Woody species diversity from proposed ecologically sensitive area of northern Western Ghats: implications for biodiversity management

The Western Ghats of India support an array of tropical forests ranging from wet evergreen to scrub formations. Several endemic and threatened plant species are located in areas other than protected areas (PAs). There is an urgent need to understand species diversity in areas other than PAs, for effective management of tropical forests. In this context, reserve forests and informal PAs of Amboli from northern Western Ghats have been investigated. Woody species composition, diversity, and stand structure were assessed by laying quadrats and transects (n=46, area=2.575ha) in closed and open canopy forest patches covering habitat heterogeneity and environmental gradient of the area. A total of 2,224 individuals (of 87 species, 68 genera, and 35 families) was enumerated. Memecylon umbellatum, Syzygium cumini, and Diospyros nigrescens were found to be the most dominant species as per importance value index. Melastomataceae was the most dominant family as per family importance value, whereas Euphorbiaceae and Rutaceae were the most speciose. Fourteen IUCN Red List assessed species and 18 species endemic to the Western Ghats were encountered. Endemic species accounted for nearly 20% of the total number of individuals sampled. Demographic profile exhibited reverse ‘J’ pattern. Average basal area was 27.02m2 per hectare. Woody species diversity of Amboli forests was found comparable with other PAs from northern Western Ghats. Amboli and the adjoining area have been proposed as ecologically sensitive and in the wake of anthropogenic and developmental pressures they experience, it calls for urgent conservation attention.


INTRODUCTION
Woody species form an important component of the forest landscape both because of their diversity and biomass. They play a vital role in shaping overall structural dynamics of the forest stands and offer various kinds of 'ecosystem services'. Of the 36 global hotspots of biodiversity, Western Ghats, extending along the western coast of India, along with Sri Lanka comprise the Western Ghats-Sri Lanka hotspot (Conservation International 2019; Myers et al. 2000). Western Ghats of India occupy the fifth position in the world in terms of economic potential of their biological resources (Ganeshaiah & Shaanker 2007). It is globally, an area of high endemism with 1,500 endemic species of which 352 are woody plant species and also houses over 4,000 medicinal plants species. WGs support an array of tropical forest types ranging from wet evergreen to scrub formations covering an area of about 1,64,284km 2 (Kasturirangan et al. 2013). Although nearly 10 percent of the Western Ghats hotspot is under formal protection, it has been pointed out that PAs in this region have historically been established on an ad-hoc basis with little attention to diversity distribution (Bhagwat et al. 2005). There is indeed a growing recognition that PAs cannot be conceived and managed as "islands" isolated from other PAs and from the rest of the landscape context (Laurance et al. 2012). Hence, there is a need to recognize high potential of informal protected areas such as sacred groves for effective conservation management (Bhagwat & Rutte 2006) that can supplement the PA diversity. The conservation management in the region needs to address the following questions: (1) do existing PAs adequately represent the biodiversity? (2) do excluded forest patches sustain more species than PAs? and (3) how many PAs are required to cover the entire gamut of biodiversity? Considering the high endemism, it is necessary and urgent to evaluate conservation potential andecosystem services of the buffer areas surrounding the PAs or other areas not included in formal PA network. CEPF (2007) report showed that NWG have presence of more fragmented forests patches than the southern Western Ghats (SWG) and are under the pressures of selective logging, excessive grazing, fire, and road construction. Though sporadic records of quantitative inventorization of forest stands from PAs of NWG area available (Kanade et al. 2008;Joglekar et al. 2015), lack of focused studies on diversity that exists outside PAs in fragmented forests is a major challenge in understanding changes in forest community under anthropogenic impacts. Understanding the spatial distribution of these forests, their conservation significance and knowledge of vegetation types thus, becomes essential for outlining effective management strategies.
The forests of Amboli area act as a transition zone between NWG and SWG. CEPF (2007) report identified Amboli region as an irreplaceable site for certain globally threatened species that lack formal protection. Four new faunal species were described from Amboli region in a span of less than five years (Satose et al. 2018). The forests of Amboli experience high developmental pressures owing to growing tourism enterprises, necessitating conservation planning, for which exploration of the region's diversity is necessary. In this paper, we have characterized the woody species diversity, composition and stand structure of Amboli forests from relatively less explored area of NWG.
Amboli (MSL=700m) is located in Sawantwadi Taluka of Sindhudurg District of Maharashtra ( Figure 1) in NWG. Although the area lies outside the formal PA network, it includes private forests, reserved forests and community owned forests spread across 659.88ha (Bharmal et al. 2011). Fragmented forests of Amboli form a mosaic of different vegetation and habitat types. Primary vegetation type is evergreen (closed canopy: >60% and height 15-20m), with stunted vegetation around lateritic outcrops (open canopy: 20-40%, height 5-8m) (Image 1). These together harbor endemic and threatened plant species and unique ephemeral flush vegetation that characterize lateritic plateaus. The area is proposed as ecologically sensitive (Maharashtra Government Resolution) and also forms a part of geographically and ecologically important Sahyadri-Konkan Ecological corridor (CEPF 2007).
It is the type locality of species like a Caecilian Gegeneophis danieli (Giri et al. 2003), Amboli Tiger Toad Xanthophryne tigerina (Biju et al. 2009), leaping frog Indirana chiravasi (Padhye et al. 2014) and water snake Rhabdops aquaticus (Giri et al. 2017). Biologists who studied the diversity of avifauna and Lepidoptera (Bharmal et al. 2011;Satose et al. 2018) concluded that the area is rich in biological diversity. Though the area has been explored in details for faunal diversity, Woody species of proposed ecologically sensitive area Tadwalkar et al.

J TT
comprehensive taxonomic floristic studies are rare (Kulkarni 1988;Almeida 1990). There is dearth of quantitative ecological studies. Amboli is a famous destination for tourists and naturalists alike due to picturesque landscapes, waterfalls and faunal sightings. But owing to the unplanned and unregulated tourism, the area witnesses encroachment into the forested landscapes, logging, and poorly planned construction.

Sampling design
Standard methods of woody vegetation analysis were followed (Ganesh et al. 1996;Sutherland 2006). Species composition and diversity were assessed by laying quadrats (n=40, size 20 x 20 m) and transects (n=6) in closed and open canopy forest patches covering habitat heterogeneity. It was ensured that the sampling plots cover significant environmental gradient of the area. Transect length varied from 500 x 5 m or 250 x 5 m or 200 X 5m depending on the patch size. Each quadrat and transect was marked by GPS. The total area sampled was 2.575ha and intensity of sampling amounts to 0.39% of sampling, which is more than a standard requirement of 0.01% for such enumerations (Shivraj et al. 2000).

J TT
Vegetation composition, stand structure and diversity assessment All woody species were enumerated for individual height and girth (≥15cm at 1.3m height above ground) measurements. Species level identification was done using regional flora (Almeida 1990;Singh et al. 2001). Endemicity and IUCN Red List status of the species were assignedby referring to standard literature (Pascal 1988;BIOTIK 2008;Singh et al. 2015; https://www.iucnredlist. org/). Data collected from quadrat and transect sampling were used to understand woody species composition and diversity. For stand structure and basal area estimates data from quadrats was used. Importance value index (IVI) and family importance value (FIV) were calculated as per Ganesh et al. (1996). For the diversity estimates, data from quadrats and transects was pooled. Diversity was estimated using Shannon's index (H') as per Magurran (2004). Compositional similarity between sampled plots was assessed by Bray Curtis similarity index calculated using PAST (version 3). Correlation analysis was performed using the R software (version 3.5.1).

(A) Woody species composition and diversity
A total of 2,224 individuals were sampled during the study representing 87 species spanning across 68 genera and 35 families. Genus Diospyros was found to be the most diverse genus with four species followed by Ixora and Ficus (represented by three species each). Fifty-six genera (82%) were represented by only one species in the sampled area. Figure 2 represents 10 most abundant genera in the sampled area with corresponding abundance.
Out of 87 species that were encountered during the study, Memecylon umbellatum was found to be the most abundant species in the area (N=501, 22.53%) followed by Mallotus phillipensis, Syzygium cumini, Diospyros candolleana, Symplocos racemosa, and Diospyros nigrescens. These six species together contributed to 56.11% of the total abundance. A long tail of singleton species was seen where, singleton species and doubleton species contributed to 24.1% (n=21) and 12.6% (n=11), respectively, to the stand structure. Persea macrantha, Homalium ceylanicum, and Mitragyna praviflora were among a few species represented by only one individual and Euonymus indicus, Lagerstroemia microcarpa, and Litsea deccanensis were represented by two individuals. Table 1 depicts various phyto-sociological attributes from the sampled plotsin the study area. The abundance in the sampled plots varied greatly from one individual (OLR2, OMD2) to 384 individuals (CCR3); whereas number of species ranged from 1 (OLR2, OMD2) to 32 (CCR3). Maximum number of woody endemic species (9) was reported from Malai Pathar (CMP4), whereas Mahadevgad road (CMD1) showed highest number of endemic individuals (59). Presence of WG endemic species, Diospyros candolleana was a notable feature in this area. Shannon index varied from 0 to 2.86 within sampled plots, '0' being recorded for two open forest plots which were represented by single individual. In order to get insights into the contribution of singleton and doubleton species in overall woody species diversity of the area, Shannon index value was plotted against the proportion of singleton and doubleton species in the sampled plots, depicted in Figure 3 (r=0.798, p<0.001).

J TT
The results showed highly significant relation indicating contribution of rare speciesin the overall diversity of the study area. Table 2 gives various species attributes of the study area. Fourteen IUCN assessed species together accounted for 15% of the total number of individuals encountered. Diospyros candolleana, listed in the Vulnerable (VU) category, was found to be one of the dominant species in the study area. Evergreen (E) is the dominant habit represented by 78% of species which are mainly distributed in closed forest patches. Eightysix percent of species showed zoochory as a dispersal mode. An attempt has also been made to assign species status (canopy / middle storey / understorey) as per the vegetation strata observed in the study area.
Cluster analysis (Figure 4) revealed that maximum species similarity of the plots was observed to be ca. 74%. Quadrats laid in Lingachi rai sacred grove area (a community owned forest), formed a cluster. This cluster exhibits low similarity with the other quadrats taken in reserve forests, private forests and Sadachi rai (a sacred grove situated in the reserved forests). It is interesting to note here that these quadrats despite being laid in the closed forests exhibit different patterns. Open forest patches showed lowest (2% to 20%) species compositional similarity with closed forest patches.

(B) Importance Value Index (IVI) and Family Importance Value (FIV)
Data collected through quadrat sampling (S=78, N=1213) was used for the estimation of IVI and FIV. Memecylon umbellatum was found to be the most dominant species as per IVI (Table 3). Though represented by only 6% of individuals, Syzygium cumini was found to be second most important species due to its high basal area followed by Diospyros nigrescens, Aglaia lawii, and Dimocarpus longan. Family Melastomataceae represented by the genus Memecylon in the study area, showed the highest FIV (56.38) due to its abundance as well as the basal area. Families Myrtaceae, Anacardiaceae, Ebenaceae, and Euphorbiaceae were found to be the other most important families as per FIV ( Figure 5). Euphorbiaceae and Rutaceae were the most speciose families with six species each followed by Lauraceae and Rubiaceae (5 species each).

(C) Stand structure
The girth class distribution showed typical reverse 'J' shaped curve ( Figure 6). First three GBH classes, i.e., 15-30cm, 30-45cm, and 45-60 cm contributed to 73% of the individuals (no. of species=70) ( Figure   7). Less than 1% individuals were represented in GBH class > 210cm. They were comprised by species such as Holigarna grahamii, Persea macrantha, Syzygium cumini, Mangifera indica, and Memecylon umbellatum. Total basal area recorded was 43.23m 2 . GBH classes (45-120 cm) contributed to highest basal area (40.99%), however, it should be noted that maximum number of individuals was found among lower GBH classes with subsequent GBH classes showing steady decrease in number of individuals (Figure 7). Basal area decreased with increasing GBH which was depicted by very low abundance. Stand basal area of Memecylon umbellatum and Syzygium cumini was around 41% of the total basal area.

(D) Endemic species diversity and abundance
Of the total number of species recorded, 18 species were Western Ghats endemics and accounted for nearly 20% of the total number of individuals sampled. Genus Diospyros (represented by two endemic species -D. candolleana and D. nigrescens) comprised of 51.8% of the endemic individuals. D. candolleana (VU) was also found to be one of the dominant species in the study area as revealed from IVI. Drypetes venusta, Knema attenuata, and Meiogyne pannosa were encountered only in the sacred groves. Sacred groves also showed presence of H. grahamii (>195cm) and Beilschmiedia dalzellii (>180cm). Such hefty individuals of these species were seldom seen elsewhere highlighting the significance of protection of sacred groves in biodiversity conservation. Endemic species richness also exhibited highly significant relation with Shannon diversity (r=0.766, p<0.001) (Figure 8). Table 4 represents various ecological attributes from  study area and compares it with similar such studies conducted elsewhere inside PAs and reserve forests of NWG.

DISCUSSION
Present study provides systematic account of woody species composition of Amboli forests. In comparison with studies from protected areas from NWG, the sampled area showed high species richness and abundance ( Table 4). Out of 35 families, Euphorbiaceae and Rutaceae were found to be diverse families of Amboli forest followed by Lauraceae and Rubiaceae. Though highly diverse, Lauraceae and Rubiaceae showed lower FIV values due to its lower density and lower basal area as similar to studies conducted in Kalakad-Mundanthurai forests of SWGs by Ganesh et al. (1996). As per FIV, Melastomataceae was found to be the most dominant family which is very similar to family dominance in Chandoli NP (Kanade et al. 2008) and Koyna WS (Joglekar et al. 2015). Puri et al. (1983) and Pascal (1988) assigned Memecylon-Syzygium-Actinodaphne (M-S-A) floristic series to evergreen forests of NWG based on the criteria of dominance-abundance-fidelity. Current study revealed Memecylon-Syzygium-Diospyros type which is found to be different from Memcylon-Syzygium-Olea type found in protected areas of NWG (Table 4). M. umbellatum, the most dominant species Woody species of proposed ecologically sensitive area Tadwalkar et al.

J TT
in the study area was represented by >20% of the total number of individuals. Similar trend was found in studies conducted in Chandoli NP and Koyna WS where M. umbellatum was represented by 27% and 34% individuals, respectively. The study area harbored 18 species endemic to the WG that accounted for around 20% of the individuals sampled. It is interesting to note that some endemic species represented in the study   area are among the most important species according to IVI. These include D. nigrescens (IVI 13.43), Holigarna grahamii (IVI 11.02) and D. candolleana (IVI 10.61). This underlines the importance of the study area in sustaining the population of endemic woody species. High proportion of endemic species was also reported by Kanade et al. (2008) from undisturbed evergreen forest patches of Chandoli NP. Similar findings were reported from Koyna WS which showed presence of 23 endemic species represented by 656 individuals (15.27%). The dominance of typical evergreen forest species such as Holigarna grahamii and Aglaia lawii, both endemic species, suggest an origin from a community differing in composition from the typical M-S-A types (Watve et al. 2003). Amboli forests showed presence of 14 IUCN assessed species (six species being VU or NT) with 15% of total individuals sampled which is comparable to Koyna WS that recorded 13 IUCN assessed species and 9% of total number of individuals (Joglekar et al. 2015).
Since the area under consideration is relatively small, we may expect high similarity among the species in the sampled plots, however, clustering with Bray-Curtis similarity plot reveals that there are unique species conferring unique composition to the plots. Closed forest patches of Lingachi rai form a separate cluster as against other closed reserved forest patches and Sadachi rai. Species like Artocarpus hirsutus, Blachia denudata, Beilschmiedia dalzellii, and Caryota urens were present in Lingachi rai with low/no occurrence in other closed forest patches. Average stand basal area of Amboli forests was 27.02m 2 /ha which was found to be comparable with other studies conducted in protected areas of NWG (Table4). Present study also showed reverse 'J' pattern of the stand structure with highest number of species and individuals in lowest GBH class (15-30 cm) (Kanade et al. 2008;Joglekar et al. 2015) while higher basal area was found to be between 45-120 cm. Typical evergreen endemic forest species like Aglaia lawii, Beischmedia dalzellii, Holigarna grahamii and ecologically important species like Ficus sp., Dimocarpus longan were present in higher GBH classes (above 180cm) indicating healthy nature of vegetation.

CONCLUSION
Studies on the vegetation analysis and biodiversity pattern are of utmost importance especially in the forest areas outside the PA network. Such areas in tropics are actively managed and modified by humans. Unplanned and uncontrolled tourism especially during monsoon, poorly planned construction and logging are some J TT of the disturbance drivers affecting floral and faunal diversity of Amboli (Image1). Floristic surveys form the primary step for carrying out ecological restoration of a particular area (Mota et al. 2017) and provide the inputs which feed large scale databases.
In this context, present study forms an important step in establishing the baseline data about woody plant diversity of the region. Closed forest patches with dominance of endemic and rare species emphasized the importance of conservation of Amboli forests in patchily distributed forests of NWG. It also revealed that the woody plant diversity in Amboli forest is comparable to other PAs from NWG. The information thus generated can be used effectively by BMC formed under the provisions of Biological Diversity Act (2002). Conserving this unique landscape rich in flora and fauna involving BMC and other stakeholders such as local community and forest department will reveal new facets of participatory conservation model that can be replicated elsewhere in the adjoining areas. www.threatenedtaxa.org The Journal of Threatened Taxa (JoTT) is dedicated to building evidence for conservation globally by publishing peer-reviewed articles online every month at a reasonably rapid rate at www.threatenedtaxa.org. All articles published in JoTT are registered under Creative Commons Attribution 4.0 International License unless otherwise mentioned. JoTT allows allows unrestricted use, reproduction, and distribution of articles in any medium by providing adequate credit to the author(s) and the source of publication.