Organic nitrogen uptake is a significant contributor to nitrogen economy of subtropical epiphytic bryophytes

Without any root contact with the soil, epiphytic bryophytes must experience and explore poor, patchy, and heterogeneous habitats; while, the nitrogen (N) uptake and use strategies of these organisms remain uncharacterized, which obscures their roles in the N cycle. To investigate the N sources, N preferences, and responses to enhanced N deposition in epiphytic bryophytes, we carried out an in situ manipulation experiment via the 15N labelling technique in an Asian cloud forest. Epiphytic bryophytes obtained more N from air deposition than from the bark, but the contribution of N from the bark was non-negligible. Glycine accounted for 28.4% to 44.5% of the total N in bryophyte tissue, which implies that organic N might serve as an important N source. Increased N deposition increased the total N uptake, but did not alter the N preference of the epiphytic bryophytes. This study provides sound evidence that epiphytic bryophytes could take up N from the bark and wet deposition in both organic and inorganic N forms. It is thus important to consider organic N and bark N sources, which were usually neglected, when estimating the role of epiphytic bryophytes in N cycling and the impacts of N deposition on epiphytic bryophytes in cloud forests.

substrata 15,20,21 . For example, Ayres, et al. 21 reported that some bryophyte species can obtain N directly from the soil. In addition to terrestrial species, many bryophytes grow on other plants as epiphytes in the canopy habitats in montane and subalpine forests 8,22,23 . The canopy habitats are usually considered to be harsh, with a variable and sporadic nutrient input, a limited storage capacity for available water and nutrients, and low physical stability, etc. 24,25 . Accordingly, epiphytic bryophytes must experience and explore poor, patchy, and heterogeneous environments. It remains unknown where and how these organisms obtain nutrients without root contact with the soil.
The N preferences of epiphytic bryophytes remain uncharacterized. The preferences for ammonium (NH 4 + ) or amino acids over nitrate (NO 3 − ) have been observed in vascular plants when different N forms are supplied in equal doses 26,27 . Some researchers have suggested that bryophytes may not have a preference in N uptake because nutrients can enter moss tissues easily through cation exchange and the proton (H + ) pump (e.g., NH 4 + and amino acids) and through cotransport (e.g., NO 3 − ) for positively charged ions 10,15,28 . Other researchers have suggested that the uptake of NH 4 + should be higher than that of NO 3 −15 29 . Current knowledge of N cycling in bryophytes and the effects of N deposition on bryophytes largely relates to inorganic N [30][31][32] . The ability of bryophytes, especially the epiphytic ones, to use organic N as a N source and the ecological significance of this has been largely neglected in past studies, although the preference of amino acids has been reported in several terrestrial bryophytes 33,34 .
Anthropogenic N deposition has been increasing globally since the 19 th century, which has triggered major changes in the dynamics of carbon (C) and N, as well as floral diversity in terrestrial ecosystems [35][36][37] . Increasing concern has been focused on the effects of enhanced N deposition on bryophytes, which were suggested to be sensitive to N pollution 11,18,38 . A previous study indicated that species richness and the cover of the epiphytic bryophyte community has significantly decreased because of increased N input. The growth and vitality of the investigated species have declined in locations with high N loads 11 . However, we still do not know the potential impacts of increased N deposition on the total N uptake and the N preference of epiphytic bryophytes. For example, do bryophytes absorb more N in response to increased N deposition, as has been suggested in some studies 39,40 ? Do bryophytes shift their N preference to increased N deposition when N is much more easily obtained? This information may provide possible explanations for the detrimental effects of the high N loads mentioned above.
Due to their particular biological nature (no cuticle barrier, lacking roots and developed vascular systems, and growing on bark, etc.) and special habitats (poor, patchy, and heterogeneous in N supply), epiphytic bryophytes are likely to have different N uptake and use strategies under natural conditions and under increasing atmospheric N deposition. Epiphytic bryophytes usually dominate the tree trunks and branches in montane forests on moist, undisturbed sites 8,41 . For example, the subtropical montane cloud forest located in the Ailao Mountain National Nature Reserve of Southwest China, which is generally characterized by persistent and frequent cloud cover at the canopy level, is especially rich in epiphytic bryophytes 42 . In total, 176 epiphytic bryophyte species have been recorded (accounting for ~30% of the total epiphytes in the study region) 41 and the total biomass of the epiphytic bryophytes is 6.7 tons per hectare (accounting for ~63% of the total epiphytes in the study region) 41,43 . A multifactor in situ manipulation experiment was carried out via the 15 N labelling technique in three coexisting and common epiphytic bryophyte species in the subtropical cloud forest. The main objectives were to: 1) Determine whether epiphytic bryophytes can take up N from tree bark or bark surface; 2) Confirm the capacity of epiphytic bryophytes to absorb organic N and quantify its amount in the total N economy of the bryophytes; 3) Address the potential impacts of increased N deposition on the N uptake dynamics of epiphytic bryophytes.

Results and Discussion
Direct uptake of N from bark. Using a 15 N labelling approach, we confirmed that epiphytic bryophytes indeed relied more on N from the air than from the bark of their hosts, but the contribution of the N from the bark should not be neglected (Tables 1 and 2; Fig. 1). Cryptogams, such as bryophytes, which can absorb water directly through their surfaces, have been traditionally suggested to be largely independent of their substrate and were thought to absorb N mostly from precipitations and biological N fixation 19,44 as well as the relocation of nutrients from dead moss tissue 45 . It is surprising that epiphytic bryophytes could use the bark N because bryophytes possess rhizoids rather than roots, which reach only a few cm into the bark surface. However, the recovery of the label in the bryophyte tissue ( Fig. S1) does not necessarily reflect active uptake by the bryophyte, since the added N could have reached the bryophyte via passive diffusion along the bryophyte shoots. Endophytic fungi could have retained the added label and get some amino N mineralization and uptake as NH 4 + . It is also possible that not all bacteria have been killed by the ampicillin once it was injected into the bark. Nevertheless, the 15 N enrichment in the bryophyte shoots suggest that the epiphytic bryophytes can acquire N from the bark and translocate it to their shoots via various pathways, e.g. passive diffusion, endophytic fungi, bacteria. Therefore, this still need further investigations in the future.
During the injection procedure, we observed a small amount of leakage of the solution that had been injected into the bark. This was probably due to the limited water-holding capacity of the bark during the injection process. A spot of the leaked solution may flow directly onto the bryophyte shoots, and thus may have been taken up at that point, which obscure the interpretation of our results. Nevertheless, the proportion of 15 N recovered in the trunk-dwelling bryophyte P. assamica after 24 h incubation was 3.2% for NO 3 − , 6.4% for NH 4 + , and 5.4% for glycine, respectively (Fig. S1). Similar results were found in the other two species, H. flabellatum and H. scalpelifolium (Fig. S1). Although lower than the air deposition counterparts (Fig. S1), the proportion of 15 N recovered in the three bryophyte species through bark injection was comparable with previously reported data on the recovery from soil by the Antarctic moss Sanionia uncinata (2% for alanine and 4% for NH 4 + ) 46 . Thus, the fact that a small amount of liquid leaked out of the bark may not significantly impact the conclusion that epiphytic bryophytes can take up N from bark.
Yet it has been demonstrated that bryophytes can take up N from bark when it is added, but is there actually a significant amount of available N in bark for bryophytes to take up? The average total N concentration in barks of the three dominant host species, i.e. Lithocarpus xylocarpus, L. hancei, and Castanopsis wattii are 7.85 ± 0.61, 8.92 ± 1.33, 8.52 ± 0.49 g kg −1 , respectively, which are higher than the average total N concentration in the surface soil (0-20 cm: 6.53 ± 0.83 g kg −1 ) (Song et al. unpublished data). Mean concentrations of total N, NH 4 + -N, and NO 3 − -N in the stemflow in the study region are 2.39 ± 1.11, 1.14 ± 0.50, and 0.42 ± 0.21 mg l −1 , respectively, which are significantly higher than that in the precipitation (0.49 ± 0.14, 0.11 ± 0.05, 0.04 ± 0.01 mg l −1 ) 47 . The above data indicate that there is actually a significant amount of available N in bark for epiphytic bryophytes to take up, which is comparable to N availability from atmospheric deposition and other substrates in the subtropical cloud forest.
Field investigations have indicated that bryophytes are the dominant epiphytic plants in this ecosystem, as they account for approximately 63% of the total biomass, which is more than any other epiphytic vegetation type (orchids, ferns, and lichens, etc.) 41 . This indicates that epiphytic bryophytes may be significant competitors for N in the stemflow and throughfall, which could have potential consequences for the plant community structure and nutrient cycling at the ecosystem level 47,48 . Since all three bryophyte species included in this study have the capacity to absorb the available N from the bark, according to the δ 15 N signals of the shoots (Fig. 1), this uptake may be common among epiphytic bryophyte species in general. Increasing evidence indicates that both the substratum and the atmosphere are important mineral sources for bryophytes, even in 'feather mosses' , which have poor soil-moss contact 46,49 . Recently, Liu, et al. 15 estimated that soil N accounted for approximately 40% of the total N in terrestrial bryophytes. If N absorption from bark is common among epiphytic bryophytes, this could partially explain their widespread distribution and importance in many montane and moist ecosystems. If this is the case, the N cycle in the studied forest ecosystem should be modified. Bryophytes  ). In this study, the contribution of organic N (glycine) ranged between 28.4% and 44.5% of the total N uptake, which was comparable with NH 4 + , but significantly higher than NO 3 − (Tables 1 and 2; Fig. 2). This was probably due to the greater cation-exchange capacity than the anion-exchange capacity of the cell walls 29,50 . No significant differences were detected for the acquisition of different N forms  Table 2. Percentages of nitrogen (N) absorbed in forms of NO 3 − , NH 4 + , and glycine by three bryophyte species under air deposition and bark injection at low, medium, and high N addition levels. Means ± SE are presented (n = 4). Capital letters after values indicate significant differences at 0.05 error probability levels among different labelled-N forms under same N levels, while lowercase letters indicate significant differences among different N levels under same labelled-N forms.

Organic N as important component of N input.
between air deposition and bark injection for all three epiphytic bryophyte species (Table 2), which demonstrated that these organisms do not shift their N preference from air to bark N sources. Our study indicated that amino acids, a small but important organic N component, as well as NH 4 + -N and NO 3 − -N can serve as important N sources for epiphytic bryophytes. Previous laboratory and field studies have also revealed that amino acids can be absorbed and utilized at substantial rates, which greatly contribute to the total N uptake and effects the N metabolism of bryophytes 15,34,51,52 ; however, this has not been studied in epiphytic bryophyte species. For example, Forsum, et al. 33 applied 15 N labelled solutions to Hylocomium splendens (Hedw.) and found that this species preferred amino acid N over NO 3 − , although the assimilation of glycine remained lower than that of NH 4 + . The mean uptake rates were 1.8 μ mol g −1 DW h −1 for NO 3 − , 3.6 μ mol g −1 DW h −1 for NH 4 + , and 3.4 μ mol g −1 DW h −1 for glycine, which indicated that the amino acids could be absorbed by the bryophytes 50 . The preference for amino acid N or NH 4 + over NO 3 − observed could be partially explained by their differences in assimilation costs. According to Liu, et al. 15 , and references therein, the assimilation cost of amino acids is expected to be lower than that of NH 4 + and much lower than that of NO 3 − , which is due to the requirement that NH 4 + must be attached to a C skeleton before use while NO 3 − requires additional reduction steps to NH 4 + . These studies demonstrated that amino acids should be one of the most cost-effective N forms that can be utilized by bryophytes. Considering that Scientific RepoRts | 6:30408 | DOI: 10.1038/srep30408 amino acids account for only a small proportion of organic N, the bioavailable fraction of organic N is expected to be much larger than that found in amino acids 53 . Although the organic N input in the studied forest has never been directly measured, the fact that the annual input of total N through precipitation (10.5 kg N ha −1 y −1 ) was ca. threefold of the sum of the two main inorganic-N forms (NO 3 − : 0.91 kg N ha −1 y −1 , NH 4 + : 2.69 kg N ha −1 y −1 ) 47 , implies that organic N may be an important N form in the studied region. The contribution of organic N to the N economy of epiphytic bryophytes might have been seriously underestimated in the past, although uncertainty exists considering some organic N, e.g. amino acids may have been mineralized and been took up as NH 4 + .
Impact of enhanced N deposition on N uptake. In this study, the shoot 15 N concentration and the N uptake rates increased significantly with increasing N concentrations (Table 1; Figs 1 and 2), but the N preferences of the three bryophyte species shifted only slightly in response to the addition of N, except for Homaliodendron flabellatum that preferred glycine under medium N addition, but shifted to NH 4 + under high N addition (Table 2). This implies that increased N deposition increases the total amount of N absorbed by the epiphytic bryophytes, but it does not alter the N preference over a short time. Due to N limitations in many ecosystems dominated by bryophytes, a slight increase in N can increase the absorption of N, which enhances the chlorophyll content of the bryophyte cells, thus increasing the photosynthetic capacity 54,55 . However, excessive N supply is detrimental to these sensitive organisms. For example, it has been demonstrated that the oversupply of N can result in an excess uptake of NH 4 + in the cell, which threatens the cell homeostasis and causes toxicity, and thus a subsequent reduction in the growth of the bryophytes 55,56 . Increased N deposition can alternatively cause increased amino acid accumulation in bryophyte tissues, which may deplete reserves of soluble carbohydrates necessary for growth 54 . As indicated in this study, high N loads resulted in excessive N uptake, which may induce biochemical disorders in bryophytes 57 .
In conclusion, this study provides clear evidence that epiphytic bryophytes can uptake N from the bark and can translocate it to their shoots. The ability to translocate the absorbed N to their shoots is of particular importance, since shoots typically have greater N demands for photosynthetic enzymes. This study highlighted that organic N, as opposed to inorganic N, contributed remarkably to the N economy of the epiphytic bryophytes. High N loads may result in excessive N uptake, which may induce biochemical disorders in bryophytes. Thus, it is important to consider organic N and bark N sources when estimating the role of epiphytic bryophytes in N cycling and the impacts of N deposition on epiphytic bryophytes in cloud forests. On account of the persistent, frequent cloud/fog cover, the presence of large, old trees, and long-term effective protection, the forest harbors abundant epiphytes, and is especially rich in epiphytic bryophytes 41,42 . The most dominant epiphytes growing on tree trunks including Homaliodendron flabellatum (Sm.) Fleisch., Plagiochila arbuscula (Brid. ex Lehm.) Lindenb., H. scalpelifolium (Mitt.) Fleisch., and P. assamica Steph. are bryophyte species 41 . Epiphytes comprise one of the most diverse and are a conspicuous element of the subtropical cloud forest, and they are also extremely important in carbon, water and nutrient cycling in these ecosystems 47,48 . For example, epiphytes in the subtropical cloud forest can fix a significant amount of N 2 ; the latest estimated annual N input fixed by epiphytic bryophytes reaches 3.89 kg N ha −1 y −1 60,61 .

Experimental design and treatments. Two mosses (H. flabellatum and H. scalpellifolium) belonging to
Neckeraceae and one liverworts (P. assamica) belonging to Plagiochilaceae were selected for this manipulation experiment as they were abundant, representative trunk-dwelling species in the study region 41,42 . In November 2014, four areas (two hectares each) in the cloud forest in Xujiaba were chosen as the experimental plots. In each plot, 48 square quadrats (20 cm × 20 cm) that were dominated by each single species (16 quadrats for each species) were marked on large trunks (Diameter at breast height > 20 cm) that were located between 1.0 m-2.0 m above the forest floor. The average total biomass of H. flabellatum, H. scalpellifolium, and P. assamica collected from each experimental quadrat were 1.01, 1.13, and 1.46 g, respectively (Fig. S2).
To study the N preference of the three epiphytic bryophytes, four treatment groups were established. In each treatment, equal proportions of glycine, NH 4 + , and NO 3 − (1:1:1) were used. Glycine was adopted because it has been widely used as a model amino acid for studies of organic N uptake by plants 62 . The first treatment was considered to be the control (Control), and no 15 N-labelled N was added. In the other three treatments, only one of the N forms was labelled with 15 N, i.e., 15  − -N labelled). The experiment was divided into two parts. First, we aimed to determine if epiphytic bryophytes could absorb the available N directly from the tree bark. For this, a subset of high N solutions (total N concentration of 12 mM: each N form at 4 mM) with the four treatment groups mentioned above (CK, Glycine-N labelled, NH 4 + -N labelled, NO 3 − -N labelled) were injected uniformly into the bark of the trees in the 48 marked quadrats (four N forms × three species × four replicates) with 5 mL syringes. Before the injection process, a steel needle (15 cm in length and 3.5 mm in diameter) was used to squeeze through the bryophyte layer and set up injection holes on barks. The trees were then injected at nine injection points (three rows × three columns) each on the bark of all quadrats to approximately a 5 mm depth using special metal frame sheets (20 cm × 20 cm) to ensure uniformity and consistency. The total amount of N injected to each quadrat corresponded to a dose of 0.21 kg N ha −1 .
The second experiment was designed to study the treatment effect of increased N addition on the N preference and uptake. Three N levels, i.e., low (total N concentration 3 mM: each N form at 1 mM), medium (6 mM: each N form at 2 mM), and high (12 mM: each N form at 4 mM), of each treatment were added. In this experiment, 5 ml of the experimental solutions with different N concentrations were sprayed uniformly over 144 experimental quadrats (three N levels × four N forms × three species × four replicates) with small pressure sprayers. The total amounts of N added were 0.05, 0.11, and 0.21 kg N ha −1 (equal to 18.25, 40.15, 76.65 kg N ha −1 y −1 ) for the low, medium, and high N levels, respectively. The above treatment levels in our simulations of N input were implemented based on background rate (10.5 kg N ha −1 y −1 ) in the study region 47 and a predicted rate measured in a comparable region, i.e. southern China: 30-73 kg N ha −1 y −1 63 . Spraying was conducted within special metal frame cubes (20 cm × 20 cm × 20 cm) to avoid the loss of the solutions through air movement. Ampicillin (10 mg L −1 ) and CaCl 2 (100 μ M) were added into the solutions to avoid the rapid decomposition of the amino glycine.
Scientific RepoRts | 6:30408 | DOI: 10.1038/srep30408 Bryophyte shoots were harvested one day after the injection or spraying events, following the procedures of Krab, et al. 51 and Rousk,et al. 46 . According to the method used by Warren 64 , all bryophyte shoots were rinsed in 50 mM KCl to remove any remaining 15 N that was still adhering to their surface, and then they were rinsed with ultrapure water, oven-dried (70 °C), and ground for 15 N isotope analysis. The total biomasses of the three target bryophyte species are shown in Fig. S2. N contents and 15 N/ 14 N ratios were determined using an isotope ratio mass spectrometer (Isoprime 100, Isoprime Ltd., Cheadle, UK) coupled with a vario PYRO cube elemental analyzer (Elementar Analysensysteme GmbH, Hanau, Germany) with a continuous flow mode. The atom% excess 15 N (APE) was calculated as the atom% 15 N difference between the bryophytes from the 15 N treated plots and from the control plots. Data analysis. According to Xu, et al. 65 , the uptake of the 15 N (mg 15 N m −2 ) by the bryophyte shoots was calculated by multiplying the N content (mg N g −1 d.w.), APE, and biomass (g m −2 ). The uptake of the available N forms corresponding to the 15 N treatment was calculated as in the following: unlabelled l abelled unlabelled labelled where m labelled is the total mass (g m −2 ) of the 15 N-labelled N injected or sprayed per quadrat and m unlabelled is the mass of available N forms measured in solution. U labelled is the uptake (g m −2 ) of 15 N from the source m labelled and U unlabelled is the uptake of the available N from the source m unlabelled . All data were subjected to normality and homoscedasticity tests before statistical analysis. Two different ANOVA models were adopted: one focused on the N source and the other focused on the N level. To compare N sources, only the high N treatment groups through both air spraying and bark injecting were compared. To compare the treatment effects of the N levels, only the air spraying treatment groups, including low, medium, and high N levels, were considered.
Multiple comparisons of the shoot δ 15 N, N uptake rates, and contributions of different N forms (percentages of N absorbed in the forms of NO 3 − , NH 4 + , and glycine) among the labelled-N forms under the same N addition levels and among different N addition levels under the same labelled-N forms were conducted using LSD's or Game-Howell's post hoc tests. All the analyses mentioned above were conducted in SPSS 16.0 (SPSS, Chicago, IL, USA), and all figures were made using SigmaPlot 12.5 (Systat Software Inc., San Jose, CA, USA).