Keywords

7.1 Introduction

The study of floristic diversity plays very crucial role to understand plant community structure and function (Patel and Dabgar 2012). Further to that, such study also provides vital scientific information for various kinds of environmental planning because plants provide basic amenities that help to sustain life on earth. The study of plant community dynamics and species diversity is an important aspect of forest ecology, which entails the contribution of various species in determining structure of specific habitat type (Patel et al. 2012). To generate the information on diversity and distribution on tree and shrub species is of primary importance for developing the conservation and management strategies (Kumar et al. 2006).

Due to geographical situation and prevailing physio-climatic condition, this region supports the distinctive gene pool distributed in various habitats such as tropical thorn forest, grassland, savannah, mangrove, saline desert and freshwater wetland in the region of our interest, i.e. western Kachchh (Patel 2012). These forests are scattered throughout the district but predominantly recorded from the western part of Kachchh, i.e. Lakhpat and part of Nakhatrana talukas. Noticeably, a large portion of such forests, i.e. about 1185 km2 area constituting 84% of thorn forest in western Kachchh, remains unprotected and thus not brought under management for maintaining vegetation cover or habitat improvement. Only 228 km2 area of these forests is confined within the Narayan Sarovar Sanctuary and has got the legal protection.

In general, due to poor canopy formation, these forests form open canopy cover that is mainly constituted by Acacia senegal and A. nilotica. The occurrence and distribution of codominant species varied greatly within the study area due to high variability within soil taxonomy, moisture regimes and site-specific geological formations. Such species include Euphorbia caducifolia, Grewia tenax, Cordia perrottetii, Salvadora oleoides, S. persica and Capparis decidua. The invasion of Prosopis juliflora is very common in these forests, which has changed the floral composition and vegetation structure. Several species of conservation significance are also reported from this area (Patel and Dabgar 2018).

Mining is a major developmental activity in western Kachchh, which promotes several other economic activities and contributes substantially in local and regional economies. Such vigorous developmental activities are causing land and ecosystem damage, which leads to the loss of vegetation cover and loss of species diversity (Huang et al. 2015). Additionally, the overburdens of coal mines when dumped in unmined areas create mine spoils, which also affect the surrounding vegetation cover and plant species diversity (Alekseenko et al. 2018).

In the past very few studies have been conducted on floristic diversity of the region, and limited information is available about the plant community structure. The studies are mainly focused to fulfil legal requirement in the form of Environmental Impact Assessment studies. Gujarat Mineral Development Corporation (GMDC) (1995a, b, c) conducted rapid EIA/EMP studies for proposed Mata-no-Madh, Akrimota and Umarsar lignite mining complex and conducted surveys around these mining complexes to assess the status of biological (flora and fauna) components. Identification afforded recognition of 9 species of trees, 15 species of shrubs and 22 species of herbs from core zone, while 34 trees, 28 shrubs, 51 herbs, 12 climber species and 9 crop species were recorded from buffer zone. From the Umarsar mining site, 9 species of trees, 12 species of shrubs, 31 species of herbs and 8 species of climbers were recorded from core zone, while 26 species of trees, 20 species of shrubs, 41 species of herbs, 9 species of climbers and 11 crop varieties were recorded from buffer zone. In Akrimota mining site, 8 tree species, 12 shrubs, 36 herb species and 6 climber species were recorded from core zone, while 26 tree species, 18 shrubs, 43 herbs, 11 climber species and 11 crop varieties were recorded from buffer zone. Also, 10 species of plants that possess medicinal value have also been recorded from this site. Further to that, in 2006 GMDC conducted a Rapid Environmental Impact Assessment for Umarsar Lignite Mine and reported 81 floral species. In the past very few studies have been conducted on the correlation of vegetation and coal mining (Cornwell 1971; Fyles et al. 1985; Prasad and Pandey 1985; Singh and Jha 1987). Hence, present work deals with the community structure and diversity pattern in different zones around the three major mining complexes to assess the impact of mining on floristic diversity.

7.2 Methodology

Floristic diversity assessment methods for various plant life forms such as trees, shrubs, climber, herbs and grasses varied greatly in terms of data collection, characterization, classification and identification (Bhatt 1993; Patel et al. 2012). Specific methods of sampling and data analysis used for evaluating different life forms of interest in the present chapter are as follows:

7.2.1 Sampling Design and Field Sampling

To understand the impact of mining on floral diversity, the diversity and distribution pattern of plant species was recorded in nine mine zones (referred to as Z1 to Z9 in present study) that represent the area under zones at the regular interval of 2 km from centre of the mining centres. After stratification, the entire area was divided into 5 × 5 km grids using Survey of India’s geo-referenced co-ordinate system. Those grid cells were further subdivided into 1 × 1 km smaller grids, and a total of 180 transects (Fig. 7.1) were laid down in select grids chosen randomly using random number table (Table 7.1). Stratified random sampling was employed to lay down samples of 1.1 km (n = 180 transects) different habitats. All transects were laid diagonally to randomly picked 1 × 1 km grid. Care was taken to have adequate sampling in each of the topographical features across lateral and vertical gradients like altitudinal range and terrain and spatially within each habitat type.

Fig. 7.1
figure 1

Sampling point location in and around coal mines of Kachchh, Gujrat

Table 7.1 Distribution of samples in study area

7.2.2 Sampling and Data Collection

To carry out enumeration of the vegetation of the study area, a simple, systematic and predetermined technique was employed. Various habitats were surveyed using transect and quadrate methods as suggested by Misra (1968) and Mueller-Dombois and Ellenberg (1974) to record the various characteristics of tree, shrub and climber species. We used 1.1-km-long transect (180 no.) for vegetation survey (Table 7.1).

Flora of Gujarat State (Shah 1978), flora of the Indian Desert (Bhandari 1990) and the Bombay grasses (Blatter and McCann 1934) were used to identify various plant species collected from study area. Collected plant specimens were deposited in the herbarium section of Gujarat Institute of Desert Ecology, Bhuj. Secondary sources including published literature such as WCMC (1994) and Red Data Book of Indian Plants (Nayar and Sastry 1990) were used to identify the rare, endangered and threatened plant species.

7.3 Data Analysis

7.3.1 Species Composition, Distribution and Diversity

Quantitative community characteristics such as density, abundance, % frequency, basal area and important value index (IVI) of each species were determined by following the methods as outlined by Misra (1968) and Mueller-Dombois and Ellenberg (1974). Distribution pattern was analysed using the value of abundance and % frequency. High frequency and low abundance of a species indicate the regular distribution; however, low frequency and high abundance indicate the contagious distribution of species (Misra 1968).

The % frequency of each species was calculated and placed into their respective frequency classes as suggested by Raunkiaer (1934). Further, species diversity and similarity indices were calculated using the following formulae:

  1. (A)

    Shannon-Wiener diversity index (H′) H =  ∑ pi ×  ln (pi)

    • Where:

      • H = index of species diversity

      • pi = proportion of total sample belonging to the ith species

      • ln = natural log1

  2. (B)

    Simpson diversity index (D):

    • D = 1−Dominance. Measure ‘evenness’ of the community from 0 to 1

  3. (C)

    Menhinick richness index – the ratio of the number of taxa to the square root of sample size

  4. (D)

    Margalef’s richness index (Rl)

$$ Rl=S\hbox{--} 1/\ln (n) $$
  1. (E)

    Dominance:

    • 1 – Simpson index. Ranges from 0 (all taxa are equally present) to 1 species dominant to the community.

    • D = sum ((ni/n)2) where ni is number of individuals of species i.

  2. (F)

    Equitability (El or J′): Shannon diversity divided by the logarithm of number of taxa.

    • This measures the evenness with which individuals are divided among the taxa present. J′ = H′/In(S)

    • where n = total number of individuals and S = total number of species

  3. (G)

    Buzas and Gibson’s evenness: eH/S

The rarefaction analysis was performed using Biodiversity Pro software (Biodiversity 1997 NHM and SAMS, http://www.nhm.ac.uk/zoology/bdpro) to cope up with the problem in comparing diversity among various zones evaluated during the present study.

7.3.2 Community Characteristics

Similarity indices: Jaccard’s similarity coefficient (Sj) is used to assess similarity of sites and uses presence/absence data. As it, Sorenson’s similarity index (SI), also looks at the similarity of pairs of habitats in terms of presence of species but gives greater preference to the species, those are common to the sites compared to those found on unique sites. Formula for calculating both of these indices are as below:

$$ Sj=a/\left(a+b+c\right)\ \mathrm{and}\ Ss=2a/\left(2a+b+c\right). $$

Where Sj = Jaccard’s similarity coefficient and Ss = Sorenson’s similarity coefficient, a = number of species common to (shared by) sites, b = number of species unique to the first site and c = number of species unique to the second site.

Further, cluster analysis, which is a hierarchical, agglomerative and polythetic technique of clustering, was used to quantify the resemblance among plant communities inhabiting various mine zones. Cluster analysis was performed on plant species abundance data obtained from transect survey using minimum variance technique, also known as Ward’s method, which has been recognized as best way to classify the ecological communities and to identify community structure (Ludwig and Reynolds 1988). The cluster analysis generated a dendrogram providing hierarchically nested groups or clusters representing distinct plant communities, which were represented by subclusters.

7.3.3 Floristic Diversity and Distribution

Results are presented in the form of the impact of mining on vegetation communities. In order to achieve the objective, the entire study area was stratified into 2 km circular zones around each mining site.

7.3.4 Overall Floristic Composition

Kachchh district falls under arid to semiarid climate which demarcated by the xerophytes or majority of annual plant species. During the present study, a total of 326 plant species (including 1 Gymnosperm, i.e. Ephedra foliata) belonging to 63 families and 210 genera were recorded (Table 7.2). Out of them 14 species are of conservation significance that include above-mentioned gymnosperm species (Table 7.3). Poaceae family was the most dominant with 45 species followed by Fabaceae and Convolvulaceae with 24 and 20 species. Twenty-four families were recorded with 24 species while 27 families for single species. Dicotyledonous plants were represented with 81% (n = 264 species) with 7% of tree/small tree, 15% of shrub/under shrub and 12% of climber/twiner/parasite and 50% of herbs, whereas monocotyledonous group was represented by 19% of species of grasses and sedges. Abundance and % frequency ratio reveal that all species were contagiously distributed, while overall plant community was heterogeneous in nature, which was revealed from the frequency diagram. Acacia senegal and Salvadora oleoides were the most dominant species with high IVI value (76.60). The mean ± SD numbers of tree, shrub and climber species were 4.18 ± 0.14, 9.12 ± 0.30 and 3.37 ± 0.17, respectively. Mean density ± SD for tree, shrub and climber were 122.58 ± 5.75, 285 ± 8.97 and 56.32 ± 3.72, respectively. Shannon diversity index values for tree, shrub and climber are with mean ± SD of 1.01 ± 0.03, 1.71 ± 0.05 and 0.88 ± 0.05, respectively. Species richness index (Menhinick index) values of tree, shrub and climber are with mean ± SD of 0.91 ± 0.029, 1.21 ± 0.03 and 1.00 ± 0.032, respectively. Species evenness for tree, shrub and climber are with mean ± SD 0.75 ± 0.01, 0.71 ± 0.01 and 0.88 ± 0.007, respectively.

Table 7.2 List of plant species reported from study area
Table 7.3 Plant species with conservation significance

7.4 Impact of Mining Activities on Plant Community Structure

7.4.1 Overall Description of Impact of Mining

Around the mining area, major Prosopis forest was found because of the high tolerance value of Prosopis juliflora against the extreme edaphic condition and some patches of Euphorbia-Salvadora and Acacia forest, but they were confronted of the tremendous pressure of extinction. Campylanthus ramosissimus, Commiphora wightii, Dactyliandra welwitschii, Helichrysum cutchicum, Heliotropium bacciferum var. suberosum, H. rariflorum, etc. are conservation significant plant species found around the mining area. Including mining activities, other disturbance was also occurring which causes the impact on the biodiversity. ANOVA results reveal that there was no significant difference in plant communities between nine zones in terms of various diversity indices, viz. number of species, density, Shannon diversity, Menhinick and Margalef species richness, evenness, etc.

Cluster analysis reveals that the greater similarity among sampling points belongs to the same kind of habitat. Four distinct groups (Z2, Z9:Z8:Z7:Z5, Z6:Z3, Z1:Z4) were clearly identified from abundance data of tree species, while on the basis of shrub species, four distinct groups (Z5, Z9:Z8:Z6, Z4:Z3, Z7:Z2:Z1) were clearly identified also (Figs. 7.2, 7.3 and 7.4).

Fig. 7.2
figure 2

Zone-wise clustering of tree species

Fig. 7.3
figure 3

Zone-wise clustering of shrub species

Fig. 7.4
figure 4

Rarefaction curve of various mining zones for tree species

7.4.1.1 Zone-Wise Description of Floristic Diversity

Zone One (Z1)

A total of 147 plant species were recorded from the zone one (Z1) including 12 trees, 26 shrubs, 15 climbers, 74 herbs and 20 grasses. Abundance and % frequency ratio reveals that all tree, shrub and climber species were contagiously distributed within area. Important value index shows that the Acacia senegal was the most dominant species (IVI = 114.68) followed by A. nilotica (101.50). Frequency diagram of the tree, shrub and climber reveals that the zone was heterogeneous in nature. Various diversity indices were obtained from the data of tree, shrub and climber. Mean ± SD numbers of species occurring in transect were 4.5 ± 2.54, 9.26 ± 4.04 and 2.66 ± 1.66 for tree, shrub and climber, respectively, while densities of the said habitat were 118.75 ± 76.10, 248.33 ± 84.80 and 48.33 ± 39.56. Shannon diversities of the tree, shrub and climber were in mean ± SD, 1.11 ± 0.47, 1.82 ± 0.53 and 0.71 ± 0.60, respectively, while Menhinick species richness values with mean ± SD were 0.98 ± 0.49, 1.29 ± 0.43 and 0.93 ± 0.30 for tree, shrub and climber, respectively. Evenness values of tree, shrub and climber with mean ± SD were 0.78 ± 0.14, 0.74 ± 0.13 and 0.93 ± 0.09, respectively (Figs. 7.5 and 7.6).

Fig. 7.5
figure 5

Zone-wise diversity indices of tree species

Fig. 7.6
figure 6

Zone-wise diversity indices of shrub species

Zone Two (Z2)

A total of 199 plant species were recorded from the zone two including 14 trees, 26 shrubs, 20 climbers, 112 herbs and 27 grasses. Tree, shrub and climber species were contagiously distributed within area, which was revealed from % frequency and abundance ratio. Acacia nilotica was the most dominant and ecologically successful species with high IVI value (142.05) followed by A. senegal (78.51). Frequency diagram of all habit tree, shrub and climber reveals that the community of the zone two was heterogeneous in nature. A number of species of tree, shrub and climber with mean ± SD were 4.81 ± 2.88, 8.31 ± 3.62 and 3.41 ± 2.77, respectively, while densities with mean ± SD were 120.62 ± 73.45, 250 ± 96.64 and 69.16 ± 64.87 of tree, shrub and climber, respectively. Shannon diversities of tree, shrub and climber with mean ± SD were 1.05 ± 0.54, 1.63 ± 0.61 and 0.85 ± 0.62, respectively, while Menhinick species richness values of tree, shrub and climber with mean ± SD were 1.02 ± 0.41, 1.17 ± 0.36 and 0.92 ± 0.32, respectively. Species evenness values with mean ± SD were 0.75 ± 0.20, 0.69 ± 0.12 and 0.89 ± 0.11 of tree, shrub and climber, respectively (Figs. 7.5 and 7.6).

Zone Three (Z3)

Two hundred and eight plant species were recorded from the mining zone three with 7 trees, 33 shrubs, 19 climbers, 117 herbs and 32 grasses. Abundance and % frequency ratio reveals the all tree, shrub and climber species were contagiously distributed within area, while Acacia nilotica was ecologically successful and well-established species with high IVI value (126.90) followed by A. senegal (100.04). Comparative results with frequency diagram reveal that the plant community was heterogeneous in nature. Mean ± SD numbers of species were 4 ± 1.55, 9.33 ± 3.39 and 3.28 ± 2.67 of tree, shrub and climber respectively, while densities were 130.66 ± 85.70, 271 ± 129.99 and 63.57 ± 68.82 for tree, shrub and climber, respectively. Shannon diversities of tree, shrub and climber with mean ± SD were 0.98 ± 0.35, 1.77 ± 0.40 and 0.81 ± 0.70, respectively, while Menhinick species richness were 0.87 ± 0.29, 1.28 ± 0.25 and 0.95 ± 0.40 of tree, shrub and climber, respectively. Species evenness values of tree, shrub and climber with mean ± SD were 0.74 ± 0.15, 0.69 ± 0.13 and 0.90 ± 0.10, respectively (Figs. 7.5 and 7.6).

Zone Four (Z4)

A total of 219 higher plant species were recorded from zone four, in which there are 12 trees, 35 shrubs, 22 climbers, 124 herbs and 26 grasses. Abundance and frequency ratio of the tree, shrub and climber reveals that all species of the said habitat were contagiously distributed within area. Prepared frequency diagram compared with the normal frequency diagram reveals that the vegetation was heterogeneous in terms of nature. In that heterogeneous vegetation, the most dominant and ecologically successful tree species was Acacia senegal with high IVI value (106.25) followed by Salvadora oleoides (81.14) and A. nilotica (76.99). A number of species occurring during the transect were with mean ± SD for tree, shrub and climber, 4.42 ± 1.64, 9.42 ± 3.89 and 3.17 ± 2.05, respectively, while densities of the tree, shrub and climber were 108.46 ± 63.25, 288.84 ± 105.96 and 46.30 ± 30.90, respectively. Shannon diversity index values with mean ± SD of tree, shrub and climber were 10.4 ± 0.42, 1.72 ± 0.65 and 0.84 ± 0.65, respectively, while the Menhinick species richness index values were 1.03 ± 0.39, 1.26 ± 0.44 and 1.07 ± 0.45, respectively. Species evenness values of the tree, shrub and climber with mean ± SD were 0.71 ± 0.17, 0.69 ± 0.14 and 0.91 ± 0.10, respectively (Figs. 7.5 and 7.6).

Zone Five (Z5)

A total of 251 plant species were recorded in which there were 9 trees, 38 shrubs, 26 climbers, 142 herbs and 36 grasses. Abundance and % frequency ratio of each species of the tree, shrub and climbers reveals that all species were contagiously distributed within area, while a comparison of the frequency diagram of the various species with normal frequency diagram reveals that the plant community was heterogeneous in terms of nature. Among the heterogeneous communities, Acacia senegal was the most dominant and well-established species with high IVI value (102.38) followed by Salvadora oleoides (101.08). Mean ± SD numbers of species occurring during transect of tree, shrub and climber were 4.5 ± 1.64, 9 ± 3.91 and 3.46 ± 2.08, respectively, while densities with mean ± SD were 139.28 ± 84.18, 290 ± 3.91 and 56.15 ± 37.26 of tree, shrub and climber, respectively. Shannon diversity values of tree, shrub and climber with mean ± SD were 1.06 ± 0.42, 1.75 ± 0.65 and 0.92 ± 0.60, respectively, while Menhinick species richness values with mean ± SD were 0.97 ± 0.40, 1.18 ± 0.41 and 1.05 ± 0.35 of tree, shrub and climber, respectively. Species evenness values of tree, shrub and climber were 0.71 ± 0.20, 0.73 ± 0.10, 0.87 ± 0.09, respectively (Figs. 7.5 and 7.6).

Zone Six (Z6)

Two hundred and forty-four total plant species were recorded from zone six including 8 trees, 39 shrubs, 27 climbers, 137 herbs and 33 grasses. Abundance and % frequency ratio reveals that all tree, shrub and climber species were contagiously distributed within area, while a comparative result of frequency diagram shows heterogeneous nature of community. The important value index (IVI) reveals that the Acacia senegal (93.78) was most dominant species followed by Salvadora oleoides (88.02). Mean ± SD numbers of species of tree, shrub and climber were 3.36 ± 1.86, 9.45 ± 4.47 and 4.09 ± 3.08, respectively, while densities were 115 ± 73.58, 320.22 ± 192.91 and 76.19 ± 69.85, respectively. Shannon diversity index mean values of tree, shrub and climber were 0.80 ± 0.54, 1.75 ± 0.65 and 1.03 ± 0.63, respectively, while Menhinick species richness index values of tree, shrub and climber were 0.81 ± 0.45, 1.22 ± 0.39 and 1.05 ± 0.37, respectively. Species evenness means ± SD value of tree, shrub and climber were 0.79 ± 0.26, 0.72 ± 0.14 and 0.85 ± 0.08, respectively (Figs. 7.5 and 7.6).

Zone Seven (Z7)

A total of 236 plant species were recorded from the zone seven including 9 trees, 41 shrubs, 19 climbers, 133 herbs and 34 grasses. Distribution pattern of the tree, shrub and climber species within area was contagious as revealed by the ratio of abundance and % frequency, while comparative results of frequency diagram show that the plant community of this zone was heterogeneous in nature. Mean ± SD numbers of species forming transect were 4.36 ± 1.30, 11 ± 3.24 and 3.38 ± 1.68 of tree, shrub and climber, respectively, while densities were 132 ± 69.96, 325 ± 95.81 and 64.72 ± 47.04, of tree, shrub and climber, respectively. Shannon diversity index mean ± SD values of tree, shrub and climber were 1.11 ± 0.30, 1.96 ± 0.47 and 0.90 ± 0.55, respectively, while Menhinick species richness index mean ± SD values were 0.93 ± 0.34, 1.36 ± 0.30 and 1.01 ± 0.39 of tree, shrub and climber, respectively. Species evenness mean ± SD values of tree, shrub and climber were 0.75 ± 0.14, 0.69 ± 0.13 and 0.85 ± 0.10, respectively (Figs. 7.5 and 7.6).

Zone Eight (Z8)

A total of 237 plant species were recorded from the community of zone eight (Z8) in which there are 10 trees, 33 shrubs, 18 climbers, 139 herbs and 37 grasses. All of the tree, shrub and climber species were contagiously distributed within area, which were results of abundance and % frequency ratio. The nature of community was heterogeneous, which was revealed by the comparison of frequency diagram to normal frequency diagram. Among this heterogeneous plant community, Salvadora oleoides was the dominant and ecologically successful species with high IVI value (92.97) followed by Acacia senegal (71.96). Mean ± SD numbers of species of tree, shrub and climber were 4.22 ± 1.89, 8.72 ± 3.56 and 3.16 ± 1.85, respectively, while densities of tree, shrub and climber were 115.27 ± 72.07, 276.38 ± 113.05 and 40.83 ± 31.63, respectively. Shannon diversity index mean values of tree, shrub and climber were 1.05 ± 0.46, 1.72 ± 0.60 and 0.90 ± 0.52, respectively, while Menhinick species richness index mean (±SD) values of tree, shrub and climber were 0.93 ± 0.31, 1.19 ± 0.41 and 1.13 ± 0.28, respectively. Species evenness mean ± SD values of tree, shrub and climber were 0.77 ± 0.13, 0.69 ± 0.13 and 0.90 ± 0.07, respectively (Figs. 7.5 and 7.6).

Zone Nine (Z9)

A total of 225 species were recorded from the plant community of zone nine in which there were 9 trees, 31 shrubs, 23 climbers, 129 herbs and 33 grasses. Abundance and % frequency ratio reveals that all tree, shrub and climber species were contagiously distributed within area, while the frequency diagram reveals that the plane community was heterogeneous in nature. Important value index reveals that the Salvadora oleoides (135.69) was the most dominant species followed by Acacia senegal (119.70) of that community. Mean ± SD numbers of species of tree, shrub and climber were 3.59 ± 1.79, 7.78 ± 5.59 and 3.36 ± 2.05, respectively, while mean ± SD density values of tree, shrub and climber were 121.13 ± 96.44, 270.65 ± 110.10 and 46.59 ± 34.37, respectively. Shannon diversity index mean (±SD) values of tree, shrub and climber were 0.92 ± 0.52), 1.35 ± 0.89 and 0.89 ± 0.60, respectively, while Menhinick species richness index mean ± SD values of tree, shrub and climber were 0.83 ± 0.25, 0.99 ± 0.60 and 1.11 ± 0.39, respectively. Species evenness mean ± SD values of tree, shrub and climber were 0.82 ± 0.12, 0.71 ± 0.16 and 0.88 ± 0.10), respectively (Figs. 7.5 and 7.6).

Vegetation of the study area is mainly composed of thorny plant species. Upper canopy consists of the most dominant plant species like Acacia nilotica, A. senegal and Salvadora oleoides while major component lower canopy mainly comprising shrubby vegetation are Prosopis juliflora, Euphorbia caducifolia, Capparis decidua, Grewia tenax and G. villosa. Higher species diversity, richness and evenness of shrubs suggest their dominance over tree community in our study area. Kachchh region has specific forest type called tropical thorn forest, which consists the thorny shrub species. Shrub vegetation is dominated by thorny shrub species, while open areas are dominated by the annual plants which are mainly supported by scanty rainfall during monsoon. Most of the species were contagiously distributed within area and create heterogeneous plant community in nature. A total of 326 plant species were recorded including trees, shrubs, climbers/twiners, herbs and grasses.

The results of impact assessment of mining activities on vegetation of the total nine circular zones were identified within 2 km width because mining causes massive damage to landscape and biological communities, creating unfavourable habitat condition for plant growth. It was found that there was more or less impact in the inner zones of mining impact areas delineated for the present study. The number of tree and shrub species drastically decreased in their number due to mining. The unfavourable habitat conditions prevailing the coal-mined area have reduced the chances of regeneration of species, thereby, reducing the number of species in the mined areas. Habitat was changed by dumping of the contaminated soil on the natural habitat and by succession of extreme condition-tolerant plant species like Prosopis juliflora that occupies the landscape. Results of the study reveal that the positive impact on the vegetation of mining and associated activities creates the anthropogenic pressure on the local plant community structure and diversity. The main tree species like Acacia senegal, A. nilotica and Salvadora oleoides face heavy threats of extinction from the area. The present study led to the conclusions that phytosociological analysis can be used as important tools for predicting the suitability of mine-spoil habitats for plant growth, which would be helpful to revegetation of the mined areas. To minimize the impact of mining on the local vegetation, conservation strategies and extensive plantation drive of the local or native plant species in contaminated mining area are urgently required by long-term monitoring of the plantation.

7.5 Conservation and Management Strategies

With the clear understanding of environmental, ecological and social component with respect to the vegetation formation in the western Kachchh and widely available management options, the following strategies are suggested as conservation measures for rare, endangered and threatened (RET) plants and ecologically sensitive areas (ESAs).

  1. A.

    In view of high floral diversity and increasing biotic pressures on vegetation, the following management strategies for NSS and LBS are suggested for the conservation of threatened floral species:

    • A detailed status survey of threatened floral species needs to be done, and the habitats supporting high floral diversity should be identified and notified as mini-core areas (MCAs) of the sanctuary.

    • Strict protection should be given to these MCAs to prevent wood cutting, grazing and collection of other natural resources.

    • Degraded habitats should be restored through restoration programmes with threatened species and their associate species.

    • Herbaceous cover (including grass cover), of both dense and sparse grasslands, provides potential habitat for seven rare herbaceous species such as Helichrysum cutchicum (470 nos.), Heliotropium bacciferum (375), Heliotropium rariflorum (325), Tribulus rajasthanensis (95) and Indigofera caerulea (24). A detailed status survey of these herbaceous covers and identification of the location where there is very high abundance was recorded.

    • There are 26 villages with the population of 10,872 located in and around the sanctuary (recent survey part of this project), and they depend on the sanctuary area for grass, firewood, fodder and other nontimber forest produce. Based on the dependency level, few villages, i.e. Ratipar, Gugariyana, Midhiyari and Amiyu, should be considered for eco-development.

    • Creation of village-level eco-development committees, distribution of fuel-efficient stoves, development of fodder and fuel wood plantation, restoration of grassland and creation of employment alternative for economically poor villagers should be included in management planning for the overall conservation of threatened biological diversity of NSS and LBS.

  2. B.

    Within the study area like Lifri, Guneri, Siyot, Haripar, Walka nana, Walka mota and Lakhapar, out of seven localities, eight species of threatened plants were recorded. Even though these RFs were under control, severe wood cutting and grazing pressures were reported from the neighbouring villages. Based on the abundance of threatened species and existing threats, the following management strategies are suggested:

    • RFs like Gugaryani, Chakrai, Mindhiari, Kaniari, Maniyara and Khirsara RFs and adjoining forest stretches should be resurveyed for the complete status assessment of RET plants.

    • Within these RFs, the areas with the abundance of above-mentioned threatened species need to be identified.

    • The identified stretches should be protected by providing natural green fence by planting Euphorbia caducifolia and Balanites aegyptiaca.

    • Local villagers should be informed to avoid grazing and cutting of trees in the potential locations identified.

    • Apart from the protection measures, the concerned forest staff should be employed for monitoring these protected sites to assess and understand the productive potential of the threatened species.

    • These protected sites in RFs should be resorted with the threatened plants, and it can act as natural seed banks for the threatened species.

  3. C.

    Biodiversity is under varying degrees of protection in the protected and normal areas, but considerable proportion of biodiversity is manifested outside these areas and is existing without any protection. Realizing the limitation in declaring more areas as protected areas, and the need of participation of local communities in conservation efforts, the concept of ESAs is gaining popularity. During this study the following areas have been reported with diverse plant species and some threatened flora. Since these areas are not having any protected area status, they are under severe biotic pressures from the local villagers. (However for these areas reported for high floral diversity, area-specific floral richness is not discussed.)

  1. I.

    Piper: It is a hilly tract located close to Narayan Sarovar and provides high diversity of floral species. A total of 108 plant species were reported during this study. Further it has been found that this hilly area provides habitat for six threatened species.

  2. II.

    Kharai: This area is dominated by thorn scrub forest holding 127 plant species. Out of these plant species, around 42 were medicinal plants and 5 of conservation significance.

  3. III.

    Punrajpur: This unprotected area supports 106 plant species in which 5 are of conservation significance.

  4. IV.

    Vayor and Vagoth: This village area, i.e. hedges of agriculture and common grazing land, harboured the total five plant species of conservation significance including highly endangered plant species Corallocarpus conocarpus.

    • Considering the high floral diversity of these village areas, it has been suggested that these microhabitats should be declared as ecologically sensitive areas (ESAs).

    • Villager’s dependency on the identified hot spots for their natural resources like fodder and fuel wood collection should be identified.

    • Many government, nongovernment (Shrujan, Ashapura Foundation, Setu-Abhiyan, KMVS and VRTI) and research organizations are currently working with the villagers of proposed ESAs. In order to reduce the biotic pressures on the identified hot spot areas, the dependent villagers should be connected with these organizations to get socio-economic benefit.

    • All the identified hot spots should be strictly prohibited for grazing and wood cutting by providing alternate resources through the ongoing developmental programmes.

    • Creation of village-level forest and resources management committees and monitoring of the ESA are suggested.