Assessment of potential reduction in likelihood of extinctions for Bukit Tigapuluh Sustainable Landscape and Livelihoods Project

FINAL REPORT

10 October 2018

Frank Hawkins, Craig Beatty, Marcelo Tognelli

IUCN-US

Washington DC

Species Threat Abatement and Restoration Metric

Assessment of potential reduction in likelihood of species extinctions for Bukit Tigapuluh Sustainable Landscape and Livelihoods Project

FINAL REPORT 10 October 2018

Aim: to use the existing approach and methodology for STAR to calculate an initial ex-ante value for the project area of BTSLLP, and to make management recommendations based on the result of that analysis.

Tasks and results:

Identify likely threatened taxa present in the general project area based on global species distribution maps and existing lists

This analysis utilizes the methods of the Species Threat Abatement and Restoration Metric (STAR) to deliver quantified estimates of the contribution that investment could make to the reductions in extinction pressures for taxa listed as ‘Threatened’ in the IUCN Red List of Threatened Species.

This analysis was predicated on the spatial extent (including a 5km buffer) of a landscape concession in proximity to Bukit Tigapuluh National Park in eastern Sumatra, .

The following 25 mammal and species, classified as ‘Near-threatened (NT)’, ‘Vulnerable (VU)’, ‘Endangered (EN)’, and ‘Critically Endangered (CR)’, whose ranges overlap with the spatial extent of the concession area, were included in this analysis.

Bonaparte’s Nightjar Sumatran Elephant Elephas Greater Slow Loris Nycticebus Caprimulgus concretus VU maximus ssp. sumatranus CR coucang VU

Greater Green Sun Bear Helarctos malayanus Whiskered Flying Squirrel Chloropsis sonnerati VU VU Petinomys genibarbis VU

Large-billed Blue-flycatcher Orbiculus Leaf-nosed Bat Temminck’s Flying Squirrel Cyornis caerulatus VU Hipposideros orbiculus VU Petinomys setosus VU

Black Partridge Melanoperdix Agile Gibbon Hylobates agilis Sumatran Surili Presbytis niger VU EN melalophos EN

Wallace’s Hawk-eagle Nisaetus Smooth-coated Otter Lutrogale Flat-headed Cat Prionailurus nanus VU perspicillata VU planiceps EN

Helmeted Hornbill Rhinoplax Manis javanica CR Bearded Pig Sus barbatus vigil CR VU Rajah Sundaic Maxomys Large Green-pigeon Treron Maxomys rajah VU Siamang Symphalangus capellei VU syndactylus EN Whitehead’s Sundaic Maxomys Brooks’s Dyak Fruit-bat Maxomys whiteheadi VU Malay Tapir Tapirus indicus Dyacopterus brooksi VU EN Marbled cat Neofelis diardi VU

Fig 1. Map of Sumatra showing location of concession

Species data on reptiles and amphibians from the IUCN Red List of Species do not indicate the presence of any threatened species in these taxa in the project area.

We have not identified Sumatran Orang-utan Pongo abelii on this list as the species is the subject of an introduction program.

Identify the threats that apply to these species

Threats that are listed in the IUCN Red List of Species that most commonly affect the species listed above include loss of due to industrial and subsistence farming, hunting, and the construction of large dams. A considerable amount of effort has been devoted to obtaining additional information on these threats. See appendix 4 for more details on data sets collected. In the context of this project, it is clear that the most important potential threats that apply to the threatened species present in the project area are hunting and loss of habitat. Our management recommendations are based on this assessment.

Verify the presence and distribution of species in project area in collaboration with project staff and advisors/ contractors

It has been difficult to establish the presence of threatened species in the project area beyond the major focal species for which surveys have been conducted, including tigers and Sumatran elephants. Camera trap data have been requested from the Frankfurt Zoological Society, but permission to use these data was refused. In addition there is no data on the distribution in the project area for many of the more threatened smaller mammals and . While recent analyses suggest that there are no threatened reptiles and amphibians in the project area, it is very likely that targeted surveys would reveal species of conservation concern.

Evaluate existing data on geographical disposition of threats, using a buffer 5 Km around the project area

Fig. 2 Map of the Bukit Tigapuluh Concession Area showing 5 km buffer, Bukit Tigapuluh National Park, and in time periods 2001-2005, 2006-2010, and 2011-2016.

Figure 2 shows the distribution of loss across the project area in three different time periods. We included a 5 km buffer around the project area in this analysis for two reasons: firstly, if management action reduced threats in the project area (for instance forest clearance) it might just displace the threat to an area adjacent, and secondly, this buffer covers the likely area that people living within the project area could make use of (for instance for hunting) on a daily basis.

The threat map shows that clearance of primary forest in areas now covered by occurred in the project area including in areas adjacent to the National Park very recently (2011-2016). Whether this forest clearance was specifically for the of rubber, or for subsistence agriculture is not known. However this threat to biodiversity is clearly something that has had very significant impacts on the conservation value of in the region in the very recent past, and will obviously need to be monitored and controlled very carefully if further biodiversity loss is not to happen. In particular It is clear that the potential for loss of habitat important for biodiversity is high in the 5 km buffer zone around the project area (see Figure 2).

It was very difficult to gather data on the spatial distribution of threats other than for forest cover change across the project area. For threats related to hunting and non-sustainable use, we considered developing a proxy map that assumes hunting pressure is correlated to the presence of human habitation and roads. However the data layers we obtained for this purpose are mostly inaccurate and outdated. For the purposes of management recommendations, we have therefore made the assumption that forest cover change and intensity of hunting are closely correlated, and that the forest cover change layer in Figure 2 represents a threat intensity map for both deforestation and hunting.

These data have been integrated into a project GIS system maintained by Marcelo Tognelli of the Joint IUCN/CI Biodiversity Assessment Unit based in Washington DC.

In collaboration with the project team, acquire project GIS polygons and land- use/conservation proposals for each

We did not have access to detailed land-use or conservation proposals for polygons within the project area, except that Blok 1 and Blok 2 are prioritised for conservation and restoration activities.

Map threats against land-use and conservation proposals for each project polygon

In the absence of detailed management proposals for each of the landuse polygons in the project area, we calculated the ex-ante STAR value by taxonomic group, species and threat, which can then be used to orient management for conservation within the overall project area.

We used the STAR approach and methodology (Appendix 3) to assess the relative change in species extinction risk that could be generated by reducing the threats to the species present in the project area. To do this, forest cover data from Global Forest Watch were used to calculate the Area of Habitat (AOH) for the threatened species listed above. AOH is the habitat available to a species within its range, as defined in the Red List of Species. The proportion of the total Area of habitat of each species present in the project area is an approximation of the proportion of the total species’ population found in the project area, and is important for the calculation of STAR. We then took the Scope and Severity Scores for each pressure (also from the Red List of Species) as it applied to each species to calculate the potential STAR yield for each pressure.

The methodology to calculate STAR is shown in Appendix 3. The tables showing the raw STAR scores for first and second order threats, broken down by mammals and birds, are shown in Appendix 1 and 2. These results are shown graphically in Figure 3 and 4 below. Bukit Tigapuluh Concession; threatened species contribution to STAR scores

Figure 3 Relative contribution of conservation of individual species to potential reduction in species extinction risk.

Figure 3 shows the relative change in global species extinction risk that could be delivered by focusing on individual species. The amount of species extinction risk reduction possible is equivalent to the proportion of the species’ habitat that is present in the project area, weighted by the threatened status of the species (see methodology, appendix 3). Hence highly threatened species with small global habitat area, of which the project area is a significant fraction (for instance Hipposideros orbiculus), are those that will generate the greatest STAR value. A value of 5 would be equivalent to removing all the threats from the entire range of a Critically Endangered species, making it Least Concern. A STAR value of 1 would approximate to reducing the threat level of a particular species from (for example) Critically Endangered to Endangered, or Endangered to Vulnerable, by reducing the threats that cause the species to have that threat rating.

The top 7 species in this list are all mammals, reflecting the higher vulnerability of mammals to hunting (larger species; Sumatran Elephant, Sumatran Surili, Agile Gibbon, Malay Tapir, Siamang) and their generally smaller ranges (smaller species; Orbiculus Leaf-nosed Bat, Brooks’s Dyak Fruit Bat). Other smaller mammals and most birds are more vulnerable to habitat loss. Bukit Tigapuluh Concession: Species threats contribution to STAR Scores

Figure 4. Relative contribution of threats to potential reduction of species extinction risk. Each higher-level category of threat is grouped by colour, according to the key under the title.

Figure 4 shows the relative change in species extinction risk that could be generated by addressing pressures that cause species to become threatened. The pressures that are making these species at risk of extinction apply across the species’ range, and at this point in the analysis (the ex-ante STAR calculation) we do not know if these pressures apply at the scale of the project area.

The size of each box shows the relative impact that reducing this pressure to zero at the scale of the project area would generate on global species extinction risk.

Figure 4 shows that the maximum reduction in extinction risk that could be generated by investment in the project area will be from eliminating further primary forest loss caused by industrial or subsistence agriculture, and by eliminating hunting, especially of mammals.

This analysis demonstrates both the main pressures and threats for most of the species of concern in this region and the extent to which interventions directed at alleviating these threats within the concession area and its 5km buffer area will contribute to a quantifiable global reduction in the threats to these species as a group. The values for each pressure and each species are expressed as proportions of the total extinction risk that could be avoided by reducing pressures across the whole of all the threatened species’ ranges. Details of these analyses are presented in Appendices 1 and 2.

For instance, interventions focused on addressing the threats posed by agro-industrial farming in this area have the potential to contribute a 1.2% reduction in the extinction risk of this group of mammals and birds (1.19% reduction attributed to mammals and 0.01% attributable to birds). These figures can be compared with global contributions to the reduction in threats to these taxonomic groups as a whole once the global data are available.

Develop set of recommendations for project implementers on maximizing STAR through modification of land use/conservation plans

We focus recommendations from two perspectives- firstly geographical, and then species/pressure.

Geographical focus

Overlaying the Area of habitat maps for all threatened species combined with the project area landuse types enables us to propose some management based on spatial priorities.

Figure 5: Concentration of Area of habitat for threatened species across the project area, and areas of focus for further research and management, in relation to conservation areas (Blok 1 and Blok 2).

In general the areas of maximum value for threatened species shown in Figure 5 (the darkest green areas) overlap the parts of the project area where conservation has been prioritised (Blok 1 and 2: combined AOH value is not shown for Blok 1) or are already conserved in Bukit Tigapuluh National Park. However there are five additional areas, labelled on the map 1-5, that have not been identified, as far as we know, for further conservation action. Area 1 is of significant importance as it is the only part of the project area that overlaps with the locally distributed and threatened Hipposideros orbiculus, the species for which the project area can contribute the maximum reduction in species extinction risk. Area 2 consist of generally intact primary forest in the 5 km buffer of the project area, in a zone where there is little other primary forest. It seems likely that this area will be vulnerable to deforestation and pressure from hunting if no action is taken, and merits a comprehensive biodiversity assessment. Area 3 is a small remnant of primary habitat in the middle of the plantation that might still retain populations of important reptiles and amphibians. Area 4 contains what appears to be riparian habitat where relict populations of amphibians might occur, and the formation may be of importance in ensuring connectivity between other areas of habitat or protecting waterbodies from sedimentation. Area 5 is a block of apparently intact primary forest outside the conservation area but within the concession and adjacent to the National Park that would seem worth further species assessment and subsequent protecting.

Species/Pressure perspective

Given the amount of time we have spent obtaining data, it has not been possible to assess the comparative status of a counterfactual site nearby, in order that the project can demonstrate progress in reducing threats against a control.

The fact that there are many threatened species of bird and mammal whose ranges overlap with the project area, but have not yet been confirmed, means that the current estimation of ex-ante STAR is hypothetical. The first recommendation for the project is therefore that targeted surveys for the threatened species that have not yet been confirmed at the project site should be scheduled.

The species that have so far been confirmed (elephant, tiger, tapir ) are mostly species with very large global ranges, for which improvement of management leading to reduction of threats will have small impacts on their likelihood of global extinction. Species that have smaller ranges will have a larger proportion of their global range covered by the project area. Investment at the project level to reduce risks of extinction of these species will therefore offer a greater STAR return on investment.

Figure 3 shows that focusing on species for which the project area holds a significant proportion of their global AOH will allow the project to generate the maximum reduction of species extinction risk. We have already identified a particular priority in Hipposideros orbiculus, a under geographical perspectives, above. Most of the other species for which global extinction risk can be significantly reduced are larger- bodied mammals, including Elaphas maximus sumatranus (treated here as a species, although in fact only a sub-species of the very widespread Asian Elephant), the Sumatran Sirili Presbytis melalophus, two gibbons (Hylobates agilis and Symphalangus syndactylus) and the Malay Tapir Tapirus indicus. A very important priority for the project to be able to deliver on the ex-ante value of STAR in reducing global species extinction risk is to assess whether these species do occur in the project area or in the 5 km buffer, and conduct assessments of the population sizes and threats that weigh on them.

From a threat perspective, the most important potential threat to larger vertebrates (mammals and birds) in Bukit Tigapuluh, is hunting and unsustainable use. In the absence of this threat, it is unlikely that many of the species listed as threatened from the site would be completely eliminated by the degradation of forest habitat, at least in the short term. According to project personnel, the management strategy of the concession operator is to completely eliminate primary forest clearance and hunting in the project area, and to promote forest restoration in the conservation areas. If this management approach is successful, measured against a baseline and compared with a counterfactual site elsewhere in the region, then the total STAR ex-ante value calculated here should be attained.

While there are species that have been identified in the project area that have small ranges, most of the smaller-bodied that could contribute significant to the site-level STAR value are not yet assessed in detail. These include reptiles, amphibians and frAoHwater fish, as well as various taxa of plant. Surveys to assess the presence of these species across the concession will likely result in the discovery of many other species with small ranges in the vicinity, some of which will have an elevated risk of extinction. Focusing on these species will permit a much more significant reduction in extinction risk.

A similar potential future increase in STAR return will be possible if greater investment is made in those species that have higher levels of risk of extinction, for instance Manis javanica (Sunda Pangolin: Critically Endangered) which has been eliminated by hunting for its scales throughout most of its historical range, and for which Sumatra may be a remaining population center. A similar situation applies to the Helmeted Hornbill, Rhinoplax vigil, (CR) which while widespread, is under very heavy and increasing pressure from hunting.

Next steps

If the BTLSSP would like to measure its ex-post contribution to reduction of extinction risk, the following steps will be necessary:

1. Confirmation of the presence of the 25 species of threatened mammal and bird whose AOH overlaps with the project area, with priority given to the species with the largest STAR scores in Figure 3 2. Data on presence of small-bodied animals, especially reptiles and amphibians, and potentially important plant taxa (e.g. palms) in the project area, and threats to these species 3. Recalculation of ex-ante STAR contribution possible by managing threats to all threatened species, including those above 4. Mapping of key threats across project area, especially related to hunting, which is hard to detect and monitor, but also including potential new threats identified in step 1 above 5. Estimation of yearly reduction in extent and severity of threats that could be accomplished by conservation investment across the project area- for instance percentage annual reduction of forest loss (or increase in forest area from restoration) 6. Baseline of STAR values calculated across the project area and broken out by land-use unit within the project area 7. Monitoring of evolution of pressures as a result of conservation investment in each land-use unit 8. Calculation of ex-post reduction of risk of species extinction delivered through pressure reduction

Appendix 1. Species Threat Abatement and Restoration Metric: First order species pressures/threats

Birds and Mammals: Birds: Mammals: Sum of return on Sum of return on Sum of the investment ex- investment ex- return on ante in the actual ante in the actual investment ex- area of habitat in area of habitat in ante in the actual 2017 2017 area of habitat in First Order Threats 2017 Agricultural & forestry effluents 0.000047 0.000047 Annual & perennial non- timber crops 0.020853 0.000046 0.020899 Commercial & industrial areas 0.000084 0.000084 Dams & water management/use 0.000217 0.000217 Domestic & urban waste water 0.000031 0.000031 Droughts 0.000002 0.000002 Fire & fire suppression 0.003810 0.000020 0.003830 Fishing & harvesting aquatic resources 0.000040 0.000040 Garbage & solid waste 0.000007 0.000007 Habitat shifting & alteration 0.000054 0.000054 Housing & urban areas 0.004652 0.004652 Hunting & trapping terrestrial animals 0.006492 0.000028 0.006520 Industrial & military effluents 0.000023 0.000023 Livestock farming & ranching 0.000077 0.000077 Logging & wood harvesting 0.005147 0.000049 0.005196 Marine & frAoHwater aquaculture 0.000023 0.000023 Mining & quarrying 0.000095 0.000002 0.000097 Oil & gas drilling 0.000010 0.000010 Problematic native species/diseases 0.000020 0.000020 Recreational activities 0.000084 0.000084 Roads & railroads 0.000089 0.000089 War, civil unrest & military exercises 0.000101 0.000101 Wood & pulp plantations 0.002513 0.000003 0.002516 Grand Total 0.044470 0.000150 0.044620

Appendix 2. Species Threat Abatement and Restoration Metric: Second order species pressures/threats

Birds and Mammals: Birds: Mammals: Sum of Sum of Sum of the return on return on return on investment investment investment ex-ante in ex-ante in ex-ante in the actual the actual the actual area of area of area of habitat in habitat in habitat in 2017 2017 Second Order Threats mammals 2017 War, civil unrest & military exercises 0.000101 0.000101 Unintentional effects: (subsistence/small scale) [harvest] 0.000021 0.000028 0.000049 Unintentional effects: (large scale) [harvest] 0.000382 0.000020 0.000402 Unintentional effects (species is not the target) 0.000599 0.000599 Type Unknown/Unrecorded 0.000054 0.000054 Trend Unknown/Unrecorded 0.003810 0.000006 0.003816 Soil erosion, sedimentation 0.000023 0.000023 Small-holder plantations 0.000278 0.000278 Small-holder grazing, ranching or farming 0.000077 0.000077 Small-holder farming 0.002904 0.000008 0.002913 Shifting agriculture 0.006036 0.000010 0.006046 Sewage 0.000023 0.000023 Scale Unknown/Unrecorded 0.000329 0.000004 0.000333 Roads & railroads 0.000089 0.000089 Recreational activities 0.000084 0.000084 Problematic native species/diseases 0.000020 0.000020 Persecution/control 0.001417 0.000002 0.001419 Oil & gas drilling 0.000010 0.000010 Motivation Unknown/Unrecorded 0.004727 0.004727 Mining & quarrying 0.000095 0.000002 0.000097 Large dams 0.000017 0.000017 Intentional use: (subsistence/small scale) [harvest] 0.000021 0.000021 Intentional use: (large scale) [harvest] 0.000021 0.000021 Intentional use (species is the target) 0.004493 0.000026 0.004519 Increase in fire frequency/intensity 0.000014 0.000014 Housing & urban areas 0.004652 0.004652 Habitat shifting & alteration 0.000054 0.000054 Garbage & solid waste 0.000007 0.000007 Droughts 0.000002 0.000002 Dams (size unknown) 0.000184 0.000184 Commercial & industrial areas 0.000084 0.000084 Agro-industry plantations 0.001930 0.000003 0.001932 Agro-industry farming 0.011913 0.000023 0.011936 Abstraction of ground water (unknown use) 0.000016 0.000016 Grand Total 0.044470 0.000150 0.044620

Appendix 3: STAR methodology

A pressure-based metric for Species Threat Abatement and Restoration The pressure-based metric has three forms:

1. Comparative ex-ante STAR. Large-scale assessment of geographic variation in the ex-ante return on investment, i.e. the potential for intervention to achieve reduction in overall pressure intensity and hence in extinction risk. For a particular site or grid square, this potential depends on the number of threatened species present, the category of extinction risk for each one, and the proportion of each species’ global population present in the site or grid square. The potential can be partitioned out using information on the relative significance (from scope and severity scoring) of each pressure for the species concerned. This produces a probabilistic map of ex-ante STAR, as it is not known whether or not a particular threat impacts a particular species at a particular site or grid square– but this is more likely for threats that apply to a large percentage of the population. This analysis can be automated for birds but cannot be carried out for other taxonomic groups yet without improvement in the information base (i.e. more complete scoring of threat scope and severity).

2. Site-focused ex-ante STAR. Focused site-scale assessment of the ex-ante return on investment, i.e. the potential for intervention to achieve reduction in overall threat intensity and hence in extinction risk. This potential depends on the number of threatened species present, the category of extinction risk for each one, and the proportion of each species’ global population present at the site. For specific pressures, the potential depends on the severity of the threat to each species and whether the site is within its scope. This analysis requires some site-specific information on the pressures present, so cannot at the moment be fully automated1. For taxa other than birds, it would also require scoring of threat severity for the pressures present.

3. Site-focused post-ante STAR. Focused site-scale assessment of the ex-post return on investment, i.e. the reduction in overall threat intensity achieved over time by addressing one or more specific pressures at the site. The reduction depends on the number of threatened species present, the category of extinction risk for each one, the proportion of each species’ population present, the severity of the pressures addressed for each species and the effectiveness of addressing those pressures. This analysis requires development of a counterfactual scenario (what would happen without intervention) and assessment of the before and after levels of targeted pressures against this counterfactual. The metric:

1 This might become possible in future, with improved comprehensiveness and documentation of site-level threats for Key Biodiversity Areas  Applies to geographic areas, either across a large scale (providing a map of ex-ante STAR) or to particular sites  Allows comparison of return on investment across sites or grid squares, and across different pressures  Allows prediction of return on investment for interventions targeted at particular threats, at particular sites  Allows assessment of achieved return on investment for interventions targeted at particular threats, at particular sites  Depends on apportioning the relative contribution of pressures to a species’ extinction risk, across the species’ range  Employs the same approach for site-focused ex-ante and ex-post measures. However, an extra step is needed to calculate ex-post measures, which involves the actual reduction in threat magnitude against predicted levels. This requires geographically-focused measures of threat magnitude, usually including a time series.

At present, the detailed threat scores needed to calculate the ex-ante metric consistently and rapidly are only documented for one major fully assessed group – birds. Birds are generally a good indicator group for overall biodiversity importance (BirdLife International 2013 – SOWB) but are far from perfect, and have overall lower levels of threat than, for example, mammals or amphibians (IUCN 2016). Birds are also usually better indicators for the terrestrial than for the marine realm. The supporting data for Red List assessments is continuously being improved, however, and it is expected that scores for scope and severity will be completed for other fully assessed groups in the near future (e.g. for mammals by 2018).

It is possible to apply the approach to other taxa even without fully documented threat scores, but this requires input of expert knowledge, and is likely to be practical only case-by-case.

Comparative ex-ante STAR

Step 1. Determine the site(s) or landscape of interest – i.e. the potential geographical targets for investment

(a)

(b)

Figure 1a and 1b. Schematic of application of the comparative ex-ante approach to a region with site of biodiversity interest (A-I, e.g. Key Biodiversity Areas), or a set of grid squares (A1-D6).

Step 2a. Determine which threatened species occur in the landscape or region Confirmed presence of species is documented for Key Biodiversity Areas in the World Database of KBAs (www.keybiodiversityareas.org). Potential presence of threatened species can be determined by a GIS analysis using Red List range maps, available for commercial use via IBAT for Business (https://www.ibatforbusiness.org/). Red List categories Near-threatened, Vulnerable, Endangered and Critically Endangered should be included. For consistency, only species in fully assessed groups should be included. As of 2017, these are birds, mammals, amphibians, certain groups of reptiles and fishes, reef-forming corals, cycads and conifers2. This is to avoid biases caused by geographical variation in the comprehensiveness of Red List assessment.

Species 1 VU Species 2 VU Species 3 EN Species 4 EN Species 5 CR

Figure 2. The hypothetical region of interest holds five threatened species, 1-5, in Red List categories ranging from Vulnerable to Critically Endangered.

2 See http://www.iucnredlist.org/about/summary-statistics#Tables_3_4 Step 2b. Determine the proportion of each threatened species’ global population (P, as a percentage) that is present at each site, or in each grid square. This may be done in different ways.

A method that can be automated is to use range as a proxy, and calculate the proportion of the mapped range intersected by each site or grid square.

This is a crude method, and more reliable results will be achieved using ‘area of habitat’ maps, i.e. maps that show where in the overall range the species may actually be present. Such maps have been developed for birds, mammals and amphibians but are not yet available in IBAT. Once AOH maps are available, this approach can also be automated.

For site-scale analyses, more detailed information is necessary. Suitable information may be available in the World Database of KBAs (www.keybiodiversityareas.org) or IUCN Red List species assessments (www.iucnredlit.org), including population estimates for particular species at particular KBAs. For species largely or entirely confined to KBAs and for which KBAs have been identified throughout their distribution, the area of each KBA as a proportion of the combined area of all KBAs identified for the species can be used as a proxy measure.

Figure 3. Site A (in the hypothetical landscape in Figure 1) holds five threatened species, with different extinction risk categories and different proportions of their global population at the site.

Step 3. For each threatened species, assess the relative contribution of individual pressures to ongoing/anticipated declines. The IUCN Red List and the World Database of KBAs use a standard, hierarchical classification of ‘direct threats’ or pressures (see Annex D). For birds, and many other assessed species, each pressure has been scored for timing (past, current, future), scope (the proportion of the population it affects) and severity (how rapid a decline the threat causes for the proportion of the population that is affected).

For the STAR, only current or future pressures are relevant. These scores provide a straightforward way to calculate the relative contribution of individual direct pressures to ongoing or anticipated declines (see Annex D). Table 1 summarises the relative weighting of different combinations of scope and severity scores, from the mean expected value of the population decline expected over ten years or three generations.

Severity Very Rapid Rapid Slow, Negligible No decline *Causing/ Declines Declines Significant Declines could cause Declines fluctuations Whole 63 24 10 1 0 10 Scope Majority 52 18 9 0 0 9 Minority 24 7 5 0 0 5

Table 1. Mean % population decline expected in 10 years or three generations for different combinations of scope and severity scores (for current and future pressures only). Pressures that cause or could cause significant fluctuations can be treated as for slow, significant declines; so can pressures of unknown scope and severity where there is credible evidence for a non-negligible effect. *For the purposes of the STAR, decline rates for fluctuating populations are arbitrarily treated the same as those for “slow, significant declines”.

Some issues:

Threatened species not subject to pressures

Species listed only under Red List criterion D (CR and EN) or D1 (VU) are listed as threatened because of their small population size without any requirement for the existence of a current or future threat. A threat-based approach to assessing STAR is not appropriate for these species, and they should be excluded from the analysis. There are only 124 bird species (8.5% of all threatened birds) classified as threatened under just criterion D. Species classed only under D2 are listed because of small range size, but only if there is a plausible future threat that could quickly reduce their status to CR or EX in a short time frame. Such species are thus amenable to this approach, if the future threat is identified and scored.

Probabilistic assessment based on overall significance of pressures

The analysis is carried out for each species as a whole. For this overview analysis, the result for each site or grid square is a probabilistic assessment based on the overall scope scores for each threat (i.e. there's a higher chance of the threat being an issue in any particular square if it affects a larger proportion of the population). Specific pressures may not always be relevant to the species at a particular site or grid square. For example, a species threatened by invasive alien species (IAS) may occur at some sites where the IAS in question are not yet present. This is addressed for specific sites by the method outlined for Element 2 (see below).

Differences between species in comprehensiveness of threat scoring

Some species are better known than others. Even among birds, the comprehensiveness of threat assessment and conservation planning is patchy. Some species have only a few pressures listed, others many – sometimes including numerous pressures of low scope or severity. One approach to this would be to ignore, or cap the contribution of, lower-impact pressures. However, an analysis of threat scores for birds has shown only a handful of cases (16 species out of 2,489 threatened and Near Threatned species: 0.6%) where the cumulative threat scores are inappropriate for the threat category (BirdLife International, unpublished data); these will be corrected in the coming months. This suggests that it is reasonable to use the full suite of identified pressures when assessing the relative contribution each threat makes to ongoing and future declines.

Interacting pressures

Some pressures may be inter-dependent or synergistic. For example, climate change or road building may be direct pressures, and also promote another direct threat through the spread of invasive alien species. This is not problematic for the STAR except in cases where addressing one threat will not provide the positive results expected, because another threat must also be addressed alongside. Such cases are likely to be unusual.

Hierarchical level of pressure classification

Which pressures to use to calculate the STAR? The IUCN/Conservation Measures Partnership (CMP) Unified Classification of Direct Threats is hierarchical. Threats are always coded at the lowest level in the hierarchy (see Annex D). For example, the first-level pressure Agriculture & Aquaculture includes four second-level pressures, covering non-timber crops, forestry, livestock and aquaculture. ‘2.1 Annual and perennial non-timber crops’ in turn covers four third-level pressures, categorised as 2.1.1. Shifting agriculture; 2.1.2. Smallholder farming; 2.1.3. Agroindustry farming; 2.1.4. Scale Unknown/Unrecorded. For calculating the comparative ex-ante STAR, it is recommended that threats are included at the most detailed level where scoring is available. However, these may later be ‘rolled up’ into higher-level categories for purposes of presentation or decision-making (making the assumption that the STAR from addressing different pressures is additive). This ‘rolling up’ may be especially useful when considering interventions and calculating the ex-post STAR (see below), as one kind of intervention, e.g. land acquisition or community conservation management, could address several kinds of pressure together.

Using these means, scope x severity scores can be filled in for each species and each pressure as in Table 2 below.

3 Threat – mean % decline expected based on scope x severity scores Species Invasive Energy & Biol Res Agri- Cl change Pollution Total species mining Use culture 1 VU 0 0 9 5 1 0 15.0 2 VU 10 0 0 0 5 1 16.0

3 EN 0 1 24 10 1 9 45.0 4 EN 10 0 18 24 0 0 52.0 5 CR 1 1 10 52 0 0 64.0

Table 2. Hypothetical example of assessing the mean population implications of scope x severity scores for particular pressures (see Table 1 above; species as in Figure X)

The importance of these scores is in assessing the relative contributions of pressures to each species’ extinction risk. The absolute values are of less relevance because species may be placed in a Red List category for a number of different reasons. Rates of population decline are one factor in a set of complex criteria, with varying thrAoHolds depending on other factors.

Using the scores in Table 2, the proportional contribution of different pressures to each species’ extinction risk can now be calculated (Table 3 below).

3 In ten years or three generations (whichever is longer)

Pressure - portion of contribution to extinction risk, r Species Invasive Energy & Biol Res Agri- Cl change Pollution Total species mining Use culture 1 VU 0.00 0.00 0.60 0.33 0.07 0.00 1 2 VU 0.63 0.00 0.00 0.00 0.31 0.06 1 3 EN 0.00 0.02 0.53 0.22 0.02 0.20 1 4 EN 0.19 0.00 0.35 0.46 0.00 0.00 1 5 CR 0.02 0.02 0.16 0.81 0.00 0.00 1 Table 3. Hypothetical example of the proportional contribution of species pressures to species’ extinction risk, calculated from the scores in Table 3 (species as in Figure X). The pressure portions add up to 1 for each species.

Step 4. Combine weighted species and threat scores to calculate the ex-ante Return on Investment for each pressure, and for each site or grid square overall

Figure 4. Schematic showing calculation of STAR component for Species 4 in Site A, for one relevant pressure (Biological Resource Use) which contributes 46% of pressures overall based on scope and severity scores.

Threat category weightings

The assumption behind the STAR is that if all identified pressures were effectively addressed, and reduced to negligible scope and severity, that would permit the species to be downlisted from its current threat category to Least Concern.

The reduction in extinction risk achieved by eliminating pressures on a Critically Endangered species is much greater than for a Vulnerable species. Without weighting species, however, these would have the same potential contribution to the STAR. Therefore it is desirable to weight species by their Red List category.

The two most obvious options for weighting are:

(a) by relative extinction risk – estimated as 0.5 for Critically Endangered species, 0.05 for Endangered, 0.005 for Vulnerable and 0.0005 for Near Threatened (Butchart et al. 2004 PLoS Biol. 2: e383);

(b) using an equal steps approach – 4 for Critically Endangered species, 3 for Endangered, 2 for Vulnerable and 1 for Near Threatened, as applied in the Red List Index (Butchart et al. 2004).

Using approach (a) would mean the STAR would be overwhelmingly influenced by a small set of highly threatened species. This is not a desirable attribute for the STAR, so the equal steps weighting approach is preferred. (A similar logic has been applied in the case of the Red List Index.)

Pressure - portion of contribution to extinction risk, r Specie Red List Populat Invasive Energy Biol Res Agri- Cl Poll- Total s category ion %, P species & Use culture change ution weight, w mining 1 VU 2 3.2 0.00 0.00 0.60 0.33 0.07 0.00 1 2 VU 2 8.2 0.63 0.00 0.00 0.00 0.31 0.06 1 3 EN 3 0.7 0.00 0.02 0.53 0.22 0.02 0.20 1 4 EN 3 23.1 0.19 0.00 0.35 0.46 0.00 0.00 1 5 CR 4 12.5 0.02 0.02 0.16 0.81 0.00 0.00 1

Table 4. Hypothetical example of the proportional contribution of species pressures to species’ extinction risk, calculated from the scores in Table 2 (species as in Figure X), with Red List category weight and population portion (% at the site) for each species.

Pressure - STAR contribution = w.P.r Specie Red List Populat Invasive Energy Biol Res Agri- Cl Poll- Total s category ion %, P species & Use culture change ution ex-ante weight, w mining STAR 1 VU 2 3.2 0.0 0.0 3.8 2.1 0.4 0.0 6.4 2 VU 2 8.2 10.3 0.0 0.0 0.0 5.1 1.0 16.4 3 EN 3 0.7 0.0 0.0 1.1 0.5 0.0 0.4 2.1 4 EN 3 23.1 13.3 0.0 24.0 32.0 0.0 0.0 69.3 5 CR 4 12.5 0.8 0.8 7.8 40.6 0.0 0.0 50.0 Total ex-ante STAR 24.4 0.8 36.8 75.2 5.6 1.4 144.2

Table 5. Hypothetical example showing calculated components of potential contribution to STAR at this site or grid square, by pressure and by species, from the data in Table 4. The maximum STAR across the whole range of a species ranges from 100 (NT) to 400 (CR); for these five species, addressing all pressures across the whole range would give an STAR of 1400. At this site, the maximum potential returns equate to the sum of the proportional ranges of species, weighted by threat, which is 144.2. Assessing the ex-ante STAR for each pressure allows the potential returns to be assessed alongside the costs and feasibility of interventions, and measurability of changes in pressure intensities (see ‘Selecting and implementing interventions’, below).

The ex-ante STAR for a particular pressure at a site or grid square with n threatened species

푛 = ∑1 푤. 푃. 푟 The total ex-ante STAR is the sum of the ex-ante STAR across all pressures.

If all pressures were addressed effectively for a single species, the STAR score for that species would range from 100 (if NT) to 400 (if CR).

In Table 5, the total ex-ante site STAR is 144.2 – equivalent in conservation terms to reducing the full suite of pressures on one CR species by 36%. However, not all pressures can usually be addressed at once. The pressures with the highest ex-ante STAR here are agriculture (75.2) and biological resource use (36.8). These are therefore likely targets for interventions for this set of species in this landscape.

Appendix 4 sources of geospatial data compiled for estimation of ex-ante STAR

- Distribution ranges of threatened species: mammals, birds, reptiles and amphibians (IUCN Red List) - GBIF locality data for several taxonomic groups - Tiger conservation landscape (International Tiger Program) - Global forest change data 2000-2016 (Global Forest Change) - Tree plantations (WRI) - Logging concessions (WRI) - Oil palm concessions (WRI) - mills (WRI) - Cities, villages, towns (OpenStreet Map) - Roads (OpenStreet Map) - Waterways (OpenStreet Map) - Land cover data for concession areas (RLU - WWF-Indonesia) - HCS and HCV polygons (RLU - WWF-Indonesia). Shapefile of HCS/V appears to be different from Figure 6 in the final report (Final report HCS LAJ- SEPT 08 2015-eng.pdf) - Rubber plantations within concession areas (RLU - WWF-Indonesia) - Buildings within concession areas (RLU - WWF-Indonesia) - Roads within concession areas (RLU - WWF-Indonesia) - Rivers within concession areas (RLU - WWF-Indonesia)

Data requested but not obtained

- Requested data on Area of habitat of mammals, birds and amphibians (University of La Sapienza, Rome); AOH maps compiled from existing data - Requested camera trap data and signs data (e.g. tracks, faeces, etc.) from the Frankfurt Zoological Society (emailed Sunarto to put us in touch with Alex Moβbrucker from FZS). Request refused. - Request polygons of conservation areas AB30 1 and 2.