Simulating ectomycorrhiza in boreal forests : implementing ectomycorrhizal fungi model MYCOFON into CoupModel ( V 5 )

Ectomycorrhizal fungi (ECM), the symbiosis between a host plant and mycorrhizal fungi, has been shown to considerably influence the C and N flux between soil, the rhizosphere and plant in boreal forest ecosystems. However, ECM is either neglected or presented as an implicit, non-dynamic term in most ecosystem models which 15 can potentially reduce their predictive power. In order to investigate the necessity of an explicit consideration of ECM in ecosystem models, we implemented the previous developed MYCOFON model into a detail process-based soil-plant-atmosphere model, CoupModel. MYCOFON explicitly describes the C and N fluxes between ECM and roots. This new Coup-Mycofon model approach (ECM explicit) is compared to two simpler model approaches, of which one contains ECM implicitly as 20 an non-dynamic N uptake function (ECM implicit) and the other represents a version where plant growth has a constant N availability (nonlim). Parameter uncertainties are quantified by using Bayesian calibration where the model outputs are constrained to current forest growth and soil conditions for four forest sites along a climate and N deposition gradient in Sweden over 100 year period. Our results show that the nonlim approach could not describe both the forest growth and soil C and N conditions 25 properly. The ECM implicit/explicit approach is able to describe current conditions with acceptable uncertainty. The ECM explicit Coup-Mycofon model provide a more detailed description of internal ecosystems fluxes and feedbacks of C and N fluxes between plant, soil and ECM. Our modelling highlights the need of incorporating ECM in current ecosystem models. We also provide a key set of posterior fungal parameters which can be further investigated and evaluated in future ECM studies. 30


Introduction
Boreal forests cover large areas on the Earth's surface and are generally considered substantial carbon (C) sinks (Dixon et al., 1994;Pan et al., 2011).The sink strength is determined through the balance between major C uptake and release processes, i.e., plant photosynthesis and both autotrophic and heterotrophic respiration, and is largely controlled by nitrogen (N) availability (Magnani et al., 2007;Högberg et al., 2017).Numerous studies have shown that soil N availability is the main driver for plant and microbial dynamics (Vitousek and Howarth, 1991;Klemedtsson et al., 2005;Lindroth et al., 2008;Luo et al., 2012;Mäkiranta et al., 2007;Martikainen et al., 1995).Thus, a proper description of N dynamics in ecosystem models is prerequisite for precisely simulating plant-soil C dynamics and greenhouse gas balance (Maljanen et al., 2010;Schulze et al., 2009;Huang et al., 2011).Ecosystem models, however, vary considerably in Published by Copernicus Publications on behalf of the European Geosciences Union.their representation of N fluxes: from very simplified presentations (e.g., the LPJ-GUESS model: Sitch et al., 2003;Smith et al., 2011) to very complex approaches that aim to capture the whole N cycle (e.g., LandscapeDNDC: Haas et al., 2012;CoupModel: Jansson and Karlberg, 2011).
Ectomycorrhizal fungi (ECM) are common symbionts of trees in boreal forests.ECM are more efficient than roots in taking up different N sources from the soil (Plassard et al., 1991), and they store vast amounts of N in their tissues (Bååth and Söderström, 1979) and can cover a large fraction of their host plants' N demand (Leake, 2007;van der Heijden et al., 2008).Further, ECM are shown to respond sensitively to ecosystem N availability and are generally considered as adaptation measures to limited N conditions (Wallenda and Kottke, 1998;Read and Perez Moreno, 2003;Kjoller et al., 2012;Bahr et al., 2013;Choma et al., 2017).Previous research showed that ECM can receive between 1 and 25 % of the plants' photosynthates and constitute as much as 70 % of the total soil microbial biomass, thus having a major impact on soil C sequestration in boreal forests (Staddon et al., 2003;Clemmensen et al., 2013).Overall, the functions and abundance of ECM fungi constitute numerous pathways for N turnover in the ecosystem and considerably influence the magnitude and dynamics of C and N fluxes.
Nevertheless, ECM have rarely been considered in ecosystem models (for an overview about modeling ectomycor-rhizal traits, see Deckmyn et al., 2014).To our knowledge, only five ecosystem models have implemented ECM to various degrees: the ANAFORE model (Deckmyn et al., 2008), the MoBiLE environment (Meyer et al., 2012), the MySCaN model (Orwin et al., 2011), and more recently the Moore et al. (2015) and Baskaran et al. (2016) ECM models (Table 1).In the ANAFORE model, ECM are described as separate C and N pools.However, this model does not distinguish between mycorrhizal mycelia and mantle.The C allocated from the host tree to ECM is simulated as a zero-order function, further regulated by nutrient and water availability.ECM can also facilitate organic matter decomposition in the ANAFORE model.The MySCaN model uses a similar approach for ECM C uptake and dynamics but does not, to our knowledge, include the influence of water availability on ECM.In both models, ECM transfer of N to the host is regulated by the C / N ratios of the plant and fungi.In the MoBiLE model, C allocation to ECM is more complex than that in the ANAFORE and MySCaN models, and the N allocation to the host by the ECM can feed back into their C gains.However, the N allocation to the host plant is described similarly to the other two models.In MoBiLE, mycorrhiza are further distinguished between mycelia and mantle, but can neither degrade organic matter nor take up organic N forms.Mycelia and mantle differ in their capacity to take up N, and the mantle has a slower litter production rate than that of mycelia.Both the Moore et al. (2015) and Baskaran et al. (2016) ECM models represent the ECM as a separate model pool and explicitly simulate ECM decomposition, but with much simpler process descriptions, and the interaction with environmental functions are neglected (Table 1).
The overall aim of this study is to improve understanding of ecosystem internal C and N flows related to symbiosis between ECM and a host tree in order to improve the model predictive power in assessment of C sequestration and climate change.This is carried out by presenting a new version of the CoupModel, coupled with an explicit description of ECM.Specifically, we implement the previously developed MYCOFON model (Meyer et al., 2010) into the wellestablished soil-plant-atmosphere model, CoupModel (Jansson, 2012).We choose the MYCOFON model because, first, it contains a very detailed description of ECM fungal C and N pools and all major C and N ECM exchange processes (i.e., litter production, respiration, C uptake, N uptake), and second, ECM can also additionally respond to the soil N availability (Table 1).Therefore, ECM growth and N uptake, both mineral and organic N forms, respond dynamically to environmental functions and plant C supply in the new Coup-MYCOFON model (Fig. 1).This detailed "ECM explicit" modeling approach (hereafter called ECM explicit) is further compared with two simpler modeling approaches -the "ECM implicit" and "nonlim" approaches -which already exist in CoupModel.The ECM implicit approach does not represent the ECM as a separate pool but incorporates ECM into the roots implicitly.Plants are thus allowed to take up additional organic N sources from soil organic pools and do not respond to environmental functions.The ECM implicit approach has been used in a similar way by Kirschbaum and Paul (2002) and Svensson et al. (2008a).The nonlim approach assumes an "open" N cycle and plant growth is limited by a constant N availability to a static fixed level (e.g., in Franklin et al., 2014).These three ECM modeling approaches constitute most of the current ECM representations in ecosystem models and are tested by four forest sites situated along a climate and N fertility gradient across Sweden (Fig. 2).Bayesian calibration is used to quantify the uncertainty of model parameters and identify key parameter sets.

Model description
The CoupModel ("Coupled heat and mass transfer model for soil-plant-atmosphere systems"; Jansson and Karlberg, 2011) is a one-dimensional process-oriented model, simulating all the major abiotic and biotic processes (mainly C and N) in a terrestrial ecosystem.The basic structure is a depth profile of the soil for which water and heat flows are calculated based on defined soil properties.Plants can be distinguished between understory and overstory vegetation, which allows simulation of competition for light, water, and N between plants.The model is driven by climate data -precipitation, air temperature, relative humidity, wind speed, and global radiation -and can simulate ecosystem dynamics in hourly, daily, and yearly resolutions.A general structural and technical overview of the CoupModel can be found in Jansson and Moon (2001) and Jansson and Karlberg (2011), and a recent overview of the model was also given by Jansson (2012).The model is freely available at www.coupmodel.com.The CoupModel (v5) is complemented with an ectomycorrhizal module (MYCOFON; Meyer et al., 2010), which allows the direct simulation of the C and N uptake processes of ECM.The MYCOFON (v1) model is described in detail by Meyer et al. (2010), and here only the key processes of plant and ECM fungal growth, N uptake, and litterfall and respiration are described.

Plant growth in CoupModel
An overview of model functions is given in the Appendix, Table A1.Plant growth is simulated according to a "radiation use efficiency approach" in which the rate of photosynthesis is assumed to be proportional to the global radiation absorbed by the canopy, but limited by temperature, water conditions, and N availability (Eq. 1, Table A1a).Assimilated C is allocated into five different plant C compartments: C root , C leaf , C stem , C grain , and C mobile .The same compartments also represent the corresponding N amounts.The "mobile" pool (C mobile , N mobile ) contains embedded reserves that are reallocated during certain time periods of the year, e.g., during leafing.Respiration is distinguished between maintenance and growth respiration, where a Q 10 function response is used (Eq.2.1, 2.2; Table A1a).Plant litter is calculated as fractions of standing biomass (Eq.3, Table A1a).

ECM fungal C and N pools
The ECM are closely linked to the trees' fine roots and consist of C and N pools.The C pool is distinguished between the mycelia, which are responsible for N uptake, and the fungal mantle, which covers the fine-root tips.The C pool is the difference between C gains from plant supply and C losses due to respiration and litter production (Eq.8.1, Table A1b).Accordingly, the fungal N pool is the result of the difference between N gains by uptake, N losses by litter production, and N transfer to the plant (Eq.8.2, Table A1b).ECM fungal C and N pools distinguish between mycelia and mantle, which is of importance when simulating N uptake (only the mycelia are able to take up N), and also when simulating litter production if the more complex approach is chosen (see Sect. 2.1.4).The ratio between mycelia and mantle is determined by the parameter FRAC MYC , which defines the fraction of mycelia C in total ECM fungal C. For all other N and C exchange processes (growth, respiration, and N transfer to plant), the separation between mycelia and mantle is disregarded.

Growth of ectomycorrhizal fungi
ECM growth is limited by a defined maximum, i.e., only a certain amount of tree host assimilates are directed to the ECM.This maximum ECM growth is determined by a potential C supply from the plant and limited by N availability (Eq.5.1, Table A1b).The C supply is defined by a constant fraction of the root C gain and is leveled off by the function f (c fungiavail ) as soon as a defined value of soil-available total N is exceeded, i.e., in the model the potential ECM growth declines with rising soil N.This scaling function is based on observations from field and laboratory experiments, which showed that the ECM biomass of mycelia and mantle can be as much as 30-50 % of fine-root biomass, and the majority of ECM decrease in abundance and functioning when the soil N levels are high (e.g., Wallander, 2005;Wallenda and Kottke, 1989;Högberg et al., 2010).The actual ECM growth is limited by the maximum growth and calculated by a predefined fraction of assimilated root C, assuming that the production of an optimum mycorrhization degree requires a certain amount of ECM biomass (Eq.5.2, Table A1b: FRAC OPT ×c frt ).This is further dependent on the N supply from the ECM to the roots, f (n supply ).The model thus follows the assumption that plants feed the ECM with C as long as their investment is outweighed by the benefits obtained (Nehls et al., 2008).A minimum C supply to prevent ECM fungi death during C shortage is guaranteed by the term during time periods when plant photosynthesis is limited and belowground C supply to root and ECM becomes zero (Eq.5.3, Table A1b).

Respiration and litter production of ectomycorrhizal fungi
Respiration is separated into two components (maintenance and growth) for both ECM and root respiration (see Eqs. 2 and 6, Table A1).Two approaches are available to simulate ECM fungal litter production that differ in complexity.The simple approach (Eq.7.1, 7.2; Table A1) uses one common litter rate L for both the fungal mantle and mycelia.Consequently, possible specific effects of the mantle and mycelia tissue on litter production are neglected.The alternative "detailed" approach (Eq.7.3, 7.4; Table A1) has specific litter rates for ECM mantle and mycelia (L M , L MYC ).This setup is recommended when investigating different biomass ratios between mycelia and mantle and their effects on overall litter production.Irrespective of the approach used for litter production, ECM have the capability to retain a defined amount of N during senescence (Eq.7.2, 7.5; Table A1b: nret fungi ).
In this study, the simple approach is applied.
2.1.5Plant mycorrhization degree, plant N uptake, and ECM fungal N transfer to plant According to field investigations, the mycorrhization degree can vary considerably between species.For spruce (Picea abies), typical mycorrhization degrees of over 90 % have been reported (Fransson et al., 2010;Leuschner, 2004).The impact of the ECM mantle on fine-root nutrient uptake has been controversially discussed, but the majority of studies indicate that the root is isolated from the soil solution, i.e., the nutrient uptake is hampered so that the plant is highly dependent on ECM supplies (Taylor and Alexander, 2005).Therefore, the mycorrhization degree is of major importance when plant-ECM-soil N exchange and plant nutrition are of interest.In the explicit Coup-MYCOFON model, mycorrhization degree is calculated as the ratio between ECM C pool and the defined optimum ECM C pool, divided by the defined optimum mycorrhization degree (Eq.9, Table A1b).It should be noted that the optimum mycorrhization degree needs to be defined with care as there is often a discrepancy between the applied root diameter in experimental studies and models: in experiments, mycorrhization degrees usually refer to fine roots ≤ 1 mm, whereas models often consider fine roots as roots with a diameter of up to 2 mm.The mycorrhizal mantle has an impact on the mineral plant N uptake.This is because plant ammonium and nitrate uptake is largely driven by the plant N demand (Eq.4.1, 4.2; Table A1) but also regulated by the N availability function (Eqs. 15,16,17;Table A1: f (n avail ), f (n mhumavail )) based on the assumption that only a certain fraction of soil ammonium and nitrate is available for plant uptake.The ECM fungal mantle reduces this availability in such a way that reduction is highest at maximum biomass.In a balanced symbiosis, the fungus provides nutrients to the plant in exchange for the plant's C supply.In the Coup-MYCOFON model, the amount of ECM fungal N transferred to the plant is determined by either the plant N demand or, if the plant N demands exceed the ECM fungal capacity, the available fungal N (Eq.10.1, 10.2; Table A1).This is the amount of "excess" N that is available after the ECM have fulfilled their defined minimum demand as calculated by the fungal C / N ratio (Eq.10.2, Table A1).This relation is again based on the assumption that the ECM fungi will only supply the plant with N as long as its own demand is fulfilled (Nehls et al., 2008).

Ectomycorrhizal fungal N uptake
In the Coup-MYCOFON model, ECM can take up both mineral and organic N.For both N forms, the potential ECM uptake is first defined.This is determined by the size of ECM C pool, the fraction of ECM C that is capable of N uptake (the mycelia, FRAC MYC ), and an uptake rate (NO 3 RATE , NH 4 RATE , NORG RATE (Eq. 11.1,11.3,11.4,11.6;Table A1b).This function is based on the assumption that only the ECM fungal mycelia can take up N. Values for NO 3 RATE , NH 4 RATE , and NORG RATE are derived from published values but with wide ranges (Table 2).The actual N uptake is dependent on the available soil N as well as the ECM N demand (Eq.11.2, Table A1).The N availability function f (n avfungi ) determines the fraction of soil N that is available for ECM fungal uptake and is controlled by the parameters NUPT ORGFRACMAX (the fraction of organic N available for uptake) and NUPT FRACMAX (the fraction of mineral N available for uptake).N availability for ECM corresponds to the plant-available N (Eq.16, Table A1), but as ECM are more efficient in the uptake of nutrients, the availability is enhanced for both mineral and organic N (Eq.17.1, 17.2, 17.3; Table A1).To prevent the ECM N demand being covered by only one N form, the parameters r NO 3 , r NH 4 , r LIT , and r HUM are included, corresponding to the ratio of nitrate to ammonium in total available soil N (litter and humus).If the potential N uptake exceeds the available soil N, the actual uptake corresponds to the available N (Eq.11.2 and 11.5; Table A1b).

Three ECM modeling approaches
Three modeling approaches of different complexity were applied in this study.The basic nonlim approach was conducted to test if plant N uptake can be described as proportional to the C demand of the plants of the respective sites.In this case, the plant N uptake is not regulated by the actual soil N availability, and N is used from a virtual source potentially exceeding the soil N availability, thus as an open N cycle.The ECM implicit approach simulates plant uptake of organic N, which is assumed to be via ECM, i.e., ECM are considered implicitly as being responsible for N uptake but www.geosci-model-dev.net/11/725/2018/Geosci.Model Dev., 11, 725-751, 2018   (1995), and Smith and Read (2008).b Set equally to trees according to Thornley and Cannell (2000).c Staddon et al. (2003) and Ekblad et al. (2013).d Högberg and Högberg (2002) and Wallander and Nilsson (2003).e Estimated.
are not physically represented in the model.The rate of the organic N uptake is determined by the plant N demand and restricted by the availability of organic N in the soil humus pools (Eq.4.4, 4.5; Table A1).Plants can also additionally take up ammonium and nitrate (Eq.4.1, 4.2; Table A1).In the ECM explicit approach, ECM are fully physically considered as described above.ECM growth interacts dynamically with plant growth and responds to changes in soil N availability and soil temperature.ECM can take up both mineral and organic N forms.

Simulated regions and database
Simulations were performed for four forests sites -Lycksele, Mora, Nässjö, and Ljungbyhed -situated along a climate and N deposition gradient in Sweden (Fig. 2).Climate and site information is given in Table 3 and the climate data were taken from the Swedish Meteorological and Hydrological Institute (SMHI).Data on forest standing stock volumes and forest management were derived from the database and practical guidelines of the Swedish Forest Agency ( 2005) and applied as previously described by Svensson et al. (2008a).Soil C content as well as soil C / N ratio, previously determined by Berggren Kleja et al. ( 2008) and Olsson et al. (2007), were used to describe soil properties in the initial model setup.
For all simulated sites and modeling approaches, the development of managed Norway spruce forests was simulated in daily steps over a 100-year period from a newly established forest to a closed mature forest.Available daily climate data  were repeated four times in order to cover the entire period, and thus climatic warming effects are not considered here.A minimum of specific regional data including the meteorological data, N deposition, and soil data were used as input values (Table 3).Otherwise, model parameters were kept identical between modeling approaches in order to evaluate the general model applicability.An overview of the parameter values is shown in Table A1d in the Appendix.For a more detailed site description and CoupModel setup, see Svensson et al. (2008a).

Observational constraints
We performed a Bayesian calibration for all modeling approaches and sites.In this study, we emphasize the models' predictability in precisely describing the long-term plant and soil developments, also aiming at maximized model flexibility.This allows us to compare the different model approaches in terms of explaining the measured data and also to investigate distributions and uncertainty of key parameters.The previous modeling study by Svensson et al. (2008a) demonstrated that the changes of soil C at these sites were rather small over a 100-year period while the soil C / N ratio showed large variabilities with different N supply assumptions.Therefore, in this study the measured C / N ratio of soil organic matter and standing stock biomass were used as observational constraints.The measured error (also called relative uncertainty in Table 4) for both the soil C / N ratio and the standing stock biomass were difficult to assume due to lack of information.An uncertainty estimate of 30 % was generally recommended under such conditions (van Oijen et al., 2005).In order to reduce the weight of values close to zero on behalf of large peaks, a minimum measured error that is 10 % of the measured value was defined in this study (Klemedtsson et al., 2008).This is also because our inten- Lycksele 64

Model parameters chosen for calibration
The different ECM modeling approaches were calibrated for a comprehensive set of key parameters, which are chosen according to their function as regulating factors of the C and N fluxes in the plant-soil-mycorrhiza continuum (Table 2).In the nonlim approach, the constant N supply parameter Con-stantNsupply for the spruce tree was a calibration parameter.
In the implicit approach, the fraction of organic N available for plant uptake (NUPT ORGFRACMAX ) was included in the calibration based on Svensson et al. (2008a).For the ECM explicit approach, all ECM fungal parameters in MYCOFON including ECM growth (C and N assimilation and uptake, C and N losses), overall N uptake and plant N supply, respiration, and littering were calibrated.For all three approaches, the humus decomposition rate (K H ), the C / N ratio of microbes (CN mic ) regulating soil mineralization and thus soil N availability, and the fraction of plant C assimilates allocated to the rooting zone (F ROOT ) regulating ECM fungal growth were additionally calibrated.
H.He et al.: Simulating ectomycorrhiza in boreal forests

Bayesian calibration of models
The prior distributions of the parameters were chosen as uniform and uncorrelated, with wide ranges of possible values (Table 2).Bayesian calibration combines the prior information about the parameters and the observational constraints on model outputs to obtain a revised probability distribution called posterior distribution (Yeluripati et al., 2009).The posterior probability of any parameter vector is proportional to the product of its prior probability and its corresponding likelihood (Eq.1).The likelihood function that determines acceptance of the parameter sets as the posterior distributions is based on the assumption that the model errors (the differences between simulated and observed values) are normally distributed and uncorrelated (van Oijen et al., 2005).Furthermore, model errors are assumed to be additive so that the log-likelihood function reads where y i is observed values, f (ω i • θ i ) is simulated values for a given model input vector ω i and parameter set θ i , σ i is standard deviation across the measured replicates, and n is number of variables measured.
To construct the posterior parameter distribution, many sets of parameter θ were sampled.In this study, candidate parameter sets were generated using the Metropolis-Hastings random-walk Markov chain Monte Carlo (MCMC) algorithm (van Oijen et al., 2005;Vrugt, 2016).Briefly, a parameter ensemble of "walkers" move around randomly and the integrand value at each step was calculated.A few number of tentative steps may further be made to find a parameter space with high contribution to the integral.MCMC thus increases the sampling efficiency by using information about the shape of the likelihood function to preferentially sample in regions where the posterior probability is high (Rubinstein and Kroese, 2016).For each simulation, the model's likelihood was evaluated for a certain parameter set.After each run, a new parameter set was generated by adding a vector of random numbers ε to the previous parameter vector: where θ i is the previous parameter vector, θ i+1 the new parameter vector, and ε random numbers.
The normally distributed random numbers ε have a mean of zero and a step length of 0.05, i.e., 5 % of the prior parameter range as proposed by van Oijen et al. (2005).After a sufficiently long iteration (referred to as the "burn-in" period), the Markov chain reaches a stationary distribution that converges to the joint parameter posterior (Ricciuto et al., 2008).Van Oijen et al. (2005) recommended chain lengths in the order of 10 4 -10 5 for modeling forest ecosystems with many observational constraints.In this trial study, we performed 10 4 runs for each ECM modeling approach and site.This is because a length of 10 4 model runs with a burn-in length of around 10 3 runs results in numerically stable results for our current considered problem.The step sizes used in this study result in acceptance rates between 25 and 50 % (Table 4), which is also generally the most efficient range for the MCMC algorithm (Harmon and Challenor, 1997).

Results
3.1 Comparison of the three modeling approaches 3.1.1General ability to reproduce tree growth and soil C / N The three modeling approaches show different accuracies in reproducing current plant growth and soil C / N ratio after calibration (Table 4b).The posterior model in the implicit and explicit approaches shows better performance of simulating soil C and N, as indicated by the soil C / N ratio, than the nonlim approach.The latter tends to simulate a lower soil C / N ratio, indicated by the negative mean errors (ME, difference between the simulated and measured values) in the posterior model (Table 4b).The ME by the nonlim approach is also 2 to 5 times higher than that when using the implicit or explicit approach (Table 4b).The nonlim approach tends to overestimate plant growth as the posterior mean of ME for plant C is always positive, while the implicit and explicit approaches tend to show an underestimation (Table 4b).
All posterior models underestimate soil C / N for the northern sites, which are generally more N limited, but gradually switch to overestimation at the southern sites.The model with the nonlim approach simulates better plant growth for the southernmost site, Ljungbyhed, than the other sites.Further, modeled plant growth at Ljungbyhed is overestimated by the implicit approach but underestimated when the explicit approach is used (Table 4b).The acceptance of model runs in posterior is higher for the nonlim (25 to 48 %) and implicit approaches (42 to 50 %), followed by the explicit approach (30 to 33 %).No major differences are found for the summed log likelihood for both calibration variables (Table 4b).

Ecosystem N and C fluxes and comparison to measured data
Modeled major ecosystem N fluxes in the posterior are shown in Fig. 3.The modeled N litterfall, uptake, and leaching fluxes differ significantly from one modeling approach to another, where the nonlim approach always gives the highest fluxes.The explicit and implicit approaches show similar modeled N fluxes for the northernmost site, Lycksele.However, the differences between these two approaches become larger when moving south where higher fluxes are simulated by the explicit approach (Fig. 3).For instance, modeled N litter production in the explicit approach increases by 1 to 30 % compared to the implicit approach, but N losses due to uptake and leaching also increase by 10 to 50 % for Lycksele and Ljungbyhed, respectively (Fig. 3).The modeled N pool sizes for these two sites also differ, where the explicit approach shows a larger mineral N in the soil and a smaller organic N pool compared to the implicit approach (Fig. 4).
Figure 5 shows the modeled major ecosystem C fluxes and comparison with previous results by Svensson et al. (2008a) and measured data from three other Swedish sites (Flakaliden, Knottåsen, and Åsa ; Fig. 2) at comparable latitudes and on comparable soils by Lindroth et al. (2008).The simulated plant gross primary production (GPP) using three approaches all show an increasing trend from the northern sites to the southern sites due to a more favorable climates and N availability for spruce forest growth.For the studied four sites, the nonlim approach simulates the highest GPP followed by the explicit approach and lastly the implicit approach.The variation in modeled GPP between the explicit and implicit approach ranges from 12 % at northernmost Lycksele site to 7 % at the southernmost Ljungbyhed site (Fig. 5).Simulated GPP in this study is generally higher than that by Svensson et al. (2008a) but comparable with the measured data from Lindroth et al. (2008).It should be noted that the GPP at the southern site, Åsa, was only measured for 1 year and thus can show large uncertainties due to annual variations.Modeled ecosystem respiration generally follows the pattern of GPP.The net ecosystem exchange (NEE) values predicted by the three approaches all show an overall atmospheric C uptake for all the sites where the explicit approach seems to have a higher uptake strength than the others (Fig. 5).Current estimates of NEE are again within the measured range by Lindroth et al. (2008), although a small net release of C was measured at Knottåsen, likely caused by the abnormal high temperature during those measured years.In addition, explicitly including ECM also increases the soil respiration for the four sites except the northernmost Lycksele site.The simulated ranges are however somehow smaller than those by Svensson et al. (2008a).
The nonlim approach generally shows much higher uncertainties in the modeled N fluxes than either the implicit or explicit approaches.The nonlim approach simulated soil N sequestration of up to 2 g N m −2 yr −1 for all the sites, but values much lower or close to zero were found when using the other two modeling approaches (Fig. 5).The simulated soil C balance by the nonlim approach also contrasts with that of soil N, in which the soil sequesters C at the northernmost site, Lycksele, but loses C at a rate of 6 to 17 g C m −2 yr −1 for the other three sites (Fig. 5).Therefore, soil C and N are not in steady state and are decoupled in the nonlim approach over the simulated 100-year period.The implicit and explicit approaches, however, show a strong coupling between soil C and N (Fig. 5).That is, for the implicit approach, Lycksele and Mora soils lose 6 and 5 g C m −2 yr −1 , respectively, while Nässjö and Ljungbyhed soils gain 3 and 13 g C m −2 yr −1 , respectively.Similarly, Lycksele and Mora lose N by 0.2 and 0.1 g N m −2 yr −1 , while Nässjö and Ljungbyhed gain N by 0.3 and 0.6 g N m −2 yr −1 .For the explicit approach, soil C and N losses at the two northern sites are slightly higher than those in the implicit approach.The respective net change in the soil C and N pools of the implicit approach corresponds well to the results by Svensson et al. (2008a), who also sug-Figure 6. Posterior parameter distributions for N uptake parameters: constant N supply rate in the nonlim approach (grey) and organic N uptake capacity in the implicit (white) and explicit (black) ECM model approaches.Distributions are presented as box plots over the prior range of variation (corresponding to the range on the x axis).Box plots depict the median (bold line), the 25th and 75th percentiles (bars), and the 10th and 90th percentiles (whiskers).
gest a small loss of soil C in the north, whereas soils in the south gain C.However, when the explicit approach is used, the soils in the south are also predicted to lose C and N. Lindroth et al. ( 2008) found a similar trend in the soil net C change as simulated by the explicit approach here, but with a higher loss rate between 24 and 133 g C m −2 yr −1 (Fig. 5).Overall, our results show that accounting for ECM in boreal forest ecosystems can have a considerable impact on the predicted C and N dynamics for both the plants and soil.

Posterior distributions of common parameters
The posterior distributions differ from the prior uniform distributions for all modeling approaches and parameters, reflecting the efficiency of Bayesian calibration (Figs. 6 and 7).The posterior constantNsupply parameter in the nonlim approach shows the lowest values at Lycksele and the highest at Ljungbyhed.This means a higher N supply is necessary at the southern sites to explain the observed tree biomass and soil C / N ratio.No significant differences in parameter values -microbial C / N ratio (CN MIC ), humus decomposition coefficient (K H ), and the fraction of C allocated to roots (F ROOT ) in the nonlim approach are found for the different sites (data not shown).The organic N uptake parameter in the implicit and explicit approaches (NUPT ORGFRACMAX ) shows an opposite pattern with the highest values for Lycksele and lowest for Ljungbyhed and larger parameter uncertainties are found for the explicit approach (Fig. 6).Parameter values for the northern sites also have a much wider range compared with the southern sites (Fig. 6), which also explains the larger simulated ME of soil C / N at the northern sites (Table 4).Both approaches demonstrate that the plant and soil conditions at the northern sites could not be simulated without an enhanced uptake of organic N.
When the implicit approach is used, the posterior humus decomposition coefficient K H shows higher values for the northern sites and decreases along the studied transect, demonstrating a modeled enhancement of organic matter decomposition and soil mineralization for northern sites (Fig. 7).A less-clear tendency towards higher values at the southern sites is identified for the fraction of C allocated to roots, the F ROOT parameter.The microbial C / N ratio (CN MIC ) parameter for both implicit and explicit approaches show similar posterior distributions for the three northern sites.However, much lower values are obtained for the southernmost Ljungbyhed site (Fig. 7), reflecting a more N-rich soil environment.Overall, parameters are less constrained and only minor differences among sites are found when the explicit approach is used (Fig. 7).

ECM fungal-specific parameters
The posterior distributions of all ECM fungal-specific parameters are constrained to log-normal or normal distributions (data not shown).The mean values of N uptake parameters (NORG RATE , NH 4 RATE , NO 3 RATE ) show a decreasing trend from the northern to southern sites (Fig. 8).This again means an enhanced ECM fungal N uptake is necessary to explain the observed soil and plant data at the more Nlimited northern sites.Similarly, lower values for the northern regions and higher values for the southern regions are also found for the minimum ECM fungal C / N ratio parameter (CN FMIN ).The optimum ratio between ECM and root C content, FRAC OPT , tends to be higher at the northern sites and lower at the southern sites, also implying a higher modeled ECM biomass at the northern sites (Fig. 8).MIN SUPL , the minimum supply of N from ECM to the host plant parameter, does not show a clear trend.Further, differences in the other ECM parameters for the four sites are minor (Fig. 8).

Correlation between parameters
An overview of correlations for all posterior model parameters can be found in Tables A2, A3, and A4.Key parameter sets showing correlation with each other (defined here as a Pearson correlation coefficient r ≥ 0.3 or ≤ −0.3) are shown in Fig. 9.When the implicit approach is used, a significant positive correlation is obtained between the humus decomposition rate, K H , and the fraction of C allocated to the rooting zone, F ROOT .The organic N uptake parameter, NUPT ORGFRACMAX , and microbial C / N ratio, CN MIC , are significantly negatively correlated, except for a weak correlation for Ljungbyhed (Fig. 9).A weak correlation between NUPT ORGFRACMAX and F ROOT is also found for the Nässjö site (Table A2).For the explicit approach, the correlation coefficients between K H and F ROOT are decreased, and there is also a weaker correlation between NUPT ORGFRACMAX and CN MIC for all sites compared to the implicit approach (Fig. 9).No clear correlation between common and ECM fungal parameters is obtained.Further, a negative correlation occurred among microbial C / N ratio, CN MIC , and the fungal N uptake rates (Norg RATE , NH 4 RATE , NO 3 RATE ), but only for the northern sites Lycksele and Mora (Table A4).A moderate correlation is found for K H and the fungal litter rate, L, for Ljungbyhed.Among fungal parameters, the N uptake rates moderately correlate to the litter production rate, L, at the northern sites, but correlations at Nässjö and Ljungbyhed are either weak or nonexistent (Table A4).

Discussion
Our new version of the CoupModel provides a detailed predictive model framework to explicitly account for ECM in the plant-soil-ECM continuum.Model comparison to two simpler ECM modeling approaches shows large variations in N dynamic simulations and that ECM and organic N uptake have to be included in ecosystem models to be able to describe the long-term plant and soil C and N development.
Our results confirm that ECM have a substantial effect on soil C and N storage and can also impact forest plant growth.But more importantly, including them in ecosystem models is both important and feasible.

Comparison of the three ECM modeling approaches
The nonlim approach in this study shows an overestimation of plant growth and also larger biases in soil N than the implicit and explicit approaches even after calibration (Table 4).Soil N is expected to reach a steady state over a period of 100 years (Svensson et al., 2008a).Therefore, the nonlim approach largely overestimates soil N sequestration, which can be attributed to the assumed "virtual" constant N uptake from the unlimited source.According to our model predictions, this virtual N fraction accounts for 20 to 30 % of the total plant N uptake.A previous CoupModel application by Wu et al. (2012) demonstrated that the nonlim approach could possibly describe short-term C and water dynamics for a Finnish forest site.The same nonlim approach was also used in Franklin et al. (2014) to simulate Swedish forest biomass growth and its competition with ECM.This seems to suggest that plant growth and the C cycle can be simulated reasonably with the nonlim approach, although a slight trend of overestimation is exhibited.However, our modeling exercise further indicates that in this simplified approach soil C and N are uncoupled (Fig. 5), and therefore this approach is not recommended for future long-term soil C and N predictions.This is also reflected in the posterior model parameter distributions in which the constantNSupply rate parameter shows primary control on the modeled plant growth and soil conditions.Other parameters have minor or no importance for the model results, reflecting an oversimplified model structure of N. Thus, the following discussion focuses on the other two modeling approaches.Moore et al. (2015) demonstrated that accounting for ECM in ecosystem models would substantially affect soil C storage, and that the impact is largely dependent on plant growth.Our study additionally shows that ECM representation in ecosystem models could further feed back into the predicted plant growth through N. When ECM are implicitly included, the model simulates a plant biomass 48 g C m −2 (average of four sites, ± SD: 86) lower compared to the measured data.When they are explicitly included, the difference becomes even larger, 185 (± 35) g C m −2 (Table 4).Including ECM explicitly in the model therefore results in decreased plant growth.This somehow differs from the general assumption that growth should be higher in mycorrhized plants, i.e., boreal forest trees, due to optimized nutrient supply (Pritsch et al., 2004;Finlay et al., 2008, see also review by Smith and Read, 2008).This discrepancy can possibly be due to (1) an enhanced root litterfall due to a higher turnover of ECM mycelia.Simulated litter production is 50 to 110 g C m −2 yr −1 higher with the explicit approach compared to the implicit approach.This could be explained by conceptually considering the ECM implicitly in the roots where the litterfall rate of roots is about 3 times lower than that of ECM (calibrated litter rate of ECM is 0.0075 d −1 , Fig. 8, whereas the litter rate of roots is 0.0027 d −1 , Tawww.geosci-model-dev.net/11/725/2018/Geosci.Model Dev., 11, 725-751, 2018 ble A1d).These two approaches thus show large differences in simulating litter production.The discrepancy could also be due to (2) an enhanced N immobilization in ECM under Nlimited conditions based on the assumption that ECM retain more N in their own biomass in response to plant allocation of newly assimilated C (Nehls et al., 2008).The increasing trend towards the northern sites shown by the constrained optimum ECM C allocation fraction parameter (Fig. 8) also indicates a higher proportional C investment by the forest plants in ECM in northern, N-limited conditions.The resulting ECM-plant competition for N could then potentially result in decreased plant N uptake and thus plant growth (Näsholm et al., 2013).Finally, the discrepancy could be due to (3) biases in simulating ECM N uptake due to model or parameter uncertainties caused by high variability among ECM species and the scarcity of direct measurements in the field (Smith and Read, 2008;Clemmensen et al., 2013).The current explicit approach implements many biotic interactions and internal feedbacks within the plant-soil-ECM continuum.However, increasing the number of processes and interactions in an already complex ecosystem model will not nec-essarily generate more reliable model predictions; as shown here, the parameters in the explicit approach have a larger uncertainty range and show high correlations with other parameters even after calibration.This is also shown by the smaller accepted ratio in the calibration (Table 4), which can be explained by model complexity, i.e., as more parameters are included for calibration, accepted combinations of parameter sets become less likely.It should also be noted that the explicit and implicit approaches show considerable difference in estimating soil respiration.Compared to the implicit approach, the explicit approach simulates a 15 % higher soil respiration for the northernmost site and 40 % for the southernmost site.The measured soil respiration at Flakaliden is 400 to 590 g C m −2 yr −1 (Coucheney et al., 2013) and 460 to 520 g C m −2 yr −1 at Åsa (Von Arnold et al., 2005) and these data generally align better with the results modeled by the explicit approach (Fig. 5).The estimated higher soil respiration is partly due to the higher litter production and consequently soil respiration in the explicit approach but also due to a higher decomposition of the old organic matter (humus) Geosci.Model Dev., 11, 725-751, 2018 www.geosci-model-dev.net/11/725/2018/ as shown by the constrained higher humus decomposition coefficient, K H , in the explicit approach (Fig. 7).This corresponds well with findings from field measurements and recent modeling studies that ECM are able to degrade complex N polymers in humus layers, thus enhancing soil N transformation under low-N conditions (Hartley et al., 2012;Moore et al., 2015;Lindahl and Tunlid, 2015;Parker et al., 2015;Baskaran et al., 2016).The modeled higher soil respiration further explains the minor losses of soil C and N at the southern sites and also a higher mineral N pool and thus higher N leaching in the explicit approach (Figs. 3 and 4).

Constrained parameters
Our constrained parameters generally indicate a shift in the role of ECM from northern to southern sites with a corresponding shift in both climate and soil conditions (Figs. 6,7,and 8).The ECM N uptake parameters show a decreasing trend with increasing soil N availability in the explicit approach.This is consistent with observations that at the northern N-limited sites, organic N uptake by ECM is highly important for plant growth, becoming less important as N availability increases southwards (e.g., Hyvönen et al., 2008;Näsholm et al., 2013).Shown by the explicit approach, the mycorrhization degree of tree roots at Lycksele and Mora (> 90 %) is much higher than that of Ljungbyhed (15 %); thus the majority of modeled N uptake is through fungal mycelia at northern sites.A similar trend is also found for the organic N uptake parameter in the implicit approach, but with a larger site-to-site difference, thus indicating a stronger response to soil N conditions (Fig. 6).This is expected as more detailed ECM processes in the explicit approach should result in more internal interactions and feedbacks and thus more resilience to the change of environmental conditions.Most ECM fungal parameters in the explicit approach are not -or only weakly -dependent on the differing environmental conditions along the modeled transect, except for the N uptake parameters (NORG RATE ) and ECM fungal minimum C / N ratio (CN FMIN ), which show different mean values (Fig. 8).As such, these parameters need to be calibrated carefully when further applying the model to other sites with different soil nutrient levels or climate conditions.
Most of our constrained parameter distributions are not sharply peaked, but instead rather flat, and few parameters show high covariance .This is, however, not a unique characteristic of the CoupModel or currently used data constraints and, indeed, has been previously demonstrated in numerous studies with ecosystem models of similar complexity (e.g., He et al., 2016;Klemedtsson et al., 2008;Wang et al., 2001).On the one hand, this generally reflects the equifinality of models (Beven, 2006), where multiple parameter sets can lead to equally good representations of the system.One the other hand, it also indicates poor identifiability of the calibrated parameters with respect to the available information for parameterization.Here, we again show that given the same data constraints, the parameter identifiability decreases with increasing model complexity (Sierra et al., 2015).In our study, the correlation between the humus decomposition coefficient, K H , and the fraction of C that is allocated to the rooting zone, F ROOT , is smaller when ECM are modeled explicitly rather than implicitly (Fig. 9).However, the correlations between the ECM fungal litter rate and ECM fungal N uptake rates, and that between fungal N uptake rates (NORG RATE ) and the microbial C / N ratio (CN MIC ) (Fig. 9) further indicate these ECM fungal parameters in the more complex explicit model cannot be identified well without adding a new dataset as additional constraint.One of the major challenges of explicitly including ECM in ecosystem models is the still sparse information about ECM, e.g., unknown turnover of ECM mycelia (Ekblad et al., 2013).Previously reported turnover rates of newly formed mycelia vary from days to weeks, even up to 10 years (Staddon et al., 2003;Wallander et al., 2004), mostly due to the high variability in ECM species and structures (see review by Ekblad et al., 2013).Additionally, root turnover rates can also vary considerably between species, soils, and climate zones (Brunner et al., 2012).Thus far, very few studies have reported parameterization of C and N cycling for ECM in boreal forests.The present model calibration thus provides a key set of ECM parameters that can be further tested by field observations, and more importantly, can act as a prior for future ECM modeling studies.

Conclusions
The key components and features of the Coup-MYCOFON model have been described.The new version of Coup-Model explicitly accounts for the links and feedback among ECM, soil, and plant.The comparison of three modeling approaches that differ in complexity demonstrates that the simple nonlim approach cannot describe the soil C / N ratio and also overestimates forest growth.When including ECM either implicitly or explicitly, both models deliver accurate long-term quantitative predictions on forest C and N cycling with simultaneous considerations of the impact of ECM on ecosystem dynamics.However, the implicit approach shows a much lower litter production and soil respiration than the explicit approach, and both approaches slightly underestimate forest growth.The ECM explicit Coup-MYCOFON model provides a more detailed description of internal ecosystem flux and feedback of C and N. The constrained ECM parameter distributions presented in this study can be used as guidelines for future model applications.Overall, our model implementation and comparison suggest that ecosystem models need to incorporate ECM into their model structure for a better prediction of ecosystem C and N dynamics.(Meyer et al., 2012, and references therein).
No. Equation × r s ε L : coefficient for radiation use efficiency; f (T l ), f (CN l ), f (E ta /E tp ): response functions to leaf temperature, leaf CN, and air moisture (see Table A1c); r s : global radiation absorbed by canopy.

Plant maintenance respiration (g C m
Plant nitrate and ammonium uptake (g N m −2 d −1 ) (only shown for nitrate, equivalent for ammonium) 4.1 and where 4.3 dem N plant = c a→i −c i→atm CN iMIN f (n NO 3 avail ): fraction of soil NO 3 available for plant uptake (see response functions Table A1d); n NO 3 soil : soil NO 3 -N content (g N m −2 ); dem N plant : plant N demand (g N m −2 d −1 ); r NO 3 : fraction of soil NO 3 -N in total mineral soil N; c a→i : plant C gain (g C m −2 d −1 ); c i→atm : respiration of respective plant compartment i (g C m −2 d −1 ); CN iMIN : defined minimum C : N ratio of each plant compartment i.
Plant organic N uptake (g N m −2 d −1 ) from the humus layer 4.4 ECM fungal C and N litter production (c fungi→lit : g ; n fungi : ECM fungal N content (g N m −2 ); L: litter rate, nret fungi : ECM fungal N, which is retained in fungal tissue; N RET : fraction of N retained in fungal tissue from senescence.
If ECM fungal growth is detailed 7.4 FRAC MYC : fraction of mycorrhizal mycelia in total fungal biomass; L MYC : litter rate of mycorrhizal mycelia; L M : litter rate of ECM fungal mantle tissue.

No. Equation
Uptake and transfer processes of ECM and plant n NO 3 pot→fungi : potential ECM nitrate uptake (g N m −2 d −1 ); r N : fraction of ammonium N and total mineral N in the soil;f (n demfungi ): N uptake response to N demand; n NO 3 soil : soil nitrate content (g N m −2 ); f (n avfungi ): N uptake response to soil availability; NO 3 RATE : nitrate-specific uptake rate (g N m −2 d −1 ); c fungi : ECM fungal biomass (g C m −2 ); FRAC MYC : fraction of mycorrhizal mycelia in total ECM biomass.
ECM fungal organic N uptake from litter and humus (given for litter, equivalent for humus with humus-specific parameter) 11.4 n lit→fungi = n litpot→fungi × r lit × f (n demfungi ) if n litpot→fungi × r lit < n litsoil × f (n litavfungi ) × r lit 11.5 n lit→fungi = n litsoil × f (n litavfungi ) × r lit if n litpot→fungi × r lit > n litsoil × f (n litavfungi ) × r lit 11.6 n litpot→fungi = LIT RATE × c fungi × FRAC MYC where n litpot→fungi is potential ECM organic N uptake from litter (g N m −2 d −1 ), r lit is fraction of litter N in total organic N in the soil, f (n demfungi ) is N uptake response to N demand, n litsoil is soil litter content (g N m −2 ), NLIT RATE is litter-specific uptake rate (g N g C −1 d −1 ), c fungi is ECM fungal biomass (g C m −2 ), and FRAC MYC is fraction of mycorrhizal mycelia in total ECM biomass.

No. Equation
Plant mineral N uptake response to N availability and ECM fungal mantle 15 f (n minavail ) = NUPT FRACMAX × e (−FM×m)   where NUPT FRACMAX is the coefficient describing fraction of soil N available, and FM is uptake reduction due to ECM fungal mantle.
Plant organic N uptake response to N availability and ECM fungal mantle (given for litter, equivalent for humus) 16 f (n litavail ) = NUPT ORGFRACMAX × e (−FM×m)   where NUPT FRACMAX is the respective uptake coefficient for N from humus (included in calibration), and FM the uptake reduction due to ECM fungal mantle.
ECM N uptake response to N availability where NAVAIL COEF is a coefficient and N minsoil is the total soil content of ammonium and nitrate (g N m −2 ).where min N Plant is the defined minimum ECM fungal N supply in plant N uptake, n fungi→plant is actual ECM N supply to plant (g N m −2 d −1 ), and n soil→plant is total plant N uptake from the mineral and organic fraction (g N m −2 d −1 ). (

Figure 1 .
Figure 1.A simplified overview of C and N fluxes among plants, mycorrhiza fungi, and the soil in the Coup-MYCOFON model.Light blue indicates the newly implemented MYCOFON model.

Figure 2 .
Figure 2. Location of the four study sites in Sweden modified from Svensson et al. (2008a).Filled cycles represent the four studied sites.Open circles are the measured sites reported in Lindroth et al. (2008) used for comparison.

Figure 3 .
Figure 3. Soil N fluxes for the nonlim (grey columns), implicit (white), and explicit (black) model approaches; same color scheme used for the other figures.Presented are the major N inputs (N deposition, total N litter production, added to the soil litter pool by fresh litter), and outputs (N uptake from the plant or ECM, N leaching).Error bars indicate the 90th percentile of accepted model runs (posterior).Units for N are g N m −2 yr −1 and g C m −2 yr −1 for C.

Figure 4 .
Figure 4. Average soil organic and mineral content in the implicit ECM model (upper graph) and explicit ECM model (lower graph) for the two sites Lycksele and Ljungbyhed.Box plots indicate the median (bold line), the 25th and 75th percentiles (bars), and the 10th and 90th percentiles (whiskers).

Figure 5 .
Figure 5. Simulated GPP, ecosystem respiration, NEE, soil respiration, change in soil C, and change in soil N for all four sites with the three ECM modeling approaches and also compared with modeled data by Svensson et al. (2008a) and measurements by Lindroth et al. (2008).

Figure 7 .
Figure 7. Posterior parameter distributions for common parameters using the implicit (a: white) and explicit (b: black) ECM approaches for four different sites from north to south.Distributions are presented as box plots over the prior range of variation (corresponding to the range on the x axis).Box plots depict the median (bold line), the 25th and 75th percentiles (bars), and the 10th and 90th percentiles (whiskers).The parameters shown are K H the humus decomposition coefficient, F Root the fraction of C assimilates distributed to the roots, ECM, and CN MIC the microbial C / N ratio.

Figure 8 .
Figure 8. Posterior parameter distributions of ECM fungal-specific parameters (from top left to bottom right): organic N uptake rate (NORG RATE ), ammonium uptake rate (NH 4 RATE ), respiration coefficient (K RM ), ECM fungal litter rate coefficient (the rate at which mycelia and mantle die and add to the soil litter pool, L), minimum ECM fungal C / N ratio (CN FMIN ), ECM minimum N supply to plant (MIN SUPL ), optimum ratio between ECM and root C content (FRAC OPT ), and N sensitivity coefficient (NAVAIL COEF ).Distributions are presented as box plots over the prior range of variation (corresponding to the range on the x axis).Box plots depict the median (bold line), the mean (black point), the 25th and 75th percentiles (bars), and the 10th and 90th percentiles (whiskers).

Figure 9 .
Figure 9. Correlation between model parameters, given as the Pearson correlation coefficient, for the implicit (white) and explicit ECM (black) approaches.(a) Correlation between humus decomposition coefficient (K H ) and the fraction of C assimilates (GPP) directed to ECM and roots (F ROOT ).(b) C / N of microbes (CN MIC ) and fraction of organic N available for uptake (NUPT ORGFRACMAX ).Correlation between ECM fungal parameters includes (c) humus decomposition coefficient (K H ) and ECM fungal litter rate (L).(d) ECM organic N uptake (NORG RATE ) and C / N of microbes (CN MIC ).

N
transfer from ECM to plant (g N m −2 d −1 ) 10.1 n fungi→plant = dem N plant if dem N plant ≤ n fungiavail n fungi→plant = n fungiavail if dem N plant > n fungiavail dem N plant : plant N demand, n fungiavail : fungal available N for transfer to plant (all g N m −2 d −1 ).10.2 n fungiavail = n fungi − c fungi CN FMAX c fungi : ECM biomass (g C m −2 ); CN FMAX : maximum C : N ratio of fungal tissue, which allows N transfer to plant.ECM fungal nitrate and ammonium uptake (given for nitrate, equivalent for ammonium with ammonium-specific parameter)11.1 n NO 3 →fungi = n NO 3 pot→fungi × r NO 3 × f (n demfungi ) if N NO 3 pot→fungi < n NO 3 soil × f (n avfungi ) 11.2 n NO 3 →fungi = n NO 3 soil × f (n avfungi ) if N NO 3 pot→fungi > n NO 3 soil × f (n avfungi ) 11.3 n NO 3 pot→fungi = NO 3 RATE × c fungi × FRAC MYC Photosynthetic response to leaf C / N ratio

Table 2 .
Maximum and minimum parameter values prior to Bayesian calibration for the nonlim, implicit, and explicit model approaches.(a) Common parameters (all three approaches).(b) Parameters of the nonlim approach.(c) ECM fungal parameters of the explicit approach.

Table 3 .
Climatic and soil data and initial settings of the four study soils applied in all model approaches.

Table 4 .
Prior values of variables used for model calibration and accepted relative uncertainty (a), and posterior model performance indicators (b): mean error (ME) between simulated and measured values, standard variation in ME (SD), and summed log likelihood of all accepted runs for simulated standing plant biomass (g C m −2 ) and soil C / N ratio after the 100-year simulation period.

Table A1 .
Model functions describing plant growth, ECM fungal growth, model parameters, and response functions of plant and ECM.Parameters are always shown with capital letters.(a) Description of plant model functions.(i is fine roots, coarse roots, stem, leaves, grain, mobile).(b) Functions describing processes related to ECM fungal growth and N exchange to plant.(c) Overview of response functions of plant and ECM fungal growth and N uptake.(d) Overview of model parameters; previous CoupModel parameters are mostly from Svensson et al. (2008a) and ECM parameters are from literature values

Table A1 .
Continued.((c frt × FRAc OPT ) − c fungi ) × f (n supply ) c a→root : C available for root and mycorrhiza growth (g C m −2 d −1 ); FRAC FMAX : maximum fraction of total root and mycorrhiza available C that is available for ECM; f (c fungiavail ): response function which relates ECM growth to N availability; c frt : total root C content (g C m −2 ); FRAC OPT : optimum ratio between root and ECM C content; c fungi : total ECM C content (g C m −2 ); f (n supply ): response function of fungal growth to the amount of N (both mineral and organic N) that is transferred from ECM to plant.fungi→atm = c mfungi→a + c gfungi→a where c mfungi→a is ECM fungal maintenance respiration and c gfungi→a is ECM fungal growth respiration (all in g

Table A2 .
Correlation between common model parameters for all simulated sites with the implicit and explicit approaches.Correlation is given as the Pearson correlation coefficient.