Composition and structure of plant communities in the Moist Temperate Forest Ecosystem of the Hindukush Mountains, Pakistan

Here, we investigated the relationship between Plant communities and the environment from the moist temperate vegetation of Lakoo mountainous forest District Swat. We sampled data from 162 sampling units (Quadrates) using 1x1m 2 for herbs 5x5m 2 and shrubs, while 10 x10m 2 for trees, systematically considering six elevation gradients between the altitudinal from 1970m to 3095m. We performed statistical analysis like Canonical correspondence analysis (CCA) and TWINSPAN (two-way indicator species analysis) for ecological assessment and clustering of plant communities. To check upon the correlation of species (CR) with topographic and edaphic variables we used statistical software PC-ORD version 7. We recognized 264 species plants belonging to thirty families. We recorded key sampling measurements of density, frequency, and cover for all these species which are vital for community description. The results showed Shannon-Winner , s, and Simpson diversity values as 19.18 and 3.17 respectively. The importance value indexes (IVI) were used to identify the leading and rare species of plant in each community or cluster group. In total we recognized eleven different communities as: Berberis- Abies-Bergeni, Picea - Indigofera - Poa,Abies- Parrotiopsis-Poa, Quercus- Viburnum-Poa, Picea- Salix-Primula, Abies- Viburnum-Poa, Viburnum- Taxus-Poa, Pinus- Viburnum-Lithospermum, Abies-Berberis-carex, Pinus- Viburnum-Poa and Parrotiopsis- Picea-Poa through hierarchical cluster analysis (TWINSPAN). CCA analysis revealed that of all studied edaphic and topographic variables altitude, silt, calcium carbonate, and organic matter were the strongest factors determining plant community diversity and composition in each microclimate of the eleven communities. Visually the vegetation of the forest was dominated by small-sized trees followed by shrubs, and regenerates indicating the stage of secondary regeneration. We found severe human interference in disturbing the existing biodiversity, which requires immediate conservation to ensure sustainable management and utilization of natural resources of the Lalkoo moist temperate forest.

associations that result from various environmental factors which are fundamental baselines in management decisions (Zhang and Zhang, 2011;Li et al., 2017;Rahman et al., 2020).
An elevation gradient is a composite environmental gradient (Zhang et al., 2006;Körner, 2007) that offers a supreme prospect for critically observing species diversity and shaping the composition of plant communities.Variations in elevation, slopes, and aspects also enhance diversity in various plant communities (Arora, 2002;Rahman et al., 2020).In tropical rain forests these gradients are measured as natural monitoring sites where drivers of community distribution patterns and ecosystem functions can be assessed over small geographical reserves (Li et al., 2017;Körner et al., 2017;Veintimilla et al., 2019).
A great concern emerged in ecological and environmental sciences about the strong relationship between biodiversity and ecological factors.Regionally Moist temperate forests with high altitudes in Hindukush-Himalaya are the most liable vegetation to environmental change (Salick et al., 2009;Shaheen et al., 2015).Among different ecological factors, elevation is the most essential ecological factor that monitors species and plant community diversity.Especially in plant community ecology, a lot of work has been carried out on the relationship between species diversity, community diversity, and elevation gradient (Donoghue, 2008;Li and Feng, 2015;Kluge et al., 2017;Lei et al., 2020).This indicates that distribution patterns and richness of plant species with elevation frequently vary along various transects and at various spatial scales in vegetation (Nogués-Bravo et al., 2008;Rowe and Lidgard, 2009).
In Pakistan, moist temperate forests like Lalkoo valley are rich biodiversity hotspots of the great mountain ranges of Hindukush-Himalaya (Ahmad et al., 2016).However, some phytosociological work regarding the establishment of plant communities along an elevation gradient in moist temperate forests has been reported in some countries around the world.The pioneering work of these includes Ahmad (2009) and Siddiqui et al. (2010) used multivariate statistical analysis to inspect the vegetation-environmental complex of blue pine forests of moist temperate zones in Pakistan.Similarly, Shaheen et al. (2015) documented species distribution patterns and community structures of conifers in moist temperate forests of the Kashmir Himalayas, and Dhyani and Dhyani (2016) reported ecosystem provisioning services from moist temperate forest ecosystems in India.We aimed to investigate changes in plant community dynamics along altitudinal and edaphic gradients in Lalkoo moist temperate forest, in Pakistan.

Introduction
Plant community assessments and structural studies of vegetation along ecological gradients are one of the most significant topics in community ecology (Peters et al., 2019;Iqbal et al., 2018).The study of plant associations and their classification is related to the field phytosociology while vegetation in the strict sense is the ecological quantification of plant resources (Timilehin et al., 2017;Ali et al., 2018;Veintimilla et al., 2019;Peters et al., 2019).Multilateral relationships exist between vegetation structure and prevailing ecological features like elevation (Sher et al., 2014 ;Haq et al., 2020), topographic heterogeneity (Zhang et al., 2006;Kuma and Shibru, 2015;Zeb et al., 2020;Mondal and Zhang 2018), water availability and soil Physiochemical characteristics (Forzieri et al., 2011;Li et al., 2017) and availability of light (Bonari et al., 2017;Peters et al., 2019).
A community is a cluster of plant species that present fundamental functional units both spatially and temporally, which have a direct relation to seasonal differences and sampling periods (Ali et al., 2018;Ilyas et al., 2015;Hussain et al., 2019).Species diversity plays a key role in biodiversity conservation supervision and most effectively as an indicator of community constancy in most of moist temperate and temperate forests (Shaheen et al., 2011a: Roy et al., 2015: Ali et al., 2018).Community structure, distribution pattern, and vegetative function are the greatest ecological features of forest ecosystems (Zhang et al., 2006;Timilsina et al., 2007;Ali et al., 2018).Understanding how diverse community processes interact is not only fascinating in itself but is also essential for predicting how communities will respond to various environmental conditions.Therefore distribution pattern of plant species in a community is a function of surrounding environmental variables (Weiher and Howe 2003;Götzenberger et al., 2012).Therefore, it is necessary for understanding plant community composition and distribution, along an elevation gradient in forest ecosystems (Zhang et al., 2006;Bonari et al., 2017;Li et al., 2017).Environmental gradients like topographic and edaphic factors (Zhang and Zhang, 2011;Khadanga and Jayakumar 2020) influence plant community structures in forests (Ullah et al., 2015;Ismail et al., 2019;Zeb et al., 2020).Research has shown that the patterns of vegetation are widely influenced by ecological factors, such as regional topography features (such as aspect and slope), edaphic factors, and anthropogenic pressure (Rahman et al., 2021).Relation of vegetation attributes, indicator values (predictable richness), and ecological gradients establish various community structures (Ullah et al., 2015;Iqbal et al., 2018;Ismail et al., 2019).In an ecosystem indicator species monitor outcoming changes in the component of plant carbonato de cálcio e matéria orgânica foram os fatores mais fortes na determinação da diversidade e composição da comunidade vegetal em cada microclima das 11 comunidades.Visualmente a vegetação da floresta foi dominada por árvores de pequeno porte, seguidas de arbustos, e se regenera, indicando o estágio de regeneração secundária.Encontramos grave interferência humana na biodiversidade existente, o que requer conservação imediata para garantir o manejo sustentável e a utilização dos recursos naturais da floresta temperada úmida de Lalkoo.

Material and Methods
The investigated area Lalkoo is a mountainous moist temperate type of forest, located in the Swat district Hindukush belt of north Pakistan.The area is 60 km from the capital city Mingora/Saidu Sharif.Geographically Lalkoo area lies within the coordinates of 35°08' 27.62"N and 72°23'09.12" E with an altitudinal range starting from about 1581m.a.s.l., and reaches the highest peak of alpine pastures of 3849 m.a.s.l.The area with a monsoon belt consists of moist temperate forests (Beg and Khan, 1974;Ullah et al., 2015).The annual temperature of the investigated area was recorded at 13.96-22.25°C.The mean annual rainfall (RF) is 1777mm.The RF was recorded in July and August with a mean total RF of 228.9mm and 220.9mm while minimum RF occurred in November and December which is recorded at 42.4mm and 78.4mm (Data source PMD; http://www.pmd.gov.pk).Lalkoo valley lies in the Sino-Japanese (SJ) phytogeographical region (Ilyas et al., 2015).The investigated area is covered with undulating slopes and ridges covered with thick vegetation in the sub-alpine zone and scrubs and meadows in alpine pastures.Current work on Lalkoo valley is mainly concerned with the establishment of plant communities and distribution patterns in correlation with environmental variables (Zhang et al., 2006;Sanders and Rahbek, 2012;Li et al., 2017).
The phytosociological work of the study area was carried out systematically in six altitudinal transects based on, species richness and diversity, habitats, and physiognomic contrasts.Density (D), Frequency (F), and Cover (C) for each plant species was recorded to attain Importance values (IV) for the establishment of various community patterns (Figure 1).Geographical coordinates of each study site were recorded using Global Position System (GPS).The formation of various plant communities was based on the highest importance values (IV) calculated for each specie following (Badshah et al., 2016;Forzieri et al., 2011;Ilyas et al., 2015;Ali et al., 2018).Total area quantification was assessed by species-area curve by using statistical software PC-ORD version 7.0 (Figure 2).
Plant communities, organization, distribution, and diversity were evaluated along an altitudinal gradient from1970 to 3120m in 54 sites in Lalkoo valley.For detailed data, collection vegetation was systematically divided into For detailed data, collection vegetation was systematically divided into 6 altitudinal transects, each one with 9 plots (resulting in 54 sampling units), and in each plot you established 3 quadrats -one for each strata for herbs, shrubs, and trees (Kent and Coker, 1992).Vegetation was analyzed by using quadrates 1 × 1 m 2 designed for herbs, 5 × 5 m 2 designed for shrubs and 10 × 10 m 2 for tree species, at each site.To each sampling site, relative and absolute values of density (D), frequency (F), and cover (C) along with their IVI values were recorded following the procedure of Curtis and McIntosh (1950) and Kamran et al. (2019).
During the study 264 plant species were documented in 162 quadrates.The slope and aspect were recorded with the help of a digital compass, elevation was measured by altimeter and the depth of soil (DS) was calculated with a 2m iron rod and ruler of 36Cm was used for analysis of depth of litter (DL).Index of Diversity (D) was measured by using methodology of Simpson (1949) and Shannon (1948) for each stand as (Equation 1): were, D stands for index of diversity, n represents the sum of individuals of a particular species, and N is the sum of individuals of all species.The Shannon Wiener index ( , H ) is the measure of the average degree of uncertainty in expecting what species of an individual selected at arbitrary from a sampling unit will fit to (Shannon 1948).

Plant community classification
Clustering describes the unsupervised learning task of partitioning observations into homogenous subsets, or clusters, to uncover subpopulation structure in a dataset.Hierarchical cluster analysis (TWINSPAN) classified 264 plant species into eleven community types by using PC-ORD 7. Figure 3.A comprehensive explanation of these community types concerning associated species of plants and ecological edaphic and topographic variables (Supplementary material 1).

Plant community diversity parameters
Shannon's diversity index for all species shows significant differences among different plant associations and presents a gradual increase from lower altitude to higher altitudes.Values obtained from the analysis of Shannon-Wiener and Simpson indices were used for measurements of diversity among different plant communities presented in Table 2. Of the total eleven plant communities, Picea-Indigofera-Poa, Quercus -Viburnum -Poa, and Berberis-Abies-Bergenia represents the lowest Shannon diversity values 0.76, 1.2 and 1.50 respectively.Hence strength of diversity become increases for certain communities like Abies-Parrotiopsis-Poa (2.42), and Abies-Viburnum-Poa(2.03),  and .However, four plant communities of the study site were significantly highest diversity values from 3.6 -3.99 for community types Abies-Berberis-Carex, Parrotiopsis-Picea-Poa, Pinus-Viburnum-Poa, and Picea-Salix -Primula that indicates overall potential diversity.
Simpson , s diversity index for all plant species shows the remarkable differences among distributed groups.These differences are indeed related to human interference like deforestation and grazing intensity.Concerning the vertical structure of the study area, this index shows a significant difference only in the Pinus-Viburnum-Lithospermum community with a maximum value D= 0.99 as presented in Table 1.However Berberis-Abie-Bergenia, Abies-Parrotiopsis-Poa(APP-III), Viburnum-Taxus-Poa (VTP-VII) and Pinus-Viburnum-Poa stands average Simpson , s diversity values from 0.99 -0.91 .More Plant community Parrotiopsis-Picea-Poa stands the lowest value of Simpson , s diversity (0.87).

Canonical correspondence analysis
For investigation, the effects of eleven environmental variables(EV) on the distribution patterns of plant communities Canonical correspondence analysis (CCA) was used (Figures 4-6).Analysis of CCA model (11-Environmental variables) revealed variation of 3.8 Axis -1, 3.3 Axis -2, and 2.9 Axis -3 alone of total unexplained variance (Figure 4    (Figure 4 and supplementary material 2) while among all environmental variables phosphorous was found negligible influence on the distribution of plant communities.
Our study also revealed that, Quercus dilatata, Abies pindrow, Pinus wallichiana and Picea smithiana had high dominance trees in terms of IVI while in shrubs were Berberis lyceum, Indigofera heterantha, and Viburnum cotinifolium and herbs were Poa alpina, and Poa annua (Table 1).

Discussion
Multivariate analysis was carried out to examine the distribution and classification of plant communities that vary greatly in Lalkoo Hindukush mountainous range Pakistan.The most important factors controlling this variation is elevation, soil texture, soil organic matter, soil pH, soil electrical conductivity, soil chemical nature topography, and so on among which elevation is a key factor that affect community distribution (Keddy, 1992;Rahman et al., 2021).
The systematic classification of plant communities is a basic tool used in vegetation ecology that is subjected for various purposes like conservation, land planning and management data reduction and synthesis that were used by scientists up to date (Diamond, 1975;Götzenberger et al., 2012;Oyola-Guzman and Oyola-Morales, 2018) however traditional classification of plant communities is typically constructed on general features of species assessed by personal experience (Yong-Chang, 2004;Qin et al., 2016).The methods applied in the present analysis provide a baseline for users to provide multiple classification procedures of plant communities for accurate information in future perspectives.The study identified 264 species of vascular plants that were classified in to eleven plant community types (clusters) with different diagnostic species in Lalkoo mountainous forest.Differences among dominant species and forest plant community compositionindicate an influence of environmental, anthropogenic, and topographic factors (Wiens and Donoghue, 2004;Khan, 2012;Idowu et al., 2020;Ohler et al., 2020;Rahman et al., 2021).The dominant plant species of the study area in all communities (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11) are Berberis, Picea smithiana, Abies pindrow, Abies pindrow, Viburnum grandiflorum Pinus wallichian, and Parrotiopsis jacquemontiana which were also reported for adjacent temperate and moist temperate forests of Pakistan as dominant species in most of the plant communities (Khan, 2012;Habib et al., 2016;Shaheen et al., 2017).Indigofera, Salix, Taxus, Bergenia, Poa, Carex, Lithospermum, and Primula is representing typical leading species of the study area that have been reported in different plant communities connected to moist temperate forests (Khan et al. 2020).
One of the basic parameters for measurement of plant community structure and composition is the importance value index (IVI), Species with the same IVI have the same or similar population structure (Lamprecht, 1989;Hundera and Gadissa, 2008) resulting in maximum IVI values ascribed to some species.In comparison to this approach, all studied species showed different ecological importance and difference in relation to their recorded values of IVI.This indicates that these nine plant species are highly ecological significance that is adapted to high ecological amplitude, disturbance, and influence of local communities in Lalkoo moist temperate forest.
Assessment of plant communities -environmental correlation multivariate technique CCA was applied.Of the environmental variables, Altitude, Clay, Silt, Sand, pH, Organic Matter, Calcium Carbonate, Nitrogen, Phosphorus Potassium, and Electrical Conductivity affect the plant community's classification and distribution.This is also documented that a species with the high cumulative percentage of variance (3.8) and environmental correlation on the ordination scale was exhibited along axis 1 .However of the total measured environmental variable used in CCA analysis the most influential factors were Calcium  Carbonate looked to be the cause for the distribution pattern of the community of Abies-Parrotiopsis-Poa, silt for Parrotiopsis-Picea-Poa community and altitude (Shaheen et al., 2011b) for Picea-Indigofera-Poa community.Likewise, outcomes were also described by Khan et al. (2019), andAhmad et al. (2016).
Altitudinal variation is a significant factor controlling plant community structure.In mountainous forests altitudinal gradient limits species distribution and plant community types (Chawla et al., 2008;Rahman et al., 2022).In this context, a lot of studies were carried out on the variation of the diversity of species along elevation gradient in mountainous forests (Lomolino, 2001).In Lalkoo hills we found a lot of indicator species with different altitudinal preferences for example Abies pindrow was found from 1900-2920m, Picea smithiana was of common occurrence in the elevation range of 2215-3045m, Pinus wallichiana most abundantly recorded from 2000-2345m, Parrotiopsis jacquemontiana ranged from lower elevation 2085m was distributed at high elevation 3095m.Quercus dilatata is distributed from mid-elevation (2368m) but shows its peak of dominancy towards high elevations (2920m).Dominant indicator shrubs Berberis lyceum and Indigofera heterantha herbs Poa annua and Primula rosea were found in range from 2000-3045m.All plant species showed significant floristic and structural heterogeneity with local ecological variables.Other studies in the Hindu Kush-Himalayas have documented similar elevation ranges for temperate ecosystems (Shaheen et al., 2012;Rahman et al., 2021).
Present findings revealed that Lalkoo forest is a potentially complex region, chiefly in terms of wide spread range of biodiversity and important portable natural resources.The area is under heavy anthropogenic and biotic pressure which needs conservation strategies for the complete restoration of Lalkoo forest.The present investigation also suggests that if human pressure remains exists in the study area, the established natural plant resources may be disappeared and will follow rapid extinction as observed in the nearby Hindukush mountainous forests of District Swat, commonly Girbanr hills Miandam hills (Hussain et al., 1997), Qalagai hills (Ilyas et al., 2012;Ali et al., 2018).Our description of plant communities, species and environmental variables is an integrated way to study diversity in the ecosystem.

Conclusion
Current work investigated 11 plant communities with potential indicator plant species in sub-alpine natural mature forests of Lalkoo Valley, Hindukush mountains Pakistan.These indicators will be helpful in the future for monitoring variations in response to ecological fluctuations that affect plant community structure and composition (Conti et al., 2017).Altitude and edaphic factors (silt, organic matter, and calcium carbonate) played a vital roles in determining the plant communities.These forests provide carbon storage, a better environment, biodiversity, and other natural resources to the indigenous communities.Our analysis also revealed that several anthropogenic and edaphic factors like deforestation, overgrazing, wood extraction, expansion of agriculture fields by the destruction of soil cover, and soil erosion are serious threats to the vegetation in the investigated area Lalkoo valley.It is strictly recommended that conservation strategies are essential to deal with restoration plans, rangeland management, fuel alternatives sources, and local community awareness for restoration of existing natural biodiversity.

Figure 2 .
Figure 2. Species area curve obtained from 264 plant species in Lalkoo valley.

Figure 3 .
Figure 3. Cluster Dendrogram indicating 11 plant association types in the Lalkoo Valley.
and supplementary material 2).The total variance ("inertia") of CCA data was 8.578.The main patterns in plant community distribution in the overall data set of CCA reflect on silt, altitude, and calcium carbonate.Silt represents a strong positive correlation on Parrotiopsis -Picea -Poa (PPP) community and a negative with Berberis-Abies-Bergenia (BAB) along Axis-2.Interestingly PPP community shows potential growth toward high altitudes (2920m) while BAB community towards lower altitudes (2085m) as presented in Figure5.While Abies-Parrotiopsis-Poa(APP) shows a strong negative correlation with Axis -1 influenced by calcium carbonate.Similarly, Picea-Indigofera-Poa community affected by altitude shows a strong positive correlation with Axis -1.Most of the plant communities were clustered around the center of the ordination axis

Figure 5 .
Figure 5. CCA plot Analysis of illustrating the influence of elevation on spreading pattern of plant communities in Lalkoo valley Swat.

Figure 6 .
Figure 6.Analysis of CCA plot illustrating correlation between calciumcarbonate and plant communities along axis-1 and 2.

Table 2 .
Calculated values of Diversity indices from Lalkoo valley Swat.
NR: Number of releves; NS: Number of species; H: Shannon index; D: Simpson index; IVI: Importance values index.