Lichens and bryophytes have increasingly been recognized to play a relevant role in global biogeochemical cycles . They are globally abundant, growing on soils, rocks and epiphytically on trees. At high latitudes, they may form extensive covers on the forest floor, and in wetlands mosses frequently represent the dominant vegetation type. In drylands, lichens and bryophytes form so-called biological soil crusts together with photosynthesizing cyanobacteria, algae, fungi and bacteria. These crusts cover vast areas in arid and semiarid ecosystems.

In a first approach, based on empirical upscaling of field measurements according to ecosystem categories, calculated that lichens and bryophytes, together with free-living cyanobacteria and algae, fix around 14.3 (Gt CO2) year-1 (3.9 Gt carbon) at the global scale. This corresponds to about 7 % of the net primary productivity (NPP) by terrestrial vegetation. In an alternative approach to the empirical upscaling of observations, utilized a process-based non-vascular vegetation model for lichens and bryophytes, called LiBry, to calculate the NPP of these organism groups at the global scale, obtaining similar results.

In addition to photosynthetic carbon uptake, lichens and bryophytes are able to fix nitrogen through symbiosis with cyanobacteria . Together with free-living cyanobacteria, their nitrogen fixation was estimated to sum up to a global value of ∼ 49 (Tg N) year-1 , which accounts for nearly half of the biological nitrogen fixation on land. The LiBry model yielded a similar estimate of up to 34 (Tg N) year-1, based on the nitrogen requirements of lichens and bryophytes determined by . Moreover, it was found in the same study that the organisms may contribute significantly to biotic enhancement of global chemical weathering, by release of weathering agents such as organic acids. Their potential for chemical weathering was derived from their phosphorus demand, assuming that they dissolve surface rocks to acquire phosphorus.

Recently, lichen- and bryophyte-related nitrogen fluxes other than fixation of nitrogen have been shown to be significant at the global scale. found that biological soil crusts, which may contain large fractions of lichens or bryophytes, emit considerable quantities of the reactive trace gases NO and HONO, accounting for ∼ 1.7 (Tg N) year-1. This corresponds to ∼ 20 % of global nitrogen oxide emissions from soils under natural vegetation .

Furthermore, showed that a large variety of lichen and bryophyte species release nitrous oxide (N2O). They estimated that the organisms emit a total value of 0.45 (0.32–0.59) (Tg N2O) year-1 at the global scale, which corresponds to 4–9 % of natural terrestrial N2O emissions . Since N2O is an important greenhouse gas and also the main depleting substance of stratospheric ozone which is still emitted today, quantifying all contributing sources is of high importance .

Absolute values of N2O release estimated by were highest for lichens and bryophytes living on the ground in the boreal zone and for epiphytic lichens and bryophytes in the humid tropics. The relative contributions of lichens and bryophytes to total ecosystem N2O emissions, however, were highest in desert and tundra biomes, due to the low emissions by other vegetation and the soil there. The high relevance of lichens and bryophytes for N2O emissions in drylands and at high latitudes is in accordance with their strong impacts on other components of the nitrogen cycle in these regions. Bryophytes, for instance, have been suggested to be the main source of nitrogen input into boreal forests through fixation from the atmosphere by cyanobacterial partners . Also in drylands, lichens and bryophytes are crucial for input of nitrogen into the ecosystem , and they may even be essential providers of nitrogen for vascular plants .

The estimate by is derived from measuring emissions of N2O by the organisms in the laboratory under a range of environmental conditions. All lichen and bryophyte species analysed by showed release of N2O. Lichens and bryophytes were shown to utilize 15N labelled NO3- but not NH4+, indicating that N2O is likely formed during denitrification. The exact process of N2O formation, however, remains largely unknown. One option is that the organisms themselves release N2O during the metabolization of nitrate, in a similar way to that suggested by for vascular plants. Another option is that bacteria growing on lichen and moss cushions are responsible for the emissions of N2O. This second option is supported by a recently published study, wherein several strains of the bacterial genus Burkholderia, which were shown to emit N2O, were isolated from the boreal peat moss Sphagnum fuscum . While describe that the substrate, which the organisms grew on, was thoroughly removed, further cleaning steps to remove potential bacterial colonies have not been conducted.

Another finding by is that N2O emissions are related to respiration by a relatively constant factor. By applying this factor and, furthermore, assuming a fixed ratio between respiration and NPP, the authors utilized the global NPP data of to obtain globally resolved N2O emissions by lichens and bryophytes. The reliability of global estimates derived from upscaling of small-scale measurements depends on the variation of the measured fluxes. The field measurements of NPP, which were extrapolated to the spatial scale of a biome by , vary by around 2 orders of magnitude. Measurements of N2O emissions by lichens and bryophytes, too, show considerable variation. Regarding biological soil crusts, several studies analysed denitrification rates to be negligible , and N2O production was calculated to constitute only 3–4 % of the N fixation rate . Other studies, however, described high denitrification rates that either increased or decreased with advancing crust development . One possibility to increase the reliability of large-scale estimates of N2O emissions by lichens and bryophytes is the application of alternative, methodically different approaches.

For this reason, we apply here the process-based non-vascular vegetation model LiBry to assess the contribution of these organisms to the global N2O budget. LiBry simulates photosynthesis, respiration and growth of lichens and bryophytes as a function of environmental conditions. To distinguish global patterns of productivity on the ground and in the canopy, the model represents these locations and their differing environmental conditions separately. We calculate respiration by lichens and bryophytes directly as a function of environmental conditions, and we derive N2O emissions based on the simulated respiration. By doing this, we obtain physiologically driven and spatially resolved data on the N2O emissions by lichens and bryophytes at the global scale. Since we estimate respiration with LiBry, we do not need to make assumptions regarding the ratio of NPP to respiration, contrary to . Furthermore, we quantify different sources of variation in N2O emissions and determine their relative importance.

The non-vascular vegetation model LiBry estimates global patterns of photosynthesis, respiration and net primary productivity of lichens and bryophytes . The model calculates these physiological processes as a function of climate and additional environmental conditions, which are provided in the form of time series of global gridded maps. Photosynthesis in LiBry is determined by ambient levels of light, CO2 and temperature according to the Farquhar approach . Respiration is simulated as a function of temperature via a Q10 relationship. Both processes also depend on the water status of the simulated lichens and bryophytes, which includes limitation of CO2 diffusion at high water content. NPP is derived from the difference between gross photosynthesis and respiration. A unique feature of LiBry is that functional diversity of lichens and bryophytes is represented by a large number of artificial species, instead of being aggregated into one or a few average functional types. The advantage of this approach is that adaptation of the organisms to differing environmental conditions is simulated in a more realistic way. Physiological processes such as photosynthesis and respiration are calculated separately for each artificial species. LiBry has been successfully applied to estimate global NPP by lichens and bryophytes and other impacts of these organisms on global biogeochemical cycles .

The model version presented here contains several extensions compared to the original version: first, an NPP-based weighting scheme was introduced, which assigns relative abundances to all artificial species that survive in a grid cell of the model in the steady state . This allows an average grid cell value of NPP based on the relative abundances of the simulated species in that cell to be derived. In the original version, grid cell NPP could only be predicted in the form of a range of values, due to unknown abundances of the species. The average grid cell NPP is close to the upper end of the range of productivity values, since the most productive simulated species are assumed to be the most abundant ones. Secondly, a dynamic disturbance scheme was implemented, which replaces the equilibrium computation of surface coverage by a monthly update of coverage . This makes the new model applicable to transient scenarios of climatic and environmental change, while the original model required the assumption of a steady state to compute coverage.

For this study, we run LiBry with an initial value of 3000 artificial species in each grid cell for a period of 600 years to reach steady state, with climatic fields and other forcing data from . Our global estimates are based on average values over the last 50 years of the simulation. We evaluate the new version of LiBry in the same way as the original one , by comparing simulated NPP to field measurements on a biome basis.

LiBry does not include an explicit representation of processes that directly result in emission of nitrous oxide. However, it has been determined experimentally by that N2O emissions by lichens and bryophytes are related to their respiration by a conversion factor of 16 ng N2O (mg CO2)-1. The conversion factor has a 90 % confidence interval of 11 to 21 ng N2O (mg CO2)-1. Since LiBry explicitly calculates respiration by lichens and bryophytes, we derive N2O emissions from simulated respiration using the conversion factor of .

The study by uses NPP of lichens and bryophytes, together with free-living cyanobacteria and algae, to estimate N2O emissions, since global upscaled data on respiration of these organisms are not available from . Thereby, assume a fixed ratio of respiration to NPP. To determine this ratio, they evaluate literature data, obtaining a rate of respiration relative to net photosynthesis of ∼ 49 % Table S6. Since measurements have been made in the sunlight but respiration continues in the dark, respiration is multiplied by a factor of 2, assuming a 12 h day. This leads to an estimated ratio of respiration to NPP which is roughly 1 : 1. To evaluate LiBry further, we compute the ratio of respiration to NPP in LiBry to assess whether the model is in agreement with these observations.

In the study of , the substrate of the samples was removed to avoid biases resulting from N2O release by microbes in the substrate. Foliose and fruticose lichens as well as mosses were collected which grew on soil, rocks and epiphytically on trees, and the authors found no variation in N2O emissions depending on the underlying substrate. Endolithic and crustose lichens were not included in that study, as for these growth forms the dry weight, which is needed for calculations, could not be determined in a reliable manner.

Variation in field measurements of N2O emissions may result not only from physiological differences between species but also from variation in climatic conditions, which can be significant at the small scale. To upscale emissions from point measurements to the large scale, it is important to quantify the relative contributions of these different sources of variation. When, for instance, the variation between species regarding their N2O emissions is small, it suffices to sample a low number of species to obtain an average emission for a certain climatic condition. LiBry can provide an indication of the relative importance of these sources of variation, since the model not only represents climate variability but also simulates diverse physiological strategies. Each grid cell of the model contains a range of surviving species at the end of the simulation and that, consequently, shows a range of N2O emissions. LiBry does not simulate spatial variation in climatic conditions within a grid cell. However, by comparing average emission rates of grid cells from different climates, it is possible to assess the relative importance of climatic conditions for variation in N2O emissions. We select five model grid cells from different ecosystem classes to analyse the relative importance of differences between species and climatic heterogeneity on variation in N2O emissions. It should be pointed out that LiBry does not compute directly N2O emissions by lichens and bryophytes, but it derives them from simulated respiration through an empirical linear relationship. Hence, differences in the sensitivities of respiration and N2O emissions to climatic conditions may lead to uncertainties in our predicted effects of climate on N2O emissions.

The global distribution of net primary productivity simulated by the updated version of LiBry is shown in Fig. . Productivity by lichens and bryophytes is highest in forested regions and lowest in deserts and agricultural regions. Hence, the spatial pattern is mainly controlled by water availability, except for cropland. In LiBry, it is assumed that lichens and bryophytes only grow on the area fraction of a grid cell which is not occupied by crops. Therefore, on a grid cell basis, regions with a high fractional cover of cropland show low productivity by lichens and bryophytes, in spite of favourable climatic conditions. The high productivity in the humid tropics mainly results from epiphytic lichens and bryophytes in the canopy, while in the boreal zone the larger fraction of productivity stems from the ground.

As a result of the dynamic surface coverage, the spatial patterns of NPP differ slightly between the new and the original version of LiBry, but the large-scale gradients remain the same. Comparing the global pattern of lichen and bryophyte NPP simulated by the new version of LiBry to an empirical estimate by shows good agreement, similar to the original version. Furthermore, the total global NPP predicted by the new LiBry differs from the original estimate due to the updated calculation of coverage. The main difference is found for the tropical forest canopy, where simulated NPP increases significantly. The total global NPP of 4.3 (Gt C) year-1 estimated by the new LiBry compares well to the value of 3.9 (Gt C) year-1 calculated by .

Comparison of simulated NPP to field measurements on a biome basis suggests that LiBry predicts realistic values of NPP for a range of ecosystems (Fig. ). In particular, simulated NPP in the tropical and the boreal forest matches well with observations, while the original version of LiBry seemed to underestimate NPP in these biomes. In the biomes desert and, to a lesser extent, tundra, LiBry seems to overestimate productivity, which may have also been the case with the original version. A potential explanation for this is that productivity in dry and cold areas may be limited not only by climatic factors but also by nutrient availability . Since photosynthesis and growth are only controlled by climatic factors in LiBry, the effect of spatial variation in nutrient availability on productivity cannot yet be simulated. It should be pointed out, however, that, except for the boreal biome, the number of field measurements is quite low and, consequently, the observation-based characteristic values for each biome are subject to considerable uncertainty.

Figure shows simulated global patterns of nitrous oxide emissions by lichens and bryophytes. Nitrous oxide emission is highest in the humid tropics and subtropics with values up to 10 (mg N2O) m-2 year-1 (Fig. a). A second region of high emissions is the boreal zone with values up to 8 (mg N2O) m-2 year-1. Dry regions show lowest values of nitrous oxide emissions, in general less than 1 (mg N2O) m-2 year-1. Considering only lichens and bryophytes which grow as epiphytes in the canopy (Fig. b), emissions in the humid tropics are around 3 times higher than in the boreal and temperate zones. Lichens and bryophytes on the ground show highest values of nitrous oxide emissions in the boreal zone, with values around 3 (mg N2O) m-2 year-1 (Fig. c). Regarding the ground, tropical and subtropical regions only partly show N2O emissions comparable to those of the boreal zone. The reason for this is low simulated productivity and coverage of lichens and bryophytes on the ground in tropical and subtropical climates, which also leads to low respiration on a grid cell level and hence to low N2O emissions.

Figure shows the simulated global spatial distribution of the ratio of respiration to NPP. The assumption of a globally constant ratio of respiration to NPP is used by to derive ecosystem-scale N2O emissions by lichens and bryophytes from their NPP. Alternatively, this ratio can be derived from the independent LiBry estimates of NPP and respiration. The simulated ratio shows a latitudinal pattern with increasing values towards the tropics (Fig. a). This results from the influence of surface temperature on respiration in combination with high nighttime temperatures in the humid tropics, which cause high respiration rates during the night. Note that high respiration relative to NPP of tropical lichens and bryophytes does not necessarily mean high respiration at the grid cell level, since net productivity and coverage may be low. Respiration by lichens and bryophytes in the canopy shows a slightly weaker latitudinal gradient than on the ground, which can be explained by efficient evaporative cooling in the canopy (Fig. b). In contrast, lichens and bryophytes on the ground usually grow within the surface boundary layer, which reduces cooling by turbulent heat transfer, leading to a strong influence of incoming radiation on surface temperature. Since radiation input increases toward the Equator, the ratio of respiration to NPP on the ground in the tropics is markedly higher than at high latitudes (Fig. c). The ratio of respiration to NPP varies from less than 1 to around 2, while most values are around 1. This means that gross primary productivity (GPP) is partitioned roughly equally into NPP and respiration, which agrees well with the observational data from .

An overview of global total values of N2O emissions, respiration, NPP and the ratio of respiration to NPP estimated by LiBry is shown in Table . Table  shows N2O emissions by lichens and bryophytes for individual grid cells from five different ecosystem classes (see also Table ). Variation in emissions between species within a grid cell is large; it can exceed 3 orders of magnitude. The variation due to climatic conditions is smaller, but it still amounts to almost 2 orders of magnitude based on the grid cells with the highest and lowest average emission rates. Comparing Table  to the global range of N2O emissions by lichens and bryophytes (Fig. ) shows that the five selected grid cells represent well the global variation in emissions due to climatic conditions. Thus, both functional diversity of the artificial species and different climatic conditions are important for variation of N2O emissions, according to the LiBry simulation.

In this study we estimate nitrous oxide emissions by lichens and bryophytes with the global process-based non-vascular vegetation model LiBry. Thereby, we derive N2O emissions from respiration fluxes which are, together with photosynthesis and net primary productivity, simulated by LiBry.

We use an updated version of LiBry which contains significant modifications with regard to the original version published in . Regarding NPP, the new version estimates 4.3 (Gt C) year-1, while the original version of LiBry predicted a range of 0.34 to 3.3 (Gt C) year-1. The increase in predicted NPP is mainly attributed to a higher simulated productivity in the tropical forest canopy, since a new disturbance scheme allows for a higher surface coverage of lichens and bryophytes there. An empirical global estimate of NPP by lichens, bryophytes, free-living terrestrial cyanobacteria and algae amounts to 3.9 (Gt C) year-1. Our new estimate is higher than that by , although LiBry does not consider free-living cyanobacteria and algae. This may be explained by the small contribution of cyanobacteria and algae to the overall global carbon uptake, which can be compensated for by minor relative changes in productivity of lichens and bryophytes . It is not straightforward to determine which number is closest to reality, since both the process-based estimate by LiBry and the empirical one by are subject to uncertainty. In the study of , for instance, it is assumed that productivity and active time are uniform within a biome. Furthermore, use a globally uniform value of surface cover fraction to scale up local field measurements of productivity to the global scale. However, values of surface coverage by lichens and bryophytes compiled by vary greatly at the small scale, which makes upscaling to larger scales challenging.

While productivity estimated by LiBry is evaluated in this study, large-scale surface coverage of lichens and bryophytes simulated by LiBry has been evaluated for regions north of 50∘ N in . It was shown that LiBry predicts realistic values of cover fraction. Moreover, values of surface cover predicted by LiBry for other regions of the world are in agreement with the estimate of . In spite of uncertainties regarding productivity and abundance of lichens and bryophytes, comparing the empirical and process-based approaches gives confidence in the order of magnitude of the LiBry simulation results.

As a 50-year steady-state average value, we estimate total N2O emissions by lichens and bryophytes of 0.27 (0.19–0.35) (Tg N2O) year-1, which is at the lower end of the range of 0.32 to 0.59 (Tg N2O) year-1 calculated by . The evaluation of LiBry regarding simulated NPP shows that our global patterns and total values of NPP are very similar to the empirical estimate by . Since use this NPP estimate by to derive N2O emissions, differences in NPP are most likely not the reason for our lower estimate of N2O emissions compared to . Instead, this may be explained by differing methods to compute respiration: while assume a globally uniform ratio of respiration to NPP of a value of 2 to estimate respiration, LiBry simulates respiration independently as a species-specific function of temperature and water status. This results in a lower global average value of around 1 for the ratio of respiration to NPP predicted by LiBry. Our estimated ratio of respiration to NPP agrees well with laboratory measurements, but it is in general difficult to compare a global ecosystem-scale value to small-scale and short-term observations.

Our simulated global pattern of N2O emissions is slightly different than that shown in , who estimate highest values in the boreal zone and only intermediate values in the humid tropics. This can be explained by their assumed constant ratio of respiration to NPP, which makes their global pattern of N2O emissions identical to that of NPP, which is shown in . In LiBry, however, the simulated ratio of respiration to NPP increases towards higher surface temperatures in the tropics (Fig. ). Furthermore, the ratio shows large spatial variation. Evaluating this simulated pattern is difficult, since estimates which are extrapolated to the large scale, such as the NPP estimate by , are not available for respiration by lichens and bryophytes. However, observed ratios of respiration to NPP of lichens and bryophytes vary considerably at the species level, as shown by, for example, . Using a constant ratio of respiration to NPP may therefore introduce a bias in the estimated spatial distribution of N2O emissions.

Small-scale measurements of N2O emissions by lichens and bryophytes may show considerable variation. The sources of this variation may be physiological differences between species, variation of associated microbial communities and/or heterogeneity in climatic conditions. We examine the relative importance for respiration of differences between species compared to climatic differences with LiBry, since the model simulates various physiological strategies and represents variation in climatic conditions at the global scale. Thereby, we assume that the relationship between respiration and N2O emissions is relatively insensitive to climatic conditions and physiological differences between species, as suggested by the experiments by . Table  shows that both differences between artificial species and different climatic conditions are important for variation of N2O emissions. Upscaling of N2O emission rates measured in the field may therefore be subject to considerable uncertainty. Modelling approaches in this direction should probably account for both interspecific variation in processes associated with N2O release by lichens and bryophytes as well as variation in climatic conditions.

Although our approach considers the most important sources of variation in N2O emissions by lichens and bryophytes, it is associated with uncertainties that should be discussed further. These uncertainties mainly result from our method to estimate respiration and from assumptions concerning the empirical relationship between respiration and N2O emissions.

Respiration and the ratio of respiration to NPP simulated by LiBry are difficult to validate, since the number of laboratory or field studies which measure not only NPP but also GPP and respiration is not very high. Moreover, long-term measurements of respiration would be required to determine the ratio of respiration to NPP. Otherwise, assumptions about the contribution of respiration in the dark to total respiration are necessary.

To obtain N2O emissions from respiration, our results rely on the laboratory incubation measurements and the calculated ratio of N2O emissions to respiration presented in . Furthermore, our approach considers effects of variation in climatic conditions on N2O emissions by lichens and bryophytes. Hence, it is necessary to discuss the sensitivity of the relationship between respiration and N2O emissions to a range of climatic conditions. As shown in Fig. 3, the relationship between respiration and N2O emissions seems to be insensitive to temperature changes for the tested species. Likewise, variations in water content have no clear effect on the relationship between N2O release and respiration Fig. S3. Although the sensitivities of N2O release to temperature and water content are similar to those of respiration across species, the relationship between N2O release and respiration shows interspecific variation. However, in spite of a large number of around 40 sampled species, the relationship shows a relatively narrow 90 % confidence interval of 11.3 to 20.7 ng N2O (mg CO2)-1 . This suggests that the mechanism of N2O release by lichens and bryophytes is similar between different species.

To analyse the relation between the production of N2O and respiratory CO2 in greater detail, measurements of both fluxes by means of online gas exchange measurements would be needed, which then could be linked to the observed water status of the organisms. Since LiBry explicitly represents the dynamic water saturation of lichens and bryophytes, this would allow a more process-based prediction of the duration and magnitude of N2O emissions. In this way, the uncertainty associated with our approach would be reduced, facilitating an improved estimate of global N2O emissions by lichens and bryophytes.

In order to assess model-based estimates of N2O emissions by lichens and bryophytes, a relatively large number of field measurements are necessary. Currently, most N2O measurements, independent of the substrate or organisms measured, generally suffer from major uncertainties, additionally to variation from functional diversity and differing climatic conditions: first, the majority of these studies have been conducted using the acetylene inhibition technique. The idea of this method is to inhibit the last denitrification step, so that the measured N2O amounts should reveal the sum of N2O and N2 release during denitrification under natural conditions. It has, however, been shown quite a while ago that this method leads to an underestimation of denitrification under oxic conditions . Secondly, the most widely used measuring technique has been the closed-chamber method, which is inexpensive and easy to use. This, however, has major shortcomings, as environmental conditions are hard to control and only limited surface areas can be measured . Furthermore, the limited temporal resolution of chamber measurements may affect estimated N2O emissions . Thirdly, depending on the environmental conditions under which the experiment is performed – particularly water, temperature, and nutrient conditions – the obtained N2O emission rates could differ widely. Thus, it is indispensable to report and consider the exact environmental conditions under which the measurements were made and to restrict natural emission data to those assessed under typically occurring natural conditions.

Respiration by lichens and bryophytes is not the only process which can be used to estimate their N2O emissions. report a relationship between nitrogen fixation and N2O release in biological soil crusts, which include lichens and bryophytes, as well as soil bacteria and algae. For this approach, however, reliable nitrogen fixation data are sparse. It is also possible to estimate the demand for nitrogen by lichens and bryophytes with LiBry with an uncertainty range of around 1 order of magnitude . However, it is not straightforward to derive realized nitrogen uptake or nitrogen fixation from this, since LiBry does not yet include processes related to nitrogen uptake or metabolization of nitrogen species. Therefore, for this study, we chose the relation between respiration and N2O release to quantify N2O emissions by lichens and bryophytes.

Our simulated global N2O emissions by lichens and bryophytes of 0.27 (0.19–0.35) (Tg N2O) year-1 amount to around 3 % of global N2O emissions from natural sources on land . This value may sound low at first glance, but it equals about 50 % of the atmospheric deposition of N2O into the oceans or 25 % of the deposition on land . Considering that N2O has a strong negative effect on stratospheric ozone and a significant warming potential as a greenhouse gas, even relatively small emissions should not be neglected in global budgets.

The study by estimates global patterns of N2O emission from soils and finds that the humid tropics contribute most to global N2O emission due to high temperature and precipitation. Our simulated pattern of global N2O emissions by lichens and bryophytes also shows a hotspot in the humid tropics, but the relative contribution of the boreal zone to the global flux seems to be higher than in . This probably results from the high simulated NPP in the boreal zone, particularly on the ground, which compensates for the lower respiration and therefore N2O emission per productivity due to low temperatures. Relative contributions of lichens and bryophytes to N2O emissions are highest for ecosystems in desert regions and at high latitudes, which agrees with the results by .

We estimate large-scale spatial patterns and global values of N2O emissions by lichens and bryophytes from a process-based model of their productivity and respiration. Our results suggest a significant contribution of lichens and bryophytes to global N2O emissions, albeit at the lower end of the range of a previous empirical estimate. Since both approaches use respiration to derive N2O emissions, our lower estimate likely results from a different method to predict respiration, compared to the empirical approach. Hence, while estimates of productivity are relatively well constrained, evaluating models with regard to estimated respiration may improve predictions of N2O emissions by lichens and bryophytes. One important finding derived from our simulation is that the ratio of respiration to NPP by lichens and bryophytes shows spatial variation and a latitudinal gradient at the global scale. This means that productivity and N2O emissions by the organisms are not necessarily correlated and that tropical regions may show higher emissions than polar regions given the same NPP. Furthermore, we show that both physiological variation among species and variation in climatic conditions are relevant for variation in respiration and, consequently, N2O emissions. Ecosystem-scale estimates of N2O emissions by lichens and bryophytes should therefore include sufficient ranges of species and climatic conditions to avoid biased results. Our results build on the empirical finding that N2O emissions by lichens and bryophytes are linearly related to their respiration. This relationship is relatively insensitive to climatic conditions and shows no large variation between species. However, the relationship is based on closed-chamber measurements. Therefore, it would be useful to perform online gas exchange measurements of N2O emissions and respiration to test the effect of climatic conditions on the relationship between N2O release and respiration. Furthermore, using alternative approaches to estimate N2O emissions by lichens and bryophytes may be helpful to constrain our approach.