Assessment of leaf morphological, physiological, chemical and stoichiometry functional traits for understanding the functioning of Himalayan temperate forest ecosystem

Leaf functional traits support plant survival and growth in different stress and disturbed conditions and respond according to leaf habit. The present study examined 13 leaf traits (3 morphological, 3 chemical, 5 physiological, and 2 stoichiometry) of nine dominant forest tree species (3 coniferous, 3 deciduous broad-leaved, 3 evergreen broad-leafed) to understand the varied response of leaf habits. The hypothesis was to test if functional traits of the conifers, deciduous and evergreen differ significantly in the temperate forest and to determine the applicability of leaf economic theory i.e., conservative vs. acquisitive resource investment, in the temperate Himalayan region. The attributes of the functional traits i.e., leaf area (LA), specific leaf area (SLA), leaf dry matter content (LDMC), leaf water content (LWC), stomatal conductance (Gs), and transpiration (E) followed the order deciduous > evergreen > coniferous. Leaf carbon and leaf C/N ratio showed the opposite pattern, coniferous > evergreen > deciduous. Chlorophyll (Chl) and photosynthetic rate (A) were highest for evergreen species, followed by deciduous and coniferous species. Also, structural equation modelling determined that morphological factors were negatively related to physiological and positively with chemical factors. Nevertheless, physiological and chemical factors were positively related to each other. The physiological traits were mainly regulated by stomatal conductance (Gs) however the morphological traits were determined by LDMC. Stoichiometry traits, such as leaf C/N, were found to be positively related to leaf carbon, and leaf N/P was found to be positively related to leaf nitrogen. The result of the leaf functional traits relationship would lead to precise prediction for the functionality of the temperate forest ecosystem at the regional scale.


Results
Variations in morphological traits. LA ranged from 34.00 to 173 cm 2 for deciduous broad-leaved species, 6.00 to 71.50 cm 2 for evergreens broad-leaved and 1.00 to 50 cm 2 for conifers in the temperate forest studied (Fig. 1A). Mean SLA was 178.83, 112.00, and 43.31 cm 2 g −1 in deciduous broad-leaved, evergreen broad-leaved, and coniferous tree species, respectively (Fig. 1B). However, LDMC was similar for conifers and evergreens broad-leaved (0.36%), but slightly higher for deciduous broad-leaved species (0.41%) (Fig. 1C). Therefore, LA and SLA were both maximum in deciduous broad-leaved species, followed by evergreen broad-leaved and www.nature.com/scientificreports/ coniferous species (Fig. 1A,B). LA and SLA differ significantly across the studied leaf habits, however, deciduous broad-leaved and evergreen broad-leaved were statistically similar and differ significantly with conifers for LDMC ( Fig. 1A,B).
Variations in chemical and stoichiometry traits. LCC ranged from 28.00 to 48.36% (mean 42.21%) in deciduous broad-leaved species, from 20 to 60% (mean 45%) in evergreen broad-leaved species, and from 30.12 to 86.44% (mean 52.17%) in coniferous tree species (Fig. 1D). Mean LNC was 2.50%, 1.60%, and 2.00% in deciduous broad-leaved, evergreen broad-leaved, and coniferous tree species, respectively (Fig. 1E). Differences in LCC and LNC were significantly different with leaf habits (Fig. 1D,E) and LPC was more or less similar in all leaf habits i.e. non-significant at 5% level (Fig. 1F). Conifer and evergreen broad-leaved species were homogeneous group for LCC and LNC, and statistically differ with deciduous broad-leaved (Fig. 1D,E). Differences in stoichiometry traits i.e. leaf C/N and leaf N/P were statistically non-significant i.e. invariant across the three leaf habits (Fig. 1L,M).

Discussion
Evergreen leaves were characterized by higher leaf construction costs, slow nutrient returns, and tougher laminae. In contrast, deciduous leaves was associated with a higher photosynthetic rate per unit leaf mass, due to their higher LNC and SLA, higher intrinsic photosynthetic capacity, and less competition for light and carbon dioxide 27 . Evergreen broad-leaved favors infertile soils, longer leaf life span, and greater shade tolerance, that reduce seasonal variance in leaf exchange. In contrast, deciduous broad-leaved are favored by high seasonality, thermal, moisture, and light conditions, which are positively correlated across seasons 28 . Morphological, chemical, physiological, and stoichiometry functional traits varied among different leaf habits in temperate forests and support for understanding the functioning of the forest ecosystem. In general, present  www.nature.com/scientificreports/ evaluation observed variation in the traits and the order was: deciduous > evergreen > coniferous. However, LCC and leaf C/N ratio showed the reverse pattern (coniferous > evergreen > deciduous). There was little variation in LNC and LPC between the three leaf habits. Furthermore, present study found higher LA, SLA, and LDMC in deciduous broad-leaved species than in conifers and evergreen broad-leaved tree species. The higher LA and SLA in the deciduous broad-leaved species might be due to higher light interception 1,29 . Also, the higher SLA in deciduous broad-leaved species in tropical and subtropical forest indicates an acquisitive plant strategy 8,30 , while lower SLA in evergreens and conifers in subtropical forest indicates a conservative plant strategy 1,31 . LCC was significantly higher and LNC lower in conifers and evergreen broad-leaved plants in comparison to deciduous broad-leaved while LPC did not differ significantly among leaf-habits. A previous study on global data synthesis reported that leaves of conifers and evergreen broad-leaved plants have a higher carbon than deciduous broad-leaved species 32 . Higher carbon may be attributed to the presence of higher lignin content in conifers in comparison to the deciduous 33 . Leaf C/N was highest for conifers; however, leaf N/P was highest for deciduous species. The leaf C/N was high than subtropical forests 8,34 , possibly due to the high absorption capacity and utilize efficiency of nitrogen in the subtropical region 8 . High LWC is also reported for deciduous trees in temperate and boreal forests 35,36 . Chl and A were highest for evergreen broad-leaved tree species, possibly due to the higher availability of light and interaction with LNC. The availability of the leaf throughout the year enables the evergreen to use the nutrients to support the new growth and control photosynthesis 18 . Gs and E of leaves were lowest in deciduous species. The variation in the leave traits might be due to a robust leaf structure of evergreen species, which resists CO 2 diffusion resulting in lower mesophyll conductance such as Gs 37-40 and E 41 . Our analysis for the evaluation of variations in leaf functional traits was across woody species in a temperate forest in the Indian Himalayan region. Among the different types of species studied, the three conifer species (Abies pindrow, Cedrus deodara, and Pinus wallichiana) had higher LCC and LDMC and are thus said to follow www.nature.com/scientificreports/ resource conservation strategy 42 . The three broad-leaved deciduous species (Aesculus indica, Pyrus pashia, and Toona ciliata) had higher SLA and LNC, indicating resource acquisition strategy. The three broad-leaved evergreen species (Euonymus pendulous, Quercus leucotrichophora, and Rhododendron arboretum) had high leaf Chl and A, and the traits leaf Chl and A did not exhibit significant differences among the three types. Although this study was limited to a specific region and climate, we do believe the results can be extended to other temperate forests dominated by the same representative species. According to leaf economic theory coniferous have a resource conservation strategy with low SLA while the deciduous, have a strategy of fast acquisition with high SLA. Moreover, less productive plants tend to have low SLA, high LDMC, and leaves with long longevity (resource conservation strategy) however in the productive environment, the plants tend to have high growth rates, high SLA, low LDMC, low longevity leaves (fast acquisition strategy). The results of our study support the predictions of leaf economic theory, which is useful for understanding of functioning and a tool for predicting the responses of the forest vegetation to environmental changes, as plant strategies are dependent on the interactions among multiple traits 1 . This study provide information for further understanding the mechanism of species coexistence and predicting which kinds of species may assemble in a particular region in response to changes in environmental conditions.
We also observed that among all the morphological traits, LDMC was having strong relationship as observed by others 7,8 . LA and SLA were corelated as LA is directly affects the SLA 43 . Among physiological traits, Gs was having the strongest relationship, probablly due to the increase in CO 2 concentration in the atmosphere as observed by others 38,40,44 . We found a negative relationship between Gs and E that might be due to the rise in  www.nature.com/scientificreports/ CO 2 level 45,46 . Our study also observed the positive relationship between Chl and A in the temperate region as reported by others 36 .

Conclusion
In the temperate forest studied, plant functional type classification explains forest ecosystem functioning. There have been few systematic studies on functional traits in temperate forest tree communities of the Himalayan region. This study investigated leaf functional traits concerning the morphology, physiology, chemical and stoichiometry component of the temperate forest tree community in this region. The results demonstrated that LA, SLA, LDMC, Gs, and E differed significantly between leaf habits, and their values followed the order deciduous > evergreen > conifers. The LCC showed the opposite pattern, conifers > evergreen > deciduous. Leaf Chl and A rates were highest for evergreen species, followed by deciduous and coniferous species. Overall, the variation in leaf functional traits affected leaf functions. Hence, species co-habiting in the same environment employ different plant adaptive strategies, i.e. conservative and acquisitive, for dealing with that environment. Moreover, variation in the functional traits among three-leaf habits largely supports the predictions of leaf economic theory.

Material and methods
Study site. The temperate forest studied lies in the Indian Himalayan region of India, in Mussoorie Forest Division, Uttarakhand (30°28′02.6″N, 78°05′47.9″E; 2277 m asl). The mean annual rainfall in the region is around 2200 mm and the mean annual temperature is 20 °C. The region experiences three main seasons, winter (October to February), summer (March to June), and rainy (July to September) 47 . Soils of the region are Leptosols, Regosols, and Cambisols, developed mostly on dolomite 48 . The natural vegetation of the area is predominantly dense mixed forest (evergreen, deciduous, and coniferous tree species). Weather patterns differentiating this temperate forest and the region's geographical features have enabled dominance of species such as oak, rhododendron, and conifers 49 .

Plant functional trait measurement. A vegetation survey was carried out using the quadrat method,
where 20 quadrats, each measuring 10 m × 10 m, were laid out in the forest to study tree characteristics. The tree species were grouped into needle-leaved conifers, broad-leaved evergreens, and deciduous angiosperms (Table 1). LA, SLA, LDMC, LWC, and Chl were estimated based on measurements on five fresh, mature, fully expanded, and healthy leaves in five individuals per species, as described elsewhere 23 (Table 2). LCC, LNC, and LPC were measured on fresh leaves collected from the forest, dried in the laboratory, crushed, and analyzed according to methods listed in Table 2. Leaf physiological traits, i.e., A, E, and Gs were measured by LICOR XT-6400 photosynthesis equipment. The youngest and fully expanded leaves were used preferentially for measuring physiological parameters and measurements were made between 9 am and 2 pm under clear-sky conditions. A total of 45 observations were made for each parameter, on nine trees (five replicates per tree) (  ANOVA was applied to test the difference among the three functional groups for each functional trait and reported with Box-plot. Structural equation modeling was applied to examine the casual relationships among leaf morphological, physiological, chemical, and stoichiometric traits. SEM was conducted by the "lavaan package" and models were visualized with JASP software "SEM package" 38 .
Theoretical settings: a priori model. A theoretical framing of causal relationships between leaf morphological, physiological and chemical traits of temperate forest ecosystems, is explained in Fig. 3. According to an a priori model for the temperate forests, morphological and chemical features are understood to have a favorable relationship, while physiological traits and morphology are understood to have a negative relationship. This negative relationship may be attributed to the fact that many physiological processes are more plastic than structural processes 50 . Indeed, the physiology of trees can change dramatically without morphological changes in the short term 50 . Moreover, the difference in plasticity between morphological and physiological features is species-specific. Shade-intolerant species, for example, have greater physiological plasticity, while shade-tolerant species have greater morphological plasticity 51 . These complexities among the physiological, morphological and chemical features were used to assess species habitat affinities 50 , where leaf area was found to be positively associated with SLA 1,8,28 , as SLA influences canopy expansion and growth by changing total leaf area per plant and thereby affecting light interception and efficiency 52 . Chl, meanwhile, has a positive relationship with A 14 , as chloroplasts acclimatize to the environment and modulate the stoichiometry of components as per the requirements of plants 53 . A has a positive relationship with E 51 for the management of leaf temperature 54,55 , and a negative relationship with Gs 35,38 . The stomata alter the aperture in response to external conditions in order to maximize the photosynthesis-water loss tradeoff 56 , and are therefore strategically linked to Chl and A. The responsiveness of leaf functional features to environmental variables, i.e. light and nutrient availability, is used by plant species to occupy environmental niches 57 . SLA is positively correlated with the relative growth rate of plants 58 , and reflects the potential rate of return on investment for a leaf intercepting light 59 . Leaf size accounts for water use efficiency and the amount of light intercepted for photosynthesis 59 , whereby SLA is positively correlated with the relative growth rate of plants 58,60 . Finally, the photosynthetic capacity of the leaves is positively associated with foliar N concentrations and specific leaf area 61 . Here we emphasize the significance of the ecological scale at which trait variation is considered, and suggests that common trait-by-trait scaling interactions should be handled with caution at regional to local scales. More specifically, PFT can be a useful predictor variable for inferring one feature from another. The results have significant implications for dynamic vegetation models at the local scale, and for www.nature.com/scientificreports/ using trait-based techniques to predict forest function at regional and local levels, depending on the availability of the data 62,63 . www.nature.com/scientificreports/