Effects of light acclimation on shoot morphology, structure, and biomass allocation of two Taxus species in southwestern China

Acclimation to changing light conditions plays a crucial role in determining the competitive capability of tree species. There is currently limited information about acclimation to natural light gradient and its effect on shoot structure and biomass in Taxus species. We examined the acclimation of the leaf and shoot axis morphology, structure and biomass allocation of Taxus yunnanensis and T. chinensis var. mairei under three different natural light environments, full daylight, 40–60% full daylight and <10% full daylight. The leaf biomass, nitrogen content per unit area, leaf carbon content per dry mass and leaf dry mass to fresh mass ratio increased with light in both species, demonstrating an enhanced investment of photosynthetic biomass and structural investment under high light. The number of leaves per unit shoot axis length and the leaf dry mass per unit shoot axis length increased with light in both species. However, the light increase did not result in the increase of the total shoot mass. T. yunnanensis produced larger leaves under low light and a higher shoot axis length per unit dry mass under high light, whereas the leaf size and biomass yield of T. chinensis var. mairei were not sensitive to light.

plants enhance the interception of light with a large leaf area per unit leaf biomass 3 and shade leaves enhance the potential relative growth rates through little physiological activity (slow respiration and light-saturated photosynthetic rates). Moreover, shoots under high irradiance conditions have a circular cross section with tightly packed leaves 18,19 , thus maximizing foliar biomass investments per unit shoot length and taking advantage of the increased irradiance for photosynthesis 13 . A quantitative understanding of biomass allocation patterns is of fundamental importance to plant ecology and evolution and is useful for agricultural and forest practices and implementation 20 .
Taxus species are well-known for their high content of taxol, which exhibits bioactivities against certain tumors 21,22 . Although most Taxus species are protected plants in many areas and many studies have been conducted on their pollination biology 23 , genetics 24 and taxol production 25 , few studies have characterized the changes exhibited by Taxus species under different light conditions 26,27 , and little research has been conducted on their shoot biomass allocation along a natural light gradient and the effects of light on the shoot structure of Taxus species are not well addressed 22 . In the present study, the acclimation of shoot structure and biomass allocation in response to a natural light gradient for two Taxus species, Taxus yunnanensis and T. chinensis var. mairei, was investigated to determine the contributions of leaf-and shoot-level adjustments in structure and biomass allocation, and to characterize the potential role of shoot structure on species competitiveness. Specifically, this study addressed (1) how shoot morphology and structure are acclimated to light environments and (2) how shoot biomass is allocated under different light environments.

Results
Leaf size and morphology along a natural light gradient. Individual leaf area (A L ) decreased with increasing light levels for T. yunnanensis but not for T. chinensis var. mairei (Fig. 1A). Leaf roundness (Fig. 1B) and leaf biomass per unit area (LMA, Fig. 1C) increased with increasing light levels in both T. yunnanensis and T. chinensis var. mairei.
T. yunnanensis had a higher A L in low light and LMA in middle and high light compared with T. chinensis var. mairei (P < 0.05). Leaf roundness did not differ between T. yunnanensis and T. chinensis var. mairei under all light conditions (P > 0.05).
Changes in foliar nitrogen and structural investments along a natural light gradient. The leaf nitrogen content per unit area (N area , Fig. 2A), leaf carbon content per unit dry mass (Fig. 2C) and leaf dry mass to fresh mass ratio (Fig. 2D) increased with light in both species. Overall, the total leaf carbon content per unit dry mass and the dry to fresh mass ratio were strongly correlated with the light level, thus suggesting that both variables reflect changes in leaf structural investments. In contrast with N area , the leaf nitrogen content per unit dry mass (N mass ) decreased with increasing light in T. yunnanensis and T. chinensis var. mairei (Fig. 2B).
T. yunnanensis had lower N mass under high light and higher leaf carbon content per unit dry mass under all light conditions compared with T. chinensis var. mairei (P < 0.05). However, there were no significant differences in N area and the dry to fresh mass ratio between T. yunnanensis and T. chinensis var. mairei under all light conditions (P > 0.05). Changes in shoot axis morphology and biomass along a natural light gradient. Shoot axis dry mass decreased with light in T. yunnanensis (Fig. 3A), whereas shoot axis length per unit dry mass increased with light (Fig. 3D). Shoot axis dry mass and shoot axis length per unit dry mass did not differ among different natural light conditions in T. chinensis var. mairei (P > 0.05) (Fig. 3A,D). Shoot axis length decreased with light in both T. yunnanensis and T. chinensis var. mairei (Fig. 3B). Shoot axis diameter was not sensitive to light in either T. yunnanensis or T. chinensis var. mairei (Fig. 3C).  Only under high light did a significant difference exist between T. yunnanensis and T. chinensis var. mairei in shoot axis dry mass and shoot axis length (P < 0.05). The shoot axis diameter and shoot axis length per unit dry mass did not differ between T. yunnanensis and T. chinensis var. mairei under any of the three light conditions (P > 0.05).
Biomass allocation between leaf and shoot axis along a natural light gradient. The total shoot mass was not associated with light in either species (Fig. 1D). However, the number of leaves per unit shoot axis length and leaf dry mass per unit shoot axis length increased with increasing light in both T. yunnanensis and T. chinensis var. mairei ( Table 1). The leaf dry mass per unit shoot axis dry mass, the number of leaves per unit shoot axis dry mass and leaf dry mass per unit total shoot mass also increased with light in T. yunnanensis, but not in T. chinensis var. mairei ( Table 1). The leaf area per unit total shoot mass did not differ among the three light conditions in T. chinensis var. mairei (P > 0.05). T. yunnanensis had the highest leaf area per unit total shoot mass under high light (P < 0.05).
The total shoot mass and leaf dry mass per unit shoot axis length did not differ between T. yunnanensis and T. chinensis var. mairei under any of the three light conditions (P > 0.05), and there were no significant differences between T. yunnanensis and T. chinensis var. mairei in the number of leaves per unit shoot axis length, leaf dry mass per unit shoot axis dry mass, number of leaves per unit shoot axis dry mass, leaf dry mass per unit total shoot mass or leaf area per unit total shoot mass, except for under high light. T. yunnanensis had a higher number of leaves per unit shoot axis length, leaf dry mass per unit shoot axis dry mass, number of leaves per unit shoot axis dry mass, leaf dry mass per unit total shoot mass and leaf area per unit total shoot mass compared with T. chinensis var. mairei under high light (P < 0.05).

Discussion
Adjustment of leaf structure and chemistry to a natural light gradient. In plants, phenotypic plasticity may be expressed at various levels, such as the shoot or leaf level; as a result, in trees, the morphological and physiological properties of leaves dramatically change in response to light gradients 18,28,29 . In the present study, smaller leaves were observed under T. yunnanensis in high light and longer leaves were observed for both Taxus species under high light. Decreases in the mean area of individual leaves have been frequently observed as light increases 13,30,31,32 but not consistently 33 . Smaller leaves improve the coupling between the leaves and the atmosphere, thereby reducing leaf temperature and transpiration 13 . Leaf coupling with the atmosphere is further enhanced through increased leaf elongation, thus decreasing the effective leaf size for a common leaf area 13 . In addition, smaller leaves increase the homogeneity of the radiation field and enhance the penetration of the canopy through penumbral radiation 34 . Although the leaves became smaller under high light, the number of leaves per unit shoot axis length was positively correlated with light for the two species (Table 1).
LMA is an important index of leaf structure that is highly correlated with the light environment and is closely associated with photosynthesis 35 . The results of the present study showed a significant increase in LMA with increasing light in the two species, similarly to findings in other studies 1,13,27,35,36 . A greater variation in LMA suggests that LMA is an important factor in the acclimation of these two species to various light conditions. Higher LMA suggests that there is more tissue in which photosynthetic reactions can occur when the light strikes the leaves. Lower and higher LMA leaves were able to use light efficiently in high light 37 . Smaller leaves have fewer cells with thicker cell walls and show lower quantum efficiency 38 ; therefore, smaller leaves may reduce heat loading. Leaves become smaller as light levels increase, thus leading to increased LMA 37 . A lower LMA in lower-light situations has been proposed to be an important adaptive characteristic allowing species to construct a larger foliar area with a given fraction of plant mass in the leaves 39 . In a simulation study, Sims et al. 40 have shown that growth is highly dependent on adjustments in LMA to low and high light. The needle N area increased with increased light for both Taxus species in the present study, reflecting the much higher variability in LMA than N mass . The close relationship between leaf N area and light level has previously been reported 13,41,42,43 and reflects the increase in the carboxylating capacity of sun-adapted leaves mediated through an increase in the Rubisco-associated nitrogen pool 41 . In contrast, the needle N mass decreased with increasing light for both Taxus species. Several studies have shown that the N mass either increases or decreases with light within tree or forest canopies. For example, the leaf or needle nitrogen concentration on a dry weight basis has been positively correlated with relative irradiance in several species, such as Eucalyptus grandis 44 and Betula pendula 43 . In contrast, the N mass decreases with increasing leaf or needle irradiance for the shrubs Corylus avellana and Lonicera xylosteum 43 , Fagus sylvatica 45 , Acer platanoides and Padus avium 46 , and Picea abies 1,47 , and this pattern reflects the different light demands of these species 42,43 . Thus, the results of the present study are consistent with the conclusion that N mass decreases with increasing leaf or needle irradiance in shade-intolerant species, such Picea abies. Moreover, many physiological processes, such as photosynthesis and growth, are closely correlated with N mass , because photosynthate production and translocation exhibit different time lags.
Leaf carbon content per dry mass and leaf dry mass to fresh mass ratio were also positively correlated with light, suggesting greater structural investments under high light. Greater structural investments in leaves acclimated to higher light enhance leaf rigidity and may improve leaf resistance to low leaf water potentials 48 , but these investments might also reduce leaf photosynthetic capacity per unit dry mass 49 . Under low light, the moisture conditions and lower levels of evaporation facilitate high leaf water content, thus leading to a lower leaf dry mass to fresh mass ratio, and maintenance of turgor, and providing a low-resource method to enhance leaf display for light capture. Leaf carbon content per dry mass and leaf dry mass to fresh mass ratio increased as the light intensity increased, leading to increased leaf mass. The variation in leaf carbon content per dry mass and leaf dry mass to fresh mass ratio under different light conditions may contribute to functional and ecological segregation 38 .

Acclimation of shoot axis morphology to natural light gradient.
In the present study, the shoot axis length decreased with increasing light. Küppers, using hedgerow species, has elegantly demonstrated differences between tree species in shoot growth in response to a heterogeneous light environment 50 . The study has also shown that sun-shoots have shorter shoot axis lengths and are less efficient in expansion into potentially unoccupied space and prevention of self-shading compared with shade-shoots. These findings are consistent with a typical foraging mechanism. According to the foraging theory, the shoot axis should be shorter, and branching frequency should be higher in branches exposed to higher resource availability. On the one hand, species reduce the shoot length under high light. In this way, smaller masses of woody support tissue are invested in the production of unit leaf tissue. The advantage of a smaller support mass is that more biomass is available for new leaf growth. This shift in biomass allocation from stem to leaf mass improves competitive capability. At the same time, relatively long shoots develop under low light, thus enabling expansion into potentially unoccupied spaces and preventing self-shading. On the other hand, species also occupy more space and efficiently intercept more light through increased branching frequency. This phenomenon reflects a capability to sense neighbors through phytochrome-based mechanisms 12 .
In contrast to other plant resources, light availability depends not only on leaf area but also on the shoot axis length. Longer shoots occupy more space and intercept more light under low light. Some studies have shown that maximum shoot growth occurs in shady conditions 51 . The opposite growth pattern has also been observed in some studies, showing that shoots are often shorter under lower light 52,53 . The adjustments of shoot axis length provide an explanation for the greater efficiency of light interception in plants grown in low light relative to open fields 54 .
Unlike the shoot axis length, the shoot axis diameter was not sensitive to light in the present study. There were no significant differences among three light environments for either species. The responses of shoot axis diameter to a heterogeneous light environment indicated that diameter growth is not a function of the acclimation of shoot axis morphology. Increased diameter growth under high light reflected only an enhanced investment of shoot axis biomass, which in turn reduced the investment of leaf biomass under steady total shoot mass and weakened photosynthesis and competitive capacity.

Acclimation of biomass allocation to a natural light gradient.
Decreased light availability changed the allocation to shoot-level biomass of Taxus species. Although total shoot mass was not associated with light in T. yunnanensis and T. chinensis var. mairei, the number of leaves per unit shoot axis length and leaf dry mass per unit shoot axis length increased with increasing light in both species, suggesting that more biomass was allocated to the leaves. In high-light environments, light dispersed over more leaves is used more efficiently in photosynthesis and less is lost to saturation 55 , whereas in low-light environments, light concentrated on a few leaves is efficiently intercepted at a minimal cost in leaf construction and maintenance 56 .
In contrast, the changes in leaf area per unit total shoot mass did not show a clear trend for either species, although T. yunnanensis had the highest leaf area per unit total shoot mass under high light. However, the leaf dry mass per unit shoot axis dry mass, number of leaves per unit shoot axis dry mass and leaf dry mass per unit total shoot mass were positively correlated with light in T. yunnanensis but not in T. chinensis var. mairei. These results showed that less biomass was invested in woody support tissue in T. yunnanensis. In this way, more new production of a unit leaf tissue can be produced at the risk of an increased self-shading of the crown. This shift in biomass allocation from shoot axis to leaf mass is also reflected in the high photosynthesizing tissue of understory saplings 3 .

Species differences in plastic adjustment to a natural light gradient.
Although the species examined in the present study are within the same genera, significant differences existed in shoot structure and biomass allocation adjustments to natural light gradients. T. yunnanensis had larger leaves under low light, greater LMA under moderate and high levels of light, higher leaf carbon content per unit dry mass under all light conditions, and lower leaf N mass under high light compared with T. chinensis var. mairei. All the above-mentioned leaf characteristics reflect changes in light interception efficiency, investment photosynthetic biomass, and leaf structural investments. Variations in leaf ecophysiology among species in different taxa have been studied to understand acclimation to light 13,32,57 , but few studies have addressed the leaf changes within the same genera 28 . The results of the present study demonstrated that even within the same genera, species exhibit distinct ecophysiological characteristics to acclimate to light. Larger leaves under low light suggest that more diffuse or direct light can be intercepted. Greater LMA under high light suggests that when this light strikes sun leaves, there will be more tissue in which photosynthetic reactions can occur. Accordingly, a higher leaf carbon content per unit dry mass shows that T. yunnanensis had greater structural investments under high light.
Except for leaf morphological and physiological adjustments, T. yunnanensis exhibited a lower shoot axis dry mass and shoot axis length, with a higher number of leaves per unit shoot axis length, leaf dry mass per unit shoot axis dry mass, number of leaves per unit shoot axis dry mass, leaf dry mass per unit total shoot mass and leaf area per unit total shoot mass compared with T. chinensis var. mairei under high light. These characteristics suggest that more biomass is allocated to the leaves in T. yunnanensis. Higher leaf biomass increased carbon fixation and biomass growth and potential relative growth rates. Differences in leaf and shoot axis morphology and biomass allocation may substantially modify the competitive potential of a species. The results of the present study indicate that T. yunnanensis is more competitive than T. chinensis var. mairei in the same light environment.

Conclusions
Positive correlations among leaf biomass, nitrogen content per unit area and light indicated an enhanced investment of photosynthetic biomass under high light and a greater leaf-level light-harvesting efficiency. The leaf carbon content per dry mass and leaf dry mass to fresh mass ratio of all species were also correlated positively with light, thus suggesting a greater structural investment under high light. Although increased light did not result in an increase in the total shoot mass, the shoot axis dry mass decreased with increasing light levels and enhanced the biomass yield in leaves and the investment of photosynthetic biomass. These structural acclimations showed that increased light resulted in the increase of the investment in photosynthetic biomass.
Species differences in structural acclimation demonstrated different foliar investment of photosynthetic biomass. Larger leaves resulted in more efficient harvesting of diffuse irradiance under low light in T.yunnanensis. In addition, the shoot axis length per unit dry mass, leaf dry mass per unit shoot axis dry mass, leaf number per unit shoot axis dry mass and leaf dry mass per unit total shoot mass increased with increasing light in T. yunnanensis, suggesting an enhanced investment of photosynthetic biomass under high light. These structural acclimations indicated that interspecific variations in shoot architectural adjustments may play an important role in differentiating species-specific photosynthetic potential.

Materials and Methods
Study species and sites. The field experiments were conducted at Jingdong Station (23°56′ -24°50′ N, 100°21′ -101°15′ E) of the Research Institute of Resources Insect, Chinese Academy of Forestry (CAF) in Yunnan Province, southeastern China. The elevation of this field station is 1200 m a.s.l. The mean annual temperature is 18.3 °C, with a range from − 2 °C in January to 37 °C in July 22 . The mean annual precipitation is 1100 mm with a seasonal distribution, primarily occurring in summer (June-September) 22 . Two Taxus species (Taxus yunnanensis and T. chinensis var. mairei) were planted by hand in 2008 under three light environments. We evaluated the shoot biomass allocation under three different natural light conditions: full daylight (high light), 40-60% full daylight (middle light) and < 10% full daylight (low light), estimated at mid-day using an LI-190SA quantum sensor (Li-Cor, Inc., Lincoln, NE, USA) 22 . The study locations were in proximity to the mean distance between sites 27 m (24-33 m). The soil in the sampled sites was latosol. Soil samples were collected under all three light conditions and subsequently tested, and there were no significant differences among the sites in terms of soil organic matter, N, P and K content, or available nutrients. This work was conducted in accordance with the forestry standards of the "Observation Methodology for Long-term Forest Ecosystem Research" of the People's Republic of China.

Plant materials and data collection.
Terminal shoots (including leaves and twigs) were all sampled under each of the three different natural light conditions. Under each natural light condition, we also randomly selected 20 individuals of similar height (188.6 ± 11.7 cm), and collected three terminal shoots from the top of each individual for analysis. Sixty individuals were sampled from each species, with 360 terminal shoots collected for analysis.
Sampled shoots were removed from trees, stored in plastic bags and transported to the laboratory in a cooling box. The number of leaves per a shoot were counted in the laboratory; the leaf fresh weight was measured with an electronic balance (± 0.0001 g); and leaf length, leaf width and leaf area were measured with a portable area meter (AM-300). The leaves were dried at 85 °C to a constant weight and then weighed to determine the leaf dry mass. All leaves from the same individual were pooled for chemical analysis. Leaf nitrogen (mass basis, N mass ) was measured using an elemental analyzer (NCS2500, Carlo Erba Instruments, Milan, Italy). Leaf carbon content was determined using a vario Macro Elemental Analyzer (Elementar Analysensysteme GmbH, Germany) 58 .
The basal and apical diameters of the shoot axis were measured (using a digital caliper with precision of 0.01 mm) in two perpendicular directions, and a mean was calculated. The shoot axis length (digital caliper with precision of 0.01 mm) and shoot axis dry mass (after drying to a constant mass at 85 °C) were also measured and recorded.
From the collected data, the following parameters were derived: individual leaf area, leaf roundness, leaf dry mass per unit area (LMA), total shoot mass, area-based leaf nitrogen content, shoot axis diameter, shoot axis length per unit shoot axis dry mass, number of leaves per unit shoot axis length, number of leaves per unit shoot axis shoot axis dry mass, leaf mass per unit shoot axis length, leaf dry mass per unit shoot dry mass, and leaf area. The individual leaf area of the shoot was calculated by dividing the total leaf area by the total number of shoot leaves; the leaf roundness was calculated by dividing the leaf length by the leaf width; the LMA was quantified as the ratio of dry leaf weight to the fresh leaf area; the total shoot mass was the sum of shoot axis mass and leaf mass; the area-based leaf nitrogen content was inter-converted via LMA (N area = N mass LMA) 13,59 ; the shoot axis diameter was considered to be the mean of the basal and apical diameters of the shoot axis; the shoot axis length per unit shoot axis dry mass was quantified as the ratio of shoot axis length to shoot axis dry mass; the number of leaves per unit shoot axis length and the number of leaves per unit shoot axis dry mass were calculated by dividing the number of leaves by shoot axis length and shoot axis dry mass, respectively; the leaf dry mass per unit shoot axis length and leaf dry mass per unit shoot axis dry mass were determined as the ratio of leaf dry mass to shoot axis length and shoot axis dry mass, respectively; the leaf dry mass per unit total shoot mass was quantified as the ratio of leaf dry mass to the total shoot mass; and the leaf area per unit total shoot mass was quantified as the ratio of leaf area to total shoot mass.

Statistical analysis.
A non-parametric analysis was conducted on leaf dry mass per unit area, leaf nitrogen content per unit area, leaf nitrogen content per unit mass, leaf number per unit shoot axis dry mass, leaf dry mass per unit total shoot mass and leaf area per unit total shoot mass for T. yunnanensis and leaf nitrogen content per unit mass for T. chinensis var. mairei. The Kruskal-Wallis analysis of variance with multiple comparisons was used, because the variance was not homogenous among different natural light conditions. For the other above-mentioned parameters, one-way ANOVA with a post hoc Tukey HSD test was used to compare parameter means among different natural light conditions. The independent-samples t-test was applied to examine species differences in terms of the parameter means among natural light conditions. All statistical analyses were performed using SPSS17.0 (SPSS Inc.). All tests were performed at a significance level of α = 0.05.