Plant Community Composition and Structure of Chenab Valley in a Part of Nanda Devi Biosphere Reserve

The present paper deals with the vegetation communities and their diversity patterns in Chenab valley, the buffer zone of Nanda Devi Biosphere Reserve (NDBR) in Chamoli District of Uttarakhand, India. A total of 42 sites were selected randomly based on the landform heterogeneity of the area. Eight forest communities with overlap among vegetation types and also various plant associations were noticed through Principal Component Analysis (PCA) following PAST program and two shrub communities identified separately were, Berberis and bamboo. The range of density in various forest communities was from 203-545 trees ha-1 and total basal area from 17.5-71.7 m2 ha-1. The range of species richness of tree layer, shrub layer and herb layer was from 2-14, 1-10 and 4-14 and diversity from 0.693-2.304, 0.5142.052 and 1.202-2.583, respectively. The distribution pattern of trees, shrubs and herbs shows that the species were evenly distributed in most of the sites and the ß-diversity of the present study area is 7.4. Rhododendron and Taxus, the undercanopy species facilitated the regeneration of Chimnobambusa falcata, while the conversion of lower girth class individuals to higher girth class individuals is steady and progressive. Though, evergreen and deciduous species had good population of seedlings and saplings, but the conversion to next girth class was very poor due to the high anthropogenic pressure. The present study reveals that the forest vegetation in Chenab valley is better than that of other parts of Nanda Devi Biosphere Reserve, for which conservation strategies have been discussed in the paper.


Introduction
The Himalaya is one of the richest biogeographic zones in India and is well known for its ecological, hydrological, aesthetic and socio-cultural values (Mani, 1974). The location, topography and climate of this complex system have endowed it with rich and diverse life form. The large scale surface removal due to geological instability and cyclic climate change has greatly influenced the floral and faunal communities. The disturbance can alter environmental conditions by changing light availability and soil condition, which are mainly due to heavy anthropogenic pressure (Fredericksen and Mostacedo, 2000). The chronic form of disturbance in these forests often led to inadequate recovery of ecosystems due to continuous exploitation of the forests (Singh, 1998). The recurrent anthropogenic intervention such as fuel wood, fodder, litter and non-timber forest products (NTFP) collection, as well as grazing, browsing and trampling largely by livestock can substantially lead the habitat alteration of the species (Pandey and Shukla, 1999).
Garhwal Himalaya has relatively a mixture of dry and moist temperate climate, which influences the growth and vitality of the forests through the water balance in the watersheds. Chamoli District in Uttarakhand state harbors the rich and varied flora, which has been of great attrac-tion for the professional collectors, ecologists and as well as amateur enthusiast. The occurrence of large number of species in the area and richness of the flora has rendered the district a botanical paradise. The lower Chenab valley, a buffer zone of Nanda Devi Biosphere Reserve forms the most prominent ecological boundary, where sub-alpine forest terminates. In many areas the presence of high altitude oak (kharsu oak, Quercus semecarpifolia), maple (Acer caesium), conifer (silver fir, Abies pindrow) and deciduous species (birch, Betula utilis) reach up to timberline. The common montane bamboo (Thamnocalamus spathiflorus) is prevalent in the canopy gaps (Puri et al., 1989).
The area above treeline is marked by a zone of stunted trees (krummholz) and shrubs with associate species like Salix, Rosa, Cotoneaster and Berberis. The structure and composition of high altitude forests in west Himalaya has been described (Champion and Seth, 1968;Singh and Singh, 1992) and richness and diversity of forest ecosystems by several workers (Adhikari et al., 1991(Adhikari et al., , 1998Bhandari et al., 1997;Dhar et al., 1997;Hussain et al., 2008;Kumar et al., 1997;Kunwar and Sharma, 2004;Rikhari et al., 1991;Saxena and Singh, 1984) in last few decades. The Garhwal Himalaya is witnessed to a faster place of modernization and development due to various activities, viz. coming up large scale dam construction, pilgrimage, widening of roads and eco-tourism. Such signs of gered and endemic species e.g., Musk deer (Moschus Chrysogaster), Himalayan Tahr (Hemitragus jemlahicus), Common leopard (Panthera pardus), Asiatic black bear (Ursus thibetanus), Serow (Nemorhaedus crispus), Royles' Pika (Ochotona roylei) among mammals and Monal (Lophophorus impeyanus), Koklas (Pucrasia macrolopha), Kalij (Lophura leucomelanos), Snow Partridge (Lerwa lerwa) and Himalayan Snowcock (Tetraogallus himalayensis) among birds. Although, agriculture has been primary source of subsistence economy for the local people of the area and enjoy unrestricted access due to traditional rights to use the natural resources, such as fuel wood and fodder collection along with non-timber forest products and use the area for livestock grazing. Exploitation of endangered medicinal plants (Dactylorhiza hatagirea, Podophyllum hexandrum, Picrorhiza kurooa) and pastel green (foliose lichen, locally known as Jhula) are the major concern of management.
The climate of the study area is typically the West Himalayan temperate and alpine type and divided in to five distinctive seasons, viz. spring (April), summer (May-June), rainy ( July-September), autumn (October-December) and winter (December-March). Precipitation is moderate over most of the year, while concentrated during monsoon and during winters the precipitation was in the form of snow and higher elevation areas experience heavy to moderate snowfall. The rocks in the area fall within the Central Himalayan Zone or Central Crystallines (Heim and Gansser, 1939). With increasing elevation the soil texture becomes finer, especially above 2000 m and a weak podzolization may occur at and around timberline (Singh and Singh, 1992).

Methods
A short reconnaissance trip for 3-4 days was made in the beginning in Chenab Valley and the adjacent forests of Thang village. The characteristic forest types were identified and within each identified forest different sites (1 hectare each) were selected for further data collection based on the extent of forest area. The site specific characters such as altitude, latitude, longitude, aspect, slope, canopy cover and condition of litter and soil were recorded. The data collection was done through systematic random sampling by laying quadrats (Misra, 1968). In each site ten, 10×10 m random quadrats were laid for the enumeration of trees (individuals >31.5 cm circumference at breast height (cbh)). Within each 10×10 m quadrat, 5×5 m quadrat for saplings (10.5-31.4 cm cbh at 1.37 m) and shrubs and four, 1×1 m quadrats for herbs were laid. The density, frequency and total basal area was calculated (Misra, 1968;Muller-Dombois and Ellenberg, 1974), however, Importance Value Index (IVI) was calculated by summing up the relative values of density, frequency and total basal area (Curtis, 1959). The Principal Component Analysis (PCA) was performed to explore the patterns of forest communities stretch and strain are now discernible in its inhabitants as also the whole environment (Chadha, 1998).
The Nanda Devi Biosphere Reserve (NDBR) in the western Himalayas is one of the important protected areas in India, has two core zones, namely Nanda Devi NP and Valley of Flowers NP. Both the NPs established in 1982 and designated as World Heritage Sites by UNESCO in 1988 and2005, respectively in the recognition of their biological and cultural diversity. An assessment of the present status of vegetation in the region is important as it will not only enable us to assess the conservation aspects of this unexplored valley, which have recently been included in Nanda Devi Biosphere Reserve, but also to understand the likely impacts of developmental activities in the area for future management of this very important buffer zone.

Study area
The study area is located in Chenab valley, the buffer zone of Nanda Devi Biosphere Reserve (NDBR) in Chamoli District of Uttarakhand. The Chenab valley and adjacent areas of village Thang are contiguous with Urgam Reserve Forest in southeast and Badrinath Forest Division in northeast. Joshimath Forest Division of NDBR and the area specified for Vishnuprayag Hydel Project are in the eastern boundary of Thang village (Fig. 1). The village Thang, which is spread over the entire study area (ca. 20 km 2 ) with its six winter hamlets, viz. Mulia (2050 m), Malia (2100 m), Kanakot (2200 m), Guar (1700 m), Darun-Ghiwani (2700 m) and Danedar (2500 m) is inhabited by 254 families with 1700 souls. According to Champion and Seth (1968) the present study area harbours forests viz., evergreen broadleaved, deciduous broadleaved, coniferous, mixed broadleaved-coniferous and sub-alpine forests. The valley and adjacent forest area rich in endan- Fig. 1. Location of study area in Nanda Devi Biosphere Reserve, Uttarakhand and the species distribution following the PAlaeontological STatistics (PAST) program (Hammer, 2002) and the data matrix used for PCA was IVI values of tree species at each site.
The species diversity was determined by using Shannon-Wiener information function (H´, Shannon and Wiener, 1963) and richness (Menhinick's Index) as given in Maguran (1988). The beta (β) diversity is computed to measure the rate of species change across the stands (Whittaker, 1960).
β diversity = Sc/s Where, Sc is the total number of species encountered in the entire study area and s is the average number of species per stand.
The data collected for circumference was pooled for deciduous and evergreen species and girth classes were made arbitrarily to know the regeneration status following Ralhan et al. (1982) and Rikhari et al. (1991).

Community composition
A total of 42 sites were randomly selected in Chenab valley of Nanda Devi Biosphere Reserve. Eight forest communities were identified through Principle Component Analysis (PCA; Fig. 2) and some of them are shown in Photo 1. Although, the overlap among vegetation types and also various plant associations were noticed in the ordination space. The first axis explained 38.82% of the variance and the second axis 11.95%. The first five major forest communities of the study area are shown with 95% ellipses. However, two shrub communities were also recorded from the study area, which were dominated by Berberis and Chimnobambusa, respectively.
The forest types are separated based on altitudinal gradient and association of tree species (Fig. 2a). It was noticed that eight major forest types occupied the study area viz. alder mixed forest, tilonj oak-mixed forest, walnut-mixed forest, mixed forest, kharsu oak forest, kharsu oak-yew forest, kharsu oak-birch forest and kharsu oak-mixed forest. Among these forests the maximum similarity was between kharsu oak dominated forests, while tilonj oak-mixed forest and alder mixed forest similarity was quite low. The species contributing more for tilonj oak-mixed forest were A. pictum, C. jacquemontii, L. umbrosa, A. indica, C. viminea, L. ovalifolia and C. australis; for alder-mixed forest the species like R. arboreum, C. deodara, P. pashia, A. pindrow, Q. leucotrichophora and A. caesium were the major contributor, however, kharsu oak forms the forest with T. wallichiana and B. utilis separately in subalpine region (Fig. 2b).

Alder-mixed forest
The total tree density of alder-mixed forest was 343 trees ha -1 , of which 26% contributed by Alnus nepalensis (Tab. 1). The total basal area of the forest was 23.9 m 2 ha -1 and Q. leucotrichophora contributed the maximum (37%) followed by A. nepalensis (26%; Tab. 1), while the IVI value of Alnus nepalensis wasmaximum (79.9). The total snag density of the forest was 33.3±13.3 trees ha -1 . The total tree sapling density of the forest was 13.3 individuals ha -1 and Ficus rumphii was the only species (Tab. 2) and the total tree seedling density was 1047 individuals ha -1 , of which A. pindrow contributed the maximum (39%) followed by C. deodara (33%). The total shrub density was 4760 individuals ha -1 , of which Reinwardtia indica accounted for 19% followed by Rubus ellipticus (18%). The total herb density of the forest was 24 individuals m -2 and Fragaria and Galium contributed the most (Tab. 2).

Tilonj oak-Mixed forest
The total tree density of tilonj oak-mixed forest was 490 trees ha -1 , of which Q. floribunda contributed 22% and the total basal area was 53.0 m 2 ha -1 , of which 39% contributed by Q. floribunda (Tab. 1). The dominant tree species was Q. floribunda with maximum IVI (82) followed by Aesculus indica (43). The total tree sapling density of the forest

Mixed oak forest
The total tree density of the forest was 483 trees ha -1 , of which 19, 17 and 14% contributed by Q. semecarpifolia, Q. floribunda and A. pictum, respectively (Tab. 1). The total basal area was 73.2 m 2 ha -1 , of which 31% and 27% contributed by Q. semecarpifolia and Q. floribunda, respectively. The dominant tree species of the forest were Q. semecarpifolia and Q. floribunda with IVI values 72 and 63, respectively. The snag density of the forest was 27±13.5 trees ha -1 . The total tree sapling density was 395 individuals ha -1 and T. wallichiana contributed the maximum (42%). The total tree seedling density was 2053 individuals ha -1 , of which T. wallichiana contributed 44%. The total shrub density was 4307 individuals ha -1 and Indigofera and Cotoneaster contributed the maximum (25% each; Tab. 2). The total herb density was 44 individuals m -2 , of which Bergenia contributed 14%.

Kharsu oak-mixed forest
The total tree density of the forest was 508 trees ha -1 , of which 22% contributed by Q. semecarpifolia and fol-was 432 individuals ha -1 , of which Litsea umbrosa contributed 43% and the total tree seedling density was 582 individuals ha -1 , of which A. pictum contributed 31% and followed by Q. floribunda (21%; Tab. 2). The total shrub

Walnut-mixed forest
The total tree density of the forest was 410 trees ha -1 and all the species contributed equally to the density. The total basal area was 25.2 m 2 ha -1 , of which 52% contributed by J. regia (Tab. 1). The dominant tree species was Juglans regia with maximum IVI (105). The snags were totally absent in the forest. The total tree sapling density was 1360 individuals ha -1 , of which Litsea umbrosa occupied 60% (Tab. 2). The total tree seedling density was 2080 individuals ha -1 and Juglans regia contributed 77%. The total shrub density was 7240 individuals ha -1 , of which Cannabis sativa contributed 36% followed by Chimnobambusa falcata (29%). The total herb density was 59 individuals m -2 and dominated by Potentilla fulgens (24%; Tab. 2).

Kharsu oak-Yew forest
The total tree density of the forest was 210 trees ha -1 , of which 48% contributed by Q. semecarpifolia and followed by Taxus wallichiana (46%). The total basal area was 41.4 m 2 ha -1 and Q. semecarpifolia contributed 79%. The dominant tree species was Q. semecarpifolia with maximum IVI (174; Tab. 1). The snag density of the forest was 93±11 trees ha -1 . The total tree sapling density was 752 individuals ha -1 , of which 40% and 32% contributed by Taxus wallichiana and Q. semecarpifolia, respectively. The total tree seedling density of the forest was 1870 individuals ha -1 , of which Taxus wallichiana contributed 75%. The total shrub density was 7450 individuals ha -1 and dominated by Cotoneaster (28%; Tab. 2). The total herb density of the forest was 43 individuals m -2 and dominanted by Iris kumaonensis (32%).

Kharsu oak-Birch forest
The tree density of kharsu oak-birch forest was 203 trees ha -1 , of which Betula utilis contributed 51% followed by Q. lowed by R. arboreum (21%; Tab. 1). The total basal area was 63.0 m 2 ha -1 , of which Q. semecarpifolia contributed 47%. The dominant tree species was Q. semecarpifolia and had maximum IVI (91). The snag density of the forest was 42±18 trees ha -1 . The total tree sapling density was 328 individuals ha -1 , of which 46% contributed by Q. semecarpifolia. The total tree seedling density was 832 individuals ha -1 , of which 30% contributed by J. regia and 28% by Q. semecarpifolia (Tab. 2). The total shrub density was 6024 individuals ha -1 , of which Chimnobambusa falcata contributed 36%. The total herb density was 25 individuals m -2 and maximum contribution was 20% by Thymus serpyllum.

Kharsu oak forest
The total tree density of the forest was 464 trees ha -1 , of which 26% contributed by Q. semecarpifolia. The total basal area was 58.7 m 2 ha -1 and Q. semecarpifolia contributed 60% (Tab. 1). The dominant tree species was Q. semecarpifolia with maximum IVI (117). The snag density of the forest was 30±21 trees ha -1 . The total tree sapling density was 513 individuals ha -1 , of which 47% contributed by Q. semecarpifolia and followed by J. regia (39%). The total tree seedling density was 2584 individuals ha -1 , of which J. regia contributed 40% followed by Q. semecarpifolia  su oak forest and birch forest (1.0). However, the zonal β diversity was highest in upper temperate (5.9) followed by sub-alpine (4.0) and lower temperate (1.9). The β diversity for the entire Chenab valley (1750-3060 m) was 7.4.

Discussion
The alder forest seems to be an early seral stage along riverine areas (Mohan and Puri, 1954), however, silver fir with some broadleaved species seems to be a late seral stage in one of the forest of the present study area in the higher altitudes, appearing as the climax and similar association is also suggested (Parker, 1942). This is mainly due to the fact that these areas become nitrogen depleted and alder is the only species to grow along riverine areas and establish themselves due to their ability to fix nitrogen (Sharma and Ambasht, 1988). In the present study, the density of alder forest is 4-5 times lower than the values reported for Pindari region, Kumaun (Adhikari et al., 1991). The density values of tilonj oak-mixed forests of present study area ranged from 350-720 trees ha -1 , which are on the higher side of the values reported for Dharamganga, Asiganga, Bhatwari and Dogadda watersheds (240-370 trees ha -1 ) by Adhikari and Rawat (2004), however, the values (760-1107 trees ha -1 ) reported for Kumaun region by  are higher than the reported values of present study. The density values of kharsu oak dominated forest in present study ranged from 264-545 trees ha -1 , which is higher than that of reported for kharsu oak by Adhikari and Rawat (2004) in Bhatwari and Asiganga watersheds (223 and 230 trees ha -1 , respectively). The density values reported for Dharamganga (362 trees ha -1 ) and Dogadda (372 trees ha -1 ) catchments by Adhikari and Rawat (2004), timberline kharsu oak forest (340 trees ha -1 ) in Tungnath by Rai et al. (2012),) and for Pindar catchment (480 trees ha -1 ) by Adhikari et al. (1995) are comparable with the present study forests and lower than that of the values reported for Tungnath by Rai et al. (2012) at subalpine region (810 trees ha -1 ).
The density values of mixed broadleaved forest in Pindari region of Kumaun (360-640 trees ha -1 ) reported by Adhikari et al. (1991) are higher than that of the present study and comparable with conifer-broadleaved forest (413 trees ha -1 ) of Sakteng WS (Adhikari, 2005). The density values of kharsu oak-birch forest (203 trees ha -1 ) of present study is comparable with the values (205 trees ha -1 and 238-250 trees ha -1 , respectively) reported for Gangotri region by Adhikari and Rawat (2004) and Nanda Devi National Park by Adhikari (2004). However, the density value of birch forest (700 trees ha -1 ) in Pindar catchment, Kumaun by Garkoti and Singh (1994) and in Sakteng WS, Bhutan (573 trees ha -1 ) by Adhikari (2005) are on the higher side of the range reported for present study. The density values reported by Adhikari (2005) in Sakteng Wildlife Sanctuary for oak-conifer forest (144 trees ha -1 ) and oak forest (270 trees ha -1 ) are on the lower side of the semecarpifolia (49%) and the total basal area was 42.6 m 2 ha -1 , of which 67% contributed by Q. semecarpifolia (Tab. 1). The dominant tree species was Q. semecarpifolia with maximum IVI (168). The snag density of the forest was 63±35 trees ha -1 . The total tree sapling density was 313 individuals ha -1 and the contribution of Q. semecarpifolia was maximum (65%). The total seedling density was 333 individuals ha -1 and dominated only by Q. semecarpifolia (Tab. 2). The total shrub density was 10240 individuals ha -1 and dominated by Juniper macropoda and Skimmia laureola (39 and 30%, respectively). The total herb density of the forest was 31 individuals m -2 and Morina and Iris contributed the maximum (26% each).

Berberis community
The total shrub density of the community was 2440 individuals ha -1 , of which Berberis chitria contributed 79% (Tab. 2). The total herb density of the community was 20 individuals m -2 and Oxalis corniculata was the dominant species (44%).

Bamboo community
The total shrub density of bamboo community was 5900 individuals ha -1 (Tab. 2) and Chimnobambusa falcata is the only species.

Species richness and diversity
The tree species richness was highest in alder mixed forest (9) followed by tilonj oak-mixed forest, mixed oak forest and kharsu oak mixed forest (7 in each), while lowest species richness was in kharsu oak-birch forest (2; Tab. 3). The tree diversity ranged from 0.693 (kharsu oak-birch forest) to 1.909 (alder mixed forest, Tab. 3). The shrub species richness was highest in tilonj oak-mixed forest (18) and lowest in bamboo community (1). The shrub diversity was highest in tilonj oak-mixed forest (2.683) followed by alder mixed forest (2.035) and lowest in Berberis community (0.834, Tab. 3). The herb species richness was highest in tilonj oak-mixed forest (20) followed by kharsu oak forest (16) and alder-mixed forest (15) and lowest in Berberis community (4). The distribution pattern of trees, shrubs and herbs shows that species were evenly distributed in most of the sites (Tab. 3).
The β diversity, a measure of rate of species change was observed and it was in following order: tilonj oak-mixed forest (3.9) > kharsu oak forest (2.8) > alder mixed forest (2.5) > kharsu oak mixed forest (1.8) > mixed forest (1.7) > kharsu oak-yew forest (1.3) > walnut-mixed forest, khar-shrub density values reported by Adhikari (2005) for oakconifer forest and conifer-broadleaved forest (6497 and 6661 shrubs ha -1 , respectively) of Sakteng WS, Bhutan are also comparable with the present study forests, while the values of oak forest (1496 shrubs ha -1 ) are on the lower side of the range reported for present study forests. A comparative account of density is given in Tab. 4.
The total basal area in alder mixed forest of present study area ranged from 14-34 m 2 ha -1 , which is quite lower than the basal area reported for alder forest in Pindar catchment, Kumaun Himalaya (Adhikari et al., 1991). The total basal area of tilonj oak-mixed forest in the present study ranged from 40-70 m 2 ha -1 , which is comparable with the total basal area reported for tilonj oak forest in Dharamganga, Bhatwari and Dogadda watersheds (49-65 m 2 ha -1 ) by Adhikari and Rawat (2004) and also comparable with range reported for the present study. The density values reported for P. wallichiana -C. deodara and A. spectabilis-Q. semecarpifolia forests (2100 trees ha -1 and 2090 tree ha -1 , respectively) by Kunwar and Sharma (2004) are quite higher than the values reported for present study.
The total density of shrubs in present study forests ranged from 1960 (alder mixed forest) -23120 shrubs ha -1 (kharsu oak forest). The shrub density of all the forests of present study is more or less similar, except tilonj oak-mixed forest, kharsu oak forest and kharsu oak-birch forest, where upper limit of density is too high. These values are comparable with the range reported for mixedbroadleaved forest and silver fir-tilonj oak mixed broadleaved forest (1880-6090 shrubs ha -1 and 6630 shrubs ha -1 , respectively) by Adhikari et al. (1991) and cypress forest (1280-4170 shrubs ha -1 ) by Adhikari et al. (1998). The Tab. 4. Comparative account of tree density (tree ha -1 ) and total basal area (m 2 ha -1 ) from different study sites in west Himalayan region  (2005)  1180 11.7 Garkoti and Singh (1994) wallichiana-C. deodara and A. spectabilis-Q. semecarpifolia forests (90 m 2 ha -1 and 152 m 2 ha -1 , respectively) reported by Kunwar and Sharma (2004) are quite higher than that reported for present study. A comparative account of total basal area of different forests is given in Tab. 4. The polynomial regression shows that the density of trees decline (r 2 = 0.52, p<0.001) along an altitudinal gradient (Fig. 3a), while there is no correlation between altitude with density of shrubs and herbs. No correlation was observed between the density of trees, shrubs and herbs. The polynomial regression shows that the total basal area of trees decline (r 2 = 0.24, p<0.01) along an altitudinal gradient (Fig. 3b). A significant positive correlation (r 2 = 0.46, p<0.001) was observed between the total basal area and density (Fig. 3c). Similar observations were also made for the forests of Kumaun Himalaya (Adhikari et al., 1991;Rikhari et al., 1997;Singh et al., 1994).
As it is reflected by the β diversity values that these forests form the lower and upper transition zones in the present study area, are comparable with the values reported the range reported for Kumaun Himalaya (34-71 m 2 ha -1 ) by Singh et al. (1994). However, the total basal area values for tilonj oak-mixed forest reported by Adhikari and Rawat (2004) in Dharamganga and Asiganga watersheds (34-36 m 2 ha -1 ) are lower than that of present study forest.
The total basal area of kharsu oak forest in the present study ranged from 42-78 m 2 ha -1 , which is higher than that reported by Adhikari and Rawat (2004) for Dharnmaganga, Bhatwari, Asiganga and Dogadda watersheds (40, 41, 36 and 36 m 2 ha -1 , respectively), however, comparable with the values reported by Adhikari et al. (1995) for kharsu oak forest in Pindar catchment, Kumaun Himalaya (76 m 2 ha -1 ) and subalpine kharsu oak forest by Rai et al. (2012) in Tungnath (65.2 m 2 ha -1 ). The total basal area of kharsu oakbirch forest is on the higher side of the range reported for present study forest than that of birch forest by Adhikari (2004) in Nanda Devi NP (13-16 m 2 ha -1 ), Adhikari and Rawat (2004) in Gangotri watershed (5 m 2 ha -1 ) and Singh et al. (1994) in Pindar catchment, Kumaun Himalaya (23 m 2 ha -1 ), as these are pure birch forests. The total basal area values of present study forests ranged from 18-72 m 2 ha -1 are comparable with the values reported for the forests of Kumaun Himalaya (23-61 m 2 ha -1 ) by Rikhari et al. (1991) and 21-84 m 2 ha -1 for tropical and temperate forests of the world by several workers (Dabel and Day, 1977;Duvugneaud and Denaeyer De-Smet, 1970;Franklin et al., 1979;Reiners, 1972;Rochow, 1972). Similarily, the total basal area reported by Bhandari et al. (1997) for temperate forests of Garhwal Himalaya (15-60 m 2 ha -1 ) is comparable with the present study forests, while the density (320-2080 trees ha -1 ) is quite higher than that reported for present study (203-545 tees ha -1 ). The total basal area values for P. Fig. 3. Population structure of different deciduous (a) and evergreen (b) species and pattern of regeneration in the entire study area proliferated in the area. The broken branches of Taxus due to heavy snowfall have led the formation of canopy gaps, which has supported bamboo growth. The alpine meadows above timberline are reservoir of many ecologically and economically (medicinal) important species, which needs adequate research and management. The impact of climatic variability can be seen by the invasion of Betula utilis and Juniperus macropoda in the meadows. The meadows like Selamghetta, Darkharak and Pilpar have added aesthetic value to the valley and showed a drastic change in plant communities and their composition. Due to heavy anthropogenic pressure in these meadows and nearby forests, which has led the habitat destruction, imposed a threat to the endangered plant species as well as the wildlife of the area. Therefore, it is felt that the livestock grazing can be controlled through the regulatory mechanism i.e. rotational grazing, as it has been a major pressure in the past, to conserve the biodiversity and sustainability of the area with the help of State Forest Department.