Soil moisture variations affect short-term plant-microbial competition for ammonium, glycine, and glutamate

We tested whether the presence of plant roots would impair the uptake of ammonium (), glycine, and glutamate by microorganisms in a deciduous forest soil exposed to constant or variable moisture in a short-term (24-h) experiment. The uptake of 15NH4 and dual labeled amino acids by the grass Festuca gigantea L. and soil microorganisms was determined in planted and unplanted soils maintained at 60% WHC (water holding capacity) or subject to drying and rewetting. The experiment used a design by which competition was tested in soils that were primed by plant roots to the same extent in the planted and unplanted treatments. Festuca gigantea had no effect on microbial N uptake in the constant moist soil, but its presence doubled the microbial uptake in the dried and rewetted soil compared with the constant moist. The drying and rewetting reduced by half or more the uptake by F. gigantea, despite more than 60% increase in the soil concentration of . At the same time, the amino acid and -N became equally valued in the plant uptake, suggesting that plants used amino acids to compensate for the lower acquisition. Our results demonstrate the flexibility in plant-microbial use of different N sources in response to soil moisture fluctuations and emphasize the importance of including transient soil conditions in experiments on resource competition between plants and soil microorganisms. Competition between plants and microorganisms for N is demonstrated by a combination of removal of one of the potential competitors, the plant, and subsequent observations of the uptake of N in the organisms in soils that differ only in the physical presence and absence of the plant during a short assay. Those conditions are necessary to unequivocally test for competition.


Introduction
Observations of plant uptake of organic nitrogen (N), particularly amino acid N, in soils with large amounts of organic N but insufficient net mineralization rates to satisfy the annual N demand (e.g., Schimel and Chapin 1996;N€ asholm et al. 1998) have refueled the debate on the competitive interaction between plants and soil microorganisms (Kaye and Hart 1997). As plants from ecosystems with relatively high net N mineralization rates also take up amino acids (N€ asholm et al. 2009), it is possible that organic N may be another shared and mutually limiting N source over which plant-microbial competition may occur, especially during periods when inorganic N is scarce.
Competition for N may come out differently depending on the access to various N sources, their concentration in the soil (Bardgett et al. 2003), and other soil conditions, such as spatial distribution of roots and microorganisms (Wang and Bakken 1997;Hodge et al. 2000;Xu et al. 2011), decomposability of the soil C (M ansson et al. 2009), and soil moisture Lipson and Monson 1998) varies from 0.2 to 7.7 in microorganisms and from 0.6 to 2.1 in plants, in arctic systems (Schimel and Chapin 1996;Henry and Jefferies 2003) and temperate grasslands (Bardgett et al. 2003). In the latter system, the microbial uptake was dominated by glycine during plant growth in May but shifted to NH þ 4 in September, possibly in response to an increased pool of easily decomposable C as plant senescence started. This preference for NH þ 4 to glycine in microorganisms was also evident in their response to glucose addition to soil, which increased respiration, at the same time as grasses in the same soil reduced their uptake of glycine more than the uptake of NH þ 4 (Dunn et al. 2006). When soil moisture conditions are constant, microorganisms show no uptake discrimination between NH þ 4 and amino acids or prefer simple amino acids to NH þ 4 , whereas most grass species and shrubs seem to prefer NH þ 4 and glycine to more complex amino acids, although species differences exist (Weigelt et al. 2005;Harrison et al. 2007;Sørensen et al. 2008).
Short drying and rewetting cycles occur frequently during a growth period in temperate forests (Ladekarl 1998;Subke et al. 2003). When the soil is drying, bacteria accumulate osmotic active internal solutes, such as free amino acids and their derivates, to maintain the internal water potential in balance with the surrounding environment (Csonka and Hanson 1991). Following rewetting, concentrations of amino acids and mineralized N increase in the soil due to cell lysis (Marumoto et al. 1982;Pulleman and Tietema 1999) and microbial decomposition of soil organic matter (Van Gestel et al. 1993;Appel 1998;Lipson and Monson 1998).
As the soil is rewetted, the efflux of easily decomposable C from roots (Neuman and R€ omheld 2001), microorganisms (Marumoto et al. 1982), and soil organic matter (Pulleman and Tietema 1999) will increase. This easily available C may fuel fast-growing microorganisms (Bottner 1985;Van Gestel et al. 1993) and increase rates of microbial N mineralization and net immobilization of NH þ 4 Pulleman and Tietema 1999;Bengtsson et al. 2003;Kaiser et al. 2011), to the detriment of N uptake in plants. Under constant and high soil moisture conditions, fast-growing microorganisms will be less active (Bottner 1985;Van Gestel et al. 1993) and possibly open a window of competition with plants for limited quantities of mineralized N.
Plants can also respond rapidly to rewetting of a dry soil and restore the NH þ 4 and NO 3 À uptake within hours to a few days after moderate-to-severe drought (Brady et al. 1995;Cui and Caldwell 1997;Buljovcic and Engels 2001). Furthermore, organic N is potentially a more costeffective N source than inorganic N (Schmidt and Stewart 1999), e.g., by also supplying carbon to a plant that has been energy limited during drought (Raab et al. 1996).
According to the resource competition model (Burkholder 1952), competition for the same and limiting resource is assumed to bear a cost to all competitors. Both plants and microorganisms qualify for part of the definition, as they can use the same sources of N. However, just like removal, introduction or both of one or more potential competitor is a necessary criterion to demonstrate competition (Schoener 1983). Solely relying on uptake data, as in most work on plant-microorganism competition for nitrogen sources so far, is not a proof for plant-microbial competition for N. But uptake data will, in combination with the removal of one or the other group of organisms, be useful in testing the second part of the condition for resource competition, namely the limitation of the availability of N by one potential competitor to the other. Excluding the microorganisms from a competition assay in soil without changing the soil conditions is not an alternative, so we made an effort to exclude the plant and yet provide the same soil conditions as if the plant had been present.
To address the issue of competition for different N forms between plants and microorganisms under varying moisture regimes, we developed two hypotheses to test in this project: 1 Microorganisms will take up less NH þ 4 in the presence than in the absence of the plants in a constantly moist soil. 2 Immediately following rewetting of a dried soil, microorganisms increase their total uptake of N, especially NH þ 4 , in response to the increased N pool in the soil, while plants turn to an increased uptake of energy providing N sources, that is, glycine and glutamate. To test the hypotheses, plant-microbial competition for NH þ 4 , glycine and glutamate, which represent less and more complex amino acids, respectively, was studied in a 24 h assay, in which the 15 N uptake in the organisms was measured in planted soils and unplanted reference soils that had been either constantly moist or dried and rewetted over a short period of time. Dual labeled amino acids ( 13 C-15 N-amino acids) were used to estimate the uptake of intact amino acids by the plants. The soil ATP content was analyzed to estimate the soil microbial biomass and activity.

Soil
The soil was classified as dystric cambisol (FAO system; soil characteristics are given in Table 1) and was sampled at Torup, southern Sweden (55°33 0 N, 13°37 0 E). Oak (Quercus robur L., 80%) and beech (Fagus sylvatica L., 20%) were dominating the tree canopy, and the understory vegetation was sparse. Litter was removed within a 20 9 20 m square, and five soil samples, each covering a 0.5 9 0.5 m square, were randomly collected from the top 5 cm, sieved through a 4-mm mesh, and then pooled. The soil C:N ratio, pH (H 2 O), and water holding capacity (WHC) were determined. The soil was stored at 2°C until planting 2 weeks after sampling.

Plant material
We used Festuca gigantea L., a common grass in the type of oak-beech forest where the soil was sampled. F. gigantea grows relatively fast (RGR = 1.0-1.4 week À1 ; Grime et al. 1988) and is often found on soils with relatively high N mineralization and nitrification rates, as indicated on a 1-9 scale by its Ellenberg N value of 6 (Ellenberg et al. 1991) and FNIS value of 8 (calculated from data in Diekmann and Falkengren-Grerup 1998). FNIS values are taking both the total amount of mineral N and the amount of NO 3 À into consideration, that is, when two soils have the same concentration of total mineral N, the soil with the highest NO 3 À concentration will get the highest FNIS value. F. gigantea is most often found to be nonmycorrhizal (Harley and Harley 1987). The plants were grown from seeds in >99% pure silica sand supplied with a nutrient solution (including 150 lmol/L NO 3 À and 100 lmol/L NH þ 4 ) for one to 2 weeks after germination and then planted in the experimental soils. Seedlings included in the experiment had a shoot height of 7-10 cm and root length of 5-8 cm.

Experimental design and 15 N addition in the competition assay
Before planting, the field moist soil was wetted with deionized water to obtain 60% WHC 1 day before planting. Transparent plastic pots with a diameter of 27 mm and a height of 100 mm were filled with 84 g of soil (62 g dry weight soil). The pots were placed in a greenhouse at a temperature of 20°C day/16°C night and 16 h of daylight (with additional light of 160 lmol m À2 s À1 during the day). Sprinklers maintained the relative air humidity at about 50%. The pots were kept at 60% of WHC by addition of deionized water.
We developed a new competition assay method, in which also the soil that was used as unplanted in the assay was planted prior to the assay, to ensure that the effect of plant presence on competition would be measured with an equal biomass and activity of microorganisms and equal concentrations of dissolved C and N in the planted and unplanted soil. For this purpose, two seedlings of F. gigantea were grown in each pot for 3 weeks with the roots of each plant enclosed with soil in nylon bags (width 15 mm, height 50 mm, and mesh size 25 lm) (Fig. 1). The bags separated the roots physically from the bulk soil but allowed, for example dissolved N and C and microorganisms to move between the bulk soil and the soil in the root bag. The root bags also enabled us to remove the plants from the soil without leaving root fragments.
After 3 weeks of incubation, the pots were divided into two main groups. One of them (40 pots) was prepared to house the plants during the 15 N competition assay. This group was divided into two subgroups, with 20 pots in each. In one of them, the root bag with the two individuals of F. gigantea was removed from the bulk soil. The plants were carefully taken out of their bag. Both individuals of F. gigantea were replanted in the hole left after the bag, the soil from the bag was returned into the hole, and the soil surface carefully compacted to avoid preferential flow of the labeled solution to be added later. The pots with the replanted plants were kept in the greenhouse to adjust soil moisture to 10% of WHC, which was obtained after 4 days. The plants were then kept at 10% of WHC for another 3 days. One hour before 15 N addition, WHC was returned to 60% by addition of deionized water.
In the other subgroup, the plants were kept in the greenhouse at a constant 60% of WHC, replanted 3 days before 15 N addition, as described, and kept at 60% of WHC until 15 N was added.
The other main group of pots, prepared to be unplanted during the assay, was also divided into two subgroups, each with 20 pots. In one of the subgroups, WHC of the soil was adjusted to 10% in 4 days and kept Table 1. The characteristics of the soil used in the experiments. The pH, soil total C and N (mg g À1 dw soil), soil C:N ratio, and field concentrations (lmol/L) of glycine, glutamate, sum of all analyzed amino acids (Total a.a.), and NH þ 4 in the soil solution sampled with Rhizon tension lysimeters (means AE SE, n = 5) when the samples for the experiment were collected. Total a.a. shown after decreasing concentrations: glutamine (43.8 lmol/L), aspartic acid, glutamate, alanine, aspargine, serine, glycine, threonine, and arginine (2.4 lmol/L). at 10% of WHC for another 3 days, as described above.
The two individuals of F. gigantea were then removed, the roots gently separated from the soil in the root bag, the soil returned into the hole left after the bag, and the soil surface gently compacted. One hour before 15 N addition, WHC was returned to 60% by addition of deionized water.
The 20 pots in the other subgroup prepared to be unplanted during the assay were kept in the greenhouse at a constant 60% of WHC until the day for 15 N addition. Then, the plants were removed from the soil as described above, the soil returned to the hole left after the bag, the soil surface gently compacted, and the plants harvested for analysis, as described below.
Ten replicate pots for each treatment received deionized water spiked with 0.05 lmol 15 N per g dw soil as 15 NH 4 Cl, sodium glutamate (H 2 15 N 13 CO[ 13 CH 2 ] 2 [ 13 CO-ONa] 2 ) or glycine (H 2 15 N 13 CH 2 13 COOH) (Cambridge Isotopic Laboratories, >98% 15 N). Four 0.5-mL injections per pot were made with a 10-cm long syringe, which was gradually withdrawn from the bottom to the top of the pot to distribute the solution evenly.
Within 1 h after first injection, half of the number of treated pots was harvested as described below, and the initial (t = 0) 15 N (atom%) in soil and plants determined. These data were used in calculating the gross uptake of N in organisms (see section Calculations) and for comparison of plant 15 N with that in plants harvested from pots prepared to be unplanted. The remaining pots were left for 24 h in a climate chamber (20°C/15°C day-night temperature) with 16 h of light of 400 lmol m À2 s À1 and a mean air humidity of 50%.
When harvested, the plants were gently removed from the soil and rinsed in a mixture of 1.0 mmol/L KCl and 0.5 mmol/L CaCl 2 for 5 min to remove 15 N from the root surface. The plants were then dried at 70°C for 24 h. The dried roots and shoots were separately weighed and ground into a fine powder using a ball mill (Retsch,Mixer Mill 200).
Samples of 10 g of gently mixed soil (soil bag and surrounding 6 mm wide casing of soil) from the upper 50 mm of a pot were taken for chloroform fumigationextraction to release the 14 N and 15 N in the microbial biomass. The extraction followed the method of Brookes et al. (1985), except that 0.4 mol/L KCl was used instead of 0.5 mol/L K 2 SO 4 , to maximize the amount of N per g salt in the soil extract. Five mL of the 0.4 mol/L KCl soil extracts from fumigated and unfumigated samples was vacuum centrifuged without heat in a centrifuge (Savant AES 1000) to remove water, and the residue was ground using a mortar and pestle. The dried plant and soil samples were then analyzed with isotope ratio mass spectrometry to determine total 15 N (atom%) and concentrations, as described below.
To estimate the potential assimilation of 13 CO 2 by plants during the 24 h incubation, control pots with F. gigantea, receiving only deionized water, were placed among the pots receiving 13 C 15 N-amino acids. No detectable 13 C enrichment was found in those plants (data not shown), suggesting that assimilation by plants of 13 Figure 1. Schematic drawing of the experimental design of the plant-microbial competition assay. The plant with its roots enclosed in a soil bag was incubated for 3 weeks in pots filled with soil. Then, the soil bag was removed in one-fourth of the pots, the plant returned to the soil, and the water holding capacity (WHC) adjusted from 60% to 10 in 4 days. At that time, the soil bag was removed in another fourth of the pots, and the plant returned to the soil. The plants with removed soil bag were incubated for another 3 days, onefourth at 10% WHC and the other fourth at the original 60%. The third-fourth of the pots were incubated for 3 weeks plus 4 days plus 3 days at the original 60% WHC, with the roots enclosed in a soil bag. The fourth-fourth of the pots was incubated for 3 weeks, then, WHC in the soil was adjusted from 60% to 10 for 4 days, and the plant with roots enclosed in a soil bag incubated for another 3 days at 10% WHC. The plant in the last two treatments was then removed from the pot and the soil in the soil bag returned to the bulk soil in the pot. which may have been released from pots treated with labeled amino acids, was negligible.

Analysis of 15 N and 13 C with mass spectrometry
The ground, dried plant material of roots and shoots and the homogenized salt from dried soil extracts were weighed in tin capsules (4 mg plant material and 80 mg salt) and analyzed for 15 N and 13 C with continuous flow isotope ratio mass spectrometry. The samples were oxidized in an ANCA-GSL elemental analyzer and passed to a 20-20 isotope ratio mass spectrometer (IRMS, PDZ Europa, UK). The amount of total N and C was quantified, after the subtraction of values from blank samples, using standard curves of 5, 50 and 100 lg N as glycine.
One reference sample (glycine) was analyzed after every fifth sample to compensate for drifting isotope ratios. The d 15 N of the reference was calibrated against atmospheric N 2 and the d 13 C against PDB. The precision of the isotopic determinations was <0.2 & for N and <0.1 & for C. The total amount of N in the soil extracts was too low (often <10 lg N per sample) to be determined with sufficient precision by the mass spectrometer. Therefore, it was determined by a Shimadzu TOC-V CPH analyzer (Kyoto, Japan) after the extracts had been diluted five times with deionized water to reduce the salt interference.
Analysis of soil amino acids and NH þ 4 The soil water was sampled with Rhizon tension lysimeters (Rhizon SMS, Wageningen, the Netherlands). Recent studies suggest that samples from small tension lysimeters give a more representative description of the soluble N pool available for plant acquisition and the contribution of amino acids relative to NH þ 4 to the pool of plant available N than common extraction methods with water or salt solutions (Inselsbacher et al. 2011). The lysimeters were inserted before the rewetting, and samples were taken 1 h after rewetting the dry soil in both the dried-rewetted and the constantly moist treatment. The pots were treated in the same way as the pots in the competition assay but without addition of 15 N and were sampled at the start of the assay. The lysimeter (diameter 2.5 mm, length 50 mm) was made of a porous polymer with a mean pore size of 0.1 lm. The small pore size in combination with the sterile sampling tube minimized the microbial activity in the samples. The lysimeter was installed diagonally into the soil core, and the soil water was sampled by connecting the lysimeter to a 10-mL pre-evacuated sterile sampling tube (Rhizon SMS). At the sampling site in the field, the soil water was also collected with the lysimeters installed diagonally in the upper 5 cm of the soil.

ATP and PLFA measurements
Adenosine triphosphate was extracted from the soil using the method described by Eiland (1985), but with some modifications. Soil (0.5 g fresh weight) was extracted with 20 mL of ice-cold 500 mmol/L H 2 SO 4 and 250 mmol/L Na 2 HPO 4 for 30 min on a rotary shaker. Following extraction, 25 lL of the soil solution was transferred to 3 mL of 250 mmol/L Tris with 4 mmol/L EDTA and a pH of 7.5. Of this buffered soil solution, 50 lL was transferred to a scintillation vial and 50 lL of luciferase-luciferin enzyme (adenosin 5 0 -triphosphate [ATP] assay mix, Sigma-Aldrich, Stockholm, Sweden) added. The light output was immediately measured for 15 s in a Beckman LS6500 scintillator equipped with a single photon meter. The amount of ATP in the samples was calculated from internal standards prepared by adding 19.5 mL of ice-cold extractant and 0.5 mL of ATP standard (5.0 lg per 0.5 mL, disodium salt, Sigma-Aldrich) to each soil sample prior to extraction.
Total extractable PLFAs (phospholipid fatty acids) were used as an indicator of microbial biomass, as described in T€ orneman et al. (2008).

Calculations
The uptake of 15 N was expressed as atom% excess after subtracting for atom% 15 N in unlabeled samples. Atom% was defined as [(R s /(R s + 1)] 9 100, where R s is the ratio of 15 N to 14 N in the sample. Atom% excess was defined as [(R s À R r )/((R s À R r ) + 1)] 9 100, where R r is the ratio of 15 N to 14 N in the unlabeled sample. The uptake was calculated by multiplying the atom% excess numbers by the total N content in the root and shoot samples and in the salt residues from fumigated and unfumigated samples. The total uptake of NH þ 4 , glycine, and glutamate was calculated by multiplying the fraction of 15 N immobilized by plants and microorganisms by the total N in the NH þ 4 , glycine, and glutamate soil pools to account for the dilution of the 15 N label by the available N pool: where N i is the gross immobilization of NH þ 4 and amino acid N, N b is the 15 N taken up by the organisms and Nt 0 / 15 Nt 0 is the ratio of total N to 15 N in the NH þ 4 , glycine, and glutamate pools within 1 h after 15 N addition. The 15 N addition resulted in approximately 10% increase in the NH þ 4 present before label application, 65% of glutamate, and 80% of glycine, respectively. Therefore, the [Nt 0 / 15 Nt 0 ] factor for the amino acids was approximated to 1. No correction was made to account for the potential internal regulation of the amino acid uptake in response to the added tracer amounts (Sauheitl et al. 2009), so the calculated uptakes may overestimate uptake at ambient concentrations. Other N transformation processes than uptake of the added and available N sources within 1 h after initiation of the assay may have occurred. For instance, added 15 NH þ 4 may have nitrified during the assay and 15 N taken up as 15 NO 3 À , hence, overestimating the uptake of NH þ 4 and mineralization of organic matter may have added unlabeled N to the NH þ 4 pool, hence underestimating the NH þ 4 uptake. No corrections were made for those or other N transformations, but we assumed that potential enrichment and dilution of the N pools during the assay would affect N uptake in both plants and microorganisms without discrimination.
The plant uptake of 13 C was calculated in the same way.

Statistical analyses
The data set was tested for equality between variances (Levene 0 s test of equality of error variances) and for normal distribution (Kolmogorov-Smirnov and Shapiro-Wilk tests of normality). The effect of plants, the dryingrewetting treatment, and the N form on microbial N uptake was tested in three-way ANOVAs. As the N form and drying-rewetting treatment interacted, two-way ANOVAs were used to evaluate the effect of drying-rewetting and constant moisture treatments separately for the microbial N uptake. Both plant and N form had significant effects on microbial N uptake in the drying-rewetting treatment, and there was a significant interaction between the N form and the planted-unplanted treatment. A oneway ANOVA was therefore used to evaluate the plant effect on microbial N uptake for each N form separately in the drying-rewetting treatment. The effect of plants and drying-rewetting on bacterial biomass/activity (ATP) and the effect of drying-rewetting and N form on plant N uptake were tested in two-way ANOVAs. The drying-rewetting had a significant effect on the plant N uptake, and there was an interaction between the N form and the drying-rewetting treatment. Therefore, the effect of each N form on the plant N uptake was tested separately in one-way ANOVAs. The relationship between mole 13 C excess and mole 15 N excess in plants exposed to 15 N-13 C amino acids was tested in a linear regression. All statistical analyses were performed in SPSS 11.0 (SPSS Sweden, Kista, Sweden).

Microbial 15 N uptake and ATP-biomass/ activity
Ammonium was the most important N source for the microorganisms when the moisture conditions were changed from a constant 60% of WHC to drying-rewetting (Fig. 2), especially in the planted soil, where the NH þ 4 uptake was almost three times higher than the uptake of glycine and glutamate. The plant effect on microbial NH þ 4 uptake was significant only in the dried and rewetted soil (P < 0.01, n = 5) ( Fig. 2A) and not affected by changes in microbial biomass and activity, as the soil ATP content was similar in all treatments (Fig. 3). The total PLFA content was not significantly affected by the drying and rewetting treatment or by the root proliferation (106 AE 24 and 97 AE 14 nmol g À1 (mean AE SE) in planted and unplanted, constantly moist soil, respectively, and 102 AE 18 and 105 AE 20 in planted and unplanted, dried and rewetted soil, respectively). The microorganisms took up more NH þ 4 than the plants in both treatments, but the ratio of microbial to plant uptake was ten times higher in the dried-rewetted soil (70.7) than in the constantly moist soil (7.6).
The microorganisms took up the same amount of N from glycine and glutamate as from NH þ 4 in the constantly moist soil, regardless of the presence or absence of the plant (Fig. 2). As the amount of glycine N and glutamate N taken up was independent of the drying-rewetting event (Fig. 2), microorganisms in all assays used 7-8 times more of each of the two amino acids than the plant. These observations together show no evidence for a limitation of the uptake of N by microorganisms imposed by the plant in constant or varying soil moisture conditions. Thus, one of the qualifications for competitive interactions by the plant on microorganisms was not fulfilled.
The microbial uptake of 15 N ranged from 14 to 36% of added 15 N, with the highest values in planted dried-rewetted soil, while the plants took up less than 2% of the added 15 N. The total plant and microbial uptake of NH þ 4 were 14% of the extractable NH þ 4 concentration in the constantly moist soil and 30% in the dried-rewetted soil (calculated from data in Figs. 1, 3 and Table 3).

Plant 15 N uptake and uptake of intact amino acids
While NH þ 4 became the dominant N source for microorganisms, it became less important to plants when the soil was dried and rewetted as compared to the constantly moist soil (Figs. 1, 3). The shift was mostly an effect of the significant decrease in the NH þ 4 uptake (P < 0.05, n = 5), but there were slight increases in the glutamate uptake as well (P < 0.01, n = 5), in the dried-rewetted soil. The allocation of 15 N and 13 C to the shoots was independent of the soil moisture conditions, ranging from 33 to 40% and 24 to 43% of total uptake for 15 N and 13 C, respectively. The root dry biomass ranged from 25.5 mg dw to 28.9, independent of the treatment.
The order of N source preference was NH þ 4 , glycine, and glutamate in the constantly moist soil, and the plant took up twice as much NH þ 4 as amino acid N (Fig. 4). This uptake was one-tenth and even less (amino acids) of the uptake by the microorganisms (cf. Figs. 2, 4). As a consequence of the more than 50% reduction in the NH þ 4 uptake as the soil was dried and rewetted, glycine and NH þ 4 became equally used N sources. The plant reduced its acquisition of N from all three sources together by an average of 28% as the constantly moist soil was dried and rewetted (cf. Fig. 4). As the microorganisms increased their N uptake by an order of magnitude at the same change of soil conditions, the second part of the condition for resource competition, namely the limitation of the availability of N by a potential competitor, may be fulfilled.
Large proportions of the labeled amino acids were taken up intact, as illustrated by the linear relationship between excess 13 C and excess 15 N in the plant tissue (Fig. 5). A slope of 2.0 for 13 C 2 -15 N-glycine and 5.0 for 13 C 5 -15 N-glutamate would be equivalent to 100% of assimilation of 15 N as intact amino acids. A slope of 1.25 for 13 C versus 15 N in glycine-treated plants in our study suggests that 62 % of the 15 N was taken up as intact glycine or, alternatively, that there was some metabolism within the plant. Likewise, the slopes of 3.23 and 4.25 for glutamate suggest that 65 and 85% were taken up intact in the constant moist and dried-rewetted treatments,  . Plant uptake of NH þ 4 , glycine, and glutamate (ng N g À1 dw soil) in the dried-rewetted and constantly moist treatment. The bars represent means and standard error (n = 5). One-way ANOVA; ns = P > 0.05, *P < 0.05, **P < 0.01. respectively (Fig. 5), or, alternatively, that there was some metabolism in the plant that varied with the soil moisture conditions. However, the mean ratio of the 13 C-to-15 N uptake was closer to the value of the intact tracer for glycine than for glutamate and for the constantly moist samples compared with the dried-rewetted (Table 2).
Soil amino acid and NH þ 4 concentrations The NH þ 4 concentrations were considerably (P < 0.001, n = 5) higher than the concentrations of any of the amino acids in all treatments, whereas glycine and glutamate concentrations were similar (Table 3). Drying and rewetting increased the concentrations of all three N forms (P < 0.001 for NH þ 4 and P < 0.05 for glycine and glutamate). The ratio of glycine and glutamate concentrations to the NH þ 4 concentration was orders of magnitude larger in the field than in the experimental soil (Tables 1  and 3). In the field samples, the total concentration of the nine most abundant amino acids was similar to the NH þ 4 concentration ( Table 1).

Discussion
Organisms qualifying as potential competitors have to use a common and limited resource (Burkholder 1952). The first requirement was fulfilled in the experiment, as both plants and microorganisms used all three N sources. However, it was difficult to determine whether any of the sources was limiting. The NH þ 4 uptake in plants and microorganisms during the 24-h assay in the constant moist soil was only about 14% of the pool at the end of the assay (cf. Figs. 1, 3 and Table 3). That seems to point to NH þ 4 unlimited conditions, in which case, the absence of a plant effect on microbial N uptake in the constant moist soil may reflect soils without N limitation. Those conditions would be inappropriate for a rejection of the first hypothesis, which suggested competition for N in the constant moist soil by impaired microbial NH þ 4 uptake in the presence of plants. However, the low degree of utilization of the NH þ 4 pool does not necessarily mean that the system was N unlimited, as the turnover time is not known.
Drying and rewetting of a soil is known to induce a burst of mineralization of organic N, resulting in high soil . The 13 C: 15 N ratio of added glycine (slope = 2.0) and glutamate (slope = 5.0) is shown by the broken lines, while the regressions calculated from observations of excess 13 C versus excess 15 N in plants are represented by unbroken lines. Glutamate-treated plants in constantly moist soil: slope = 3.23, R 2 = 0.95, P < 0.01, and dried-rewetted soil: slope = 4.25, R 2 = 0.87, P < 0.05, and glycine-treated plants in constantly moist soil: slope = 1.25, R 2 = 0.89, P < 0.05, and dried-rewetted soil: slope = 0.70, R 2 = 0.48, P = 0.19 (not shown). concentrations of NH þ 4 (Kieft et al. 1987;Van Gestel et al. 1993;Pulleman and Tietema 1999) and amino acids (Lipson and Monson 1998). Our experiment was no exception. Notwithstanding the fact that the soil NH þ 4 concentration increased, there was strong evidence for the drying and rewetting treatment favoring the NH þ 4 uptake in microorganisms, especially in the presence of plants. The microbial uptake increased to almost 30% (cf. Figs. 1, 3and Table 3) of the NH þ 4 pool at the beginning of the 24-h assay, at the expense of the uptake in the plants, which decreased by 60%. The reduced plant uptake of NH þ 4 in the dried and rewetted soil may depend on superior uptake by fast-growing bacteria and microfungi, which have higher surface-to-volume ratios and growth rates than plants, triggered by rhizodeposition and exudates from the plants (Henriksen and Breland 1999;Neuman and R€ omheld 2001;Kaiser et al. 2011). Those changes in the NH þ 4 uptake pattern in combination with the simultaneous shift toward an amino acid dominated N uptake in the plants, supposedly to partly compensate for the negative effects of the increased microbial NH þ 4 immobilization, may be taken as a pretext for a temporary competition for NH þ 4 under conditions of net mineralization exceeding immobilization.
The microbial NH þ 4 uptake was stimulated by the drying-rewetting treatment also in the absence of roots, suggesting that concentrations of easily degradable substrate with high C:N ratios increased when the soil was dried and then moistened, possibly by release of physically protected organic matter (Van Gestel et al. 1993;Pulleman and Tietema 1999) and extraction of C when water was added. This result is consistent with previous studies that also found increased microbial 15 NH þ 4 immobilization from 24 h (Pulleman and Tietema 1999;Bengtsson et al. 2003) to 1 week ) after rewetting a dry soil. The lack of correspondence between variations in microbial N uptake and the ATP content in soil may depend on the insensitiveness of the ATP measurement to subtle variations in microbial activity (Contin et al. 2000). In a study by Raubuch et al. (2002), soil drying and rewetting was found to lower the ATP content in the soil but increase the respiration rates compared with a constantly moist soil, suggesting a decrease in microbial biomass rather than a lowered activity. Compared with our experiment, their soils were dried fast (15 h) and at high temperature (40°C), which may have caused cell lysis and consequently a greater reduction in microbial biomass and ATP content.
Amino acid N is potentially a more cost-effective N source under conditions when the leaf net CO 2 uptake decreases because of stomatal closure, for example during drought (Cornic and Massacci 1996;Schmidt and Stewart 1999). Drought causes a decline in the activity of some enzymes, for example nitrate reductases (Cornic and Massacci 1996), and induces others, such as root proteases, facilitating amino acid uptake (Kohli et al. 2012). That may provide an additional explanation to the higher amino acid uptake in the dried-rewetted soil. The plant seemed to have a preference for glycine-to-glutamate, which is in agreement with previous studies on preferences for less complex amino acids (Lipson et al. 1999;Weigelt et al. 2005;Harrison et al. 2007), whereas the microorganisms showed no discrimination between the two. This suggests that F. gigantea used the different amino acids independent of the immobilization by the microorganisms. The short-term uptake of glycine-15 N was similar in F. gigantea and grass species from other ecosystems (Henry and Jefferies 2003;N€ asholm et al. 2009), but the drying-rewetting effect on the amino acid uptake was opposite to that found in the alpine sedge Kobresia myosuroides (Lipson and Monson 1998). That may have been the result of a longer drying period (45 days) for the soil with K. myosuroides, causing more damage to the roots than in our study.
The lower 13 C: 15 N ratio in F. gigantea in the dried and rewetted treatments compared with the constant moist soil can be explained by either an uptake of mineralized amino acid-15 N parallel to the intact amino acids, or by a higher amino acid catabolism and respiration by plants in the dried-rewetted treatment. Under conditions when carbohydrates are in low supply, for example when water is limiting, plants can deaminate glutamate and catabolize glycine and use the C as an energy source (Buchanan et al. 2000). The 15 NH þ 4 , produced from the deamination, may be used to synthesize new amino acids, while some of the 13 C may be lost by respiration, thus reducing the Table 3. Soil solution (sampled with Rhizon tension lysimeters) concentrations of NH þ 4 , glycine, and glutamate (lg 9 10 À2 N g À1 dw soil) in the different treatments at the start of the 24-h competition assay.