Nitrogen fixation in mosses is the primary input of N in pristine ecosystems. Nitrogen deposition is expected to increase due to increased human influence with the presumed outcome that N2 fixation will decrease. However, N deposition consists of ~20% organic N, and we have very limited knowledge about how this influences N2 fixation in mosses. This study suggests that organic N has very different and often positive effects on N2 fixation. Below we discuss the details and consequences of these findings.
Effects of N additions on moss-associated N2 fixation
Nitrogen additions increased N content in the mosses, which indicated that the mosses absorbed the exogenous N. Hence, we can address our hypotheses. We hypothesized (H1) that N2 fixation in two common moss species would decrease with increasing N additions. However, this was only partly confirmed, as N2 fixation response to N additions was highly dependent on N form and moss species. The realistic NH4NO3 addition (0.4-4 kg N ha−1 yr−1) followed the expected pattern with lower N2 fixation rates with increasing rates as also seen in boreal forest (Gundale et al. 2011; Rousk & Michelsen 2016). Interestingly, addition of alanine and urea had positive or no effects on N2 fixation in this experiment (see below). Usually, the nitrogenase enzyme is depressed by its end product, i.e., ammonia. Besides, NH4NO3 causes acidification (Tian & Niu 2015), and low pH can inhibit N2 fixation in mosses (Alvarenga & Rousk 2021), as a result of suppressed nitrogenase enzyme activity (see below).
According to our second hypothesis (H2), more complex N forms, alanine and urea, showed a higher threshold for inhibiting N2 fixation and often benefitted N2 fixation at low rates compared to NH4NO3, as NH4NO3 inhibited activity already at 2-4 kg N ha−1 yr−1 for P. schreberi and around 10 kg N ha−1 yr−1 for S. capillifolium, while 10 and 20 kg N ha−1 yr−1 with alanine addition for P. schreberi and S. capillifolium, respectively, and urea inhibited N2 fixation only at the highest rate (20 kg N ha−1 yr−1) for both species (Figs. 2, S1). Surprisingly, realistic additions of alanine and urea did not inhibit N2 fixation in either species, but rather, promoted N2 fixation. A possible explanation is that the organic N forms were allocated towards growth by both cyanobacteria and moss and therefore, did not inhibit N2 fixation (Krausfeldt et al. 2019; Liu et al. 2013; Rawson 1985). Indeed, phycocyanin concentration, as a measure of cyanobacterial abundance, did increase in both species after adding organic N (only a trend in P. schreberi) but not after adding inorganic N. Moreover, organic nitrogen could also act as a carbon source for cyanobacteria (Krausfeldt et al. 2019), which could save the cost for photosynthesis and allow more energy to be invested to fix N2. Another positive, but indirect effect of organic N, is the increase in moss pH that can promote N2 fixation activity (e.g. (Alvarenga & Rousk)). Further, amino acids can be absorbed and utilized directly by moss and cyanobacteria, and the assimilation cost of amino acids is considered to be lower than that of NH4+ and much lower than that of NO3− (Liu et al. 2013; Song et al. 2016). Urea likely offers the greatest energetic advantage because urea hydrolysis by urease results in the production of two N containing molecules (Herrero et al. 2001). Moreover, the breakdown of urea results in the release of CO2 as a by-product, which can be incorporated into photosynthesis, reducing the reliance on active uptake (Glibert et al. 2014). Hence, different uptake – and metabolism mechanisms lead to the diverse response patterns of N2 fixation towards different types of N. This is also reflected in the higher moss tissue N content in both investigated moss species after organic than inorganic N additions, suggesting different uptake strategies or requirements for inorganic vs. organic N (Krab et al. 2008; Liu et al. 2013).
Under the extreme N addition scenario, all three N forms inhibited N2 fixation, although inorganic N had the strongest effect. The inhibition could be explained by pH stress and toxicity caused by high N concentrations. In this study, high rates of N additions changed the pH optimum for N2 fixation – above or below the pH optimum in the control samples, depending on moss species. The pH optimum for N2 fixation in both species was lower than the pH of 5.9 ~ 6.2 found by Smith (1984) and may be due to different moss species investigated. High rates of added NH4NO3 decreased pH while alanine and urea increased moss pH. Rawson (1985) found that several amino acids affected nitrogenase and appeared to be toxic at high concentrations in culture (10mM). The hydrolysis of urea produces two ammonia molecules, which can be protonated to form two NH4+ molecules and cause an increase in pH (Carlini & Ligabue-Braun 2016; Herrero et al. 2001; Veaudor et al. 2019), thus under the extreme urea addition, high NH4+ accumulation and high pH (average pH in both species was 5.67 after urea addition) may account for inhibition of N2 fixation. Even though we found inhibition of N2 fixation by alanine and urea at high rates, in natural ecosystems, organic N contributes 20-30% to total N deposition (Cornell 2011; Violaki et al. 2010), which means in natural ecosystems organic N inputs are much lower than our extreme addition rates and could promote N2 fixation. Besides, in natural environments available N is always complex and mixed. A preference to take up different N forms such as ammonium, nitrate, amino acids, urea should therefore be taken into account when considering the N addition effects on N2 fixation (Andersen et al. 2020; Liu et al. 2013).
Recovery of N2 fixation after N additions
Since organic N did not inhibit N2 fixation in the realistic N addition scenario, recovery from the N stress can strictly speaking not occur. Nonetheless, given that we expected an inhibition of N2 fixation by all N forms, we deprived the mosses of N for 2 weeks, and we found inhibition after the extreme N addition scenario, we still define the N2 fixation rates during the N deprivation period as recovery rates. In line with our third hypothesis (H3), we found evidence for higher recovery of N2 fixation after lower N additions than in the extreme N addition scenario. Also, a higher recovery rate was found with urea addition than with NH4NO3 or alanine addition in both moss species during the N deprivation after the realistic scenario but not after the extreme scenario, which may be because low rates of urea additions increased moss pH (5.22~5.52 for P. schreberi and 4.68~4.92 for S. capillifolium) more than NH4NO3 and alanine addition, creating an environment conducive to N2 fixation. Urea also provides both C and N to cyanobacteria (Krausfeldt et al. 2019). Recovery after N additions under the realistic scenario suggests that the cyanobacteria may have down-regulated N2 fixation during the N additions, since N2 fixation is an energy consuming process (Sohm et al. 2011; Turetsky 2003). As soon as N availability is decreasing (N deprivation phases), cyanobacteria start fixing N2 again. After extreme N additions, however, N2 fixation did not recover. Previous studies showed that recovery from high N loads needs a longer time or N needs to be actively removed via e.g. rinsing (Rousk et al. 2014a; Rousk & Michelsen 2016). Therefore, recovery from N loads is possible, if N input remains below a certain threshold.
Moss species-specific responses to N addition
Although acetylene reduction rates were higher in S. capillifolium than in P. schreberi in this study, the rates still likely underestimate the actual N2 fixation as in Sphagnum, the most dominant diazotrophs are methanotrophs (Bragina et al. 2013; Leppänen et al. 2015), whose activity is suggested to be inhibited by acetylene. Yet, the high acetylene reduction rates in Sphagnum in our study indicate either a high abundance of cyanobacteria present, which are not inhibited by acetylene, or not all methanotrophs are inhibited by acetylene during the incubation period (Rousk et al. 2018).
Throughout the experiment, P. schreberi and S. capillifolium had comparable average nitrogenase activity in the control and urea treatments. However, nitrogenase activity in P. schreberi dropped 3 times after NH4NO3 addition and halved after alanine additions compared to S. capillifolium. The cumulative N2 fixation rates in P. schreberi with NH4NO3 addition was suppressed but not in S. capillifolium, and we only found a positive relationship between phycocyanin and N addition rates in S. capillifolium in the alanine and urea treatment. All these results only partly supported H4, stating that P. schreberi would be more sensitive to N additions than S. capillifolium. Nevertheless, response differences between the moss species could be identified. The higher sensitivity of N2 fixation in P. schreberi towards increased inorganic N may be due to the different colonization locations of cyanobacteria in the two moss hosts. Cyanobacteria colonize the leave surface of P. schreberi (DeLuca et al. 2002), causing them to be exposed to the environment directly. Sphagnum mosses harbor microorganisms both on the surface and inside their hyaline cells. Hyaline cells provide a relative stable living space where diazotrophs are protected from N stress (Bragina et al. 2012). Also, S. capillifolium, and Sphagnum species in general, are common in boreal peatlands with usually low pH (Kostka et al. 2016; Turetsky et al. 2012), which could explain why N2 fixation in this species was less responsive to extreme NH4NO3 additions leading lower pH (Fig. S2) compared to P. schreberi. Higher regression slopes between N content and N addition rates found in P. schreberi also suggests that P. schreberi took up more of the added N and could be therefore more sensitive to N inputs than S. capillifolium, leading to P. schreberi had lower threshold toward N inputs (Fig. 1).