1 Introduction

Global biodiversity is under massive pressure due to changes in land-use patterns, climatic scenarios and socio-economic components (Keenan et al. 2015). Extreme developmental activities over the last few decades have adversely affected the forest ecosystem and left behind forests that are highly fragmented and noncontiguous (Mir et al. 2020), which hinders the patterns of plant dispersal, encourages genetic drift and induces invasion of alien species (Laurance et al. 1998). Anthropogenic activities have escalated the rate of habitat destruction, according to a several research (Kardol and Wardle 2010). In general, increasing population and pollution coupled with climate change throughout the world had severely affected the global biodiversity. Several man-made activities also altered the diversity and distribution of many species of Himalayas (Dawson et al. 2011). Unsustainable extraction (for fuel and fodder) from the forests, establishment of Hydro-project stations, rapid urbanization, over grazing in the core areas of a reserve forest, forest fire, etc. resulted in the sharp decline of biodiversity in the Indian Himalayan region (IHR). These reasons also influence the growth, distribution and regeneration of Quercus oblongata.

In the IHR, there are some studies on ecology, floristic diversity, species composition and distribution, forest vegetation had been done so far (Dhar et al. 1997; Samant et al. 1998; Joshi and Samant 2004; Samant and Joshi 2004; Pant and Samant 2012; Sharma and Samant 2013; Devi et al. 2019; Lal and Samant 2019; Barman et al. 2021). However, only a few studies concern regarding population ecology and habitat distribution modelling (Barik and Adhikari 2012; Adhikari et al. 2012, 2018; Yang et al. 2013; Lal and Samant 2015, 2017, 2019; Paul et al. 2018; Lal et al. 2020) in the northwestern Himalayas had been carried out. The fact is that sufficient and adequate datasets on the Chamba district’s current biodiversity are either unavailable or dispersed which is a matter of concern. Thus, conservation of depleting populations of Ban oak is a social-ecological challenge and requires an interdisciplinary approach. Therefore, a quantitative assessment of forest vegetation is essential for understanding the forest ecosystem structure, composition, and diversity, and for developing critical forest management and biodiversity conservation strategies with policies to protect the environment (Condit et al. 1998).

Himalayan ecosystems are known for the representative, natural and unique forests (Singh et al. 2021), supporting many ecologically and economically important multipurpose tree species (Paul et al. 2018). The northern Himalayas are a biodiversity hotspot comprising diverse vegetation types (Champion and Seth 1968) ranging from tropical moist deciduous forest to temperate and sub-alpine forests, grasslands, alpine scrub, and meadows (Marchese 2015; Roy et al. 2015). Temperate broadleaved evergreen forests are an important natural resource and a major component of sustenance of mountain natives of Himachal Pradesh, which is especially dominated by Oak. Quercus oblongata D. Don (commonly known as Ban oak) in Himachal Pradesh is a medium sized (Joshi and Juyal 2017), evergreen and keystone tree species (Dhyani et al. 2020) of Fagaceae family found between 1000 and 2500 m amsl (Rana et al. 1989; Singh and Singh 1986) with high abundance around 2000 m amsl (Singh et al. 2016a, 2016b) and dominant, climax tree species in the moist temperate forests of northwestern Himalaya (Troup 1921) (Fig. 1). Forests of Ban Oak are rich in soil organic matter, have high biodiversity, source of fuel wood and fodder, water holding capacity, and provide ecosystem services (like provisioning, regulating, supporting, etc.) that supports human well-being of mountain natives (Dhyani et al. 2020).

Fig. 1
figure 1

(Source: GBIF)

Distribution of Q. oblongata throughout the Himalayan Region

Full size image

Forest regeneration is a fundamental feature of forest composition and structure of the species. The presence of sufficient number or populations of seedlings, saplings, and mature plants signifies successful regeneration of forest species (Singh et al. 2021; Dasila et al. 2020). Presence of seedlings and saplings under a canopy of mature forest vegetation also indicates the future structure of that forest community (Austin 1977). The size class distribution of populations within a structure allows determination of regeneration, distribution, and future stability of tree species population in forest communities (Singh et al. 2021). Determining the distribution of seedlings, saplings, and adults, along with analysis of size class distributions, is a part of regeneration analysis and helps understand changes in soil parameters because nutrients in the soil are dependent on the vegetation structure and functioning of the forest (Han et al. 2011). Thus, a variation in soil nutrients may lead to an alteration in forest structure (Singh et al. 2021).

Population assessment, conservation prioritization and species reintroduction are the successful ecological techniques for recovering and maintaining the position of dwindled species populations, degraded habitats and ecosystems (Kuzovkina and Volk 2009; Zai et al. 2009; Rodriguez-Salinas et al. 2010; Polak and Saltz 2011; Adhikari et al. 2012). Ecological Niche Modelling (ENM) helps in identifying the suitable habitats, i.e., very high, high, medium and low potential areas for a species to grow in a geographic region and thus plays a key role in in situ conservation and ecosystem restoration (Adhikari et al. 2012; Singh et al. 2021). In view of this, studies are being carried out in other parts of the Indian Himalayan Region (IHR) (Lal and Samant 2015, 2017, 2019; Chettri et al. 2018; Sharma et al. 2018; Paul et al. 2018; Lal et al. 2020).

Climate change is identified as the influencing factor in shaping the future of any forest vegetation. Climate has a huge influence in forest ecosystem (Chaturvedi et al. 2011; Birch 2014; Upgupta et al. 2015; Chakraborty et al. 2018; Devi et al. 2018; Kumar et al. 2019) where changing climatic conditions is one of the most powerful drivers (Dhyani et al. 2020) and major concern of biodiversity loss affecting mountainous forest ecology (Dale et al. 2001; Ellison et al. 2017; Seidl et al. 2017). In general, studies focusing on the composition, structure, and function of forests had been done so far throughout the northwest Himalaya (Kalakoti et al. 1986; Singh and Singh 1986; Dhar et al. 1997; Joshi and Samant 2014; Rawat et al. 1999; Pant and Samant 2008, 2012; Rana and Samant 2009, 2010; Sharma and Samant 2013; Devi et al. 2019; Lal and Samant 2019; Barman et al. 2021). However, focused studies on ban oak populations have not been carried out so far in Himachal Pradesh. Therefore, in view of the various anthropogenic pressures and changing climate scenario, we carried out population assessment of ban oak forests to identify environmentally suitable areas for reintroduction and guide government policies to protect this species.

2 Materials and methods

Study area

– Chamba district is extremely rich in flora and fauna due to its varied topography and climate and situated between north latitude 32° 11′ 30″ and 33° 13′ 06″ and east longitude 75°49′ 34″ and 77° 03′ 30″, with an estimated area of 6,522 km2 and is surrounded on all sides by lofty hill ranges (Fig. 2). The territory is wholly mountainous with altitude ranging from approximately 600–6300 m amsl. Chamba the land of lord Shiva is famous for its untouched natural beauty. The district has Dalhousie, Khajjiar, Chamba Town, Pangi and Bharmour as main tourist destinations. There are five lakes, five wild life sanctuaries and countless number of temples. The summer season in Chamba starts from the middle of April and lasts till the last week of June. Rains in Chamba start in the month of July, when the monsoon breaks-in, and continue till late August or mid-September. The winter season in Chamba starts in the month of December and lasts till the month of February. The vegetation mainly comprises of sub-tropical, temperate, sub-alpine and alpine types.

Fig. 2
figure 2

Map of the study area

Full size image

Site selection, surveys, sampling, identification, and data analysis

– The sites were selected and field surveys were conducted on each and every accessible aspect along the topographical gradient between 812 and 2266 m amsl in the ban oak forests. Information on the habitat type and site characteristics, i.e., altitude, latitudes, and longitudes of each sampled plot were obtained with the help of Global Positioning System (Garmin Montana 650). The slope and aspects were measured by Clinometer (some sites by Abney Level) and compass, respectively. The habitats were identified based on the physical characteristics and dominance of the vegetation. Sites having a closed canopy with high proportions of humus were considered shady habitats, and sites having high content of moisture as moist habitat; with boulders on > 50% of the ground cover as bouldery habitats; containing rocks as rocky habitat; and along the streams and riverbanks riverine habitat (Samant and Joshi 2004). The field surveys and samplings were conducted in the month of May 2019.

A plot of 50 × 50 m was laid. Trees, saplings and seedlings were sampled randomly by 10 quadrats of 10 × 10 m, shrubs by 20 quadrats of 5 × 5 m and herbs by 20 quadrats of 1 × 1 m size. For the collection of data from these quadrats standard ecological methods (Curtis and McIntosh 1950; Greig-Smith 1957; Kershaw 1973; Mueller-Dombois and Ellenberge 1974; Dhar et al. 1997; Joshi and Samant 2004). Species diversity (H') was calculated by using the Shannon–Wiener information index (Shannon and Wiener 1963). The concentration of dominance (Cd) was calculated using Simpson’s index (Simpson 1949). The total count of species was considered as species richness.

According to field data, tree individuals were grouped into arbitrary Circumference at Breast Height (CBH) classes. The individuals with CBH of 30.5 cm or more were considered as trees, those with CBH of 10.5–30.4 cm as saplings, and those with CBH < 10.5 cm as seedlings (Dhar et al. 1997). Shrubs were considered as woody species having several branches arising from their base, and herbs were considered as those species having aerial parts surviving for one season, though their underground parts, i.e., roots/rhizomes/bulbs, etc. may remain alive during other seasons (Joshi and Samant 2004).

Three vegetation layers, i.e., tree, shrub and herb were analyzed for species richness, diversity, frequency, density, abundance, basal area (BA), total basal area (TBA), important value index (IVI) and regeneration of tree species. The abundance data of different sites were pooled to get community averages in terms of density, total basal area and IVI. Data analysis was done following Kersaw (1973); Dhar et al. (1997) and Samant et al. 2002. The size class distribution of trees, saplings, and seedlings of Q. oblongata was also done based on the CBH range of the tree (Singh et al. 2021). For the size class distribution profile of Q. oblongata, 12 CBH classes were identified, namely, < 10.5 cm, 10.5–31.4 cm, 31.5–52.5 cm, 52.6–73.5 cm, 73.6–94.5 cm, 94.6–115.5 cm, 115.6–136.5 cm, 136.6–157.5 cm, 157.6–178.5 cm, 178.6–199.5 cm, 199.6–220.5 cm and ≥ 220.6 cm. These values were used to predict the future occurrence or dominance of the species.

Assessment of Physico-chemical properties of Soil

– Soil samples were collected from each studied plot of 50 × 50 m. Soil was cored up to 20 cm depth. Five soil samples, four from the corners and one from the center of each plot were collected, pooled and mixed properly to make a composite sample. The air-dried soil samples were assessed for further tests and analysis. Soil pH was measured using pH meter in 1:5 mixture of soil and distilled water, moisture content was recorded as percentage (%) difference in fresh and dry soil weight, % organic carbon and organic matter were analyzed as described by Walkley and Black method (Walkley and Black 1934), available nitrogen by Kjeldahl method (Subbiah and Asiijah 1956), available phosphorus by Olsen’s extraction method (Olsen et al. 1954) and available potassium by flame photometer (Allen et al. 1974; Jackson 1958). By plotting all of the data along with the first two principal components, principal component analysis (PCA) was used to extract the most important information from the multivariate dataset.

Species distribution/habitat modelling

– Q. oblongata is one of the dominant species in northwestern Himalaya, and hence, the distribution of species is very wide in the state having more dominance in 7 districts (namely Chamba, Mandi, Kullu, Shimla, Solan, Sirmour and Dharamshala region of Kangra). The habitat modelling (Fig. 3) was performed assuming that the species habitat suitability is influenced significantly by the climatic data (abiotic conditions) derived from temperature and precipitation. Fourteen primary distribution records of Q. oblongata were collected through field surveys. The coordinates of all occurrence points were recorded in decimal degree to an accuracy range of 3–5 m using a Global Positioning System.

Fig. 3
figure 3

The methodology of eco-distribution mapping based on MaxEnt modeling

Full size image

MaxEnt (Maximum Entropy Modelling) version 3.4.3 was downloaded from the portal (https://biodiversityinformatics.amnh.org) and used for modelling (Phillips et al. 2006) with a set of 19 bioclimatic variables (Phillips and Dudik 2008). The climatic variables used to construct our ENM were defined by Terribile et al. (2012) from the factorial analysis of 19 bioclimatic variables based on a correlation matrix among them in which were chosen variables that reduced problems of collinearity. These variables were mean diurnal range, temperature seasonality, mean temperature of driest quarter, precipitation of driest month, precipitation seasonality and precipitation of warmest quarter. ENMs do not differ significantly between different global circulation models (GCM) (Barbosa et al. 2019). In this sense, the variables were estimated for the community climate system model 4 (CCSM4) which was downloaded from World Clim portal (http://www.worldclim.org/current) having 30 arc sec (equivalent to 900 m at equator, ⁓ 1 km) resolution for the present and two future periods averaged at years 2050 and 2070.

The bioclimatic variables (Supplementary table S3) were derived from the surface temperature and precipitation observations which were widely used as environmental predictor variables in species distribution model (SDM) studies. The MaxEnt model fits a probability density function based on the values of environmental variables (bioclimatic and topographic data) observed at the species occurrence locations. Topographical data such as altitude and aspect were derived from the Shuttle Radar Topographic Mission (SRTM). Elevation data for a digital elevation model (DEM) of 30 m resolution were downloaded from the US Geological Survey (USGS) portal (https://earthexplorer.usgs.gov). Remotely sensed data on bioclimatic variables (Hijmans et al. 2017) in highest resolution (i.e., ⁓ 1 km) were downloaded in BIL format and converted to ASCII raster grids with the help of ArcGIS version 10.1. Resampling DEM was used to match the bioclimatic data’s coordinate reference system and resolution. MaxEnt generates response curves for each predictor variable. In highlighting the relative influence of each predictor variable, the Jackknife method was used (Fielding and Bell 1997). Before running the model, the variables were checked for multi-collinearity to avoid bias (Dormann et al. 2007; Elith et al. 2010). High correlations (> 0.70) between bioclimatic variables would indicate that datasets have a similar broad scale spatial pattern. In addition, taking all default variables could lead to spurious results or obscure the effect of the most important variables (Supplementary table S4).

Records used to build the ENM were split randomly into two groups: one for calibration with 75% of the data and the other for evaluation with 25%. To validate model robustness, 15 replicated runs were performed for the species with a threshold rule of 10 percentile training presence, which omitted all regions with habitat suitability lower than that for the lowest 10% of the occurrence records. In the replicated runs, a cross-validation technique was employed, where samples were divided into replicate folds, and each fold was used for test data. MaxEnt would estimate the importance of the variables with percent contribution and permutation importance values. Percent contribution would represent how much the variable contributed to the model, based on the path selected for the particular run. The desired values along with the average model results were incorporated in ArcGIS to match the selected threshold using reclassification. Permutation importance was determined by changing the predictor values between presence and background points and observing how they affected the area under the curve (AUC). The permutation importance would include the final performance of the model, rather than the path used in an individual run, which would be useful for evaluating the importance of a particular variable. The contribution of each variable to the habitat model of Q. oblongata was calculated using the built-in software or “Jackknife test,” which was used to determine the dominant bioclimatic factors responsible for the potential distribution of the species (Li et al. 2016).

Based on the AUC values, the quality of the model was evaluated and graded following Thuiller et al. 2005, as poor (AUC < 0.8), fair (AUC 0.8–0.9), good (AUC 0.9–0.95), and very good (AUC 0.95–1.0). The trained MaxEnt model was supplied with future environmental data representing the climate variability in order to project the future habitat suitability maps. Except for bioclimatic variables, the topographic variables were assumed to remain constant for the future projections. Arc GIS version 10.1 were used to generate predictive maps of species distribution and potential areas of distribution and reintroduction. In the present study, 10 percentile training presence logistic threshold was applied to stipulate 10 percentile minimum threshold.

Biomass Estimation

– The diameter at breast height (DBH) for each tree was measured at 1.37 m above the surface. The density of trees under each DBH class was used to determine the total tree biomass value. To estimate the above ground biomass (AGB), the biomass value per tree under different DBH classes were estimated using regression equation of Chambers et al. (2001).

$$\begin{aligned} & {\text{AGB}}:\;{\text{ln }} = \, - 0.{37} + \, 0.{\text{333 lnD}} + 0.{933 }\left[ {{\text{ln}}\left( {\text{D}} \right)} \right]^{{2}} - \, 0.{122 }\left[ {{\text{ln}}\left( {\text{D}} \right)} \right]^{{3}} \\ & {\text{Below}}\;{\text{Ground}}\;{\text{Biomass}}\;\left( {{\text{BGB}}} \right):\;{\text{AGB }} \times \, 0.{29} \\ & {\text{Total}}\;{\text{Biomass}}\;{\text{Density}}\;\left( {{\text{TBD}}} \right):\;{\text{AGB }} + {\text{ BGB}} \\ \end{aligned}$$

Total carbon density (TCD) was estimated by multiplying the factor 0.5 with TBD value.

Numerical and statistical analysis

– All numerical analyses were done in MS-Excel and principle component analysis using paleontological studies (PAST) package.

3 Results

Spatial pattern and distribution

– In total, 14 sites representing 3 habitats and 6 aspects between 812 and 2266 m amsl and with 14–65° slope were sampled from the Chamba district. Most sites (5, each) were represented by dry and shady habitats, followed by moist (4) habitat. Six (6) sites were represented by West aspect, followed by southeast (3), east (2) and northeast, northwest and southwest (1, each) aspects. The slopes varied from 14° to 65°. Physical characteristics, dominant species, aspect, altitude, and slope of the populations of Q. oblongata are presented in Table 1.

Table 1 Site characteristics and associated tree species of Quercus oblongata D. Don populations in the studied populations
Full size table

Species composition of studied populations

– Total 188 species (Trees: 30; Shrubs: 33 and Herbs: 125) belonging to 162 Genera and 66 Families were recorded from 14 sampled sites. Maximum species (88) (Trees: 06, Shrubs: 12 and Herbs: 70) was recorded in Dumas-Tissa, followed by Chaned (81 species; Trees: 13, Shrubs: 11 and Herbs: 57); KKWS-I (70 species; Trees: 06, Shrubs: 08 and Herbs: 56); Banikhet (69 species; Trees: 08, Shrubs: 12 and Herbs: 49) (Supplementary Fig. S1). Dominant genera were Erigeron, Ficus, Persicaria and Quercus (3 species, each); followed by Adiantum, Anaphalis, Arisaema, Aster, Carex, Clematis, Digitalis, Galium, Geranium, Hypericum, Impatiens, Indigofera, Isodon, Pinus, Rubus, Trifolium and Viola (2 species, each). Dominant family was Compositae (19 species); followed by Leguminosae (14 species); Lamiaceae (13 species); Poaceae (12 species); Rosaceae (8 species); Orchidaceae and Polygonaceae (7 species, each); Ranunculaceae and Rubiaceae (5 species, each); Amaranthaceae, Acanthaceae, Moraceae, Pinaceae and Pteridaceae (4 species, each). Monotypic families were Adoxaceae, Apocynaceae, Aspleniaceae, Linaceae, Oleaceae, Onagraceae, Saxifragaceae, Scrophulariaceae, Smilacaceae, Thelypteridaceae, Thymelaeaceae, Verbenaceae, Zingiberaceae.

Population ecology

– Overall, site-wise species richness, total basal area (TBA), species diversity (H'), and concentration of dominance (Cd) of tree, shrub, and herb layers are presented in supplementary table S2. The richness of trees ranged from 2 (Jot and Hatli-Dramman) to 8 (Bhujnoi-Gate); shrubs from 4 (Khajjiar-I and Khajjiar-II) to 12 (Banikhet and Dumas-Tissa. each); herb richness ranged from 29 (Chaned) to 47 (Banikhet). Total trees density varied from 270 Ind ha−1 (Jot) to 750 Ind ha−1 (Banikhet), and the density of Q. oblongata varied between 90 Ind ha−1 (Hatli-Dramman) to 580 Ind ha−1 (Bhujnoi-Gate) (Supplementary Fig. S2). Total shrubs density varied from 1102 Ind ha−1 (Hatli-Dramman) to 2689 Ind ha−1 (Dummas-Tissa), and total herbs density varied from 31 Ind m−2 (Hatli-Dramman) to 85 Ind m−2 (KKWS I). Altitude wise density of trees, saplings and seedlings in studied populations is given in Fig. 4. TBA varied from 11.18m2 ha−1 (Hatli-Dramman) to 55.67 m2 ha−1 (Dummas-Tissa). TBA of Q. oblongata varied between 8.13 m2 ha−1 (Kalatop Khajjiar Wildlife Sanctuary: KKWS-V) and 55.67 m2 ha−1 (Dummas-Tissa). The Cd of Q. oblongata varied between 0.31 (KKWS-IV) and 0.99 (Bhujnoi-Gate). Cd for shrubs ranged from 0.0069 (Jot) to 0.1422 (Dummas-Tissa), and herbs from 0.02 (Dumas-Tissa) to 0.30 (Bhujnoi-Gate). H' of trees varied from 0.07 (Dummas-Tissa) to 0.49 (KKWS-V); shrubs from 0.58 (Khajjiar-II) to 2.16 (Dummas-Tissa); and herbs from 2.21 (KKWS-I) to 4.17 (Dumas-Tissa). The shady habitat represented maximum density (536 Ind ha−1), followed by moist (521 Ind ha−1) and dry (418 Ind ha−1) habitats. TBA was highest in shady habitat (55.32 m2 ha−1), followed by moist (41.99 m2 ha−1) and dry (37.65 m2 ha−1) habitat.

Fig. 4
figure 4

Altitude wise density of trees, saplings and seedlings in studied populations

Full size image

Regeneration pattern

– Total seedlings density ranged from 80 Ind ha−1 (Jot and Hatli-Dramman) to 370 Ind ha−1 (Dumas-Tissa). Seedlings density of Q. oblongata ranged from 40 Ind ha−1 (Jot) to 140 Ind ha−1 (Chaned). Total saplings density ranged from 30 Ind ha−1 (Hatli-Dramman) to 160 Ind ha−1 (Dumas-Tissa). Saplings density of Q. oblongata ranged from 30 Ind ha−1 (Banikhet and Hatli-Dramman, each) to 110 Ind ha−1 (Dumas-Tissa) (Table 2). Seedlings density and saplings density were significantly correlated (p ≤ 0.01, r = 0.63, n = 14) throughout the sites.

Table 2 Population wise total trees density, seedlings and saplings of Quercus oblongata D. Don populations
Full size table

Size class distribution profile of Q. 0blongata

– The size class distribution diagram of Q. oblongata deviated slightly from the reverse J shape structure that is considered to be an indication of good regeneration status. Individuals with smaller diameter class < 10.5 cm (i.e., seedlings) and 10.5–31.4 cm (i.e., saplings) were relatively high in number, but those with diameter class between 31.5–52.5 cm were relatively fewer. Furthermore, the number of individuals in the medium girth class (52.6–73.5 cm and 73.6–94.5 cm) was higher than one would expect in a reversed J distribution. The lowest number of individuals were recorded in the 157.6–178.5 cm girth class (Fig. 5).

Fig. 5
figure 5

Distribution of Q. oblongata individuals under different diameter size classes

Full size image

Physico-chemical properties of soil

– Among the populations, pH was found to be slightly acidic across all elevations due to high litter decomposition and organic matter; it ranged from 6.1 (Banikhet) to 6.9 (Hatli-Dramman) which is considered as satisfactory for the growth of most of the plants in natural ecosystem. Moisture content was maximum in KKWS-IV (41.79%) and minimum in Chaned (13.25%). Total nitrogen of soil was found lowest at Chaned population (1.05%) and the highest at KKWS-V population (5.98%). Soil organic matter varied from 7.86% (Chaned) to 19.91% (KKWS-III); the total organic carbon of soil from 2.51% (Dumas-Tissa) to 9.69% (Banikhet); available phosphorus varied from 13.46 kg ha−1 (KKWS-II) to 71.22 kg ha−1 (KKWS-V); and available potassium varied from 18.42 to 51.13 kg ha−1 (Fig. 6). Except for moisture content and potassium, which were found in the opposite direction, it reveals that most of the soil parameters were varied in similar ways. Population wise PCA plot has indicated that Jot, KKWS II, Chaned, Bhujnoi-Gate being at the right east corner of the ordination plot, has unique set of gradients for dominance and thus does not share its habitat with other species. Similar characteristics were observed for KKWS-III and KKWS-V, which was positioned at eastern side of ordination plot (Fig. 7).

Fig. 6
figure 6

Physico-chemical profile of soil in different populations

Full size image
Fig. 7
figure 7

Distribution of the populations represented by two principal components in PCA

Full size image

Species distribution/habitat modelling

– The output generated by MaxEnt predicted suitability of habitats of the species on the scale from 0 to 1, with the least suitable areas represented by 0, and the most suitable ones represented by 1. The habitat distribution modelling calibration test for Q. oblongata showed satisfactory results. Receiver operating characteristic (ROC) curve displays sensitivity and 1-specificity with AUC for training (0.908) and test (0.885) where average test AUC for replicated runs was 0.825 ± 0.02. Based on permutation importance, mean temperature of wettest quarter (BIO_09) was the most influential and had a permutation importance of 71.40%, followed by DEM, with a permutation importance of 11.47%, mean diurnal range (BIO_02) (7.47%), slope (7.38%), temperature seasonality (BIO_04) (5.97%), precipitation of driest month (BIO_14) (3.75%), precipitation seasonality (BIO_15) (2.38%) and precipitation of warmest quarter (2.03%). The most important variable based on percent contribution was temperature seasonality with a contribution of 43.70% to the habitat model, followed by precipitation of driest month (31.45%), mean temperature of driest quarter (15.64%), DEM (7.50%), mean diurnal range (4.20%), slope (1.70%) and precipitation seasonality (0.36%). Precipitation of warmest quarter (0.03%) had the least influence on the MaxEnt Model (Table 3). The Jackknife plot for comparing the climatic variables in model prediction is depicted in Fig. 8. During field surveys, it was observed that Q. oblongata survived better only in temperate climatic conditions, and that the species preferred precipitation in the coldest quarter and grew mostly in moist habitats between 1600 and 2200 m amsl. This was validated by the model output. In the ROC analysis, performance was measured by the AUC (Fig. 9a–d).

Table 3 Environmental variables and their percent contributions for predicting the potential distribution of Quercus oblongata D. Don
Full size table
Fig. 8
figure 8

Results of Jackknife test highlighting the importance of environmental variables for Quercus Oblongata in Chamba District

Full size image
Fig. 9
figure 9

a Omission rate and predicted area as a function of cumulative threshold; b Test omission rate and predicted area as a function of cumulative threshold averaged over the replicate runs; Model performance; c ROC curve displays sensitivity and 1-specificity with AUC for training (0.908) and test (0.885); and d Average test AUC for replicated runs 0.825

Full size image

Potential habitat distribution area

– In the study area, maximum habitats with high suitability thresholds were distributed in Kalatop Khajjiar Wildlife Sanctuary area, Banikhet and Dumas-Tissa area (Fig. 10). Field surveys revealed that the predicted potential habitats were mostly located in the shady and shady moist habitats of the studied populations in Chamba district. Of the total 6498 km2 area, 1879 km2 (28.91%) area was completely not suitable for natural regeneration or reintroduction while 1247 km2 (19.19%) area was low suitable, 1507 km2 (23.19%) area was moderately suitable and 1865 km2 (28.70%) area with high suitability of reintroduction, natural growth and development of the Q. oblongata population (Supplementary Table S2).

Fig. 10
figure 10

Suitable probability distribution map of Q. oblongata in Chamba district

Full size image

Future projection of climatically suitable areas

– In the CCSM4 model, future projections of habitat suitability maps for Representative Concentration Pathway (RCP) 4.5 and RCP 8.5 (2050 and 2070, respectively) were to some extent close to the present distribution. According to the current findings, if climate change occurs at expected levels (RCP 4.5 and RCP 8.5), the specie’s geographic location will expand relative to its present potential distribution, RCP 4.5 anticipates an increase of 0.4 percent in habitat suitability area by 2050 and 1.4 percent by 2070, respectively, for the high habitat suitability area. 1.5 percent increase is projected in 2050, and 1.1 percent decrease is expected in 2070, according to the RCP 8.5 estimate. However, when compared to the present prediction, the potential high suitability showed higher rate of increase in both scenarios (RCP 4.5 and 8.5) for 2050, after which in scenario (RCP 8.5, 2070), it showed a decreasing trend (Fig. 11).

Fig. 11
figure 11

Potential areas of Q. oblongata’s occurrence and distribution in Chamba district in future IPCC RCPs (Representative Concentration Pathways) for the years 2050 and 2070

Full size image

Biomass estimation

– Total above ground biomass density (TAGBD) ranged from 13.01 to 164.94 Mg ha−1. Maximum TAGBD was found in Banikhet. Above ground biomass density for Q. oblongata (AGBD_QO) was ranged from 10.52 to 110.80 Mg ha−1. Maximum AGBD_QO was found in Dumas-Tissa. Total below ground biomass density (TBGBD) ranged from 3.77 to 61.53 Mg ha−1. Maximum TBGBD was found in KKWS-I. Below ground biomass density for Q. oblongata (BGBD_QO) was ranged from 3.05 to 52.84 Mg ha−1. Maximum BGBD_QO was found in KKWS-I. Total Biomass Density (TBD) ranged from 16.79 to 212.78 Mg ha−1. Maximum TBD was found in Banikhet. Biomass density for Q. oblongata (BD_QO) was ranged from 13.57 to 142.94 Mg ha−1. Maximum BD_QO was found in Dumas-Tissa. Total carbon density (TCD) ranged from 8.39 to 106.39 Mg ha−1. Maximum TCD was found in Banikhet. Carbon density for Q. oblongata (CD_QO) was ranged from 6.79 to 71.47 Mg ha−1. Maximum CD_QO was found in Dumas-Tissa (Table 4).

Table 4 Values of biomass estimation in Chamba district, northwestern Himalaya
Full size table

Correlation of ecological components

– The statistical analysis revealed a positive correlation between density of seedlings and saplings (R2 = 0.5428, r = 0.7367, n = 14, p ≤ 0.05); TBA and altitude (R2 = 0.1214, r = 0.3484, n = 14, p ≤ 0.05); density of trees and seedlings (R2 = 0.1523, r = 0.3902, n = 14, p ≤ 0.05); density of trees and Cd (R2 = 0.1491, r = 0.3861, n = 14, p ≤ 0.05) Total trees density showed positive correlation with density of Q. oblongata (R2 = 0.5599, r = 0.7482, n = 14, p ≤ 0.05). The species richness showed negative correlation along the altitude (R2 = 0.1761, r = −0.4196, n = 14, p ≤ 0.05), that revealed decreased species richness with the increasing altitude attributed to severe climatic conditions at the higher elevations making unfavorable for the survival of various species. Negative correlation observed between seedlings density and altitude (R2 = 0.4204, r = −0.6483, n = 14, p ≤ 0.05) and density of tress with density of shrubs (R2 = 0.1116, r = 0.3340, n = 14, p ≤ 0.05) (Fig. 12a–h).

Fig. 12
figure 12

a Correlation between saplings density and seedlings density; b Correlation between species richness and altitude; c correlation between seedlings density and altitude; d correlation between total basal area and altitude; e correlation between trees density and seedlings density; f correlation between trees density and concentration of dominance; g correlation between trees density and shrubs density and h correlation between total trees density and density of Q. oblongata

Full size image

4 Discussion

Natural resources are essential to human survival, and the northwestern Himalayas are provided with a wide variety of floristic diversity that is significant from an ecological and economic standpoint, including Ban oak. Ban oak’s natural populations are being depleted due to an increase in demand and the unsustainable exploitation for foraging. Structural diversity of forest vegetation, species composition and distribution patterns are the main focus of ecological studies (Rana et al. 2021). Conducting appropriate ecological research and raising public awareness are the need to protect the biodiversity components of any biogeographic area and are essential for the conservation of plant species and for understanding forest dynamics (Barik and Adhikari 2012; Polak and Saltz 2011). One of the most significant indicators that can be used to assess ecosystems at various scales is the floral variety and species composition of a forest community (Rana et al. 2021). Numerous indicators, including the Shannon index, species density, species diversity, and concentration of dominance, may be used to anticipate the complexity of plant communities and comprehend the dynamics of a forest (Rana et al. 2021). Ecological studies provide detailed information on the homeostatic capacity of any forest ecosystems to unforeseen environmental changes (Christopher 2020).

The northwestern Himalaya including Chamba district of Himachal Pradesh is blessed with diverse floristic diversity of ecological and economical importance including Ban oak. Decent species variety within the same forest type is shown by the presence of 188 plants at the floor of ban oak forests in the analyzed populations. This might be attributed to the distinctive geography of the northwestern Himalayas and to environmental factors that are conducive to the species' development (Rana et al. 2020). In comparison to past studies done in the Himalayan zones, the number of species found in the Chamba is lower. Kais Wildlife Sanctuary recorded 607 vascular plant species in total (Lal and Samant 2019), Manali Wildlife Sanctuary reported 637 vascular plant species (Rana and Samant 2009), and Cold Desert Biosphere Reserve reported 354 medicinal plant species (Singh 2007). The threat on ban oak forests is evident from this outcome. The presence of a large number of species in the older research may be attributable to protected regions, where there is less anthropogenic activity.

Species diversity is defined as the number of different species present in an ecosystem, i.e., species richness and relative abundance of each of those species. Because the variation across various communities is influenced by the variety of both the microclimate and the macroclimate, long-term variables like community stability and evolutionary time influence the species diversity (Verma et al. 2004). The study site with more species variation tends to be more stable and resilient than the sites which are dominated by monotypic communities. According to previous findings like Lal and Samant 2019 (H’: 0.27–2.17) and Rana et al. 2021 (H’: 1.40–2.10), the reported range of tree diversity was lower in the present study (0.07–0.49). While for the shrubs it is also reported low (0.58–2.16) comparing other studies (Rana et al. 2021; Rana and Samant 2009; Pant and Samant 2012). This result illustrates the pressure on ban oak forests and urge conservation needs. In case of herbs, species diversity (2.21–4.17) was comparable than the earlier reported values (Samant et al. 2002 (2.74–4.13); Lal and Samant 2019 (2.88–4.26)). Low protection and conservation efforts, unsustainable harvesting, significant human activity, and a lack of adequate soil cover are further factors that contribute to the area's low species diversity.

The study was conducted only in the accessible habitats having < 65° Slope; thus, some sites having accessibility issues. Only representative sites were selected from the district which is mainly dominated by ban oak. The maximum numbers of populations were represented by shady and moist habitat indicating that these habitats form the best platform for the overall development of the species. The high values of basal area for shady moist habitat indicated the availability of sufficient nutrients and the favorable environmental conditions for development. This is due to the sparse distribution of individuals and less competition for nutrients. Distribution of Q. oblongata is mostly found in south aspects which is similar to the trends that are found in other parts of the IHR (Devi et al. 2019; Lal and Samant 2019; Barman et al. 2021) and probably due to the favorable ecological landscapes. The study reveals that the maximum density of Q. oblongata is found at an elevation of 2000 m ± 500 m amsl and clearly states the growth preference of Ban oak along an altitudinal gradient and this growth preference was also recorded by Dhyani et al (2020). These aspects, elevation and habitats are responsible for proper establishment of the species.

Q. oblongata is a dominant species of a highly diverse forest ecosystem with its associate tree species viz., Rhododendron arboreum, Lyonia ovalifolia, Pyrus pashia, Quercus floribunda, Aesculus indica and Neolitsea pallens (Dhyani et al. 2020). The rich diversity of the herbs, shrubs, pteridophytes, and epiphytes is also noticeable in Ban oak forests. The highest H' of trees on moist habitat and southern aspect, along with the highest H' of shrubs in moist slopes and southwestern aspect and the highest H' of herbs in shady habitat and west and southwest aspects, revealed the suitability of these topographic gradients for Q. oblongata in the studied populations, hence require priority attention for conservation. An examination of the dispersion pattern revealed that, in the case of trees, most species exhibited uniform distribution, with a few species exhibiting random distribution (Singh et al. 2016a, 2016b). In general, it has been observed that density, basal area, total basal area, species richness, species diversity and concentration of dominance varied from habitat to habitat and aspect to aspect. The regeneration status of a tree species is determined by the presence of seedlings and saplings in the forests (Joshi and Samant 2004, 2014). The results of correlations are comparable with the previous studies conducted in the other regions of the Western Himalaya (Sharma and Samant 2013).

The capacity for regeneration of a species that make the forest, both spatially and temporally, determines the forests wealth. Old trees die and are continuously replaced by new ones during the regeneration of a forest (Malik and Bhatt 2016). The structure of a forest vegetation is categorized by the presence of sufficient number of seedlings, saplings, and young trees. Adequate number of seedlings and saplings depicts “good” regeneration trend, inadequate number of seedlings and saplings indicate “poor” regeneration, and complete absence of seedlings and saplings indicates “no” regeneration (Singh et al. 2016a, b). In the present study, the occurrence of the moderate number of Q. oblongata individuals under sapling classes and young tree classes shows that the populations of this species will continue to dominate in the future also. However, frequent monitoring of the populations will help in understanding the dynamics of the populations of Q. oblongata in the context of anthropogenic factors and climate change (Thadani and Ashton 1995). The poor recruitment and establishment of seedlings of the species in the recorded populations may be due to high tree canopy coverage and lack of microhabitat, high litter accumulation and low light availability affecting the seed germination of Q. oblongata. Total seedlings density ranged from 80 to 370 Ind ha−1 and total saplings density ranged from 30 to 160 Ind ha−1 and is comparable with the previous studies (Pant and Samant 2012; Singh et al. 2021; Paul et al. 2019). Discontinuous distribution pattern of Q. oblongata seedlings was observed along with increasing altitude and changing aspects. The maximum studied population was recorded with low seedling density. The gap in the distribution of Q. oblongata seedlings might be due to high heterogeneity. Other factors responsible for the discontinuous recruitment of seedlings may be high canopy cover and aspect. Absence of seedlings density in the southwest and northeast aspects also supports the fact that aspect plays a crucial role in species distribution (Samant et al. 2002).

The edaphic factors, moisture content and nutrients are required for the survival of most of the species. Spatial variations in soil properties in forests are mainly due to the rooting pattern and litter accumulation. Soil moisture and temperature are limiting and controlling factors for the significant variation in topographical characters and vegetation pattern. Increase in moisture content with elevation might be due to decrease in temperature and increase in tree canopy cover and availability of water sources (Dasila et al. 2020). It was assumed that soil moisture availability affects the regeneration of species in ban oak forests (Dhyani et al. 2020). However, inadequate soil moisture might be responsible for seed germination and recruitment of seedlings and saplings in the studied populations. Seedlings are more sensitive to moisture deficiency than deep-rooted mature plants, which might be attributed to slow initial growth. Thus, lack of suitable microhabitat for seed germination and adverse climatic conditions and anthropogenic pressures might be responsible for lower recruitment and establishment of seedlings in the studied populations. In the plot clustering could be clearly observed showing close association of habitats within the cluster and the sharing of same. The close association of some habitats suggests that these species are affected in nearly the same way and concentration by the available biotic and abiotic factors, and as a result, their importance values are nearly identical. In addition, altitude influenced the dispersion of these species in the study area. Populations such as Jot, KKWS-IV, Dumas-Tissa and KKWS-V did not share the relation/habitats/niches with any individually and uniquely in the study area, they differed from other populations. High tree density and dense canopy would provide sufficient humus and soil moisture, which is required for seed germination. However, the establishment and survival of seedlings and saplings would depend on other biotic and abiotic factors such as weather conditions, anthropogenic activities, grazing, richness, and ecosystem properties (Kumar and Ram 2005). The availability of nitrogen would depend, to a large extent, on the quantity and properties of organic matter (De Hann 1977). Therefore, the high amount of organic matter in the forest might also be the reason for richness of nitrogen content (Singh et al. 2021).

The prediction of habitat shifts using ENM/SDM tools is essential for orienting the human-assisted efforts in recruitment and regeneration of the slow growth of ban oak with appropriate Silviculture efforts, especially in the degraded oak forest patches (Dhyani et al. 2020). For habitat suitability modelling, MaxEnt Model is very suitable, and the overall accuracy of the model is high. Maxent was efficiently used to predict the current distribution and changes in the future distribution patterns in response to the climate change of ecologically and economically important ban oak. Multi-collinearity test was performed to choose the variables most appropriately, which helps in modelling, eco-distribution mapping and prediction of the species across the geographical space (Supplementary Table 4). An AUC value of 0.908 is closer to 1.0 which indicates better model performance. The AUC test (0.908) values were comparable to the values earlier reported for Drepanostachyum falcatum (0.978) in Himachal Pradesh (Samant and Lal 2015); Carpinus viminea (0.954) and Corylus jacquemontii (0.938) in Himachal Pradesh (Paul et al. 2018, 2019), Quercus oblongata (0.876) in central Himalaya (Dhyani et al. 2020), Quercus semecarpifolia (0.902) and Betula utilis (0.912) in Great Himalayan National Park (Singh et al. 2021), Taxus contorta (0.905) in Himachal Pradesh (Chauhan et al. 2022). It is important to note that AUC values tend to be lower for species that have broad distribution scope. The highly suitable, suitable, and moderately suitable classes can be considered for the reintroduction (in situ conservation) of the species (Singh et al. 2021). The areas being identified as having high suitability (i.e., highly suitable and suitable areas) showed high densities of trees, saplings, and seedlings. The potential habitat distribution map can help in categorizing the potential distribution regions, especially in the highly suitable and suitable areas. This suggests a need for development of conventional propagation protocols for mass multiplication, establishment, and maintenance of nurseries for quality planting material and dissemination to the local inhabitants and the Forest Department for plantation in the highly suitable, suitable, and moderately suitable areas of district Chamba.

Forests act as net carbon sinks and are one of the main sources of fixing of atmospheric carbon. The adverse climatic conditions along the altitudinal gradients considerably affect the growth and development of trees, therefore further carbon accumulation in temperate forests, and lower biomass at high altitudes could be due to the cold temperature. Studied populations were mainly dominated by moist temperate oak forest and play a significant role in carbon sequestration as tree biomass is one of the main carbon sinks. Nutrient supply become more limiting factor to the plants at higher altitudes thereby limiting forest productivity and biomass accumulation. The mature forest ecosystem accounts for higher carbon stock than young and juvenile forests. Maximum carbon sink was observed in Kalatop Khajjiar Wildlife Sanctuary and adjoin studied populations also indicated to promotion of the carbon management through afforestation in northwestern Himalaya. In the present study, above ground biomass density for Q. oblongata was ranged from 10.52 to 110.80 Mg ha−1 and below ground biomass density ranged from 3.05 to 52.84 Mg ha−1. The values of AGB are comparatively very low and BGB more or less comparable with the study reported by Bisht et al. (2022) in central Himalaya for ban oak (AGB: 204.6–497.2 Mg ha−1; BGB: 58.4–130.20 Mg ha−1).

The protection of forest biodiversity remains a major challenge in the management of forest resources in the era of climate induced vulnerable ecosystems. Botanical and ecological data can provide basic information about the many alterations in any forest vegetation and biodiversity over the period of time. Main implications for the conservation are inadequate knowledge of forest botany among the foresters and other professionals is leading to the plantations of wrong species causing unorganized management of the forest areas (Rana et al. 2021), lack of financial assistance and accessibility, less awareness among the mountain natives, lack of infrastructure and monitoring, and finally lack of implementation of official orders in ground levels that ultimately increase less conservation. Vegetation ecology investigates the species composition and their sociological interaction in communities (Mueller and Ellenberg 1974). Biodiversity is mostly considered for human sustenance, secure economy and societal benefits. Biodiversity of oak forests has been continuously declining throughout the northwestern Himalaya during the past decades, despite its recognized importance to sustainable livelihood, human well-being, ecosystem functioning, services, resistance and resilience under climate change. The best approach to protect and conserve the existing mixed broadleaved oak forests, which is a life support to millions of rural inhabitants in the northwestern Himalaya, is by supporting natural course of regeneration, and conserving and reducing pressure on soil seed bank, in the light of climate vulnerabilities (Dhyani et al. 2020). This is where local communities will have a very important role to play in conservation and protection of Ban oak forests.

Extraction of Ban oak from its natural ecosystem is growing at an alarming rate due to its multiple purpose and many ecosystem services which ultimately creates huge pressure on natural regeneration in the wild (Rawal et al. 2012). This study provides comprehensive information on physical characteristics, richness, density, total basal area, species diversity, concentration of dominance, regeneration, pattern along with the details of physico-chemical properties of soil, size class distribution, and habitat suitability modelling and population ecology of Q. oblongata. The study would not only help in eco-restoration of the habitats, but also in recovering the species population and improving its in situ conservation (Paul et al. 2018). However, the changing environmental conditions, especially climate change coupled with continued anthropogenic pressure, may lead to reduced populations of Q. oblongata and replacement by other species with no or very low economic value (Samant et al. 2006).

If properly used, the phytosociological research may be very useful and important for managing the biodiversity in the area. The vegetation indices may be used to forecast forest ecosystem structure and comprehend vegetation dynamics. Measures that are specific to communities and habitats are crucial for the preservation of biological variety. Therefore, there is an urgent need to create awareness among the native communities, forest department, NGOs, and other stakeholders on status, soil requirement, economic values and conservation of the species. Frequent monitoring of the natural habitats and development of conventional propagation protocols for mass multiplication and establishment of seedlings in in situ and ex situ conditions are suggested for conservation and sustainable utilization of the species. This would serve the local communities for prosperity in future.