INTRODUCTION
The island of Borneo, encompassing parts of Indonesia, Malaysia, and Brunei, is a global conservation priority (Myers et al. Reference Myers, Mittermeier, Mittermeier, da Fonseca and Kent2000; Olson & Dinerstein Reference Olson and Dinerstein2002; Stattersfield et al. Reference Stattersfield, Crosby, Long and Wege1998), but the lowland portions of the island of Borneo have suffered from deforestation, forest fire and conversion to estate crops (Curran et al. Reference Curran, Trigg, McDonald, Astiani, Hardiono, Siregar, Caniago and Kasischke2004; Langner et al. Reference Langner, Miettinen and Siegert2007). The central upland portions of the island are more remote and generally less suitable for industrial exploitation, and these areas remain in relatively healthy ecological conditions (Rautner et al. Reference Rautner, Hardiono and Alfred2005).
Initiatives to preserve the ecological intactness of the central areas of Borneo while the opportunity still exists have gained traction in the last ten years. In February 2007, the governments of Brunei Darussalam, Indonesia and Malaysia signed the Heart of Borneo (HoB) Declaration, which commits the three governments to a single conservation vision to ensure the effective management of forest resources and conservation of a network of protected areas, productive forests and other sustainable land uses in the upland HoB (Stone Reference Stone2007). The major partners in this process are the local and national government, in particular the agencies involved in land-use planning but the participation of the industrial sector (such as oil palm and mining companies), committed to integrate development with conservation of natural resources, is considered crucial.
The initial delineation of the HoB encompassed all the interior highlands and mountains of the headwaters of the major rivers of Borneo. Later boundaries were expanded to include the foot hills and important intact adjacent lowlands (Fig. 1).
The island of Borneo, indicating extent of forest area by type and outlining the area known as the Heart of Borneo.
There have been calls from within the conservation community for rigorous programme evaluation of conservation efforts, including the measurement of counterfactuals (Ferraro & Pattanayak Reference Ferraro and Pattanayak2006; Pullin & Knight Reference Pullin and Knight2001, Reference Pullin and Knight2009), yet examples of even basic monitoring of large conservation programmes are difficult to find. Existing large conservation monitoring efforts include the Everglades ecosystem (NRC [National Research Council] 2006, 2008, 2010, 2012), Chesapeake Bay (Boynton Reference Boynton, Bailey, Moesel and Moore2008) and the Great Barrier Reef (State of Queensland 2011). Some of these monitoring programmes represent the extensive and costly efforts of developed national and state governments. It is a challenge to develop a concise set of indicators which represent the health of such large and diverse systems. In the Everglades, it took several years of iterative research work and deliberations to reduce the initial set of system indicators from 54 (NRC 2008) to ten key indicators (NRC 2012). Yet such efforts to develop practical ecological indicators for large high priority ecosystems are desperately needed in developing countries, where resources are tight despite pressing conservation need.
The ultimate measure of the success or failure of conservation initiatives will be the ecological health of the ecosystems including their key species and sustainability of institutions which support that ecological health. There are no comprehensive data about the current state of the natural systems in the HoB. Development of a monitoring framework that could characterize both the ecological health and conservation status of the HoB would be the first step, followed by collection of data using the monitoring framework. The objectives of this study are to: (1) develop a set of practical indicators that might be representative of the ecological status of the HoB and could be monitored at appropriate time intervals; (2) develop a larger set of indicators that are representative of the overall conservation status of the HoB; and (3) analyse data for those indicators to assess the current overall ecological and conservation health of the HoB.
METHODS
For our monitoring framework, we applied an elaboration and modification of an existing framework (Gordon et al. Reference Gordon, Parrish, Salzer, Tear, Pace-Aldana, Meffe and Carroll Groom2006; Parrish et al. Reference Parrish, Braun and Unnasch2003), which is a subset of a comprehensive method for conservation planning, continuously developed over the course of twenty years of application (Poiani et al. Reference Poiani, Baumgartner, Buttrick, Green, Hopkins, Ivey, Seaton and Sutter1998). The generic version of this strategic conservation planning framework is known as the Open Standards for the Practice of Conservation (Conservation Measures Partnership 2007; Dietz et al. Reference Dietz, Brown and Swaminathan2010; Schwartz et al. Reference Schwartz, Deiner, Forrester, Grof-Tisza, Muir, Santos, Souza, Wilkerson and Zylberberg2012).
The Open Standards planning framework directs practitioners about how to select focal conservation features (concise combination of the predominant ecosystems, natural communities, keystone and/or threatened species) and associated indicators. The indicators we used were initially selected by a small group of experts (as identified by the World Wildlife Fund) on the natural history and conservation of Borneo (Appendix 1, see supplementary material at Journals.cambridge.org/ENC). The process of selecting indicators consisted of three phases: (1) selecting conservation features, or conservation ‘targets’ that were representative of the ecosystems of the area (such as tropical lowland forest or riparian systems); (2) identifying the ecological attributes of the above conservation features that must be in place for good ecological function (such as the extent of ecosystems, connectivity of ecosystems and species population size); and (3) identifying measurable indicators for the above ecological attributes (such as the area of an ecosystem or estimated number of individuals of a species).
The conservation targets selected were chosen because they were believed to best represent:
• the predominant ecosystem types (as an areal percentage) on the island. These were usually identified first because they represented most of the species in the HoB.
• keystone species or area-sensitive species that may cross several ecosystem types. Such keystone or area-sensitive species capture elements of minimum habitat block size and connectivity.
• one or two species or species groups that are subject to anthropogenic threats (for example poaching) that are not covered by previously selected conservation features.
An emphasis was placed on narrowing the set of key indicators to a maximum of 8–12 features.
The ecological attributes of the conservation targets, such as elements of the size, condition or landscape context of the selected ecosystems and species, were next to be considered. Such attributes are aspects of a target's biology or ecology that, if present, define a healthy target and, if missing or altered, would lead to the outright loss or extreme degradation of that target over time (Conservation Measures Partnership 2010). For example, if tropical lowland forest was one of the selected conservation targets, potential ecological attributes include the extent of tropical lowland forest, the physical or age structure of the forest, and presence and distribution of native species.
For each of the ecological attributes, a practical measurable indicator was identified if known. An emphasis was placed on limiting the number of indicators and on identifying practical indicators that were already being measured or that could be measured at low cost.
Subsequently, a small team of field researchers (Appendix 1, see supplementary material at Journals.cambridge.org/ENC) attempted to collect meaningful data for each of the consensus indicators, across the HoB. If data for a particular indicator were considered insufficient or unrepresentative, that indicator was temporarily omitted until better methods or data became available. After data were collected, a larger group of experts was consulted for their opinions on the usefulness of the indicators, the associated data and the rating thresholds (Appendix 1, see supplementary material at Journals.cambridge.org/ENC).
In addition to measures of ecological health in the HoB, two additional classes of conservation-related measures were discussed by the expert team: direct threats to natural systems and conservation management indicators.
The experts used objective criteria to analyse and rate (Appendix 2, see supplementary material at Journals.cambridge.org/ENC) the anticipated level of a number of direct threats to biodiversity in the HoB over the next 10 years. Each combination of potential and current threats versus biological target was analysed on four-point scales of scope, severity and irreversibility. A combined threat rating was then calculated for each threat-biological target combination and all threats were combined using Miradi conservation planning software (Appendix 2, see supplementary material at Journals.cambridge.org/ENC; Kapos et al. Reference Kapos, Balmford, Aveling, Bubb, Carey, Entwistle, Hopkins, Mulliken, Safford, Stattersfield, Walpole and Manica2008; Conservation Measures Partnership 2010). The result was then sorted by the software program to reveal those threats that were most serious across the HoB. This process clearly indicated the highest threats and the most threatened targets.
We define conservation management as the policies, regulations, institutions and capacities to manage biodiversity and natural resources. Though all of these aspects are of interest, the most important indicators are the extent to which different natural ecosystems are covered by legal protected areas and the extent to which those protected areas are effectively managed. The second conservation management indicator, management effectiveness, is typically represented by indices such as rapid assessment and prioritization of protected area management (RAPPAM; Ervin Reference Ervin2003) or the monitoring effectiveness tracking tool (METT; Stolton et al. Reference Stolton, Hockings, Dudley, MacKinnon and Whitten2007).
A consideration for collecting information about the extent of the major ecosystems was the availability of a cloud-free mosaic remote sensing coverage for all of Borneo (SarVision 2008). The indicators were developed and organized by the dominant and critical ecosystem types and species, which are described later.
An important aspect of the monitoring framework was that it attempted to provide context for the raw data collected for each indicator. A unique set of objective criteria was developed for each indicator so that every indicator could be rated on a four-point scale of viability (Parrish et al. Reference Parrish, Braun and Unnasch2003) with the following definitions (Conservation Measures Partnership 2010): ‘very good’, namely ecologically desirable status, requires little intervention for maintenance; ‘good’, namely indicator within acceptable range of variation, some intervention required for maintenance; ‘fair’, namely outside acceptable range of variation, requires human intervention; and ‘poor’, namely restoration increasingly difficult, may result in extirpation of target.
By combining the indicators (Appendix 3, see supplementary material at Journals.cambridge.org/ENC; Conservation Measures Partnership 2010), statements could be made about particular components of the HoB's natural systems or about the HoB as a whole. The investigations can be reproduced at appropriate time intervals. The assessments presented here are currently scheduled to be repeated by the end of 2012.
RESULTS
The expert workshop produced a short list of indicators organized by the most important ecosystems and species (Appendix 4, see supplementary material at Journals.cambridge.org/ENC). While all of the identified indicators are considered to say something important about the ecological health of the HoB, not all of the indicators are currently practical (Table 1) and most but not all of the indicators were measured during the first phase (Table 2). Overall, enough information has been collected to make some broad statements about the vital state/health of the HoB.
Sources of historic and current ecosystem extent in the HoB
Rated indicators of ecological health of the HoB (of which sufficient data was available).
The selected indicators (and the rationale for choosing them) and the results of the data gathering and analysis are each described. The first few indicators (extent of lowland forest, upland forest, montane forest, heath forest, limestone forest, peat swamp forest and river ecosystems) historically covered the vast majority of the HoB area. Indicators that provide information about the current extent of natural vegetation types were a consensus choice for all of the experts consulted.
Lowland forest
At one time the dominant ecosystem of lower lying areas of Borneo, the great dipterocarp forests of Borneo are among the island's most important ecological features, originally occupying 27% of the current HoB. The extent of this ecosystem was considered to be important by the expert indicator team and this was the only practical indicator developed for lowland forest. While most of the island's lowland forest was found outside of the current boundaries of the HoB, some contiguous areas of relatively intact lowland forest were included when the HoB was delineated. Forest condition was another potential factor identified, but while current light detection and ranging (LIDAR) technology permits detailed mapping of forest condition (Asner et al. Reference Asner, Powell, Mascaro, Knapp, Clark, Jacobson, Kennedy-Bowdoin, Blaji, Paez-Acosta, Victoria, Secada, Valqui and Hughes2010), the costs are prohibitively expensive and make widespread application in Borneo unlikely. An updated map of the extent of lowland rainforest ecosystems indicated that 63% of remaining historic lowland forest remains within the HoB (classified as ‘good’).
Upland forest
The structure and ecology of upland rainforest, also known as hill dipterocarp forest (Whitmore Reference Whitmore1984; Table 1) are similar to those of lowland rainforest The upland rainforest historically occupied c. 42% of the current HoB (Table 2).Extent of the upland forest was considered a critical indicator, and broad assessment of forest condition of all upland forest is not possible at this time. Borneo's upland forest lies within the relatively intact HoB, most (82%) of the forest is still standing and in many cases is considered to be in primary condition. Upland forest in the HoB is thus rated as ‘very good’.
Montane forest
Montane forest (Table 1) historically occurred over c. 18% of the HoB. As altitude increases, the forest still contains dipterocarps, but only at certain terrain types. Compared to altitudes below 800 m, the height of the canopy is reduced to c. 25 m on average, tree buttresses are small, uncommon, or both, cauliflory is rare, large lianas are usually absent and epiphytes are more common than in lowland and upland forests. These areas have a very high rate of local endemism for animal and plant species (Stattersfield et al. Reference Stattersfield, Crosby, Long and Wege1998; Wikramanayake et al. Reference Wikramanayake, Dinerstein, Loucks, Olson, Morrison, Lamoreux, McKnight and Hedao2002). Extent of the upland forest was considered the most important and practical critical indicator, and forest condition measurement is technically feasible but impractical at this time (Table 2).
Nearly all of Borneo's montane forest ecosystems are located within the HoB, and the remaining 89% of the historic montane forests are generally still in primary condition, rated ‘very good’.
Heath forest
Heath forest, known as kerangas in Borneo, is typically found on raised sandstone terraces near the coast and sandstone plateaus inland; the heath soils are well-drained, acidic and coarse, with a superficial covering of organic material (MacKinnon et al. Reference MacKinnon, Hatta, Halim and Mangalik1996). Heath forests have distinct structure and vegetation, with shorter and smaller trees than typical rainforest (Kartawinata Reference Kartawinata, Singh and Gopal1978; Whitmore Reference Whitmore1984). Heath forest historically occupied c. 2% of the HoB. Spread throughout the HoB are many small (< 500 ha) areas of heath forest, often with unique species compositions. Several areas have been burned in the last 25 years and, due to the poor soils, recovery and restoration of the forests is extremely difficult. A viable long-term extent of heath forest within the HoB is proposed to be c. 60% of its historic areal extent. Since existing heath forest in the HoB is 48% of its historic coverage, this indicator was rated as ’fair’.
Limestone forest
Limestone forests (Table 1) historically occupied < 0.5% of the HoB. The unique soil characteristics result in what appear to be unique floral compositions (Anderson Reference Anderson1965). There are relatively few tree species, but overall plant species richness is high, with high levels of plant endemism (MacKinnon et al. Reference MacKinnon, Hatta, Halim and Mangalik1996). Limestone forest tends to be dry due to excellent soil drainage, and is therefore susceptible to fire. It does not recover quickly from burning and large-scale disturbance (Anderson Reference Anderson1965).There is not a distinctive vertebrate fauna associated with this forest, which typically occurs as blocks within a larger expanse of rainforest, the same vertebrates being found in both. However, there appear to be distinctive invertebrate faunas. Within the HoB, 79% of historic limestone forest areas remain, resulting in a ‘good’ rating. Degradation of limestone ecosystems is usually irreversible (Whitmore Reference Whitmore1984).
Peat swamp forest
Peat swamp habitats historically occupied 8–12% of the island of Borneo (MacKinnon & Artha Reference MacKinnon and Artha1981) and 3% of the HoB. They are generally found in low-lying areas and poorly-drained high-altitude depressions near coasts, rivers and/or lakes. Lowland peat swamps are rain fed, domed and not generally subject to flooding. The peat itself is usually > 50 cm thick and may be up to 20 m thick. The soil and the water emerging from it are acidic and poor in nutrients, especially calcium. A restricted, though relatively impoverished, flora is indicative of peat swamp forests (Anderson Reference Anderson1963; Brunig Reference Brunig1973). Most peat swamps exhibit a marked zonation, with greater tree species richness and canopy height toward the outside of the swamp (Anderson Reference Anderson1963; Whitmore Reference Whitmore1984).Since peat swamp ecosystems occur mainly in the wide coastal lowlands, many of those areas have either been converted or are highly disturbed. Only 2.1% of the peat swamp forests in Kalimantan are regarded as pristine (Miettinen & Liew Reference Miettinen and Liew2010). Interior peat swamp areas within the HoB are generally not highly disturbed, and 72% of historic peat swamp within the HoB is still present, resulting in a rating of ‘good’.
River ecosystems
Rivers are heavily used by local people and industry for transportation and food. They are considered to be rich in species (Abell et al. Reference Abell, Thieme, Sabah Perez and Petry2008), yet there is very little information available about the in-stream conservation status. Nearly all of the freshwater swamp forests within the HoB occur in the riparian zones. The focus for indicators was on the watersheds and riparian zones of the major rivers, where the selected indicator was the percentage of relatively intact forest in the watershed of each of four major rivers: the Kapuas, Barito, Mahakam and Kayan rivers. The southern and western lowland portions are generally converted, while in the northern part conditions are somewhat better; the overall rating for this indicator was ‘fair’.
Aspects of the following species were selected as indicators because they are either keystone species or are sensitive indicators of human activities. The keystone species add minimum forest habitat block size, associated species and connectivity considerations, and the sensitive indicators add threat considerations.
Rhinoceros
The only locations where there is still an opportunity for survival of the Borneo Rhinoceros are in a few protected areas in Sabah (Malaysia), partly outside of the HoB. The occasional records from other areas in Borneo concern most likely only single individuals and no viable populations. The World Wide Fund for Nature (WWF) in Sabah keeps records of all rhinoceros sightings and is currently compiling updated documentation.
The population size of Borneo rhinoceros was chosen as an indicator because this large-bodied species is slow to reproduce and is a sensitive indicator to overhunting and disturbance. The population of Sumatran rhinoceros in Borneo is believed to total < 40 individuals (Zafir et al. Reference Zafir, Payne, Mohamed, Lau, Sharma, Alfred, Williams, Nathan, Ramono and Clements2011), and the current rating for this species is ‘poor’. In fact, the proposed future rating for this species is only ‘fair’, since that is estimated to be the best long-term scenario, requiring constant management to maintain rhinoceros in the HoB.
Bornean elephants
There is some question as to whether the elephants on Borneo are indigenous, though they are likely to be a relict subspecies (Cranbrook et al. Reference Cranbrook, Payne and Leh2008). Elephants’ main habitat in Borneo is dipterocarp forest; they are considered ecological engineers, and thus a keystone species (Wright & Jones Reference Wright and Jones2006).
The elephants of Borneo, which are mostly living in Sabah, have suffered from habitat conversion over the last decades. Population size, distribution and connectivity are the major indicators for measurement. The size of the Bornean elephant population was chosen as an indicator. The ability of elephants to live side by side with humans is also an indication of man's ability to accommodate them through various conservation activities. The total population size remains relatively stable, since elephant herds move from destroyed habitats into intact habitats, but these are becoming overpopulated (Alfred et al. Reference Alfred, Ambu, Nathan and Goossens2011), and the rating for population size is ‘fair’.
Due to their similar ecological roles and use of habitat, and uneven distributions, the expert team grouped elephants and banteng together for indicator development as forest edge herbivores. Banteng are found in secondary rainforests and sub-humid forests on Borneo (Hedges & Meijaard Reference Hedges and Meijaard1999). However, there is limited information available about the status and distribution of banteng in Borneo.
Orang-utan
Historically, the Bornean orang-utan occupied most lowland peat swamp and the lower upland ecosystems across Borneo. The population has strongly decreased particularly because many of their habitats were converted in the last decades (Payne & Prudente Reference Payne and Prudente2008). While orang-utans can survive in degraded forest, it is not clear if these are sink populations destined to decline, and where extensive deforestation has occurred orang-utans may raid human crops and consequently pay with their lives (Meijaard et al. Reference Meijaard, Sheil, Nasi, Augeri, Rosenbaum, Iskandar, Setyawati, Lammertink, Rachmatika, Wong, Soehartono, Stanley and O'Brien2005). Their sensitivity and conspicuousness makes them a good indicator of forested landscape conditions. Hunting for the pet trade is also an issue. The majority of the orang-utan populations live outside the HoB boundaries, however the HoB also harbours a number of large populations, mainly in lowland and peat swamp forests in West and Central Kalimantan and in Sabah. The populations of orang-utans in the HoB are relatively stable, and lie within protected areas. The densities in lowland and in peat swamp forests of the HoB are rated as ‘fair’. The long-term scenario is to improve the rating for orang-utan populations in peat swamp forests to ‘good’, which could be achieved through the many conservation efforts in the Danau Sentarum area.
Bornean clouded leopard
The forest-dependent Bornean clouded leopard (Neofolis diardi diardi) is one of the top predators of Borneo and the largest cat species of the island. This animal likely plays a major role in regulating healthy populations of monkeys, ungulates and smaller mammals. Clouded leopards are very hard to observe or even catch on camera traps, but are probably present in low densities in most of the remaining large forested areas of Borneo.
Observations of clouded leopards are made in logged secondary forest, but it is unclear if this is a satisfactory habitat or not (Wilting et al. Reference Wilting, Fischer, Bakar and Linsenmair2006). A viable population of clouded leopards of at least 50 individuals would need a minimum of between 350 and 625 km2 of uninterrupted forest blocks (Wilting et al. Reference Wilting, Fischer, Bakar and Linsenmair2006). Clouded leopards were considered to be a useful indicator of an area-sensitive species in the HoB. Based on the availability of forest areas that meet the minimum size criteria, identified during the GIS analysis, the viability rating for viable populations of this species is therefore considered ‘good’.
Endemic Nepenthaceae
It was decided to develop an indicator for the suite of rare and endangered endemic pitcher plants (Nepenthaceae), most of which are limited to a particular and often fragile habitat (Clarke Reference Clarke1997). The presence of these plants in historic areas was chosen as an indicator because the areas where they are found are known, and because these species are vulnerable to habitat disturbances and wildlife trade, and are indicative of other sensitive plant groups in the HoB. The relatively high level of attention to these rare species provides some form of protection, and the estimated rating for endemic Nepenthaceae, based on 80% of historical locations being viable, is ‘good’.
Threat and conservation management indicators were discussed as a way to supplement the ecological indicators by providing a more complete picture of the trends in conservation threats.
Direct threats to natural systems in the HoB
The top threats for the decade 2007–2017 were considered to be: industrial conversion of natural forests; illegal logging; legal commercial but unsustainable rates of timber extraction; forest fire; mining; and overhunting and collecting.
The highest ranking threat to the integrity of the HoB is industrial conversion of natural forests. Conversion is mainly occurring to establish oil palm plantations and pulp wood plantations. There are 770 000 ha of active concessions within the HoB in Sabah (Malaysia) and 830 000 ha of active and newly allocated concessions within the HoB in Kalimantan (WWF 2011). The most affected ecosystem is the lowland rainforest. Most locations are at the edges of the HoB, though some inland areas of the HoB in West Kalimantan are also allocated. The experts rated the threat of industrial conversion of natural forests from 2007 through 2017 as ‘very high’ and the expert ranking for illegal logging from 2007 to 2017 was ‘high’.
Forest fires are also considered a major threat. These devastating events were found to occur mainly at the edges of the HoB and the main ecosystems affected were heath forests, lowland forests and peat swamp forests. Forest fire danger from 2007 to 2017 was rated as ‘very high’ by the experts.
Mining is a threat in Kalimantan that has been somewhat underestimated so far, probably because of lack of data. The major type of mining in the Hob is for coal, which is always accomplished through open pit mining. The best data available indicates that there are more than 1 100 000 ha of coal concessions within the HoB, of which 980 000 ha are in the research or exploration phase, indicating potential future growth in the impact of mining (WWF 2011). The expert rating for the threat of mining in the HoB for 2007–2017 was ‘high’.
There are very few data on the level of poaching and overharvest. The experts’ opinion was that this threat deserved a rating of ‘high’ for 2007–2017.
Note that, because of the relatively short time window of ten years to 2017, climate change was not analysed as one of the current top threats.
The overall threat level for the HoB, combining all of the individual threat ratings (including the minor threats, which are not discussed here), is ‘medium’. This level indication was derived using the Miradi-based threat rating analysis (Appendix 2, see supplementary material at Journals.cambridge.org/ENC). Major threats of industrial forest conversion and mining currently currently exist mainly at the edges of the HoB, but are likely to expand further inland without conservation management intervention.
Conservation management in the HoB
With the possible exception of montane forest, representation of the HoB's habitats (Table 3) is generally lower than recommended by conservation science (Soule & Sanjayan Reference Soule and Sanjayan1998). A comprehensive survey of the management effectiveness of the HoB's protected areas has yet to be carried out, but is a very high priority.
Protected area representation of each ecosystem.
DISCUSSION
Of the initial 29, later revised to 23, biological indicators identified by the expert team, sufficient information could be gathered for 17 indicators. Some of the unfulfilled indicators may be assessed within the next few years, but others may have to wait until suitable methodologies and resources can be developed. Several indicators remain mere ideas about what needs to be known (Table 2), and it is hoped that practical methods and resources will emerge to capture that information. These metrics will undoubtedly improve in the next few years.
The results for the individual indicators of the ecological health of the HoB in 2007–2008 were mixed (Table 2). Broadly speaking, the ecological systems are doing well, with some exceptions. Ratings were ‘fair’ to ‘very good’ within the HoB. In the HoB, montane and upland forest ecosystems were doing well. Lowland forest and peat swamp forest were also doing well (though this was not the case in Borneo as a whole). Heath forest was only ‘fair’, and there was little prospect for improvement given the depleted soils that form the substrate for this forest type; this highlights a need for targeted protection of heath forest within the HoB. The natural forest cover in the major river watersheds is also cause for some concern, and efforts to restore cover and connectivity of riparian forest would seem to be a priority. Development of a more comprehensive assessment of the state of the HoB's freshwater biodiversity is another priority.
Among the key species, elephant populations were assessed as ‘fair’ to ‘good’. Efforts to maintain and restore connectivity for sub-populations continue. Rhinoceros populations were assessed as ‘poor’, and the prospects are not encouraging. However, a recent proposal (Zafir et al. Reference Zafir, Payne, Mohamed, Lau, Sharma, Alfred, Williams, Nathan, Ramono and Clements2011) suggested bringing all remaining rhinoceros to Tabin Wildlife Reserve where their interaction and, hopefully, breeding could be protected. More comprehensive information on other key species, such as orang-utans, banteng and clouded leopard, would be useful.
As already stated, the extent of forest cover in the HoB is severely threatened by several factors unless sustainable land-use policies are implemented. The Indonesian Ministry of Forestry estimates that Indonesia has been losing between 1.6–2.8 million ha yr−1 in recent years due to illegal logging and land conversion (United Nations Office on Drugs and Crime 2011).
For protected area representation, the oft-cited policy-driven goal is 10% of any country should be subject to protection (IUCN 1993; World Commission on Environment and Development 1987). However, conservation science indicates that larger percentages (> 30%) (Margules et al. Reference Margules, Nicholsis and Pressey1988; Saettersdal & Birks Reference Saettersdal and Birks1993; Soule & Sanjayan Reference Soule and Sanjayan1998; Svancara et al. Reference Svancara, Brannon, Scott, Groves, Noss and Pressey2005) are increasingly likely to be required to support biodiversity, ecosystem services and buffer against changing climate. The protected areas representation of most of the major ecosystems in the HoB should be enlarged in order to obtain these goals, as only three ecosystems have more than 10% of the areas protected.
CONCLUSIONS
So, are the data for the indicators sufficient to provide an accurate picture of the ecological health of the HoB? The overall biodiversity rating for the HoB is ‘good’. We believe that this information, along with supplemental information on threats and conservation management, is a fair assessment of the conservation status of the HoB. Better information is required on a number of aspects of the HoB's ecology, and further development will lead to continued improvements in the dataset. These initials indicators provide a sufficient baseline for tracking the future ecological health of the HoB. It should also be noted that the ecological health of Borneo as a whole is not so positive, given the extensive forest conversion that has already occurred in the lowlands (Langner et al. Reference Langner, Miettinen and Siegert2007).
Further effort should concentrate on periodical (for example every 2–4 years) assessment of the existing indicators and generating better information for some of the indicators, the priorities being: forest conditions; freshwater biodiversity indices by major watershed/river; bearded pigs densities over time; clouded leopard densities; banteng densities; key plant families or groups; the location and condition of very small areas of heath forest; and the location and condition of very small areas of limestone forest.
Despite the encouraging transboundary HoB Declaration, the pressure on the remaining forests and associated species of the HoB has never been greater. The good news is that the governments of the HoB have pledged to protect and restore the natural resources of the HoB for the benefit of local, national and even global constituencies. Their bold commitments, supported by other governments, multilateral agencies and non-governmental organizations, provide great promise that a balanced configuration of land use will eventually protect and maintain natural systems and the ecosystem services that they provide. We hope that our development and assembly of baseline conservation indicators and data will assist in the tracking of the ultimate conservation of this globally important area.
ACKNOWLEDGEMENTS
We thank the Sall Foundation (for funding the initial HoB measures work), A. Shapiro (for GIS-based analyses and mapping), the late Mrs H. Tobing (for the overall coordination of data collection and analysis in Indonesia), Ian Kosasih and A.Tomasek (for their overall support), J. Moore, A. Salim and T. Barano (for Borneo-wide environmental assessment and sharing baseline data), L. Curran, W. Giesen, G. Limberg, J. Payne, E. Meijaard, D. Sheill, R. Stuebing, A.Tjiu (the experts who provided invaluable input and without whom this study would not have been possible), and colleagues working in Kalimantan (data collection and good times in remote locations).
