In cold environments as polar and alpine regions the edaphic nitrogen is mainly available as organic compound, imposing metabolic restrictions to the biological mineralization of nitrogen (Shaver and Chapin, 1980; Pietr et al., 1983; Atkin, 1996). To cope with the inorganic N scarcity, plants take advantage of symbiotic interaction with microorganisms (e.g., root mycorrhizal symbionts and root endophytes), as a strategy to enhance their nutritional status (Hobbie et al., 2000; Newsham, 2011; Acuña-Rodríguez et al., 2020). The benefit of the interactions, as described in plant-mycorrhyzae interactions from the Arctic tundra (Hobbie and Högberg, 2012), are related to the capacity of the microbial symbionts to mineralize complex organic N compounds into inorganic forms like ammonium (NH₄⁺) and nitrate (NO₃^–), which are easily absorbed by the plant’s roots. In consequence, the plant-microorganisms association increases N acquisition and enhances the ecophysiological performance of plants (Hobbie and Hobbie, 2008).

The microbiome associated with the Antarctic vascular flora is dominated by the ascomycetous fungi known as dark septate endophytes or DSE (Upson et al., 2009b; Newsham, 2011; Ruotsalainen, 2018). These symbiotic fungi, usually found in the roots, can enhance plant nutrient acquisition, particularly N and P (Newsham, 2011; Hill et al., 2019). However, the shift from organic to inorganic N as nutrient source seems to alter the effect of some root DSE in their host plants, either positively or negatively. This is similar to what has been found for the plant-mycorrhizae interaction of the Arctic tundra in which the role of mycorrhizae on the net plant N-uptake decrease if inorganic N become more available (Hobbie et al., 2000; Johnson et al., 2010). As shown by Upson et al. (2009a), under controlled conditions, four out of six DSE strains had positive effects on shoot and root biomasses of Deschampsia antarctica (Poaceae) individuals only when grown using organic N as nutrient source. When supplied with inorganic N, some detrimental effects on the plant were observed (Upson et al., 2009a), presumably because both plant and fungi compete for soil resources, shifting the plant-DSE association from beneficial to negative for the host. Thus, the positive role of DSE root-symbionts on their host plants’ performance is still not conclusive and appears to be highly dependent on the environmental conditions (Newsham, 2011; Acuña-Rodríguez et al., 2020).

Among the terrestrial ice-free areas that allow the life of vascular plants in Maritime Antarctic, those that harbor ornithogenic soil, represent a particular edaphic environment due to their extremely high N concentration (Pires et al., 2017). During the summer, the animal N input produces a patchy spatial distribution of edaphic N, which concentrates around colonies (Bölter et al., 1997; Park et al., 2007). For example, it has been estimated that in Maritime Antarctica, total soil N could vary from highly enriched (N_tot = 14.9–8.8 g kg^–1) surrounding animal colonies, to highly depleted (N_tot = 0.5–0.17 g kg^–1) approximately 800 m away from the colony’s influence (Bölter et al., 1997; Łachacz et al., 2018). Furthermore, the composition of the N pool can also vary drastically depending on the distance to these colonies. The rapid mineralization of animal urea not only raise local ammonium⁺ concentrations in the presence of water, but also produces a volatile N source through the emanation of gaseous ammonia (Pietr et al., 1983), which can be exported up to 1 km away from the bird colonies, depending on the local topography and wind dynamics (Erskine et al., 1998; Bokhorst et al., 2019a). This inorganic N input, spontaneously mineralized from animal-originated N-forms, has been related to the greater performance of lowland coastal plant populations compared with those from more inland locations (Androsiuk et al., 2015). Thus, given that the composition of the N-pool (i.e., organic or inorganic) is known to alter the effect of microbial symbiotic on plants (beneficial or costly), it can be predicted that in ornithogenic N-enriched soils, N-acquisition by Antarctic vascular plants might not be exclusively attributed to the role of symbiotic microorganisms.

Several studies have tested this hypothesis using the isotopic fractionation that occurs during the biological N mineralization in some fungal symbiont-plant associations (Benavent-González et al., 2019 and references therein). Given the natural existence of two stable isotopes of nitrogen (¹⁴N and ¹⁵N), the proportion of the heavier isotope in both the N source (soil) and N products (i.e., plant and fungal tissues), has been proposed to be affected by the active role of fungal symbionts in the process of N uptake (Högberg, 1997). For example, during the acquisition of organic N mycorrhizal fungi is prone to retain ¹⁵N-enriched N, while ¹⁵N-depleted N is transferred to the plant hosts (reviewed in: Hobbie and Högberg, 2012). Hence, in this plant-fungi interaction model, the intermediate step of acquiring N through the fungal symbiont generates low δ¹⁵N values in foliar tissues compared to the isotopic signature of the soil N source (Michelsen et al., 1998; Hobbie et al., 2000). Nevertheless, unlike most plant communities, the microbiota associated to the roots of the Antarctic plants is dominated by DSE instead of mycorrhizal fungi (Upson et al., 2009b). Antarctic endophytes and mycorrhizal fungi, however, seem to play a similar ecological role enhancing nutrient acquisition and N in particular (Hill et al., 2019; Acuña-Rodríguez et al., 2020).

The main goal of the present study was to explore the role of fungal endophytes on the N biological mineralization and plant N-acquisition processes when inorganic N is not limiting. We specifically addressed two questions: (i) is the organic N-mineralization in the rhizosphere of two vascular Antarctic plants enhanced by the presence of root endophytes under N-enriched conditions? and (ii) does root endophytes participate in the N-uptake of these plant species when inorganic N is not a limiting factor? To answer these questions we specifically measured: (a) the percentage of NH₄⁺ accumulated in the soil through time in plants inoculated and non-inoculated with root fungal endophytes to determine the relevance of this symbiotic association on the process of N biological mineralization, (b) the differences in biomass accumulation between those inoculated and non-inoculated plants and (c) the δ¹⁵N isotopic signature in foliar tissues, we determined if fungal endophytes actively participate in the process of N-acquisition when inorganic N is not a limiting factor. By answering these questions, we are able to evaluate if the root fungal endophytes maintain their positive role as N-uptake enhancers for their hosting plants when grown in N-rich ornithogenic soils, such as those found in some Antarctic habitats.

Microscopy analyses demonstrated that E+ individuals were progressively colonized by DSE both extra and intracellularly. Considering that at the beginning of experiments there was no evidence of root colonization, the root infection process was successfully (Data not shown). By the end of the N-mineralization experiment (60 days), the percentage of infested roots in C. quitensis inoculated with P. chrysogenum reached 88.5 ± 1.6% and was 91.2 ± 0.9% in D. antarctica plants inoculated with P. brevicompactum. Relative to the temporal dynamic of the available NH₄⁺ in the substrate of the experimental plants, GAMM models revealed for C. quitensis and D. antarctica a significant increase in time among the rhizospheric soil of both E− and E + plants (Supplementary Table 1). However, despite this general increase among all experimental groups, there was a significant influence of the infection status in both species, and particularly in C. quitensis, where E+ plants showed greater contents of NH4+ in their rhizospheres if compared with E− individuals (Figure 2). This can be easily observed in the absence of confidence interval overlapping in C. quitensis. By contrast, in D. antarctica the fitted splines for E+ and E− individuals appear close to each other, such as to do not appear statistically different in some time periods toward the end of the experiment (Figure 2).

For both species, the enhanced availability of inorganic N in the form of NH₄⁺ in soil of E+ individuals may explain their higher average dry biomass at the end of the experiment relative to E− plants (Figure 3). In this sense, a significant biomass increase of 34 and 23% was found for both C. quitensis and D. antarctica in E+ individuals, relative to their respective axenic E− counterparts [endophyte treatment: F_{(1, 24)} = 114.12; p≤ 0.0001]. However, there was no significant interaction between endophyte treatment and species, meaning that the effect of endophytes on plant biomass was similar in both C. quitensis and D. antarctica (Figure 3).

In relation to the ¹⁵N isotopic signature of the foliar tissues, significant differences were found between experimental groups (E+ and E− plants) in both species (Figure 4). The average δ¹⁵N values obtained showed that, relative to the isotopic fractionation in the initial substrate (δ¹⁵N = 3.67), the foliar tissue of C. quitensis and D. antarctica individuals from both endophyte treatments were depleted in ¹⁵N. However, the fractionation among E− plants (C. quitensis: 3.05 ± 0.51‰; D. antarctica: 2.71 ± 0.49‰) was far lower than in E + plants (C. quitensis: −2.31 ± 0.89‰; D. antarctica −0.35 ± 0.86‰). This suggests that for both species, the inoculated root endophytes were significantly involved in the process of N-uptake. Furthermore, the interaction term in the two-way ANOVA was statistically significant [endophyte treatment × species: F_{(1, 36)} = 25.27; p < 0.0001] with N fractionation being significantly greater in C. quitensis than in D. antarctica, but only in E+ plants (Tukey test, p < 0.05; Figure 4).

Our results indicate that the presence of the studied root endophytes significantly favored organic N mineralization in the rhizospheric soil associated with Antarctic vascular plants Colobanthus quitensis and Deschampsia antarctica. Additionally, endophytes favor N-uptake independently of the availability of NH₄⁺ –an inorganic and easily assimilable N-source– in both species. Previous studies on Antarctic vascular plants have shown their capacity to modify the quality and composition of the soil organic N pool (Roberts et al., 2009), and their potential to obtain free amino-acids and small peptides from the soil (Hill et al., 2011, 2019). However, it has been recently demonstrated, at least for D. antarctica, that much of these rhizospheric dynamics of N-transformation and uptake in this plant rely on the activity of their fungal root endophytes (Hill et al., 2019). In this sense, the present results are complementary to those of Hill et al. (2019), who demonstrated the participation of endophytes in the uptake of small peptides by D. antarctica under controlled conditions. The greater accumulation of NH₄⁺ in the soils of E+ plants of both species found in our study suggest that, together with their capacity to metabolize amino-acids in an early stage of organic matter decomposition, the rhizospheric mineralization of organic N forms like urea is also enhanced by root fungal endophytes. Nevertheless, increase of NH₄⁺ over time in soil with plants (E−) also occurred, that could be explained by direct hydrolysis of urea in the soils or even as result of the mineralization performed by plants itself. Urease activity has been reported for other root endophytes (Jumpponen et al., 1998; Narisawa, 2017), and it is likely that the species used in this study also have the same metabolic capability. On the other hand, the improvement in the mineralization of the organic N source by endophytes could be related to the higher biomass found in E+ individuals relative to E− plants at the end of the experiment. Even though it is not possible through our experimental design to define which specific N uptake pathway was favored by the fungal symbiont, its presence definitively promotes the incorporation of N into the plant hosts. However, of all the possible N forms, NH₄⁺ is the most plausible compound incorporated by these Antarctic plants after endophytic mineralization from urea.

Fungal endophtytes could also explain the higher N uptake efficiency that has been observed in Antarctic vascular plant species, particularly D. antarctica, when the inorganic N availability increases in the soil (Rabert et al., 2017). Indeed, the preference of D. antarctica for NH₄⁺ as its main N source has been demonstrated, even when other inorganic N forms like NO₃^– were available in a wide range from low- to high-levels (Rabert et al., 2017). In contrast, C. quitensis did not show any substrate preference when exposed to similar concentrations of these inorganic N compounds (Rabert et al., 2017). This may explain the pattern of NH₄⁺ accumulation in soils found in this study, which was more evident in C. quitensis than in D. antarctica. Moreover, NH₄⁺ accumulation in soils was higher in C. quitensis toward the end of the experiment (days 30 and 60) compared to D. antarctica. Thus, the higher efficiency of D. antarctica in acquiring NH₄⁺ could explain the lower accumulation of this substrate in the soil, even under the improved mineralization promoted by the fungal inoculation.

It is important to highlight that the presence of root endophytes significantly changed the ¹⁵N isotopic signature in the foliar tissues of both species, demonstrating the active participation of this endophytic fungi in the process of N uptake by the host plant roots. Despite the ¹⁵N signature found in leaves of E+ and E− plant tissues from both species appear to be depleted relative to the substrate, this effect was significantly larger in leaves of inoculated (E+) individuals, particularly in C. quitensis. The slightly depleted, and still positive, ¹⁵N signal observed in the foliar tissues of E− plants is consistent with plants being grown on a ¹⁵N-enriched substrate, which is typical of ornithogenic soils (Zhu et al., 2009). This is because the process of ammonia volatilization that occurs spontaneously in the presence of water after an input of uric acid in the soil, strongly discriminates against the heavier N isotopes, increasing its proportion in the soil substrate as the lighter isotope leaves the soil pool as volatile NH₃ (Erskine et al., 1998; Bokhorst et al., 2019b). For this reason, among E− plants, which acquire N without the aid of microbial symbionts, the isotopic signal in their tissues was similar to those of the substrate. Contrastingly, infected individuals (E+) of both species showed a negative isotopic ¹⁵N signature, indicating a larger depletion of the heavier isotope in the assimilated N, presumably by the N-fractionation generated by the fungal symbiont. This mineralization process, which should be analogous to those exerted by mycorrhizal fungi in Arctic plant species, produces ¹⁵N-enriched fungal tissues, while transferring ¹⁵N-depleted nitrogen forms to the plant host (Hobbie and Colpaert, 2003; Hobbie and Högberg, 2012). This would suggest that the δ¹⁵N signature in the endophyte biomass should also be enriched in ¹⁵N. However, due the anatomical distribution of the fungal endophytes inside the root tissues, it was not possible for us to measure this signature in the fungal biomass.

Several studies have estimated the proportion of N isotopes among the Antarctic biota, highlighting the role of marine-derived N on the fertilization of terrestrial ecosystems in relation to their proximity to active mammal and bird colonies (Erskine et al., 1998; Park et al., 2007; Bokhorst et al., 2019a,b). Nevertheless, these values could be highly variable depending on the local conditions. For example, Park et al. (2007) reported in the surroundings of Palmer station in Biscoe Point δ¹⁵N values of 11.2 and 11.0‰ for C. quitensis and D. antarctica, respectively, which showed also a small depletion in ¹⁵N respective to a 13.4‰ found in the soil (Park et al., 2007). However, in a similar study, Lee et al. (2009) found that the ¹⁵N isotopic signatures of D. antarctica from Barton peninsula (King George Island) varied between 0.4 and 4.5‰, depending on how influenced the plants were by the local bird nesting sites. In the light of this, the isotopic signatures found here appear particularly depleted in ¹⁵N (negative values for both species) probably because in our experimental setup we did not reproduce the continuous input of enriched ¹⁵N produced by animal colonies in the field and because the experimental soil was retrieved from a zone without marine animal influence.

It is important to acknowledge that experimental and laboratory conditions are drastically different from the field. For example, by accelerating the rate of N uptake process because growth chambers cannot mimic the exact interaction between temperatures, relative humidity, and radiation experienced by plant in natural conditions. Nonetheless, this do not override the positive effect of fungal endophytes in process uptake here. Similar to the plant-mycorrhiza model, a depleted isotopic signature in the leaves is a clear evidence of the fungal symbiont mediation in the N assimilation by the Antarctic host plants. However, our results suggest that the effect of fungal endophytes for N uptake is most pronounced for C. quitensis than for D. antarctica. This is because despite δ¹⁵N of foliar tissue in both species was significantly depleted relative to their E− counterparts, the fractionation between the substrate and the foliar tissues was lower in E+ plants of D. antarctica (δ¹⁵N_fract = 4.02), than E+ plants of C. quitensis (δ¹⁵N_fract = 5.98). It has been demonstrated that D. antarctica has the metabolic capability to incorporate small organic N-forms like amino acids and short peptides directly from the soil (Hill et al., 2019); a process that seems to fractionate less against the heavier isotope than the endophytic fungi does, leaving a less depleted signature in the plant tissue. However, this was not assessed in this study. Further research is needed to fully understand how fungal symbionts module different pathways of N acquisition and their relative relevance for each Antarctic vascular plant species.

Among cold environments the genus Penicillium has been observed in soil permafrost and ice-caps (Gunde-Cimerman et al., 2003; Zucconi et al., 2012). But it is also present in different Antarctic substrates such as oligotrophic (Godinho et al., 2015), ornithogenic (McRae et al., 1999), and the active layer of soil permafrost, in which spores of the two species studied here were present (Kochkina et al., 2014). In addition, some Penicillium species were found in different tissues of the Antarctic flora, including rhizoids of the liverwort Cephaloziella varians (Newsham, 2010) and shoot of the moss Bryum argenteum (Bradner et al., 2000). Nevertheless, has been poorly demonstrated the role of fungal endophytes (e.g., Penicillium spp.) in the nitrogen uptake assessed by the isotopic modulation and/or fractionation rates.

Based on our experimental results, we build a conceptual model (see Figure 5) that illustrates the effects of DSE in the nutrient acquisition in the two native vascular Antarctic plants. In the absence of DSE endophytes, E− plants seem to mainly uptake enriched N-sources, either from the enriched NH₄⁺ previously present in the field soil samples, or from the small organic compounds (e.g., amino acids and short-chain peptides) that Antarctic plant species may be capable to uptake (Hill et al., 2019). A proportion of the urea-derived NH₄⁺ (δ¹⁵N -1.5‰), which hydrolyzed spontaneously at the acidic conditions (pH 5.8), found in the soil, could also be uptaken due to the high affinity of plants for this N form, particularly by D. antarctica (Rabert et al., 2017). However, despite the presence of this ¹⁵N-depleted NH₄⁺ source in the substrate of all experimental plants, the ¹⁵N signature in the final tissues of E− plants from both species was less fractionated (δ¹⁵N 2.7–3.1‰), yet, partly depleted relative to the soil substrate (δ¹⁵N 3.7‰). By contrast, the symbiotic interaction left a signature in the foliar tissues of E+ plants that was far more ¹⁵N-depleted (δ¹⁵N −0.4 to −2.3‰) than E− plants relative to the ¹⁵N in the initial substrate, such as has been previously proposed (Hobbie and Högberg, 2012; and references therein). In this sense, this isotopic signature strongly suggests that a large proportion of the N taken up, is preferentially managed through endophytes-mineralized N compounds, probably in the form of NH₄⁺. This may be supported by the higher mineralization registered in the soils from E + individuals from both species compared with their axenic counterparts.

In conclusion, here we corroborate that despite being grown under rich N soils, DSE exert a positive effect in the N-uptake of the two Antarctic vascular plants. This effect was mediated both, by the enhanced availability of inorganic N sources in the substrate such as NH₄⁺, but also by the active participation of fungal endophyte in the process of N-uptake, as suggested by the isotopic signature encountered in the foliar tissues of these plant species. Although, further research is needed to determine the specific routes by which fungal endophytes fulfill this role, here we identify some promising avenues of research to accomplish such a goal.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

IA-R, CT-D, and MM-M designed and performed the experiments. IA-R, AG, and CA analyzed the data. All authors wrote and reviewed the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.