B. R. Ramesh, M. H. Swaminath, Santoshgouda V. Patil, Dasappa, Raphaël Pélissier, P. Dilip Venugopal, S. Aravajy, Claire Elouard, and S. Ramalingam. 2010. Forest stand structure and composition in 96 sites along environmental gradients in the central Western Ghats of India. Ecology 91:3118.

Abstract: This data set reports woody plant species abundances in a network of 96 sampling sites spread across 22000 km² in central Western Ghats region, Karnataka, India (74°15' – 75°40' E; 15°15' – 13°30' N). Due to its varied climate and diverse topography, the study area, which is part of the Western Ghats–Sri Lanka biodiversity hotspot, supports a wide array of non-equatorial tropical habitats including wet evergreen, moist and dry deciduous intact as well as degraded forests and scrublands. These formations, floristically moderately rich and diversified, are characterized by a lower rate of endemism than in the southern part of the Western Ghats. This data paper provides abundance and girth data for 76813 trees and lianas of 446 species collected in 96 sampling sites during 1996–1997. A total of 61965 individuals >10 cm girth at breast height (gbh) were recorded in 96 1-ha macroplots, while 14848 individuals <10 cm gbh but >1 m height were sampled in three 0.1-ha microplots located within each macroplot. Additional data regarding the stand structure (average canopy height, percent canopy cover, number of strata) and the level of degradation are available for the macroplots, along with environmental data derived from other sources and analyses, such as soil types, rainfall, length of the dry season, and altitude. These data have been used to produce ecological research papers as well as to elaborate conservation value maps and recommendations toward sustainable management of the forests of the central Western Ghats region.

Key words: biodiversity sampling plots; central Western Ghats; India; Karnataka; mesoscale plot network; plant species abundances; tropical forest types.

INTRODUCTION

Patterns of floristic diversity at the mesoscale (Holt 1993) deserve particular attention as they are of direct relevance to operational management and conservation of natural resources. Particularly in tropical regions where the natural vegetation persists only as fragments, data on forest stand structure and composition gathered across large scale environmental and/or degradation gradients, aid in defining and mapping forest types (Condit 1996). Such data also enables understanding forest succession and dynamics (Sheil et al. 2000) as well as testing hypotheses about the ecological processes through which species diversity is maintained (e.g., Couteron et al. 2003, Munoz et al. 2007). However, if vast areas of the humid tropics have been sampled by forest services, large-scale reliable floristic data are generally scarce and limit ecological studies.

This data paper documents the abundance of woody plant species in a network of 96 sampling sites that covers an area of about 22000 km² in the central Western Ghats region, Karnataka, India, which belongs to the Western Ghats–Sri Lanka biodiversity hotspot (Myers et al. 2000). The sampling plots, laid out in 1996–1997 in the framework of a collaborative project between the Karnataka Forest Department and the French Institute of Pondicherry, are 1-ha in size and sample a wide array of habitats including wet evergreen, moist and dry deciduous, intact as well as degraded forests and scrublands. A total of 61965 individuals of 400 trees and lianas species with girth at breast height (gbh) equal or above 10 cm were recorded in the 1-ha plots (macroplots), while 14848 individuals of 334 species less than 10 cm dbh but more than 1 m height were recorded within 3 0.1-ha subsamples (microplots) in each macroplot. Additional data on stand level structural parameters (mean canopy height, canopy cover, etc.), disturbance level as well as bioclimatic variables (rainfall, dry season length, etc.) are also provided.

This data set has already been used to publish ecological research papers (Munoz et al.2007, 2008, Ramesh et al.2010b) as well as to derive conservation value maps and recommendations towards sustainable management of the forests of the central Western Ghats region (Ramesh and Swaminath 1999, Ramesh et al. 1999, 2009a,b).

METADATA

CLASS I. DATA DESCRIPTORS

A. Data Set Identity. Karnataka Biodiversity Sampling Plots, Western Ghats, India

B. Data Identification Code. IFP_ECODATA_KBSP_v4

C. Data Set Descriptors.

1. Originators.

B. R. Ramesh ([email protected]) and S. V. Patil, Ecology Department, French Institute of Pondicherry, 11 St. Louis Street, Puducherry 605001, India;

M. H. Swamminath and Desappa, Forest Research Institute, Karnataka Forest Department, Aranya Bhavan, 18th Cross, Malleswaram, Bangalore 560003, India.

2. Abstract. This data set reports woody plant species abundances in a network of 96 sampling sites spread across 22000 km² in central Western Ghats region, Karnataka, India (74°15' to 75°40' E; 15°15' to 13°30' N). Due to its varied climate and diverse topography, the study area, which is part of the Western Ghats–Sri Lanka biodiversity hotspot, supports a wide array of non-equatorial tropical habitats including wet evergreen, moist and dry deciduous intact as well as degraded forests and scrublands. These formations, floristically moderately rich and diversified, are characterized by a lower rate of endemism than in the southern part of the Western Ghats. This data paper provides abundance and girth data for 76813 trees and lianas of 446 species collected in 96 sampling sites in 1996–1997. A total of 61965 individuals >10 cm girth at breast height (gbh) were recorded in 96 1-ha macroplots, while 14848 individuals <10 cm gbh but >1 m height were sampled in 3 0.1-ha microplots located within each macroplot. Additional data regarding the stand structure (average canopy height, percentage of canopy cover, number of strata) and the level of degradation are available for the macroplots, along with environmental data derived from other sources and analyses, such as soil types, rainfall, length of the dry season, and altitude. These data have been used to produce ecological research papers as well as to elaborate conservation value maps and recommendations towards sustainable management of the forests of the central Western Ghats region.

D. Key Words: biodiversity sampling plots; central Western Ghats; India; Karnataka; mesoscale plot network; plant species abundances; tropical forest types.

CLASS II. RESEARCH ORIGIN DESCRIPTORS

A. Overall Project description.

1. Identity. Assessment and conservation of forest biodiversity in the Western Ghats of Karnataka, India.

2. Originators.

B. R. Ramesh ([email protected]), Ecology Department, French Institute of Pondicherry, 11 St. Louis Street, Puducherry 605001, India;

M. H. Swamminath, Forest Research Institute, Karnataka Forest Department, Aranya Bhavan, 18th Cross, Malleswaram, Bangalore 560003, India.

3. Period of Study. 1996–1999.

4. Objectives. This project responded to a need for an integrated, comprehensive resource management approach, through the development of new policies to protect and conserve the ecologically and economically valuable forests of the central Western Ghats region of Karnataka, India. Based on regional level vegetation mapping, landscape scale analyses of forest status and fine scale studies on impacts of disturbance on the forest structure and composition, the project has elaborated conservation value maps and recommendations towards the sustainable management of forests.

5. Abstract. Within the framework of the Western Ghats Forestry Project (http://www.odi.org.uk/projects/98-99-tropical-forestry/projects/3275.htm), the Forest Research Institute of the Karnataka Forest Department (KFD), in the mid 1990s, has established a network of 102 1-ha Biodiversity Sampling Plots covering a wide range of vegetation in the central Western Ghats region of Karnataka State, India. The original project, among other objectives, aimed at delineating zones for management considerations based on forest status and distribution, along with details on anthropogenic pressures. Between 1996–1999, under the aegis of the project "Assessment and Conservation of Forest Biodiversity in the Western Ghats of Karnataka, India" funded by the Fonds Français pour l'Environnement Mondial (FFEM; http://whc.unesco.org/en/ffem), the botanists of the French Institute of Pondicherry (FIP) collaborated with KFD to review the taxonomic identification of the plant species in 96 of the 100 × 100 m macroplots established by KFD (six plots located in plantations were not considered). They also enumerated all woody plants below 10 cm gbh but above 1 m height in three microplots of 10 × 10 m located within each macroplot. The FFEM project comprised of three separate components: The FFEM project comprised of three separate components: (1) an update of the forest maps of South India in order to assess the forest cover changes between 1977 and 1997, which revealed a loss of ca. 12% of the forest area in 20 years, concentrated in zones already degraded and fragmented (i.e., outside the "reserve forests"), and attributed primarily to the extension of coffee plantations; (2) an assessment of tree biodiversity from the network of plots reported here, which showed that a strong interaction between bioclimatic and anthropogenic (disturbance regime) factors predicted the major forest types in the region; and (3) an in-depth study of the impact of selective logging on the wet evergreen forest of the Kadamakal Reserve Forest, which revealed that a minimum rotation of 30–40 years between two successive harvests was necessary to allow the forest to recover in terms of biomass, structure and floristic composition. Results from these studies have yielded conservation value maps and priorities for forest conservation (Ramesh and Swaminath 1999, Ramesh et al. 2009a,b) as well as updated 1:250,000 forest maps of South India (Pascal et al. 1982–1997a,b,c) and a few scientific papers (Pascal et al. 1998, Pélissier et al. 1998, Guero and Ramesh 1999, Ramesh et al. 1999).

B. Brief subproject description.

1. Site description.

• Site type. Network of forest sampling plots covering various non-equatorial tropical forest types in various degradation stages.

• Geography. The study area is located between 74°15' E to 75°40 E' and 15°15' to 13°30' N coordinates, falling within the administrative boundaries of Uttara Kannada, Shimoga, and Chikmagalur districts of Karnataka State in south-west India. The total area covered is estimated to be ca. 21970 km² of the central Western Ghats (WG) region, and extends from the coastal plain of the Arabian Sea to the humid hill zone (locally called Malnad) at the foot of the Western Ghats 'great escarpment', and to the Karnataka plateau (locally called Maidan), which overlooks the coastal lowlands and recedes towards the eastern upland region. Sampling plots were located up to 80 km and 125 km inland, in the northern and southern portions of the study area, respectively (Fig. 1).

FIG. 1. Location map of the 96 Biodiversity Sampling Plots in central Western Ghats Region, Karnataka, India (KBSP network). Solid isolines correspond to mean annual rainfall; dashed isolines to mean number of dry months in a year, a month being considered as dry according to Pascal (1982), i.e., when rainfall in mm is less than two times the mean temperature in °C.

• Habitat types. The plots encompass a wide array of habitats from hyper-wet evergreen forests (rainfall ≥ 8000 mm/yr) to dry deciduous forests (rainfall < 1000 mm/yr). Soils are predominantly lateritic from the coastline, up the escarpment and across the Karnataka plateau. The humid hill zone of the Malnad consists of two main categories of lateritic soil, namely 'reworked' lateritic soils and impoverished ferrallitic soils (Bourgeon 1989).

• Geology. In the southern part of the WG, bedrock is composed of Archean rocks from the Precambrian shield, with a prevalence in the study area of very old metamorphic material enriched in metals (iron, manganese, copper, plumb, gold), which constitutes the Dharwar system (Pascal 1984, 1988).

• Climate. The WG relief barrier orographically exacerbates the summer monsoon rains and is also responsible for a steep environmental cline from the plateau edge into the interior upland. It follows that the combination of the onset of southwest monsoon with the orographic effects of the escarpment of the Ghats determines strong regional bioclimatic gradients (Pascal 1982, Gunnell 1997). The windward side of the WG receives heavy rains as the monsoon progresses from west to east. Between the west coast and over the escarpment, a parallel stretch of 60–100 km, the rainfall varies between 2000 to more than 5000 mm/yr. To the interior region a rapid diminishing of rainfall to less than 1000 mm/yr/ is observed within a distance of 10–50 km. Showers prior to and following the monsoon, augment the total rainfall received at the transitional zone. Between the coastline and the crest of the Ghats, at elevations above 800 m, mean coldest month temperature is 23°C, while in the hilly terrains at medium elevations (800–1400 m) it varies between 16 and 23°C. At last, the differential timings in arrival and withdrawal of the monsoon rains results in a increase from 4 to 7 months in the length of the dry season from south to north of the study region. In the west-east direction, the length of the dry season also rapidly increases from 4 to 7 months, due to a sharp decrease in rainfall beyond the crest of the Ghats (see Fig. 1).

2. Sampling design.

• Design characteristics. The initial sampling scheme was designed by KFD based on the perception of local level officers. However, when FIP was involved in the project, the selection of sites was rationalized with respect to scientific knowledge on regional vegetation and habitat types, as depicted on FIP forest maps, sheets # 1-Belgaum-Dharwar-Panaji (Pascal et al. 1982–1997a), # 2-Shimoga (Pascal et al. 1982–1997b), and # 3-Mercara-Mysore (Pascal et al. 1982–1997c), and their explanatory booklet (Pascal 1986). In Uttara Kannada district, the plots were laid in forests close to villages in order to enable the monitoring by Village Forest Committees. However, in Shimoga and Chikmagalur districts, the plots are more widespread. Finally, although the sampling scheme was stratified to cover all forest types in the region and tentatively their various degradation stages, it was not however rigorously balanced in terms of number of plots per strata.

• Permanent plots. The sampling design encompasses 96 sampling sites, since six sites of the initial KBSP network located in plantations were not considered in this subproject. In each site a 1-ha macroplot (100 × 100 m) was delineated and marked with stone posts every 20 m along the boundary. In each macroplot, three 0.1-ha microplots (10 × 10 m) were established across the diagonal axis in order to study the regeneration (Fig. 2).

FIG. 2. Layout of 1-ha macroplots with three 0.1-ha microplots along a diagonal (in gray).

• Data collection. Within each macroplot, all live individual trees and lianas that were equal and above 10 cm girth at breast height (gbh) were tagged with serial numbers, botanically identified, and measured. For multi-stem trees, all the stems above 10 cm gbh were measured and were retained as multi-stem girth data for single individual tree. Other macroplot level details recorded included the visual estimation of canopy height (m), canopy cover (%), number of strata, and degradation stage. In the microplots, all the individual woody plants less than 10 cm gbh and more than 1 m height were enumerated and identified. Based on literature (Pascal 1986, 1988) and botanical expertise of the first author (BRR), all recorded species in the macro- and microplots were later classified into different habits (trees with various potential heights, shrub and liana), as evergreen or deciduous in nature, and as endemic of the Western Ghats or not (based on Ramesh et al. 1997).

3. Research methods

• Field/Laboratory method. KFD staff laid out the plots. They delineated boundaries of macroplots, tagged the trees and lianas with numbered metal labels and recorded tree gbh. The lists were passed to FIP staff that precisely relocated each site from Survey of India 1:50,000 scale toposheets (GPS equipment was not available at FIP at the time) and conducted the botanical identifications (see below) and stand level characterization (visual assessment of average canopy height, percentage of canopy cover, and number of strata). The level of degradation was assessed using a double codification. The first one relates to the stand physiognomy visually assessed in the field according to three levels of structural integrity with respect to potential climax: 1-slightly disturbed (dense forest), 2-moderately disturbed (woodland to savanna woodland according to potential forest type), 3-highly disturbed (tree savanna to scrub woodland and tickets according to potential forest type). The second codification is based on the percentage of individuals in a plot that belonged to deciduous species, which is an indicator of floristic integrity (see Pascal 1984, 1986 and Ramesh et al. 2010b) for potential wet evergreen forests (i.e., for all forest types except ATT and LTD; see IV.B). It is coded in three classes: 1-less than 33%, 2-between 33 and 66%, and 3-more than 66% of deciduous individuals. However, primary moist (LTD) and dry deciduous (ATT) forest types always harbour more than 66% of deciduous individuals, so that the level of degradation of potential deciduous forest types is only assessed from structural integrity. It follows that undisturbed wet evergreen forest plots hypothesized to be close to a typical climax stage are coded as structurally slightly disturbed (structural integrity=1) with less than 33% of deciduous individuals (floristic integrity=1), all other combinations referring to more or less degraded stages, either floristically, structurally or both. Undisturbed deciduous forests are coded as structurally slightly disturbed (structural integrity=1) and are not coded regarding floristic integrity (Table 1).

TABLE 1. Codification of degradation stages of potential evergreen and deciduous forest types according to structural and floristic integrity.

Potential climax type	Floristic integrity	Structural integrity
		1	2	3
Evergreen	1	close to intact
	2	degraded
	3
Deciduous	NA	close to intact	degraded

FIP staff also installed three 0.1-ha microplots in each macroplots (see Fig. 2), in which they enumerated and identified all woody plants with gbh < 10 cm and height > 1 m. Recorded rainfall data extracted from the Annual Rainfall Reports (Directorate of Economics & Statistics, Government of Karnataka, Bangalore), for a total of 336 meteorological stations located in the neighborhood of the plots and followed over a period of at least 10 years since 1940, was used for the interpolation of the total rainfall values for the field plots. Interpolation was performed through ordinary kriging method available within the spatial analyst extension in ArcView 3.1 (ESRI, 1992–2000). A semi-variogram with 1.3 m interval lag was computed for 5 different models: circular, spherical, gaussian, exponential, and linear. Based on semi-variograms, circular model (R² = 0.996) was selected for the actual interpolation. Interpolated rainfall values were derived for each sampling site from the 12 closest stations within a 12 km radius. These data proved to have a much better predictive power regarding vegetation than the 1-km resolution WorldClim data (Ramesh et al. 2010b). Dry season length and soil types were derived from digitized, geo-referenced versions of Pascal (1982) bioclimatic maps and Bourgeon (1989) soil map.

• Taxonomy and Systematics. Most the botanical identifications were made in the field by FIP botanists and with the help of Pascal and Ramesh (1987) botanical key based on vegetative characters. In doubtful cases, specimens were collected and identified at FIP Herbarium (HIFP), which is the taxonomic reference for this data set. Only interesting specimens with either flowers or fruits collected from sampling plots or surrounding areas were kept to enrich FIP's herbarium and assigned with a code that refers to the sampling plot. A large majority of plants were identified to species level, including three morphospecies (Ipomea sp1, Strobilanthes sp1, and Vitaceae sp1). Fourteen individual trees were identified to genus level only (Cissus sp. and Syzigium sp.), while 320 small-sized palms were included within a unique Calamus sp. genus, though 4 to 5 different species are known in the region. The identity of only 59 individuals out 61965 in the KFD lists could not be determined by FIP staff. These individuals, which are categorized as unidentified, are all lianas, for which either we were not able to collect samples or we collected only vegetative stages. Samples with doubtful identification are kept with individual's number for further verification.

• Permit history. Permission to carry out the fieldwork during the project (1995-1998) was delivered by the Government of India through a Memorendum of Understanding between FIP and KFD.

• Legal requirements. Any new visit to the sampling sites is subject to authorisation delivered to FIP staff by KFD.

• Project personnel. The macroplots were laid out by KFD staff under the supervision of M. H. Swaminath (Principal Investigator and Director of the KFD Forest Research Institute) and Desappa (associated investigator and KFD Officer). FIP Ecology Department staff laid out the microplots, conducted stand level characterizations and all botanical identifications: Dr. B. R. Ramesh (Principal Investigator and expert botanist); Dr. S. V. Patil and Dr. C. Elouard (associated investigators and botanists); S. Aravajy and S. Ramalingam (technical assistance in fieldwork and botanical identifications). Staff of the FIP Ecology Department is in charge of the herbarium specimens and database management.

CLASS III. DATA SET STATUS AND ACCESSIBILITY

A. Status.

1. Latest Update. Data collection was completed in 1997 and stored into a dedicated database since 1999. Rainfall originally derived from Pascal (1982) bioclimatic map, have been reworked in 2007 from the meteorological records available, leading to the definition of new bioclimatic classes (Venugopal 2008, Ramesh et al. 2010b; see III.B.3).

2. Latest Archive date. 2010.

3. Metadata status. Metadata are not complete regarding details of the field method and instrumentation used by KFD staff to establish the plots. It is complete for the part of the work conducted by FIP staff (botanical identification, stand level characterization, installation of microplots, database implementation and data storage and derivation of environmental variables from external sources).

4. Data verification. The database has been screened for missing, erroneous and out of bounds entries, which have been checked with the original field data sheets. In addition, occasional cross verifications were performed in the field during later visits to selected sites, by comparing random new observations with plot's listings.

B. Accessibility.

1. Storage location and medium. FIP ecological data archives (http://www.ifpindia.org/). Digital versions and copies of the database are stored in several locations at FIP.

2. Contact person. Dr. B. R. Ramesh, French Institute of Pondicherry, 11 St. Louis Street, 605001 Pondicherry, India, tel. +91 413 2334 168, fax +91 413 2339 534. Email: [email protected].

3. Copyright restrictions. None.

4. Proprietary restrictions. None, the data are free to use for further analyses, with due citation to this Data Paper.

CLASS IV. DATA STRUCTURAL DESCRIPTORS

A. Data set file.

1. Identity. Data set is downloadable as a single archive, KPSP_v4_Archive.zip (148 Ko) which contains the following data files (with a corrected version of Macroplot_data.txt available separately, see list below):

2. Size. The number of rows and columns includes headers; size is given for uncompressed files.

3. Format type. The data files are in ASCII text, tab delimited.

4. Header information. Headers corresponding to variable names (see IV.B.) are included as first row in the data files.

5. Alphanumeric attributes. Mixed.

6. Special characters.

7. Authentication procedures.

B. Variable definitions. The variables are listed in the order they appear in each data file. Variable names are headers included as first row in the data files.

CLASS V. SUPPLEMENTARY DESCRIPTORS

A. Data acquisition.

1. Data forms. Initial census data (tree numbers and gbh measurements) for the macroplots were taken by KFD staff and passed to FIP staff as paper lists, one for each macroplot. Macro- and microplots species identifications as well as plot level details were taken on specific empty data sheets prepared in advance. An example of such data sheet is given in Ramesh and Swaminath (1999; see also Ramesh et al. 2009a).

2. Data entry/verification procedures. Data were computerized by FIP technicians who participated in the field measurements, one person reading the field sheet to another one entering the data in the computer while repeating what he heard (see also V.A.1 above). Further cross-verification procedures of the database have been implemented once the data entry completed.

B. Archiving. Original data sheets are stored at FIP Botany Lab. One master copy of the complete database and an extraction of the present data set are stored on the FIP server with several copies on computers and CDs at FIP.

C. Publications and results. A few related studies have been carried out by FIP staff and collaborators from this data set or parts thereof: Ramesh and Swaminath (1999), Ramesh et al. (1999), Belna (2006), Munoz et al. (2007, 2008), Venugopal (2008), Ramesh et al. (2009a,b, 2010b).

D. History of data set usage. Part of this data set was requested in 2009 by I. Parmentier, post-doctoral fellow at Laboratoire d'éco-éthologie évolutive, Université Libre de Bruxelles, CP160/12, Av. F. D. Roosevelt 50, 1050 Brussels, Belgium.

VI. LITERATURE CITED

Belna, K. 2006. Analysis of the floristic variation across the 1-ha plots laid out in the Western Ghats of Karnataka for the FFEM project. French Institute, Pondicherry, Internal Report.

Bourdillon, T. F. 1908. Forest trees of Travancore. The Travancore Government Press, Trivandrum, India.

Bourgeon, G. 1989. Explanatory booklet on the reconnaissance soil map of forest area - Western Karnataka and Goa. Travaux de la Section Scientifique et Technique, Hors Série n°20. French Institute, Pondicherry, India.

Caius, J. F. 1986. The Medicinal and Poisonous Plants of India. Scientific Publishers, Jodhpur, India.

Chandrabose, M., and N. C. Nair. 1987. Flora of Coimbatore. Bishan Singh Mahendra Pal Singh, Dehra Dun, India.

Condit, R. 1996. Defining and mapping vegetation types in mega-diverse tropical forests. Trends in Ecology & Evolution 11:4–5.

Couteron, P., R. Pélissier, D. Mapaga, J.-F. Molino, and L. Tellier. 2003. Drawing ecological insights from a management-oriented forest inventory in French Guiana. Forest Ecology and Management 172: 89–108.

ESRI, 1992–2000. ArcView 3.1. Environmental Systems Research Institute, Redlands, Washington, USA.

Guero, M. C., and B. R. Ramesh. 1999. Forest vegetation maps to Western Ghats, GIS@Development 3:42–64.

Gunnell, Y. 1997. Relief and climate in south India: the influence of the Western Ghats on the current climate pattern of penisular India. International Journal of Climatology 17:1169–1182.

Holt, R. D. 1993. Ecology at the mesoscale: the influence of regional processes on local communities. Pages 77–88 in R. D. Ricklefs and D. Schluter, editors. Species diversity in ecological communities-Historical and geographical perspectives. University of Chicago Press, Chicago, Illinois, USA.

Jain, S. K. 1981. Glimpses of Indian Ethnobotany. Oxford and IBH Publishing, New Delhi, India.

Keshava Murthy, K. R., and S. N. Yoganarasimhan. 1990. Flora of Coorg. Vimsat Publishers, Bangalore, India.

Myers, N., R. Mittermeier, C. Mittermeier, G. Da Fonseca, and J. Kent. 2000. Biodiversity hot-spots for conservation priorities. Nature 403:853–858.

Munoz, F., P. Couteron, B. R. Ramesh, and R. S. Etienne. 2007. Estimating parameters of neutral communities: from one single large to several small samples. Ecology 88:2482–2488.

Munoz, F., P. Couteron, and B. R. Ramesh. 2008. Beta-diversity in spatially implicit neutral models: a new way to assess species migration. The American Naturalist 172:116–127.

Pascal, J.-P. 1982. Bioclimates of the Western Ghats. Travaux de la Section Scientifique et Technique, Hors Série n°17. French Institute, Pondicherry, India.

Pascal, J.-P. 1984. Les forêts denses humides sempervirentes des Ghâts occidentaux de l'Inde : écologie, structure, floristique, succession. Travaux de la Section Scientifique et Technique n° 20. French Institute, Pondicherry, India.

Pascal, J.-P. 1986. Explanatory booklet on the forest map of south India (sheets: Belgaum-Dharwar-Panaji, Shimoga, Mercara-Mysore). Travaux de la Section Scientifique et Technique, Hors Série 18. French Institute, Pondicherry. http://hal.archives-ouvertes.fr/hal-00444285/fr/

Pascal, J.-P. 1988. Wet evergreen forests of the Western Ghats of India: ecology, structure, floristic composition and succession. Travaux de la Section Scientifique et Technique n° 20bis. French Institute, Pondicherry, India.

Pascal, J.-P., and B. R. Ramesh. 1987. A field key to the trees and lianas of evergreen forests of the Western Ghats (India).Publication du Departement d'Ecologie n° 23. French Institute, Pondicherry, India.

Pascal, J.-P., S. Shyam-Sunder, and V. M. Meher-Homji. 1982–1997a. Forest Map of South India (Mercara-Mysore). Travaux de la Section Scientifique et Technique, Hors Série 18a. French Institute, Pondicherry. http://hal.archives-ouvertes.fr/hal-00444285/fr/

Pascal, J.-P., S. Shyam-Sunder, and V. M. Meher-Homji. 1982–1997b. Forest Map of South India (Belgaum-Darwar-Panaji). Travaux de la Section Scientifique et Technique, Hors Série 18b. French Institute, Pondicherry. http://hal.archives-ouvertes.fr/hal-00444285/fr/

Pascal, J.-P., S. Shyam-Sunder, and V. M. Meher-Homji. 1982–1997c. Forest Map of South India (Shimoga). Travaux de la Section Scientifique et Technique, Hors Série 18c. French Institute, Pondicherry. http://hal.archives-ouvertes.fr/hal-00444285/fr/

Pascal, J.-P., R. Pélissier, M. E. Loffeier, and B. R. Ramesh. 1998. Floristic composition, structure, diversity and dynamics of two evergreen forests plots in Karnataka state, India. Pages 507–511 in F. Dallmier, and J. A. Comiskey J. A., editors. Forest biodiversity research, monitoring and modelling. Man and the Biosphere Series vol. 20, The Parthenon Publishing Group, Paris, France.

Pélissier, R., J.-P. Pascal, F. Houllier, and H. Laborde. 1998. Impact of selective logging on the dynamics of a low elevation dens moist evergreen forest in the Western Ghats (South India), Forest Ecology and Management 105:107–119.

Rama Rao, R., and M. Sahib. 1914. Flowering plants of Travancore. The Government Press, Trivandrum, India.

Ramesh, B. R., and J.-P. Pascal. 1997. Atlas of endemics of the Western Ghats (India). Distribution of tree species in the evergreen and semi-evergreen forests. Publication du Departement d'Ecologie n° 38. French Institute, Pondicherry, India.

Ramesh, B. R., and M. H. Swaminath. 1999. Assessment and conservation of forest biodiversity in the Western Ghats of Karnataka, India. French Institute, Pondicherry, India, and Fonds Francais de l'Environment Mondial, Paris, Project Report.

Ramesh B. R., M. H. Swaminath, and M. Seetharam. 1999. Strategies for the conservation of biodiversity of the Western Ghats. My Forest 35:9–14.

Ramesh, B. R., M. Seetharam, M. C. Guero, and R. Michon. 2009a. Assessment and Conservation of Forest Biodiversity in the Western Ghats of Karnataka, India. 1. General Introduction and Forest Land Cover and Land Use Changes (1977-1997). Pondy Papers in Ecology n°6. French Institute, Pondicherry. http://hal.archives-ouvertes.fr/hal-00408263/fr/

B. R. Ramesh, M. H. Swaminath, S. V. Patil, S. Aravajy, and C. Elouard. 2009b. Assessment and Conservation of Forest Biodiversity in the Western Ghats of Karnataka, India. 2. Assessment of Tree Biodiversity, Logging Impact and General Discussion. Pondy Papers in Ecology n°7. French Institute, Pondicherry. http://hal.archives-ouvertes.fr/hal-00408305/fr/

Ramesh, B. R., N. Ayyappan, P. Grard, J. Prosperi, S. Aravajy, and J.-P. Pascal. 2010a. Western Ghats v.1.0 - A multimedia identification system of evergreen species of the Western Ghats, India. Collection Ecologie 46, French Institute, Pondicherry. http://www.biotik.org/downloads/WGhats.zip

Ramesh, B. R., P. D. Venugopal, R. Pélissier, S. V. Patil, M. H. Swaminath, and P. Couteron. 2010b. Mesoscale patterns in the floristic composition of forests in the central Western Ghats of Karnataka, India. Biotropica, doi 10.1111/j.1744-7429.2009.00621.x.

Sheil, D., S. B. Jennings, and P. Savill. 2000. Long-term permanent plot observations of vegetation dynamics in Budongo, a Ugandan rain forest. Journal of Tropical Ecology 16:765–800.

Subramanian, K. N. 1995. Flora of Thenmala. International Book Distributors, Dehra Dun, India.

Subramanian, K. N., N. Venkatasubramanian, and V. K. Nallaswamy. 1987. Flora of Palghat. Bishen Singh Mahendra Pal Singh, Dehra Dun, India.

Venugopal, P. D. 2008. Regional patterns in the floristic composition and structure of forests in the Western Ghats, India. Internal Report, French Institute, Pondicherry, India.