Stem productivity in relation to nitrogen concentration and carbon isotopic composition (δ¹³C) in leaves of hinoki cypress (Chamaecyparis obtusa Endlicher) plantations in Shikoku district, Japan

Abstract

The stem productivity of the hinoki cypress (Chamaecyparis obtusa Endlicher) in relation to plant nitrogen status and water-use efficiency was investigated in the Okuono (OKU) and Karakawa (KRK) areas in Shikoku district, Japan, where abundant precipitation occurs. The nitrogen concentration and carbon isotopic composition (δ¹³C) in the leaves were used as indexes of plant nitrogen status and water-use efficiency, respectively. The leaf nitrogen concentration increased with decreasing soil carbon/nitrogen (C/N) ratio and with increasing soil pH. There was a marginally significant negative correlation between leaf δ¹³C and soil water content in the KRK area, but leaf δ¹³C in the OKU area did not correlate with the soil water condition, and increased on the upper slope. The results suggest that hinoki trees in the KRK area have higher water-use efficiency (high leaf δ¹³C) under lower soil water conditions. In the OKU area, meanwhile, leaf δ¹³C in the upper slope was higher due to adaptation to adverse conditions. When 12 plots in two areas were included, the mean height and stem increments increased with increasing leaf nitrogen concentration and with decreasing leaf δ¹³C. These findings suggest that nitrogen acquisition is a primary factor for stem productivity in the areas concerned but the productivity of some forests is restricted by the soil water condition or other conditions, as indicated by the high value of leaf δ¹³C. The measurement of nitrogen concentration and δ¹³C in leaves can provide us with valuable insights into the relative importance of nitrogen, water and other conditions on stem productivity in the two areas.

Keywords:

• hinoki cypress, nitrogen, soil, stem increment, water use efficiency.

Introduction

The hinoki cypress (Chamaecyparis obtusa Endlicher) is a popular coniferous plantation species in Japan, which accumulates considerable carbon in its biomass (Fang et al. 2005; Fukuda et al. 2003). Previous studies have shown that stem productivity or height growth (site index) in Japanese forests is related to topography (Enoki et al. 1996; Tateno and Takeda 2003; Zushi 2006) and soil type (Inoue et al. 1973; Mashimo 1960). These studies suggest that soil water and nitrogen availability are the primary factors that determine stem productivity.

The utilization of water resources can be evaluated by the leaf carbon isotopic composition (δ¹³C). This is widely used as an index of water-use efficiency because the leaf carbon (C) isotopic composition is a function of both the supply of carbon dioxide (CO₂) to the C fixation sites and photosynthetic capacity (chloroplast demand for CO₂) (Farquhar et al. 1982). Higher δ¹³C in leaves would indicate higher water-use efficiency (higher ratio of assimilation to leaf conductance). Previous studies have shown that plants growing on ridges have a higher δ¹³C than those in valleys (Garten and Taylor 1992; Hanba et al. 2000). These studies suggest that δ¹³C in the leaves is a useful indicator of water limitation by plants. However, variations in leaf δ¹³C have not been determined for hinoki cypress plantations.

Nitrogen is one of the resources that limit the productivity of temperate forests (Vitousek and Howarth 1991). Previous studies have shown diverse relationships between soil nitrogen availability and stem productivity: positive (Pastor et al. 1984; Reich et al. 1997), no significant relationship (Binkley et al. 2003; Zushi 2003), and high stem productivity despite low soil nitrogen availability (Joshi et al. 2006; Inagaki et al. 2004, 2008). The reason for the absence of a clear relationship between soil nitrogen availability and stem production is not known, but the difference in soil water conditions may affect the nitrogen-productivity relationship. If trees are limited by the soil water condition, increased soil nitrogen availability should have less effect on stem productivity. Therefore, the relative importance of water and nitrogen resources to stem productivity should be evaluated, but information for hinoki cypress plantations is scarce.

In this study, we investigated the stem productivity and utilization of nitrogen and water resources by hinoki cypress plantations at both high (900–950 m) and low (320–370 m) altitudes in Shikoku district. These areas receive high annual precipitation (>2900 mm). Nitrogen concentration and δ¹³C in the leaves were used as indexes of the plant nitrogen status and water-use efficiency, respectively. The objectives of this study were to determine: (1) whether leaf nitrogen and δ¹³C are related to soil conditions, and (2) whether stem productivity is related with leaf nitrogen concentration and δ¹³C. From these relationships, we discuss the applicability of δ¹³C of hinoki leaves in areas of high precipitation.

Materials and Methods

Study site

The study was conducted in two areas at different altitudes in Kochi Prefecture, Shikoku district, southern Japan. The Okuono area (OKU) is located at a higher altitude (N33° 41', E133° 15', 900–950 m ASL), and the mean annual temperature and precipitation are 9.8°C and 2948 mm, respectively, obtained using 1-km mesh data from the Digital National Land Information database of the Ministry of Land, Infrastructure and Transport, Japan. The Karakawa area (KRK) is located at a lower altitude (N32° 50', E132° 52', 320–370 m ASL), with mean annual temperature and precipitation of 14.1°C and 2964 mm, respectively. Although the mean annual temperature is 4.3°C higher at the lower altitude, there is abundant precipitation at both altitudes. The soil in these stands was clayey Dystrudept (Soil Survey Staff 2010) derived from schist in the OKU and KRK areas respectively. At both altitudes, there were even-aged hinoki cypress plantations, and within each area, six study plots (20 × 20 m) were established (Fig. 1 and Table 1). The plot number increased from higher to lower altitude. In the OKU area, plots OKU1 and OKU2 were located on the upper slope while the OKU6 plot was located on a convex relief position at a lower altitude (Table 1) and the remaining three plots (OKU3-5) were located on the middle slope. According to the classification of forest soil in Japan (Forest Soil Division 1976), the soil types in OKU1 and OKU6 were a drier subtype of moderately moist brown forest soil (B_D(d) type) while the other four plots were moderately moist brown forest soil (B_D type). In the KRK area, a total of six plots were selected on the upper (KRK1), middle (KRK 2–4), and lower (KRK5–6) slopes. The soil type on the upper slope was classified as B_D(d) while the middle and lower slope plots were B_D.

Figure 1. Location of study areas. Six plots (20 × 20 m) were established in each area.

Table 1. Topography and soil properties in study plots

Plot	Topography	Soil Type*
OKU (high altitude)
OKU1	Upper slope	BD(d)
OKU2	Upper slope	BD
OKU3	Middle slope	BD
OKU4	Middle slope	BD
OKU5	Middle slope	BD
OKU6	Convex relief	BD(d)
KRK (low altitude)
KRK1	Upper slope	BD(d)
KRK2	Middle slope	BD
KRK3	Middle slope	BD
KRK4	Middle slope	BD
KRK5	Lower slope	BD
KRK6	Lower slope	BD
*Forest soil division (1976). OKU, Okuono area; KRK, Karakawa area.

Soil properties

In each area, samples of the surface soil were taken at a depth of 0–5 cm (in which fine roots were distributed densely) from five randomly selected sites within the study plot from July to August 2007. A 100-ml soil core was collected at a depth of 0–5 cm and surface soil at the same depth was collected for soil solution measurement. Five soil samples were combined into a single sample. The weight of the field-moist sample was measured, and the sample was dried at 105°C for 48 h. The volumetric water content was calculated as the weight of water in the soil (the difference between the field-moist and dried sample) divided by the volume of the soil core. Although the soil water condition was measured once in summer, we confirmed that the difference of soil water content between plots can be used as a representative value because the seasonal and yearly fluctuation of soil water content was sufficiently minor (Miyamoto et al. unpublished data; Shinomiya et al. 2006).

Soil samples were passed through a 2-mm sieve and the sieved soil was analyzed for carbon and nitrogen concentrations, measured using an NC analyzer (NC-22F, Sumika Chemical Analysis Service, Osaka, Japan). The field-moist samples were used for pH analysis. The soil pH was determined for a soil-water suspension (1 : 2.5 w/w) using a pH meter (TOA pH meter HM35V, Tokyo, Japan).

Leaf properties

Fresh leaves were collected via a slingshot in July 2007. Five trees were randomly selected per plot and three samples per individual were collected from the upper part of the crown. The three fresh-leaf samples collected from an individual tree were then combined into a single sample. The samples were dried for 48 h at 75°C and analyzed for nitrogen content and δ¹³C. The nitrogen concentration in the leaf samples was measured using an NC analyzer as described above, while δ¹³C in the leaves was measured using an on-line elemental analyzer (NC 2500; CE instruments, Milan, Italy) coupled with an isotope ratio mass spectrometer (MAT252; Thermo Electron, Bremen, Germany). The results are expressed as δ values, defined as ‰ deviation from standard reference materials:

where R = ¹³C/¹²C and values for standards are ¹³C of Pee Dee belemnite (PDB) = 0 by definition. DL-α-alanine (δ¹³C = − 23.5‰) was used as the running standard. The standard deviation of replicated measurements was less than 0.10‰.

Tree census

Hinoki trees were planted after clear-felling of previous hinoki in 1977 and 1974 in the OKU and KRK areas respectively, with initial tree density of approximately 3000 trees ha⁻¹. The area was weeded several times. In the OKU area, low intensity thinning was conducted in 1997, although not in the KRK area. The diameter at breast height (DBH) and tree height were measured in September and March 2007 in the OKU and KRK areas respectively, on trees aged 30 and 32 years old. Stem volume was estimated from the height and DBH of each tree by an equation (Forestry Agency 1970). Conversion to stem biomass was performed by multiplying the stem volume by the average wood density of hinoki (0.45 Mg m⁻³) (Inagaki et al. 2009). The mean annual increment of tree height, DBH and stem biomass was calculated as division by forest age due to the difference in age between the OKU and KRK areas.

Statistical analysis

The difference in the measured properties of the two areas was analyzed by t-test, while Spearman rank correlation coefficients were calculated for the soil and plant properties and the stem growth properties were predicted by nitrogen concentration and δ¹³C in leaves using multiple regression analysis. All statistical analyses were performed with JMP software (Ver. 5.01; SAS Institute 2002).

Results

Soil properties

The soil in the OKU area had higher total carbon and nitrogen concentrations and lower pH compared to that in the KRK area (p < 0.05, Table 2). The soil C/N ratio did not differ between the two areas (p > 0.05). Soil C/N ratio in the OKU area was high in the convex relief plot (OKU6, Fig. 2), while that in the KRK area decreased from the upper to lower slope (Fig. 2). Volumetric water content in the OKU area was lower in the convex relief (OKU6) plot than in the other five (Fig. 2). In the KRK area, volumetric water content was relatively constant in all plots, but slightly higher in plots KRK2 (middle slope) and KRK5 (lower slope, Fig. 2). The volumetric water content was not clearly related to the soil type at either altitude (Table 1, Fig. 2).

Figure 2. Soil and leaf properties in the Okuono area (OKU) and Karakawa area (KRK). Plot numbers are shown in Fig. 1. For leaf properties, means with standard deviation are shown. C, carbon; N, nitrogen; WC, water content; δ¹³C, carbon isotopic composition.

Table 2. Properties of surface soil (0–5 cm) in hinoki cypress plantations

		OKU		KRK		t-test
Properties		Mean	SD	Mean	SD	p
Vol. water content	m^3 m^−3	0.335	0.039	0.209	0.018	0.001
Soil pH		4.43	0.22	4.84	0.12	0.002
Total C concentration	g kg^−1	143.5	38.0	100.7	21.5	0.041
Total N concentration	g kg^−1	8.0	1.3	6.0	1.4	0.034
C/N ratio		17.7	2.0	16.9	1.7	0.36
C, carbon; N, nitrogen; OKU, Okuono area; KRK, Karakawa area; SD, standard deviation.

Leaf properties

The stand characteristics and leaf properties are shown in Table 3 and Fig. 2, respectively. The leaf nitrogen concentration did not differ significantly between the two areas (t-test, p > 0.05), but in the KRK area, it increased from the upper to lower slope. The leaf nitrogen concentration in the OKU area was higher at the lower slope (OKU4 plot) and lower at the convex relief position at the lower altitude (OKU6 plot). Leaf nitrogen concentration increased with increasing soil pH (Table 4) and with decreasing soil C/N ratio (Fig. 3a). Leaf δ¹³C did not differ between the two areas (t-test, p > 0.05). In the KRK area, leaf δ¹³C was lower on the middle slope (KRK2 plot) than in the other plots, and leaf δ¹³C in the OKU area was higher on the upper slope (Fig. 2). The relationship between soil water content and δ¹³C in leaves in the two areas showed different trends (Fig. 3b). In the KRK area, there was a marginally significant negative correlation between leaf δ¹³C and soil water content (Spearman rank correlation, r_s  = −0.795, p = 0.06), but leaf δ¹³C in the OKU area was not related to soil water content (r_s  = 0.429, p = 0.41). There was no significant correlation between nitrogen concentration and δ¹³C in leaves (r_s  = −0.410, p = 0.19).

Figure 3. Relationship between the soil carbon/nitrogen (C/N) ratio and leaf nitrogen (N) concentration (a) and between the soil water content and leaf carbon isotopic composition (δ¹³C) (b) OKU, Okuono area; KRK, Karakawa area.

Table 3. Stand characteristics of hinoki cypress forests

		OKU		KRK		t-test
		Mean	SD	Mean	SD	p
Stand density	n ha^−1	2408	197	2233	523	0.93
Mean DBH	cm	15.5	1.4	17.0	2.1	0.58
Mean height	m	11.2	1.3	13.3	1.2	0.21
Stem biomass	Mg ha^−1	120	27	151	19	0.14
Mean DBH increment	cm yr^−1	0.37	0.04	0.42	0.04	0.89
Mean height increment	m yr^−1	0.52	0.05	0.53	0.07	0.52
Mean stem increment	Mg ha^−1yr^−1	4.0	0.9	4.7	0.6	0.30
DBH, diameter at breast height; OKU, Okuono area; KRK, Karakawa area; SD, standard deviation.

Table 4. Spearman rank correlation coefficient between leaf and soil properties

	Leaf δ^13C	Leaf N	Soil pH	Soil C	Soil N	Soil C/N
OKU area (n = 6)
Leaf N concentration	0.54
Soil pH	0.17	0.38
Soil C concentration	0.26	−0.43	−0.78†
Soil N concentration	0.26	−0.43	−0.78†	1.00**
Soil C/N ratio	0.37	−0.20	−0.70	0.94**	0.94**
Soil vol. water content	0.43	0.71	0.06	−0.14	−0.14	−0.09
KRK area (n = 6)
Leaf N concentration	0.41
Soil pH	−0.52	−0.12
Soil C concentration	0.12	0.31	−0.12
Soil N concentration	0.24	0.70	−0.13	0.90**
Soil C/N ratio	−0.32	−0.94**	0.03	−0.26	−0.64
Soil vol. water content	−0.79†	0.20	0.46	0.09	0.20	−0.31
All sites (n = 12)
Leaf N concentration	0.41
Soil pH	0.04	0.66*
Soil C concentration	0.05	−0.41	−0.74**
Soil N concentration	0.07	−0.33	−0.74**	0.99**
Soil C/N ratio	−0.05	−0.64*	−0.36	0.34	0.26
Soil vol. water content	−0.27	−0.38	−0.70**	0.62*	0.63*	0.10
**p < 0.01, *p < 0.05, †p < 0.10. C, carbon; N, nitrogen; OKU, Okuono area; KRK, Karakawa area.

Prediction of stem productivity by leaf properties

The mean annual increment of height, DBH, and stem volume did not differ for the OKU and KRK areas (t-test, p > 0.05, Table 3) and all the data were grouped for regression analysis. The mean height increment increased with increasing leaf nitrogen concentration (p = 0.001, Table 5). Multiple regression analysis showed increased mean height increment with increasing leaf nitrogen concentration (p = 0.001) and decreasing leaf δ¹³C (p = 0.05, Fig. 4a). The mean DBH increment increased with increasing leaf nitrogen concentration (p = 0.03), but the effect of leaf δ¹³C was not significant in multiple regression analysis (p > 0.05). Mean stem increment increased with increasing leaf nitrogen concentration (p = 0.019, Table 5). In multiple regression analysis, mean stem increment increased with increasing leaf nitrogen concentration (p = 0.003) and with decreasing leaf δ¹³C (p = 0.02, Table 5, Fig. 4b).

Figure 4. Measured vs. predicted mean height and stem increment using multiple regression relationships: mean height increment = −0.73 −0.031 (leaf carbon isotopic composition (δ¹³C)) + 0.025 (leaf nitrogen (N)); mean stem increment = −26.9 − 1.00 (leaf δ¹³C) + 0.36 (leaf N). OKU, Okuono area; KRK, Karakawa area.

Table 5. Results of multiple regression analysis. Standard partial correlation coefficients of each factor (β) and adjusted coefficient of determination of the model are shown

	Leaf N		Leaf δ^13C		Model
	β	p	β	p	R2adj	p
Mean height increment
	0.895	0.001	–	–	0.762	0.001
	–	–	−0.106	0.74	−0.088	0.74
	0.938	0.001	−0.291	0.05	0.828	0.001
Mean DBH increment
	0.633	0.03	–	–	0.340	0.03
	–	–	−0.114	0.73	−0.086	0.73
	0.680	0.02	−0.246	0.35	0.338	0.06
Mean stem increment
	0.665	0.019	–	–	0.386	0.019
	–	–	−0.372	0.23	0.052	0.23
	0.764	0.003	−0.521	0.02	0.633	0.005
N, nitrogen; δ^13C, carbon isotopic composition; DBH, diameter at breast height.

Discussion

Soil properties

The soil carbon concentration was greater in the OKU area (higher altitude) than in the KRK area (lower altitude) (Table 2), which was primarily attributable to the difference in mean annual temperature between the areas; organic matter decomposes more rapidly in the KRK area and at a higher mean annual temperature. The soil carbon/nitrogen (C/N) ratio was higher on the upper slope in the KRK area and convex relief at the lower altitude in the OKU area (Fig. 2). Previous studies showed that the soil C/N ratio decreased from the upper to lower slope (Enoki et al. 1996; Tateno and Takeda 2003) and a higher soil C/N ratio was found on the ridge (Enoki et al. 1996). The high soil C/N ratio in the OKU6 indicates the soil condition of the site is dry enough even at the lower altitude, indicating that the shape of the relief, i.e., concave or convex, was more important than the altitude along the slope. The soil water content was higher in the OKU area than in the KRK area (Table 2). Because rainfall was high in both areas, lower evapotranspiration at lower temperatures would be a primary reason for higher water content in the OKU area.

Leaf properties

The nitrogen (N) concentration in the leaves was related to soil pH and soil C/N ratio (Table 4) and was lower on the upper slopes in the KRK area and on the convex relief in the OKU area (Fig. 2). Several studies have shown a similar pattern along a topographic gradient (Nakanishi et al. 2009; Tateno and Takeda 2010). These findings suggest that leaf N concentration can be determined by soil chemical properties such as soil pH and C/N ratio in the site but vary with micro-topographic relief within the area.

In the KRK area, there was a marginally significant negative correlation between leaf δ¹³C and soil water content (Fig. 3b). The result suggests that hinoki trees have higher water-use efficiency (higher leaf δ¹³C) under limited water conditions. This pattern was consistent with the findings of previous studies (Garten and Taylor 1992; Hanba et al. 2000). However the effect of soil water conditions on water-use efficiency was very weak because the KRK area received high annual precipitation (> 2900 mm) and soil water conditions were modest. By contrast leaf δ¹³C in the OKU area was not related to soil water content (Fig. 3b). Because the soil water content in the OKU area was relatively high, soil water conditions should be sufficient for hinoki trees and the variation of soil water gradient was narrow due to the size of the study area concerned. The variation of leaf δ¹³C in the OKU area might also be caused by the presence of adverse conditions for growth other than soil water limitations.

Previous studies have reported that leaf δ¹³C is positively correlated with altitude (Cordell et al. 1998; Hultine and Marshall 2000; Körner et al. 1991; Sparks and Ehleringer 1997). There are several mechanisms to explain the observed pattern, but all found a close correlation between leaf thickness and leaf δ¹³C. Therefore the morphological adaptation of the leaves is one possible reason for higher leaf δ¹³C on the upper slopes in this study although no morphological properties were measured in this study. We conclude that the control factor of leaf δ¹³C should differ between the two areas: in the KRK area, leaf δ¹³C is marginally related to soil water content, while in the OKU area, leaf δ¹³C was related to slope position rather than soil water content.

Stem growth and resource utilization

The mean height, DBH, and stem increment were positively correlated to leaf nitrogen concentration (Table 5). These results suggest that the uptake of nitrogen indicated by the leaf nitrogen concentration was a primary factor in determining the stem productivity in the two study areas. In addition, the utilization of water resources indicated by leaf δ¹³C was also a significant factor in determining height and stem increments (Table 5, Fig. 4). These findings indicate that information on leaf nitrogen concentration and δ¹³C can provide valuable insights into the relative importance of soil nitrogen, soil water and other conditions on stem productivity in the two areas. The results also suggest that leaf δ¹³C is a useful indicator of stem productivity as well as leaf nitrogen concentration, even in areas of high precipitation. Future studies that measure leaf δ¹³C as well as nitrogen concentration from a broad range of environmental conditions will contribute to the understanding of stem productivity in relation to nitrogen and water utilization.

Acknowledgments

We are grateful to the staff of Shikoku Research Center, Forestry and Forest Products Research Institute for their valuable assistance with this study. We also thank the Shikoku Regional Forest Office, Forest Agency of Japan for permission to study in the national forest. We thank Dr Akira Kagawa in Forestry and Forest Products Research Institute for his support with isotope analysis. This study was financially supported in part by research grant #200701 of Forestry and Forest Products Research Institute of Japan.