Leaf and bud demography and shoot growth in evergreen and deciduous trees of central Himalaya, India
- First Online:
- Cite this article as:
- Negi, G.C.S. Trees (2006) 20: 416. doi:10.1007/s00468-006-0056-4
- 150 Views
Leaf and bud demography and shoot growth were studied in 10 evergreen (ES) and 15 deciduous (DS) tree species occurring between 600 and 2200 m elevation in the central Himalayan mountains in India. Results were analyzed to help explain why ES prevail in the vegetation of this region, even though the number of ES is no greater than for DS. Although each species had its own pattern with regard to leaf and bud demography and seasonality of shoot extension and radial growth, it was possible to group the species on the basis of shoot growth phenology. In most species, leaves emerged during March-April, at the onset of warm and dry summer season. The ES recruit leaves in shoots more rapidly than the DS. Across all species, peak number of leaves per shoot (5.8–20.7), peak leaf area per shoot (116.2–1559.2 cm2), peak number of vegetative buds per shoot (1.9–14.5), bud survival per shoot (23–84%), shoot extension growth (6.4–40.8 cm) and shoot extension period (13–30 weeks) varied considerably. The peak leaf area per shoot (587.7 vs. 246.7 cm2) and shoot extension growth (19.3 vs. 11.2 cm) were significantly greater for DS than for ES, and these two functional groups of species were clearly separable with regard to shoot growth characteristics.
Results indicate that rapid recruitment of leaf crop in the shoots, longer leaf life-span, and access to ground water due to deep roots were some of the advantages, the ES had over the DS, that may have likely enable them to maintain growth for a longer period in this region of warm winters and longer winter day length as compared to temperate climates. In the shallow rooted DS, shoot growth seems to be much affected by a seasonal drought in winter and they are likely to be affected more in the event of failure of monsoon rains in this region.
KeywordsCentral HimalayaDeciduous treesEvergreen treesLeaf and bud demographyShoot growth phenologyWarm temperate climate
Seasonality in phenology occurs (Hilty 1980), suggesting that other climatic factors such as, photoperiod (e.g. Wright and Van Schaik 1994), temperature (e.g. Arroyo et al. 1981), and rainfall (e.g. Opler et al. 1976) may be important determining phenology (Marques et al. 2004). In the central Himalayan region of India, the most notable feature of the climate is the monsoon pattern of rainfall that has a strong influence on plant adaptation and regional ecosystem processes (Zobel and Singh 1997). In this region, most forests are dominated by evergreen species (ES) (both broad-leaved and conifer); in spite of that, the number of deciduous species (DS) is as many as that of ES in the floristic composition (Singh and Singh 1987). Most of the ES have a leaf life-span of about 1-yr and show a simultaneous leaf fall and emergence in March-April. A majority of the DS of the region shed their leaves during winter, which corresponds to the dry season. In this region in both the ES and DS, peak activity of leaf emergence starts at the onset of dry and hot summer weather, and new foliage is completely developed long before (1–3 months) the onset of the rainy season (Negi and Singh 1992). In the sub-tropical humid seasonal forests in north-eastern part of Himalaya, leaf emergence in a majority of trees takes place towards the end of the dry season that is followed by a long rainy season (Shukla and Ramakrishnan 1982). In contrast to the Himalaya, leaves of species with determinate growth in summer-dry temperate climates are fully expanded before the height of drought (e.g. Zobel 1974). Although temperature (Longman and Jenik 1974) and photoperiod (Frankie et al. 1974) have been suggested as the possible triggering factors of leaf emergence, the leaf formation in trees during a continued dry season has not been assessed properly. An equally important question, which has been a concern of regional ecologists, is why ES dominate the forest vegetation of this region, while DS are of minor importance (Singh et al. 1990; Dhaila 1991). In these studies, the superiority of ES over DS has been explained mainly by the ability of ES to maintain photosynthesis year-round. Because this region is located in sub-tropical latitudes, day length and temperature do not limit phenology as much as in temperate regions (Singh and Singh 1987). Previous studies indicate that the distribution of DS is generally limited to mesic and fertile sites along the water courses (Troup 1921; Champion and Seth 1968), and a delay in the commencement of rainy season may affect leaf growth of DS more than the ES (Negi and Singh 1992). It is generally recognized that seasonal drought can favour deciduous leaves, and that infertile soils can favour long-lived evergreen leaves (Givnish 2002).
Oak and pine, the two major forest forming genera of the study area, have been reported to differ in their basic strategy of countering water stress (Zobel and Singh 1995). Tewari (1998) measured water potential of these two species in this region and found that deep rooted oak conducted water at lower water potential (0.5–0.6 MPa) than does pine. It implies that shallow rooted pine avoids severe water stress by closing stomata. Zobel et al. (2001) found lowest water potential at high elevations in this region. Poudyal et al. (2004) studied water relations and drought response in five dominant forest tree species in Nepal and found significant variation among sites, species and season in predawn and midday water potential and monthly soil water potential at 15 cm depth for all species. In a subtropical savanna, Nelson et al. (2002) found that DS were more deterministic and constrained in their growth responses to increased moisture availability than ES. Baker et al. (2002) reported that, despite similar rainfall, the soil water availability during dry season in an evergreen forest site was more than in the semi-deciduous forest in a tropical rain forest of Ghana.
Leaf habit (e.g. evergreen, deciduous) and leaf life-span are important plant life history traits, which are widely regarded as important attributes in the classification of plant functional types (Chapin and Tryon 1983; Reich et al. 1997). These traits have been linked to a number of plant and ecosystem-level processes, including photosynthetic capacity, growth rate, primary productivity and nutrient cycling (e.g. Shaver 1981). The potential use of demography to describe and predict the growth of plant parts has been frequently pointed out (e.g. Bazzaz and Harper 1977; Maillette 1982; Kudo 1995). Each species has a unique pattern of leaf demography (Chabot and Hicks 1982; Nilsen et al. 1987), which interacts with leaf physiology to control plant growth (McGraw and Antonovics 1983). The seasonal leaf population size is controlled by patterns of leaf production and mortality, which are often not clearly separated in time (e.g. Gill and Mahall 1986). Also, seasonal changes in leaf area are critical for fluxes of carbon, solar energy and water in forests. For example, photosynthesis, transpiration, respiration and light interception can be directly related to leaf area (Gholz et al. 1976).
Two contrasting patterns of bud and shoot development are generally recognized in temperate trees (Tomlinson 1978). In one pattern, found in most conifers, the vegetative shoot is fully “preformed” in the resting bud (Lanner 1976). In the other, at least some shoots are not entirely preformed in the resting buds and a portion of the leaves are formed in the season that the shoot extends (“neoformed”) (Halle et al. 1978). Lechowicz (1984) reported that, with the onset of spring in the temperate deciduous forests of eastern North America, tree leaves do not all emerge in perfect synchrony. It is less clear what causes phenological differences among tree species in a particular region (e.g. Ahlgren 1957). Even within single forests, leaf emergence in spring varies over several weeks among co-existing native trees, and species produce leaves at quite different rates during the rest of the growing season. This has been reported for Populus (Critchfield 1960), Fraxinus (Gill 1971) and Quercus spp. (Reich et al. 1978). Temperate trees exhibit periodicity of both extension and radial growth, which is clearly correlated with seasonal fluctuations in climate (Zimmermann and Brown 1971). Continuous stem shrinkage, even for several years during rainless periods, has been reported for many trees (Kozlowski et al. 1962). A greater stem shrinkage in DS compared to ES have been observed by many workers in the central Himalayan region (Singh et al. 1990; Dhaila et al. 1995) and elsewhere (e.g. Fraser 1956; Winget and Kozlowski 1965), following a severe drought (Baker et al. 2002). This shrinkage is highly correlated with atmospheric water deficits (Hinckley et al. 1978s, Reich and Borchert 1982). A few studies in this region (e.g. Ralhan et al. 1985; Rawal et al. 1991), and the north-eastern Himalayan region (e.g. Boojh and Ramakrishnan 1982; Shukla and Ramakrishnan 1982) have described some phenological traits of the forest vegetation and has stressed the need to carry out more studies. Phenology has recently emerged as an important focus of ecological research and could have a contribution to climatic change studies (Schwartz 1999).
In this study, seasonal patterns of recruitment and mortality of leaves and vegetative buds, shoot elongation and shoot diameter growth in 25 tree species (10 ES and 15 DS) of the central Himalayan region were recorded. Objectives were: (I) to recognize differences in leaf and bud demography and shoot growth phenology of ES and DS; and (II) to understand how shoot growth phenology is linked with the preponderance of ES in the vegetation of this region. It was expected that from this comparative study of these two major functional groups of species, a better insight would be produced about the phenological adaptations of plants in response to climate of the region.
Description of the study area
Study site and vegetation
List of species selected for study in the Central Himalayan region
Growth form and leaf persistence
Leaf life-span (days)*
Shorea robusta Gaertn. f
Chir pine forest
Pinus roxburghii Roxb
Myrica esculenta Ham. ex. D. Don
Quercus leucotrichophora A. camus
Acer oblongum Wall. ex Dc
Rhododendron arboreum Smith
Quercus floribunda Don
Ilex dipyrena Wall
Machilus duthei King
Quercus lanuginosa Don
Litsea umbrosa Nees
Forest gaps, landslide sites, early successional
habitats & plantations
Sapium insigne Trim
Engelhardtia colebrookiana Lindl.
Sapium sebiferum Roxb
Alnus nepalensis D.Don
Salix alba L
Populus deltoides Bartr
Aesculus indica Colebr
Fraxinus micrantha Lingelsh
Populus ciliata Wall
Around cropfield areas
Bauhinia variegata Linn
Ficus palmata Forsk
Grewia optiva Roxb
Prunus cerasoides D.Don
Pyrus pashia Buch.-Ham
Climate of the study area
This region has certain characteristic climatic features. Though it falls under sub-tropical latitude, the abrupt rise in mountains creates a temperature comparable to that of a temperate climate. Areas above 2000 m elevation receive frequent snowfall in winter. However, the most notable feature of the climate is the monsoon pattern of rainfall with about three-quarters of the annual rainfall (1500–2500 mm) concentrate in the rainy season (mid June–mid September). This period is most favourable for plant growth, for it is also warm. In spite of high annual rainfall, early summer (preceding the monsoon period) and winter are relatively dry, generally with <10 mm monthly rainfall, and potential evapotranspiration that is often in excess of precipitation (Singh and Singh 1992). Data obtained from the State Observatory at Naini Tal (located at 1850 m elevation in the study area) for the study years (1985–86) indicate that within the elevation transect of 600–2200 m the mean monthly temperature ranges from 13–32°C at 600 m, 8–21°C at 1500 m, and 6–20°C at 2200 m elevation. A rise of 270 m in altitude corresponds to a fall of 1°C in the mean temperature up to about 1500 m, above which the fall in temperature is more rapid (Singh and Singh 1987). The year is divisible into three seasons: the winter, a cold and relatively dry season extending from mid-December through February; the summer, a warm and dry season extending from mid-April to mid-June; and the rainy, a warm and wet season extending from mid-June to mid-September. Transitional periods between summer and winter and between winter and summer are referred to as autumn and spring, respectively. Considering that the conditions are severely dry at aridity index less than 1.25 and moderately dry between 1.25–1.88 (Koppen 1931), all the months with the exception of rainy months were severely dry at low-to-mid altitude sites (≤1450 m) and January, March, April and November at the high elevation sites (≥1750 m altitude). Data on the sunshine hours indicate that for March and April, when plant growth is initiated, the mean sunshine was 8.14 h and 9.14 h per day, respectively. During the rainy season the mean monthly sunshine ranges from 2.01–3.27 h per day, and during winter it ranges from 5.56–8.05 h per day.
Forest soils of this area investigated by Khanna (1986) reveal that forest soils lack horizontal profiles. The soil is sandy loam, with sand percentage decreasing from lower to higher altitudes (79–61%). The pH ranges from 6.0–6.8, and gradually declines with increasing altitude. Soil organic carbon and total nitrogen increase with increasing altitude. Data on total N and P content of soil indicate that the forest soils are not poor in nutrients. However, the nutrient pool size may be limiting because of the lack of deep soils on most of the forested hill slopes. Seasonal pattern of soil moisture (at 0–10 and 10–30 cm depths) measured for the site of lowest elevation follows that of the aridity index mentioned earlier. Tewari (1998) found soil water potential at 10 and 60 cm soil depths significantly different across all the forest types (sal, chir pine and oak) and seasons in the study area. And the soils at 60 cm depth were moister than at 10 cm depth.
Leaf and bud demography and shoot growth measurements
Twenty five tree species (10 ES and 15 DS) were selected for study (Table 1). Out of the 25 species, two (P. deltoids and S. alba) were exotic species, maintained by the Forest Department as pure plantations. The ES were divisible into four canopy and six sub-canopy species. The DS generally occurred as isolated individuals, or they formed small patches in gaps of natural forests or in landslide areas (early successional habitats) and around crop fields. All the DS (except S. robusta) were sub-canopy species. Except P. roxburghii (a conifer), all the species were broad-leaved.
Phenological records on leaf emergence and leaf drop were made from February 1985 to May 1987 in a forest site where the trees of a given species were extensively distributed (Table 1). At each of the sites the number of forest stands (area about 1 km2) varied from one to five, depending upon occurrence of a species. In each site, 50–100 trees of similar maturity were marked for each species. The sites were visited at a weekly interval during the periods of peak phenological activity. During the remaining period the visits were less frequent. If a given phenophase was observed in 5–10% individuals of a species, it was considered to have initiated, and the species was considered to be in that phenophase as long as that phenophase was represented by at least 5–10% individuals (cf. Ralhan et al. 1985). To determine the demography of leaves and buds, and shoot extension and radial growth for each of the 25 species, 10 average sized mature trees (dbh >31.5 cm), having a similar degree of crown development were selected across the five study sites. In each of the selected individuals, 100 vegetative buds (distributed equally in upper, middle and lower crown positions) were marked in spring 1985. Fifty new shoots (originating from marked buds) were harvested randomly at monthly intervals, and observed for number of leaves and buds and shoot extension and radial growth. In this way it was possible to record the periodical changes in number of leaves and buds and shoot extension and radial growth precisely. The number of buds surviving till next successive growth period and producing new vegetative shoots was considered bud survival relative to the peak number of vegetative buds produced in the previous year. Shoot length was measured to the nearest 1 mm and the diameter to the nearest 0.1 mm by a micrometer in two directions (at right angles). Seventy five leaves (of the major spring cohort) were selected randomly from these harvested shoots to measure leaf area by leaf area meter (LI 3000 A, LI-COR, USA). Leaf area per shoot was computed monthly by multiplying mean leaf number times mean leaf area. Mean leaf life-span was adapted from Negi (1989) who has calculated it by considering the time interval (days) between peak leafing and peak leaf drop on a leaf population basis in a forest stand marked under this study. In this study, the classical leaf survivorship curve, tagging leaves and their periodical censuses (e.g. Williams-Linera 2000) was not attempted to determine leaf life-span due to logistic constraints. The data were statistically analyzed (Snedecor and Cochran 1967).
Annual variations in leaf emergence of trees
Comparison between evergreen and deciduous species in regard to time of leaf initiation in year day, with January 1 being the day 1
Study site and altitude
Growth initiation (in year day)
Kalona (1350 m)
Kailakhan (1950 m)
Government House (2000 m)
Kilbury (2200 m)
Average for all forest species
In 20 of the 25 species studied, vegetative bud-break (leaf emergence) occurred during March-April when air temperature rises steadily. Leaves emerged as early as in late February in three DS, F. palmata, P. pashia and S. alba, and as late as in mid-May in B. variegata and S. insigne (both DS), when conditions were warm and dry. In many species leaf emergence lasted 3–4 weeks. However, in two DS (P. ciliata and P. pashia) leaf emergence extended over 7–8 weeks, and in five DS only about two weeks (Fig. 2). P. cerasoides was an unusual species, for it produced new leaves and shoots during autumn when most deciduous species were shedding their leaves.
All species had a unimodal pattern of leafing in spring. However, two-three minor leaf flushes were also recorded in most of the ES and half of the DS during rainy season through winter. In general, these minor leaf flushes contributed ≈20% to the total leaf population. Of the 25 species studied, 13 show minor leafing activity either or both during rainy season and autumn (Figs. 3 and 4). All the three oak species showed third minor leafing, occurring during the autumn. All the ES except for A. oblongum, L. umbrosa, M. duthei, M. esculenta and R. arboreum (all sub-canopy species) showed minor leafing activity. Minor leafing was also recorded in 7 DS (B. variegata, G. optiva, P. cerasoides, P. pashia, S. sebiferum, and two exotic species P. deltoids and S. alba) during the rainy season. The number of native species having minor leafing increased with increasing elevation (r=0.968; P<0.01).
Certain characteristic of leaf and bud growth in the shoots of 25 central Himalayan trees (±SE). Values in parentheses are the number of species falling in each group
Percent leaf population attained in Ist month of bud-break
Peak leaf pool size (number per shoot)
Period of stable leaf pool size (weeks)
Peak bud pool size (number) per shoot
Bud survival per shoot (%)
Mature leaf area (cm2) per shoot
Leaf area attained in Ist month of bud break (%)
Leaf mortality (leaf drop) began as early as mid-rainy season in nine of the 13 DS, in autumn in A. nepalensis, E. colebrookiana, F. palmata and G. optiva (all DS), and in January (winter) in all the ES (Figs. 3 and 4). Peak leaf mortality in the DS was recorded between November and February, while in the ES it was during March-April, which coincided with new leaf and shoot formation. Leaf mortality was spread over a maximum of nine months in Q. leucotrichophora and Q. floribunda, and a minimum of two months in F. palmata. The minor leafing activity induced a minor leaf drop in the older shoots of some species (viz., I. dipyrena, L. umbrosa, M.duthei, M. esculenta, P. roxburghii and all the three oaks).
The peak leaf area per mature shoot was found smallest in I. dipyrena (116.2 cm2) and largest in S. robusta (1559.2 cm2). On an average the peak leaf area per shoot for DS (587.7 cm2) was significantly greater (P<0.02) than for the ES (246.7 cm2). Accomplishment of peak leaf area in shoots in the ES was rather slow and accomplished one month later than the DS. Per cent leaf area per shoot accomplished after one month of leaf emergence was 63.3% for ES and 67.6% for DS (Table 3).
In most species bud recruitment progressed from shoot base to shoot apex during April-May, i.e., nearly one month after the initiation of leaf emergence (Figs. 3 and 4). In 17 species (half of them were ES), bud recruitment was completed within a 2–3 months, whereas it lasted 4–6 months in the remaining species. In P. roxburghii, R. arboreum and S. insigne (all early successional species), bud recruitment began during the rainy season (long after leaf emergence had been completed), and it was localized to the shoot apex. A few basal leaves were devoid of buds in all species. Total number of buds produced per shoot was lowest (1.9) in P. roxburghii and highest (14.5) in S. alba. This value was almost the same for ES (6.7) and DS (7.8). The peak number of buds and leaves were positively correlated (r=0.697; P<0.01). The bud population remains stable for a minimum of 13 weeks in some DS with a short leaf life-span (P. ciliata and S. insigne) and for a maximum of 30 weeks in A. nepalensis, E. colebrookiana and P. cerasoides (all DS with leaf life-span about one year; Table 1). In this respect ES and DS (21.4 vs. 21.7 weeks) were not different.
Bud mortality began in late rainy season in P. cerasoides, S. sebiferum, F. micrantha and P. ciliata (all DS) and during autumn in half of the ES (including all oaks and their sub-canopy counterparts) and some DS, and during winter in A. nepalensis, B. variegata, E. colebrookiana, G. optiva (all DS), and in P. roxburghii and R. arboreum (both ES). In general, bud mortality started from the shoot base and progressed towards the shoot apex. Bud survival ranged from 23% in M. esculenta to 84% in P. roxburghii. The average bud survival for ES and DS was almost equal (42% vs. 43%) (Table 3). Bud mortality was almost equal in ES and DS (mean = 57.3%), but occurred rapidly in DS. Bud mortality in DS coincided with the decline in temperature in autumn and winter drought, whereas it was more delayed in ES. Bud mortality seems to be increased also by a shaded environment, for the sub-canopy ES had a markedly lower bud survival (37%) than the canopy ES (49%). Vegetative buds survived and produced new shoots only at the shoot apex (sylleptic branching) in some of the early successional species (e.g. P. roxburghii, R. arboretum, S. insigne).
Twig extension and radial growth in 25 Central Himalayan trees (±SE). Values in parentheses are the number of species falling in each group
Percent (of the total shoot length) attained in Ist month of bud-break
Shoot elongation period (weeks)
Shoot length at full extension (cm)
Percent (of the total) diameter attained in Ist month of bud break
Maximum shoot diameter (mm)
Percent winter shrinkage of shoot diameter
Shoot diameter growth
The percentage of the total shoot radial growth realized within one month of shoot extension ranged between 50% in S. alba and 92% in P. roxburghii. This value was almost similar for ES (74%) and DS (70%) (Table 4). The duration of shoot diameter increment varied from two months in P. roxburghii and M. esculenta to 7–9 months in L. umbrosa and R. arboreum (Figs. 5 and 6). The peak shoot diameter of Q. floribunda (2.8 mm) was the minimum and that of P. roxburghii (18.8 mm) was the maximum. The mean shoot diameter for both ES and DS (5.6 vs. 4.9 mm) was similar. Shoots of all species shrank generally beginning in autumn in DS and in winter in ES. The shrinkage ranged between 7% in Q. floribunda and 28% in S. insigne (Table 3). The average shoot shrinkage for DS (18.9%) was conspicuously greater than for the ES (14.2%). Peak radial growth attained by the shoots coincided with the rainy season in most species, followed by a decline in diameter, which was significantly greater (P<0.05) for DS compared to ES. Reduction in shoot diameter was primarily due to shrinkage of intact tissue rather than loss of outer twig tissues.
The influence of cold winter on the life of plants, reflected in suppression of growth, is a characteristic feature of temperate climates (Maruyama 1978), which is different from the tropical climates where harsh summer with low soil moisture may suppress the leaf emergence till the onset of rainy season (Frankie et al. 1974; Longman and Jenik 1974). In this respect the central Himalayan region somewhat resembles the warm temperate parts of the world (Kozlowski 1972), as most of the ES and DS of this region produce leaves during warm and dry summer season, long before (1–3 mo) the onset of rainy season. Characteristics that distinguish the Himalaya from most temperate forest regions include a climate with highly concentrated summer rainfall, mild winters for the latitude, high sun angles due to the subtropical latitude and high elevation (which increase potential insolation), and low annual variability in day length (Zobel and Singh 1997). Therefore, with respect to the growth initiation in the forest vegetation, this region represents a transition between the typical temperate and tropical climates.
The proximate control of initial leaf emergence in most temperate deciduous trees is usually the cumulative thermal sum (degree-hours, degree-days) to which buds are exposed after a prerequisite cold period (Kramer and Boyer 1995). Studies (Njoku 1963; Lawton and Akpan 1968) have implicated day length and air temperature (Frankie et al. 1974; Hanninen 1995) increase as the inducer of leaf flushing, which holds true for the present study area where peak activity of bud break and leafing takes place during March–April when photoperiod and temperatures are increasing (Negi 1989). Further, the role of isolated rain showers to initiate leafing by replenishing water content has also been frequently emphasized (e.g. Borchert et al. 2002). In the study sites, though the soil moisture continues to be low from October to mid-June, the long dry spells are broken by isolated rain showers (average monthly rainfall is 42.9 and 61.5 mm around 2000 m elevation, and 31.1 and 6.7 mm in March and April, respectively in lower altitudes), possibly facilitating leaf emergence during March-April. Reich and Borchert (1982) reported that a storm of 20 mm rainfall could revive the water potential sufficiently to support growth in Tabebuia neochrysantha. It is also possible that redistribution of water stored in roots of trees may help in raising the water potential of shoots. Poudyal et al. (2004) observed a decrease in wood water content before the onset of leaf flushing, that suggests withdrawal of stored stem water during active phenological development, which may contribute to rehydration of active tissues (Borchert 1994). The establishment of new foliage shortly before the warm and wet rainy season is likely to optimize photosynthetic gain (Rivera et al. 2002).
Of special interest is the leaf phenology of ES, which dominate the forests of this region, out competing DS in mature communities of most habitats (Singh and Singh 1987). ES retain leaves throughout the winter and exhibit simultaneous leaf fall and emergence in summer resembling the “leaf-exchanging type” species of tropical forests (Longman and Jenik 1974). Retaining leaves throughout a year enables the ES to utilize the same unit of nutrients to support the new growth (Fife and Nambiar 1982) and maintain some photosynthesis throughout the winter (Saeki and Nomoto 1958). This strategy makes a tight circulation of nutrients in the ecosystem, a characteristic feature of late successional communities (Vitousek and Reiners 1975) occupied by the ES in this region. Although the DS of this region has significantly greater leaf area per shoot compared to ES (587.7 vs. 246.7 cm2), and they are known to have higher photosynthetic efficiency (Murphy and Lugo 1986; Bhadula et al. 1995), the leaf mass values are reported to be 3.7–8.6 Mg/ha for deciduous forests and 10.0–28.2 Mg/ha for evergreen broad-leaved forests, and the net primary productivity increases linearly with both leaf mass and leaf mass duration (Singh et al. 1994).
In this region drought could be stressful for most species, particularly during peak summer (May and early June). In a study on water relations of the three major forest types of the present study area (i.e. sal, chir pine and oaks), Tewari (1998) found that soil water potential (60 cm depth) remain high up to autumn because of the continuing effect of the concentrated monsoon rainfall. It declines rapidly in subsequent winters to an extent that trees may need to draw water from deeper soils, and rises in spring (higher than the tree water potentials), enabling trees to draw water from shallow soil layers. Subsequently during summer soil water potential declined. The ES either raise or stabilize their water potential even against desiccating soils when they produce leaves. However, the degree of rise is generally lower than in some DS of the region (Singh et al. 2000). In the neighbouring country Nepal, Poudyal et al. (2004) reported that plant water potential declined with increasing elevation, whereas leaf conductance increased. Deep rooted species typically have higher (less negative) predawn shoot water potential, than shallow rooted species because soil water availability increases with depth (Abrams 1990; Myers et al. (1997).
In most species, leaf production and shoot growth was rapid, about 80% or more leaves appearing in the first month following bud-break; and this period tended to become shorter with increasing elevation (r=0.08). It seems that the basic difference in duration and rate of leaf recruitment is due to “preformed in buds behaviour”. Therefore, many of the species growing at the higher altitudes in the study area (e.g. F. micrantha, L. umbrosa, M. duthei) produced ≈ 80% leaves in first month following bud break, and were thus comparable to the “short flush” species of temperate region (cf. Lanner 1976; Rook and Corson 1978). A few DS, such as A. nepalensis and P. deltoides (leaf production within one month following bud-break ≈ 55%), recorded distinctly lower values. Further, in some of the DS (e.g. A. nepalensis, G. optiva, P. pashia and S. alba) shoot elongation and leaf production continues for a longer period (16–30 weeks), emphasizing their “free growth” behaviour of shoots, as has been recorded for some species in the north-eastern Himalaya (Shukla and Ramakrishnan 1982), some temperate trees (e.g. Acer, Betula, Eucalyptus and Populus) (Pook 1984), and subtropical savanna, generally after small precipitation events (Nelson et al. 2002). These species thus possess two types of leaves: early leaves, which emerge in the spring, and late leaves (short lived), which emerge in summer through rainy season; a pattern similar to Betula platyphylla (Kudo 1995). Also in the present study almost all the ES and half of the DS produced some leaves in two to three minor episodes (multiple leaf flushing) other than the major leaf flushing in spring. Poudyal et al. (2004) reported that water relations influence tree phenology. For example, Q. lanata, which has multiple flushing, maintained high predawn water potential in summer despite severe drought. Unlike a bimodal pattern of leaf production, as has been reported in tropical climate (e.g. Mulkey et al. 1992), all the species studied by us show a unimodal pattern of leaf production.
Most investigations on the phenology of DS and ES have concluded that the former commence growth earlier than the latter (e.g. Mooney and Dunn 1970; Gray 1983). However, this difference may be related to a difference in microsites occupied by these species within the community (Gill and Mahall 1986). In this study a weak trend of earlier commencement of growth in DS was observed when ES and DS co-existing at the same site were compared (Table 2). Contrary to it leaf flushing was delayed by one week in DS than the ES in the timberline vegetation of this region (around 2500 m altitude) (Rawal et al. 1991). Earlier and rapid leaf expansion (2.61 vs. 0.89 cm2/d) in DS than in ES as reported for these species (Negi and Singh 1992) is considered typical of northern temperate trees and shrubs (Kozlowski 1971; Chapin and Tryon 1983). In an oak forest of this region it was found that DS complete shoot extension and leaf expansion significantly earlier than the ES (Dhaila et al. 1995). In a tropical montane cloud forest of Mexico, patterns of leaf emergence were similar for DS and ES, and more than 50% leaf expansion occurred in less than a month for most of the DS but required more than a month for the ES (Williams-Linera 2000). ES and DS have different rooting patterns and they experience distinctive soil moisture and nutritional regimes (Sobrado 1986). In this region, the shallow rooted DS have a growth pattern closely coupled with warm and wet season, whereas the deep rooted ES enable the plants to use soil water during the dry season also (Dhaila et al. 1995). The gradual stem shrinkage distributed over 3–5 months in the ES indicate that they are less vulnerable to drought. If arrival of monsoon were to be delayed, which is quite common in this region, the ES with better adaptation to drought would have an advantage over the DS, in which nutrient acquisition from soil appears to be a wet season dependent phenomenon (Negi and Singh 1993). The wet conditions are known to improve soil nutrient status (Yavitt et al. 2004). It seems that drought stress is not severe enough in our study area to give advantage to leaflessness and summer deciduous phenology, thus explaining the dominance of ES in this region (Tewari 1998).
The evergreen leaf habit is often considered advantageous in climatic regions where precipitation and temperature allow year-round photosynthesis, with the deciduous habit increasing in prevalence as the unfavourable season (e.g., cold winter or summer drought) increases in length and/or severity (Chabot and Hicks 1982; Kikuzawa 1995). The incidence of DS in the tropics is reported to increase with increasing seasonality of rainfall (e.g. Whitmore 1979; Hall and Swaine 1981). In our study area the difference between the mean June and January temperatures were 19 and 14°C at the lowest and highest sites of the transect, respectively. These temperature differences are lower than 20°C, which is the lower limit of the difference found by Wolfe (1979) to correspond to the occurrence of deciduous forests in relatively less dry areas. In the present study area of subtropical latitudes, the ES bear leaves throughout the year, but like DS bear the cost of annual replacement of old leaves by new leaves. A situation of less marked contrast between favourable and non-favourable periods, with respect to temperature, seems to favor the leaf characters of ES. The temporal variation in leaf initiation, leaf longevity (Negi and Singh 1992) and canopy development observed in this study, together with varied photosynthetic activity (Bhadula et al. 1995), and a pronounced vertical rooting (Dhaila et al. 1995) may be factors that allow for the continued coexistence and relatively high woody plant diversity in the region (Singh et al. 1994). As has been reported by Nelson et al. (2002), this study indicate that the woody plants in this seasonally dry subtropical ecosystem do not adhere to the broad generalizations of leaf habit-leaf longevity relationships developed in highly seasonal temperature, high latitude or tropical ecosystems; but rather, more closely resemble patterns observed in non-seasonal, tropical environments (Kikuzawa 1978; Givnish 2002).
This study was carried out under the supervision of Prof. S.P. Singh, FNA, Head, Botany Department, Kumaun University, Naini Tal, India. Author is thankful to Prof. D.B. Zobel of Oregon State University, USA for reviewing a draft copy of this paper, and to the Director, G.B. Pant Institute of Himalayan Environment and Development, Kosi-Almora for providing facilities.