Plant and Soil

, Volume 297, Issue 1, pp 127–137

Interactions of plant species mediated plant competition for inorganic nitrogen with soil microorganisms in an alpine meadow

  • Minghua Song
  • Xingliang Xu
  • Qiwu Hu
  • Yuqiang Tian
  • Hua Ouyang
  • Caiping Zhou
Regular Article

DOI: 10.1007/s11104-007-9326-1

Cite this article as:
Song, M., Xu, X., Hu, Q. et al. Plant Soil (2007) 297: 127. doi:10.1007/s11104-007-9326-1

Abstract

Sources of competition for limited soil resources, such as nitrogen, include competitive interactions among different plant species and between plants and soil microbes. We hypothesized that plant interactions intensified plant competition for inorganic nitrogen with soil microorganisms. To test these competitive interactions, one dominant species (Kobresia humilis Serg) and one less abundant gramineous herb (Elymus nutans Griseb) in an alpine ecosystem were selected as target species to grow under interactions with their neighboring plants and without interaction treatments in field plots. 15N-labeled ammonium and nitrate were used to quantify their partition between plants and soil microorganisms for 48 h after tracer additions. Responses of K. humilis to interactions from their surrounding plants were negative, while those of E. nutans were positive. Species identity, inorganic nitrogen forms, and plant interactions significantly affected the total amount of nitrogen utilization by soil microorganisms and plants. Although plant interactions have negative effects on nitrogen uptake of K. humilis, there is an increase in microbial immobilization of nitrogen under presence of its neighbors. For E. nutans, facilitation from surrounding plants is in favor of their nitrogen uptake. Compared with K. humilis, competition for \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) was less intensive between E. nutans and microorganisms. \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by soil microorganisms and plants were not more than or much lower than their utilization of \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) under different interaction treatments. These results suggested that the partitioning of inorganic nitrogen between plants and soil microorganisms is mediated by plant–plant interactions and interactions between plants and soil microorganisms.

Keywords

Competition Facilitation \( ^{{15}} {\text{NH}}^{ + }_{4} \)\( ^{{15}} {\text{NO}}^{ - }_{3} \) 15N recovery The Tibetan Plateau 

Introduction

Nitrogen (N) is a primary nutrient limiting plant growth in most terrestrial ecosystems (Vitousek and Howarth 1991), especially in alpine ecosystems, where N is mineralized slowly due to low temperatures. Hence, the competition for nutrients between soil microorganisms and plant species has been considered to be particularly pronounced (Jonasson et al. 1996, 2001). It is assumed that soil microorganisms are the superior competitors for N because of their major role in the mineralization processes as well as their unique attributes compared with plant roots, such as large surface-area, volume ratios and rapid growth rates (Kaye and Hart 1997; Hodge et al. 2000).

Most short-term field 15N experiments showed that plants and soil microbes are mutually limited by, and consequently compete for, inorganic N (Jackson et al. 1989; Zak 1990). Competition for N between plant species and soil microorganisms is a complex process affected by abiotic and biotic factors (Buchmann 2000). Abiotic factors, such as temperature, precipitation and atmospheric CO2 concentrations are among the most important physical factors. Moreover, plant detrital C:N ratios influence soil microorganisms activity, and thus affect competition for N between plants and soil microorganisms. Soil inorganic N forms (NO3-, NH4+) may also influence plant-microorganisms competition for N. Biotic factors, such as species identity and plant–plant interactions, control the utilization of N between plants and soil microorganisms (Brooker 2006). Recent research has shown that N uptake by blue oak (Quercus douglasii) seedlings is significantly suppressed by an annual grass. However, the competition between blue oak seedlings and soil microorganisms is much less (Cheng and Bledsoe 2004). This implies that plant–plant interactions play a key role in regulating the partition of inorganic N between plants and soil microorganisms (Cheng and Bledsoe 2004). In addition, competition for N between plants and soil microorganisms may be an important mechanism controlling N limitation to plants (Reynolds et al. 2003). Other experimental results support the idea that plants and soil microorganisms compete for inorganic N in forest and grassland ecosystems. These experiments also show that plant removal cause an increase in nitrification, while uptake of N by heterotrophic microorganisms is unaffected by plant N uptake (Jackon et al. 1989; Zak 1990; Norton and Firestone 1996; Schimel et al. 1989).

Although numerous studies have been done to evaluate N partitioning between plants and soil microorganisms, most studies in this regard have been focused on either plant N acquisition or microbial N immobilization (Killham 1994; Stark and Hart 1997; Farley and Fitter 1999). So far, very few have focused on evaluating the role of plant interactions in inorganic N partition between plants and soil microorganisms. Hence, studies on how the plant interactions mediated N partition between plants and soil microorganisms can facilitate a better understanding of the mechanisms in N cycling at ecosystem level.

The Qinghai-Tibet Plateau covers about 2.5 million km2 with an average altitude of more than 4,000 m. About 35% of its area is occupied by alpine meadows (Zheng 2000), where low temperature causes low rates of organic matter decomposition (Zhou 2001). Most of the N is bound in organic form (Cao and Zhang 2001). Net N mineralization rates in these meadows are microbial immobilization during the growing season (−26.79 mg m−2 d−1). Most alpine plant species growing on the plateau form prostrate rosettes, with slow growth rates. They are extremely clumped under the stress of abiotic conditions. Previous field investigations on plant–plant interactions suggested that the balance between competition and facilitation among the plants in the alpine meadow ecosystems are species specific, and the net interactions varied from competition to facilitation. The interactions mediated the nutrient utilization of plants (Song et al. 2006). But it is unclear if the interactions also influenced the partitioning of inorganic N between plants and soil microorganisms. In this study, an experiment was carried out which manipulated the competitive environment for two target plant species, and evaluated whether plant interactions mediated plant–soil microorganisms partitioning of inorganic N. The aims of this research were (1) to determine if there is competition for inorganic N between soil microorganisms and plants in the alpine meadows; (2) to test if there is any significant difference in utilizing different forms of N between plants and soil microorganisms. Integrating these results, we were able to test the hypothesis that plant interactions intensified plant competition for inorganic N with soil microorganisms.

Materials and methods

Study site and target species

The experiment was performed at the Haibei Research Station of Alpine Meadow Ecosystem, the Chinese Academy of Sciences, located in the northeast of the Qinghai-Tibet Plateau (37°32′N, 101°15′E). The average altitude of this area is 3,240 m asl. The annual precipitation was 560 mm, 85% of which was concentrated in the growing season (from May to September). The mean annual temperature was −1.7°C. The soil is classified as Mat Cry-gelic Cambisols (Chinese soil taxonomy research group 1995) corresponding to Gelic Cambisol (WRB 1998), and a detailed description is shown in Table 1. The study area is dominated by Kobresia humilis Serg. The common species include Stipa aliena, Elymus nutans Griseb, Saussurea superba Anth., Festuca ovina Linn., Gentiana lawrencei Burk.var farreri T.N.Ho, Gentiana straminea Maxim., Potentilla nivea Linn., Potentilla saundersiana Royle, Scirpus distigmaticus Tang et Wang, Kobresia pygmaea C.B. Clarke in Hook, and Carex sp (Zhou 2001). The vegetative coverage of the surface is over 95%. Rooting depth is shallow, with over 90% of roots concentrated in the upper 15 cm of the soil (Zhou 2001). Plant growth is N limited, with marked growth responses occurring in August.
Table 1

Characteristics of the upper 10 cm of soils at the study site, Haibei, Qinghai-Tibet Plateau

Soil characteristics

pH

8.0 ± 0.1

Bulk density (g cm−3)

0.70 ± 0.05

C:N ratio

19.6 ± 0.3

SOC (kg m−2)

11.8 ± 0.3

Total soil N (kg m−2)

0.60 ± 0.04

Microbial biomass N (g m−2)

6.5 ± 0.3

DON (g m−2)

1.8 ± 0.1

Extractable inorganic N (g m−2)

1.4 ± 0.4

Data (means ± S.E.) are shown (n = 6–8).

Two species (Kobresia humilis Serg and Elymus nutans Griseb) that exhibit different growth forms and phenological rhythm were selected as target plants in this study. The dominant species K. humilis belongs to Cyperaceae, with rosette growth forms. They sprout in the middle of April in early spring, and bloom and fruit in early June. The less abundant species E. nutans, a Gramineae, has an erect growth form, sprouts at the end of May, and blooms and fruits in August. Distinctively, K. humilis is a perennial clonal species with a phalanx rhizome. E. nutans is a perennial clonal species with a guerilla rhizome.

Experimental design

In May 2004, a block uniform in species composition and cover was selected in a typical K. humilis meadow. Fifteen main-plots (3 × 3 m) were arranged with 2-m wide buffer zones between them. Then each plot was divided into four subplots (0.75 × 0.75 m), with 0.5-m wide buffer zones on all sides. This provides a randomized complete block design with two plant species and two interaction treatments as fixed factors nested within the plots, two treatments (target plant grown under no interactions with their neighbors; target plant grown under interactions with their neighbors) were established for each target species. In total, 15 replicates were assigned to each treatment for each species.

For each plant species, two interaction treatments were manipulated. Root interactions between target species and their neighbors were hampered by root exclusion tubes. For shoot treatments, we tied down the shoots of the neighboring plants, using thin bamboo netting (mesh size: 2 × 2 cm). The detailed description is as follows: (1) no interaction for target plants (N), target plants growing without neighboring above- and below-ground parts (netting and tube); (2) with interaction for target plants (F), target plants growing in contact with neighboring shoots and roots (no netting and no tube).

On May 26–28, 2004, a hole 7.5 cm in diameter and 20 cm in depth was excavated in the center of each subplot. In each plot, two subplots were randomly selected, the holes of which were installed using PVC tubes (7.5 cm in diameter and 20 cm in length), whereas the other two subplots were not installed using PVC tubes. The holes were then refilled with root-removed, sieved soils (<2 mm) excavated from the same holes, and the surrounding vegetation was left untouched.

The two target plant species were transplanted from the alpine meadow on May 29, 2004, Seedlings with similar size (individuals of each of the two species were size-standardized by selecting the same number of leaves, same length of roots, with similar original biomass) were chosen for each of the two species, then they were transplanted into the designated spots. In each plot, one individual of each target species was transplanted into one hole with a PVC tube (without interaction) and another one was transplanted into the hole without a PVC tube (with interaction). For the shoot interaction-free treatments we tied down the shoots of the neighboring plants using netting, whereas for shoot interaction treatments we left the neighboring plants unmanipulated (Wilson and Tilman 1991; Twolan-Strutt and Keddy 1996; Cahill 1999, 2002a, b, 2003). The transplanted plants were watered daily for 5 days to reduce transplant shock.

15N injection

On August 20, 2004, 15 main-blocks were selected for the 15N experiment. Five main-blocks were injected with \( ^{{15}} {\text{NO}}^{ - }_{3} \) solution (Na15NO3 98.36 at.% 15N, 57.21 mmol·l−1), and another five were injected \( ^{{15}} {\text{NH}}^{ + }_{4} \) solution [(15NH4)2SO4 98.27 at.% 15N, 37.31 mmol·l−1). The remaining five main-blocks were injected with H2O as the control treatments. At each of three injection points in each pot, we inserted the needle to 8 cm depth, released 0.8 ml of solution to produce a relatively uniform distribution of 15N from 0 to 20 cm depth. A total of 2.4 ml solution was injected in each pot. The amount of 15N injected into individual pots was calculated to produce an average concentration of \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) 3.32 mg N kg−1 and of \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) 2.55 mg N kg−1 soil. The soil inorganic N concentration was 7.37 ± 0.40 mg N kg−1 in the natural alpine meadows. Considering \( {\text{N}} - {\text{NH}}^{ + }_{4} \) concentration is much lower than \( {\text{N}} - {\text{NO}}^{ - }_{3} \) concentration in natural alpine meadow soil, concentration of \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) injected is higher than concentration of \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \), which can reduce the influence of higher \( {\text{N}} - {\text{NO}}^{ - }_{3} \) concentration impeding utilization of \( {\text{N}} - {\text{NH}}^{ + }_{4} \).

Data collection

Plant materials and soils (20 cm depth) were collected for 48 h after 15N injections. All target plants were separated into shoots and roots. The roots were rinsed briefly with water, then for 30 min with 0.5 mmol l−1 CaCI2 solution, and again with distilled water. These plant materials were oven-dried (60°C for 48 h), weighed, ground and analyzed for total N and at.% 15N. Soil samples were sieved to 2 mm to remove coarse fragments for measuring moisture, pH, microbial biomass N and exchangeable inorganic N. Dried subsamples were ground to a fine powder and, after removal of carbonates by diluted HCI, used to measure soil organic carbon (SOC) and total N.

Soil moisture was measured by the gravimetric method (105°C, 24 h). Soil pH values were measured using a glass electrode with a 1:2 soil-to-water ratio. Total N was measured by Kjeldahl digestion with a salicylic acid modification (Pruden et al. 1985), and SOC was measured following the method described by Kalembasa and Jenkinson (1973). Microbial biomass N was estimated by a chloroform fumigation-direct extraction technique (Brookes et al. 1985; Davidson et al. 1989). Total N in 0.5 M K2SO4 extracts (1:4 soil: extractant) was also determined by Kjeldahl digestion of a salicylic acid modification (Pruden et al. 1985), whereas \( {\text{NH}}^{ + }_{4} - {\text{N}} \) and \( {\text{NO}}^{ - }_{3} - {\text{N}} \) were measured by steam distillation with MgO, using Devarda’s alloy to reduce \( {\text{NO}}^{ - }_{3} \) to \( {\text{NH}}^{ + }_{4} \) (Bremner 1965). Dissolved organic nitrogen (DON) was calculated as the difference between total N and exchangeable inorganic N in the extracts. All plant, soil and diffusion samples were analyzed for total N and at.% 15N by continuous-flow gas isotope ratio mass spectrometry (CF-IRMS).

Calculations and statistical analyses

Target plant biomass in the presence of their neighbors relative to that of absence of their neighbors was measured as competitive response (CR). CR was calculated for interactions with their neighbors, using the following equations:
$${\text{Competitive}}\,\,{\text{ response:}}\,\,\,\,\,\,\,{\text{CR = ln }}{\left( {{{\text{FB}}} \mathord{\left/ {\vphantom {{{\text{FB}}} {{\text{NB}}}}} \right. \kern-\nulldelimiterspace} {{\text{NB}}}} \right)}$$
(1)
where NB is mean biomass of the target plants without interaction with their neighbors (N treatment), FB is the mean biomass of target plants exposed to interaction with their neighbors (F treatment). CR was calculated for all plants that could potentially interact with their neighbors. The biomass under interaction (F treatment) was compared to the mean target plant biomass under no interaction (N treatment) of the same species. Negative values indicated net competitive responses of target species to their local neighbors, and positive values indicated net facilitative responses. The competitive response metric is similar to both the response ratio described by Goldberg et al. (1999), and competition intensity (CI) used by other researchers (Wilson and Tilman 1991, 1995; Twolan-Strutt and Keddy 1996), where TCR = ln(1 − CI).

Plant 15N uptake (PN) can be calculated by multiplying the mass of N in samples (Nsample) by the at.% 15N difference between plants from 15N added (at.% 15Nadded) and from control plots (at.% 15Ncontrol) in the following equation: PN = (at.% 15Nadded −  at.% 15Ncontrol) × Nsample.

Soil microbial biomass N can be calculated from 1-day CHCI3 fumigation measurements by the following equation: BN = (1-day CHCI3-N) / 0.54 (Brookes et al. 1985) where BN is the soil microbial biomass N and CHCI3-N is the amount of total N extracted by 0.5 mol l−1 K2SO4 from soil immediately after fumigation, minus that extracted from a non-fumigated soil at the time fumigation commenced. The 15N immobilized in microbial biomass was calculated as the difference in the 15N recovered in non-fumigated and fumigated soil samples (i.e. 15NMBN = 15NFum·NFum − 15NExtr·NExtr).

15N recovery percentage (%) for both plants and microbial biomass was calculated as:
$$\begin{array}{*{20}c} {^{{{\text{15}}}} {\text{N}}\,\,{\text{recovery}}\,\,{\text{\% }} = \left( {{\text{the amount}}\,\,\,^{{15}} {\text{N}}\,\,{\text{taken up by plants}}} \right.} \\ {{\text{or microorganisms}}{{\text{ }}\,} \mathord{\left/ {\vphantom {{{\text{ }}\,} {\,{\text{total }}^{{15}} {\text{N added into }}}}} \right. \kern-\nulldelimiterspace} {\,{\text{total }}^{{15}} {\text{N added into }}}} \\ {\left. {{\text{per plant plot}}} \right)\, \times 100} \\ \end{array} $$

We examined the effects of species identity, interaction treatments on total plant biomass, using general linear models (GLM) with species identity and competition treatments as fixed factors in a randomized complete block ANOVA. Biomass was ln transformed to satisfy the assumption of normally distributed data in ANOVA. To test whether plant interactions affect plant competition for N with soil microorganisms, data were analyzed using three-way ANOVA with species identity, N forms and plant competition treatments as main factors, respectively. ANOVA were used to test the differences in different forms of inorganic N uptake by K. humilis serg., E. nutans Griseb and soil microorganisms under target plants experiencing interaction treatments, respectively. All statistical analyses were performed using SPSS10.0.

Results

Plant interactions with their neighbors

Both plant species identity and interaction treatments affected plant growth. Total biomass significantly differed between species and among different interaction treatments (Table 2). The significant species identity × interaction treatments indicated that plant–plant interactions are species specific. The interactions with surrounding plants led to a decrease in biomass for K. humilis (Fig. 1a). On the contrary, total biomass of E. nutans increased under presence of their neighbors than in the absence of their neighbors (Fig. 1a). Responses of K. humilis to interaction with their neighbors were negative and competitive, while those of E. nutans were positive and facilitative (Fig. 1b).
Table 2

ANOVA results for ln-transformed target plant total biomass

Source

DF

Type III SS

F

P

Species

1

86.381

349.427

0.000

Competition treatments

3

23.239

94.007

0.000

Species × competition treatments

3

3.821

15.457

0.003

Species identity and competition treatments served as two fixed effects.

Fig. 1

Mean (±S.D.) total biomass for two target species measured after 12 weeks of growth under two interaction treatments (a), and Mean competitive response for two target species (b). For a with None representing plants growing in the absence of their neighbors and Full representing plants growing in the presence of their neighbors. Different letters indicate significant differences between treatments (LSD test, P < 0.05). n = 15 for each treatment. For b competitive response is the natural log of the proportional growth of target plants when grown with neighbors, compared against growth without neighbors (Eq. 1). Positive values reflect facilitation, negative values reflect competition

15N recovery

\( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{{\text{ + }}}_{{\text{4}}} \) and \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{{\text{ - }}}_{{\text{3}}} \) recovery by plants and soil microorganisms were species specific (Table 3). In addition, N partition between plants and soil microorganisms was mediated significantly by interactions of target plant species with their surrounding plant species (Table 3).
Table 3

ANOVA results of the total inorganic nitrogen utilization by soil microorganisms and plant species

Source

DF

Type III SS

F

P

Species

1

0.186

8.404

0.005

Inorganic N forms

1

0.542

24.487

0.000

Competition treatments

3

0.215

3.239

0.028

Species × Inorganic N forms

1

0.0216

0.977

0.327

Species × competition treatments

3

0.206

3.096

0.033

Inorganic N forms × competition treatments

3

0.0109

0.164

0.920

Species × Inorganic N forms × competition treatments

3

0.0155

0.233

0.873

Species identity, nitrogen forms, and competition treatments served as three fixed effects.

\( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) recovery by plant species and soil microorganisms

For dominant species K. humilis, the competition from their surrounding plants had negative effects on inorganic N uptake. \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) recovery by K. humilis was higher under absence of their neighbors than in the presence of their neighbors (Fig. 2a). By comparison, \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) recovery by soil microorganisms showed an opposite trend, which was lower under target plants without competition from their neighbors than that of target plants with competition (Fig. 2a).
Fig. 2

Competition for \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) (a, b) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) (c, d) between plant species (K. humilis and E. nutans) and soil microbes under target plants experiencing interactions with their neighbors and without interaction treatments respectively, with None representing plants growing in the absence of their neighbors and Full representing plants growing in the presence of their neighbors. Different letters indicate significant differences between interaction treatments (LSD test, P < 0.05). n = 5 for each treatment

For less abundant species E. nutans, the facilitation from their surrounding plants had positive effects on inorganic N uptake. \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) recovery was higher under E. nutans with interactions from their neighbors than that under E. nutans without interactions from their neighbors (Fig. 2b). However, soil microbial immobilization for \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) was not significantly different whether the neighbors of E. nutans were present or absent (Fig. 2b).

\( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by plant species and soil microorganisms

Partition of \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) between K. humilis and soil microbes showed the same patterns as that of partition of \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) in the presence of surrounding plants (Fig. 2c). \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by K. humilis was higher in the absence of surrounding plants than in the presence of surrounding plants. While \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by soil microorganisms was higher under target plants with competition than those without competition from their neighbors (Fig. 2c). For both E. nutans and soil microbes, \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery was higher under E. nutans with interactions with their neighbors than without interactions (Fig. 2d).

Differences in utilization of \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) between plants and soil microorganisms

Plant interactions mediated the partitioning of \( {\text{NH}}^{ + }_{4} \) and \( {\text{NO}}^{ - }_{3} \) between plants and soil microorganisms. Under competition treatments, \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \) uptake by K. humilis was higher than that of \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) uptake. But under no competition treatments, \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) uptake by K. humilis were not significantly different. \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \) uptake by soil microorganisms was higher than that of \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) uptake with or without competition (Fig. 2a, c).

\( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \) uptake by E. nutans was higher than that of \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) uptake with or without competition. There were no significant differences in \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) uptake by soil microorganisms under E. nutans with interactions with their neighbors. While \( ^{{{\text{15}}}} {\text{N}} - {\text{NO}}^{ - }_{3} \) uptake by soil microorganisms was higher than that of \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) uptake under E. nutans without interactions with their neighbors (Fig. 2b, d).

Discussion

The relationship between soil microorganisms and plants are simultaneously mutualistic and competitive (Harte and Kinzig 1993). As decomposers, soil microorganisms are indirectly responsible for the bulk of terrestrial vegetation’s annual nutrient demand (Schlesinger 1991). In turn, plant matter is the major source of photosynthetically fixed carbon for decomposers. However, microorganisms and plants also compete for soil nutrients (Hodge et al. 2000). Our results supported the hypothesis that plant–plant interactions mediated plant competition for nitrogen with soil microorganisms. K. humilis suffered from competition from their neighbors, the competition had a negative effect on its competition for inorganic N with soil microorganisms, but the increase C source supply from neighboring plants may improve the activity of soil microorganisms, which facilitates their immobilization for inorganic N. But for E. nutans, the facilitation from surrounding plants favored their growth and their uptake for inorganic N. Compared with K. humilis, the competition for inorganic N between E. nutans and soil microorganisms was less intense.

Plant interactions

Our study showed that the species had different responses to interactions from their surrounding plants. Although the competition from surrounding plants had a negative effect on the growth of K. humilis, E. nutans grew somewhat better under the presence of their surrounding plants. Our results are consistent with previous research suggesting that plant–plant interactions are species specific (Callaway 1998; Choler et al. 2001). Root characteristics, including growth rate, biomass, fine root density and surface area, are primary plant factors affecting belowground competition (Casper and Jackson 1997). K. humilis is a dominant species in alpine meadows, with a rosette growth form. The belowground buds sprout at the end of April. The maximum growth rate occurs in May and June. At the beginning of August, there is a flush of nutrients and resources are stored underground for the next year (Zhou 2001). The high fine root density and rapid growth rate of K. humilis makes it a superior competitor for available soil N. When neighboring plants were absent, total 15N recovery by K. humilis increased significantly (increased by 135 and 457% for \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \), respectively). E. nutans is relatively less common in alpine meadows, with erect growth forms. E. nutans sprouts in the middle of May. The developed guerilla rhizome warrants its foraging for belowground resources (Zhou 2001). E. nutans made a positive response when growing with their neighbors. When neighboring plants were present, 15N recovery by E. nutans increased significantly (increased by 147 and 170% for \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \), respectively). At high elevations, the stressful conditions, such as low temperature or wind scouring may limit plant growth more than resource availability for some species. Amelioration of these severe stresses by neighbors may favor growth more than competition for resources with the same neighbors impairs growth (Callaway et al. 2002).

In interaction treatments, PVC tubes were used to separate root interaction between target plant and their neighbors. The PVC tubes may not be “pot bound” for growth of target plants. One reason is that plants growing in the alpine ecosystems are very small, and their growth rate is also slow. Roots of replanted plants are small at the beginning of our experiment. There are enough places for root development within PVC tubes with 7.5 cm diameter during the growing season. When we collected the plant, we noticed that there was not the bounding effect, which was confirmed further by the low value in total biomass. Another reason is although PVC confined plant growth, plants can get the necessary resource by deepening and extending their root. The key limiting factors to plant growth in alpine ecosystem is stressful environmental conditions, but not resource capture.

N partition between plants and soil microorganisms

In the alpine meadow on the Qinghai-Tibet Plateau, N is limiting to plant production (Cao and Zhang 2001), and there is evidence that N limits the size of N pools in soil microorganisms (Xu et al. 2003). The negative net N mineralization rate suggested that soil microorganisms were N-limited (Zhou 2001). Previous nitrogen addition experiments showed that both plant biomass and microbial biomass increased significantly, which indicated N-limitation in the alpine meadow ecosystem on the Qinghai-Tibet Plateau (Qiu and Du 2004). Our results suggested that assimilation of inorganic N into plant and soil microbial biomass is mediated by plant–plant interactions. On the one hand, soil microorganisms compete with plants for inorganic N. On the other hand, soil available C sources from plants also promoted microbial activities. The potential for microbial N immobilization is directly related to the amount and quality of the C sources. For example, many studies showed that plant root exudates were one of the most important C sources. Exudation of mucilage, exoenzymes, organic acids, sugars and amino acids from roots of vascular plants is a common phenomenon (Whipps 1990; Marschner 1995) and some perennial grasses allocate as much as 20% of their assimilated C to the process (Bokhari 1977). These exudates are readily broken down by heterotrophic microbes that grow in close association with the root surfaces (Clarholm 1985). Increased C exudation and the highly labile exuded substrate were quickly utilized and incorporated into a growing rhizospheric microbial population. In turn, increased heterotrophic microbial activity yielded more available N, which was associated with elevated N uptake and photosynthesis of plants (Hamilton et al. 2001).

Our results showed that the competition for N between K. humilis and soil microorganisms was obvious. \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by K. humilis were higher with their neighbors absent than when present. On the contrary, \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) and \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by soil microorganisms were higher under the presence of the neighbors around K. humilis than under the absence of the neighbors. But the competition for N between soil microorganisms and neighboring plants of K. humilis was less obvious. A possible explanation may be in that neighboring plants might have both positive and negative effects on soil microorganisms. Reduction in competition for N following neighbour removal not only gets rid of competition for N with soil microorganisms (Kaye and Hart 1997; Cheng and Bledsoe 2004), but this possibly also excludes the C-supply from neighboring plants (Scott and Binkley 1997; Schweitzer et al. 2004). For E. nutans, they utilized less inorganic N compared with K. humilis, the competition for N between E. nutans and soil microorganisms was less obvious. \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) recovery by soil microorganisms has no significant differences whether the neighboring plants of E. nutans are present or absent. The possible reasons may be that reduction in competition for N following neighbour removal offsets the positive effects of increased C-supply from neighboring plants. While \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) recovery by soil microorganisms was higher with the neighbors of E. nutans present than when absent, suggesting the C-supply from neighboring plants promoted activity of heterotrophic soil microorganisms (Johnson 1992; Norton and Firestone 1996; Schimel et al. 1989). Moreover, net N mineralization rate in situ was much lower for K. humilis than that of E. nutans, whether with or without interaction with their neighbors (Fig. 3), which further supported the conclusion that competition for inorganic N between K. humilis and soil microorganisms was much intensive than that of between E. nutans and soil microorganisms.
Fig. 3

Net N mineralization rate for K. humilis and E. nutans under the target species with interactions from their neighbors and without interactions from their neighbors in August

Utilization of \( {\text{NH}}^{ + }_{4} \) and \( {\text{NO}}^{ - }_{3} \) by plants and soil microorganisms

Our results showed that \( {\text{NH}}^{ + }_{4} \) utilization by soil microorganisms and plants is much lower or not more than their utilization of \( {\text{NO}}^{ - }_{3} \) under interaction treatments, despite the concentration of \( ^{{{\text{15}}}} {\text{N}} - {\text{NH}}^{ + }_{4} \) injection, was higher than that of \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) ( \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) 3.22 mg N kg−1 and \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) 2.55 mg N kg−1 soil). Jackson et al. (1989) obtained a similar result in short-term 15N experiments. The result was consistent with the conclusion that both plants and soil microorganisms take up more \( {\text{NO}}^{ - }_{3} \) than \( {\text{NH}}^{ + }_{4} \) in the long-term (within 2 months) 15N additions experiment in the same alpine meadow (Xu et al. 2003). The reason may be related to the mobility of \( {\text{NH}}^{ + }_{4} \) and \( {\text{NO}}^{ - }_{3} \). Much of the available cation \( {\text{NH}}^{ + }_{4} \) can be absorbed by soil organic matter or solid particles, while mobile \( {\text{NO}}^{ - }_{3} \) can be moved to roots by mass flow (Nye and Tinker 1977). Thus, it is easier for plants and soil microorganisms to take up more \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) than \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \). In addition, differential preferences of plants for nitrate and ammonium may also be one of the important reasons (Kronzucker et al. 1997). Species grown on \( {\text{NO}}^{ - }_{3} \) is frequently superior to those grown on \( {\text{NH}}^{ + }_{4} \), and higher concentration of \( {\text{NH}}^{ + }_{4} \) sometimes cause toxicity to plants (Malhi and Nyborg 1988; Pearson and Stewart 1993). In our experiment, injection of Na+ in the nitrate treatments may cause an increase in cation concentration in soil. But the influence might be mild. Since injected Na+ is little, which concentration (3.91 mg Na+ kg−1) is over 1,000 times lower than Na+ concentration in natural alpine meadow soil (5.6 × 103 mg Na+ kg−1).

Our results also suggested that plant–plant interactions mediated the distribution of different forms of N between plant species and soil microorganisms. Without competition, the dominant species K. humilis took up more \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \), and soil microbial biomass immobilized less \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \). The amount of \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) was distributed almost evenly between K. humilis and soil microorganisms under K. humilis without competition from surrounding plants. But competition from surrounding species reduced \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) uptake by K. humilis, and promoted \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \) uptake by soil microorganisms. The possible reason was that competition might decrease N availability for plants, but increase C availability for soil microorganisms. The increased available soluble C from exudation of surrounding species might have promoted soil microbial activity. For the less abundant species E. nutans, the uptake of any form of N is significantly lower than that of soil microbial immobilization. Facilitation from surrounding species is in favor of N uptake (\( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \) and \( ^{{15}} {\text{N}} - {\text{NO}}^{ - }_{3} \)) by E. nutans, and at the same time exudation from surrounding species can promote heterotrophic soil microbial immobilization for \( ^{{15}} {\text{N}} - {\text{NH}}^{ + }_{4} \). Therefore, the balance of distribution of N between E. nutans and soil microorganisms depended on plant–plant interactions and interaction between plants and soil microorganisms. For the dominant species K. humilis, they have stronger a ability to obtain different forms of inorganic N (\( {\text{NH}}^{ + }_{4} \) and \( {\text{NO}}^{ - }_{3} \)) in the soil than that of the less abundant species E. nutans.

Our results showed that plant interactions mediated plant competition for N with soil microbes. However, the effect of increased exudation by plants on rhizospheric processes, to our knowledge, has not been addressed. Consequently, the occurrence of these positive or negative plant–plant interactions and plant–microbe interactions under field conditions, i.e., in the alpine meadows that balance the species-diversity and competition for N between plants and soil microbes, still needs to be investigated. In addition, previous studies showed that the competition for inorganic N between plants and soil microorganisms is time-dependent, and that plant and soil microbe N assimilation have different seasonal patterns (Jaeger et al. 1999). Therefore, how plant–plant interactions mediate temporal partitioning of N between plants and microorganisms needs further investigation. Nevertheless, our results suggest that understanding plant physiological processes may need to be placed within an ecosystem context that includes the intimate interactions between plants and their associated heterotrophic microbial community.

Acknowledgements

We appreciate Dr. Martin Werth, from Institute of Soil Science and Land Evaluation, University of Hohenheim, Germany, for his meticulous work in improving the language usage of our manuscript. This research was funded by the National Natural Science Foundation for Young Scientists of China (30600070), and the National Basic Research program of China (Grant no.2005CB422005), and the Key Project of the Chinese Academy of Sciences (KZCX3-SW-339-04). Our experiments were performed in accordance with the current laws of our country.

Copyright information

© Springer Science+Business Media B.V. 2007

Authors and Affiliations

  • Minghua Song
    • 1
  • Xingliang Xu
    • 1
  • Qiwu Hu
    • 1
  • Yuqiang Tian
    • 1
  • Hua Ouyang
    • 1
  • Caiping Zhou
    • 1
  1. 1.Institute of Geographical Sciences and Natural Resources Researchthe Chinese Academy of SciencesBeijingPeople’s Republic of China

Personalised recommendations