between plants. For conversion of inorganic to organic N, NH4+ is directly assimilated by plants, whereas NO3 must first be reduced to NO2 and then to NH4+ in roots and/or shoots before it can be assimilated. Reduction of NO3 requires energy, C and protons (Marschner, 1995; Forde & Clarkson, 1999). Similar assimilatory mechanisms occur in mycorrhizal fungal mycelia (Smith & Read, 1997; Hawkins et al., 2000; Martin & Plassard, 2001). As a consequence, NH4+ fluxes within fungal and plant tissues, or between plants through hyphal connections, might require less energy and be cost effective, leading to more rapid fluxes or translocations of NH4+ than NO3. We also found one report that NO3 could be toxic to Eucalyptus (Shedley et al., 1995), and two other reports that NO3 inhibited root-hair development in Casuarina (Kohls & Baker, 1989; Arnone et al., 1994). The presence of NH4+ often depresses NO3 uptake when both are supplied together, so many plants appear to take up NH4+ preferentially (Marschner et al., 1991; Paul & Clark, 1996; Kreuzwieser et al., 1997; Nordin et al., 2001). AM-mediated one-way N transfer (15NH4+ alone or 15NO3 alone) or EM-mediated one-way N transfer (15NH4+ alone) from N2-fixing to non-N2-fixing plants has been reported (Bethlenfalvay et al., 1991; Frey & Schuepp, 1992, 1993; Arnebrant et al., 1993; Ekblad & Huss-Danell, 1995; Johansen & Jensen, 1996; Martensson et al., 1998). No previous studies compared N transfer when plants were supplied with ammonium nitrate labelled as 15NH414NO3 and as 14NH415NO3.

In summary, biomass and N content were increased by mycorrhization. Mycorrhizal networks have the capacity to mediate significant N transfer among interconnected plants. Net N transfer occurred between mycorrhizal Casuarina/Eucalyptus pairs. Our results showed that, compared with the nonnodulated nonmycorrhizal pairs, amounts of N transfer were greater in the nonnodulated mycorrhizal pairs, and substantially greater in the dual-nodulated mycorrhizal/mycorrhizal pairs, whether 15NH4+ or 15NO3 was applied. Nitrogen transfer was greater between mycorrhizal pairs when 15NH4+, rather than 15NO3, was supplied. Unexpectedly, we measured much greater net N transfer to Casuarina than to Eucalyptus when Casuarina was both nodulated and mycorrhizal, and Eucalyptus was also mycorrhizal. Mycorrhizas alone had less impact on N accumulation in non-N2-fixing Eucalyptus than in N2-fixing Casuarina, despite the fact that Eucalyptus had about twice the mycorrhizal colonization of Casuarina. When Casuarina was both nodulated and mycorrhizal, and when its partner Eucalyptus also became mycorrhizal through hyphal interconnections, biomass and N content were maximized in both plant species; plant performance was similar between donors and receivers through mycorrhizal-mediated N transfer. Our data agree with previous reports that the main effect of mycorrhization on plant N status was in N2-fixing plants (Hayman, 1986; Vassilev et al., 2001). Our data also support the hypothesis that N2-fixing species are high N-demand plants, and N fixation may have evolved to help fuel an N-rich lifestyle (McKey, 1994; Sprent, 1994, 2001). This suggests that mycorrhization, together with N availability, plays a vital role in N redistribution between plants, resulting in resource sharing or facilitation. Thus N transfer could be important for plants that encounter either N excess or N limitation (Vitousek et al., 1997; Anon., 1998; Moffat, 1998). Nitrogen transfer might also reduce the success of invasive species and promote coexistence and biodiversity (Read, 1997; van der Heijden et al., 1998). This may have practical consequences for environmental restoration or phytoremediation (Franco et al., 1997; Pattinson et al., 2004; Pennisi, 2004). For these reasons, more research is warranted on two-way N transfers mediated by mycorrhizal networks.

Acknowledgements

For valuable comments on the manuscript, we thank three anonymous reviewers and Dr Alastair H. Fitter, the New Phytologist editor; and Dr D. Southworth, Department of Biology, Southern Oregon University, Ashland, OR, USA. This research was supported by Postgraduate Research Scholarships from the Department of Education, Science and Training (DEST), Canberra, and the University of Queensland, Brisbane, Australia; and a Postdoctoral Fellowship to X. He from NSF grant ‘Biocomplexity in the Environment: Phase I’ (DEB-9981711 to Dr C. S. Bledsoe), USA. Dr He is grateful to the Education Department of Yunnan Province, China for permission to study overseas.

Ancillary