the importance of monitoring biological diversity and its

GREEN et al. 41

Tropical Ecology 50(1): 41-56, 2009 ISSN 0564-3295 © International Society for Tropical Ecology www.tropecol

The importance of monitoring biological diversity and its application in

Sri Lanka

MICHAEL J.B. GREEN*, RIC HOW, U.K.G.K. PADMALAL & S.R.B. DISSANAYAKE

Department of Wildlife Conservation, 382 New Kandy Road, Malabe, Sri Lanka

Abstract: Sri Lanka, together with the Western Ghats in southern India, is one of the 34 global hotspots for biological diversity. Much of this diversity, in terms of species, is endemic to the island. While much of Sri Lanka’s flora and fauna has been collected and described prior to the 20th century, quantitative information about their distribution and status is limited, and many new species continue to be discovered. The Department of Wildlife Conservation has recently embarked on an ambitious initiative to benchmark and subsequently monitor key taxonomic groups of plants and animals that occur within its protected areas. This is in line with the Convention on Biological Diversity, which requires Member States to identify and monitor the components of biological diversity important for its conservation and sustainable use. This paper considers the importance of monitoring biological diversity, reviews current methodologies and describes the approach developed for protected areas in Sri Lanka.

Resumen: Sri Lanka, junto con los Gates Occidentales del sur de la India, es uno de los 34 puntos calientes de la diversidad biológica del planeta. Mucha de esta diversidad, en términos de especies, es endémica de la isla. Si bien gran parte de la flora y la fauna de Sri Lanka fue recolectada y descrita antes del siglo XX, la información cuantitativa sobre su distribución y estatus es limitada, y se siguen descubriendo muchas especies nuevas. El Departamento de Conservación de la Vida Silvestre se embarcó recientemente en una iniciativa ambiciosa para establecer la información básica sobre los grupos taxonómicos clave de plantas y animales que están presentes en sus áreas protegidas y su monitoreo subsiguiente. Este proyecto responde a la Convención sobre Diversidad Biológica, la cual requiere que los Estados firmantes identifiquen y monitoreen los componentes de la diversidad biológica que son importantes para su conservación y uso sostenible. Este artículo reflexiona sobre la importancia de monitorear la diversidad biológica, revisa las metodologías actuales y describe el enfoque desarrollado para las áreas protegidas en Sri Lanka.

Resumo: O Sri Lanka, juntamente com os Ghates ocidentais no sul da Índia, é um dos 34 pontos quentes para a diversidade biológica. Muita desta diversidade, em temos de espécies, é endémica da ilha. Enquanto muita da flora e fauna do Sri Lanka foi colectada e descrita antes do século 20, a informação quantitativa da sua distribuição e status é limitada, e muitas novas espécies continuam a ser descobertas. O Departamento de Conservação de Vida Selvagem embarcou recentemente numa iniciativa ambiciosa para definir o estado actual e monitorizar, subsequentemente, os grupos taxonómicos chave das plantas e animais que ocorrem no interior das áreas protegidas. Isto está em linha com a Convenção para a Diversidade Biológica, a qual requer aos Estados Membros a identificação e monitorização dos componentes importantes para a sua conservação e uso sustentado. Este artigo considera a importância da monitorização da diversidade biológica, revê as metodologias actuais e descreve a abordagem desenvolvida para as áreas protegidas do Sri Lanka.

* Corresponding Author; 33 The Close, Norwich, NR1 4DZ, United Kingdom; e-mail: [email protected]

42 BIODIVERSITY MONITORING IN SRI LANKA

Key words: Biodiversity, benchmark, baseline, gradsect, hotspot, inventory, monitor, sample, survey, Sri Lanka.

Introduction

What is biological diversity?

Biological diversity, that is, the variety of life on Earth1 resulting from millions of years of evolution and thousands of years of cultivation of plants and domestication of animals, can be described at many levels, from DNA and genes to species, population, communities and ecosystems. It is often considered in terms of the diversity of genes, species and ecosystems – three fundamental and hierarchically-related levels of biological organisation (WCMC 1992).

Identifying and monitoring biological diversity is a huge and potentially infinite task given its variability in time and space and its spectrum of levels. For most practical purposes of conserving biological diversity in situ (i.e. in the wild, rather than ex situ in ‘captive’ collections of plants and animals), it is sufficient to focus on species and ecosystems. Only in a few cases where species are limited to one or two very small population may it be necessary to monitor diversity at the genetic level to mitigate against potential inbreeding and extinction. This paper focuses largely on species diversity, the variety of living organisms, including micro-organisms, fungi, plants and animals. It is measured by the total number of species within a given area of study. Species diversity is linked to the survival of ecosystems and the services that they provide, many of which directly benefit human welfare. When a species disappears, it can affect a whole ecosystem, as can the appearance of an invasive or exotic species.

1 It is defined in the Convention on Biological

Diversity as “... the variability among living organisms from all sources including, inter alia, terrestrial, marine and other aquatic ecosystems and the ecological complexes of which they are part; this includes diversity within species, between species and of ecosystems.” [Article 2]

Why identify and monitor biological diversity?

Understanding biological diversity in terms of the processes by which ecosystems and their components function, be it at community, species, population or genetic levels, is critical to informing its sustainable use and safeguarding it for the benefit of future generations. Given that biological diversity is dynamic, continually evolving and changing in response to biotic and abiotic fluctuations and other environmental pressures, it is necessary to record in time and space (i.e. benchmark) its status quo and, subsequently, monitor that status quo in order to identify changes and assess their impacts. Such impacts may require intervention or mitigation measures to safeguard the future conservation, including sustainable use, of biological diversity. Crucially important is the need to identify species present in areas of natural habitat ahead of any changes in land use in order to assess what diversity may be lost from a locality. This is particularly pertinent to tropical ecosystems, where levels of endemism tend to be higher than in more temperate regions and, consequently, the risks of species becoming globally extinct may be greater.

States Parties to the Convention on Biological Diversity are required to identify and monitor the components of biological diversity important for its conservation and sustainable use (Article 7). These components, as defined in Annex I of Article 7, include: ecosystems and habitats, species and communities; and genomes and genes of social, economic, cultural, scientific or conservation importance. The provisions apply to the components of biological diversity both within (in-situ) and outside of (ex-situ) their natural habitats.

Sampling biological diversity

Species diversity, as measured by inventories, is the ‘stock in trade’ for selection of areas for conservation, many management decisions, nearly all environmental impact assessments and most political discussions on biodiversity. While

GREEN et al. 43

surrogates may be found for defining ecosystem diversity (such as forest types or vegetation associations) and genetic diversity is more frequently described only for larger charismatic or economically important species, diversity at the species level invariably underpins considerations of biological diversity in both scientific and public debates.

Description of species diversity is generally confined to well researched groups, such as vascular plants and larger vertebrates, with little attention given to the remaining 99% of biodiversity that is made up of non-vascular plants, fungi, invertebrates and micro-organisms (Ponder & Lunney 1999). The completeness of inventories compiled for these frequently monitored charismatic groups are greatly influenced by scale, with global or continental diversity generally well described by species richness (Myers et al. 2000), regional or landscape diversity less well described (Gomez de Silva & Medellin 2001) and local diversity often poorly described (How 1998; Thompson et al. 2003).

Questions concerning the derivation, accuracy, adequacy and interpretation of inventories are a central issue to consider when determining the relevance of species lists to the conservation and management of biological diversity.

Sampling considerations

An understanding of the strengths and limitations of the methodology employed to generate species inventories is essential to their interpretation. Methodologies vary enormously, particularly between plants which are sedentary and animals which are mobile to varying degrees. Plant species richness is usually determined or counted in absolute terms with all individuals, species or assemblages in an area measured in totality, while animal species richness can only be estimated by one of several approaches that take into account the movement or behaviour of the specific group being assessed.

Vertebrate groups, which represent the ’charismatic’ taxa that form the majority of the faunal species inventories currently available for biodiversity assessment, all require different sampling methodologies. Thus, inventories should only be interpreted after taking into account:

• methodological constraints with respect to the taxonomic group being assessed;

• timing and duration of the survey; and • expertise of the recorders in identifying the

taxa and their experience with the methodologies.

Timing, duration and replication of sampling

The distribution and abundance of most faunal species vary temporally in relation to daily, seasonal or longer-term cycles or fluctuations in their environments. This variation reflects the genetic diversity of species as well as the diversity of communities to which they belong. Thus, sampling of the different faunal groups should be undertaken at times of day or night and during seasons of greatest activity. In the case of plants, sampling is best done during the flowering seasons or, at least, avoiding the driest times of year, to facilitate the ready identification of species.

Shortcomings in interpreting these temporal variations are compounded by the fact that different responses may be expected in different ecoregions, where driving environmental variables, such as day-length, rainfall and temperature, operate at different times and scales. Frequently, biodiversity surveys are of such short duration that less than 50% of the total assemblage is recorded and this percentage is generally highly skewed in favour of the more common species (How 1998).

In assessing species diversity it is essential that consideration is given to variation in activity and distribution over sampling time. There are no ‘hard and fast’ rules concerning the number of replicates needed to accurately assess variation but 35 replicates per sampling unit of classification has been shown to be necessary to generate precise and unbiased estimates of diversity evenness indices (Payne et al. 2005).

Expertise

An additional constraint applicable to species inventories relates to the expertise of the recorder. Lists based on observation alone, without supporting evidence from voucher specimens, are subjective and have to be assessed with respect to the competence of the observer to correctly identify the species. This is particularly relevant to birds and amphibians where the experience of the observer is crucial to the species specific identification of calls, behaviour and, in the case of


mammals, indirect evidence such as tracks, droppings and feeding signs.

Inventory interpretation

Species inventories are used in a wide variety of contexts and analysed in many different ways. As Boero (2001) states: “Look at species lists in a standard ecological paper and check how accurate are the data sets that, often, are analysed with the most sophisticated statistical packages”. Many inventories are merely compendiums of information collected previously from an area and, therefore, contain all the inherent limitations associated with the use of non-quantitative methodologies. Inventories generated from museum or herbarium collections are cases in point. Similarly, unwarranted emphasis is regularly placed on threatened species, for which surveys are rarely adequate, with the result that these threatened taxa form the basis of resource allocation for species recovery plans, weight the criteria for reserve design, constrain development or exploitation and underpin many state-of-the-environment decisions (Possingham et al. 2002). However, for the better studied taxonomic groups, such as birds and mammals and to lesser extent flowering plants, there may be sufficient historical records by way of species inventories to identify and evaluate long-term and broad-scale changes in the composition and distribution of the flora or fauna.

Sampling adequacy and inventory evaluation

The detection of faunal species is highly variable and dependent on the complex relationship between an animal’s behaviour and its environment. Consequently, there is a high degree of variation in determining the composition of species in floral or faunal assemblages. Assessments of faunal assemblages are often made using the most conservative information available - the basic species inventory. Balmer (2002) provides evidence that analyses devoid of species’ relative abundances do not reveal the real ecological patterns in the data, based on an analysis of assemblage differences using similarity measures that considered both species presence and abundance. It is particularly important that analyses include species recorded only once at one or more locations, as this may provide valuable

information about rare species most at risk from threats such as environmental change and disturbance. Many ecological packages allow for this type of analysis.

Species accumulation curves or alternative methods (Diserud & Aagaard 2002; Smith et al. 2000; Thompson et al. 2003), based on either sampling effort or captures, are essential for an appraisal of the completeness of an inventory. For example, species accumulation curves were used to identify forests which were not adequately surveyed during the National Conservation Review, thereby highlighting the need for further sampling prior to any decisions being taken about their future management (Green & Gunawardena 1997). Other approaches to determining the adequacy of inventories are based on the probability of capturing new species using capture-recapture algorithms (Nichols et al. 1998).

The area of inventory evaluation is expanding rapidly in the literature with suggestions that auto similarity (Cao & Larsen 2002) and sensitivity analysis (Freitag & van Jaarsveld 1998) provide more accurate measures of species richness than estimates or comparisons based on sample size.

Long-term monitoring of biological diversity is expensive and rarely undertaken. However, repeated surveys provide fundamental information on seasonal, annual and other cyclical changes that cannot be obtained by other means. Such studies are necessary for obtaining a detailed inventory of threatened, rare or uncommon species in a location (How & Cooper 2002).

Alternatives to species inventories

Umbrella, focal, indicator or keystone species (Lambeck 1997), have been proposed as surrogates to replace more time-consuming and expensive surveys to inventory entire taxonomic groups. While some of the literature on the effectiveness of these alternative approaches remains conjectural (Simberloff 1998), it has been shown for woody plants in Sri Lanka, for example, that higher-taxon richness can be used as a surrogate for total species richness (Balmford et al. 1996a, 1996b).

The use of surrogates has been taken even further using the concept of congruence, whereby knowledge about the diversity of one taxonomic group is interpreted to reflect the diversity of other groups. MacNally et al. (2002) advise caution in using this approach, having found a lack of

GREEN et al. 45

congruence in diversity between taxonomic groups within a specific area (i.e. high diversity in one taxonomic group does not necessarily correlate with high diversity in another). At the bioregional scale in Western Australia, for example, taxa assemblages differ markedly according to the underlying climatic, edaphic and environmental conditions (McKenzie et al. 2000).

Finally, at regional scales, there is recent literature to suggest that species richness can be predicted from environmental variables (Trakhtenbrot & Kadmon 2006). Such an approach has considerable attraction in terms of costs and time, with a potentially important role in informing regional policy and decision-making processes.

Sri Lanka - a case study

Sri Lanka, together with the Western Ghats in southern India, is a global hotspot for biological diversity (Myers et al. 2000), 34 of which are currently recognised (Mittermeier et al. 2005). These 34 hotspots are defined regions where 75% of the planet’s most threatened mammals, birds and amphibians survive within habitat covering just 2.3% of the Earth’s surface. To qualify as a hotspot, a region must meet two criteria: it must contain at least 1,500 species of vascular plants ( > 0.5 percent of the world’s total) as endemics; and it must have lost at least 70% of its original habitat due to the impact of human activities.

Knowledge about Sri Lanka’s wealth of species is summarised in Table 1 for some of the better known and more conspicuous taxonomic groups of plants and animals. Levels of endemism are extremely high for many of these groups, an extreme example being freshwater crabs. All 51 species are endemic and 26 species are known from only one locality (Bahir & Pethiyagoda 2006). Much of this diversity is found in the montane, submontane and lowland rain forests of the wet zone and moist monsoon forests of the intermediate zone. It is also here that levels of endemism are highest, especially among flowering plants (Dassanayake et al. 1980-2000) and some of the less mobile faunal groups, such as tree-frogs, agamid lizards and skinks (Bambaradeniya 2006).

While much of Sri Lanka’s flora and fauna had been collected and described prior to the 20th century, many new species continue to be

discovered and await scientific description, particularly among some of the invertebrate and smaller vertebrate groups. Spiders, for example, are poorly researched: 501 species are known but the actual number might exceed 4,000 species (Benjamin & Bambaradeniya 2006). Among vertebrates, much recent collection and taxonomic review of amphibians has resulted in a near doubling of this fauna from 53 to 102 species, based on published descriptions. More new species await description: the total size of this group is currently estimated at about 140 species (Pethiyagoda et al. 2006).

Much, but by no means all, of Sri Lanka’s biological diversity is represented within its system of protected areas, administered principally by the Department of Wildlife Conservation and the Forest Department. Such protected areas, in theory though not always in reality, have been selected for their wealth of biological diversity and are safeguarded under various national legislation for long-term conservation purposes. Knowing what and how biological diversity is distributed within a protected area is essential for informing its future management, as well as providing context to the planning and management of the country’s entire protected areas system. As relatively little is known about biological diversity within many of Sri Lanka’s protected areas, let alone outside protected areas, a priority is to identify and monitor the components of biological diversity within such areas designated for its conservation, particularly since so much species diversity is endemic to the island.

Recent surveys of in situ biological diversity

The most comprehensive nationwide survey to date is the National Conservation Review of forest biodiversity, undertaken in 1991-1996 by the Forest Department. Woody plants and selected animal groups within all natural forests of 200 ha or more, except those occurring in the politically inaccessible northern and eastern parts of the country, were surveyed (Green & Gunawardena 1997). Totals of 1,153 woody plant, 64 butterfly, 71 mollusc, 32 amphibian, 65 reptile, 139 bird and 37 mammal species were recorded, representing a significant proportion of the diversity of each of these groups as summarised in Table 1. Other more recent surveys have been carried out in


greater detail for specific taxonomic groups, such as plants at Peak Wilderness (Singhakumara 1995a) and Kanneliya-Dediyagala-Nakiyadeniya (Singhakumara 1995b), amphibians and reptiles in the Knuckles (de Silva 2005) and an island-wide survey of land snails (Naggs et al. 2005), but none covers the range of plant and animal groups, using quantitative methods, employed by the National Conservation Review.

The task of identifying further and monitoring the components of Sri Lanka’s biological diversity is urgent because there is evidence that much is being lost rapidly. Deforestation and degradation of forests, for example, has reduced closed canopy forest from 84% in 1888 to 24% in 1992 (Legg & Jewell 1995), the causal factors being encroachment and alienation of forest land for

settlement and agriculture (Forestry Planning Unit 1995). Loss of habitat is accompanied by a reduction in the distribution and abundance of species and, in the case of endemics, may even result in their global extinction. Reference to the 14 volume Revised Handbook to the Flora of Ceylon (Dassanayake et al. 1980-2000), for example, shows that 133 (4.4%) of Sri Lanka’s 3,044 indigenous species have not been accessioned by the National Herbarium at Peradeniya since the beginning of the 20th century (R. Pethiyagoda, pers. comm. 2007). Of these, 63 are endemic, which amounts to 6.9% of the country’s 919 endemic species. While not necessarily extinct, there is a serious lack of knowledge about the conservation status of a significant component of the flora, underlining the

Table 1. Recorded diversity of Sri Lanka’s better known flora (Dassanayake et al. 1980-2000; Senaratna 2001; MENR 2002) and fauna (Bambaradeniya 2006).

Taxonomic groups No.

species No.

endemics %

endemics Notes

INDIGENOUS FLORA

Algae 896 unavailable

Fungi 1,920 unavailable

Lichens 110 39 28.2

Mosses 575 unavailable

Liverworts 190 unavailable

Ferns and allies 314 57 18.2

Gymnosperms 1 0

Angiosperms 3,044 919 30.2 Excludes 1,083 exotic species

INDIGENOUS INVERTEBRATE FAUNA

Bees 148 21 14.2

Dragon/damselflies 120 57 47.5 Includes 4 new endemic species to be described

Aphids 84 2 2.4

Ants 181 1 0.6

Butterflies 243 20 8.2

Ticks 27 1 3.7

Spiders 501 unavailable Total number of species likely to exceed 8,000

Freshwater crabs 51 51 100.0 Of the 7 genera, 5 are endemic to Sri Lanka

Land snails 246 204 82.9 Excludes 18 exotic, mostly agricultural species

VERTEBRATE FAUNA

Freshwater fish 82 44 53.7 Excludes 22 species introduced since 1982

Amphibians 102 88 86.3 Many more amphibian species await description

Reptiles 184 105 57.1 Includes 5 marine turtle and 13 marine snake species

Birds 482 25 5.2 Includes 220 breeding residents

Mammals 91 16 17.6 Excludes 16 introduced and 27 marine mammal spp.

GREEN et al. 47

importance of routinely monitoring such biological diversity.

Biodiversity Baseline Survey

The Department of Wildlife Conservation has recently embarked on an ambitious initiative, known as the Biodiversity Baseline Survey, to benchmark and subsequently monitor the diversity of species within key taxonomic groups that occur within protected areas. The design and sampling protocols developed for this Survey are outlined below, together with their application in a selection of protected areas. Full details of the methodology can be found in the field manual (Green et al. 2007a).

Benchmarking biological diversity in

Sri Lanka’s protected areas

Scope

The Biodiversity Baseline Survey, ongoing since mid-2006, is being undertaken in terrestrial and aquatic habitats of seven protected areas, chosen on account of their high and differing biodiversity:

• Bundala National Park • Ritigalala Strict Natural Reserve • Horton Plains National Park • Udawalawe National Park • Minneriya National Park • Wasgomuwa National Park • Peak Wilderness Sanctuary The following six taxonomic groups were

selected for purposes of the Survey on the basis of being (a) well known and of general interest to scientists and managers, (b) relatively easy to survey systematically and identify, and (c) potentially of value to protected areas management:

• Mammals • Reptiles • Birds • Freshwater fish • Amphibians • Vascular plants

Stratified and gradsect sampling of terrestrial

habitats

Knowledge about the distribution of the different habitats within an area provides a sound

basis for stratifying the design of a survey at an appropriate scale. Habitat maps were available for this purpose (e.g. EML 2006), having been derived from overlays of a suite of geographic and environmental variables that include geology, soil, altitude, aspect, land use and vegetation. The habitats identified from these overlays are too numerous for purposes of this Survey, so sampling focuses on the main vegetation types, while taking into account the spatial distribution of other environmental variables such as geology, soils, slope, altitude and streams for aligning transects, as well as roads for purposes of access. Horton Plains National Park, for example, can be stratified into five main vegetation types, each of which is sampled by aligning transects along the gradients of such variables (Fig. 1).

This is similar to the gradsect approach used in the National Conservation Review (Green & Gunawardena 1997). Gradsects are essentially transects aligned along environmental gradients, enabling a maximum diversity of taxa to be sampled with a minimum effort.

Quantitative integrated sampling design for

terrestrial habitats

Sampling of the different taxonomic groups is: (a) quantitative to ensure that measures of species richness and abundance are robust and able to be statistically analysed; and (b) integrated in design to provide the basis for examining relationships between plant and animal species, assemblages or taxonomic groups within the same sampling units and to make most efficient use of time, energy and other resources spent marking out and sampling transects.

The design is also simple rather than sophisticated, so that it can be readily applied to other protected areas and repeated for monitoring purposes by Department of Wildlife Conservation staff and others engaged by the Department in such work in the future.

Units for sampling terrestrial vascular plants and vertebrates are described in Box 1. The key design features of this integrated, quantitative approach to sampling terrestrial species diversity are as follows: • Each habitat type is sampled by a minimum of

four replicate transects of 1 km, taking into account as much of the environmental variation (notably in geology, soil, aspect and


altitude) as practicable, within the constraints of time and access to the area.

• Quadrats, measuring 100 m x 5 m, are located at 150 m intervals along 1 km transects (i.e. 4 quadrats per 1 km transect).

• Each quadrat is divided into 10 m x 5 m plots, within which each taxonomic group is sampled. Vascular plants are recorded in every plot of each quadrat; vertebrate taxonomic groups are recorded from within certain plots in either all quadrats (birds) or alternate quadrats (amphibians, reptiles, mammals).

• Field work logistics are based on sampling a total of 4 km of transects within a four-day period. Four days/nights are considered to be the minimum period of time necessary to effectively sample (trap) small mammals at a given location (see Wijesinghe & Brooke 2005). This governs the length of field trips, the duration of which is a multiple of four days. In

practice this amounts to the following sampling rates and intensities: o Vascular plants: four quadrats (i.e. 1 km)

per day. o Amphibians and reptiles: five alternate

plots within each of two quadrats per day (i.e. 1 km per day), using the Quadrat Clearing Technique (Heyer et al. 1994).

o Birds: two Variable Circular Plots (VCPs) at the beginning and end of each of four quadrats surveyed early morning and evening (i.e. 1 km per day). Birds are recorded directly or indirectly from their songs over a period of 10 minutes, within three zones of radius 0-10 m, >10-20 m and >20 m.

o Mammals: eleven small aluminium box traps (Sherman or Elliott) for small mammals (e.g. rodents) at 10 m intervals and two larger Tomahawk traps for small

Fig. 1. Horton Plains National Park – habitats are sampled within each vegetation type while also aligning transects along other environmental gradients, such as slope, soil and geology, as shown diagrammatically by the two lines superimposed on the vegetation map.

GREEN et al. 49

carnivores at the beginning and end of each of two quadrats along four transects per day (i.e. 4 km per day, repeated over four days/nights).

Note that the two quadrats sampled for mammals alternate with those sampled for amphibians and reptiles to minimise disturbance to one taxonomic group while recording another.

The relationship between these different sampling units with respect to the different taxonomic groups is shown diagrammatically in Fig. 2. The way in which transects were laid out in the field varied according to the terrain, distribution of the habitat and access, as illustrated in Fig. 3.

A total of 120 days of field work was undertaken during an initial six-month period of the Biodiversity Baseline Survey in 2006-07, based on an average of two 10-day field trips (including travel time) per month. Sampling intensity amounted to a total of 304 quadrats distributed along 76 km of transects in four protected areas,

each of which was stratified into 4 or 5 habitat types (Green et al. 2007b-e).

Stratified sampling of aquatic habitats

Aquatic habitats are stratified into river sub-basins, using topographic survey maps of an appropriate scale. As for the survey of terrestrial habitats, a minimum of four replicates are sampled within each river sub-basin. Of the total of 106 river sub-basins, 38 (36%) were surveyed during the initial six-month period of field work in 2006-07.

Quantitative sampling design for aquatic

habitats

Replicate sampling of the head, mid- and lower reaches of at least four rivers or streams within each sub-basin of a protected area is undertaken for fish and various measures of water quality that include pH, conductivity, dissolved oxygen, total dissolved solids, turbidity and temperature. Other

Box 1. Description of sampling units

Transects – 1 km long

Four replicate transects, 1 km in length, are located within each defined habitat (e.g. vegetation type). A transect may cover more than one habitat type in areas comprising a mosaic of highly fragmented habitats.

Where necessary or appropriate, transects may be extended from 1 km to up to 4 km in order to cover several vegetation types along a major environmental gradient, such as altitude. Such transects are essentially gradsects.

Quadrats – 100 m x 5 m

The length of a quadrat is defined for the duration of the sampling period by a centre line of highly visible nylon cord, marked at 10 m intervals by tape. A thin cane or pole, 2.5 m in length, is aligned perpendicular to this central cord in order to determine the width of the quadrat. Cords are removed at the end of the sampling period (four-days).

The ends of each 100 m length of quadrat line are georeferenced, using a GPS (Global Positioning System), and permanently marked for purposes of future monitoring, using paint and/or a thick nylon cord tied loosely around a branch of vegetation (not applicable in grasslands).

Quadrats may be aligned along a single straight line 1 km in length, or in a square (250 m x 250 m) or rectangle (350 m x 150 m). Regardless of transect shape, quadrats must be 150 m apart. This criterion is particularly important with respect to mobile taxa, such as birds, to minimise the potential for recording the same individual in adjacent quadrats. Sometimes, where access is very restricted or where the habitat is very linear in its distribution (e.g. riverine vegetation), it may be necessary to establish four parallel quadrats, spaced at least 150 m apart, either side of an access route.

Plots – 10 m x 5 m

The subdivision of quadrats into plots, by marking the central nylon cord at 10 m intervals, is critical. It enables the reptile/amphibian Quadrat Clearing Technique to be applied at 20 m intervals, small mammal traps to be positioned at 10 m intervals and, most importantly, it provides the basis for examining potential relationships between plant and animal (amphibian, reptile or small mammal) species and assemblages.


large water bodies, such lakes, reservoirs and swamps, are similarly sampled. Fish are sampled using sweep, throw, gill and seine nets, rod and line, and by snorkelling for a standard period of

time, or until it is apparent that additional species are unlikely to be recorded from the sampling station.

Fig. 2. Location of quadrats within a transect and Variable Circular Plots (VCP) within a quadrat. Quadrats A and C are sampled for mammals, quadrats B and D for reptiles (not to scale).

Fig. 3. Quadrats within a transect (1 km) are aligned in a single line (a), in parallel (b, d) or in a square (c), depending on the distribution of the habitat, access to it and terrain. Quadrats are aligned perpendicular to environmental gradients, such as altitude (as shown by contours in b and d) to maximise sampling of biological diversity (not to scale).

GREEN et al. 51

Analysing the results

The types of analyses applied to a biodiversity sampling regime depend entirely on the objectives of the study, the methodologies used and the data collected and available for analysis. For purposes of the Biodiversity Baseline Survey, data analysis is undertaken in the following sequence:

(i) Data are checked for various recording and entry errors.

(ii) The effectiveness of sampling is assessed using species accumulation curves.

(iii) Species diversity is measured using both alpha and beta statistics.

(iv) Observed species diversity is interpreted in the areas of interest.

Data screening

Data screening is a basic process that is vital for all biological survey work. At best, analyses are worthless if the data used are incorrect. At worst, it is wasteful of time and, more critically, can lead to completely erroneous conclusions upon which management or policy decisions may be based. Such outcomes must be avoided at all costs.

It is essential, therefore, to check all data for errors, such as incorrect coding of sites, miss-spellings of species names, incorrect recording of numbers and erroneous data entry. Many techniques can be used but two-way pivot tables (matrices) of species by quadrat or habitat type, containing the numbers of recorded individuals, is one very effective method.

Data appraisal - assessing effectiveness of sampling

The first step in data appraisal, after screening and verifying the data, is to evaluate the number of species recorded in relation to the effort expended in encountering them. The effort expended can be measured by the number of individuals encountered, the number of units sampled or the time taken (e.g. number of days). A graph of the cumulative number of species encountered plotted against effort provides a species accumulation or discovery curve that enables an investigator to assess the relative completeness of the sampling regime.

Some examples of species accumulation curves are provided in Fig. 4. Inadequately sampled assemblages continue to show a steep incline in the number of species encountered as sampling effort increases, as in the case of birds at Ritigala

and Horton Plains; while a well sampled assemblage shows a distinct plateau in the cumulative number of species with additional sampling, as evident for plants at Wasgomuwa and to a less extent Peak Wilderness (Fig. 4).

Caution should prevail in the analysis and interpretation of data from locations that are poorly sampled for their composite species assemblages. Where there is no evidence of any plateau in the cumulative number of species, further sampling is necessary. If this is not possible because of lack of time or other resources, it is imperative that the inadequacy of the sampling is highlighted in all analyses and reports to ensure that any management and policy decisions take account of this constraint.

Several predictive models are also available for estimating the potential number of species in a defined sampling unit. These are based variously on the functions of the number of species seen in only one or two samples (Chao 2, Jacknife), the number of species that have only one or two individuals in the entire pool of samples (Chao 1), or the proportions of samples that contain each species (Bootstrap). These estimates of potential species diversity can be generated from ecological packages such as Primer-E (Clarke & Warwick 2001) and Estimate S (Colwell 2005).

Species diversity measures

The core analyses of any conservation study are the estimates of biological diversity. A sample often comprises subgroups, referred to here as units, such as particular localities, regions or habitat types. As a consequence, two levels of diversity are commonly considered: alpha diversity, which is the diversity within a unit; and beta diversity, which is the amount of compositional variation between units.

Alpha [α] diversity

The best known measurement of diversity within a unit is Species Richness (S), which is the number of species in the unit. It is an intuitive and well-established measure that is widely used by field ecologists because of its simplicity and ease of calculation. It is also readily appreciated and easily communicated to colleagues, decision makers and the general public.

However, there is a wide and growing literature on other measures of diversity that have appeal because they are less dependent on sample


size or effort and take into account the distribution of species’ abundances in the unit under consideration. The generality and applicability of such measures have been evaluated by Jost (2006) and his recommendations have been followed in the Biodiversity Baseline Survey by adopting two indices of alpha diversity, in addition to Species Richness. These, calculated using the program Primer-E (Clarke & Warwick 2001), are as follows:

• Shannon Entropy (H’) calculated from exp

( ∑=

−S

i

ii pp1

ln ), where, pi is the proportion

of individuals of the ith species. • Simpson Concentration (D) calculated from

1/∑=

S

1i

2

ip .

It is clear from plots of these three measures for seven habitats at Wasgomuwa National Park, for example, that Species Richness is correlated with both diversity indices in all cases except grassland. Furthermore, both species diversity indices (H’ and D) are strongly correlated (Fig. 5). Comparisons of alpha diversity are a valuable means of assessing the relative conservation importance of different units (e.g. protected areas, habitats, transects).

Beta [β] diversity

Beta diversity quantifies the amount of compositional variation and relationships between units of interest, such as habitats within a locality. There are numerous ways of measuring beta diversity that are dependent on concepts or measurements of underlying sources of compositional variation (see Magurran 2003).

For purposes of the Biodiversity Baseline Survey, a widely-used measure, the matrix of the Bray-Curtis coefficient of similarity, is calculated

Fig. 4. Examples of species accumulation curves for plants (left) and birds (right) in various protected areas, based on records from the Biodiversity Baseline Survey.

Fig. 5. Measures of alpha diversity in Wasgomuwa National Park, using Species Richness (S), Shannon Entropy (H’) and Simpson Concentration (D). Dry Mixed Evergreen Forest (DMEF) is classified into dry, intermediate and wet types.

GREEN et al. 53

(see Ludwig & Reynolds 1988; McCune & Grace 2002). As information in similarity matrices is usually very difficult to understand, various analytical procedures are used to assist with interpretation, usually by means of graphical representations. Principal Coordinates Analysis (known by several other names including ordination, metric scaling, seriation and multidimensional scaling) was adopted during the initial stage of the Biodiversity Baseline Survey, using the statistical package Genstat (Genstat 2006). Currently, similarity matrices are examined with non-metric Multidimensional Scaling, using the ecological package Primer-E. Both of these approaches present the similarity matrix in two (or more) dimensions. The relative advantages of the latter are discussed by Clarke (1993).

An example of non-metric Multi-dimensional Scaling, shown in Fig. 6, reveals some important features of floristic data collected from Peak Wilderness National Park. First, Submontane Forest quadrats fall into two distinct sub-groups, indicating this may be a heterogeneous habitat type. Similarly, there are two sub-groups of Lowland Forest, one of which is similar to Lowland Disturbed Forest. Montane Forest appears similar to one of the Submontane Forest sub-groups but care should be exercised because these two sub-groups may actually differ on a third axis. Finally, Secondary Submontane Forest, although only represented by four quadrat samples, are quite widely spread, suggesting that they may constitute a heterogeneous group.

Principle Coordinates Analysis and non-metric Multidimensional Scaling are powerful tools to validate, or otherwise, the original stratification of habitat types upon which field sampling is based. As demonstrated in Fig. 6, habitat types assigned to quadrats on the basis of vegetation maps invariably prove to be an oversimplification and some re-classification may be necessary. Also, outliers in the analyses need to be reviewed with respect to various modifying processes (e.g. timber extraction, fire, species enrichment) that may explain deviations from expected species composition.

Conclusions

Species inventories are central to the identification and description of biodiversity

hotspots at a variety of scales (Brooks et al. 2001); they play an important role in selection of areas for conservation (Cabeza & Moilanen 2001; Margules & Pressey 2000); and also provide the basis of most environmental impact assessment protocols. However, the value of species inventories is proportional to their completeness and accuracy. These key measures are seldom highlighted in the presentation of inventories and, together with a statement of methods employed, are essential for their assessment and interpretation.

Species inventories can be enhanced when greater emphasis is given to improving identification skills and the quality of data presented for interpretation. The practice of vouchering is desirable for most taxonomic groups as it provides: ‘hard evidence’ of the identity of the species; baseline information on species biology; and, importantly, an opportunity to evaluate species changes over time.

Species inventories are decidedly more informative if accompanied by quantitative data on species abundance, since this permits a more accurate and enhanced level of assessment of both the methodology and comprehensiveness of the inventories.

No single index of diversity adequately depicts species diversity in a unit. However, there is generally a strong correlation between all indices once sample sizes become large. This highlights the importance of extensive sampling.

Fig. 6. Beta diversity for vascular plants at Peak Wilderness Sanctuary. Each point represents a quadrat. The Stress value of 0.09 indicates that the two dimensional plot provides a good approximation of the distances between quadrats, in terms of similarity in species composition.


The use of surrogates for assessing biological diversity may be helpful if resources (time and funds) are scarce but their limitations must be understood and fully acknowledged. Environmental correlates can inform assessments of biological diversity but they cannot replace detailed inventory surveys.

In Sri Lanka, a hotspot for biological diversity, significant progress has been made in identifying diversity at species levels in systematic, quantifiable ways at national scales, beginning with the National Conservation Review in 1991-1996 and followed up more recently with the Biodiversity Baseline Survey described here. Such extensive surveys provide the basis for future monitoring, which is a crucial next step towards the Government of Sri Lanka meeting its obligations under the Convention on Biological Diversity.

Acknowledgements

The Biodiversity Baseline Survey is part the Protected Area Management and Wildlife Conservation Project, funded by the Asian Development Bank, World Bank-Global Environment Facility and the Government of the Netherlands. It has been undertaken in two phases for the Department of Wildlife Conservation, Ministry of Environment and Natural Resources, Sri Lanka: initially by ARD Inc. in association with Infotechs Ideas Pvt. Ltd and Greentech Consultants Pvt. Ltd and subsequently, during an extension to the Project, by Infotechs Ideas Pvt. Ltd in association with Greentech Consultants Pvt. Ltd.

Various members of the Biodiversity Baseline Survey team contributed to the development of methods for sampling the different taxonomic groups, further details of which can be found in the field manual (Green et al. 2007a): B.M.P. Singhakumara (plants), S.M.D.A.U. de Alwis (fish), P.N. Dayawansa (amphibians and reptiles), D. Weerakoon (birds), M.R. Wijesinghe (non-volant mammals) and W.B. Yapa (bats).

References

Bahir, M.M. & R. Pethiyagoda. 2006. Conservation of Sri Lankan freshwater crabs. pp. 77-83. In: C.N.B. Bambaradeniya (ed.) The Fauna of Sri Lanka:

Status of Taxonomy, Research and Conservation. The World Conservation Union, Colombo, Sri Lanka & Government of Sri Lanka.

Balmer, O. 2002. Species lists in ecology and conservation: abundances matter. Conservation Biology 16: 1160-1161.

Balmford, A., M.J.B. Green & M.G. Murray. 1996a. Using higher-taxon richness as a surrogate for species richness: I. Regional tests. Proceedings of the Royal Society of London B 263: 1267-1274.

Balmford, A., A.H.M. Jayasuriya & M.J.B. Green. 1996b. Using higher-taxon richness as a surrogate for species richness: II. Local applications. Proceedings of the Royal Society of London B 263: 1571-1575.

Bambaradeniya, C.N.B. (ed.). 2006. The Fauna of Sri Lanka: Status of Taxonomy, Research and

Conservation. The World Conservation Union, Colombo, Sri Lanka & Government of Sri Lanka.

Benjamin, S.P. & C.N.B. Bambaradeniya. 2006. Systematics and conservation of species in Sri Lanka: Current status and future prospects. pp. 70-76. In: C.N.B. Bambaradeniya (ed.) The Fauna of Sri Lanka: Status of Taxonomy, Research and

Conservation. The World Conservation Union, Colombo, Sri Lanka & Government of Sri Lanka.

Boero, F. 2001. Light after dark: the partnership for enhancing expertise in taxonomy. Trends in Ecology and Evolution 16: 266.

Brooks, T., L. Hannah, G.A.B. da Fonseca & R.A. Mittermeier. 2001. Prioritizing hotspots, representing transitions. Trends in Ecology and Evolution 16: 673.

Cabeza, M. & A. Moilanen. 2001. Design of reserve networks and the persistence of biodiversity. Trends in Ecology and Evolution 16: 243-248.

Cao, Y., D.D. Williams & D.P. Larsen. 2002. Comparison of ecological communities: the problem of sample representativeness. Ecological Monographs 72: 41-56.

Clarke, K.R. 1993. Non-parametric multivariate analyses of change in community structure. Australian Journal of Ecology 18: 117-143.

Clarke, K. R. & R. M. Warwick. 2001. Changes in Marine Communities: An Approach to Statistical

Analyses and Interpretation. 2nd edn. PRIMER-E, Plymouth.

Colwell, R. K. 2005. Estimate S: Statistical Estimation of Species Richness and Shared Species from Samples. Version 7.5. User's Guide and application. [http://purl.oclc.org/estimates]

Dassanayake, M.D. & F.R. Fosberg. (eds.). 1980-1991. A Revised Handbook to the Flora of Ceylon. Volumes I-VII. Amerind Publishing Company, New Delhi.

GREEN et al. 55

Dassanayake, M.D. & W.D. Clayton. (eds.). 1996-2000. A Revised Handbook to the Flora of Ceylon. Volumes X- XIV. Oxford and IBH Publishing Company, New Delhi.

Dassanayake, M.D., F.R. Fosberg & W.D. Clayton. (eds.). 1994-1995. A Revised Handbook to the Flora of Ceylon. Volumes VIII-IX. Amerind Publishing Company, New Delhi.

de Silva, A. (ed.). 2005. The diversity of the Dumbara Mountains. The Knuckles Massif, Sri Lanka: with special reference to its herpetofauna. Lyriocephalus Special Issue. Journal of the Amphibian and Reptile Research Organisation of Sri Lanka (ARROS) 6: 1-382.

Diserud, O.H. & K. Aagaard. 2002. Testing for changes in community structure based on repeated sampling. Ecology 83: 2271-2277.

EML. 2006. Habitat Maps for Selected Areas in Sri Lanka: Horton Plains National Park. Department of Wildlife Conservation, Ministry of Environment and Natural Resources, Colombo.

Forestry Planning Unit. 1995. Sri Lanka Forestry Sector Master Plan. Ministry of Agriculture, Lands and Forestry, Battaramulla, Sri Lanka.

Freitag, S. & A.S. van Jaarsveld. 1998. Sensitivity of selected procedures for priority conservation areas to survey extent, survey intensity and taxonomic knowledge. Proceedings of the Royal Society, London 265: 1475-1482.

Genstat. 2006. Genstat for Windows. Release 9.1. VSN International, Oxford.

Gomez de Silva, H. & R.A. Medellin. 2001. Evaluating completeness of species lists for conservation and macroecology: A case study of Mexican land birds. Conservation Biology 15: 1384-1395.

Green, M.J.B. & E.R.N. Gunawardena. 1997. Designing an Optimum Protected Areas System for Sri Lanka's

Natural Forests. 2 Volumes. Environmental Management in Forestry Developments Project, Forest Department, Government of Sri Lanka, Colombo.

Green, M.J.B., S.M.D.A.U. de Alwis, P.N. Dayawansa, R. How, U.K.G.K. Padmalal, S. de Silva, B.M.P. Singhakumara, D. Weerakoon, M.R. Wijesinghe & W.B. Yapa. (eds.). 2007a. Biodiversity Baseline Survey: Field Manual. Department of Wildlife Conservation, Ministry of Environment and Natural Resources, Colombo. [http://203.143.23.34/BBS/bbs. html]

Green, M.J.B., S.M.D.A.U. de Alwis, P.N. Dayawansa, R. How, B.M.P. Singhakumara, D. Weerakoon & M.R. Wijesinghe. (eds.). 2007b. Biodiversity Baseline Survey: Horton Plains National Park. Department of Wildlife Conservation, Ministry of Environment

and Natural Resources, Colombo. [http://203.143. 23.34/BBS/bbs.html]

Green, M.J.B., S.M.D.A.U. de Alwis, P.N. Dayawansa, R. How, B.M.P. Singhakumara, D. Weerakoon & M.R. Wijesinghe. (eds.). 2007c. Biodiversity Baseline Survey: Peak Wilderness Sanctuary. Department of Wildlife Conservation, Ministry of Environment and Natural Resources, Colombo. [http://203.143.23.34/ BBS/bbs.html]

Green, M.J.B., S.M.D.A.U. de Alwis, P.N. Dayawansa, R. How, B.M.P. Singhakumara, D. Weerakoon, M.R. Wijesinghe & W.B. Yapa. (eds.). 2007d. Biodiversity Baseline Survey: Ritigala Strict Natural Reserve. Department of Wildlife Conservation, Ministry of Environment and Natural Resources, Colombo. [http://203.143.23.34/BBS/bbs.html]

Green, M.J.B., S.M.D.A.U. de Alwis, P.N. Dayawansa, R. How, B.M.P. Singhakumara, D. Weerakoon, M.R. Wijesinghe & W.B. Yapa. (eds.). 2007e. Biodiversity Baseline Survey: Wasgomuwa National Park. Department of Wildlife Conservation, Ministry of Environment and Natural Resources, Colombo. [http://203.143.23.34/BBS/bbs.html]

Heyer, W.R., M.A. Donnelly, R.W. McDiarmid, L.C. Hayer & S. Foster. (eds.). 1994. Measuring and Monitoring Biological Diversity: Standard Methods

for Amphibians. Smithsonian Institute Press, Washington D.C.

How, R.A. 1998. Long-term sampling of a herpetofaunal assemblage on an isolated urban bushland remnant. Journal of the Royal Society of Western Australia 81: 143-148.

How, R.A. & N.K. Cooper. 2002. Terrestrial small mammal fauna of the Abydos Plain in the north-eastern Pilbara, Western Australia. Journal of the Royal Society, Western Australia 85: 71-82.

Jost, L. 2006. Entropy and diversity. Oikos 113: 363-375. Lambeck, R. J. 1997. Focal species: a multi-species

umbrella for nature conservation. Conservation Biology 11: 849-856.

Legg, C. & N. Jewell. 1995. A 1:50,000-scale forest map of Sri Lanka: the basis for a National Forest Geographic Information System. Sri Lanka Forester Special Issue (Remote Sensing): 3-21.

Ludwig, J. A. & J.F. Reynolds. 1988. Statistical Ecology. John Wiley and Sons, New York.

MacNally, R., A.F. Bennett, G.W. Brown, L.F. Lumsden, A. Yen, S. Hinkley, P. Lillywhite & D. Ward. 2002. How well do ecosystem-based planning units represent different components of biodiversity? Ecological Applications 12: 900-912.

Magurran, A.E. 2003. Measuring Biological Diversity. Blackwell, Oxford.


Margules, C.R. & R.L. Pressey. 2000. Systematic conservation planning. Nature 405: 243-253.

McCune, B. & J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, Oregon, USA.

McKenzie, N.L., G.J. Keighery, N. Gibson & J.K. Rolfe. 2000. Patterns in the biodiversity of terrestrial environments in the Southern Carnarvon Basin, Western Australia. In: A. A. Burbidge, M.S. Harvey & N. L. McKenzie (eds.) Biodiversity of the Southern Carnarvon Basin. Records of the Western Australian Museum, Supplement 61: 511-546.

MENR. 2002. State of the Environment in Sri Lanka. A National Report prepared for the South Asian Association for Regional Cooperation. Environmental Economics and Global Affairs Division, Ministry of Environment and Natural Resources, Colombo.

Mittermeier, R.A., P.R. Gil, M. Hoffman, J. Pilgrim, T. Brooks, C.G. Mittermeier, J. Lamoreux & G.A.B. da Fonseca. 2005. Hotspots Revisited: Earth’s

Biologically Richest and Most Threatened

Terrestrial Ecoregions. Conservation International, Washington D.C.

Myers, N., R.A. Mittermeier, C.G. Mittermeier, G.A.B. da Fonseca & J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853-858.

Naggs, F., D. Raheem, K. Ranawana & Y. Mapatuna. 2005. The Darwin Initiative project on Sri Lankan land snails: Patterns of diversity in Sri Lankan forests. The Raffles Bulletin of Zoology 12: 23-29.

Nichols J. D., T. Boulinier, J.E. Hines, K.H. Pollock & J.R. Sauer. 1998. Estimating rate of local species extinction, colonisation, and turnover in animal communities. Ecological Applications 8: 1213-1225.

Payne, L.X., D.E. Schindler, J.K. Parrish & S.A. Temple. 2005. Quantifying spatial pattern with evenness indices. Ecological Applications 15: 507-520.

Pethiyagoda, R., K. Manamendra-Arachchi, M.M. Bahir & M. Meegaskumbura. 2006. Sri Lankan amphibians: Diversity, uniqueness and conservation. pp. 125-133. In: C.N.B. Bambaradeniya (ed.) The Fauna of Sri Lanka: Status of Taxonomy, Research and Conservation. The World Conservation Union, Colombo, Sri Lanka & Government of Sri Lanka.

Ponder, W.F. & D. Lunney. (eds.). 1999. The Other 99%: the Conservation and Biodiversity of Invertebrates.

Royal Zoological Society of New South Wales, Mossman, New South Wales.

Possingham, H. P., S. J. Andelman, M.A. Burgman, R.A. Medellin, L.L. Master & D.A. Keith. 2002. Limits to the use of threatened species lists. Trends in Ecology and Evolution 17: 503-507.

Senaratna, L.K. 2001. A Checklist of the Flowering Plants of Sri Lanka. MAB Check List and Handbook Series, Publication No. 22. National Science Foundation, Colombo.

Simberloff, D. 1998. Flagships, umbrellas and keystones: is single species management passé in the landscape era? Biological Conservation 83: 247-257.

Singhakumara, B.M.P. 1995a. Floristic Survey of Adam’s Peak Wilderness. Forest Department (Forestry Information Service), Ministry of Agriculture, Lands and Forestry.

Singhakumara, B.M.P. 1995b. Ecological assessment of Kanneliya, Dediyagala- Nakiyadeniya (KDN) Forest

Complex. Department of Forestry and Environmental Science, University of Sri Jayewardenepura, Nugegoda.

Smith, W., A. Solow & C. Chu. 2000. An index of the contribution of spatial community structure to a species accumulation curve. Ecology 81: 3233-3236.

Thompson, G. G., P. C. Withers, E. R. Pianka & S.A. Thompson. 2003. Assessing biodiversity with accumulation curves; inventories of small reptiles by pit-trapping in Western Australia. Australian Ecology 28: 361-383.

Trakhtenbrot, A. & R. Kadmon. 2006. Effectiveness of environmental cluster analysis in representing regional species diversity. Conservation Biology 20: 1087-1098.

Wijesinghe, M.R. & M. de L. Brooke. 2005. Impact of habitat disturbance on the distribution of endemic species of small mammals and birds in a tropical rain forest in Sri Lanka. Journal of Tropical Ecology 22: 661-668.

World Conservation Monitoring Centre. (WCMC). 1992. Global Biodiversity: Status of the Earth’s Living

Resources. Chapman & Hall, London.

the importance of monitoring biological diversity and its

Documents