Jump to ContentJump to Main Navigation
OrangutansGeographic Variation in Behavioral Ecology and Conservation$

Serge A. Wich, S Suci Utami Atmoko, Tatang Mitra Setia, and Carel P. van Schaik

Print publication date: 2008

Print ISBN-13: 9780199213276

Published to Oxford Scholarship Online: May 2009

DOI: 10.1093/acprof:oso/9780199213276.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 19 September 2017

Orangutan population biology, life history, and conservation

Orangutan population biology, life history, and conservation

Perspectives from population viability analysis models

Chapter:
(p.311) CHAPTER 22 Orangutan population biology, life history, and conservation
Source:
Orangutans
Author(s):

Andrew J. Marshall

Robert Lacy

Marc Ancrenaz

Onnie Byers

Simon J. Husson

Mark Leighton

Erik Meijaard

Norm Rosen

Ian Singleton

Suzette Stephens

Kathy Traylor-Holzer

S. Suci Utami Atmoko

Carel P. van Schaik

Serge A. Wich

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199213276.003.0022

Abstract and Keywords

Orangutan populations are particularly susceptible to local extinction due to hunting, habitat loss, and fragmentation because they live at low population densities, grow slowly, and reproduce rarely. This chapter uses Population Viability Analysis (PVA) to consider the conservation implications of orangutan life history and population biology. First, a baseline model that incorporates the best available orangutan life-history data is presented. This model is then used to examine how plausible variation in model parameters, changes in the intensity of human-induced threats, and different conservation and management interventions would affect the probability of orangutan population persistence. The effects of existing threats on the extinction risk of specific orangutan populations on Borneo and Sumatra are also modelled. Finally, the conservation and management implications of this modeling exercise are considered.

Keywords:   population viability analysis, life history, conservation biology, orangutans

                   Orangutan population biology, life history, and conservationPerspectives from population viability analysis models

Photo © Perry van Duijnhoven

(p.312) 22.1 Introduction

Populations of many of Southeast Asia's rainforest vertebrates are in decline due to hunting, habitat loss, and forest degradation and fragmentation. These threats are deterministic processes as they directly increase mortality or decrease fecundity, thereby lowering population growth rates. As remaining forest patches shrink and become increasingly fragmented, animal populations become smaller and more isolated. These small populations are subject to the additional threats of stochastic processes (random genetic, demographic, or environmental fluctuations that decrease rates of survivorship or reproduction) and become vulnerable to local extinction (Soulé 1980; Soulé and Wilcox 1980; Caughley 1994). Population Viability Analysis (hereafter PVA) uses mathematical simulations to estimate extinction probabilities of animal populations subject to various deterministic forces and stochastic events (e.g., Soulé 1987; Boyce 1992; Nilsson 2004). It provides a tool for identifying populations at risk of extinction, assessing the effects of different types of threats, and evaluating the potential efficacy of conservation interventions (Lacy 1993). While not without its share of controversy (e.g., Boyce 1992; Mann and Plummer 1999; Shaffer et al. 2002), PVA has become a cornerstone of modern conservation science and wildlife management (Brook et al. 2000; Sjögren-Gulve and Ebenhard 2000; Beissinger 2002).

PVA models, coupled with empirical data, have identified several attributes that make species particularly susceptible to extinction. Population size is the strongest risk factor: species living in small populations are more vulnerable to extinction than are species living in large ones—a statement that has become axiomatic in conservation biology (Soulé and Wilcox 1980; Diamond 1984; Soulé 1987). Large-bodied species are particularly vulnerable because both intrinsic and environmental extinction risks have independent effects that rise sharply in species of .3 kg body mass (Cardillo et al. 2005). Species with slow reproductive rates are also unusually susceptible to local extinction because their populations are slow to recover from reductions in size (Terborgh 1974; Cox 1997). In addition, species with limited geographic ranges are more extinction-prone because they tend to be more specialized, less adaptable, and less able to survive reductions in habitat area (Brown 1995; Johnson 1998; Harcourt and Schwartz 2001). Finally, frugivores may be vulnerable because they must move between habitat patches during periods of resource scarcity (Terborgh and Winter 1980). Each of these factors suggests that orangutan populations might be expected to be especially prone to extinction. They live in highly fragmented landscapes (Rijksen and Meijaard 1999; Fuller et al. 2004; Goossens et al. 2005a) and most of their populations are small, with no populations exceeding 6500 animals (Rijksen and Meijaard 1999; van Schaik et al. 2001; Singleton et al. 2004). Orangutans are among the heaviest extant arboreal primates (Wheatley 1987) and have one of the slowest life histories of any mammal (Galdikas and Wood 1990; Knott 2001; Wich et al. 2004b, see Chapter 5). Their historic range extended to southern China and included most of the South East Asian mainland less than 20,000 years ago. However, presently, the geographic range of both Bornean (Pongo pygmaeus) and Sumatran (P. abelii) orangutans is each limited to a single island. Finally, orangutans are frugivorous and are known to move widely across the landscape in response to food scarcity (MacKinnon 1974; Singleton and van Schaik 2001; Buij et al. 2002). In sum, orangutans have a constellation of demographic, physiological, behavioral, and ecological attributes that render them highly vulnerable to extinction.

In this chapter we use PVA to consider the conservation implications of orangutan ecology, life history and population biology (see Box 22.1 for details and procedure). We present baseline models that incorporate the best available field data and that seem to accurately describe the dynamics of typical populations of Bornean and Sumatran orangutans in the absence of human-induced threats. We then use these model to examine how plausible variation in model parameters, changes in the intensity of human-induced threats (e.g., reduced habitat area or quality, greater frequency of disease, poaching), and different conservation and management interventions would affect the probability of population persistence. We briefly discuss models examining the effects of existing (p.313) threats on the extinction risk of specific orangutan populations on Borneo and Sumatra. Finally, we consider the conservation and management implications of this modeling exercise.

This chapter is largely based on information contained in the reports on a preliminary PVA data assessment and analysis session held in Singapore between 9–11 August 2003 (Ancrenaz et al. 2003) and a full Population and Habitat Viability Assessment (PHVA) workshop held in Jakarta between 15–18 January 2004 (Singleton et al. 2004). (In September 2005 a follow-up workshop for Sumatran orangutans was convened by Conservation International in Berastagi, North Sumatra. Participants expanded the modeling and conservation planning efforts of the 2004 PHVA workshop to develop a Sumatran Orangutan Conservation Action Plan.) The PHVA workshop is a process that brings together a broad range of stakeholders—including field biologists, land and wildlife managers, and also often local people or other resource users who would be impacted by conservation actions—to assess status, threats, and conservation options in order to develop a conservation plan that can be implemented and will be successful. The PHVA workshop uses PVA models as a core for compiling data, analysing trends, and assessing options, but the PHVA also includes considerations that go beyond the quantitative modleling of the PVA (Lacy 1993/4).

Both meetings drew on results and analyses presented at the first orangutan PHVA meeting held in January 1993 in Medan (Leighton et al. 1995). Full descriptions of the analyses and data sets described here can be found in the reports describing these meetings.

22.2 Results

22.2.1 The baseline models

The baseline models impose K through density dependent reproduction. At low to moderate population densities, the models produce low positive population growth (rSUMATRA= 0.015; rBORNEO= 0.025, 0.020, 0.015 in cases of best, medium, and worst natural mortality, respectively). These rates span what would be plausible rates of population growth for large, long-lived, slow-breeding apes. Population growth rates were reduced under crowded conditions (N = K): rSUMATRA= –0.002; rBORNEO= 0.001, –0.004, –0.009 in cases of best, medium, and worst natural mortality, respectively. Thus, the density-dependent curves we used led to essentially stable populations in the best quality habitat, while the populations would be projected to decline to a lower equilibrium size in poorer quality habitats. Although the two groups modeled mortality rather differently, both results were biologically reasonable. Moreover, when we removed all constraints on K from our models, they still stabilized at the desired population size, implying that population regulation in our model resulted from density dependence (as desired), not the upper limit K imposed on the model. When we started the population with N below 1000, the population increased to about 1000 and then stabilized (with the expected random annual fluctuations). These tests gave us confidence that we had constructed a solid baseline model that provided a reasonable representation of the dynamics of ‘ typical ’ orangutan populations and was an improvement on previous models (e.g., Leighton et al. 1995).

22.2.2 Model exploration

N and K. Figure 22.2 shows the projections for 10 simulated Bornean populations with N = 50, 250, and 1000 individuals, using mortality schedules for medium quality habitats. Population size showed the predicted strong effect on extinction probability. The smallest populations were highly unstable, going extinct following catastrophe-induced declines rather than recovering. Populations in habitats suitable for 100 orangutans sometimes persisted through the simulations, but showed high fluctuations in size (not shown). Populations in habitats with K of 250 or larger always persisted (at least in these small samples of simulations) and showed reduced fluctuations as the population sizes were increased. Plots for Bornean populations in high- and low-quality habitats and for Sumatran populations showed similar patterns.

We also examined the effects of N on probabilities of population persistence (i.e., the proportion of 500 simulated populations that survived), the (p.314) (p.315) (p.316) (p.317) (p.318) mean population sizes (i.e., the means of the 500 projections), and the mean genetic diversity (as a proportion of the starting gene diversity). We calculated values for each of these indices after 50 years, 100 years, and 1000 years (Table 22.3).

Patterns were similar between Borneo and Sumatra. Not surprisingly, given orangutans’ long lifespans, even small populations were relatively stable over the first half century. After 1000 years small populations lost a substantial amount of genetic diversity and were subject to high extinction probabilities. Although it is not known how much genetic variation orangutan populations will need to maintain high individual fitness (i.e., lack of inbreeding depression) and population adaptability to environmental change, a commonly used goal for endangered species management is to strive to maintain gene diversity of at least 90% of the initial level (Soulé et al. 1986). By this criterion, the populations with N =500 remained genetically healthy.

Hunting. We examined models with 1%, 2%, and 3% additional mortality in all age classes, running 500 iterations with populations of 250 individuals of Bornean orangutans (Fig. 22.3). In the best quality habitats, annual hunting rates of 1% did not cause population extinction (but did decrease population size), while even this low level of hunting caused declines to extinction if natural mortality is at the levels estimated for less than optimal habitat. Higher rates of hunting are unsustainable even in the highest quality habitats. Although the Sumatran group did not explicitly model its effects, hunting does occur on Sumatra and its effects are expected to be similar to the effects of hunting on Bornean populations.

(p.319)

                   Orangutan population biology, life history, and conservationPerspectives from population viability analysis models

Figure 22.2 Plots of 10 VORTEX simulations for Bornean populations in medium-quality habitats with starting population sizes (N) of 1000 (a), 250 (b) and 50 (c) individuals. All y-axes show population size, but scales differ. Plots for Sumatra and in habitats of different quality in Borneo show similar patterns.

Logging. High annual rates of habitat loss (15% and higher) result in certain extinction in all Sumatran orangutan populations within 50 years. At this rate only about 1–4% of orangutan habitat on Sumatra remains after 20 years. Moderate rates of logging (5–10% annually) drive most populations to extinction with 100 years; although initially large populations (n = 1000) persist, they consist of very few individuals at 100 years and are not viable. Low rates of logging can be sustained for 100 years, although all populations eventually go extinct within several hundred years. An annual loss of 1% results in a 63% reduction in Kover 100 years, while a 3% loss removes over 95% of the habitat in 100 years.

Inbreeding. Our model explored the effects of variation in the number of lethal equivalents in small populations (n = 100) of Bornean orangutans. In the best quality habitats, inbreeding had relatively small effects (e.g., 6.0 lethal equivalents reduced population size by 28%, Table 22.4).

In medium- or poor-quality habitat, inbreeding depression caused extinctions if lethal equivalents were ≥4.06, the value that has been estimated from zoo records. The impact of inbreeding depression is greatest in those scenarios with higher natural mortality, because the populations are less able to withstand moderate reductions in infant survival.

22.2.3 Modeling of specific populations

22.2.3.1 Sumatra: initial analyses

Of the 13 remaining orangutan populations on Sumatra, only seven are estimated to have 250 or more individuals. Of these seven relatively large populations, six are believed to be subject to 10–15% annual habitat loss due to logging. All six populations disappeared within 50 to 100 years in our simulations. Only the West Batang Toru population was sufficiently large (about 400 individuals) and had a sufficiently low estimated rate of habitat loss (2% annually) to persist for more than 150 years. However, even this population was pushed to extinction within 275 years. Assuming the life history parameters and estimated continued logging rates and impacts used in this model are realistic, simulation results suggest that the Sumatran orangutan population may decline by about 97% in the next 50 years (mean projected population size of 234 individuals) and that Sumatran orangutans may be extinct within 300 years (Fig. 22.4). In contrast, if logging and hunting of orangutans could be halted today, the number of Sumatran orangutans expected to remain in 50 years would be about 6570. It is unlikely that logging could be eliminated immediately in Sumatra. A more realistic timeline might be to end all logging within = years. Projections under this management scenario suggest that about 2758 orangutans would still remain after 1000 years, probably in 5–9 different populations. Although a delay of = years in ending (p.320)

Table 22.3 The effects of population size on population viability after 50, 100, and 1000 years. Results are mean values obtained from 500 iterations. For Sumatran orangutan populations only one mortality schedule was used; for Bornean orangutans values are computed for populations experiencing mortality rates in high-, medium-, and low-quality habitats

Input parameters

Starting N

50 years

100 years

1000 years

Mortality schedule

PE

N

GD

PE

N

GD

PE

N

GD

Sumatra

50

0

41

96

1

36

92

99

7

40

100

0

83

98

0

78

96

64

28

59

250

0

210

99

0

203

99

2

142

85

500

0

417

100

0

404

99

0

342

93

1000

0

839

100

0

808

100

0

732

97

1500

0

1269

100

0

1206

100

0

1149

98

2500

0

2085

100

0

2020

100

0

1947

99

Borneo

High-quality habitat

50

0

48

97

0

46

94

27

18

48

100

0

98

98

0

96

97

0

81

74

250

0

249

99

0

246

99

0

235

89

500

0

501

100

0

494

99

0

482

95

1000

0

1000

100

0

1006

100

0

974

97

Medium-quality habitat

50

0

44

96

0

41

93

87

2

33

100

0

91

98

0

87

97

1

60

67

250

0

229

99

0

226

99

0

209

88

500

0

460

100

0

452

99

0

433

94

1000

0

924

100

0

916

100

0

896

97

Low-quality habitat

50

0

40

96

0

36

92

99

0

25

100

0

82

98

0

77

96

44

17

56

250

0

207

99

0

202

98

0

167

85

500

0

421

100

0

416

99

0

377

93

1000

0

838

100

0

832

100

0

787

96

PE, % probability of extinction; N, mean population size; GD, genetic diversity of initial gene diversity.

logging might not result in species extinction, it could lead to over a 50% reduction in the number of orangutans that can be maintained in Sumatra under the conditions modeled. Therefore, quick action to reduce and stop logging can have long-term implications for orangutan populations.

22.2.3.2 Sumatra: subsequent analyses

In the September 2005 follow-up workshop for Sumatran orangutans, more refined estimates were made for the type (illegal logging, legal logging, encroachment), rate, and duration of habitat alteration and their estimated impacts on carrying capacity for orangutans. These resulted in an estimated loss of 40% of total carrying capacity for Sumatran orangutans over the next 20 years. These revised model results indicate that habitat loss and other factors will cause Sumatran orangutan populations to decline about 18–25% in the next 10 years (depending upon conditions) and about 50% over the next 50 years. Of the 7 habitat units that were then estimated to contain 250 or more individuals, = are projected to retain suitable habitat to sustain at least 250 orangutans in the future and were associated with long-term viability in VORTEX model projections. The remaining eight populations are subject to population decline and risk of extinction, depending upon the degree of threats to each population. Opening of additional logging concessions, population fragmentation, and hunting would lead to additional population decline and risk of extinction.

22.2.3.3 East Kalimantan

Orangutans persisted in each of the six major habitat units through 1000 years in the simulation, (p.321)

                   Orangutan population biology, life history, and conservationPerspectives from population viability analysis models

Figure 22.3 Impacts of hunting on population size. Traces show the effects of different hunting rates on VORTEX models of Bornean orangutan populations in high-quality (a), medium-quality (b), and low-quality habitats (c). In each plot lines show, from top to bottom, the effects of hunting that removes 0%, 1%, 2%, and 3% of the population annually. Lines plot mean population size of 500 iterations. Each population started with 250 individuals.

assuming no habitat loss from logging. However, populations of fewer than 300 individuals lost more than 10% of their genetic diversity, and declined in population size toward the end of the 1000 year period due to inbreeding.

22.2.3.4 Central Kalimantan

Overall, the populations in the major Central Kalimantan habitat units are sufficient in size that they would be expected to remain large and genetically healthy if the habitat remains and if hunting or other direct threats are avoided. Two populations are expected to experience habitat loss and reduction in K (Arut Belantikan, 60% decline, and Samba-Kahayan, 20% decline). Two other populations, Mawas and Sebangau, may steadily lose habitat until the orangutan populations are extirpated. An increase in available habitat is possible at Tanjung Puting, which would be expected to lead to increase in population size there.

22.2.3.5 West Kalimantan and Sarawak

Six of the habitat units in West Kalimantan and Sarawak are estimated to be sufficiently large to be capable of continuing to support demographically and genetically healthy populations. A smaller habitat unit at Bukit Baka (N = 175 individuals) also appears able to persist, although with diminished genetic diversity.

22.2.3.6 Sabah

The survey work of Ancrenaz and colleagues provides evidence that some populations in managed forests (e.g., Tabin, Trus Madi, and Sabah Foundation) are likely currently below K (Ancrenaz et al. 2004b, 2005). In contrast, the population at Kulamba is estimated to be at a size that is more than double the carrying capacity of the habitat. Similarly, the populations in the fragmented forests of Lower Kinabatangan are thought to be above K, and to currently have an excess of males, due to the sex bias in dispersal between fragments. We examined the impact of complete and partial isolation of fragments. When completely isolated, the smallest fragments do not contribute to the long-term populations in the province. Low rates of dispersal among fragments (as low as 1% to 3%; i.e., 7 to 21 animals moving between fragments annually) do provide considerable stability to the metapopulation. However, for such dispersal to occur, orangutans would have to be able to move safely among the forest fragments. If there were high mortality during dispersal, then the effect of 1–3% attempted dispersal events could be the same as 1–3% hunting—steady decline of the currently large population to extinction.

Two final notes. First, there are many small patches of forest on Borneo that contain very small populations of orangutans. These populations, smaller than any we modeled, would be very (p.322)

Table 22.4 Impacts of varying levels of the severity of effects of inbreeding, with 0.0, 2.0, 4.06 (the number, derived from zoo data, that we used in our baseline model), or 6.0 lethal equivalents, in Bornean orangutan populations with initial and maximum population sizes of 100 individuals. Results are presented using the three different mortality schedules corresponding to high-, medium-, and low-quality habitats

Input parameters

Lethal eqiuvalents

50 years

100 years

1000 years

Mortality schedule

PE

N

GD

PE

N

GD

PE

N

GD

High-quality habitat

0.00

0

99

98

0

98

97

0

97

76

2.00

0

98

98

0

98

97

0

91

76

4.06

0

98

98

0

96

97

0

81

74

6.00

0

97

98

0

95

97

0

66

72

Medium-quality habitat

0.00

0

90

98

0

90

97

0

91

73

2.00

0

91

98

0

88

97

0

78

71

4.06

0

91

98

0

87

97

1

60

67

6.00

0

91

98

0

87

97

25

27

62

Low-quality habitat

0.00

0

82

98

0

79

96

0

80

68

2.00

0

83

98

0

78

96

4

57

64

4.06

0

82

98

0

77

96

44

17

56

6.00

0

81

98

0

75

96

93

1

55

PE, % probability of extinction; N, mean population size; GD genetic diversity of inital gene diversity.

                   Orangutan population biology, life history, and conservationPerspectives from population viability analysis models

Figure 22.4 Mean population size of all surviving orangutan populations in Sumatra over the next 100 years given estimates current rates of logging.

vulnerable to extirpation and are unlikely to add anything to long-term viability of P. pygmaeus populations without active management and relocation of individuals. Second, some of the forest areas that we considered to be single ‘ habitat units’ (e.g., areas in central and west Kalimantan) are partly to severely fragmented. It is not known if orangutans can move among these forest fragments. If not, populations in these forest units will be much less stable and less secure than they appear to be in our models.

22.3 Discussion

22.3.1 Summary and general considerations

We used the best available data from long-term field studies to estimate basic vital rates for orangutan populations. We used these estimates as parameters in a PVA to assess extinction probability of populations on Borneo and Sumatra under current conditions, and to ascertain the sensitivity of the populations to various threats. Our analyses improved on the model used at the 1993 PHVA (p.323) held in Medan because we incorporated more realistic measures of density dependence, considered Sumatra and Borneo separately, modeled the trajectories of actual orangutan populations, and used a longer set of long-term data to estimate parameters. Despite these extensions and refinements, our results generally confirm those of the earlier PHVA workshop (see Leighton et al. 1995).

The Sumatran and Bornean working groups at the 2004 PHVA workshop used different parameters for some models. Some of this variation reflects differences in field data. Differences in field data are presumably partly due to random sampling errors associated with small field samples and partly due to real differences between the two Pongo species. More data are required before we can assess the relative importance of these two sources of variation. In other instances, experts in the Sumatran and Bornean group provided different informed estimates for certain parameters. These differences partly reflect perceived differences in habitat quality between the two islands (Wich et al. in review; Chapter 7), which may affect Pongo population demography on the two islands (Wich et al. 2004b, Chapter 5). In our models, P. pygmaeus had lower mortality and faster reproductive rates, and consequently exhibited more rapid population growth than P. abelii. In general, these differences were small and the results for the two orangutan species are largely congruent.

Although we used the best available data from field studies of wild orangutans, it is important to remember that most studies are conducted in high-quality habitats. Therefore estimates of key demographic and life-history parameters might provide unrealistically optimistic estimates of population viability (Marshall in press). As noted above, it is also possible that demographic rates vary across the species’ ranges. We need data from multiple studies to help document the extent of this variation, the degree of flexibility of the two species, and the relationships between habitat characteristics and orangutan demography and population dynamics.

It is also important to recognize that our models for Bornean orangutans assume that the habitat units will remain largely unchanged and will not be subjected to stresses larger than (or even, in some cases, as large as) those that they are currently experiencing. Yet many of these forests will be cleared or badly degraded unless urgent and forceful action is taken soon. These models should be seen not as predictions of what will happen, but rather as projections of the expected stability of the existing large populations of orangutans if the habitat units are preserved and other threats (e.g., hunting) are eliminated.

Our modeling shows that it may take a long time for the consequences of population fragmentation and isolation to have noticeable effects. Even populations that are most likely destined for extinction in several hundred years would appear to be stable after 50 or even 100 years (see Table 22.3). This is particularly worrying, given that our current methods for monitoring orangutan populations (i.e., nest counts) are relatively imprecise and inaccurate. It would probably be decades before we would realize that a population was in decline, at which time it may be too late to save the population without prohibitively high costs. This suggests that we need better ways to monitor orangutan populations, and that we take a preventative perspective on orangutan protection.

22.3.2 Threats and conservation actions

The biggest threats to orangutans are well known: habitat loss and hunting. Selective logging, especially at high levels, degrades habitat, but its effects on orangutan populations are less severe than complete loss of habitat or poaching (see Chapter 6). We therefore discuss only these two most severe threats below.

Loss of habitat due to clear-cutting or conversion to oil palm reduces the carrying capacity of forest fragments. Although the loss of habitat may be easily visible on satellite images, its effects on orangutan populations may not be immediately obvious. Orangutan densities may temporarily be extremely high in areas near logging operations due to compression effects (e.g., Marshall et al. 2006; see Chapter 6), potentially leading to dangerous overestimates of population size. In these compressed populations (i.e., those above K), we expect more intense intraspecific competition, which will depress fecundity and/or elevate mortality, and in time reduce population size. In other (p.324)

                   Orangutan population biology, life history, and conservationPerspectives from population viability analysis models

Figure 22.5 Mean population size of all surviving orangutan populations in Sumatra over the next 1000 years given estimates current rates of logging (bottom line), cessation of logging in 5 years (middle line), and immediate cessation of logging (top line).

words, although clear-cutting may not directly kill many orangutans, it does create an unavoidable ‘ extinction debt ’ (sensu Tilman et al. 1994). Clearly the complete removal of orangutan habitat has major negative effects, and additional clear-cutting should be prohibited from around the limited number of large orangutan populations that will be the last strongholds of wild orangutans.

Our models indicate that even very low hunting rates have strongly deleterious effects, a result confirmed by recent field surveys (Marshall et al. 2006). Simulation results suggest that in the best-quality habitats 1% annual hunting rates may be sustainable, but 2% or 3% annual offtake rates drive all populations to extinction. Nobody would claim to be able to detect the difference between 1% and 2% hunting rates in the field, yet the future of orangutan populations hangs on this minute difference. Therefore, current legal bans on hunting of orangutans should be strictly enforced. The current numbers of orangutan infants (and associated deaths of their mothers) estimated to be removed annually for the pet trade are far higher than is sustainable. Nijman (2005) estimated that between 200–500 infant orangutans were lost to the pet trade annually in Kalimantan alone, with an associated loss of 3–4% of reproductive females annually. Additional killings of orangutans for food, ritual purposes, or in response to crop raiding would further accelerate decline.

The biggest threats to orangutans apply to orangutan populations on both islands, although hunting is a larger threat in some parts of Borneo than Sumatra, and much less orangutan habitat remains on Sumatra than on Borneo. On the most basic level, hunting could be addressed effectively simply by the enforcement of existing laws. Enforcement of existing laws would also halt the loss of habitat in formally protected areas, although it is interesting to note that although killing of orangutans is forbidden anywhere, destruction of their habitat, which leads to their death, is legal outside protected areas. Despite the importance of and urgent need for enforcement, simultaneous implementation of complementary conservation interventions is likely to yield more favorable outcomes.

Issues of natural resource management and conservation are inherently multifaceted and a broad set of initiatives, incentives, and enforcement is required to adequately address them. The twoPHVA working groups discussed a wide range of options: stopping construction of roads in protected areas, funding existing conservation projects, improving law enforcement, monitoring of orangutan populations, reconnecting isolated populations with corridors, developing education and outreach programs, maintaining concession moratoriums indefinitely for legal logging, establishing helicopter patrols, encouraging the participation and collaboration of local non-governmental organizations, (p.325) promoting forest restoration, providing incentives for people to move out of important orangutan habitat, removal of illegal settlers in formally protected areas, working more closely with local governments and traditional community leaders, developing alternative income-generating activities for local people, establishing new research sites, developing innovative tourism opportunities, using remote sensing to monitor forest loss, initiating international and national media campaigns, and building local capacity. Issues related to the conservation of orangutan populations are broad in scale, both temporally and spatially, but also require solutions appropriate to local conditions. Finding and implementing these solutions will require substantial political will and investment at all levels.

22.3.3 Sumatra

It is clear that P. abelii is in more immediate threat of extinction than is P. pygmaeus. In comparison with Borneo, fewer Sumatran orangutans live in the wild, less suitable habitat remains, and forest is being lost at faster rates (Wich et al. 2003a). In addition, much of the current wild Sumatran orangutan population lives in Aceh, a province in which political unrest and environmental catastrophe have severely hampered conservation efforts in recent years. Fortunately, the political unrest is over and as a result there is again hope for implementation of conservation initiatives. The Ladia Galaska road project threatens to further divide one of Sumatra's largest remaining orangutan populations. It was against this backdrop that the Sumatra group considered how the conservation and management of orangutans on the island could be best effected. They concluded that many of the recommendations listed above could be usefully applied to the management and protection of Sumatran orangutan populations. Several additional possibilities were considered: gaining World Heritage Site status for the Leuser ecosystem, continuing efforts to connect the Trumon-Singkil and west Leuser populations, reconnecting the west and east Leuser habitat units, and reactivating research at Ketambe. We feel that implementation of these recommendations must be considered a major priority in global efforts to preserve wild populations of great apes.

22.3.4 Borneo

At the end of our PHVA meeting, the Borneo working group was faced with the prospect of conveying a complicated message. On one hand, there are far more wild orangutans on Borneo than was previously thought. On the other hand, far more orangutans have been lost in Borneo over the last century than we had previously realized. This message is at once both sobering and hopeful, and indicates that the first orangutan population counts probably underestimated their true abundance by several orders of magnitude. The current situation in Borneo is not as dire as in Sumatra, but the plight of P. abelii provides a glimpse of what the future for P. pygmaeus may hold. Very few orangutan habitats in Borneo are immune from logging, and most are also subjected to hunting, either directly for meat and pets, or indirectly when displaced orangutans are killed as agricultural crop pests. Only timely and effective interventions will prevent Bornean populations from facing the imminent extinction that now confronts Sumatran ones.

22.3.5 PVA and the realities of orangutan conservation

Population Viability Analyses have been part of the conservation biologist's toolkit for over two decades. Over this time, more sophisticated analytic techniques and more fully developed theoretical considerations have been incorporated into PVA (e.g., Gilpin and Soulé 1986; Beissinger and Westphal 1998; Sjögren-Gulve and Ebenhard 2000; Sæther and Engen 2002; White et al. 2002), changes which are assumed to have increased the accuracy and precision of our estimates of extinction probabilities. Despite these refinements, PVA results are only as accurate as the data upon which they are based (Lindborg and Ehrlén 2002). Estimates of mean vital rates are notoriously difficult and time-consuming to obtain for wild vertebrate populations (Thomas 1990; Münzbergová and Ehrlén 2005), and the equally important measures of variance in these rates are even more difficult to assess (Beissinger and Westphal 1998; Sæther and Engen 2002). This could scarcely be more true than for orangutans, who live at low population densities, (p.326) have long lifespans, reproduce slowly, and for whom few long-term data exist. In addition, climate, governance, and land-use policies are likely to change in important but unknown ways in the next decades in Indonesia and Malaysia, limiting the extent to which we can project current models into the future.

We recognize that our estimates of basic life history variables and the intensity of important threats are not as accurate nor as precise as we would like. However, we also recognize that unless current threats are quickly and drastically reduced, Sumatran orangutans will be extinct long before we have the chance to collect such specific data. Conservation biology, the new ‘ dismal science ’ , is full of dire predictions and urgent calls to arms. In few cases is the situation more dire and the need for action more urgent than for Sumatran orangutans. Unless bold and courageous action is taken soon, we will have stood by as one of our closest living relatives went extinct in the wild.

We conclude this chapter by summarizing the main results and indicating management implications emerging from the analyses.

  1. 1. Indonesian and Malaysian forests are disappearing and orangutans on both islands are becoming increasingly isolated into forest fragments. Populations in fragments for which we have good data are in decline. Orangutans are particularly vulnerable to population decline and local extinction because of their life histories. The biggest threats are hunting and habitat loss.

  2. 2. Our initial exploration of some scenarios representing typical populations on Borneo suggests that orangutan populations restricted to habitats capable of supporting only about 50 animals can persist for a considerable number of years, but are unstable and vulnerable to extirpation. Habitats capable of supporting more than 250 orangutans appeared necessary to ensure good demographic and genetic stability.

  3. 3. At typical population densities, 500–1000 km2 of habitat is required to maintain a demographically and genetically healthy population of Bornean orangutans. As some Sumatran forests appear to support higher orangutan densities, a slightly smaller area may be sufficient on Sumatra. Regardless, relatively few protected areas contain 500 km2 of good orangutan habitat. This means that protection and management of orangutans outside formally protected areas is required to maintain viable populations.

  4. 4. Simple calculations indicate that orangutan populations under no external threats can grow at a maximum of 2% annually (rmax ∼ 0.02). Probably very few wild populations achieve this maximum theoretical rate. This means that the loss only a few individuals per year can mean the difference between persistence and extinction.

  5. 5. Orangutan life history and population biology result in strong time lags between insults and detectable reduction in population size. Once we detect that a population is in decline, it may be too late to save it. Therefore, a precautionary approach is necessary to avert the extinction of orangutan populations.

Acknowledgments

We gratefully acknowledge the hard work and contributions of all who helped to make the 1993, 2003, and 2004 orangutan PHVA exercises a success by participating, contributing data or expertise, or assisting with organization and funding. In particular, we thank the IUCN SSC Conservation Breeding Specialist Group, the Chicago Zoological Society, Orangutan Foundation–UK, Dr. W.T.M. Smits, and The Schmutzer Primate Center for financial and logistical assistance. We would also like to acknowledge the Indonesian Institute of Sciences (LIPI) and Directorate General for Nature Conservation (PHKA) for their cooperation and for granting permission to conduct research and conservation activities in Indonesia. We thank two anonymous reviewers for helpful comments on this chapter. An executive summary of the 1993 Orangutan PHVA and the full report of the 2004 Orangutan PHVA are available at http://www.cbsg.org. Copies of the VORTEX program and manual are available at http://www.vortex9.org/vortex.html.