Different relationships of fine root traits with root ammonium and nitrate uptake rates in conifer forests

ABSTRACT Nitrogen (N) uptake by fine roots of trees is important for understanding root physiological function in forest ecosystems. However, direct investigations of in situ ammonium and nitrate uptake rates are limited. Thus, we aimed to clarify inorganic N uptake rates among tree species and determine the factors controlling N uptake through relationships with fine root traits in cool temperate coniferous forests. Using a solution depletion method for measuring N uptake, we observed the relationship of N uptake rate in the form of NH4 + and NO3 – by an intact root system with root morphological traits, such as root diameter, specific root length (SRL), and root tissue density (RTD). The coniferous roots in this study preferred NH4 + to NO3 −. Across species, there were significant relationships between NH4 + uptake and diameter, SRL, and RTD; in contrast, only RTD had a significant impact on NO3 − uptake. Relationships between N uptake rates and root morphological traits differed between NH4 + and NO3 –. We found that the relationship of inorganic N uptake with morphological traits depended on the characteristics of the N form adsorbed through soil and on tree N assimilation efficiency. Our results make a breakthrough in the understanding of root physiological function and the prediction of fundamental N acquisition strategies.


Introduction
Nitrogen (N) is one of the most essential and limiting factors for plant growth and development in temperate forest ecosystems (Vitousek and Howarth 1991;LeBauer and Treseder 2008). The existing N content in forest ecosystems is approximately 5,300 kg ha −1 , and the amount of N stored in the soil is 90% (Bormann et al. 1977). In soil, complex organic polymers in litter are decomposed and converted into bioavailable N-containing monomers (such as amino acids, amino sugars, and nucleic acids) that can be used by either plants or microorganisms (Schimel and Bennett 2004). The N content is mineralized by microorganisms, producing ammonium (NH 4 + ), and transformed to nitrate (NO 3 − ) by nitrification. Trees acquire almost all their N demand through root uptake. Generally, NH 4 + and NO 3 − as inorganic N are the major N sources for trees (Haynes et al. 1978). NH 4 + and NO 3 − differ in their assimilation efficiency and solubility in soil (Lambers and Oliveira 2019). While NO 3 − is readily available owing to its high solubility in soil water, plants must use energy to reduce NO 3 − before incorporation into amino acids. Thus, it may be more efficient for plants to take up NH 4 + because it can be immediately incorporated into amino acids. However, NH 4 + is an immobile form of N, and is not always available for uptake because it is strongly adsorbed onto soil particles. Our crucial concern is the form and amount of N acquired by the tree root from the soil as a nutrient use strategy. In addition, as inorganic acquisition by trees is a form of major N flow in forest ecosystems (Bormann et al. 1977), N uptake by tree roots is important for understanding N cycles.
Previous studies have directly measured the N uptake of plant roots and evaluated the effects of tree species (Gessler et al. 1998;Templer and Dawson 2004). For example, fine roots of American beech (Fagus grandifolia Ehrh.) seedlings took up more NO 3 − than NH 4 + , while sugar maple (Acer saccharum Marsh.) preferred to take up NH 4 + over NO 3 − (Templer and Dawson 2004). In warm and cool temperate forests, the uptake preferences of chemical N forms vary among plant species (Gao et al. 2020). In temperate and tropical forests, Liu et al. (2017) reported that tree species commonly took up more NH 4 + than NO 3 − , but the contributions of each N form were different. As there is inadequate data on N preferences in tree species, we still do not know the background mechanisms of N uptake characteristics across tree species.
An approach based on root functional traits can be one of the most useful for understanding the physiological mechanisms of tree fine roots (McCormack et al. 2017). For example, absorptive roots are generally characterized by small diameter, high specific root length (SRL: ratio of root length to mass), and/or low root tissue density (RTD: ratio of root mass to volume), resulting in longer and thinner roots that are considered to have greater capacity for water and nutrient acquisition when compared with shorter and thicker roots (Comas et al. 2002;Comas and Eissenstat 2004;Hodge 2004). These traits would reflect the availability of resources and indicate physiology, lifespan, and foraging strategy (Tjoelker et al. 2005;Ostonen et al. 2011;Adams et al. 2013;Makita et al. 2015Makita et al. , 2016, allowing comparisons across species (Comas and Eissenstat 2009;Liese et al. 2017;Yahara et al. 2019).
In the present study, we focused on the relationship between root N uptake and the morphological traits of fine root systems, including SRL and RTD. Based on differences in the solubility of N forms in soil, we hypothesized that NH 4 + is more easily acquired by roots with higher foraging ability (lower diameter, higher SRL, and lower RTD), because it is immobile in soil water. We also hypothesized that NO 3 − uptake would have weak relationships with morphological traits because of its high solubility. Determining these relationships would be helpful for understanding the generality of their impacts on root N uptake from soil.
To achieve the direct field measurement of N uptake by fine roots, we used the in situ depletion method for measuring N uptake in solution by roots (Bassirirad et al. 1999;Socci and Templer 2011;Campbell et al. 2014). This depletion method for net N uptake can quantify the change in N solution by soaking fine roots in a known N solution. It is suitable for evaluating fine roots with a thinner size, which have high physiological activity (Makita et al. 2009(Makita et al. , 2015. Overall, we aimed to relate net N uptake rates in the fine root system with morphological traits under forest field conditions at the interspecific level. Our objectives were: (1) to explore differences among species in terms of uptake rates for each N form; (2) to examine the relationships of fine root NH 4 + and NO 3 − uptake with root morphological traits, including SRL and RTD; and (3) to identify common patterns across species.

Study site
The study was conducted in Terasawayama Station Research Forest (35°53 N, 138°02E; 960-1,450 m a.s.l.) in Nagano, Japan. The annual temperature and precipitation are 9.8°C and 1,469 mm, respectively. In the study site, the soil net mineralization rate was 233.6 mg N kg −1 year −1 and the net nitrification rate was 204.1 mg N kg −1 year −1 (Hosokawa et al. 2012). This is a cool-temperate forest, with an area of 228 ha for coniferous plantations. The forest consists of plantations in larch (Larix kaempferi (Lambert) Carrière), red pine (Pinus densiflora Sieb. et Zucc.), Japanese cypress (Chamaecyparis obtusa Sieb. et Zucc.), and Japanese cedar (Cryptomeria japonica D. Don), which were selected as the target species. In Japan, these four tree species have been widely planted for timber products and cover approximately 70% of the national plantation area in Japan (Forestry Agency 2018). Larix kaempferi is a deciduous conifer, and Pinus densiflora is an evergreen conifer that is associated with ectomycorrhizal (ECM) fungi. The C. obtusa and C. japonica are evergreen conifers that are associated with arbuscular mycorrhiza (AM) fungi. At this site, studies have been conducted on root functional traits Yahara et al. 2019) and root exudation (Akatsuki and Makita 2020) in several woody species.

Field experiment of root N uptake measurement
We established a 30 × 30 m plot for each species at the site and selected eight mature trees per plot. The mean diameter at breast height of the mature trees was 38.5 cm. The root uptake of NH 4 + and NO 3 − by the target tree species was measured on seven sunny days in August 2020, corresponding approximately to the peak of aboveground growth. The procedure for root N uptake was as follows: intact root systems of the target tree species were exposed at a depth of 10 cm from the top of the soil, including the organic layer, which was attached to the mother root. The root systems of the target tree species were distinguished by their morphological traits, such as color, diameter, branching pattern, root tips, and mycorrhizal type . Living root systems were distinguished from dead roots based on their color and elasticity. The soil was removed from the larger root system by carefully washing with deionized water. Larger root systems were re-buried until the N uptake measurements (generally 24 h later). The re-burial process was used for root exudate measurement (Phillips et al. 2008) and was intended to allow additional time for roots to recover from any potential injury or stress sustained during the excavation and rinsing process.
After re-burial, two intact root segments (<2 mm diameter, approximately 10 cm in length along the main axis) were selected for N uptake measurements. They were rinsed with de-ionized water, gently wiped with a paper towel, and inserted into a tube (10 mL). Next, 10 mL of N mixed solution (containing 200 μmol L −1 NH 4 Cl and KNO 3 ) was injected into the tube ( Figure 1). We determined the concentration of the N solution by referring to Japanese forest soil datasets (Urakawa et al. 2015) and previous studies that used a depletion method (e.g. Socci and Templer 2011). The remaining root parts outside the tube were covered with wet kitchen paper to maintain root moisture. To shade the sample, the whole root system was wrapped with aluminum foil and covered with a shading sheet. The root segments soaked in the N solution were incubated for 90 min (Socci and   Campbell et al. 2014). After incubation, the root segments were collected from the N solutions and sampled. The N solution samples were immediately filtered using syringe filters (0.45 μm; Membrane Solutions Limited, Plano, TX, USA) and placed in centrifuge tubes. The root samples were stored in zipper storage bags. In total, we evaluated 64 samples of solutions and fine roots (4 tree species × 16 replications). Simultaneously, we prepared control samples of the same NH 4 Cl and KNO 3 concentrations without roots as a blank. All root and N solution samples were placed in an icebox and transported to the laboratory.

Measurement of N concentration in solution
In the laboratory, the NH 4 + and NO 3 − concentrations of the solution samples were measured using a microplate technique (Shand et al. 2008). The N solution samples were colored by the Berthelot reaction for NH 4 + and by the vanadium oxidizing agent and Griess reagent for NO 3 − . The concentration of each N solution sample was measured using a microplate reader (Multiskan Sky, Thermo Fisher Scientific, Cambridge, MA, USA). It is important to note that the experimental error in the analytical method was <2.0% for NH 4 + and <0.2% for NO 3 − . To decrease the experimental error, each experiment was carried out in triplicate.
For the calculation of in situ N uptake, the difference in N solution concentration before (N control ) and after (N solution ) the measurement was multiplied by the solution volume (V solution ) and divided by the dry weight of the root sample (DW root ) and incubation time (T). The calculation of each N from the uptake rates by the root (N uptake ) was calculated as follows: where N uptake (μmol N g −1 h −1 ) is the uptake rate of NH 4 + or NO 3 − by the fine roots; DW root (g) is the dry weight of the root sample; N control (μmol L −1 ) is the NH 4 + or NO 3 − concentration of the control N solution (before measurement); N solution (μmol L −1 ) is the NH 4 + or NO 3 − concentration of the N solution samples (after measurement); V solution (L) is the volume of the N solution; and T (h) is the incubation time.

Root morphological analysis
The root samples were scanned using a double-lamp bed scanner (GT-X980; EPSON, Nagano, Japan). We analyzed the total root length, volume, and mean diameter using a WinRHIZO Pro 2013a (Regent Instruments, Quebec, Canada); the settings were as follows: image mode: 16-bit Gray (Std); resolution: 600 dpi; scale: 100%. After scanning, the root samples were dried at 50°C for >48 h to a constant mass and weighed. The total root length, volume, and dry mass were used to calculate the SRL (m g −1 ) and RTD (g cm −3 ) of the samples.

Soil sampling and measurement
The soil samples were collected next to the location of root sampling. For the analysis of chemical characteristics, six replicate soil samples were taken per plot from the soil surface (0-10 cm) by a shovel and transported to the laboratory keeping in cooler box. In the laboratory, the samples were immediately sieved through a 2-mm mesh; then, 8 g of soil was mixed with 40 mL of 2 M KCl solution, and shaken for 12 h. Next, the shaken soil samples were filtered using ADVANTEC 5C filters (ADVANTEC, Tokyo, Japan). The NH 4 + and NO 3 − concentrations of the filtered solutions were measured using the same process as the N solution measuring method in this study, and the concentration of the samples was converted to unit soil dry weight. The other soil samples were air-dried for two weeks, and the pH and electrical conductivity (EC) were determined in a 1:5 dried soil: distilled water mixture using a pH & EC meter (LAQUA D-210PC-S, Horiba, Kyoto, Japan). Then, we measured the soil C and N concentrations (%) using NC analyzer (Flash EA 1112, Thermo Fisher Scientific, Cambridge, MA, USA), and calculated soil CN ratio.
For the analysis of water content and bulk density, the soil samples were taken from the soil surface using a 100cc core sampler. In the laboratory, fresh soil was weighed and dried at 70°C for >48 h to a constant mass. After the dry soil weight was determined, the gravimetric soil water content (%) was calculated per unit mass of oven-dried soil. In addition, the dry soil samples were sieved through a 2-mm mesh for bulk density. Finally, the bulk density, which is the weight of dry soil divided by the total soil volume, was calculated.

Statistical analysis
The effect of tree species on NH 4 + and NO 3 − uptake rates and root traits were tested using the Brunner-Munzel test, which robustly identifies differences in variance between different probability distributions (Brunner and Munzel 2000) and multiple tests by Benjamini and Hochberg (1995). To examine common pattern across species, the relationships between N uptake rates and root traits (diameter, SRL, and RTD) in all pooled data were analyzed using regression analysis. All statistical analyses were performed using R version 3.6.2 software (R Core Team 2019).

Soil characteristics
The means of the soil pH, EC, and CN ratio in each plot ranged from 4.58-5.97, 37.4-104.7 μS cm −1 , and 15.14-26.09, respectively ( Table  1). The soil NH 4 -N and NO 3 -N concentration ranges were 9.12-11.69 mg kg −1 and 0.12-6.23 mg kg −1 , respectively. The soil NH 4 + concentration was higher than the NO 3 − concentration in all the plots. The soil water content and dry bulk density ranges were 14.9-35.0% and 332-509 g L −1 , respectively.

Root uptake rates of NH 4 + and NO 3 −
The uptake rates of NH 4 + and NO 3 − by fine roots in the four coniferous species were 2.37-14.55 μmol N g −1 h −1 and −1.59-2.09 μmol N g −1 h −1 , respectively (Figure 2). The fine roots in all the species absorbed more NH 4 + than NO 3 − . Comparing the N uptake capacity among the species, the NH 4 + uptake rate of L. kaempferi was significantly higher than those of the other tree species (Figure 2a). The NO 3 − uptake rates of C. obtusa and C. japonica tended to be higher than those of L. kaempferi and P. densiflora (Figure 2b).

Fine root morphology
In terms of root morphological traits, the mean root diameter of root samples across the four coniferous species ranged from 0.26 to 0.69 mm, while SRL ranged from 8.72 to 34.10 m g −1 , and RTD ranged from 0.21 to 0.70 g cm −3 . There were significant differences in the measured root morphological traits among the species (Table 2; P < 0.05).

Relationships of NH 4 + and NO 3 − uptake rates with fine root traits
The NH 4 + uptake rate was significantly higher when mean diameter was smaller (P < 0.001) and SRL was higher (P < 0.001) across species (Figure 3a, b), and had a significantly positive relationship with the RTD (P < 0.001; Figure 3c). In contrast, NO 3 − uptake rates had no relationship with the mean diameter or SRL (Figure 3d, e), but were negatively related to RTD (P < 0.05; Figure 3f).

Discussion
We observed the potential inorganic N uptake rates of fine roots in four coniferous forest species, and our findings provide direct evidence of the relationship between root morphological traits and nutrient acquisition. Using influx in the solution depletion experiments, the NH 4 + and NO 3 − uptake rates of this study ranged from 2.37 to 14.55 and −1.59 to 2.09 μmol N g −1 h −1 , respectively (Figure 2). These values overlap with those of other studies in a cool-temperate forest (Socci and Templer 2011) and in laboratories (Gruffman et al. 2014). For example, Socci and Templer (2011) reported that NH 4 + uptake rates of sugar maple (Acer saccharum) and red spruce (Picea rubens) ranged from 7.05 to 10.59 μmol N g −1 h −1 , and that NO 3 − uptake rates ranged from −36.38 to 2.49 µmol N g −1 h −1 in the depletion method. In addition, when comparing the NH 4 + and NO 3 − uptake rates in all the target tree species, the NH 4 + uptake rates were over 30 times higher than those of the NO 3 − (Figure 2). This result is consistent with the findings of previous studies in boreal (Socci and Templer 2011), temperate (Liu et al. 2017), subtropical (Liu et al. 2020), and tropical (Liu et al.  2017) forests. In this study, the fine roots of the coniferous species preferred to take up N in the form of NH 4 + from the soil. Overall, we found a clear distinction in the specific N preference through N uptake between NH 4 + and NO 3 − in a small root segment in a cooltemperate coniferous forest.
When looking at NH 4 + uptake capacity as a feature of higher N preference in the study species, the NH 4 + uptake rate of L. kaempferi was significantly higher than that of the other tree species (Figure 2a). This finding is probably attributable to the difference of evergreen and deciduous species. Deciduous species have been reported to show higher capacity of C assimilation (Baldocchi et al. 2010) and higher photosynthetic N-use efficiency (Bai et al. 2015) than do the evergreen species. Given that photosynthetic assimilation in trees is closely linked with belowground C allocation (Hartmann et al. 2020), it would be related to the physiological functions of roots, such as the rate of N uptake (McCormack et al. 2015). Kayama et al. (2009) reported that deciduous conifer seedlings exhibit N uptake rates higher than do evergreen conifers. Consequently, deciduous species in growing season of this study were more acquisitive with N resource uses than evergreen species. However, information is limited on varied physiological functions of fine roots arising from differences in their ecological characteristics between evergreen and deciduous tree species in forests (Wang et al. 2021). Therefore, the effects of different tree species on N uptake rates by roots need further exploration.
To understand how N uptake rates by tree fine roots differed between NH 4 + and NO 3 − , we focused on their morphological traits. Surprisingly, the relationships between N uptake rate and root morphological traits across species were different for NH 4 + and NO 3 − (Figure 3). The NH 4 + uptake rate by the fine roots was negatively correlated with mean diameter and positively correlated with SRL. In contrast, the NO 3 − uptake rate had no relationship with the mean diameter and SRL. In general, both root diameter and SRL are known parameters of the ability to absorb nutrients and water (McCormack et al. 2015). Thinner roots and higher SRL could allow the efficient exploitation of soil resources and aid in resource foraging (Bowsher et al. 2016), which would be highly beneficial for the absorption of N by plants (Ryser 2006;Holdaway et al. 2011). Indeed, in L. kaempferi in this study, the thinner roots with higher SRL led to higher uptake rates of NH 4 + ( Figure 2, Table 2). These results reveal different factors controlling species-specific N uptake through fine root morphological changes between N forms.
The relationship between N uptake and morphology in our study may suggest that different soil adsorptions between NH 4 + and NO 3 − can affect the magnitude of diameter and branching of the roots. In soil, NH 4 + is strongly adsorbed by soil particles because the surface of soil particles has a negative charge (Lambers et al. 2019). The roots must modify the length and diameter for high foraging ability and may be able to access NH 4 + in soil, resulting in a higher NH 4 + uptake rate. In contrast, NO 3 − has high solubility in soil water because soil particles electrically repel it (Lambers Figure 3. Relationship between fine root NH 4 + or NO 3 − uptake rate and (a) mean root diameter (mm), (b) specific root length (SRL; m g −1 ), and (c) root tissue density (RTD; g cm −3 ) of Larix kaempferi (yellow squares), Pinus densiflora (red circles), Chamaecyparis obtusa (blue triangles), and Cryptomeria japonica (light blue diamonds). The black lines across the plots indicate a significant relationship among all species (P < 0.05). et al. 2019). In other words, NO 3 − is easily diffused to the root surface through water flow, realizing NO 3 − uptake without dependency on root morphological adaptation for NO 3 − . The roots can absorb the NO 3 − solution more easily in the soil. Thus, the characteristics of soil adsorption in the field are reflected in the difference in the mobility of different N forms and, thus, are essential for improving the specific nutrient acquisition by fine roots.
Furthermore, the higher uptake rates of NH 4 + by fine roots can be attributed to differences in the N assimilation process of plants (Lambers et al. 2019). Plants must use energy to reduce NO 3 − before being incorporated into amino acids. In contrast, NH 4 + can be immediately incorporated into amino acids to reduce energy. Thus, it could be more efficient for plants to take up NH 4 + than NO 3 − . In addition, as mentioned above, most of the NH 4 + in soil is usually adsorbed by soil particles. In this study, we used a depletion method to evaluate the potential capacity of net N uptake among four coniferous species. Inter-specific differences were clearly observed in the NH 4 + -absorbing capacity due to differences in root morphology (Figure 3). We suggest that root morphology is optimized for the absorption of NH 4 + , which is immobilized in soil. This finding highlights a key factor that enhances the ability to absorb soil N and to achieve favorable growth in a variety of environments. Thus, the roots of coniferous species may exhibit a foraging strategy in response to soil N status and then acquire and utilize N forms that are abundant in the soil and facilitate efficient assimilation.
Interestingly, there were significant relationships between the RTD and N uptake rates of both NH 4 + and NO 3 − but in different directions. The RTD was positively correlated with NH 4 + and negatively correlated with NO 3 − (Figure 3c, f). From this result, we imply that RTD is assumed to follow a trade-off between rapid growth rate/resource capture and high survival/resource conservation through NH 4 + and NO 3 − uptake. In general, fine roots with high RTD have been associated with slow growth, low water permeability, and low nutrient availability, while high resistance to external stress and longer life (Weemstra et al. 2016). The RTD and N uptake relationships in our study may be helpful for interpreting N acquisition strategies with potential functional trade-offs. However, it is still unclear whether the relationship between root RTD and N uptake is reflected in the form of NH 4 + and NO 3 − . It is likely that the NO 3 − uptake rate is affected by nitrate reductase activity, which is an indicator of NO 3 − assimilation and the NO 3 − utilization ability of plants (Koyama et al. 2008). Further studies are needed to explore the general mechanisms of N uptake function.

Conclusions
The results of this study confirm that the fine roots of coniferous species prefer the NH 4 + form of N over NO 3 − ( Figure 2). In addition, the relationships between N uptake rates and root morphological traits differ between NH 4 + and NO 3 − (Figure 3). The coniferous species in this study might optimize their root morphological traits in response to the availability of NH 4 + and NO 3 − . Root traits are ideally known to be related to nutrient uptake function (McCormack et al. 2015). The relationships between root nutrient uptake and root traits have been discussed both acquisitively and conservatively (Weemstra et al. 2016); however, the actual situation may be more complex and the relationships may change for different forms of nutrients, such as NH 4 + and NO 3 -. Indeed, we found that relationships between nutrient uptake and morphological traits depend on the characteristics of the essential mineral elements. Further research on speciesspecific N uptake in the field is required to fully understand the relationships between root N uptake and their control over morphological processes in various species and environments. This study offers new insight into the in situ N uptake of tree fine roots, and provides a breakthrough in our understanding of root physiological function and the prediction of the fundamental N acquisition strategy of trees.
Overall, we evaluated the net NH 4 + and NO 3 − uptake rates using an in situ depletion method in coniferous forests. To understand tree N uptake under forest field conditions, we propose to combine information regarding specific root N uptake and soil N availability. An approach by such depletion method could be helpful for identifying the potential species-specific function of small root segments of even the terminal root (<0.5 mm). On the other hand, stable isotope 15 N tracer method for gross root N uptake is suitable for N acquisition from soil in natural conditions, because it can applied to the whole roots without soil disturbance (Averill and Finzi 2011;McKane et al. 2002;Gallet-Budynek et al. 2009). Future studies on combined approaches could lead to a better understanding of species-specific root function and ecosystem function for heterogeneous soil nutrients.