Impact of rural community harvesting practices on plant biodiversity in Kudremukh National Park, India

Abstract

The Kudremukh National Park in the central Western Ghats (India) is a mega-biodiversity hotspot. However, the dependence on forests of tribal and non-tribal settlements in the core area of the park has resulted in forest fragmentation, posing a threat to the endemic flora. The study focuses on the disturbed forest ecosystem in the park. Using the belt transect method, we studied the vegetation structure and floristic composition of forests in tribal and non-tribal settlement areas. We compared species diversity, richness, dominance and stand quality of the park with an undisturbed nearby forest. Due to harvesting practices of the rural communities, the percentage of light-demanding species was higher in tribal and non-tribal forest compared those in with the undisturbed forest. Differences in species composition were largely due to extensive establishment of light-demanding or pioneer forest species. The pioneer forest species pose a threat to endemic species of the region by suppressing seedling growth and establishment. Our observations will help forest functionaries to prepare site-specific restoration plans. Primary forest species such as Myristica dactyloides, Palaquium ellipticum, Garcinia gummi-gutta and Poeciloneuron indicum can be used in the restoration process. These species help to maintain forest ecosystem stability and improve provision of forest ecosystem services.

Keywords:

• ecosystem services
• Western Ghats
• tribal settlement
• Shannon–Wiener diversity
• pioneer species

Introduction

The Convention on Biological Diversity (WRI et al. 1992) identified mechanisms that affect current levels of biodiversity, nearly all of which include significant human components. The extraction of timber and fuelwood by rural communities has hitherto been within the carrying capacity of surrounding forests. With the increase in human population, these activities have exceeded the carrying capacity of the forests. It is, however, inferred that harvesting of timber from a natural forest affects its structure and dynamics (Whitmore and Sayer 1992). Young forests or plantations that develop soon after harvesting generally contribute to higher biodiversity (Kimmins 1992).

The rain forests of the Western Ghats mountains are unique because of their geographical location, stable geological history, stable climate, heavy rainfall and good soil conditions that support a variety of tropical forest ecosystems (Pascal 1988). Rich in biodiversity, this area harbours numerous medicinal plants and important genetic resources for the wild relatives of grains, fruits and spices (Bawa et al. 2007). The heterogeneity of habitat and vegetation patterns that are found in this area provides ideal conditions for the occurrence of many amphibian species (Krishnamurthy et al. 2001). In recent years, mining activity, coffee and other plantations have caused large-scale encroachment of Kudremukh National Park (KNP) (Hussain 2003), exposing protected areas to human interventions, proving to be a major threat to the vast plant genetic resources of the region. The regeneration of shade-loving primary forest species in the Western Ghats region is particularly affected through ongoing human activities (Parthasarathy 1999).

The study area, Karchar, is located in the KNP of Chikkamaglur district, in the central Western Ghats of South India and is the home of the Gowdlu tribe. There is sparse distribution of families – nearly three families per 1 km². Communities residing here date back to as early as AD 4, practicing shifting cultivation, hunting and utilising natural forest resources for their livelihoods (Gadgil and Subhas Chandran 1988). In recent decades, non-tribal communities have also become dependent on these forests for sustenance and have converted part of forest area into Areca plantations. This intensification of agricultural production to meet the needs of the rapidly increasing population is the result of the dependence of non-tribal communities on forests for sustenance (Venkatesh Nayak et al. 2000).

In this study we evaluated the impact of different intensities of anthropogenic activity on the composition, diversity and stand quality of forest tree species in the KNP. We compared results from two differently managed sites (tribal and non-tribal plantation) with that of undisturbed forest.

Study area and climate

The study area lies within the KNP Western Ghats of Karnataka state (South India). It covers an area of 600.34 km² and lies between 13°1′–13° 29′ N and 75° 0′–75° 30′ E. The soils are mainly Inceptisols and Ultisols (Bourgeon 1989). The climate is typical tropical, with an annual rainfall ranging from 600 to 800 cm. The maximum temperature varies from 21°C to 34°C during April–July, while the minimum temperature ranges from 12°C to 18°C between January and May.

Material and methods

Vegetation studies

For vegetation sampling, a belt transect was used to investigate the distribution of individuals in relation to a particular area (Hill et al. 2005). Belt transects of 500 m × 10 m were laid in three sites, which differed in the way in which they were utilised and managed: tribal settlements (communities dating back to AD 4, practicing shifting agriculture, hunting and utilising forest resources for livelihoods), non-tribal settlements (communities dependent on forests for sustenance and who have converted part of the forest area into Areca plantations) and intact areas (undisturbed forest). The total sample area at each plot was 0.15 ha. Each plot was divided into 15 sub-plots measuring 10 m × 10 m, laid at 10 m distances inside the belt transect. The girth at breast height of mature trees (diameter at breast height (DBH) > 30 cm) and saplings (DBH between 10 cm and 30 cm) and collar diameter of seedlings (girth below 10 cm) were recorded.

The data on vegetation were analysed for relative frequency, density and dominance, and their sum represents the importance value index (IVI) of the species (Kershaw 1973). Species diversity was calculated using the formula of Shannon and Wiener (1963) as

and Simpson's index (Simpson 1949) as

where H and C are the Shannon index of general diversity and Simpson's index of dominance; n_i  = number of individuals of each species; and N = total number of individuals of all species.

Determination of stand quality

Using a literature survey (Gamble 1928; Chandrashekaran 1960; Rai 1979; Rai and Proctor 1986; Chandrashekara and Ramakrishnan 1994), we categorised species into three groups. Each group was assigned a numerical value, viz., Pioneer index 1 (primary species) for the group requiring a small gap for regeneration; Pioneer index 2 (early secondary) for the species whose seedlings establish in small gaps but need small- to medium-sized gaps to grow; and Pioneer index 3 (late secondary) for the group of strong light demanders. We adopted the procedures from Whitmore (1989) and Chandrashekara (1998) to indicate intensity of disturbance.

where, RISQ is the Ramakrishna index of stand quality; n_i and N are IVI of the species and community, respectively. The RISQ can vary from 1.0 to 3.0, where 3.0 indicates high disturbance.

Results

Characteristics of the vegetation

Our study recorded 63 tree species belonging to 53 genera and 28 families from 45 sub-plots in 0.45 ha. In the tribal plot, we found 33 species, with Lophopetalum wightianum (85.7 IVI) as the dominant mature tree, followed by Dimocarpus longon (24.6) and Persea macarantha (16.8). In the non-tribal plot, we recorded 29 species, with dominance of Aglaia anamallayana (47.6) followed by D. longon (42.3) and Mimusops elengi (27.2). Poeciloneuron indicum was abundant in the undisturbed forest (106.6), followed by Cleistanthus travancorensis (20.1) and Syzygium rubicundum (18.6), with a total of 34 tree species. The shade-loving primary tree species such as Myristica dactyloides, Hopea canarensis, Calophyllus polyanthus and Agrostachis indica were recorded only in undisturbed forest plots. However, light-demanding tree species such as Mallotus philippensis, Macaranga peltata and Callicarpa tomentosa were only recorded in disturbed plots (Table 1).

Table 1. Pioneer index (PI) and importance value index (IVI) of tree species in three different areas of Kudremudh National Park, South India

		Importance value index (IVI)
Species	PI	Tribal plot	Non-tribal	Undisturbed
Actinodaphne bourdillonii Gamble	1	2.156	NF	4.41
Agrostachis indica Dalz	1	NF	NF	5.21
Aglaia anamallayana (Bedd.) Kosterm.	1	9.23	47.68	NF
Antidesma menasu Tul.	1	2.345	2.90	3.75
Apodytes dimidiata Meyer ex Arn	1	3.041	3.66	6.52
Aporosa lindleyana (Wt.) Baill.	2	5.678	2.79	NF
Artocarpus heterophyllus Lamk.	2	NF	NF	3.01
Artocarpus hirsutus Lamk.	2	5.246	21.49	NF
Calophyllus polyanthus Wall. ex Choisy.	1	NF	NF	13.94
Callicarpa tomentosa (L.) Murray	3	2.084	NF	NF
Carallia brachiata (Lour.) Merr.	2	NF	4.46	2.59
Celastrus sp.	1	NF	2.22	NF
Chassalia ophioxyloides (Wall.) Craib	1	NF	NF	1.18
Chionanthus malabarica (Wall. ex G. Don) Bedd.	1	4.196	NF	NF
Cinnamomum malabathrum (Burm. F.) Blume	1	10.162	NF	1.25
Cleistanthus travancorensis Tablonszky	2	NF	NF	20.14
Dimocarpus longon Lour.	2	24.623	42.35	NF
Diospyros sylvatica Roxb.	1	NF	4.70	10.15
Elaeocarpus serratus L.	1	2.38	NF	2.88
Elaeocarpus tuberculatus Roxb.	1	16.513	8.01	14.9
Euodia lunu-ankenda (Gaertn.) Merr.	2	NF	NF	1.26
Ervatamia heyneana (Wall.) Cooler	3	4.168	2.51	NF
Euonymus indicum Heyne ex Wall.	2	NF	NF	1.17
Ficus sp.	2	NF	18.21	NF
Fahrenheita zeylanica Heyne ex Wall.	1	NF	NF	1.35
Flacourtia montana Grah.	2	NF	6.06	NF
Garcinia gummi-gutta (L.) Robs.	1	7.49	NF	NF
Garcinia morella Desr	1	NF	2.10	NF
Glochidion zeylanicum (Gaertn) A. Juss.	2	5.734	NF	NF
Holigarna arnottiana Hk f.	1	10.818	18.92	NF
Holigarna grahamii Kurz.	1	11.543	19.68	4.82
Holigarna nigra Bourd.	1	4.221	NF	NF
Hopea canarensis Hole	1	NF	NF	16.7
Knema attenuata (H.K. & Thomas) Warb	1	NF	10.92	4.39
Kydia calycina Roxb	3	4.58	NF	NF
Ligustrum gamblei Ramam.	2	2.28	NF	NF
Litsea floribunda (Bl.) Gamble	2	8.81	5.82	8.22
Litsea ghatica Saldhana	2	6.26	NF	NF
Litsea laevigata (Nees) Kosterm.	2	NF	2.16	NF
Lophopetalum wightianum Arn.	1	85.79	7.8	NF
Macaranga peltata (Roxb.) Muell-Arg	3	4.25	5.07	NF
Magaratia indica (Dalz.) Airy Shaw	1	NF	5.59	1.48
Mastixia arborea (Wt.) Bedd	1	NF	5.07	4.04
Maytenus rothiana (Walp.) Lobreau-Cal.	1	2.12	NF	NF
Mallotus philippensis (Lamk.) Muell.-Arg.	3	5.84	2.40	NF
Meyna laxiflora Robyns	3	2.38	NF	NF
Mimusops elengi L.	2	NF	27.23	NF
Myristica dactyloides Gaertn	1	NF	NF	10.36
Neolitsea scrobiculata (Meisn.) Gamble.	2	4.11	NF	NF
Nothopegia beddomei Gamble.	2	3.07	2.29	2.38
Olea dioica Roxb.	2	NF	2.72	NF
Palaquium ellipticum (Dalz.) Baillon	1	NF	NF	8.72
Persea macarantha (Nees) Kosterm.	2	16.82	11.48	3.27
Poeciloneuron indicum Bedd	1	NF	NF	106.63
Scheflera rostrata (Wt.) Harms	1	NF	NF	1.34
Symplocus racemosa Roxb.	1	NF	NF	2.79
Symplocus cochinchinensis Roxb.	1	NF	NF	1.17
Syzygium cumini (L.) Skeels	1	8.70	8.88	NF
Syzygium lanceolatum (Lam.) Wt. & Arn.	2	NF	NF	4.42
Syzygium rubicundum Wt. & Arn.	1	NF	NF	18.64
Terminalia bellirica (Gaertn.) Roxb.	3	5.19	NF	NF
Toona ciliata Roemer	3	4.64	NF	4.93
Trichilia canaroides (Wt. & Arn.) Benth.	2	NF	NF	1.39
Note: NF, not found in that forest plot.

Overall tree species richness data indicate that higher numbers of species or all size classes were found in undisturbed forest, whereas tree density of the regenerative class (seedlings) was higher in disturbed forest plots, indicating the presence of more light-demanding species (Table 2). Moreover, the diversity index value for mature trees (DBH > 30 cm) was higher in tribal and non-tribal forest (Table 2). Higher number and diversity of seedlings was recorded in the undisturbed forest. The Simpson's index was higher in the undisturbed forest; tribal and non-tribal sites had a lower index (Table 2).

Table 2. Vegetation characteristics of tree species of three areas of Kudremukh National Park, South India

	Experimental plots
Characteristics	Tribal	Non-tribal	Undisturbed
Number of species
Mature trees	33	29	34
Saplings	27	28	28
Seedlings	17	15	29
Density per hectare
Mature trees	641	604	440
Saplings	145	112	409
Seedlings	4566	4427	457
Shannon–Wiener diversity index
Mature trees	4.17	4.07	3.74
Saplings	4.58	4.08	4.13
Seedlings	3.06	3.53	3.80
Average	3.94	3.89	3.89
Dominance index
Mature trees	0.10	0.07	0.17
Saplings	0.06	0.09	0.14
Seedlings	0.16	0.12	0.15
Average	0.11	0.09	0.15
Ramakrishna index of stand quality (RISQ)
Mature trees	1.52	1.55	1.16
Saplings	1.91	2.22	1.19
Seedlings	1.89	1.41	1.17
Average	1.77	1.72	1.17

Stem density of trees at sapling stage was low in tribal and non-tribal plots, while seedling density remained high. The RISQ value of the disturbance index (Table 2) for mature trees, saplings and seedlings ranged from 1.16 to 1.19, with a low value in the undisturbed forest. The RISQ values in tribal and non-tribal forest ranged from 1.406 to 2.221, with the highest value for the sapling stage in non-tribal forest.

Stand quality

The percentage of primary species (shade species) such as M. dactyloides, Palaquium ellipticum, P. indicum and so on among tree, sapling and seedling stages was higher in the undisturbed forest (Table 3), while the percentage of light-demanding tree species such as M. peltata, Terminalia bellirica, M. philippensis and so on was higher in the tribal and non-tribal forest. In the undisturbed forest, primary tree species accounted for 81.86% and occupied a basal area of 86.83%, while such values did not exist in the tribal forest where primary tree species accounted for 49.5% and occupied a basal area of 84.63%.

Table 3. Contribution of primary, late secondary and early secondary trees to total basal area and number in Kudremukh National Park, South India

	Contribution to number of trees			Contribution to basal area
Parameters	Undisturbed	Tribal	Non-tribal	Undisturbed	Tribal	Non-tribal
Mature trees
Primary	81.86	49.5	52.45	86.83	84.63	39.78
Late secondary	16.17	36.27	42.64	13.16	12.05	42.68
Early secondary	Not found	14.21	4.9	Not found	3.3	17.53
Saplings
Primary	76.47	33.17	43.13	72.19	36.17	55.52
Late secondary	18.62	50	55.88	25.86	49.55	43.44
Early secondary	2.94	16.82	0.98	1.93	14.28	1
Seedlings
Primary	80.75	15.68	31.37	70.55	16.83	27.27
Late secondary	18.07	81.37	66.49	21.17	74.25	64.7
Early secondary	1.12	2.94	2.13	0.26	8.91	8.02
Note: Primary species – require a small gap for regeneration; early secondary species – seedlings can become established in small gaps but need small- to medium-sized gaps to grow; late secondary species – strong light demanders.

Discussion

It has been shown that human activities influence the quality of stands, microclimate, nutrient cycling and composition of forest species (Kappelle et al. 1995). One of the characteristic features of the humid tropical forest ecosystem is its high species richness (Parson and Cameron 1974). The number of species recorded here is comparable to that found in Swamy and Proctor (1994) for species richness in the forest surrounding the KNP. The stem density for the mature trees in the undisturbed forest appears to be quite low (440 trees ha^–1) compared with the study of Ghate et al. (1998), which recorded a mean density of 419 trees ha^–1 for the Western Ghats in a closed-canopy evergreen forest. However, the range of stem density recorded in the tribal and non-tribal plots for sapling and mature tree phases is closer to the observations of Pascal (1988), Chandrashekara and Ramakrishnan (1994), varying from 604 to 4566 trees ha⁻¹, which is also more than in undisturbed plots. This is due to the lower canopy cover in the tribal and non-tribal plots, which causes higher light transmission, stimulating germination and rapid growth of pioneer species (Wales 1972; Levenson 1981; Ranney et al. 1981; Lopez de casenave et al. 1995; Jose et al. 1996).

More variation in vegetation composition is observed between the undisturbed, non-tribal and tribal plots. Only 15.8% of species are similar between the undisturbed and non-tribal forest plots, where tribal and non-tribal forest plots had 40.8% species similarity (Figure 1). The species similarity is less than 50% in the human-dominated plots, which implies that anthropogenic pressures had a serious impact on primary forest species, viz. C. travancorensis, Euonymus indicum, H. canarensis, P. ellipticum, P. indicum and Scheflera rostrata in the national park. These species were only found in undisturbed plots (Table 1). The density of saplings (20–30 cm DBH) was comparatively lower in the tribal and non-tribal inhabited areas than in the undisturbed forest due to the utilisation of saplings by inhabitants in these areas for fencing around settlements (protection from wildlife), roofing, preparing agricultural implements, firewood, green foliage (for Areca plantations and other cash crops) and so on.

Figure 1. Tree species similarity among sites in Kudremukh National Park, South India.

Although early- and late-secondary seedlings are regenerating in the undisturbed forest, these seedlings do not prevent native species regeneration because they do not establish into mature trees (Primack et al. 1985; Silva et al. 1995). There were more light-demanding species in the tribal forest. This could be attributed to vegetation changes following extensive canopy disturbance in the past and the recent past. A similar result was recorded in Kibale National Park, Uganda (Chapman and Chapman 1997), where creation of larger gaps facilitated colonisation by the light-demanding species.

The mature tree diversity index was higher in tribal and non-tribal forests than in the undisturbed forest. The presence of light-demanding species contributed to the increased diversity in the tribal and non-tribal forests. However, the values were lower than those recorded in tropical rain forest of Barro Colorado Island (Knight 1975) and the Silent Valley (Singh et al. 1981), India; but higher than those in Kerekatte forest near Sringeri, near the park (Swamy and Proctor 1994). Significantly higher diversity values were observed for the seedling phase in the undisturbed forest, indicating an increase in regeneration. The diversity of the seedling stage was comparatively lower in tribal and non-tribal forest, indicating low regeneration. This could be attributed to activities such as leaf litter collection, livestock grazing and so on which disturb the regeneration of primary forest tree species. Values of the dominance index of seedlings in tribal plots were higher than in non-tribal forests due to the establishment of light-demanding species through disturbance. In undisturbed forest, light-demanding species were comparatively less abundant in all stages.

The two experimental plots had a higher disturbance index, as they are located within a gap of 1–2 km around human settlements. Anthropogenic intervention has affected the regeneration strength of primary species (shade-loving, light-sensitive species) in the human-affected areas. A higher RISQ value for saplings and seedlings than mature trees indicates repeated disturbance or failure of shade-tolerant species to establish (Chandrashekara and Sankar 1998). Due to the canopy openness and more light availability the late-secondary saplings and seedlings are more common in tribal and non-tribal plots, resulting in a loss of the regeneration capacity of primary forest species. Hence it is recommended to plant primary forest species such as Garcinia gummi-gutta, Calophyllum apetalum, M. dactyloides and so on in degraded forest fringes close to tribal and non-tribal settlements. It is also important to undertake awareness programmes among local people on the importance of conserving this biodiversity hotspot and restoring forest fringe areas in the protected areas. The Forest and Wildlife Departments, along with other government departments, should implement programmes such as providing energy- efficient cooking devices, sustainable harvesting of forest products and so on to reduce pressure on the protected area, which may have a direct ecological impact on the forest ecosystem in the future.

Conclusion

This study highlights the distribution of plant resources in the dynamic landscapes of the Western Ghats region. The forest ecosystem in both settlement areas supported higher plant diversity than the undisturbed forest, with less shade-loving primary forest species. This clearly indicates that human settlements close to rain forests directly impact regeneration of primary forest species, resulting in changes in the composition of the forest. Our observations will help forest functionaries to prepare site-specific restoration plans. Primary forest species, such as M. dactyloides, P. ellipticum, G. gummi-gutta and P. indicum, could be used in the restoration process. These species help to maintain forest ecosystem stability as well as improving the provision of forest ecosystem services.

Acknowledgements

The authors thank the Karnataka Forest Department for granting permission to carry out the work in Kudremukh National Park and their invaluable help in site selection. The valuable comments and suggestions from anonymous reviewers are gratefully acknowledged.