Keywords

1 Introduction

Tropical forests are biodiversity-rich centres on earth and harbour approximately two-thirds of all living organisms (Hughes et al. 1997) including 96% of tree species (Poorter et al. 2015). The tropical forest biome comprises of diverse ecosystems between Tropic of Cancer and Tropic of Capricorn and spans across the Americas, Asia, Africa and Australia, and has the richest biodiversity with a unique environment. According to Miles et al. (2006), dry deciduous forest covers 6% (1,048,700 km2) of the tropics. Around 40% of the earth’s subtropical area is occupied by open or closed forest and of which 42% are tropical dry deciduous forest, 33% are moist forest and 25% are wet forest (Murphy and Lugo 1986). About 54.2% of the world’s tropical dry forests are in South America and the rest are equally distributed in North and Central America, Eurasia, Africa, Southeast Asia and Australasia (Miles et al. 2006). In southeast Asia, 30% of forests in the mainland are classified as dry forest (Blackie et al. 2014).

Tropical forests of Asia, especially Eastern and Western Ghats are under threat owing to human activities and are consequently, being replaced by inferior species (Bahuguna 1999). Tropical rain forests are extensively studied compared to dry forests (Losos and Leigh 2004). Although tropical dry deciduous forests are highly degraded and converted to other land uses, they had little attraction among the researchers and the general public (Bullock et al. 1995; Rundel et al. 1995). Tropical dry deciduous forests are rich in economically important species and are known to provide high potential timber revenue (Mohapatra and Tewari 2005). Hence, there is a growing interest on dry forests in the recent past (Miles et al. 2006). Dry deciduous forests are one of the most exploited ecosystems in the world (Murphy and Lugo 1986; Janzen 1988; Gentry 1992), as they are more prone to fire in the dry season (Giriraj et al. 2010). Documentation of biodiversity patterns is essential to prioritize areas for conservation programmes (Villasenor et al. 2007). Information on structure and composition of tropical dry deciduous forests is needed to conserve and restore these threatened ecosystems. The quantitative floristic analysis of the forest provides the necessary information for future planning and management (Phillips et al. 2003).

India, being a mega-diverse country, covers about 2% of the global forest area and is one of the richest repositories due to the presence of different types of vegetation and they hold unique flora and fauna. Tropical forest in India occupies 86% of the total forest cover (Singh and Singh 1988), of which 54% are dry deciduous, 37% are moist deciduous and the remaining is wet-/semi-evergreen (Kaul and Sharma 1971). The Eastern Ghats are a fragmented hill-chain, starting from Odisha to Tamil Nadu. Studies exploring the structure and composition of forests in Eastern Ghats of Tamil Nadu are limited (Kadavul and Parthasarathy 1999a, b, 2000; Chittibabu and Parthasarathy 2000a, b, 2001; Jayakumar et al. 2002; Pragasan and Parthasarathy 2010; Muthumperumal and Parthasarathy 2013; Sundarapandian et al. 2015). However, studies on plant diversity in tropical dry deciduous forest Sathanur reserve forest is almost nil except for preliminary result output of us (Gandhi and Sundarapandian 2014a, b). Hence present study is intended to study the vegetation structure and species composition of the tropical dry deciduous forest at Sathanur reserve forest of Eastern Ghats. Furthermore, an attempt was also made to understand the factors responsible for spatial variation in plant diversity.

2 Materials and Methods

2.1 Study Area

Sathanur reserve forest (longitude 78°51′10″ and latitude 12°4′48″), a part of Chennakesava hills, Tamil Nadu, India, spread over 870 ha (Fig. 1) was chosen for the present study belonging to a part of Eastern Ghats, India. The Eastern Ghats experience heavy pressure due to illegal logging, collection of fodder, fuelwood, medicinal plants, etc. and thereby are losing its vegetation at an alarming rate (Jayakumar et al. 2002). The vegetation of this region is dry deciduous forest type (Type 7/CI of Champion and Seth 1968) based on the Champion and Seth (1968) classification of Indian forests.

Fig. 1
figure 1

Location of 30 one-hectare study plots (indicated by pink dots with numbers) in the Sathanur reserve forest of Eastern Ghats, India

Full size image

Sathanur Reserve forest receives a bimodal pattern of rainfall, with maximum rain during north-east monsoon (September–December) and very less and inconsistent rainfall during the south-west monsoon (May–July). The average annual rainfall for 44 years (1972–2015) was 965.49 mm and mean monthly maximum temperatures ranged between 28 and 37 °C while mean monthly minimum temperatures varied from 19.6 to 26.8 °C (Fig. 2). The major soil types in the district are red loam and black soil and the red loam soil is predominantly found in Sathanur reserve forest (NADP 2008). The texture of the soil was sandy loam in most of the plots while the plots near to the rivulet were sandier than the other study sites. However, pebbles are abundant in the soil in most of the plots.

Fig. 2
figure 2

Mean monthly rainfall (44 years) and temperature (study period) of the Sathanur reserve forest, Eastern Ghats

Full size image

2.2 Vegetation Analysis

Thirty square plots of 1 ha each were laid approximately at 500 m intervals in the Sathanur reserve forests during the period of November 2013 to February 2015 (Fig. 1) which were further sub-gridded into 10 m × 10 m size (100 m2) quadrats as easy, workable units. All the individual plants with ≥10 cm GBH were enumerated and their girth were measured at 1.37 m from the ground level. In multi-stemmed trees, GBHs were measured separately, after which basal area was calculated and summed up. Within each plot, 50 quadrats of 5 m × 5 m were laid in a systematic sampling method to enumerate shrubs and climbers. Similarly, fifty quadrats of 1 m × 1 m each were laid for herbs in each 1 ha plot and studied during October and November (peak growth period). Shrubs and herbs, the diameter was measured at 3 cm above the ground of each individual using vernier caliper. Plant specimens were collected and identified with confirmation of taxonomist using floras (Gamble and Fischer 1987; Matthew 1991). The cut stems were enumerated in all the plots and the disturbance index was computed based on the number of cut stems divided by the total number of stems including cut stems (Rao et al. 1990). The vegetation data collected in each plot was analyzed for analytical and synthetic quantitative characteristics. The diversity indices were computed using the freely downloadable PAST 3.1 program (version 3.1; Øyvind Hammer, Natural History Museum, University of Oslo). Importance value index (IVI) was the sum of the values of relative frequency, relative density, and relative basal area (Curtis and McIntosh 1950). The abundance to frequency (A/F) ratio for different species was determined by following Whitford (1949). The ratio indicates regular (<0.025), random (0.025–0.050), and contagious (>0.050) distribution pattern.

Some plots (plot nos. 1–10) in the present study were laid near roads, human settlement or the agricultural fields which are more vulnerable to human exploitation. Although the study area is a reserve forest, locals frequently cut trees and collect firewood, lop branches and graze their cattle. Illegal selective cutting of Chloroxylon swietenia for fencing, agricultural tools and other domestic purposes and Albizia amara for firewood are quite frequent in this forest. Plots 11–20 are relatively less disturbed than the other plots. Study plots (plot nos. 21–30) were laid on both the sides of a rivulet from Ponnaiyar river. In general, the plots near the rivulet are also subjected to human disturbances. These plots had a rocky terrain. People regularly use the rivulet for day to day activities. In addition, it also serves as a source of drinking water for cattle and hence, these plots are also under high anthropogenic pressures. Soil moisture and pH values showed wide variation among the study plots. Soil pH ranged from 5.9 to 7.1 and the moisture content ranged from 1.28 to 24.0%. The mean soil bulk density value of the three layers ranged from 1.21 to 1.82 g/m3. Coarse fragment (>2 mm size) fraction in the soil showed wide variations (33.9% of samples contain coarse fractions in all the three depths; 10.2% samples contain coarse fractions in surface (0–10 cm) and middle (10–20 cm) layers; 15.3% samples contain coarse fractions in middle and bottom layers (20–30 cm); 6.7% samples have coarse fractions only in the bottom layer) among the plots, samples and depths. However, 33.9% of the samples collected did not have any coarse fragment (Gandhi 2016; Gandhi and Sundarapandian 2017).

3 Results

3.1 Species Richness and Diversity

A total of 210 plant species belonging to 163 genera and 63 families were enumerated from 30 one-hectare plots in the tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats. Among these, trees and herbs formed the major proportion represented by 75 (35.71%) species each followed by climbers (15.71%) and shrubs (12.86%). The total species richness of study plots ranged from a minimum of 35 species/ha to a maximum of 84 species/ha with a mean of 61.5 ± 2.5 species/ha (Table 1). Among different life-forms, the species richness varied from 7 to 28 species/ha for trees; 10–36 species/50 m2 for herbs; 5–14 species/0.125 ha for shrubs and 3–23 species/0.125 ha for climbers. We encountered a total of 17,525 tree stems (≥10 cm GBH) from 30 one-hectare study plots. The density ranged from 336 stems/ha to 1075 stems/ha with a mean of 584 ± 38 stems/ha. For climbers and shrubs, it ranged from 9–252 individuals/0.125 ha and 247–969 individuals/0.125 ha respectively. The density of understory species varied considerably from 875 to 6567 individuals per 50 m2. The total basal area registered by trees across the study plots was 561.3 m2/ha and on individual plot, their basal area ranged from 7.2 to 43.1 m2/ha, with a mean of 18.7 ± 1.5 m2/ha. Similarly, for shrubs and climbers, it ranged from 0.16 m2/ha to 0.8 m2/ha and 0.06 m2/ha to 3.3 m2/ha respectively. The mean basal area for herbs in the study plots was 3.95 ± 0.3 m2 per 50 m2 and it ranged from 0.95 to 6.6 m2 per 50 m2. The diversity index, Shannon value ranged from 0.72–2.3 for tree species; 0.81–1.9 for shrubs; 1.16–2.8 for herbs and 0.24–2.8 for climbers (Table 2). Dominance index of tree species ranged from 0.14 to 0.7 with a mean of 0.4. Fisher’s alpha index ranged from 1.25–6.36 for trees, 0.75–2.86 for shrubs, 1.23–5.4 for herbs and 0.76–8.1 for climbers.

Table 1 Summary of plant diversity inventory carried out in 30 one-hectare plots of tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats (T—tree, S—shrub, H—herb, C—climbers)
Full size table
Table 2 Diversity indices of four dynamic plant life forms in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats (T—tree, S—shrub, H—herb, C—climbers)
Full size table

3.2 Species Composition

Among 75 tree species enumerated, Albizia amara was the most dominant tree species represented by 9871 individuals which is almost 56% of the total stems enumerated from 30 one-hectare study plots (Table 3). The other dominant tree species in the study plots were Chloroxylon swietenia (3171 individuals), Azadirachta indica (553 individuals), Pongamia pinnata (477 individuals) and Acacia catechu (442 individuals).

Table 3 Life form abundance in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats
Full size table

The top five species together comprised 82.8% of the total tree species’ density (Table 3). In contrast, 32 species including Strychnos nux-vomica, Gmelina asiatica and Cassia fistula were represented by <10 individuals and four species viz. Borassus flabellifer, Cordia monoica, Dalbergia oliveri, and Delonix elata were represented by just one individual across the study plots. Lantana camara (4661), Clausena heptaphylla (3970) and Tarenna asiatica (3186) were the predominant shrub species encountered across the study plots. These three species together contributed to 81% of the total shrub species density. However, 11 species among 27 were represented by <10 individuals and four species viz. Allophylus serratus, Cadaba fruticosa, Grewia abutilifolia and Ziziphus rugosa were represented by a single individual in all the study plots. The herbaceous community of the study plots was dominated by Sida cordifolia (22,038 individuals), Ageratum conyzoides (13,510 individuals) and Sida acuta (9245 individuals). Among 75 species, two species including Alysicarpus monilifer and Amaranthus spinosus were represented by single individual and seven species viz., Acanthospermum hispidum, Agave americana, Barleria lupulina, Blepharis boerhaviifolia, Commelina diffusa, Euphorbia thymifolia and Hyptis suaveolens were encountered with <10 individuals.

Pterolobium hexapetalum (899 individuals) was the predominant climber species across the study plots followed by Combretum albidum (262 individuals) and Acacia caesia (157 individuals). The top five climber species comprised 70.41% of the total liana density across the study plots. Six species including Ampelocissus tomentosa, Carissa spinarum, Dioscorea bulbifera, Maerua oblongifolia, Pisonia aculeata and Toxocarpus kleinii were represented by mono individuals and three species such as Cissampelos pareira, Jasminum angustifolium and Trichosanthes cucumerina were represented by less than ten individuals. Alien invasive species such as Lantana camara (100%) and Ageratum conyzoides (93%) were the most commonly distributed and dominant species in shrub and herbaceous communities. Another exotic invasive species Prosopis juliflora was also recorded in 43% of plots.

3.3 Importance Value Index and Family Diversity

Albizia amara was the predominant tree species with regard to IVI value across all the 30 one-hectare study plots followed by Chloroxylon swietenia and Acacia catechu (Tables 4, 5, 6 and 7). Among shrub species, Lantana camara, Clausena heptaphylla and Tarenna asiatica scored greater IVI values, whereas Sida cordifolia, Ageratum conyzoides and Sida acuta were among the predominant herb species in terms of IVI values. Pterolobium hexapetalum registered the highest IVI values among climber species followed by Combretum albidum and Acacia caesia.

Table 4 Importance value index of top ten tree species in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats
Full size table
Table 5 Importance value index of top ten shrub species in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats
Full size table
Table 6 Importance value index of top ten herb species in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats
Full size table
Table 7 Importance value index of top ten climber species in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats
Full size table

Euphorbiaceae was the most speciose plant family across the study plots represented by 17 species followed by Fabaceae and Poaceae with 13 species each (Table 8). Plant families such as Euphorbiaceae and Fabaceae were represented in all the four life-forms studied. However, 24 families were represented by just one species in the present study.

Table 8 Family-wise contribution of all life forms (tree, shrub, herb, lianas) in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats (T—tree, S—shrub, H—herb, C—climbers)
Full size table

3.4 Species-Area Curve

Species-area curve for plant diversity of tropical dry deciduous forest of all life forms (trees, shrubs, herbs, climbers) of 30 one-hectare plots pooled data showed the number of species in all the plots increased initially, steeped up to 13 ha followed by gradual rise, until the sampling reached 23 ha after which, it flattened (Fig. 3). This indicates that sampling of 23 ha is sufficient to record the species richness of the study area. Hence, coverage of 30 ha sampling has been done in the present study exhibits the current status of Sathanur reserve forest.

Fig. 3
figure 3

Species-area curve of plant species (all life forms) in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image

3.5 Diameter Class Distribution

The density and species richness of trees decreased with increasing tree diameter class (DBH) across the study plots (Figs. 4 and 5). This pattern was also consistent with dominant tree species like Albizia amara, Chloroxylon swietenia, Pongamia pinnata, and Azadirachta indica which showed more or less similar trend in diameter class distribution (Fig. 6). Overall, the juvenile population of trees (3.2–10 cm DBH) contributed 50.3% of the total tree density, whereas, adult trees constituted 49.7%.

Fig. 4
figure 4

Diameter class-wise distribution of species richness and abundance of trees in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image
Fig. 5
figure 5

Diameter class-wise distribution of species richness and density of juvenile population of tree species in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image
Fig. 6
figure 6

Diameter class-wise distribution of dominant tree species in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image

We screened all the multi-stemmed tree individuals in 30 one-hectare plots and observed a maximum of nine branches below 1.37 m height (Fig. 7). The presence of multiple stems was more prevalent among Albizia amara and Pongamia pinnata. For calculating the disturbance index, we screened the presence of cut stems and damaged stems in each one-hectare study plot and found that a total of 1335 were removed from the base and 122 individuals were damaged. Disturbance index across the study plots ranged from 0.01 to 0.25 (Fig. 8). The most number of cut stems in the study plots belonged to Albizia amara and Chloroxylon swietenia followed by Acacia catechu and Pongamia pinnata.

Fig. 7
figure 7

Distribution patterns of species in a community in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image
Fig. 8
figure 8

Contribution of multi-stemmed trees to no. of individuals and species richness in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image

3.6 Distribution Pattern

The A/F ratio of species in plots indicates the species distribution patterns. Overall, 93.83% of species (of all life forms—trees, shrubs, herbs and climbers) showed contagious distribution, 5.31% showed a random distribution and 0.86% showed the regular distribution (Fig. 9). In tree community, 91.55% showed contiguous distribution, 8.27% showed random and 0.18% showed regular distribution patterns. In shrub community, 81.68% of shrub species showed the contiguous distribution while 13.55% showed random and 4.76% showed regular distribution patterns. Similarly, contiguous distribution was more common in herb (98.89%) and climber (97.02%) life forms.

Fig. 9
figure 9

Disturbance index of the study plots in tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats

Full size image

4 Discussion

The plant species richness recorded from 30 (1 ha) plots in the tropical dry deciduous forest of Sathanur reserve forest, Eastern Ghats was 210, of which, trees and herbs occupied a significant proportion (35.71% species each) followed by lianas (15.71%) and shrubs (12.86%). The observed species richness in the present study plots (35–84 species/ha) is greater than those reported in a tropical dry deciduous forest of northern Eastern Ghats (18–55 species/ha; Devi and Behera 2003) and in line with the findings of Reddy et al. (2008a, b) and elsewhere (Hubbel 1979; Jimenez et al. 2016). However, total species richness encountered in this study is much lower than those values (173–218 species/ha) reported by Reddy et al. (2011) from tropical moist deciduous forests of Eastern Ghats, Andhra Pradesh. Plant diversity is directly influenced by many factors including climate, topography, soil and natural/anthropogenic causes (Behera et al. 2016). All the 30 one-hectare study plots of Sathanur reserve forest are located within the radius of 10 km and an altitude ranging from 192 to 250 m, hence, the climate cum topography of study plots largely remain the same. The human disturbance varied considerably among the study plots due to closeness to the settlement, agricultural field and rivulet, roads and path intrusions. The plots located near to rivulet have more soil moisture and air humidity due to water flow. The plots laid along roadsides and near human settlements are subjected to anthropogenic pressure which has resulted in an increase in cut stems. Seepage from the agricultural land to forest area also increases soil moisture content in the study plots which are close to them. These microclimatic variations among the plots could be the reason for the significant spatial variation in species richness and density among the plots even though they are located within 10 km radius. The low species richness recorded in the present study compared to the Western Ghats and some regions of Eastern Ghats could be ascribed to low and erratic rainfall pattern, anthropogenic disturbances and extensive grazing.

Tree species richness is vital to tropical forest biodiversity because trees directly or indirectly support almost all other life forms (Huston 1994). Tree (for individuals ≥10 cm DBH) species richness recorded in the current study plots ranged 7–28 species/ha, which is much lower than previous studies conducted in other parts of Eastern Ghats, for example, Kadavul and Parthasarathy (1999a, b) reported 42 to 47 species/ha in Kalrayan hills and 33–50 species/ha in Shervarayan hills. Similarly, Chittibabu and Parthasarathy (2000a) encountered 26–54 species/ha in Kolli hills of Eastern Ghats. Reddy et al. (2008a) reported 60–69 species/ha in dry deciduous forests of Eastern Ghats. Premavani et al. (2014) estimated 34–48 species/ha in tropical dry forests of central Eastern Ghats. The relatively lower species richness values obtained in the present study when compared to the other studies (Palomino and Alvarez 2008; Powers et al. 2009; Rao et al. 2015; Naidu and Kumar 2016; Sahoo et al. 2017) could be attributed to the greater extent of anthropogenic activities prevailing in the study plots including stem cutting and stem removal in addition to low rainfall.

The species richness of climbers registered in this study (3–23 species/0.125 ha) is comparable to other areas of Eastern Ghats i.e., Kolli hills (2–17 species/ha; Chittibabu and Parthasarathy 2001), tropical dry deciduous forest of Srilankamalla wildlife sanctuary, Andhra Pradesh (6–27 species/ha; Mastan et al. 2015). However, the value obtained in the present study is lower than in tropical dry deciduous forest of Nallamalai–Seshachalam–Nigidi hills of Eastern Ghats (21–29 species/ha; Reddy et al. 2008a), in the Vindhyan tropical dry deciduous forest (37 species/ha; Sharma and Raghubanshi 2010) and in Mudumalai Wildlife Sanctuary, southern India (53 species/ha; Joseph et al. 2008). The species richness of climbers is comparable even though this study site has low species richness (includes all life forms) in general which could be attributed to anthropogenic perturbation as observed by Dewalt et al. (2000), Laurance et al. (2001) and Schnitzer and Bongers (2002). According to Gerwing and Uhl (2002) and Schnitzer et al. (2004), human activities including logging create canopy gaps, which lead to more light exposure that subsequently results in the successful establishment of climbers and lianas.

The density of climbers enumerated in the present study was 569.6 individuals/ha (range: 9–252 individuals/0.125 ha). This value is comparable with other reports from tropical forest of Eastern Ghats such as Kolli hills (12.5–56 individuals/ha; Chittibabu and Parthasarathy 2001) and Bobiri forest reserve, Ghana (152–280 individuals/ha; Addo-Fordjour et al. 2009). The wide variations in the climber density among plots could be due to different levels of anthropogenic pressures associated with canopy opening and other microclimatic conditions. The low density of climbers in few plots observed in the present study may be due to selective tree felling as stated by Chittibabu and Parthasarathy (2000b).

Species richness of understory vegetation in 30 one-hectare plots in the tropical dry deciduous forest of Sathanur reserve forest was 102 (fifty 25 m2 quadrats for shrubs and fifty 1 m2 quadrats for herbs in each one-hectare plot). Comparatively, the value obtained in the present study is moderately higher than the other parts of Eastern Ghats and elsewhere (52 species/0.32 ha (eight hundred 4 m2 quadrats) in tropical evergreen forest in the Kolli hills, Chittibabu and Parthasarathy (2000b); 84 species/4 ha in tropical dry deciduous forest in Odisha, Sahu et al. (2007); 32–93 species/0.05 ha (twenty 25 m2 quadrats) in riparian environments of Atlantic forests in Rernambuco, Brazil, Gomes-Westphalen et al. (2012)). However, this is lower than several reports of tropical forests in India (155 species/1.2 ha (three thousand 4 m2 quadrats in 30 ha plots) in tropical evergreen forest in Anamalais, Western Ghats, Annaselvam and Parthasarathy 1999; 170 species/3 ha (shrubs + herbs) in tropical dry deciduous forest in Nallamalai-Seshachalam-Nigidi hills of Eastern Ghats, Andhra Pradesh, Reddy et al. 2008a; 107 species/2.04 ha (shrubs + herbs) in tropical dry deciduous forest in Similipal biosphere reserve, Odisha, Reddy et al. 2008b; 269 species in tropical forest tract of Sileur-Maredumilli hills of north Eastern Ghats, Reddy et al. 2011). The wide variations in understory species richness may be influenced by various ecosystem processes like nutrient cycling, decomposition of organic matter, hydrological cycle, soil formation, species composition of overstory, canopy cover, light penetration, different levels of disturbances, etc., as suggested by Singh et al. (2014). Comparison of understory diversity is very difficult because of the variations in the area of sampling, size of quadrats, number of quadrats etc. as suggested by Chittibabu and Parthasarathy (2000a, b). However, the value recorded in the present study is well within the range (Annaselvam and Parthasarathy 1999). The population density of understory species (shrubs and herbs) varied considerably among the species as well as among the plots. In shrub community, the three dominant species occupy 81% of populations. Lantana camara population contributed 32% of shrub composition. In the herbaceous community, five dominant species were represented by 62,318 individuals (57%) in all the study sites. In the herbaceous community, annuals were more predominant than that of perennials. Similarly, annuals were more predominant in the herbaceous community in tropical evergreen forests of Anamalais, Western Ghats as observed by Annaselvam and Parthasarathy (1999). In contrast, perennials dominated the understory community in Amazonian rainforest (Poulsen and Balslev 1991). The prevalence of annuals in the herbaceous community as observed in this study could be largely due to seasonal variation i.e. hot summer (herbaceous vegetation is completely dried off and rejuvenate in rainy season) and overgrazing. The shrub species richness is low in few study plots, owing to tree species dominance. The presence of Lantana camara, an alien weed, in all the studied plots indicates that it would have probably invaded the Sathanur reserve forest several years back and had successfully established itself (Behera and Misra 2006). The low density of herbaceous vegetation in the inner plots is because of low insulation on the forest floor as the canopy is closed.

The species diversity is dependent on the capability of species to adapt, which increases as the community becomes more stable. Species diversity is brought about by species interaction like competition as well as niche variation (Pianka 1966), which are prominently expressed in the tropical regions because of high temperature and humidity (Ojo and Ola-Adams 1996). Shannon’s index is generally higher for tropical forests (Knight 1975), whereas in Indian forests, the reported range was 0.83–4.1 (Singh et al. 1981; Sundarapandian 1997). In the present study, Shannon’s index value ranged from 0.716–2.343 for tree species; 0.812–1.948 for shrubs; 1.157–2.8 for herbs and 0.243–2.796 for climbers. It is very difficult to compare diversity indices due to variations in the sampling location and uneven plot dimensions. The greater dominance index could be due to the mono-species dominance exhibited by Albizia amara in this forest ecosystem.

The density of species is directly dependent on species richness (Denslow 1995; Condit et al. 1998). The extent of tree density contributes as much to the forest’s functional diversity, ecological processes and ecosystem services (Gopalakrishna et al. 2015). The mean tree density of 584 stems/ha registered in the present study is closer to the Amazonian average (597 stems/ha; Lewis et al. 2004) and Bornean (Asia) average (602 stems/ha; Slik et al. 2010), and 28.8% higher than the tropical forest average (425 stems/ha; Lewis et al. 2013) of Africa. Similarly, the value recorded in the present study is in line with those reported by Pragasan and Parthasarathy (2010) in the southern Eastern Ghats (457 stems/ha); Reddy et al. (2008b) in Similipal biosphere reserve (568 stems/ha) and Sahu et al. (2007) who reported 591 stems/ha in tropical dry deciduous forest, Odisha. However, the mean tree stem density values in the present study are lower than the findings of Kadavul and Parthasarathy (1999a, b) who reported 815 stems/ha in Shervarayan hills of southern Eastern Ghats. Similarly, Reddy et al. (2008a, 2011) reported 735 stems/ha and 709 stems/ha respectively in the tropical dry deciduous forest, Andhra Pradesh. The observed tree species density in the present study is higher than the findings of Sahu et al. (2012), Premavani et al. (2014), and Sahu et al. (2016) with 443 stems/ha, 360–526 stems/ha and 479 stems/ha respectively in the northern Eastern Ghats. Similarly, low stand density was recorded from other tropical forests of the world: Brazil (420–777 stems/ha; Campbell et al. 1992), Costa Rica (448–617 stems/ha; Heaney and Proctor 1990) and Malaysia (250–500 stems/ha; Primack and Hall 1992). Therefore, the observed density of trees in the present study can be considered modest when compared to the similar forest types in the Indian Eastern Ghats. Tree density may be influenced by anthropogenic activities and soil properties.

Top ten tree species including Albizia amara and Chloroxylon swietenia dominated the entire stand, contributing to 90.01% of the total individuals. Tree density differences among the plots could be due to the efficiency of seed dispersal and its establishment as well as resource exploitation levels by locals as suggested by Kadavul and Parthasarathy (1999a). Nevertheless, mono-dominance of species like Albizia amara in all the study plots shows their inherent ability to thrive in varied environmental conditions and in disturbed areas.

The mean basal area of tree community was 18.71 m2/ha in dry deciduous forests of Sathanur reserve forest that is modest when compared to the dry tropical forest in Vindhyan hills (Jha and Singh 1990). The value (7.22–43.05 m2/ha) obtained in the present study is well within the range of tropical dry forests in other parts of India (range 7–23.2 m2/ha, Jha and Singh 1990; mean 29.0 m2/ha, Reddy and Ugle 2008; range 8.15–41.17 m2/ha, Sahu et al. 2008; range 8.6–26.9 m2/ha, Reddy et al. 2008a; range 30–39 m2/ha, Reddy et al. 2011; 6.86 m2/ha, Sahu et al. 2012; range 12.98–33.3 m2/ha with mean of 25.82 m2/ha, Naidu and Kumar 2016) and elsewhere (Murphy and Lugo 1986; Lieberman and Lieberman 1987; Campbell et al. 1992). However, our mean value is less than the pantropical mean of 32 m2/ha (Dawkins 1959), Amazonian average (29 m2/ha; Lewis et al. 2004), Bornean (Asia) average (37.1 m2/ha; Slik et al. 2010), African average (31.5 m2/ha; Lewis et al. 2013) and other forests of Eastern Ghats (Kadavul and Parthasarathy 1999a, b). Similarly, the mean value recorded in the present study is lower than mean basal area values reported by several others in tropical forests of Western Ghats (Singh et al. 1981; Sundarapandian and Swamy 2000). The wide variations in the basal area among the 30 one-hectare plots obtained in the present study indicate that these plots were subjected to different levels of anthropogenic pressure. Exceptionally few plots have high values of the basal area, which indicates that those plots have more mature trees whereas, in some plots that had low basal area, there were many juveniles and very few mature trees. This is probably due to greater biotic disturbances in the area as suggested by Thakur (2015).

A total of 63 families were observed in this tropical dry deciduous forest. The most speciose families are Euphorbiaceae and Poaceae, followed by Fabaceae and Rubiaceae. Interestingly, similar findings were reported by Pragasan and Parthasarathy (2010) in tropical deciduous forests of the Eastern Ghats, where Euphorbiaceae, Rubiaceae and Moraceae were the most dominant families. Borah et al. (2016) also found that Euphorbiaceae was the dominant family in tropical forests of Barak Valley, Assam. Several others also observed the similar results (Ifo et al. 2016; Naidu and Kumar 2016). Hence, it can be noted that there is a similarity in family composition of forests in tropical environments.

Diameter class frequency exhibited an L-shaped curve for the trees and the data is in line with many other reports from Eastern and Western Ghats (Sundarapandian 1997; Kadavul and Parthasarathy 1999a). This is the typical characteristic of a tropical forest. The DBH size class distribution showed a decline in the number of individuals from lower class to higher class, indicating expanding population. The stem density decreased with increase in diameter class of trees as observed in the present study, which is in agreement with other reports (Lieberman et al. 1985; Swaine et al. 1987; Campbell et al. 1992; Swamy et al. 2000; Sundarapandian and Karoor 2013). This type of distribution indicates that this forest has a good potential for regeneration. Species richness also decreased with increase in diameter class. A similar trend was exhibited by the dominant species. Greater proportion (81.9%) of stems belonged to lower diameter class (≥3.2 cm −<10 cm). This is so because of growth of coppices from illegal cutting of adult stems for firewood and domestic purposes. This is the same case with many other dry forests where lower diameter class individuals are more in number. The greater density of low diameter class individuals is primarily due to open canopy (Manokaran and La Frankie 1990).

A/F ratios indicate species distribution patterns in a community. According to Odum (1971), generally, contagious distribution is the most common pattern in nature; while random distribution is restricted to very homogeneous microclimates and regular distribution prevails where competition among the population exists. Species distribution patterns vary due to differences in microclimate, habitat heterogeneity, dispersal ability and allelopathy (Kandari et al. 2011). Understanding the distribution patterns would be useful to develop management strategies in these forests that are under pressure.

Human activities and cattle grazing in forest ecosystems have changed the diversity, structure and functions of ecosystems (Sundarapandian and Swamy 2000; Swamy et al. 2000; Sundarapandian and Karoor 2013; Sundarapandian et al. 2015). The effect of anthropogenic disturbances on forest features would be plot-specific (Htun et al. 2011). Some plots (plot nos. 1–10) in the present study are near roads, human settlements or the agricultural fields which are easily accessible to human exploitation. The tree species richness was found to be low in these plots (plot nos. 1–10 except for 1, 7 and 10) compared to other study plots while shrub and herb species richness are observed to be more in these plots. The lower number of tree species may be due to several kinds of anthropogenic perturbations. Although the study area is a reserve forest, localites frequently cut trees and collect firewood, lop branches and graze their cattle. Illegal selective cutting of Chloroxylon swietenia for fencing, agricultural tools and other domestic purposes and Albizia amara for firewood are quite frequent in this forest. This kind of selective cutting may result in coppicing of those species which could affect forest species composition and stand structure. This has resulted in more density of both species in the plots near to the road, agriculture field and settlements which enhance the tree density in these plots. Due to greater tree density in these plots, the density of the herbaceous community is comparatively low here. Study plots (plot no. 21–30) are located on both sides of the rivulet. In general, the plots near the rivulet also have lower species richness; this could also be attributed to human disturbance and edaphic factors. The study area has a rocky terrain that would alter the structure of the forests. People regularly use the rivulet for day to day activities. In addition to that, this is a source of drinking water for cattle and hence, these plots were also under high anthropogenic pressures. The present study reveals that the edaphic variations and anthropogenic disturbance alter the microclimate among the plots which could be the reason for the significant spatial variation in species richness and density among the plots even though they are located within 10 km radius.

Anthropogenic perturbation in tropical dry deciduous forests of Sathanur Reserve forests creates niche space for ruderal weeds and alien invasive species to colonize and establish. Ruderal weeds and alien invasive species were the dominant understory community in all the study plots. Herbaceous community population in the study plots were dominated by native ruderal species like Sida cordifolia, Sida cordata and Sida acuta. Generally, these native ruderal species occur abundantly in the first year of the fallow-land of agroecosystems, wastelands subjected to frequent disturbances, adjacent to roads and rivulet, and moderately shaded and open areas of forests. The greater density of these ruderal weeds implied that these study plots are still under a certain level of disturbance. Alien invasive plants, Lantana camara and Ageratum conyzoides were observed in 100% and 93% of plots respectively. In addition to that, another exotic invasive species Prosopis juliflora was also registered in 43% of plots. This successful colonization and establishment of alien invasion revealed that these study plots are either under disturbance or have canopy opening. Many studies have confirmed that the natural or anthropogenic perturbations pave way for a conducive environment for the establishment of invasive plants (Whitmore and Burslem 1996; Sundarapandian 1997; Sundarapandian and Karoor 2013). In shrub community, 32.18% of the population is contributed by exotic invasive species. Similarly, in the herbaceous community, exotic invasive species contribution is 13.33%. The present study reveals that the understory vegetation of tropical dry deciduous forest at Sathanur Reserve forest is dominated by ruderal weeds and exotics. It indicates that this forest is under the threat of anthropogenic pressure even though it has been declared as a reserve forest. However, this forest ecosystem restores rich flora similar to other tropical dry forests of the Eastern Ghats and central India. To impede the plant invasion, timely measures are to be adopted to eliminate invasive species in order to retain and conserve the native diversity.