Nitrogen Recovery from Clear-Cut Forest Runoff Using Biochar: Adsorption–Desorption Dynamics Affected by Water Nitrogen Concentration

Forest regeneration operations increase the concentration of nitrogen (N) in watercourses especially outside the growing season when traditional biological water protection methods are inefficient. Biochar adsorption-based water treatment could be a solution for nutrient retention. We studied the total nitrogen (TN) and nitrate–nitrogen (NO3−–N) adsorption–desorption properties of spruce and birch biochar. The adsorption test was performed under four different initial concentrations of TN (1, 2, 3, and 4 mg L−1) using forest runoff water collected from ditch drains of boreal harvested peatland. The results showed that the TN adsorption amount increased linearly from the lowest to the highest concentration. The maximum adsorption capacity was 2.4 and 3.2 times greater in the highest concentration (4 mg L−1) compared to the lowest concentration (1 mg L−1) in spruce and birch biochar, respectively. The NO3−–N adsorption amount of birch biochar increased linearly from 0 to 0.15 mg NO3−–N g biochar−1 when the initial concentration of NO3−–N increased from 0.2 to 1.4 mg L−1. However, in spruce biochar, the initial concentration did not affect NO3−–N adsorption amount. The results indicate that concentration significantly affects the biochar’s capacity to adsorb N from water. The desorption test was performed by adding biochar extracted from the adsorption test into the forest runoff water with low TN concentration (0.2 or 0.35 mg L−1). The desorption results showed that desorption was negligibly small, and it was dependent on the TN concentration for birch biochar. Therefore, biochar can be a complementary method supporting water purification in peatland areas.

− -N adsorption amount of birch biochar increased linearly from 0 to 0.15 mg NO 3 − -N g biochar −1 when the initial concentration of NO 3 − -N increased from 0.2 to 1.4 mg L −1 . However, in spruce biochar, the initial concentration did not affect NO 3 − -N adsorption amount. The results indicate that concentration significantly affects the biochar's capacity to adsorb N from water. The desorption test was performed by adding biochar extracted from the adsorption test into the forest runoff water with low TN concentration (0.2 or 0.35 mg L −1 ). The desorption results showed that desorption was negligibly small, and it was dependent on the TN concentration for birch biochar. Therefore, biochar can be a complementary method supporting water purification in peatland areas.
Abstract Forest regeneration operations increase the concentration of nitrogen (N) in watercourses especially outside the growing season when traditional biological water protection methods are inefficient. Biochar adsorption-based water treatment could be a solution for nutrient retention. We studied the total nitrogen (TN) and nitrate-nitrogen (NO 3 − -N) adsorption-desorption properties of spruce and birch biochar. The adsorption test was performed under four different initial concentrations of TN (1, 2, 3, and 4 mg L −1 ) using forest runoff water collected from ditch drains of boreal harvested peatland. The results showed that the TN adsorption amount increased linearly from the lowest to the highest concentration. The maximum adsorption capacity was 2.4 and 3.2

Introduction
Leaching of N from drained peatland forests is remarkably higher (Nieminen et al., 2017) than from upland forests (Finér et al., 2021), and especially forest regeneration in peatlands imposes a threat to the quality of surface waters (Nieminen et al., 2005). Hytönen et al. (2020) estimated that 700,000 ha of peatland forests will reach regeneration age within the following 10 years in Finland, which is an increasing challenge for water protection. Large varieties of different water protection methods, such as sedimentation pits, sedimentation ponds, and peatland buffer areas, have been used to mitigate the nutrient export load to watercourses (Kalvīte et al., 2019;Marttila & Kløve, 2010), but all these are rather ineffective in reducing dissolved nutrient export loads particularly outside the growing season (Joensuu et al., 2002;Liljaniemi et al., 2003;Nieminen et al., 2005). Therefore, additional water purification methods are needed. Methods based on adsorption can be suitable for this purpose, as they are capable of eliminating both organic and inorganic pollutants from water also outside the growing season (Battas et al., 2019). Adsorption-based water purification methods have been shown to be effective in the treatment of municipal wastewaters and urban and agricultural runoff (Singh et al., 2018).
Biochar has been shown to be an effective adsorbent (Laird et al., 2010;Jin et al., 2016;Wang et al., 2015;Liu et al., 2020;Saarela et al., 2020). Biochar is formed when organic material is heated under limited oxygen concentration in the pyrolysis process. Biochar is stable, highly aromatic, and carbon-rich material (Beesley et al., 2011;Gaunt & Lehmann, 2008), and its adsorption capacity is based on high porosity, large specific surface area (Accardi-Dey & Gschwend, 2003;Kookana, 2010), and surface charges Liu et al., 2020;Mohan et al., 2014). The adsorption capacity of biochar varies with the properties of the raw materials, the pyrolysis temperature, and other manufacturing parameters (Liu et al., 2020).
Adsorption-based water protection has several challenges. A characteristic feature of nutrient export from forest areas is high water volume and concentration that is rather low compared to that in wastewater or in agricultural and urban runoff waters (Oliveira et al., 2018), and the nutrient concentration fluctuates seasonally (Mattsson et al., 2015). Therefore, the adsorbent must recover nutrients from a rather low concentration, high water volume, and in the presence of high dissolved organic matter concentration and should persist the adsorption under fluctuating nutrient concentrations. It has been reported that the adsorption amount increases with increasing element concentration in water (Ahmadvand et al., 2018;Saarela et al., 2020). This implies that the adsorption amount is improved when the runoff nutrient concentrations increase in the seasonal cycle. However, less attention has been paid to investigate the effect of decreasing concentration on the possible desorption of N compounds. Desorption occurs when adsorbed nutrients are released from the surface of biochar. Knowing the desorption of the recovered N is crucially important when biochar is applied as a water protection tool. Only a few studies before have examined the utility of biochar in peatland forest areas Saarela et al., 2020), and according to our knowledge, this is the first study to address the N adsorption-desorption processes in peatland runoff water. The current experiment was conducted to examine the effect of N concentration on the adsorption-desorption behavior of spruce and birch biochars in the water collected from a ditch draining a clear-cut peatland forest area. The main objectives are as follows: 1. To investigate the effect of TN concentrations in water on the adsorption amount 2. To determine adsorption kinetics under different TN concentrations in water 3. To quantify the desorption amount of recovered TN in birch and spruce biochar 2 Materials and Methods

Study Site and Water Collection
The study site is located in Heinävesi, in Eastern Finland (62°29′56.0″N, 28°35′56.0″E). The mean annual 1 3 temperature and precipitation in the area are 3.6 °C and 638 mm, respectively. The site was a drained fertile spruce fen that was clear-cut in August 2018. In the clear-cut, a total of 319 m 3 (124 m 3 ha −1 ) Norway spruce (Picea abies Karst.) and 239 m 3 (93 m 3 ha −1 ) silver birch (Betula pendula Roth.) were harvested. The total catchment area was 31.72 ha, and the area of 2.57 ha was clear-cut. The depth of the peat layer was 0.5-0.7 m, and it was underlain by clay. Water for the adsorption and desorption laboratory experiments was collected from a ditch that drains the clear-cut area. The water was collected outside the growing season (November 2019) when the TN concentration was 4.6 mg L −1 . Water was stored at + 4 °C before the experiments.

Biochar Properties
We used commercial biochar (Carbofex Ltd, Tampere, Finland) manufactured from Norway spruce (Picea abies L. Karst.) and silver birch (Betula pendula Roth) wood chips. The characteristics of biochars are shown in Table 1. Electrical conductivity (EC) and pH of biochars (on a 1:2.5 v:v biochar/water solution) were measured using WTW pH/cond 340i and WTW pH 340 sensors (WTW GmbH, Weilheim, Germany), respectively. N concentrations of biochars were analyzed with an elemental analyzer (Vario Max CN elemental analyzer, Elementar Analysensysteme GmbH, Germany). The specific surface area was measured using the N 2 adsorption method (Micromeritics Flowsorb II 2300, 1986).

Adsorption Experiment
To study the effect of N concentrations on the adsorption, runoff water was diluted to concentrations of 1, 2, 3, and 4 mg L −1 TN by adding Milli-Q water (Appendix, Fig. 7). Before the experiment, the water temperature was allowed to stabilize to + 21 °C. Five grams (dry mass) of spruce or birch biochar was added to 1 L of runoff water. Each concentration contained three biochar replicates and one blank control without biochar. The jars were covered with aluminum foil and placed on a platform shaker (New Brunswick™ Innova® 2300, Eppendorf Nordic A/S, Denmark). Glass jars were shaken at 110 rpm for 10 days, and 30 mL water sample was taken from each jar at the following points of time: 0, 2, 5, 25, 48, 72, 125, and 173 h from the beginning of the experiment. The water samples were filtered through a filtration assembly with Whatman GF/F glass microfiber filters (pore size 0.45 µm, GE Healthcare Bio-Sciences, Marlborough, MA, USA) and stored at + 4 °C until further analysis.
TN concentrations were determined with multi N/ C® 2100 (Analytik Jena AG, Jena, Germany). The concentrations of NO 3 − -N and NH 4 + -N were determined colorimetrically applying the methods by Miranda et al. (2001) and Fawcett and Scott (1960), respectively.
N adsorption amount (Q t i ) to biochar was computed from the observed concentration change in the water (Eq. 1): where A t p is the cumulative adsorption amount of , v t −1,t is water volume at a time interval from t − 1 to t (L), and m biochar is the dry mass of biochar in the jars.
The effect of the initial TN concentration on the TN adsorption for birch and spruce biochar was studied using the cumulative adsorption amount of TN at the end of the experiment (A 173h ) using linear mixedeffect models. Biochar type was used as fixed effects, replicates were considered random effects, and the 1 3 independent variable was the initial concentration. The model was formulated as follows: where A 173,pi represents the adsorption amount of substance p (TN, NO 3 − -N) to the biochar i (Bb, Sb, unit mg g −1 biochar) at time 173 (h), α p,i and β p,i are fixed parameters for substance p and biochar i, C ini,p is the initial concentration (mg L −1 ) for substance p. ε pij is a residual term for substance p and biochar i and replicant j with the expectation value of 0. The performance of the model was expressed as information criteria (AIC, BIC, -log likelihood, RMSE, and R 2 ).
Adsorption kinetics (adsorption rate and maximum adsorption capacity) were determined by fitting a commonly used integral form of pseudo-first order (Eq. 3) and pseudo-second order (Eq. 4) adsorption models (Largitte & Pasquier, 2016) to the data.
where A p is the cumulative adsorption of substance p (TN, and NO 3 − -N mg g −1 biochar) at time t (h). k ad1,p is the adsorption rate in pseudo-first order model (h −1 ) and k ad2,p in pseudo-second order model (mg g −1 h −1 ). Q max1,p and Q max1,p are the maximum adsorption capacity at equilibrium (mg g −1 biochar) for substance p.

Desorption Experiment
We studied the desorption of TN in a two-stage experiment consisting of adsorption and desorption phases (Appendix, Fig. 8). In the adsorption phase, we stirred 5 g of birch and spruce biochar with 1 L of runoff water in four different TN concentrations and three replicates. The runoff water with the original TN concentration of 4.6 mg L −1 was diluted with Milli-Q water to approximate concentrations of 1 mg L −1 , 2 mg L −1 , 3 mg L −1 , and 4 mg L −1 of TN. The duration of the adsorption phase was 72 h. At the beginning and at the end of the adsorption phase, the TN concentration of water was determined, and the amount of adsorbed TN (A TN ) to the biochar was calculated using Eq. 1. The biochar was filtered and divided into two samples with a mass of 2.5 g. For the desorption phase, the biochar samples were placed into glass bottles and stirred with the runoff water diluted to 0.2 mg L −1 and 0.35 mg L −1 using Milli-Q water in 250-mL total volume. The bottles were shaken at 110 rpm for 9 days; after which, water samples were taken and the TN concentration was analyzed. The desorption amount was calculated as follows: where D ci is TN desorption amount for biochar i (Bb, Sb) in concentration c (0.20 and 0.35 mg L −1 ), c end is solution TN concentration at the end of desorption phase (mg L −1 ), c ini is the solution TN concentration at the beginning of the desorption phase (mg L −1 ), v is the solution volume (L), and m biochar is the biochar mass in the bottle. The effect of the initial TN concentration on the TN desorption amount for birch and spruce biochar was examined by the cumulative desorption amount of TN at the end of the experiment (D 216h ) using linear mixed-effects models (Eq. 6). Biochar type, replicates, and the initial concentration were considered fixed effects, random effects, and the independent variable, respectively, as follows: where D 216,i , represents the desorption amount of TN to the biochar i (mg g −1 biochar) at time 216 (h), α and β are fixed parameters for birch biochar, α S and β S are fixed parameters for spruce biochar expressed as the difference from the birch fixed parameters, and X ini is the initial concentration (0.2 and 0.35 mg L −1 ) for TN. In the case of birch biochar, α S and β S are zero. The amount of adsorbed TN at the end of the adsorption phase of the desorption experiment (mg g −1 biochar), and ε ij is a residual term for TN in biochar i and replicate j with an expectation value of 0. The performance of the model was reported as information criteria (AIC, BIC, -log likelihood, RMSE).
Adsorption of TN was detected in all biochar treatments, and the adsorption amount during the experiment depended on the initial TN concentration (Fig. 2, Eq. 2, Table 2). Adsorption amount increased linearly from 0.05 to 0.2 mg TN g −1 biochar for spruce and from 0.07 to 0.3 mg TN g −1 biochar for birch (Fig. 2). Birch biochar adsorbed more TN than the spruce biochar ( Table 2).
The adsorption kinetics of TN were calculated based on the concentration differences between consecutive water samplings (Table 3). There was rapid adsorption over the first few days; after which, the adsorption slowed down (Fig. 3). The initial concentration and type of biochar did not affect k ad (Table 3). Q max was increased with the increasing initial TN concentration. According to the pseudo-first order and second order models, the Q max was about 3-4 times and 2 times higher for initial concentrations of 4 mg L −1 and 1 mg L −1 , respectively. Birch biochar had a higher Q max than spruce biochar (Table 3). The adsorption process was more consistent in the higher concentrations (Fig. 3).
The concentrations of NH 4 + -N were below the detection limit (0.02 mg L −1 ) throughout the experiment; and therefore, it was not possible to examine  1 3 the concentration changes and adsorption parameters for NH 4 + -N.

Desorption Experiment
In the adsorption phase, biochar was shaken in initial TN concentrations of 1 mg L −1 , 2 mg L −1 , 3 mg L −1 , and 4 mg L −1 ; after which, the biochars were transferred to lower TN concentrations (0.2 and 0.35 mg L −1 ) to study the desorption. TN desorption amount of birch biochar depended on the initial TN concentration in the adsorption phase, and the desorption was significantly different between the spruce and birch biochar (Table 5A, B). In 0.35 mg L −1 , the (Table 5B) spruce kept adsorbing TN also during the desorption phase of the experiment indicated by the negative values of   (Fig. 6B). Birch biochar had a predominantly higher TN desorption amount when the TN concentration was higher than 1.8 mg L −1 (Fig. 6A,  B). The birch biochar desorption was associated with the TN adsorption in smaller initial concentrations (Fig. 6C). Birch biochar released TN for adsorption higher than 0.04 mg g −1 biochar (Fig. 6C, 6D). The desorption difference for spruce biochar was not significant in smaller and higher concentrations (Table 5C, D, Fig. 6). TN desorption amount of birch biochar was associated with the TN adsorption (Table 5C, D). For spruce, the TN desorption amount in 0.2 mg L −1 was very low (Fig. 6A, C). The desorption process was more consistent for birch biochar in both 0.2 and 0.35 mg L −1 (Fig. 6). This suggests that the desorption process was dependent on the initial concentration.

Discussion
The current study showed that birch and spruce biochar effectively adsorbed TN and NO 3 − -N in forest runoff waters. Our results support previous studies by Saarela et al. (2020) andKakaei Lafdani et al. (2020). The adsorption amount of spruce and birch biochar increased with the increasing TN concentration in water (Figs. 2 and 3), which is probably due to the increased N availability in the solution and increased TN concentration gradient along the flow path from the solution to external surfaces and finally to internal surfaces of biochar (Lagritte & Pasquier, 2016;Ahmadvand et al., 2018). The results are consistent with the results obtained by Ahmadvand et al. (2018) and Saarela et al. (2020), showing that with an increase in the initial N concentration of the solution, the adsorption amount of biochar tends to increase. The low initial concentrations of NH 4 + -N (< 0.02 mg L −1 ) was the reason why we could not determine the adsorption parameters for this compound, but based on the literature, it is known that biochar is an effective adsorbent for NH 4 + -N (Wang et al., 2015;Takaya et al., 2016;Yin et al., 2017). The surfaces of carbon-based materials are generally negatively charged (Krishnan & Haridas, 2008;Yao et al., 2011), which is thought to make them poor adsorbents Table 5 Parameter estimates, standard errors of the mean, degrees of freedom, p-values, and performance criteria of linear mixed-effects models for TN desorption (Eq. 6) in the concentration of 0.2 mg L −1 treatments (left: A and C) and 0.35 mg L −1 treatments (right: B and D). The parameters α and β represent birch biochar and the parameters for spruce biochar are obtained as (α + α S ) and (β + β S ) Bold values are statistically significant.    (Table 4, Fig. 5) took place which indicates that biochar is capable of adsorbing anions, which is consistent with the results of Chintala et al. (2013) and Cai et al. (2016). In general, adsorption of ions onto the biochar surfaces can be categorized into 4 different classes: precipitation (salt precipitation), chemical adsorption (chemical reaction with surface functional groups), physical adsorption (entrapped in the solution present in interior pores), or electrostatic adsorption (cation or anion exchange to charged functional groups on biochar surfaces) (Fidel et al., 2018;Haider et al., 2016;Noel et al., 2007). Biochar's capacity to adsorb NO 3 − -N is often attributed to its high surface area and porosity and positively charged functional groups (Fidel et al., 2018). NO 3 − -N adsorption has been shown to increase with increasing biochar pyrolysis temperature and decreasing pH, likely reflecting an anion exchange adsorption mechanism (Fidel et al., 2018). Previous studies have reported mixed results about the biochar capacity to adsorb anions. Some studies have shown that biochar also effectively adsorbed NO 3 − -N (Ahmadvand et al., 2018;Chintala et al., 2013) through its numerous functional groups, such as carboxylic, hydroxyl, lactone, and ketone groups (Cai et al., 2016). Chintala et al. (2013) reported that with decreasing liquid pH, biochar can adsorb more NO 3 − -N because due to protonation reaction, the biochar surfaces become more positively charged. However, when the solution contains high concentrations of other anions, the competitive adsorption decreases the adsorption of NO 3 − -N (Chintala et al., 2013). Nevertheless, the inability or weak ability of biochar to adsorb N compounds has been reported as well (Hollister et al., 2013). Yao et al. (2012) reported that among 13 different kinds of biochars, only 2 types of biochars were able to adsorb NO 3 − -N. Gai et al. (2014) showed some NO 3 − -N was even released from the biochar. The inconsistency among the results can be explained by various preparation conditions, different characteristics of biochar produced from differing biomass, as well as various initial concentrations of N compounds in the solution.
The interaction between the adsorbents and adsorbates at equilibrium can be described using kinetic models. Pseudo-first and second-order kinetic models consider chemical adsorption mechanisms between the adsorbate and adsorbent and provide valuable information on adsorption rate and maximum 1 3 adsorption capacity (Zhao et al., 2018). In the current study, both models showed rather similar performance against the experimental data (Fig. 3, Table 3). The maximum adsorption capacity at equilibrium, Q max , was increased with increasing TN concentrations in the solution. Similar results were also reported by other researchers where maximum adsorption capacity increased with the increasing initial concentration of N (Milmile et al., 2011;Olgun et al., 2013). The pseudofirst order model results showed that the adsorption rate, k ad , of spruce biochar decreased by increasing Q max (Table 3). The pseudo-second order kinetic model assumes that the rate-limiting step is the interaction between adsorbate and adsorbent, and it is usually used to describe chemical adsorption (Zhen et al., 2015). According to the kinetic models' results, the Q max1 was about 3-4 times and Q max2 was 2 times higher for TN concentration 4 mg L −1 compared to 1 mg L −1 .
Our results imply that birch biochar has predominantly higher Q max than spruce biochar (Fig. 3) which is consistent with the results obtained by Saarela et al. (2020). Despite the higher specific surface area of spruce biochar (320 m 2 g −1 ) compared to birch biochar (260 m 2 g −1 ), the results demonstrated that Q max was higher with birch biochar than spruce biochar (Table 3). This indicates that N maximum adsorption capacity is not exclusively dependent on the biochar surface area Takaya et al., 2016;Zhang et al., 2012Zhang et al., , 2014. Q max increased with the increasing TN concentration, which can be due to N availability in solution with higher initial concentrations (Lagritte & Pasquier, 2016).
The desorption amount was significantly higher in birch biochar than spruce biochar, which agreed with the higher birch biochar adsorption amount. In contrast, spruce biochar showed negative desorption values which indicated TN adsorption instead of desorption. This shows that spruce biochar was not saturated. TN desorption amount increased by increasing the water TN concentration for birch biochar, which is coherent with the results by Chintala et al. (2013). Mechanisms behind the desorption are connected to lower bond energy with more exchangeable ions in water, entrapment of organic compounds in micropores or pore deformation mechanisms (Braida et al., 2003;Chintala et al., 2013;Dechene et al., 2014;Loganathan et al., 2009;Zhang et al., 2010). Differences between adsorption and desorption isotherms are also reported by Zhu and Selim (2000).
Using biochar in the removal of N compounds from runoff water in peatland forests can be difficult to predict because of the temporal variation in TN concentrations. The current study showed that despite the maximum adsorption of TN to birch biochar (0.309 mg TN g biochar −1 ), only 1.29% was desorbed (0.004 mg TN g biochar −1 ). The amount of released TN by the spruce biochar was only about 0.9% (0.002 mg TN g biochar −1 ) of maximum TN absorbed (0.208 mg TN g biochar −1 ). The results illustrated that the desorption rate of TN from spruce and birch biochar has been negligible, which makes biochar a promising water protection tool especially in watersheds where the aquatic ecosystems are exposed to high nutrient loads from peatlands.

Conclusions
This study examined the N adsorption and desorption behaviors of biochar in peatland forest runoff water. Our results showed efficient adsorption of both organic and inorganic N from peatland runoff waters. Despite the efficient adsorption, only a small fraction of the adsorbed TN was released through desorption when the biochar was exposed to water with low TN concentrations. The seasonal fluctuation of nutrient concentrations emphasizes the need for an effective water protection tool in peatland forestry to which adsorption-based nutrient removal with biochar seems to be an effective water protection tool. Therefore, biochar provides an optional way to runoff water purification in forested peatland areas for times when traditional water protection methods are inefficient.

3
Hospital. This study was funded by Marjatta and Eino Kolli Foundation and the Doctoral Programme in Forests and Bioresources (Early Stage Researcher funding), University of Eastern Finland. Writing was performed out with funding from the Academy of Finland to strengthen university research profiles in Finland for the years 2017-2021 (funding decision 311925) and the REFORMWATER project (funding decision 326818).

Availability of Data and Material
Data is publicly available in the supplementary materials.
Code Availability N/A.

Conflict of Interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. Shaking water samples at 110 rpm for 9 days Taking water samples and analysing N concentration at the end