Root Morphology and Secretion of two subtropical tree species to NH4+-N and NO3–-N Deposition

Background: Both NH 4+ and NO 3– are capable of greatly inuencing plants’ growth and biomass. However, the belowground responses of subtropical trees to either NH 4+ or NO 3– deposition remain poorly understood. Here, we discuss how these two forms of N deposition can affect root development, and experimentally analyzed how they could impact nitrogen and phosphorus absorption in two types (broadleaved with a brous root system vs. conifer with a tap root system) of subtropical tree species. Results: In a greenhouse in southern China, 1-year-old S. superba and P. massoniana seedlings grown on P-limited and P-normal soil were treated with NaNO 3 and NH 4 Cl solutions of 0, 80, and 200 kg N ha –1 year –1 , corresponding to the control, N80, and N200 groups, respectively. Root phenotype characteristics and metabolism ability were measured after 8 months of growth. The results showed that the root morphology and physiology variables differed signicantly between the two species under different N and P treatments. Although S. superba had a larger quantity of roots than P. massoniana, both its root growth rate and root absorption were respectively lower and weaker. N addition differentially affected root growth and activity as follows: (1) NO 3– -N80 and NH 4+ -N80 increased root growth and activity of the two species, but NH 4+ -N80 led to thicker roots in S. superba; (2) NO 3– -N200 and NH 4+ -N200 had inhibitory effects on the roots of P. massoniana, for which NH 4+ -N200 led to thinner and longer roots and even the death of some roots; and (3) NH 4+ -N could promote metabolic activity in thicker roots (> 1.5 mm) and the NO 3– -N was found to stimulate activity in thinner roots (0.5–1.5 mm) in the brous root system having a larger quantity of roots, namely S. superba. By contrast, NO 3– -N and NH 4+ -N had an opposite inuence upon functioning in the tap root system with a slender root, namely P. massoniana. Conclusion: We conclude P. massoniana has a much higher root absorption eciency; however, nitrogen deposition is


Abstract
Background: Both NH 4 + and NO 3 are capable of greatly in uencing plants' growth and biomass. However, the belowground responses of subtropical trees to either NH 4 + or NO 3 deposition remain poorly understood. Here, we discuss how these two forms of N deposition can affect root development, and experimentally analyzed how they could impact nitrogen and phosphorus absorption in two types (broadleaved with a brous root system vs. conifer with a tap root system) of subtropical tree species.
Results: In a greenhouse in southern China, 1-year-old S. superba and P. massoniana seedlings grown on P-limited and P-normal soil were treated with NaNO 3 and NH 4 Cl solutions of 0, 80, and 200 kg N ha -1 year -1 , corresponding to the control, N80, and N200 groups, respectively. Root phenotype characteristics and metabolism ability were measured after 8 months of growth. The results showed that the root morphology and physiology variables differed signi cantly between the two species under different N and P treatments. Although S. superba had a larger quantity of roots than P. massoniana, both its root growth rate and root absorption were respectively lower and weaker. N addition differentially affected root growth and activity as follows: (1) NO 3 --N80 and NH 4 + -N80 increased root growth and activity of the two species, but NH 4 + -N80 led to thicker roots in S. superba; (2) NO 3 --N200 and NH 4 + -N200 had inhibitory effects on the roots of P. massoniana, for which NH 4 + -N200 led to thinner and longer roots and even the death of some roots; and (3) NH 4 + -N could promote metabolic activity in thicker roots (> 1.5 mm) and the NO 3 --N was found to stimulate activity in thinner roots (0.5-1.5 mm) in the brous root system having a larger quantity of roots, namely S. superba. By contrast, NO 3 --N and NH 4 + -N had an opposite in uence upon functioning in the tap root system with a slender root, namely P. massoniana.
Conclusion: We conclude P. massoniana has a much higher root absorption e ciency; however, nitrogen deposition is more bene cial to the root growth of S. superba.

Background
The root is a vital organ in terrestrial plants that can x the plant into the soil and absorb nutrients and water from the soil to promote its growth, especially via ne roots [] . Much research has shown that root morphology and activity can greatly impact nutrient absorption, yet nutrition can also shape the roots' phenotype and in uence its physiological activity [−] . For example, when the soil is rich in nutrients, the lateral root (LR) and total root length undergo increased growth, whereas in barren soil root growth is poor [−] . This contrasts with studies drawing the opposite conclusion, such when the total root volume and root secretions are decreased in rich soil [] . These contradictory ndings are mainly due to the nutrient elements' type, forms, distribution, concentrations, and the species studied [−] .
Nitrogen (N) and phosphorus (P) are the two fundamentally important elements in plants and both directly in uence root growth [8,10−11,] . In the subtropics, however, soil P is usually present at low levels, particularly in forest soils in China [] . To overcome this limitation, plants growing in this region usually have a specialized root morphology and secretion mechanism to promote their phosphorus acquisition e ciency (PAE); the latter de ned as the ability of the plant to acquire P from the soil [11] . For example, depending on the species, the plant will increase the root base branch angle and density of its lateral roots and root hairs to form a shallow root system, or increase the clustering of roots and colonization by arbuscular mycorrhiza (AM) or ectomycorrhizal (ECM) fungi, to increase its PAE [−] . Recent research suggests that ECM fungi are most bene cial to plants in N-limited environments, while AM fungi dominate in P-limited environments [17−] . At the same time, because the rate of atmospheric deposition of nitrogen (N) is high and accelerating in subtropical forests of southern China [], the nitrogen acquisition e ciency (NAE) of plants could also be seriously affected. Zhu et al. reported an average deposition ux of N in China of 13.69 ± 8.69 kg ha − 1 a − 1 [] and this deposition trend is expected to increase in future decades, such that the production of reactive N will reach 270 Tg N a − 1 by 2050 [−] . Furthermore, nitrogen oxide (NO x ) and ammonia (NH 3 ), the major pollutants contributing to N deposition, may be transformed into a variety of secondary pollutants, such as nitrate (NO 3 − ) and ammonium (NH 4 + ), of which NH 4 + -N is the predominant form and NO 3 − -N is the largest increased by the accelerating rates of N deposition in China [21] . All of them could variously affect plants' growth and alter their roots' phenotype and its physiological activity [−] . In particular, evidence suggests ammonium stimulates LR branching and nitrate simulates LR elongation [25] . Nevertheless, either dysplasia or physiological dysfunction of the root system will severely affect plant growth and development [] .
Schima superba and Pinus massoniana are the predominantly cultivated tree species in secondary subtropical forest stands and plantations in southern China. Generally, S. superba is a broadleaf tree species with a well-developed brous-root system, while P. massoniana is a coniferous tree species with a taproot system. Both trees are well known for their ability to grow and thrive under barren forest soil [−] . Previous research has shown these two species could be affected by not only P and N concentrations but also N forms, and species-speci c responses in terms of their aboveground growth state have been observed [28] . For example, P. massoniana is better suited to a low-P environment, whereas S. superba bene ts from more N addition, especially of NH 4 + . A moderate Root morphology and physiology differed signi cantly between the P and species, and the thicker roots' length, nutrient absorption, and physiological activity were signi cantly in uence by the added N ( Table 1, Fig. 1). Although the root quantity of S. superba was higher than that of P. massoniana, the former's root growth rate was lower than the latter's (  Low-P soil The root morphology and physiology variables of the two tree species were all positively stimulated by N (Fig. 1). For example, under the low-P treatment, NO 3 − could increase root growth by 157.5% and root activity by 143.6% in P. massoniana (Table 2); in S. superba, the corresponding increases were 48.7% and 106.9%. When the level of NO 3 − increased from 80 to 200 kg N ha − 1 year − 1 , the root morphology of S. superba, such as root length, surface area, and volume, continued to expand (growth ratio increased nearly 32.0%), but that of P. massoniana grew slower (growth ratio decreased nearly 31.2%). The high NO 3 − level began to have inhibitory effects on the roots of P. massoniana, especially the ne roots (Table 2). Compared with NO 3 − , an increased NH 4 + concentration could promote the root volume and quantity of S. superba, whose roots became much thicker (AD growth ratio increased from 5.5-9.5%; RL growth ratio decreased from 35.6-23.3%). By contrast, applying NH 4 + -N200 inhibited the root growth of P. massoniana, whose roots became thinner and longer (AD growth ratio decreased by 11.9% to − 10.2%; the RL growth ratio increased from 38.2-64.3%).
Although the root morphology of P. massoniana was poor under the low-P treatment, its PAE and NAE were 8.3 and 3.3 times higher than those of S. superba, respectively, and NO 3 − promoted greater root absorption of both species (Table 3, Fig. 1). NH 4 + increased the sensitivity of P. massoniana to P, which diminished its PAE and NAE. Generally, under an increased N concentration, the root activity of S. superba was reduced, but the NO 3 − increased the OX content 4.8-fold to promote root absorption, whereas the NH 4 + enhanced the secretion to promote the PAE, such as the SApase increasing 1.2-fold (Table 3).
Compared with S. superba, the NO 3 − -N200 increased the root secretion of P. massoniana, such as that of SApase increasing 1.6-fold and OX increasing 8.5fold; they were all higher than activity under the NO 3 − -N80 treatment. However, both the PAE and NAE were lower under NO 3 − -N200 than NO 3 − -N80 (Table 3).
The NH 4 + -N200 signi cantly decreased the PAE and NAE of P. massoniana. The two tree species grow better under high-P soil with less nutrient stress (Fig. 1). The 80 kg N ha − 1 year − 1 was an N level evidently favorable to S. superba, despite its root growth and secretion decreasing by 13.0% and 30.0%, respectively, since its PAE and NAE remained higher than that of N0 (Tables 2 and 3).
When the N level increased to 200 kg N ha − 1 year − 1 , the roots grew fast and their secretion was also increased.  (Table 2).
Both root growth and activity were increased by NO 3 − to a lesser extent under high-P than low-P soil. Although the NO 3 − -N200 increased the OX concentration 6.3-fold in P. massoniana, its PAE and NAE were lower and root activity weakened ( Table 3). The NH 4 + -N200 did not produce the same inhibition effect as the NO 3 − -N200 in P. massoniana, instead increasing its SApase, OX, PAE, and NAE by 1.6-, 18.4-, 1.3-, and 1.7-fold compared with N0, respectively.
The relationship between root morphology, secretion, and absorption The root PAE and NAE were always positive correlated with root morphology and secretion. However, under low-P soil and N addition treatment, the correlations of PAE and NAE with root morphological variables weakened, especially under the NH 4 + addition (Fig. 1). For example, the correlation coe cient of PAE with the root length of S. superba was r = 0.22 (p = 0.38) and r = 0.30 (p = 0.23), respectively, under the N0 and NO 3 − -N80 treatments in low-P soil, yet it was r = 0.02 (p = 0.94) under NH 4 + -N80. High-P soil signi cantly increased the correlation strength between the PAE and NAE with root morphological variables. The ne root (D ≤ 1.0 mm) portion was signi cantly correlated with PAE and NAE, especially when the N level increased to N200. Under the N80 treatment in high-P soil, the PAE and NAE generally had weak relationships with root morphology. Compared to S. superba, both PAE and NAE were much more positively correlated with root secretion of P. massoniana in low-P soil, and the moderate N addition strengthened this relationship (Fig. 1).
We used these 15 root variables in a PCA, whose results showed that PC1 mainly encompassed the root morphological characteristics, while PC2 mainly represented root absorption, with PC3 corresponding to root secretion (Fig. 2). Phosphorus and species signi cantly in uenced the relationship between root absorption and root morphology (Fig. 2B). At a higher P level, the roots grew faster and absorbed more N and P, with the root quantity higher in S. superba than in P. massoniana. However, the root absorption was stronger in P. massoniana than S. superba. The root growth and absorption of S. superba responded positively to N addition, whereas P. massoniana only performed positively under a moderate N level (N80) and N form, which depended on the P level (Fig. 2B).
The relationship between root absorption and secretion was more complex and in uenced more by N addition (Fig. 2C). At a higher P level, moderate N addition could promote root secretion and absorption in P. massoniana; however, root secretion was poor in S. superba. Although a higher N addition, especially NH 4 + , promoted root secretion in P. massoniana this did not increase its root absorption, which remained similar to that in low-P soil (Fig. 2C).

Discussion
The trees S. superba and P. massoniana are both well known for their ability to grow and thrive under adverse environmental conditions, especially low-P soil.
Research has shown that S. superba forms a brous root system and P. massoniana forms a tap root system [27][28] . Our previous study demonstrated that P.
massoniana was more suited to low-P soil than S. superba [28] . In the present study, we found the root phenotypic and physiological characteristics of P. massoniana had signi cantly lower values than those of S. superba; however, the PAE and NAE of P. massoniana signi cantly exceeded that of S. superba (Tables 2 and 3). This may be due to the ectomycorrhizal fungi and molecular mechanism of higher P e ciency in vivo of P. massoniana [−] . Work by Zhang et al. revealed four members of the Pht1 phosphate transporter protein family encoding phosphate transporters in Masson pine (PmPTs), whose up-regulated genes could improve P nutrition [31] . By contrast, S. superba survives in barren soil by increasing its root quantity and exudates to promote both its PAE and NAE [11] .
Nitrogen addition led to the coarsening of roots and a greater root surface area (Table 2). For example, compared with N0, the average AD and SA increased 4.4% and 19.6% in S. superba and 5.4% and 58.2% in P. massoniana, respectively, when N was added. The thickening and increasing surface area of roots was more pronounced in P. massoniana, driven by roots with a diameter over 1.0 mm (Fig. 3). With an increasing N level, the D = 0-0.5 mm roots grew fast in P. massoniana, and their N absorption was correspondingly higher (growth ratio of PAE, GRP P. massoniana = 80.8%, GRN S. superba = 101.8%; growth ratio of NAE, GRN P. massoniana = 71.1%, GRN S. superba = 50.7%) when compared with N0; however, the D > 0.5 mm roots grew faster in S. superba and their P absorption was higher. This pattern in the results indicated that N addition promotes ne root growth and their absorption of elements depends on tree species identity. Although the rate of root growth was not faster in S. superba than in P. massoniana, the former's nutrient absorption was increased and the N addition enhanced its P utilization [−] . The ne root fraction differed between the two species, in that for the 0-0.5 mm root, the proportion was 74.9-80.9%, with an average of 77.9% in S. superba, whereas it was 68.6-83.7%, with an average of 73.4% in P. massoniana (Table 2, Fig. 3). By contrast, for the 0-1.5 mm range roots, their proportions were stable overall (S. superba: 96.4-98.7%, 97.6%; P. massoniana: 96.6-99.3%, 97.8%). This was mainly affected by different N forms and we describe this situation using a root growth model for the two tree species under low-P soil conditions when N is added (Fig. 4). Our results suggested NH 4 + could affect the relationship between the > 1.5 mm roots and nutrient absorption in S. superba, whereas the relationship between > 0.5 mm roots and nutrient absorption was affected by NH 4 + in P. massoniana, whose height growth aboveground was much more apparent (Table 4, Fig. 4). Compared with NH 4 + , we nd that NO 3 − had an opposite effect on the relationship between ne roots and absorption, and the stem diameter growth aboveground was more notable [28] . We speculate that NH 4 + could promote thicker root activity and that NO 3 − may have stimulated the thinner root activity in the brous root plants with a larger quantity of roots, whereas NO 3 − and NH 4 + have the opposite function in a tap root system with slender roots. There was a negative relationship between the growth of roots 0-0.5 mm in diameter and both PAE and NAE after the N addition in both species. The roots consume over 50% carbon xation on root consumption, including new root growth, ion uptake and transport, and maintenance processes, for which the maintenance of respiration is linearly related to root fresh weight [] . When S. superba is growing in its preferred P and N environment, it decreases its proportion of 0-0.5 mm roots in the root system and remains in a low − consumption mode [8,11,34−] . For P. massoniana, this negative relationship was likely caused by abiotic stress and the death of the 0-0.5 mm roots ( Table 2).
The root-secreted acid phosphatase and other organic acids may promote P absorption, and they may also in uence the roots' physiological activity. Since oxalic acid is the most abundant organic acid in the two tree species, we chose to study this acid herein. N addition increased the acid phosphatase and oxalic acid secretion in P. massoniana, and this may be an important reason for why this plant's PAE was enhanced. The S. superba plants use oxalic acid for main auxiliary absorption (data not published), and only the highest N level addition was able to augment its secretion (Table 3). This result suggested that S. superba was mainly through the active absorption combined with root morphological changes.

Conclusions
Increasing the N and P concentrations in soil promotes the root growth and secretion in young P. massoniana and S. superba. The root quantity was larger in S. superba but the root growth rate and absorption were stronger in P. massoniana. The root growth and absorption of S. superba are better suited to N addition, while P. massoniana fares better under moderate N addition. Different N forms in uenced the relationships between ne root proportion and both PAE and NAE. NH 4 + promoted thicker root (> 1.5 mm) activity whereas NO 3 − stimulated thinner root (0.5-1.5 mm) activity in the brous root system plants with a larger quantity of roots. By contrast, NO 3 − and NH 4 + have opposite effects on a tap root system with slender roots. The results revealed contrasting effects of different nitrogen forms on the root growth between conifers and broad-leaved trees species, which has implications for their forest operation and management.

Experimental conditions
Seeds of S. superba and P. massoniana were collected from seedling station, who undertook the formal indentation and deposition, in Zhejiang province, China. The experiment research was complied with institutional and national guidelines. All eld studies were performed in accordance with the local legislation in China and complied with the convention on the trade in endangered species of Wild Fauna and Flora. The experiment and sampling undertaken in this study were conducted in our long-term cooperation base, which was National Masson Pine base of Laoshan Forest Farm in Zhejiang province, China.
They gave us the permissions on arranging the experiment and sampling.
The simulation experiment was conducted as described in Zhang et al., whose design consisted of two species (S. superba and P. massoniana), crossed with two levels of soil P (low-P = 1.1 and high-P = 25 mg P · kg − 1 ), three levels of N (N0 = 0, N80 = 80, and N200 = 200 kg N ha − 1 year − 1 ), and two forms of N (NH 4 + and NO 3 − ) [28] . Seeds of S. superba and P. massoniana were pregerminated on a seedbed containing normal forest soil (on 20 March 2013). About a month later, when the seedlings were 2-3 cm tall, two seedlings per container were transplanted into plastic containers (18 cm in diameter, 25 cm tall; on 20 April 2013). These seedlings grew in a greenhouse without climate control in Zhejiang Province, China, where they were watered every other day. Only one seedling was retained in each container, to ensure a one seedling/container planting density throughout the experiment. With 12 replicates per treatment combination, this experiment used a total of n = 240 individual plants (2 levels of P ⋅ 2 species ⋅ 5 tests for each N form ⋅ 12 replicates).
The added N, NH 4 Cl, and NaNO 3 , was sprayed into each treatment plot (1.1 m ⋅ 0.9 m, 24 containers per plot) once a month, for 6 months, beginning when the leaves had started to bud. All seedlings were watered with distilled water until their harvest during the rst week of November 2013. Pests and weeds were controlled manually.

Measurements
At their harvest (mid November 2013), six plants from each treatment were randomly selected and the biomass of their roots, stems, and leaves was determined [28] . The roots were carefully separated from the soil and rinsed with water for later measurements, and these seedlings from each treatment were placed in a black-out ask container with 100 ml of deionized water. After 12 h, the root exudates were collected and vacuum concentrated at 40 °C using a

Declarations
Ethics approval and consent to participate The experiment research was complied with institutional and national guidelines. All eld studies were performed in accordance with the local legislation in China and complied with the convention on the trade in endangered species of Wild Fauna and Flora.

Consent for publication
Not applicable.

Availability of data and material
Not applicable.

Competing Interests
The authors declare that they have no known competing nancial interests or personal relationships that could have appeared to in uence the work reported in this paper. The plant specimens used in our study are not an endangered species. RZ and ZZ conceived and designed the study. RZ and YW performed the experiments. RZ wrote the paper. RZ and ZZ reviewed and edited the manuscript. All authors read and approved the manuscript.