Inorganic and Organic Nitrogen Acquisition by a Fern Dicranopteris dichotoma in a Subtropical Forest in South China

The fern Dicranopteris dichotoma is an important pioneer species of the understory in Masson pine (Pinus massoniana) forests growing on acidic soils in the subtropical and tropical China. To improve our understanding of the role of D. dichotoma in nitrogen (N) uptake of these forests, a short-term 15N experiment was conducted at mountain ridge (MR, with low N level) and mountain foot (MF, with high N level). We injected 15N tracers as 15NH4, 15NO3 or 15N-glycine into the soil surrounding each plant at both MR and MF sites. Three hours after tracer injection, the fern D. dichotoma took up 15NH4 + significantly faster at MF than at MR, but it showed significantly slower uptake of 15NO3 − at MF than at MR. Consequently, 15NO3 − made greater contribution to the total N uptake (50% to the total N uptake) at MR than at MF, but 15N-glycine only contributed around 11% at both sites. Twenty-four hours after tracer injection, D. dichotoma preferred 15NH4 + (63%) at MR, whereas it preferred 15NO3 − (47%) at MF. We concluded that the D. dichotoma responds distinctly in its uptake pattern for three available N species over temporal and spatial scales, but mainly relies on inorganic N species in the subtropical forest. This suggests that the fern employs different strategies to acquire available N which depends on N levels and time.


Introduction
Nitrogen (N) is a major limiting element in many terrestrial ecosystems [15,32]. In the past three decades a large number of plants have been identified to have the capacity to directly take up organic N, mainly in the form of amino acids from soil solution [3,11,21,22]. Therefore, uptake pattern of different N species may be an important mechanism responsible for species coexistence in plant communities [19]. Numerous studies on plant N acquisition of organic and inorganic N species have been conducted in subtropical and tropical forests, most focusing on tree species [1,27,28,29,36,37,38], bryophytes and lichens [6,34] as well as some epiphytes [8,35]. While there was one study investigating N acquisition by a tropical fern [39], little is done to study organic and inorganic N uptake by subtropical ferns.
Dicranopteris dichotoma is an important terrestrial fern and is widely distributed in southern China as an early-stage colonizer of acidic and oligotrophic soils [9,40]. D. dichotoma is characterized by rapid clonal growth and often forms a dense understory layer in humid subtropical and tropical forests ( Figure 1). Numerous studies have suggested that Dicranopteris species can influence many ecological processes in these forests, such as soil erosion, nutrient cycling, tree regeneration, and plant community succession [5,25,26,41]. Recently, understory removal experiments showed that the Dicranopteris-dominated understory can form favorable soil microclimates and acts as a major driver of soil biota and ecological processes in forest ecosystems [18,40]. However, N acquisition mechanisms of D. dichotoma remain unknown in these tropical forests, and clarification of D. dichotoma N uptake patterns could improve the mechanistic understanding of its role in these subtropical and tropical forests.
To investigate organic and inorganic N acquisition by D. dichotoma, we selected two different habitats in a subtropical forest ( Figure 1): one located at the mountain ridge (MR) and the other located at the mountain foot (MF). Compared to the MF site, the MR site is characterized by heavy soil erosion and relatively lower N availability. Numerous studies have demonstrated that plants under high available N levels show higher N uptake rates [7,30]. Therefore, we hypothesized that D. dichotoma would have greater N uptake capacity at high N level than at low N level. Additionally, we hypothesized that D. dichotoma could acquire more nitrate than ammonium and organic N because it is more mobile in soil solution [31]. To test these hypotheses, we conducted a short-term 15 N labeling experiment in both D. dichotoma communities with different available N levels.

Study Site
This study was carried out at Xingguo Soil Erosion Observation Station (which belongs to the Institute of Geographic Sciences and Natural Resources, Chinese Academy of Sciences) in Xingguo County (26u309N, 115u 289E, 80 m above sea level) of Jiangxi Province, southern China, where no additional special permission for the research site was needed given that the site is owned by the institute and a long-term research permission from the local government exists since 1993. Moreover, our study did not involve any endangered or protected species. Two sites were selected along a mountain slope ( Figure 1): one was located at the mountain ridge (MR) and the other was located at the foot of the mountain (MF). The understory was dominated by D. dichotoma at both sites. The Pinus massoniana trees at both sites had been planted more than 20 years before the study. In the MR sites, the trees are very small because of heavily eroded soil. In contrast, the trees grow better in the MF site. Soil was classified as loamy Lixisol and some properties are presented in Table 1, showing higher nutrient concentrations at the MF site than at the MR site.
Thirty-six similar sized clusters of D. dichotoma were randomly selected at both sites. They were divided into three groups with each group including 12 individual plants. Each of the groups was labeled either with 15  NO 2 , 95.0 atom% 15 N), by injecting 10 mg N g 21 dw soil. The tracers were injected into the soil at 3 cm depth in a pattern representing the three points of a triangle, with 5 cm length between points and the plant at the centre of the triangle. Six clusters were harvested 3 h after the tracer was injected. The remaining 6 clusters for each treatment were harvested 24 h after the tracer was injected. These harvested plants were classified into roots and shoots. An additional 6 clusters were injected with the same amount of water following the same pattern and harvested as the control after 24 h. The roots were put into 0.5 mM CaCl 2 solution for 30 min. Then, they were rinsed with purified H 2 O and dried at 75uC. These plant materials were weighed for biomass. Dried roots and shoots were ground to a fine powder using a ball mill (MM200, Retsch, Germany) for the measurements of N content and 15 N/ 14 N ratios. Fresh soil (upper 10 cm) was collected for nutrient analysis.
Soil NO 3 -N and NH 4 + -N were determined in 0.5 M K 2 SO 4 extracts on an auto-analyzer (AA3, Bran-Luebbe, Germany). Soil glycine concentrations were measured by high performance liquid chromatography (Waters 515, Waters Inc., USA) in the same extracts [23]. Aliquots of ground plant material (about 2 mg) and soil (about 40 mg) were weighed into tin capsules for analysing organic C, total N and 15 N/ 14 N ratios using isotope ratio mass spectrometry (IRMS, MAT 253, Finnigan MAT, Germany), with a Flash EA1112 interfaced by ConFlo III to the IRMS. Soil pH was measured using a glass electrode on a 1:2 soil-to-water ratio by weight.

N Uptake Calculation
Atom% excess 15 N (APE) was calculated as the atom% 15 N difference between plants from 15 N treated and from control plants. 15 N uptake by plants was estimated by calculating the 15 N excess of each plant part (biomass6%N/1006APE/100 for shoot and root individually) and then summing them up and dividing this number by root biomass and expressed as mg 15 N g 21 dw root h 21 .
The standard errors of means are presented in figures and tables as a variability parameter. T-test was used to compare the  difference in N uptake rates and soil characteristics between sites. Tukey HSD test was used to compare the contribution of three N species to total N uptake between MR and MF. Three-way ANOVA was performed to test the effects of site, N species and time and their interactions on N uptake rates. Data transformation of Ln (data) was applied to meet preconditions of variance homogeneity and normal distribution before ANOVA analysis. All differences were tested for significance at P,0.05 and all statistical analysis were performed on SPSS 17.0 software package (SPSS Inc., Chicago, IL, USA).

Results
Concentrations of soil organic C and total N were consistently but not significantly higher at MF than MR (Table 1). Compared to MR, concentrations of NH 4 + and glycine were significantly higher at MF (Table 1). NH 4 + was the dominant N species among the three N species, while nitrate and glycine-N concentrations were comparably low at both sites. Total biomass of D. dichotoma, including shoots and roots was significantly lower at MR than at MF (P,0.05) (Fig. 2). D. dichotoma at MR also had significantly lower ratios of root to shoot than at MF (MR vs MF: 1.1560.11 vs 1.7660.17) (Fig. 2).
Results of three-way ANOVA showed significant effects of site, N species and their interactions on N uptake rates but no significant effect of time (Table 2). Three hours after tracer injection, the fern D. dichotoma took up 15 (Fig. 3).
Twenty-four hours after tracer injection, uptake rates of 15 Nglycine by D. dichotoma did not change at both sites (Fig. 3). Surprisingly, uptake rates of 15    (0.2760.04 mg 15 N g 21 dw root h 21 ). By comparison, uptake rates of 15 NO 3 2 remained unchanged at MR, but they significantly increased at MF (Fig. 3, T-test, P = 0.034).
Three hours after tracer injection, the fern D. dichotoma preferentially took up 15 NO 3 2 at MR (50% of total N uptake) while it preferentially took up 15 NH 4 + at MF (69% of total N uptake). By comparison, 15 NH 4 + contributed 40% and 69% to the total N uptake at MR and at MF, respectively. The contribution of 15 NO 3 2 was significantly higher at MR than at MF. The contribution of 15 N-glycine was around 11% to the total N uptake at either of the sites (Fig. 4).
Distinct N uptake pattern of D. dichotoma was observed at 24 h after tracer injection. D. dichotoma preferred 15 NH 4 + at MR, which contributed 63% to the total N uptake, while the fern preferred 15 NO 3 2 at MF, which contributed about 47% (Fig. 4). The contribution of 15 N-glycine uptake was significantly higher at MF than at MR (14.8% vs 6.6%) (Fig. 4).

Discussion
We investigated inorganic and organic N acquisition patterns by the terrestrial fern D. dichotoma at two habitats in a subtropical forest using a short-term 15 N labeling experiment. A previous study in a tropical forest showed that amino acid uptake of a lowlight understory terrestrial fern (Danaea wendlandii) reached up to about 330 mg N g 21 dw root h 21 [39]. In this study, the terrestrial fern D. dichotoma showed a much lower uptake rate for glycine-N, only around 0.1 mg 15 N g 21 dw root h 21 . One possible explanation is that Watkins [39] tested uptake of amino acids using excised roots in solution while this study was performed in the field. Roots likely had more opportunity to take up glycine-N in solution than in soil. Available soil N concentrations will be changed through microbial competition and due to microbial mineralization of this organic tracer. Low glycine-N concentrations in the soil (Table 1) also reflect this situation. Besides, the injected soil volume was certainly smaller as the volume of soil extracted to collect the roots afterwards, i.e. many of the collected roots never came into contact with 15 N. Additionally, excised roots may have been more efficient compared to the roots used in this study, which included more inefficient roots not responsible for nutrient uptake. Nonetheless, it has to be noted that excised roots will loose lose uptake potential over time due to C starvation, because the phloem sugar import has been cut off after root excision.
Our hypothesis that D. dichotoma would have greater N uptake capacity at high N level than at low N level was partly supported by our data. Three hours after tracer injection, the fern D. dichotoma demonstrated faster uptake rates of 15 NH 4 + at MF than at MR. One possible explanation could be ascribed to good water status at MF (Table 1), which may enhance the diffusion rate of NH 4 + from soil solution to root surface [2,10]. However, 24 h after tracer injection, the fern D. dichotoma growing at MR showed higher uptake rates for the dominant N form NH 4 + . On the basis of the difference in root biomass between MR and MF (Fig. 2), the efficiency of 15 NH 4 + acquisition by roots at MR is higher than at MF. This indicates that the optimal performance of NH 4 + transporters with high affinity on the root surface [12,24,33] could be more important than soil moisture in these soils where NH 4 + is the dominant N species. Besides, strong microbial competition for 15 NH 4 + could be responsible for low uptake rates by roots at MF [13].
Although soil NO 3 2 concentrations were considerably lower compared to NH 4 + (Table 1), the fern D. dichotoma showed higher uptake rates for 15 NO 3 2 and its uptake was similar to 15 NH 4 + uptake rates in some cases (e.g., 3 h after tracer injection at MR and 24 h after tracer injection at MF). One possible explanation is that NO 3 2 is more mobile in soil solution [31]. These results indicate that our second hypothesis was partly supported by our observations. This could be explained by the fact that NH 4 + is the dominant N species in both habitats (Table 1), and therefore the fern D. dichotoma prefers NH 4 + . The preference for NH 4 + was also observed in the tropical terrestrial fern (D. wendlandii) despite high NO 3 2 concentration [39]. The contribution of the three N species to the total N uptake strongly relies on time and site (Table 2, Fig. 4). Under low N levels (MR), D. dichotoma preferentially take up NO 3 2 3 h after tracer injection, but shifted to the dominant N species (NH 4 + ) 24 h after tracer injection. Under high N levels (MF), D. dichotoma preferred the dominant N species (NH 4 + ) at 3 h after tracer injection and shifted to NO 3 2 at 24 h after tracer injection. The distinct uptake pattern for these available N species observed over temporal and spatial scales suggests that the D. dichotoma employs different strategies to acquire available N, depending on N levels and time. This could be ascribed to rapid regulation of N uptake modulated by a variety of biotic and abiotic factors in addition to tracer ( 15 N) dilution of the available soil N pool, such as the carriers of nitrate, ammonium and amino acids located at the surface of the roots [16,17,20]; soil supply rates of available N [4,14]; delivery of N to the rhizosphere through mass flow and diffusion [10]; as well as competition with soil microorganisms [13]. Further investigations should be focused on how interactions between these biotic and abiotic factors affect N uptake by D. dichotoma for a better understanding of the underlying mechanisms. Table S1 T-test for individual biomass allocation of Dicranopteris dichotoma growing at Mountain Ridge (MR) and Mountain Foot(MF) sites. Supplementary data, including the characteristics of top soil (0-10 cm), individual biomass allocation of Dicranopteris dichotoma, uptake rate of different N species and their contribution to total N uptake rate. And statistical analysis results are also included inside. (XLS)