Impact of Oil Plam Plantations on biodiversityJambi, Central Sumatra, Indonesia

“Best bet” Land-use Systems

Thematic reports

Impact of different land uses on biodiversity

Biodiversity and Productivity Assessment for Sustainable Agroforest Ecosystems

 

Unique id: IDA1AMZB

Source file: D:\Projects\ASB\ASB Country and Thematic reports\Above ground biodiversity assessmet WG\PART G.xml

 

Authors: A.N. Gillison, N.Liswanti

 

--------------------------------------------------------------------------

 

 

 

Period:  3 – 9 December 1999

 

 

 

 

Funding agency:  ACIAR (Australia). CIFOR code: R-BIO-16-1-ICR02

 

 

Summary

 

The island of Sumatra contains some of the highest levels of biodiversity on earth. The rapid expansion of Oil Palm (Elaeis guineensis) plantations has resulted in widespread removal of lowland rain forest and dramatically reduced indigenous biodiversity.  Plantation tenure can be categorised into a) Large estate >10,000ha; b) Medium estate (5000-10,000ha); c) Small holder (10-60ha) and d) Small holder (<10ha). A brief survey was undertaken to identify whether differences in biodiversity pattern under Oil Palm could be related to tenure.  A series of 20 paired sites were located that represented a broad range of tenure classes and plantations aged from two to twelve years. In each site a 40x5m plot was established and vegetation and site physical data recorded as in previous surveys (Gillison and Liswanti, 1999).  Compared with natural forest, richness in plant species and plant functional types (PFTs) under Oil Palm is reduced by at least 75% with massive loss of indigenous and endemic woody and herbaceous species. Casual observations also suggest equivalent losses in many animal groups.  Pattern analysis suggests that while there is an ecological phase shift in plantations over fours years old when canopy closure is achieved, there is also a close correspondence between tenure system and biodiversity. Although plant-based biodiversity is dramatically reduced overall, in plantations of equivalent age, richness under small holder tenure tends to be higher that that in large and medium sized estate holdings. While further research is needed to confirm these findings it apparent that if remaining biodiversity is to be conserved planning must provide for forest and stream corridors to be maintained within plantations. The establishment of large estates has greatly reduced the chances of recolonisation of native species due to the critical loss of source material. Present indications suggest  biodiversity loss of this magnitude will have a serious medium to long term impact on environmental services with associated loss in quality of life and, ultimately, reduction in profitability.  More comprehensive surveys are urgently needed in collaboration with landowners and regional planning agencies to more accurately gauge the extent of potential impact and to provide baseline information for more sustainable development.

 

 

 

1.   Introduction

 

A key research thrust of the ICRAF-led consortium on Alternatives to Slash and Burn (ASB) is to generate tools that can be used by managers and planners to seek options for sustainable management. This is particularly relevant to ensuring sustainable management under scenarios of unforecasted change in both the physical and social environments such as el Nińo events and market shocks. In determining appropriate options it is necessary to understand first, the nature of the underlying natural resource and second the likely impact of land management on this, in particular biodiversity and profitability. This report covers a biophysical investigation of the impact of Oil Palm plantation establishment in a wide area of lowland JambiProvince in Central Sumatra.  Of interest to management is the comparative impact of different tenurial systems from smallholder,(small area) to corporate ownership (large areas).  Sumatra is one of the world’s epicentres of biodiversity. Comparisons between lowland tropical rain forest in Jambi Province Sumatra and rain forested landscapes in humid tropical West Africa and the Western Amazon basin using the same survey techniques have shown the Jambi rain forest to be higher in species (180 vascular plant species in a 40x5m plot in Jambi versus 103 species in the Iquitos area of the Peruvian Amazon) (Gillison et al. 1996; Gillison, 1999). A multi-taxa, intensive biodiversity baseline study in JambiProvince also revealed similarly high trends in many faunal groups especially insects and birds (Bignell et al. 1999; Gillison, 1999; Jepson and Djarwadi, 1999; Jones et al.1999; Watt and Zborowski, 1999).

 

Despite the  extremely high biological diversity, land clearing continues unabated. Formerly pristine rainforest and older rubber plantations and ‘jungle’ rubber are being rapidly converted to Oil Palm plantations on a massive scale. A recent study by Fox et al., (2000) (Box 1) indicates the scale of this replacement.

 

 

 

Box 1

Oil Palm Plantations in Indonesia

 

“The establishment of Oil Palm plantations in the 1990s has been a major factor in land clearance in Indonesia. The conversion of forest land to Oil Palm plantations is not confined to land designated as conversionForest but frequently, in practice, includes land designated as production forest. Exact figures on current Oil Palm planting vary. One-third of Oil Palm has been planted in the last five years (BPS, 1997:112). One estimate sets the area already planted at 2.2 million hectares, of which as much as 1.5 million hectares is already in production. Another estimate sets planting slightly higher at 2.4 million hectares, of which approximately 443,000 hectares are held by state-run enterprises; 824,000 hectares are controlled by smallholders, and the rest, 1.133 million hectares, is owned by large companies (Potter and Lee, 1998). According to official figures (Tampubolon 1998), as of September 1997, there were 294 registered plantation companies scheduled to develop further plantations on 2.5 million hectares of land over the next four to five years. Table 1 provides data on the distribution of oil plantations throughout Indonesia. Prior to the recent rush to plant palm oil, most plantations were in Sumatra and most of these plantations were in North Sumatra and Riau. The area planted to palm oil was 800,000 hectares.  By 1998, new plantations had been established in all provinces of Sumatra but most substantially in Riau, Jambi and South Sumatra. Total planting on Sumatra had increased to 2.2 million hectares. At the same time, over 500,000 hectares of new plantations had been established in Kalimantan with approximately half of these plantations in West Kalimantan. Elsewhere in Indonesia, only about 150,000 hectares of land were planted with Oil Palm, half of this in South Sulawesi.”

 

Source: Fox et al. (2000). 

 

 

The visible effects of such large scale clearance can be seen in extreme gully erosion and widespread infilling and exposure of streams and drainage systems in newly planted areas (Figs. 5,6).  While these tend to be covered by regrowth the impact on plant and animal habitat is traumatic. Whereas small-holder areas tend to retain a mosaic of forest cover, these are removed by corporate holdings so that the gene pool of remaining endemic flora and fauna is dramatically reduced. The implications from this are that even if the Oil Palm plantations were abandoned and allowed to return to forest, there would be few opportunities for re-establishment of native oplants and animals.

 

According to Fox et al. (2000) if current plans are to proceed, additional allocations will bring the total land under palm oil in Indonesia to over 9 million hectares. There is a clear priority focus on the three provinces of Riau, Jambi and South Sumatra in Sumatra and on the two provinces of West and Central Kalimantan in Kalimantan with plans still be formulated for extensive palm oil plantations in Irian Jaya.  At the time of the present investigation there have been no comprehensive studies of plant and animal distibution in Central and South Sumatra. References to previous surveys are contained in Part C of the present Above-ground biodiversity section of the ASB report series (Gillison et al, 1999). 

 

 

Table 1.  Oil Palm Plantations in Indonesia

 

 

 

(1)

 

(2)

 

(3)

 

Region

 

Oil Palm Area

Mid-1980

 

Land Planted to Oil Palm Plantation

 

Land Scheduled for Oil Palm Plantation

Aceh

 

41,100

 

206,405

 

165,305

N. Sumatra

 

550,400

 

612,617

 

62,217

W. Sumatra

 

0

 

137,952

 

137.952

Riau

 

102,200

 

606,165

 

503,965

Jambi

 

30,400

 

236,059

 

205,659

S. Sumatra

 

79,100

 

309,761

 

230,661

Bengkulu

 

2,600

 

57,006

 

54,406

Lampung

 

0

 

74,530

 

74,530

W. Kalimantan

 

0

 

279,535

 

279,535

C. Kalimantan

 

0

 

110,376

 

110,376

S. Kalimantan

 

0

 

93,902

 

93,902

E. Kalimantan

 

0

 

78,938

 

78,938

N. Sulawesi

 

0

 

0

 

0

C. Sulawesi

 

11,800

 

18,036

 

6,236

S. Sulawesi

 

0

 

83,215

 

83,215

SE Sulawesi

 

0

 

0

 

0

W. Nusa Tenggara

 

1,800

 

21,502

 

19,702

Maluku

 

0

 

0

 

0

Irian Jaya

 

23,300

 

31,080

 

7,780

INDONESIA

 

842,700

 

2,957,079

 

2,114,379

Source: MoFEC, 1998 (quoted by Fox et al. 1999)

 

 

 

2.   Methods

 

A field visit was undertaken from  3 to 9 December 1999 to the heartland of the Oil Palm industry of JambiProvince. Following a field reconnaissance and disussion with local Government and industry representatives in the Bunga-Tebo and Muara-Bungo areas a series of 20 sites were located to represent as far as possible an age sequence of plantations across small-holder and corporate estate operations ranging from less than one hectare to more than 50,000 ha.  These were located according to a nested hierachy of environmental gradients using the gradsect approach of Gillison and Brewer (1985) that has been shown to be useful in rapid survey ol plants and animals (Austin and Heyligers, 1991; Wessels et al. 1998) Difficulties were encountered in selecting sites that represented within-region variability and this led to extensive travel into areas with poor road access.  Plantations at ages 2,4,5,7,10 and 12 years were sampled (Table 2). At each site  paired 40x5m plots were placed to accommodate local differences in terrain and site treatment. Particular care was taken to record history of use of insecticides, pesticides and artificial fertilisers. A field proforma revised for rapid survey for plant-based biodiversity (Gillison, 1988,1999) was used to record key site physical variables (geo-reference using a GPS; elevation (m) with aneroid digital altimeter; slope %, and aspect (deg.) , soil type, soil depth and litter depth together with terrain position). Vegetation structure included mean canopy height (m), crown cover percent (total cover, cover of woody plants and cover of non-woody plants), cover-abundance of woody plants < 1.5m tall, basal area ( m2ha-1 ), and furcation index (position of the primary break-point in a woody stem relative to total height of the individual). In each plot all vascular plant species were identified in situ by a botanist. Voucher specimens were collected for every Plant Functional Type (PFT) in every plot and later indentified at the Herbarium Bogoriense. PFTs were recorded according to a specific classification protocol (Gillison, 1981,1988; Gillison and Carpenter, 1997). PFTs are essentially morphological attributes that can be combined to describe an individual in terms of its adpative morphology. Functional attribues and functional types are becoming increasingly useful in providing additional insights into ecosystem functioning (Martinez, 1996; Woodward et al. 1996) and PFTs have been shown to be closely correlated with certain animal groups and soil nutrients including above-ground carbon (Bignell et al. 1999; Gillison, 1999; Jepson and Djarwadi, 1999; Hairiah and van Noordwijk, 1999; Jones et al.1999; Watt and Zborowski, 1999; Gillison and Alegre, 2000 – unpubl.). This is an improvement on other surveys in humid tropical countries where only limited correlates have been obtained between taxa (cf. Lawtonet al.1998). Information about site history, ownership, tenure and management practice were also recorded.  The use of PFTs has also been shown by Vanclay et al.  (1997) to be potentially useful in estimating site productivity potential for stands of mixed species of tropical timber trees.

 

Data were compiled using the recently developed VegClass software package (Gillison and Carpenter, 2000). VegClass facilitates data compilation and storage using a data structure and menu based on the rapid field survey proforma. The Windows©  based package is capable of tabulating and graphing specified variables in one or more plots and has the capacity of generating PFT diversity indices using the commonly used Shannon-Wiener, Simpson’s and Fisher’s Alpha measures (Magurran, 1996; Gillison et al. 2000, unpubl.)  In addition the VegClass produces on demand a Plant Functional Complexity (PFC) measure that reflects the  total complexity of PFT combinations in any one plot.  The data summaries can be exported to industry-standard packages such as Microsoft Excel©  and Microsoft Access© databases. The data were analysed using standard regression procedures (Minitab Ver. 12.21). An exploratory data analysis package PATN (Belbin, 1992) was also used to generate classifications using a Gower metric with a polythetic agglomerative fusion strategy (unweighted pair-group averaging) and gradient analysis using multidimensional scaling (semi-strong-hybrid) with a two-vector solution. Previous research in multi-taxa baseline studies in Sumatra and Thailand (Gillison, 1999; Gillison and Liswanti, 1999) showed that the best plant-based correlates with key fauna and soil nutrients were mean canopy height, basal area, total plant species richness, total PFT richness and a Species/PFT richness ratio. While all plant-based variables were used in the regression analyses only the above five variables were used in the cluster analyses.  By using the same MDS procedure to extract one instead of two vector solutions, it is possible to derive a ranked index of values that can be matched against land use type.  This has been found useful in studies in other tropical countries where plant-based biodiversity indicators have been sought using similar approaches (Gillison, 1999). The same procedure was undertaken in this study using the same variables. In each 40x5m plot species and PFT data were recorded cumulatively along 8 (5x5m) segments. Using these cumulative data and the VegClass software it was possible to assess sample representativeness by generating species:area and PFT:area curves.

 

All sites were clearly marked and carefully geo-referenced for follow-up studies of soil nutrients and as a basis for studying related profitability (T.Tomich and M. van Noordwijk pers. com.). Site records have been left with the ICRAF Office in Muarabungo as well as at CIFOR in Bogor.

 

 

3.   Results

 

All vascular plant species including plant families and genera and PFTs are listed in Annex 2.  All plot data are stored in electronic format at CIFOR, Bogor and are backed up in storage discs. We found that management treatment of plots varied considerably.  Whereas estate-owned plantations followed more-or-less ordered regimes of fertiliser application and tending to farmers who had the responsibilty of tending two hectare lots, independent small-holders often with less than one hectare conduced a varied regime of cattle grazing mixed with often erratic fertiliser application and weeding. Those ‘small-holders’ with large plantations (>60,000 ha) managed their holdings according to a less rigid regime of tending and fertiliser application than estate-owned enterprises (see Table 3).

 

Both the classification (Fig. 1) and ordination (MDS) (Fig.2) show a clear separation of sites based on age.  The results confirm an ecological observation in the field that after four years canopy closure occurs with the Oil Palms. This represents a significant phase-shift in the under-canopy with a reduction in herbaceous cover and in species and PFT composition and richness. Two to four-year old plantations with open canopies offer a wider range of ecological niches than the older ‘closed-canopy’ plantations. This is reflected in a corresponding increase in species/PFT ratios in the latter where more species occupy fewer PFTs (Table 4, Fig.3).  Despite the close correlation between plantation age and richness in plant species and PFTs, the highest plant-based biodiversity occurs in large holdings (>60,000 ha) operated by private individuals. Similarly sized plantations owned by estates were poorer in both species and PFTs while medium-sized, estate-owned plantations contained the lowest biodiversity overall (Figs. 3 & 4). 

 

The asymptotes shown in the species:area, PFT:area and species/PFT:area curves (Annex 3) indicate that with very few exceptions, all plots are sufficiently representative of the target plantations. The curves with highest slopes for species and PFTs are those under small-holder management. This indicates higher alpha-diversity for small-holder versus large estate holdings.

 

 

 

4.   Disussion and conclusions

 

Due to the relatively low number of samples it was not possible to account for local variations in fertiliser application, weeding regimes and grazing intensity by cattle and buffalo. A greater density of samples will be required to accommodate these effects. Despite these shortcomings the results illustrated here are consistent with our observations in the fileld. The trends in impact on overall plant-based biodiversity are also very clear – namely that estate-owned plantations incur the highest impacts on biodiversity. This is no doubt due to the more intense tending prescriptions involving pesticides and weedicides. Compared with the generally widespread forested land use mosaics observed three years earlier in the same general area, the impact observed on the landscape during the survey was dramatic (Figs, 5,6) with massive reduction in indigenous species and an increase in invasive weeds, particularly Asteraceae and Melastomataceae. Plantations 10 years and older were, on the other hand reservoirs of many fern species (Fig. 6) that became established in the axils of dead Oil Palm fronds. Many of these (e.g. Dicranpoteris linearis, Nephrolepis biserrata, Pteris ensiliformis, Pyrrosia lanceolata, Vittaria ensiformis) represent ‘weedy’ species. We conclude that with the almost complete removal of stocks of native species of both plants and anilmals that there is little potential for rehabilitation. The widespread and ongoing conversion of forest to Oil Palm throughout the Jambi lowlands and other areas of central and southern Sumatra has already resulted in a loss of biodiversity of monumental proportions. The infilling of streams, the lack of forested buffer strips along streams, widespread erosion and reduction in water quality is pervasive.  The immediate and long term impact of this management activity will be to greatly reduce those environmental services that are critical to the landscape maintenance, ecosystem health and, ultimately to profitability and human welfare.

 

International concern has been widely expressed about the need for efficient management guidelines and indicators that can be used to help sustain both biodiversity and agricultural productivity (Reid, et al. 1993 ; World Bank, 1995) and in a way that is consistent with the National Biodiversity Management Strategy of Indonesia (Government of Indonesia, 1993). The present study has highlighted the need for adequate surveys prior to and during large-scale land clearing and establishment of monoculture crops such as Oil Palm.  It appears certain that without immediate Government intervention the biodiversity heritage of all or most of the Sumatran lowlands will be lost within the next four to five years. It is of paramount urgency that baseline studies be conducted in the remaining foothills and uplands of the BukitBarisanRange and other upland areas of Sumatra before similar management impacts are experienced. Discussions with the Ministry of Forestry including the Department of Conservation and the National Parks Service indicate that these agencies are willing to support such investigations.

 

 

5.   References

 

Fox, J., Wasson, M. and Applegate G. (2000). Forest Use Policies  and Strategies in Indonesia: A Need for Change. Draft Document prepared for World Bank. (CIFOR).

Sunderlin William D. and Ida Aju Pradnja Resosundarmo. (1996). Rates and Causes of Deforestation in Indonesia: Towards a Resolution of the Ambiguities. Center for International Forestry Research, Occasional Paper No. 9.

Potter, L and Lee, J. (1998).  Oil-Palm in Indonesia: Its role in Forest Conversion and the Fires of 1997/98, WWF Indonesia Forest Fires Project.

Austin, M.P. and Heyligers, P.C. (1991). New approaches to vegetation survey design: gradsect sampling. In: Nature Conservation: Cost Effective Survey and Data Analysis (C.R.Margules and M.P. Austin, eds.) pp. 31-37. CSIRO, Australia.

Belbin, L. 1992. PATN Pattern Analysis Package: Technical Reference. CSIRO Div. Wildlife and Ecology, Canberra.

Bignell, D.E., Widodo, E., Susilo, F.X. and Suryu, H. (1999). Soil macrofauna. Ground-dwelling ants, termites, other macroarthropods and earthworms. In: Gillison, A.N., Liswanti, N.L., (Eds.), An intensive biodiversity baseline study in Jambi province, Central Sumatra, Indonesia. In: Gillison, A.N. (coordinator), Above-ground biodiversity assessment working group summary report 1996-99. Impact on biodiversity of different land uses. Alternatives to slash and burn project. ICRAF, Nairobi, pp. 77-106.

Gillison, A.N. 1981. Towards a functional vegetation classification. In: A.N. Gillison and D.J. Anderson (eds.)  Vegetation Classification in Australia. CSIRO and ANU Press, Canberra. pp. 30-41.

Gillison, A.N. 1988. A plant functional proforma for dynamic vegetation studies and natural resource surveys. Tech. Mem. 88/3. CSIRO Div. Water Resources, Canberra.

Gillison, A.N., (compiler). 1999. Above-ground biodiversity assessment working group summary report 1996-98. Goal 2: impact on biodiversity of different land uses. Bogor, Indonesia: Center for International Forestry Research

Gillison, A.N. and Alegre, J.C. (unpubl.). The use of plant functional attributes in characterising plant biodiversity and land use impact in a forested land use mosaic in the Peruvian Amazon basin.

Gillison, A.N. and Brewer, K.R.W. (1985). The use of gradient directed transects or gradsects in natural resource surveys. J. Environ. Manage. 20: 103-127.

Gillison, A.N. and Carpenter, G. (1997). A plant functional attribute set and grammar for dynamic vegetation description and analysis. Funct. Ecol. 11: 775-783.

Gillison, A.N., Liswanti, N.L., (eds.), 1999. An intensive biodiversity baseline study in Jambi province, Central Sumatra, Indonesia. In: Gillison, A.N. (coordinator), Above-ground biodiversity assessment working group summary report 1996-99. Impact on biodiversity of different land uses. Alternatives to slash and burn project. ICRAF, Nairobi, pp. 41-53.

Gillison, A.N., Liswanti, N. and Arief-Rachman, I. (1996). Rapid Ecological Assessment, KerinciSeblatNational Park Buffer Zone, Central Sumatra: Report for Plant Ecology. CIFOR Working Paper No. 14., Bogor, Indonesia.

Gillison, A.N., Carpenter, G., Thomas, M.R.  (2000) Plant functional diversity and complexity: two new complementary  measures of species diversity. (Unpubl.)

Government of Indonesia: State Ministry of Environment (1993). Indonesian NationalStrategy on the Management of Biological Diversity. pp. 33. Jakarta.

Hairiah, K. and van Noordwijk, M. 1999.  Soil properties and carbon stocks. In: Gillison, A.N., Liswanti, N.L., (eds.), An intensive biodiversity baseline study in Jambi province, Central Sumatra, Indonesia. In: Gillison, A.N. (coordinator), Above-ground biodiversity assessment working group summary report 1996-99. Impact on biodiversity of different land uses. Alternatives to slash and burn project. ICRAF, Nairobi, pp. 143-154.

Jepson, P, Djarwadi  (1999).  Birds. In: Gillison, A.N., Liswanti, N.L., (Eds.), An intensive biodiversity baseline study in Jambi province, Central Sumatra, Indonesia. In: Gillison, A.N. (coordinator), Above-ground biodiversity assessment working group summary report 1996-99. Impact on biodiversity of different land uses. Alternatives to slash and burn project. ICRAF, Nairobi, pp. 41-53.

Jones, D. Susilo. F.X, Bignell, D.E. and Suryo, H. (1999). Terrestrial Insects: Termites. Species Richness, Functional Diversity and Relative Abundance of Termites under Different Land Use Regimes. In: Gillison, A.N., Liswanti, N.L., (Eds.), An intensive biodiversity baseline study in Jambi province, Central Sumatra, Indonesia. In: Gillison, A.N. (coordinator), Above-ground biodiversity assessment working group summary report 1996-99. Impact on biodiversity of different land uses. Alternatives to slash and burn project. ICRAF, Nairobi, pp. 107-116.

Lawton, J.H., Bignell, D.E., Bolton, B., Bloemers, G.F.,  Eggleton, P., Hammond, P.M., Hodda, M., Holt, R.D., Larsen, T.B., Mawdsley, N.A., Stork., Srivastiva, D.S. , Watt, A.D.  (1998).  Biodiversity inventories, indicator taxa and effects of habitat modification in tropical forest. Nature, 391: 72-76.

Magurran, A.E. (1988). Ecological Diversity and its Measurement. Croom Helm, Lond.

Martinez, N. D. (1996). Defining and measuring functional aspects of biodiversity. In K.J. Gaston, (ed.) Biodiversity: a biology of numbers and difference. Blackwell Science, Oxford, pp. 114-148.

Reid, W.V., McNeely, J.A., Tunstall, D.B. , Bryant, D.A. and Winograd, M. (1993). Biodiversity indicators for policy makers. World Resources Institute, WashingtonD.C.

Vanclay, J.K., Gillison, A.N. and Keenan, R.J. (1996).  Using plant functional attributes to quantify site productivity and growth patterns in mixed forests. For. Ecol. Manage. 94: 149-163.

Watt, A.D. and Zborowski, P. (1999). Canopy insects: Canopy arthropods and butterfly survey: Prelliminary report. In: Gillison, A.N., Liswanti, N.L., (Eds.), An intensive biodiversity baseline study in Jambi province, Central Sumatra, Indonesia. In: Gillison, A.N. (coordinator), Above-ground biodiversity assessment working group summary report 1996-99. Impact on biodiversity of different land uses. Alternatives to slash and burn project. ICRAF, Nairobi, pp. 57-76.

Wessels, K.J., Van Jaarsveld, A.S., Grimbeek, J.D. and Van der Linde, M.J. (1998). An evaluation of the gradsect biological survey method. Biol. Cons.7: 1093-1121.

Woodward, F.I., Smith, T.M and Shugart, H.H. (1996). Defining plant functional types: the end view. In: Plant Functional Types:their relevance to ecosystem properties and global change. T.M. Smith, H.H. Shugart and F.I. Woodward, eds. pp. 355-359. CambridgeUniversity Press, Cambridge.

World Bank, Global Environment Coordination Division, Land, Water and Natural Habitats Division (1995). Mainstreaming  Biodiversity in Development: A World Bank Assistance Strategy for Implementing the Convention on Biological Diversity. Pp. 29 (Annexes I-IV). Environment Department Paper No. 29. Biodiversity Series.

 

 

 

Table 2.   Site location and physical features

 

 

Site

Symbols

location

Date