Introduction

Plants produce, store, invest, and lose carbon compounds during growth (Würth et al. 2005). The pool size of the mobile carbon, with the greatest fraction commonly present as nonstructural carbohydrates (NSC, largely starch and sugars), reflects the balance between carbon gain by photosynthesis and demand by growth and metabolism (Chapin et al. 1990; Körner 2003; Würth et al. 2005). In recent years, NSC in plant organs has been successfully used to mirror plant overall carbon supply status (Hoch et al. 2002; Hoch and Körner 2003; Körner 2003; Shi et al. 2006). Additionally, carbon allocation to growth or storage may be determined by the trade-off between growth and survival of plants (Kitajima 1994; Poorter and Kitajima 2007), therefore NSC has been used to indicate species growth strategy (Poorter and Kitajima 2007; Bansal and Germino 2008).

Various external and internal factors, including drought (Würth et al. 2005), low temperatures (Körner 2003; Hoch et al. 2003; Shi et al. 2006), atmospheric CO2 and pollutant concentrations (Körner et al. 2005), light availability (Würth et al. 1998; Niinemets et al. 2004, 2005; Poorter and Kitajima 2007), and reproduction or growth (Oleksyn et al. 2000), change the temporal or spatial patterns of NSC in plant species. It is well known that light availability largely varies with depth of canopy in a forest, and more light is available to upper than lower canopy (Niinemets and Tenhunen 1997; Tateno and Kawaguchi 2002; Prasolova et al. 2003), which leads to higher NSC in upper compared with middle and lower canopies as a result of the relative higher photosynthetic capacity in upper canopy, i.e., “more light, more NSC” (Meletiou-Christou et al. 1994; Würth et al. 1998; Herrick and Thomas 1999; Niinemets et al. 2006). However, most of these works were limited to low-elevation region; information on spatial variability of NSC in tree species under abiotic stress, such as at alpine timberline environment, is scarce, which limits the use of NSC to assess the carbon supply status or growth strategy in this area.

The Tibetan Plateau, the third pole of the world, is the ideal and an important place for exploring the structure and function of natural ecosystems and their response to climate change due to the fragility and sensitivity of ecosystems under the extreme environmental conditions driven by high-altitude climates. Alpine timberline of Sergyemla Mountain on the southeast edge of the Tibetan Plateau is one of the highest-altitude timberlines in the world, with the tree limitation near 4,500 m. Abies georgei and Juniperus saltuaria are two typical timberline species dominating in this region, and the timberline belongs to abrupt type because timberline ecotone is very narrow and the upper part of closed forest and tree limit are mostly coincident. NSC measurement of the two species will be useful to assess the carbon supply status and to indicate the balance between carbon assimilation and resumption of plants. However, no data of NSC pattern seem to be available for these two species except for a study of Shi et al. (2008), in which only altitudinal variation of mature leaf NSC was examined. To our knowledge, the current study is the first to assess the spatial pattern at the extreme alpine environment on the Tibetan Plateau. Our goal was to study how leaf NSC concentrations would vary with vertical canopy heights and leaf age in both A. georgei and J. saltuaria within the harsh timberline area, compared with plants in other environments.

Materials and methods

Study site

The study site (29°35′56″N, 94°36′38″E, 4,350 m. a.s.l.) was located on Sergyemla Mountain, located 50 km from Linzhi Town in southeast Tibet, China. The climate is characterized by high precipitation, cloudy days, cold winters, and cool summers. Mean annual temperature is −0.73°C with a minimum annual temperature of −31.6°C and maximum temperature of 24°C (Xu 1995; He et al. 2009). Average annual precipitation is 1,134.1 mm, with 75% of the annual average amount in June to September (Xu 1995; He et al. 2009). The soil is sour brown soil which is slightly acidic (mean pH 5.5). Total and available soil N are 4.5 and 0.03 mg/g, total and available soil P are 0.34 and 0.003 mg g−1, and total soil C is 63.3 mg g−1, respectively (Li et al. 2009). The study site is dominated by A. georgei and J. saltuaria, two typical timberline species, evergreen shrubs including Rhododendron pingianum and Rhododendron wardii, and deciduous shrubs including Salix oritrepha, Ribes glaciale, and Sorbus rehderiana. It is interesting to note that, although the two species grow in the same region, they are distributed on different slopes, with A. georgei dominating on northwest-facing slope and J. saltuaria on southeast-facing slope. A. georgei forest has a mean canopy height of 11.5 m, diameter at breast height (DBH) of 33.9 cm, and stem density of 459 stems/ha, whereas J. saltuaria forest has a mean canopy height of 5.9 m, DBH of 27.6 cm, and stem density of 633 stems/ha. The upper limit of A. georgei is 4,350 m, whereas it is 4,500 m for J. saltuaria. In this study, we sampled trees of A. georgei and J. saltuaria in the range of 4,300–4,350 m and 4,370–4,410 m, respectively.

Leaf sampling

Due to significant seasonal change occurring in NSC (Hoch et al. 2002; Bansal and Germino 2008), we carefully completed the leaf sampling at approximately the same time. A total of six A. georgei (aged 100–200 years) and six J. saltuaria (aged 200–300 years) trees were selected for leaf sample collection in the middle and late growing season, on 28–30 July 2005, after the completion of the current-season shoot growth. Increment cores were extracted at a height of 1.45 m in the trunk and tree age was determined by counting the annual rings in each core. Mean tree height and mean DBH for selected A. georgei were 13.2 ± 1.0 m and 33.5 ± 5.7 cm, respectively, whereas they were 6.1 ± 0.7 m and 30.7 ± 6.7 cm for J. saltuaria, respectively (data are mean ± standard deviation, SD).

At each sampling date, six needle age classes (current year, and 1-, 2-, 3-, 4-, and 5-year) were collected from the upper, middle, and lower canopy positions (13.1 ± 0.8 m, 7.8 ± 1.1 m, and 3.3 ± 0.6 m, respectively) for A. georgei by climbing trees. The apex of a long shoot produces a yearly growth increment that bears a single age class of needles. We counted needle age classes from shoot tips, which were validated by counting tree rings at the base of the shoots (Luo et al. 2005). For J. saltuaria, however, it was difficult to clearly determine leaf age, therefore we only sampled the current- and last-year leaves based on distinct differences between these two kinds of leaves. As for A. georgei, leaves were also sampled from three canopy positions (6.1 ± 0.7 m, 3.8 ± 0.5 m, and 1.1 ± 0.7 m, respectively). All needle/leaf samples were collected from the south side of trees. Each sample, composed of more than 200 needles/leaves, was immediately stored in a cool box and killed in a microwave oven (30 s at 600 W) within 6 h after collection. Thereafter, leaf samples were oven-dried to constant mass at 75°C and ground finely, and then subsamples were taken for nonstructural carbohydrate analysis.

Analysis of NSC

Nonstructural carbohydrates, defined here as the sum of free sugars (glucose, fructose, and sucrose) plus starch, were analyzed based on an enzymatic digest method with subsequent high-performance liquid chromatographic (HPLC) determination. Ground plant material was extracted with boiling, distilled water for 30 min. The soluble fraction was then analyzed for glucose, fructose, and sucrose using HPLC determination. In a second step, the insoluble material (including starch) was dialyzed at 4°C for 12 h immediately in order to remove mobile carbohydrate compounds, and then incubated for 20 h at 40°C with the crude enzyme (Clarase, a fungal amylase, Miles Laboratory Inc., Elkhart, Ind., USA) as described by Würth et al. (2005). After centrifugation, the supernatant plant extract was treated and analyzed in the same way as above, i.e., HPLC method. According to the glucose concentration, the starch concentration could be determined. Carbohydrates other than starch, sucrose, fructose, and glucose are not covered here, but they commonly contribute little to the total carbohydrate pool (Hoch and Körner 2003). Soluble sugars, starch, and total NSC were expressed on a dry-matter basis.

Statistical analysis

Two-way analysis of variance (ANOVA) with sampling canopy height and needle/leave age as two main factors was performed to test the differences in sugar, starch, and NSC concentrations between sampling heights as well as leaf ages. By using two-way analysis of variance the interaction between sampling canopy height and leaf age was also determined.

Results

NSC concentrations of A. georgei and J. saltuaria

We pooled the NSC data across sampling canopy heights and needle/leave ages to compare the NSC of A. georgei and J. saltuaria. The results indicated that there were large differences in the sugar, starch, and NSC concentrations between the two alpine timberline species growing in the same area (P < 0.001). A. georgei had obviously higher concentrations of sugar, starch, and NSC compared with J. saltuaria (7.89 ± 0.99% versus 5.25 ± 0.89% for sugars, 8.40 ± 0.74% versus 6.18 ± 0.68% for starch, and 16.29 ± 1.18% versus 11.44 ± 1.18% for NSC).

Variations of NSC concentrations along sampling heights

With the sampling height variation from upper to lower canopy, no significant difference in sugars, starch or NSC in A. georgei was found for current- or 1-, 2-, 3-, 4- or 5-year needles except a weakly significant difference of soluble sugars existing between upper and lower canopy (Tables 1, 2). Also, there was no significant variation between different canopies in current- and last-year leaves for J. saltuaria (Tables 1, 2).

Table 1 Needle/leave sugar, starch, and NSC concentrations (% dry mass) of different ages for A. georgei and J. saltuaria in relation to canopy position
Full size table
Table 2 Two-way ANOVA analyses of the effects of needle/leaf age and sampled canopy on sugars, starch, and NSC concentrations in A. georgei and J. saltuaria
Full size table

Variations of NSC concentrations along needle/leaf ages

Because no significant effect with canopy heights was found, the sugar, starch, and NSC changes along needle/leaf age were reported based on the pooled data across three canopies. Needle age had significant effect on sugar, starch, and NSC concentrations for A. georgei (Fig. 1, Table 2). With increasing needle age, sugars showed very significant increase from current-year to 2-year needles, followed by a weak but significant decline from 2- to 5-year needles (Fig. 1a). NSC had a similar pattern as sugars with needle age (Fig. 1c). For starch, 1-year needle was significantly higher than current-year needle, with a gradual decrease from 1- to 5-year needles (Fig. 1b). There were significant differences in starch and NSC between current- and last-year leaves for J. saltuaria (Fig. 2, Table 2). However, no obvious difference in sugars existed between current- and last-year leaves (Fig. 2, Table 2).

Fig. 1
figure 1

Variations of sugar (a), starch (b), and NSC concentrations (c) with needle age of A. georgei. Data are presented mean ± SD. Different letters indicate significant differences between needle age at the 0.05 level. The value for each age needle is the pooled data across three canopies due to no significant effect of sampling canopy height

Full size image
Fig. 2
figure 2

Variations of sugar, starch, and NSC concentrations in last- and current-year leaves of J. saltuaria. Data are presented mean ± SD. Different letters indicate significant differences between last- and current-year leaves at the 0.05 level. The value in this figure is the pooled data across three canopies due to no significant effect of sampling canopy height

Full size image

Discussion

Carbon status of two alpine timberline species

Compared with the needle/leaf NSC of timberline or treeline species reported by previous studies (on average, 6.13% in Hoch et al. 2002, 8–14% in Hoch and Körner 2003, and 12.5% in Piper et al. 2006), the NSC concentrations in this study are relatively higher, especially for A. georgei. Although the study area is situated at one of the highest-altitude timberline in the world and the two dominant species sampled in this study almost reach their upper limits, their needle/leaf NSC concentrations remained relatively high even under the extreme cold environment. According to the existing evidence that NSC could mirror a tree’s carbon balance, the relative abundance of NSC may indicate no carbon limitation under the present CO2 status, which is consistent with the conclusion in a study on altitudinal pattern of tree NSC in eastern Himalayas by Shi et al. (2008) and other latest studies which confirmed the growth limitation by analyzing NSC at different altitudes (Hoch et al. 2002; Hoch and Körner 2003; Piper et al. 2006; Shi et al. 2006). In addition, theoretical models suggest that carbon investment in storage (higher NSC) is an advantageous strategy in habitats with frequent stress or disturbance (Iwasa 1997; Poorter and Kitajima 2007). For plant species growing in extremely high-altitude environments characterized by frequent low temperature, allocation of carbon to storage would enhance plant endurance to harsh climate at the expense of plant growth (Kitajima 1994; Smith and Stitt 2007; Bansal and Germino 2008). The relatively higher NSC concentrations may imply a special growth strategy of these two species under extremely high-altitude timberline area compared with plant species in other environments, i.e., allocating more carbon to storage rather than using it on plant growth.

Compared with J. saltuaria distributed on the southeast-facing slope, A. georgei on northwest-facing slope had higher sugar, starch, and also NSC concentrations, which agrees with previous study on NSC measurement of Cupressacceae (Juniperus saltuaria, J. tibetica) and Pine (Abies georgei) (Shi et al. 2008). The difference between J. saltuaria and A. georgei in the present study may be related to the adaptation of the two species to different microclimatic conditions. Previous studies have verified that, in a harsh environment, plants prefer to store more carbon so as to increase their survival rather than spending it on growth (Iwasa 1997; Smith and Stitt 2007; Poorter and Kitajima 2007). Situated on the northwest-facing slope, where soil temperature is about 1°C lower than southeast-facing slope, as well as relatively lower light availability (unpublished data), A. georgei was placed in an unfavorable condition compared with J. saltuaria. In this case, carbohydrate reserves are especially important to enhance tolerance of harsh environment (e.g., low temperature or light on northwest-facing slope). However, it is very difficult to fully understand species-specific growth strategy under different slope environments only using the NSC concentrations of leaf samples. Further study should measure relative growth rate with associated leaf ecophysiological characteristics, NSC storage in other organs besides leaves, and also environmental factors at the two slopes to confirm specific growth strategy of these two species.

The canopy effect

Unexpectedly, no significant difference in NSC between sampling canopy heights for current- to 5-year needles was found in A. georgei except a weakly significant difference in sugars between upper and lower canopy. Also, there was no evident variation in NSC between canopy heights in current- and last-year leaves of J. saltuaria. Many previous studies have reported the vertical pattern of NSC along canopy heights and have confirmed “more light, more NSC” in tree species (Meletiou-Christou et al. 1994; Würth et al. 1998; Herrick and Thomas 1999; Niinemets et al. 2006). In accordance with our study, Li et al. (2001) found that neither slope exposure (south versus north) nor position within the crown (upper, middle, and lower part of crown) had effect on NSC in Pinus cembra at the alpine treeline in the Swiss Central Alps. Unfortunately, we did not measure available light for different vertical canopy heights due to the lack of efficient measuring apparatus. However, the timberlines dominated by J. saltuaria and A. georgei in this study belong to abrupt type, and even at the upper timberline, almost closed canopies are found due to the tree heights and densities. In addition, we found significant vertical variations of leaf traits (i.e., specific leaf area, N and P content per leaf area) from lower to upper canopy (Li et al. 2009), which may be determined directly by light availability (Evans and Poorter 2001; Prasolova et al. 2003). Therefore, we expect higher light availability in upper canopy, followed by middle and lower canopies. In fact, more light in upper than lower canopy has been found in many studies, especially for closed canopies (Niinemets and Tenhunen 1997; Tateno and Kawaguchi 2002; Prasolova et al. 2003).

No significant variations of NSC along canopy gradients may be as a result of different exports for canopy-dependent leaves/needles, because the leaf NSC not only reflects the net photosynthetic carbon uptake but may be related to the distribution of photosynthetic products (Chapin et al. 1990; Hoch and Körner 2003; Körner 2003; Würth et al. 2005). By measuring tissue-specific δ13C, Schleser (1992) has indicated that tree organs at lower or middle canopy height were not wholly supplied by the corresponding segment leaves at the same height, which indicates that carbohydrates produced by upper-canopy leaves may be partially transferred to lower or middle organs of trees, whereas for lower or middle leaves the transfer of carbohydrates is less. Also, Li et al. (2007) confirmed the transfer of carbohydrates from upper-canopy leaves of timberline species in this area by using stable isotope methods. The vertical pattern of NSC may be related to the extremely harsh climate at the high-altitude timberline area, in which transfer of carbohydrates from upper leaves is more than for middle or lower leaves. In addition, leaves in lower canopy could decrease carbon loss by decreasing respiration (Björkman 1981). Higher carbon transfer accompanied by higher carbon loss due to respiration for upper leaves may account for the lack of difference in NSC with canopy height. However, this is a preliminary explanation, and powerful evidence for distribution pattern of photosynthetic carbon should be verified by stable isotope labeling studies such as that described by Körner et al. (2005).

Leaf/needle effect

In general, significant effect of leaf/needle age on sugar, starch, and NSC concentrations was found in A. georgei and J. saltuaria. For A. georgei, carbohydrates showed very significant increase from current-year to 1- or 2-year needles, followed by a weak decline. Likewise, current-year leaf had lower starch and NSC compared with last-year leaf for J. saltuaria. Increases in NSC concentrations with increasing needle/leaf age (current-year versus 1- or 2-year needles for A. georgei in Fig. 1, current-year versus last-year leaves for J. saltuaria in Fig. 2) have been found in previous studies for evergreen conifers and explained by decreasing activities of enzymes responsible for carbon translocation, or decreasing ability of phloem to export photosynthates with needle aging (Hendrix and Huber 1986; Egger et al. 1996; Niinemets 1997). Similarly, Cabálková et al. (2007) found that needle NSC showed significant variations with increasing needle age from current year to 2 years old and explained the differences by the fact that older needles support the development of young shoots. However, related references are mostly focused on relatively young leaves (Hendrix and Huber 1986; Niinemets 1997; Cabálková et al. 2007). In fact, NSC concentrations did not exhibit a constantly increasing pattern with needle age. For needle age greater than 2 years in this study, NSC including starch and soluble sugars showed a gradual decrease (Fig. 2). Robakidze and Bobkova (2003) also found that 1- to 2-year-old needles have higher monosaccharides than 3- to 9-year-old needles in Siberian spruce. Generally, a balance exists between carbon acquisition in photosynthesis and carbohydrate export from leaves (Hendrix and Huber 1986). With increasing leaf age, photosynthetic capacity of leaf indeed largely declines accompanied by decrease of carbon translocation (Niinemets et al. 2004, 2005). The weak decrease of NSC with needle age from 2 to 5 years may suggest that the carbon acquisition in photosynthesis may decrease more significantly than the rate of carbohydrate export as needle age exceeds 2 years. Of course, this assumption needs to be confirmed by measuring photosynthetic characteristics and activities of enzymes responsible for carbon translocation in the future. Additionally, the accumulation of structural carbohydrates (i.e., cellulose and lignin) dilutes NSC (Li et al. 2001) because lignification of the needle hypodermis proceeds until at least the third year of development (about 2 years old) (Niinemets 1997).