orangutan€¦ · if the aza bornean orangutan population continues its current breeding rate (2.7...
TRANSCRIPT
Population Biologist Brent Johnson, [email protected] AZA Animal Program Leader Lori Perkins, [email protected] Regional Studbook Keeper & AZA Vice-Animal Program Leader Megan Elder, [email protected] AZA Ape TAG Chair Tara Stoinski, [email protected] AZA Ape TAG Vice-Chair Tracy Fenn, [email protected]
Orangutan (Pongo pygmaeus +
Pongo abelii) AZA Animal Programs
Population Viability Analysis Report
August 19, 2014
2014 Population Viability Analyses for AZA Orangutan Animal Program 1
TABLE OF CONTENTS
EXECUTIVE SUMMARY ...................................................................................................................................2
POPULATION VIABILITY ANALYSIS (PVA) APPROACH ....................................................................................3
REPORT OVERVIEW ........................................................................................................................................4
HYBRID POPULATION VIABILITY UNDER CURRENT MANAGEMENT ..............................................................5
BORNEAN POPULATION HISTORY AND CURRENT STATUS ............................................................................6
BORNEAN MODEL SCENARIOS .......................................................................................................................9
BORNEAN POPULATION VIABILITY UNDER CURRENT MANAGEMENT ........................................................10
BORNEAN ALTERNATE MODEL SCENARIOS .................................................................................................11
SUMATRAN POPULATION HISTORY AND CURRENT STATUS .......................................................................13
SUMATRAN MODEL SCENARIOS ..................................................................................................................16
SUMATRAN POPULATION VIABILITY UNDER CURRENT MANAGEMENT .....................................................17
SUMATRAN ALTERNATE MODEL SCENARIOS ..............................................................................................18
MANAGEMENT ACTIONS .............................................................................................................................19
CONCLUSIONS ..............................................................................................................................................19
ACKNOWLEDGEMENTS ................................................................................................................................20
LITERATURE CITED ........................................................................................................................................21
DEFINITIONS .................................................................................................................................................22
APPENDICES .................................................................................................................................................23
2014 Population Viability Analyses for AZA Orangutan Animal Program 2
EXECUTIVE SUMMARY Population Viability Analyses (PVA) are being conducted by Lincoln Park Zoo and Population Management Center researchers through
funding from the Institute of Museum and Library Services (IMLS). The project team uses ZooRisk 3.80 (Earnhardt et al. 2008), a PVA
modeling software, to examine what would happen to AZA populations if current conditions remain the same (the baseline scenario), and
then assess the impact of changes in reproductive rates, space availability, imports/exports, and other potential management actions
(alternate scenarios). Model scenarios for this population were developed with members of the Association of Zoos and Aquarium (AZA)
Ape Taxon Advisory Group (TAG) during summer of 2014.
POPULATION HISTORY/CURRENT STATUS Bornean orangutans (Pongo pygmaeus) have been consistently held in AZA institutions since 1940. Through a combination of importations
of animals from outside of AZA and zoo births, the population grew to a size near 75 individuals by 1969. The population currently includes
84 individuals. In the past 10 years, the population has had an average of 2.7 births/year. Current gene diversity is high (97.2%) and
inbreeding is low (average inbreeding coefficient of 0.0004).
Sumatran orangutans (Pongo abelii) have been consistently housed in North American institutions since 1928, although the population size
remained low (<30 individuals) until 1960. After that year, the population grew through a combination of importations of animals from
outside of the formally managed population and zoo births to a peak size of 105 individuals in 1997. The population averaged 2.8
births/year in the past 10 years and currently includes 85 individuals. At present, the Sumatran orangutan population has high gene
diversity (97.6%) and low inbreeding (average inbreeding coefficient of 0.0056).
For each orangutan population, managers have been moving toward a standard of parent rearing and more prolonged time that offspring
spend with their mother. They intend to maintain a minimum of 8-year interbirth intervals among females (except when infant mortalities
occur) to allow them to fully raise their young.
PVA RESULTS Model results indicate that hybrid orangutans will age out of AZA institutions in approximately 32 years, which could increase spaces for
the other orangutan populations. If the AZA Bornean orangutan population continues its current breeding rate (2.7 births/year in the past
10 years), it is expected to have a 4% chance of extinction or approximately 16 individuals remaining in 100 years. However, increasing
breeding is predicted to allow the AZA Bornean orangutan population to remain stable over the next 100 years. For the population to
maintain its current size, it would need to produce an average of ~4 births per year over the next 10 years. If the population produces ~5
births/year, it could fill 115 potential spaces (including half of those currently occupied by hybrids) in approximately 32 years.
Under its current breeding rate (2.8 births/year in the past 10 years), the formally managed Sumatran orangutan population is predicted to
decline to an average of 50 individuals over the next 100 years. Increasing breeding is predicted to also allow the Sumatran orangutan
population to remain stable over the next 100 years. To maintain its current size, the population should produce an average of ~4 births
per year over the next 10 years. By producing ~4 births/year over the next decade and ~5 births every following year, the population could
fill 115 potential spaces in approximately 33 years. Increasing breeding rates (to either maintain the current population size or fill potential
spaces) is also expected to maintain the genetic health of each orangutan population over the next century.
MANAGEMENT ACTIONS The AZA Orangutan Animal Programs should consider the following changes to current management strategies:
Re-allocate spaces currently occupied by hybrids as they become available: Because spaces are expected to become available as hybrid
orangutans age out of AZA institutions, managers should ensure that these spaces are allocated to Bornean and Sumatran orangutans to
maintain each of these populations at a larger, stable size with the best possible genetic health.
Increase breeding rates: For each orangutan population, breeding should be increased from ~3 births per year to a minimum of ~4 births
each year to maintain the current population size. To fill spaces currently occupied by hybrid orangutans as they become available, the
Bornean population should produce ~5 births per year. The Sumatran population, however, should produce ~4 births each year for the
next decade and ~5 births every following year in order to do so. To achieve these higher breeding rates, it may not be possible to
immediately implement the recommended 8-year interbirth intervals. If higher breeding rates could be maintained, each population may
become demographically stable and maintain its genetic health.
2014 Population Viability Analyses for AZA Orangutan Animal Program 3
FULL REPORT
POPULATION VIABILITY ANALYSIS (PVA) APPROACH
A Population Viability Analysis (PVA) is a model that projects the likely future status of a population. PVAs are used to
evaluate long-term demographic and genetic sustainability and extinction risk, identify key factors impacting a population’s
dynamics, and compare alternative management strategies.
This PVA utilizes ZooRisk, a computer software package that models the future dynamics of a cooperatively-bred population
using that population’s age and sex structure, mortality and reproductive rates, and genetic structure (Earnhardt et al., 2008).
ZooRisk is individual-based, meaning it tracks every animal (current and future) in the population over time. It also includes
stochasticity, the randomness in mortality, fecundity, and birth sex ratios among individuals, which is especially important for
small populations. Because of this stochasticity, we run each model many times, allowing us to determine the range of
potential outcomes a population could experience under a given set of conditions.
The most powerful use of PVAs is to compare a baseline scenario, reflecting the population’s likely future trajectory if no
management changes are made, to alternate scenarios reflecting potential management changes. For zoo and aquarium
populations, these alternate scenarios typically involve varying breeding rates (probability of breeding, or p(B)), potential
space for the population, importation or exportation rates, mortality rates, or genetic management strategies. These
comparisons can help evaluate the relative costs and benefits of possible management actions. Because the future can be
uncertain and difficult to predict, model results are most appropriately used to compare between scenarios (e.g. relative to
each other) rather than as absolute predictions of what will happen.
Full documentation on ZooRisk can be found in the software’s manual (Faust et al., 2008); complete details on the modeling
approach for this PVA, including data sources, parameter values, and model setup, can be found in Appendix A.
2014 Population Viability Analyses for AZA Orangutan Animal Program 4
REPORT OVERVIEW
Orangutans in the entire, formally managed population (i.e., at AZA institutions and Toronto Zoo) have historically included
individuals of both the Bornean (Pongo pygmaeus) and Sumatran species (Pongo abelii), as well as hybrids of the two and
individuals of unknown origin (Fig. 1). Today, relatively even numbers of Bornean and Sumatran orangutans are held among
these institutions (84 Bornean orangutans and 85 Sumatran orangutans), with 45 remaining individuals being hybrids. In this
report we very briefly present a scenario pertaining to the hybrid population, and then separately describe the population
history, status, and model results for the Bornean and Sumatran orangutan populations.
0
50
100
150
200
250
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Sumatran
Bornean
Hybrid
Unknown
Year
Nu
mb
er
of
Ind
ivid
ual
s
Figure 1. Number of orangutans in the entire, formally managed population since 1940.
2014 Population Viability Analyses for AZA Orangutan Animal Program 5
HYBRID POPULATION VIABILITY UNDER CURRENT MANAGEMENT
Currently, 45 hybrid orangutans (21 males, 24 females) are housed at 23 AZA institutions. The majority of these individuals
are currently above 25 years old (Fig. 2). Because hybrids exist in zoos only as a result of breeding that occurred prior to the
recognition of two distinct orangutan species, the Ape TAG plans to manage the hybrid population without any further
reproduction. In this management scenario, the hybrid population can be expected to age out of zoos in approximately 32
years (Fig. 3, Table 1).
Table 1. Baseline model results.
SCENARIO
DEMOGRAPHICS
Initial Population
Size Size in Year 25
1 Size in Year 100
1
Probability of
Reaching Spaces
(%)
Probability of
Extinction (%)
Median Years until
Reaching
Extinction
A p(B) = 0% 45 2 ± 1 0 0 100 32 1
Mean value + 1 Standard Deviation, taken across 1000 iterations. If an iteration goes extinct that value is not included in the calculation, so results may only reflect a few iterations
in scenarios with a high probability of extinction.
Figure 2. The age distribution of the total hybrid population (21.24) within AZA institutions.
0
10
20
30
40
50
60
70
80
1950 1970 1990 2010 2030 2050 2070 2090 2110
Nu
mb
er
of
Ind
ivid
ual
s
Year
A) p(B) = 0% Historic
Figure 3. Historical and projected mean population size with no future
breeding. Projected results are averaged across 1000 model iterations.
2014 Population Viability Analyses for AZA Orangutan Animal Program 6
Table 2. Summary of demographic statistics for the AZA population. Population Sizes Total (male.female.unknown)
Current population 84 (33.51.0)
Potentially breeding population 79 (33.46.0)
Annual rates over the last decade Mean (min-max)
Population Growth (Lambda, λ) 1.006 (0.932-1.072)
Births 2.7 (1-7)
Deaths 2.6 (0-7)
Imports 0.4 (0-4)
Exports 0
BORNEAN POPULATION HISTORY AND CURRENT STATUS
Demographics Based on the international orangutan studbook (Elder,
2014), Bornean orangutans have been consistently held in
AZA institutions since 1940. Initial growth of the population
was driven exclusively by importations of wild-caught
individuals (Figs. 4 and 5a). (Note that throughout this
section, “imports” are animals entering AZA institutions
which may be coming from non-AZA institutions such as the
private sector, EAZA, other zoo regions, or from the wild,
and conversely “exports” are animals exiting AZA institutions
and going into non-AZA populations, such as other zoo
regions or being released into the wild). Regular breeding
of Bornean orangutans in AZA zoos began in 1960, when the
population size was 39 individuals. Since 1969, the
population has remained above approximately 70
individuals (Fig. 4). The population peaked in size at 89
individuals in 2009.
Over the last 10 years (2004-2013, based on the studbook currentness date), the Bornean orangutan population has
increased slightly in size from 78 to 84 individuals and has experienced an average annual growth rate of 0.6%, although
specific annual growth rates have ranged from -6.8% to +7.2% (Table 2). The population has averaged 0.0% annual growth in
the past five years. In the past decade, the population has averaged 2.7 births per year (which corresponds to each female
having an 11.5% probability of breeding in a given year;
see Appendix A) and 0.4 imports each year.
As of April 26th
, 2014 (the studbook currentness date), the
Bornean orangutan population consisted of 84 animals (33
males, 51 females). Five females are excluded from the
breeding population, based on being beyond the
reproductive age window (14-42; see Appendix A) or being
sterile. This leaves a potentially breeding population of
79 (33.46) individuals. See Appendix B for a complete list
0123456789
10
1940 1950 1960 1970 1980 1990 2000 2010
Nu
mb
er
of
Ind
ivid
ual
s
Year
Births Imports
0123456789
10
1940 1950 1960 1970 1980 1990 2000 2010
Nu
mb
er
of
Ind
ivid
ual
s
Year
Deaths Exportsa) b)
Figure 5. Number of a) births and imports and b) deaths and exports in the population since 1940. Imported animals entering AZA institutions may be coming from non-AZA institutions (private sector, EAZA, other zoo regions, or from the wild); exported animals may be going to other institutions outside of AZA in North America, other regions, or being released into the wild.
0
10
20
30
40
50
60
70
80
90
100
1940 1950 1960 1970 1980 1990 2000 2010
Nu
mb
er
of
Ind
ivid
ual
s
Year
Total Males Females
Figure 4. Number of Bornean orangutans in AZA institutions since 1940.
2014 Population Viability Analyses for AZA Orangutan Animal Program 7
of the individuals included in the model and their reproductive status.
The population has an age structure that is largely columnar with many empty age classes (Fig. 6a). There are also few
individuals in lower age classes due to somewhat low reproduction in the past seven years. The potentially breeding
population currently displays a female-biased sex ratio of 1.4 females per male (Fig. 6b). Orangutans generally breed best
when held in groups with 1 to 3 females per male, therefore the current sex ratio is appropriate for the management of this
population.
Figure 6. The population’s age distribution within AZA institutions, divided into a) the total population 84 (33.51.0) and b) the potentially breeding population 79 (33.46.0).
Genetics The Bornean orangutan population’s pedigree is 100%
known (Table 3), and no analytical assumptions were
necessary. The managed population is descended from 50
founders with no additional potential founders remaining in
the population. Current gene diversity is estimated to be
97.23%. As gene diversity falls, reproduction may become
increasingly compromised by lower birth weights, greater
neonatal mortality, and other negative factors.
The population has a low mean inbreeding level of 0.0004 (a mean inbreeding coefficient of 0.0625 is equivalent to mating
between first cousins with no prior inbreeding). One of the largest genetic threats to small populations is the potential for
inbreeding depression, in which breeding between close relatives results in reductions in fecundity or litter size, increases in
infant mortality, and other detrimental effects (DeRose and Roff, 1999; Koeninger Ryan, et al., 2002; Ballou and Foose, 1996;
Reed and Frankham, 2003). Given the low level of inbreeding in the Bornean orangutan population, the small number of
individuals with inbreeding values above 0 prevented us from statistically testing for possible effects on infant mortality. It is
unlikely that inbreeding is impacting the population at this time; however, inbreeding may become a threat in the future and
should therefore be avoided. Because modeling inbreeding depression adds an additional layer of complexity to
interpretation of results, we have not included a “standard” inbreeding depression effect in the PVA models. There are
several strategies that can delay the effects or lower the probability of inbreeding depression including pairing based on
mean kinship and importing and breeding unrelated individuals (Ballou and Lacy, 1995). These strategies were modeled for
Table 3. Summary of starting genetic statistics for the population.
Percentage of pedigree known 100%
Genetic Diversity (GD) 97.23%
Population mean kinship (MK) 0.0277
Mean inbreeding (F) 0.0004
Mean generation time (T) (years) 20.9
Number of generations in 100 model years 4.8
a) b)
2014 Population Viability Analyses for AZA Orangutan Animal Program 8
all scenarios except those without genetic management. Because inbreeding depression is not included, readers should
consider that model results may be optimistic if inbreeding depression begins to impact the population.
Management ZooRisk can include a space limitation on population growth, reducing breeding as the population approaches the potential
space limit. This mimics the way a population manager would recommend fewer pairs when at capacity. To accurately model
the “potential space” for a population, we use either a) the projected spaces in 5 years based on a TAG’s Regional Collection
Plan (RCP) or, if that value is unavailable or inappropriate, b) the current population size + 10% or 10 individuals, whichever is
greater. These values are also placed in context of current institutional interest by the Species Coordinator. The Ape TAG’s
most recent space survey set the target size for Bornean orangutans over 4 years old at 85 individuals (AZA Ape Taxon
Advisory Group, 2007), however some additional spaces could become available if the population grows. Given this fact, and
because animals under 4 four years old still occupy space in ZooRisk, we allowed for 100 potential spaces in our model.
The AZA Bornean orangutan population is currently designated as a Green Program. The population is housed at 25
institutions, with 21 having at least one potential breeding pair.
Since the late 1980s, parent rearing has become the primary means of raising young Bornean orangutans in zoos (Fig. 7).
Additionally, there has been an increasing trend in the interbirth intervals among females raising offspring. Managers intend
to have interbirth intervals no shorter than 8 years except when infant mortalities occur to ensure proper social development
of the offspring. If reproductively viable females in the population were to maintain an average interbirth interval of 8 years,
the female probability of breeding (p[B]) in any given year would be equal to 12.5% (see appendix A for more information on
p[B]).
Figure 7. Interbirth intervals related to each rearing style over time.
0
2
4
6
8
10
12
14
16
1950 1960 1970 1980 1990 2000 2010
Inte
rbir
th In
terv
al (
Ye
ars)
Year
Parent Hand Foster
Supplemental Unknown Trend (Parent)
2014 Population Viability Analyses for AZA Orangutan Animal Program 9
BORNEAN MODEL SCENARIOS
Model scenarios were created to reflect what would happen if current management approaches continued (baseline) and to
address potential alternate management strategies (Table 4). Model setup is described more fully in Appendix A. Scenarios
are described in more detail in the following sections.
Table 4. ZooRisk Model Scenarios
Scenario Name Scenario Description p(B) Spaces
Baseline Scenario:
B. Baseline; p(B) = 11.5% Reproduction to match past 10 years (2.7 births/year) 11.5% 100
Alternate Scenarios:
C. p(B) = 19% Reproduction required to sustain current population size 19% 100
D. p(B) = 25%; spaces = 115 Reproduction required to fill hybrid spaces as they become
available (i.e., approaching 115 spaces over the next 32 years) 25% 115
p(B) = Probability of Breeding
2014 Population Viability Analyses for AZA Orangutan Animal Program 10
BORNEAN POPULATION VIABILITY UNDER CURRENT MANAGEMENT BASELINE MODEL PVA RESULTS
We examined the viability of the AZA Bornean orangutan
population under current management: 11.5% female
probability of breeding (equivalent to 2.7 births/year over the
past 10 years) and no imports or exports. The population
currently has a target size of 85 for individuals older than 4
years old, however some additional spaces could become
available if the population grows. Because of this flexibility in
the availability of spaces, and the fact that animals younger
than 4 years old hold space within our models, we allowed for
100 potential spaces in the baseline scenario. See Appendix A
for full details on all model parameters.
If the current breeding rate continues, the Bornean orangutan
population has a 4% chance of reaching extinction (Table 5). If
it does not go extinct, the population is expected to have
approximately 16 individuals remaining in 100 years (Table 5,
Fig. 8). In this scenario, the population would retain 84.4% gene diversity and fairly low inbreeding (average inbreeding
coefficient (F) = 0.055) after 100 years, although inbreeding would begin approaching the equivalent of mating between first
cousins (F = 0.063). The population is predicted to have a Vulnerable risk status in zoos and aquariums due to its declining
gene diversity (see Appendix C).
Table 5. Baseline model results.
SCENARIO
DEMOGRAPHICS GENETICS
OVERALL
POPULATION
STATUS2
Initial
Population
Size
Size in
Year 251
Size in
Year
1001
Probability
of Reaching
Spaces (%)
Probability
of
Extinction
(%)
Initial
GD (%)
GD
Retained
in Year
1001
Initial F
F Retained
in Year
1001
B p(B) = 11.5% 84 60 ± 9 16 ± 10 1 4 97.2 84.4 ± 8.8 0.0004 0.055 ±
0.034 Vulnerable
GD = gene diversity, F = inbreeding coefficient 1 Mean value + 1 Standard Deviation, taken across 1000 iterations. If an iteration goes extinct that value is not included in the calculation, so results may only reflect a few iterations
in scenarios with a high probability of extinction. 2ZooRisk uses five standardized tests to give a summary risk score for each scenario, from Low Risk (most secure), Vulnerable, Endangered, to Critical (least secure). For more details
on this score, see Appendix C.
0
20
40
60
80
100
1980 2000 2020 2040 2060 2080 2100
Nu
mb
er
of
Ind
ivid
ual
s
Year
B. Bornean baseline; p(B) = 11.5%
Historic
Potential Space
Figure 8. Historical and projected mean population size under the baseline model scenario. Projected results are averaged across 1000 model iterations.
2014 Population Viability Analyses for AZA Orangutan Animal Program 11
BORNEAN ALTERNATE MODEL SCENARIOS
We explored the impacts of altering the probability of breeding for the Bornean orangutan population in order to maintain
the population at its current size or increase the population to fill all potential spaces. For the population to sustain its
current size of 84 individuals for the next 100 years (Fig. 9, Table 6), it needs to produce an average of ~4 births/year (p(B) =
19%, scenario C; Fig. 10). Under this breeding rate, the population is expected to maintain high gene diversity (94.3%) and
fairly low inbreeding (mean inbreeding coefficient (F) = 0.043) over the next 100 years, although inbreeding would begin
approaching the equivalent of mating between first cousins (F = 0.063). Note that a p(B) = 19% is approximately equivalent
to reproductively viable females in the population maintaining an average interbirth intervals of 5.3 years. This interval could
account for infant mortality, meaning that females who lost infants would likely have shorter interbirth intervals while those
that had surviving infants could have longer ones.
We estimate that 115 potential spaces could be available for the Bornean orangutan population after hybrids age out of AZA
zoos in approximately 32 years (scenario A; Table 1), which is equal to the population’s target size of 85 individuals plus ~8
spaces for offspring and half of spaces previously occupied by hybrids (~22 spaces). If the population produces an average of
~5 births/year, it could fill the 115 potential spaces in approximately 32 years (p(B) = 25%, scenario D; Figs. 9 and 10). A p(B)
= 25% is equivalent to every reproductively viable female in the population maintaining an average interbirth interval of 4
years (which would be an average of intervals with the first offspring living and the first offspring not surviving). In this
management scenario, the population would maintain high gene diversity (95.1%) and fairly low inbreeding levels (F = 0.039)
over the next century (Table 6). In both increased breeding scenarios, the population is projected to retain a Low Risk status
in zoos (Appendix C).
Table 6. Model results for all Bornean orangutan scenarios.
SCENARIO
DEMOGRAPHICS GENETICS
OVERALL
POPULATION
STATUS2
Initial
Population
Size
Size in
Year 251
Size in
Year
1001
Probability
of Reaching
Spaces (%)
Probability
of
Extinction
(%)
Initial
GD (%)
GD
Retained
in Year
1001
Initial F
F Retained
in Year
1001
B p(B) = 11.5% 84 60 ± 9 37 ± 13 1 0 97.2 84.4 ± 8.8 0.0004 0.055 ±
0.034 Vulnerable
C p(B) = 19% 84 90 ± 10 86 ± 16 85 0 97.2 94.3 ± 0.8 0.0004 0.043 ±
0.009 Low Risk
D p(B) = 25%;
spaces = 115 84 104 ± 7 114 ± 4 100 0 97.2 95.1 ± 0.2 0.0004
0.039 ±
0.007 Low Risk
GD = gene diversity, F = inbreeding coefficient 1 Mean value + 1 Standard Deviation, taken across 1000 iterations. If an iteration goes extinct that value is not included in the calculation, so results may only reflect a few iterations
in scenarios with a high probability of extinction. 2ZooRisk uses five standardized tests to give a summary risk score for each scenario, from Low Risk (most secure), Vulnerable, Endangered, to Critical (least secure). For more details
on this score, see Appendix C.
2014 Population Viability Analyses for AZA Orangutan Animal Program 12
0
20
40
60
80
100
120
1940 1970 2000 2030 2060 2090
Nu
mb
er
of
Ind
ivid
ual
s
Year
C. Bornean; p(B) = 19%D. Bornean; p(B) = 25%; space = 115Potential SpaceHistoric
Figure 9. Historical and projected mean population size under alternate
model scenarios. Projected results are averaged across 1000 model
iterations.
0
1
2
3
4
5
6
7
2000 2020 2040 2060 2080 2100
Nu
mb
er
of
Hat
che
s
Year
C. Bornean; p(B) = 19%
D. Bornean; p(B) = 25%; space = 115
Past 10 Years
Figure 10. Number of total births in the AZA population in the past decade and projected mean number of births under alternate model scenarios. Projected results are averaged across 1000 model iterations.
2014 Population Viability Analyses for AZA Orangutan Animal Program 13
Table 7. Summary of demographic statistics for the population. Population Sizes Total (male.female.unknown)
Current population 85 (36.48.1)
Potentially breeding population 68 (34.33.1)
Annual rates over the last decade Mean (min-max)
Population Growth (Lambda, λ) 0.996 (0.932-1.049)
Births 2.8 (1-4)
Deaths 3.2 (0-7)
Imports 0.2 (0-1)
Exports 0.2 (0-1)
SUMATRAN POPULATION HISTORY AND CURRENT STATUS
Demographics Based on the international orangutan studbook (Elder,
2014), Sumatran orangutans have been consistently held in
AZA zoos and one non-AZA partner institution (Toronto Zoo)
since 1928. However, the population size remained low
(<30 individuals) until 1960 (Fig. 11). Regular breeding of
Sumatran orangutans within the formally managed
population began the following year (Fig. 12a). Initial
population growth was driven by importations of wild-
caught individuals, as well as zoo births, and the population
reached over 80 individuals in 1969 (Fig. 11). (Note that
throughout this section, “imports” are animals entering the
formally managed population which may be coming from
outside institutions such as the private sector, EAZA, other
zoo regions, or from the wild, and conversely “exports” are
animals exiting the formally managed population and going
into other populations, such as other zoo regions or being released into the wild). In 1997, the Sumatran orangutan
population size peaked at 105 individuals.
Over the last 10 years (2004-2013, based on the studbook currentness date), the formally managed population has decreased
slightly in size from 91 to 85 individuals and has experienced an average annual growth rate of -0.4%, although specific
annual growth rates have ranged from -6.8% to +4.9% (Table 7). The population has averaged +1.2% annual growth in the
past five years. In the past decade, the population has had
an average of 2.8 births per year (which corresponds to
each female having a 14.8% probability of breeding in a
given year; see Appendix A), as well as an average of 0.2
imports and 0.2 exports each year.
As of April 26th
, 2014 (the studbook currentness date), the
Sumatran orangutan population consisted of 85
individuals (36 males, 48 females, 1 unknown).
Seventeen animals (2 males, 15 females) are excluded
0
2
4
6
8
10
12
14
16
1920 1940 1960 1980 2000
Nu
mb
er
of
Ind
ivid
ual
s
Year
Births Imports
0
2
4
6
8
10
12
14
16
1920 1940 1960 1980 2000
Nu
mb
er
of
Ind
ivid
ual
s
Year
Deaths Exportsa) b)
Figure 12. Number of a) births and imports and b) deaths and exports in the population since 1928. Imported animals entering the formally managed population may be coming from outside institutions (private sector, EAZA, other zoo regions, or from the wild); exported animals may be going to other institutions outside of the formally managed in North America, other regions, or being released into the wild.
0
20
40
60
80
100
120
1920 1940 1960 1980 2000
Nu
mb
er
of
Ind
ivid
ual
s
Year
Total Males Females
Figure 11. Number of Sumatran orangutans in the formally managed
population since 1928.
2014 Population Viability Analyses for AZA Orangutan Animal Program 14
from the breeding population, based on being beyond the reproductive age window (10-45 for males, 14-45 for females; see
Appendix A), having health or behavioral issues, or being sterile. This leaves a potentially breeding population of 68
(34.33.1) individuals. See Appendix B for a complete list of the individuals in the model and their reproductive status.
The formally managed population has an age structure that is largely columnar with many empty age classes (Fig. 13a). The
potentially breeding population currently displays an even sex ratio (Fig. 13b). Orangutans generally breed best when held in
groups with 1 to 3 females per male, therefore the current sex ratio is appropriate for the management of this population.
However, managers are concerned about the loss of females from the potentially breeding population, which could cause a
male-biased sex ratio.
Figure 13. The population’s age distribution within AZA and partner institutions, divided into a) the total population 85 (36.48.1) and b) the potentially
breeding population 68 (34.33.1).
Genetics The Sumatran orangutan population’s pedigree is 100%
known (Table 8), and no analytical assumptions were
necessary. The managed population is descended from 50
founders with no additional potential founders remaining in
the population. Current gene diversity is estimated to be
97.58%. As gene diversity falls, reproduction may become
increasingly compromised by lower birth weights, greater
neonatal mortality, and other negative factors.
The population has a low mean inbreeding level of 0.0056 (a mean inbreeding coefficient of 0.0625 is equivalent to mating
between first cousins with no prior inbreeding). One of the largest genetic threats to small populations is the potential for
inbreeding depression, in which breeding between close relatives results in reductions in fecundity or litter size, increases in
infant mortality, and other detrimental effects (DeRose and Roff, 1999; Koeninger Ryan, et al., 2002; Ballou and Foose, 1996;
Reed and Frankham, 2003). Given the low level of inbreeding in the Sumatran orangutan population, the small number of
individuals with inbreeding values above 0 prevented us from statistically testing for possible effects on infant mortality. It is
unlikely that inbreeding is impacting the population at this time; however, inbreeding may become a threat in the future and
should therefore be avoided. Because modeling inbreeding depression adds an additional layer of complexity to
Table 8. Summary of starting genetic statistics for the population.
Percentage of pedigree known 100%
Genetic Diversity (GD) 97.58%
Population mean kinship (MK) 0.0242
Mean inbreeding (F) 0.0056
Mean generation time (T) (years) 24.7
Number of generations in 100 model years 4.0
a) b)
2014 Population Viability Analyses for AZA Orangutan Animal Program 15
interpretation of results, we have not included a “standard” inbreeding depression effect in the PVA models. There are
several strategies that can delay the effects or lower the probability of inbreeding depression including pairing based on
mean kinship and importing and breeding unrelated individuals (Ballou and Lacy, 1995). These strategies were modeled for
all scenarios except those without genetic management. Because inbreeding depression is not included, readers should
consider that model results may be optimistic if inbreeding depression begins to impact the population.
Management ZooRisk can include a space limitation on population growth, reducing breeding as the population approaches the potential
space limit. This mimics the way a population manager would recommend fewer pairs when at capacity. To accurately model
the “potential space” for a population, we use either a) the projected spaces in 5 years based on a TAG’s Regional Collection
Plan (RCP) or, if that value is unavailable or inappropriate, b) the current population size + 10% or 10 individuals, whichever is
greater. These values are also placed in context of current institutional interest by the Species Coordinator. The Ape TAG’s
most recent space survey set the target size for Sumatran orangutans over 4 years old at 85 individuals (AZA Ape Taxon
Advisory Group, 2007), however some additional spaces could become available if the population grows. Given this fact, and
because animals under 4 four years old still occupy space in ZooRisk, we allowed for 100 potential spaces in our model.
The AZA Sumatran orangutan population is currently designated as a Green Program. The formally managed population is
housed at 28 institutions, including one non-AZA institution (Toronto Zoo). Nineteen institutions have at least one potential
breeding pair.
Since the late 1980s, parent rearing has become the primary means of raising young Sumatran orangutans in zoos (Fig. 14).
Additionally, there has been an increasing trend in the interbirth intervals among females raising offspring. Managers intend
to have interbirth intervals no shorter than 8 years except when infant mortalities occur to ensure proper social development
of the offspring. If reproductively viable females in the population were to maintain an average interbirth interval of 8 years,
the female probability of breeding (p[B])in any given year would be equal to 12.5% (see appendix A for more information on
p[B]).
Figure 14. Interbirth intervals related to each rearing style over time.
0
5
10
15
20
25
30
1930 1940 1950 1960 1970 1980 1990 2000 2010
Inte
rbir
th In
terv
al (
Ye
ars)
Year
Parent Hand Foster
Supplemental Unknown Trend (Parent)
2014 Population Viability Analyses for AZA Orangutan Animal Program 16
SUMATRAN MODEL SCENARIOS
Model scenarios were created to reflect what would happen if current management approaches continued (baseline) and to
address potential alternate management strategies (Table 9). Model setup is described more fully in Appendix A. Scenarios
are described in more detail in the following sections.
Table 9. ZooRisk Model Scenarios
Scenario Name Scenario Description p(B) Spaces
Baseline Scenario:
E. Baseline; p(B) = 14.8% Reproduction to match past 10 years (2.8 births/year) 14.8% 100
Alternate Scenarios:
F. p(B) = 18% Reproduction required to sustain current population size 18% 100
G. p(B) = 25%; spaces = 115 Reproduction required to fill hybrid spaces as they become
available (i.e., approaching 115 spaces over the next 32 years) 25% 115
p(B) = Probability of Breeding
2014 Population Viability Analyses for AZA Orangutan Animal Program 17
SUMATRAN POPULATION VIABILITY UNDER CURRENT MANAGEMENT
BASELINE MODEL PVA RESULTS
We examined the viability of the formally managed
Sumatran orangutan population under current
management: 14.8% female probability of breeding
(equivalent to 2.8 births/year over the past 10 years) and no
imports or exports. The population currently has a target
size of 85 for individuals older than 4 years old, however
some additional spaces could become available if the
population grows. Because of this flexibility in the
availability of spaces, and the fact that animals younger than
4 years old hold space within our models, we allowed for
100 potential spaces in the baseline scenario. See Appendix
A for full details on all model parameters.
Under its current breeding rate, the Sumatran orangutan
population is expected to decline to approximately 50
individuals over the next 100 years (Fig. 15). However, the
population would retain high gene diversity (92.2%) and fairly low inbreeding (average inbreeding coefficient (F) = 0.049)
after 100 years, although inbreeding would begin approaching the equivalent of mating between first cousins (F = 0.063;
Table 10). The population is predicted to maintain a Low Risk status in zoos and aquariums over the next century (see
Appendix C).
Table 10. Baseline model results.
SCENARIO
DEMOGRAPHICS GENETICS
OVERALL
POPULATION
STATUS2
Initial
Population
Size
Size in
Year 251
Size in
Year
1001
Probability
of Reaching
Spaces (%)
Probability
of
Extinction
(%)
Initial
GD (%)
GD
Retained
in Year
1001
Initial F
F Retained
in Year
1001
E p(B) = 14.8% 85 67 ± 9 50 ± 20 6 0 97.6 92.2 ± 2.1 0.0056 0.049 ±
0.015 Low Risk
GD = gene diversity, F = inbreeding coefficient 1
Mean value + 1 Standard Deviation, taken across 1000 iterations. If an iteration goes extinct that value is not included in the calculation, so results may only reflect a few iterations
in scenarios with a high probability of extinction. 2ZooRisk uses five standardized tests to give a summary risk score for each scenario, from Low Risk (most secure), Vulnerable, Endangered, to Critical (least secure). For more details
on this score, see Appendix C.
0
20
40
60
80
100
1920 1950 1980 2010 2040 2070 2100
Nu
mb
er
of
Ind
ivid
ual
s
Year
E. Sumatran baseline; p(B) = 14.8%
Historic
Potential Space
Figure 15. Historical and projected mean population size under the baseline model scenario. Projected results are averaged across 1000 model iterations.
2014 Population Viability Analyses for AZA Orangutan Animal Program 18
SUMATRAN ALTERNATE MODEL SCENARIOS
We explored the impacts of altering the probability of breeding for the Sumatran orangutan population in order to maintain
the population at its current size or increase the population to fill all potential spaces. For the population to sustain its
current size of 85 individuals for the next 100 years (Fig. 16, Table 11), it needs to produce an average of ~4 births/year (p(B)
= 18%, scenario F; Fig. 17). Under this breeding rate, the population is expected to maintain high gene diversity (94.2%) and
fairly low inbreeding (F = 0.047) at 100 years, although inbreeding would begin approaching the equivalent of mating
between first cousins (F = 0.063). Note that a p(B) = 18% is approximately equivalent to reproductively viable females in the
population maintaining an average interbirth interval of 5.6 years. This interval could account for infant mortality, meaning
that females who lost infants would likely have shorter interbirth intervals while those that had surviving infants could have
longer ones.
We estimate that 115 potential spaces could be available for the Sumatran orangutan population after hybrids age out of AZA
zoos in approximately 32 years (scenario A; Table 1), which is equal to the population’s target size of 85 individuals plus ~8
spaces for offspring and half of spaces previously occupied by hybrids (~22 spaces). If the population produces an average of
~4 births/year over the next decade and ~5 births/year every following year, it could fill the 115 potential spaces in
approximately 33 years (p(B) = 25%, scenario G; Figs. 16 and 17). A p(B) = 25% is equivalent to every reproductively viable
female in the population maintaining an average interbirth interval of 4 years (which would be an average of intervals with
the first offspring living and the first offspring not surviving). In this management scenario, the population would maintain
high gene diversity (95.3%) and fairly low inbreeding (F = 0.040) over the next 100 years (Table 11). In either increase
breeding scenario, the Sumatran orangutan population is projected to retain a Low Risk status in zoos (Appendix C).
Table 11. Model results for all Sumatran orangutan scenarios.
SCENARIO
DEMOGRAPHICS GENETICS
OVERALL
POPULATION
STATUS2
Initial
Population
Size
Size in
Year 251
Size in
Year
1001
Probability
of Reaching
Spaces (%)
Probability
of
Extinction
(%)
Initial
GD (%)
GD
Retained
in Year
1001
Initial F
F Retained
in Year
1001
E p(B) = 14.8% 85 67 ± 9 50 ± 20 6 0 97.6 92.2 ± 2.1 0.0056 0.049 ±
0.015 Low Risk
F p(B) = 18% 85 80 ± 11 89 ± 14 74 0 97.6 94.2 ± 0.8 0.0056 0.047 ±
0.011 Low Risk
G p(B) = 25%;
spaces = 115 85 100 ± 9 114 ± 3 100 0 97.6 95.3 ± 0.2 0.0056
0.040 ±
0.008 Low Risk
GD = gene diversity, F = inbreeding coefficient 1
Mean value + 1 Standard Deviation, taken across 1000 iterations. If an iteration goes extinct that value is not included in the calculation, so results may only reflect a few iterations
in scenarios with a high probability of extinction. 2ZooRisk uses five standardized tests to give a summary risk score for each scenario, from Low Risk (most secure), Vulnerable, Endangered, to Critical (least secure). For more details
on this score, see Appendix C.
2014 Population Viability Analyses for AZA Orangutan Animal Program 19
MANAGEMENT ACTIONS The AZA Orangutan Animal Programs should consider the following changes to current management strategies:
Re-allocate spaces currently occupied by hybrids as they become available: Because spaces are expected to
become available as hybrid orangutans age out of AZA institutions, managers should ensure that these spaces are
allocated to Bornean and Sumatran orangutans to maintain each of these populations at a larger, stable size with
the best possible genetic health.
Increase breeding rates: For each orangutan population, breeding should be increased from ~3 births per year to a
minimum of ~4 births each year to maintain the current population size. To fill spaces currently occupied by hybrid
orangutans as they become available, the Bornean population should produce ~5 births per year. The Sumatran
population, however, should produce ~4 births each year for the next decade and ~5 births every following year in
order to do so. To achieve these higher breeding rates, it may not be possible to immediately implement the
recommended 8-year interbirth intervals. If higher breeding rates could be maintained, each population may
become demographically stable and maintain its genetic health.
CONCLUSIONS This model is a scientifically-sound comprehensive tool to be used by population managers for assessing future directions for
the animal program. This PVA report is provided to the AZA community and others to integrate into management of the
important species within our care. The PVA model results are intended to provide the necessary data to make science-based
decisions.
Our model results illustrate that, under current management practices, each orangutan population can be expected to
decline. Increasing breeding rates by approximately one or two births each year could ensure that each population maintains
a stable size over the next 100 years. Additionally, allocating spaces currently occupied by hybrids to Bornean and Sumatran
orangutans will help to better maintain the demographic and genetic stability of each of these populations. The AZA
Orangutan Animal Programs should follow these recommended management strategies in order to keep both orangutan
populations on the path towards long-term sustainability within AZA institutions.
0
20
40
60
80
100
120
1920 1950 1980 2010 2040 2070 2100
Nu
mb
er
of
Ind
ivid
ual
s
Year
F. Sumatran; p(B) = 18%
G. Sumatran; p(B) = 25%; space = 115
Figure 16. Historical and projected mean population size under alternate
model scenarios. Projected results are averaged across 1000 model
iterations.
0
1
2
3
4
5
6
2000 2020 2040 2060 2080 2100
Nu
mb
er
of
Hat
che
s
Year
F. Sumatran; p(B) = 18%
G. Sumatran; p(B) = 25%; space = 115
Figure 17. Number of total births in the population in the past decade and projected mean number of births under alternate model scenarios. Projected results are averaged across 1000 model iterations.
2014 Population Viability Analyses for AZA Orangutan Animal Program 20
ACKNOWLEDGEMENTS A meeting was conducted on June 19 2014 to discuss the Orangutan population and was attended by the following:
Brent Johnson, Population Biologist , Lincoln Park Zoo, [email protected]
Megan Elder, AZA Orangutan Animal Program Studbook Keeper, Como Park Zoo and Conservatory, [email protected]
Lauren Mechak, Assistant Population Biologist , Lincoln Park Zoo, [email protected]
Jennifer Michelberg, AZA Orangutan Population Advisor, [email protected]
Lori Perkins, AZA Bornean and Sumatran Orangutan SSP Coordinator, Zoo Atlanta, [email protected]
Adrienne Savrin, Research Assistant, Lincoln Park Zoo, [email protected]
Joseph L. Simonis, Post-doc at Alexander Center for Applied Population Biology, Lincoln Park Zoo, [email protected]
This report was also reviewed by:
Candice Dorsey, Director of Animal Programs, Association of Zoos and Aquariums, [email protected]
Megan Elder, AZA Orangutan Animal Program Studbook Keeper, Como Park Zoo and Conservatory, [email protected]
Lisa Faust, Vice President of Conservation and Science, Lincoln Park Zoo, [email protected]
Tracy Fenn, AZA Ape TAG Vice-Chair, Jacksonville Zoo and Gardens, [email protected]
Lauren Mechak, Assistant Population Biologist , Lincoln Park Zoo, [email protected]
Lori Perkins, AZA Bornean and Sumatran Orangutan SSP Coordinator, Zoo Atlanta, [email protected]
Adrienne Savrin, Research Assistant, Lincoln Park Zoo, [email protected]
Joseph L. Simonis, Post-doc at Alexander Center for Applied Population Biology, Lincoln Park Zoo, [email protected]
Tara Stoinski, AZA Ape TAG Chair, Zoo Atlanta, [email protected] If you have any questions about the report, please contact Lisa Faust at [email protected] Analyses in this report utilized the International Orangutan (Pongo pygmaeus + Pongo abelii) Studbook current to 26 April 2014 (Elder, 2014) and was performed using Poplink 2.4 and ZooRisk 3.8. Funding provided by Institute of Museum and Library Services (IMLS) LG-25-11-018 and MG-30-13-0065-13 to the Association of Zoos and Aquariums. Cover photo: courtesy of Megan Elder Citation: Johnson, B., Perkins, L., Elder, M., Stoinski, T., Fenn, T. Orangutan (Pongo pygmaeus + Pongo abelii) AZA Animal Program
Population Viability Analysis Report. Lincoln Park Zoo, Chicago, IL.
The contents of this report including opinions and interpretation of results are based on discussions between the project team
and do not necessarily reflect the opinion or position of Lincoln Park Zoo, Association of Zoos and Aquariums, and other
collaborating institutions. The population model and results are based on the project team’s best understanding of the
current biology and management of this population. They should not be regarded as absolute predictions of the population’s
future, as many factors may impact its future status.
2014 Population Viability Analyses for AZA Orangutan Animal Program 21
LITERATURE CITED AZA Ape Taxon Advisory Group (TAG). 2014. Ape Advisory Group Regional Collection Plan (RCP) – 3rd Edition.
Ballou, J. D., and R. C. Lacy. 1995. Identifying genetically important individuals for management of genetic variation in pedigreed populations. Pages 76-111 in J. D. Ballou, M. Gilpin, and T. J. Foose, eds. Population Management for Survival and Recovery: Analytical Methods and Strategies in Small Population Conservation. Columbia University Press, New York. Ballou, J. D., and T. J. Foose. 1996. Demographic and genetic management of captive populations. Pages 263-283 in S. Lumpkin, ed. Wild Mammals in Captivity. University of Chicago Press, Chicago.
DeRose, M. A., and D. A. Roff. 1999. A comparison of inbreeding depression in life-history and morphological traits in animals. Evolution 53: 1288-1292. Earnhardt, JM, Bergstrom, YM, Lin, A, Faust, LJ, Schloss, CA, and Thompson, SD. 2008. ZooRisk: A Risk Assessment Tool. Version 3.8. Lincoln Park Zoo. Chicago, IL. Faust, LJ, Earnhardt, JM, Schloss, CA, and Bergstrom, YM. 2008. ZooRisk: A Risk Assessment Tool. Version 3.8 User’s Manual. Lincoln Park Zoo. Chicago, IL. Michelberg, J., Perkins, L., Elder, M. Orangutan (Pongo pygmaeus + Pongo abelii) AZA Species Survival Plan® Green Program
2014 Population Analysis and Breeding & Transfer Plan. AZA Population Management Center, Chicago, IL. Zoo Atlanta,
Atlanta, GA.
Elder, M. 2014. Orangutan (Pongo pygmaeus + Pongo abelii) AZA Studbook. Como Park Zoo and Conservatory, St. Paul, MN. Reed, D. H., and R. Frankham. 2003. Correlation between fitness and genetic diversity. Conservation Biology 17: 230-237.
2014 Population Viability Analyses for AZA Orangutan Animal Program 22
DEFINITIONS Age Structure: A two-way classification showing the numbers or percentages of individuals in various age and sex classes. Current Gene Diversity (GD): The proportional gene diversity (as a proportion of the source population) is the probability that two alleles from the same locus sampled at random from the population will not be identical by descent. Gene diversity is calculated from allele frequencies, and is the heterozygosity expected in progeny produced by random mating, and if the population were in Hardy-Weinberg equilibrium. Founder: An individual obtained from a source population (often the wild) that has no known relationship to any individuals in the derived population (except for its own descendants). Inbreeding Coefficient (F): Probability that the two alleles at a genetic locus are identical by descent from an ancestor common to both parents. The mean inbreeding coefficient of a population will be the proportional decrease in observed heterozygosity relative to the expected heterozygosity of the founder population. Mean Kinship (MK): The mean kinship coefficient between an animal and all animals (including itself) in the living, zoo born population. The mean kinship of a population is equal to the proportional loss of gene diversity of the descendant (zoo born) population relative to the founders and is also the mean inbreeding coefficient of progeny produced by random mating. Mean kinship is also the reciprocal of two times the founder genome equivalents: MK = 1 / (2 * FGE). MK = 1 - GD. Mean Generation Time (T): The average time elapsing from reproduction in one generation to the time the next generation reproduces. Also, the average age at which a female (or male) produces offspring. It is not the age of first reproduction. Males and females often have different generation times. Percent Known: Percent of an animal's genome that is traceable to known Founders. Thus, if an animal has an UNK sire, the % Known = 50. If it has an UNK grandparent, % Known = 75% Population Viability Analysis (PVA): A PVA is a computer model that projects the likely future status of a population. PVAs are used for evaluating long-term sustainability, setting population goals, and comparing alternative management strategies. Several quantitative parameters are used in a PVA to calculate the extinction risk of a population, forecast the population’s future trajectory, and identify key factors impacting the population’s future. Potential Space: In the context of a Regional Collection Plan, the ‘potential space’ selected for each Program within the RCP, which may be based on available spaces for that species, desired spaces the TAG wishes to allocate, the size needed to maintain a viable population, or some combination of those factors. In the context of the ZooRisk modeling work, the potential space is a model parameter that can be set at any level, including the size listed in the RCP or a higher or lower size based on other criteria. Probability of Breeding [p(B)]: Female p(B) is the age-specific probability that a female will have at least one offspring in a year. For example, p(B) = 25% is equivalent to females producing an offspring once every 4 years. Within the reproductively viable age classes, all p(B) were set at a hypothetical constant value corresponding with an interbirth interval, which varied depending on the model scenario. Using a constant value means that all reproductively viable females would have the same chance of reproduction regardless of age. Qx, Mortality: Probability that an individual of age x dies during time period. Regional Collection Plan (RCP): document developed by Taxon Advisory Group (TAG) to describe species managed under their TAG, level of management with explanations, and evaluation of Target Population Sizes for each managed species. Risk (Qx or Mx): The number of individuals that have lived during an age class. The number at risk is used to calculate Mx and Qx by dividing the number of births and deaths that occurred during an age class by the number of animals at risk of dying and reproducing during that age class. The proportion of individuals that die during an age class is calculated from the number of animals that die during an age class divided by the number of animals that were alive at the beginning of the age class (i.e.-"at risk"). Stochastic Model: A model that includes random chance and variation in model parameters (e.g. randomly select if an individual will breed). Stochastic models will produce many different outcomes each time the model is run due to this variation. Models are typically run for many iterations to fully explore the trajectory a population might take. ZooRisk is a stochastic model. Taxon Advisory Group (TAG): There are several different TAGs and each oversees a broad group of animals (e.g. Antelope TAG, Small Carnivore TAG). Each TAG consists of several programs. The TAG contains experts including studbook keepers, program leaders, the TAG chair, and other advisors. TAGs evaluate the present conditions surrounding a broad group of animals (e.g., marine mammals) and then prioritize the different species in the group for possible captive programs.
2014 Population Viability Analyses for AZA Orangutan Animal Program 23
APPENDICES
APPENDIX A. MODEL SETUP AND METHODS Analyses in this report were performed using PopLink 2.4, ZooRisk 3.8, and R statistical package. Complete documentation on
ZooRisk’s modeling approach and setup can be found in the software manual (Faust et al., 2008). These tables document the
basic file information (Table A1) key model variables, give some context for how ZooRisk utilizes them in the model and how
the PVA modeling team applies them, give values used in model scenarios, and describe data sources (Tables A2-A4).
Additionally, age and sex specific mortality rates are documented in Tables A5-A7.
For clarity, most figures in this report show the mean population size across multiple iterations. Model results such as mean
population sizes, levels of gene diversity (GD), and inbreeding (F) are averaged across 1000 model iterations; if some model
iterations go extinct, these values are only averaged over extant, surviving iterations. Where relevant, results are reported
on medium-term (25 year), and long-term (100 year) time frames. Results such as the probability of reaching the potential
space or extinction are based on the percentage of iterations that hit that target at least once over the 1000 years. Where
applicable, ± 1 standard deviation is included; large values represent wider variability in model results.
Table A1. Information regarding studbook utilization in PopLink 2.4 and ZooRisk 3.8.
File Information
Studbook Currentness Date 26 April 2014
Studbook Name Orangutan2014
Studbook Keeper for analytical database Megan Elder
Studbook Keeper for demographic database Megan Elder
ZooRisk Project Name Orangutan_2014 Table A2. Variable used in ZooRisk 3.8 scenarios for hybrid orangutans, including a description of each variable, the value, and the source.
Model Variable Description/Details of How Variable is Used in Model Value in Model Scenarios
Source
Demographic Settings
Living Population
ZooRisk uses a starting population to initiate each model scenario, incorporating data on each individual’s pedigree, age, sex, and reproductive status. Any animals unable to breed due to age, medical issues, or sterilizations can be designated as non-reproductive in the model, which removes them from the potentially breeding population. The model assumes that any new animals (either births or imports) are potentially reproductively viable; this may be an optimistic assumption.
See included individuals in Appendix B, Table B1
Studbook data; animals in AZA as of 27 Apr 2014, UDF: SP = X
Male and Female Age-Specific Mortality Rates (Qx)
Probability that an individual of age x dies during time period. Each year, the model stochastically determines whether each individual lives or dies based on that individual’s age- and sex-specific Qx.
See Table A5. No modifications were made from extracted Qx values.
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = X
Infant Mortality Rates (Qx)
Mortality rate for infants 0-1 used in the model (as described above). Infant mortality is a vital rate that is sensitive to changes in husbandry and also may be a life stage that is vulnerable to inbreeding depression.
Male = 20% Female = 11%
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = X
Maximum Longevity
The maximum age individuals are allowed to live to in the model (if they haven’t died before that age). The model values were based on female SB# 942 and male # 941.
Male = 46 Female = 46
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = X
Reproductive Age Classes
Age classes in which females or males could potentially be paired for breeding in the model
Male 10 42 Female 1 42
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Annual Number of Offspring
When a female within the model is selected to reproduce in a given model year, ZooRisk uses these frequencies to stochastically determine the number of offspring she
1 offspring = 99% 2 = 1%
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = X
2014 Population Viability Analyses for AZA Orangutan Animal Program 24
produces. Note that this distribution will be different than a litter size distribution if a species can have multiple clutches/litters within a year.
Birth Sex Ratio (BSR)
The model stochastically assigns a sex for any offspring created in the model. We extract historical studbook data and test for bias towards males or females in the sex ratio using a χ2 test. This evaluates whether the population is significantly different than 50% males, 50% females. For this population, the extracted birth sex ratio was not significantly biased. The extracted value was 0.5612 (χ
2 =
2.9388, df = 1, p > 0.05).
0.5 Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = X
Female Probability of Breeding [p(B)]
P(B) is the age-specific probability that a female will have at least one offspring in a year. For example, p(B) = 0.25 is equivalent to females producing an offspring on average once every 4 years, or 25% of reproductively available females breeding in any given year. Historical studbook data include many females who never had access to a male and were non-reproductive for management rather than biological reasons. It is difficult to use extracted p(B) data to determine what a population could do if all individuals were in breeding situations or the population was truly trying to increase reproduction. Because of this, model scenarios use simplified hypothetical levels of p(B) to evaluate the impact of changes in reproduction. To set the baseline p(B), we extracted the average number of births/year over the past decade from the studbook and identified a p(B) level in the model that would produce, on average, an equivalent number of projected births over the first 10 model years. Alternate scenarios used higher or lower p(B) levels. In the model, all p(B) were set at the same value within all female reproductive age classes. Using a constant value means that all reproductively viable females would have the same chance of reproduction regardless of age. This population produced 0 births/year on average over the past decade, which was used to calibrate the baseline p(B).
0.00
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Genetic Settings
Genetic Management
ZooRisk can model a randomly breeding population or genetic management by mean kinship pairings and other genetic criteria, mimicking the way that AZA populations are managed to maintain gene diversity (GD) (Ballou & Lacy, 1995). This allows ZooRisk to more accurately project the amount of gene diversity retained through genetic management.
GM by mean kinship: ON
Modeling team decision
Inbreeding Depression
Inbreeding depression can be challenging to incorporate into PVA models because of uncertainty about which populations and life history traits will be affected, and at what inbreeding level they will become affected. Due to this uncertainty and since modeling inbreeding depression adds an additional layer of complexity to interpretation of results we have not included a “standard” inbreeding depression effect in the PVA models. The PVA includes management by mean kinship and measurements of final mean inbreeding levels to help users understand the potential future levels of inbreeding even with careful management. Readers should consider that model results
OFF Modeling team decision
2014 Population Viability Analyses for AZA Orangutan Animal Program 25
may be optimistic if this species would be susceptible to inbreeding depression now or in the future.
Breeding group ratio
ZooRisk can simulate breeding groups of 1 male: multiple females. If a population has few reproductive-aged animals and/or very few breeding age males, this can impact demographic results by limiting the number of pairs/groups that can be formed. It can also impact genetic results, as it influences how pairings by mean kinship are made (see manual for more details).
1 male: 1 female
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Number of Years Between Pairing
ZooRisk reshuffles the pairings with this frequency; a breeding group is left together for this number of years unless an individual group member dies, in which case another individual is shuffled in. A group is left together even if they may no longer be optimally paired by MK because of subsequent births.
28
Modeling team decision Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Other Model Settings
Potential Space
ZooRisk has the option of including a space limitation on population growth. This limitation reduces breeding in the model population as it approaches the space limit, mimicking zoo management. For example, a Program Leader may begin to recommend fewer breeding pairs if available spaces for a population become limited. To determine an appropriate space limitation for the models, the PVA team, in consultation with the AZA Wildlife Conservation and Management Committee (WCMC), developed the approach of using the number of projected spaces in 5 years based on a Taxon Advisory Group’s (TAG) Regional Collection Plan (RCP). If that number is unavailable or unsuitable (i.e. if the population is already close to or larger than that space), the team will use the current population size + 10% or 10 individuals, whichever is greater. In some instances when more information is available, i.e. a more recent survey by the Species Coordinator, we will use this value.
None Modeling team decision
Approach to space limit
The model can allow the population to grow/decline immediately to its space limit (if birth rates will allow it), but it is more likely that increases/decreases in space will occur gradually as new institutions join or leave a program.
Approach space limit gradually over 5 years
Modeling team decision
Number of Years How far into the future the model projects. 100 Modeling team decision
Number of Iterations
Since stochastic models have inherent variation, each model run (or iteration) will produce a slightly different population trajectory, and the model is run many times to reflect the full potential variation a population could experience.
500 Modeling team decision
Extinction Threshold
Size at which the population will be considered extinct. 0 Modeling team decision
Importations/ Exportations
ZooRisk can model removal or addition of individuals into the population. An importation event will bring a specific number of individuals (of a specified sex and age) that are completely unrelated to the current population (potential founders) into the population in a specified year. These might be individuals coming from the wild or from non-AZA institutions in other regions or in North America.
Exportations can be used to model reintroductions into the wild or transfer of individuals outside of the AZA population. ZooRisk can model two types – a simple export that selects a specific number of individuals of the designated sex and age classes in the specified year, or a threshold export that will select the 'extra' individuals above some threshold to export in the specified year.
0 imports and 0 exports
Studbook data, filters = AZA, 1 Jan 2004 1 Dec 2013, UDF: SP = X
2014 Population Viability Analyses for AZA Orangutan Animal Program 26
Table A3. Variable used in ZooRisk 3.8 scenarios for Bornean orangutans, including a description of each variable, the value, and the source.
Model Variable Description/Details of How Variable is Used in Model Value in Model Scenarios
Source
Demographic Settings
Living Population
ZooRisk uses a starting population to initiate each model scenario, incorporating data on each individual’s pedigree, age, sex, and reproductive status. Any animals unable to breed due to age, medical issues, or sterilizations can be designated as non-reproductive in the model, which removes them from the potentially breeding population. The model assumes that any new animals (either births or imports) are potentially reproductively viable; this may be an optimistic assumption.
See included individuals in Appendix B, Table B2
Studbook data; animals in AZA as of 27 Apr 2014, UDF: SP = B
Male and Female Age-Specific Mortality Rates (Qx)
Probability that an individual of age x dies during time period. Each year, the model stochastically determines whether each individual lives or dies based on that individual’s age- and sex-specific Qx.
See Table A6. No modifications were made from extracted Qx values.
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = B
Infant Mortality Rates (Qx)
Mortality rate for infants 0-1 used in the model (as described above). Infant mortality is a vital rate that is sensitive to changes in husbandry and also may be a life stage that is vulnerable to inbreeding depression.
Male = 25% Female = 13%
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = B
Maximum Longevity
The maximum age individuals are allowed to live to in the model (if they haven’t died before that age). The model value for female was based on SB#s 449, 660, and 510, all of which are dead. The value for males was estimated in consultation with population managers.
Male = 43 Female = 52
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = B Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Reproductive Age Classes
Age classes in which females or males could potentially be paired for breeding in the model
Male 10 42 Female 1 42
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Annual Number of Offspring
When a female within the model is selected to reproduce in a given model year, ZooRisk uses these frequencies to stochastically determine the number of offspring she produces. Note that this distribution will be different than a litter size distribution if a species can have multiple clutches/litters within a year.
1 offspring = 99% 2 = 1%
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = B
Birth Sex Ratio (BSR)
The model stochastically assigns a sex for any offspring created in the model. We extract historical studbook data and test for bias towards males or females in the sex ratio using a χ2 test. This evaluates whether the population is significantly different than 50% males, 50% females. For this population, the extracted birth sex ratio was not significantly biased. The extracted value was 0.5612 (χ
2 =
2.9388, df = 1, p > 0.05).
0.5 Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = B
Female Probability of Breeding [p(B)]
P(B) is the age-specific probability that a female will have at least one offspring in a year. For example, p(B) = 0.25 is equivalent to females producing an offspring on average once every 4 years, or 25% of reproductively available females breeding in any given year. Historical studbook data include many females who never had access to a male and were non-reproductive for management rather than biological reasons. It is difficult to use extracted p(B) data to determine what a population could do if all individuals were in breeding situations or the population was truly trying to increase reproduction. Because of this, model scenarios use simplified hypothetical levels of p(B) to evaluate the impact of changes in
0.115 (Baseline) Alternate scenarios = 0.19, 0.25
Studbook data, filters = AZA, 1 Jan 2004 1 Dec 2013, UDF: SP = B
2014 Population Viability Analyses for AZA Orangutan Animal Program 27
reproduction. To set the baseline p(B), we extracted the average number of births/year over the past decade from the studbook and identified a p(B) level in the model that would produce, on average, an equivalent number of projected births over the first 10 model years. Alternate scenarios used higher or lower p(B) levels. In the model, all p(B) were set at the same value within all female reproductive age classes. Using a constant value means that all reproductively viable females would have the same chance of reproduction regardless of age. This population produced 2.7 births/year on average over the past decade, which was used to calibrate the baseline p(B).
Genetic Settings
Genetic Management
ZooRisk can model a randomly breeding population or genetic management by mean kinship pairings and other genetic criteria, mimicking the way that AZA populations are managed to maintain gene diversity (GD) (Ballou & Lacy, 1995). This allows ZooRisk to more accurately project the amount of gene diversity retained through genetic management.
GM by mean kinship: ON
Modeling team decision
Inbreeding Depression
Inbreeding depression can be challenging to incorporate into PVA models because of uncertainty about which populations and life history traits will be affected, and at what inbreeding level they will become affected. Due to this uncertainty and since modeling inbreeding depression adds an additional layer of complexity to interpretation of results we have not included a “standard” inbreeding depression effect in the PVA models. The PVA includes management by mean kinship and measurements of final mean inbreeding levels to help users understand the potential future levels of inbreeding even with careful management. Readers should consider that model results may be optimistic if this species would be susceptible to inbreeding depression now or in the future.
OFF Modeling team decision
Breeding group ratio
ZooRisk can simulate breeding groups of 1 male: multiple females. If a population has few reproductive-aged animals and/or very few breeding age males, this can impact demographic results by limiting the number of pairs/groups that can be formed. It can also impact genetic results, as it influences how pairings by mean kinship are made (see manual for more details).
1 male: 1 female
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Number of Years Between Pairing
ZooRisk reshuffles the pairings with this frequency; a breeding group is left together for this number of years unless an individual group member dies, in which case another individual is shuffled in. A group is left together even if they may no longer be optimally paired by MK because of subsequent births.
28
Modeling team decision Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Other Model Settings
Approach to space limit
The model can allow the population to grow/decline immediately to its space limit (if birth rates will allow it), but it is more likely that increases/decreases in space will occur gradually as new institutions join or leave a program.
Approach space limit gradually over 5 years
Modeling team decision
2014 Population Viability Analyses for AZA Orangutan Animal Program 28
Potential Space
ZooRisk has the option of including a space limitation on population growth. This limitation reduces breeding in the model population as it approaches the space limit, mimicking zoo management. For example, a Program Leader may begin to recommend fewer breeding pairs if available spaces for a population become limited. To determine an appropriate space limitation for the models, the PVA team, in consultation with the AZA Wildlife Conservation and Management Committee (WCMC), developed the approach of using the number of projected spaces in 5 years based on a Taxon Advisory Group’s (TAG) Regional Collection Plan (RCP). If that number is unavailable or unsuitable (i.e. if the population is already close to or larger than that space), the team will use the current population size + 10% or 10 individuals, whichever is greater. In some instances when more information is available, i.e. a more recent survey by the Species Coordinator, we will use this value. The 2007 Ape TAG RCP set the desired spaces for Bornean orangutans over 4 years old at 85. However, animals younger than 4 hold spaces in the model, and some more spaces may become available for Bornean orangutans if the population grows slightly.
100 Alternate scenario = 115
Modeling team decision Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Number of Years How far into the future the model projects. 100 Modeling team decision
Number of Iterations
Since stochastic models have inherent variation, each model run (or iteration) will produce a slightly different population trajectory, and the model is run many times to reflect the full potential variation a population could experience.
500 Modeling team decision
Extinction Threshold
Size at which the population will be considered extinct. 0 Modeling team decision
Importations/Exportations
ZooRisk can model removal or addition of individuals into the population. An importation event will bring a specific number of individuals (of a specified sex and age) that are completely unrelated to the current population (potential founders) into the population in a specified year. These might be individuals coming from the wild or from non-AZA institutions in other regions or in North America. Exportations can be used to model reintroductions into the wild or transfer of individuals outside of the AZA population. ZooRisk can model two types – a simple export that selects a specific number of individuals of the designated sex and age classes in the specified year, or a threshold export that will select the 'extra' individuals above some threshold to export in the specified year.
p(B) calibration = Imports: 1:1 ratio of males to females ages 10-30, mimicking amounts from the past decade: 2009- 4 Scenarios = 0 imports and 0 exports
Studbook data, filters = AZA, 1 Jan 2004 1 Dec 2013, UDF: SP = B Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Table A4. Variable used in ZooRisk 3.8 scenarios for Sumatran orangutans, including a description of each variable, the value, and the source.
Model Variable Description/Details of How Variable is Used in Model Value in Model Scenarios
Source
Demographic Settings
Living Population
ZooRisk uses a starting population to initiate each model scenario, incorporating data on each individual’s pedigree, age, sex, and reproductive status. Any animals unable to breed due to age, medical issues, or sterilizations can be designated as non-reproductive in the model, which removes them from the potentially breeding population. The model assumes that any new animals (either births or imports) are potentially reproductively viable; this may be
See included individuals in Appendix B, table B3
Studbook data; animals in AZA as of 27 Apr 2014, UDF: SP = S
2014 Population Viability Analyses for AZA Orangutan Animal Program 29
an optimistic assumption.
Male and Female Age-Specific Mortality Rates (Qx)
Probability that an individual of age x dies during time period. Each year, the model stochastically determines whether each individual lives or dies based on that individual’s age- and sex-specific Qx.
See Table A7. No modifications were made from extracted Qx values.
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = S
Infant Mortality Rates (Qx)
Mortality rate for infants 0-1 used in the model (as described above). Infant mortality is a vital rate that is sensitive to changes in husbandry and also may be a life stage that is vulnerable to inbreeding depression.
Male = 23% Female = 22%
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = S
Maximum Longevity
The maximum age individuals are allowed to live to in the model (if they haven’t died before that age). The model value for female was based on SB#s 3 and 366, both of which are dead. The value for males was estimated in consultation with population managers.
Male = 50 Female = 57
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = S
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Reproductive Age Classes
Age classes in which females or males could potentially be paired for breeding in the model
Male 10 45 Female 1 45
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Annual Number of Offspring
When a female within the model is selected to reproduce in a given model year, ZooRisk uses these frequencies to stochastically determine the number of offspring she produces. Note that this distribution will be different than a litter size distribution if a species can have multiple clutches/litters within a year.
1 offspring = 99% 2 = 1%
Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = S
Birth Sex Ratio (BSR)
The model stochastically assigns a sex for any offspring created in the model. We extract historical studbook data and test for bias towards males or females in the sex ratio using a χ2 test. This evaluates whether the population is significantly different than 50% males, 50% females. For this population, the extracted birth sex ratio was not significantly biased. The extracted value was 0.561 (χ
2 =
3.2190, df = 1, p > 0.05).
0.5 Studbook data, filters = AZA, 1 an 1 8 27 Apr 2014, UDF: SP = S
Female Probability of Breeding [p(B)]
P(B) is the age-specific probability that a female will have at least one offspring in a year. For example, p(B) = 0.25 is equivalent to females producing an offspring on average once every 4 years, or 25% of reproductively available females breeding in any given year. Historical studbook data include many females who never had access to a male and were non-reproductive for management rather than biological reasons. It is difficult to use extracted p(B) data to determine what a population could do if all individuals were in breeding situations or the population was truly trying to increase reproduction. Because of this, model scenarios use simplified hypothetical levels of p(B) to evaluate the impact of changes in reproduction. To set the baseline p(B), we extracted the average number of births/year over the past decade from the studbook and identified a p(B) level in the model that would produce, on average, an equivalent number of projected births over the first 10 model years. Alternate scenarios used higher or lower p(B) levels. In the model, all p(B) were set at the same value within all female reproductive age classes. Using a constant value means that all reproductively viable females would have
0.148 (Baseline) Alternate scenarios = 0.18, 0.25
Studbook data, filters = AZA, 1 Jan 2004 1 Dec 2013, UDF: SP = S
2014 Population Viability Analyses for AZA Orangutan Animal Program 30
the same chance of reproduction regardless of age. This population produced 2.8 births/year on average over the past decade, which was used to calibrate the baseline p(B).
Genetic Settings
Genetic Management
ZooRisk can model a randomly breeding population or genetic management by mean kinship pairings and other genetic criteria, mimicking the way that AZA populations are managed to maintain gene diversity (GD) (Ballou & Lacy, 1995). This allows ZooRisk to more accurately project the amount of gene diversity retained through genetic management.
GM by mean kinship: ON
Modeling team decision
Inbreeding Depression
Inbreeding depression can be challenging to incorporate into PVA models because of uncertainty about which populations and life history traits will be affected, and at what inbreeding level they will become affected. Due to this uncertainty and since modeling inbreeding depression adds an additional layer of complexity to interpretation of results we have not included a “standard” inbreeding depression effect in the PVA models. The PVA includes management by mean kinship and measurements of final mean inbreeding levels to help users understand the potential future levels of inbreeding even with careful management. Readers should consider that model results may be optimistic if this species would be susceptible to inbreeding depression now or in the future.
OFF Modeling team decision
Breeding group ratio
ZooRisk can simulate breeding groups of 1 male: multiple females. If a population has few reproductive-aged animals and/or very few breeding age males, this can impact demographic results by limiting the number of pairs/groups that can be formed. It can also impact genetic results, as it influences how pairings by mean kinship are made (see manual for more details).
1 male: 1 female
Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Number of Years Between Pairing
ZooRisk reshuffles the pairings with this frequency; a breeding group is left together for this number of years unless an individual group member dies, in which case another individual is shuffled in. A group is left together even if they may no longer be optimally paired by MK because of subsequent births.
31
Modeling team decision Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Other Model Settings
Potential Space
ZooRisk has the option of including a space limitation on population growth. This limitation reduces breeding in the model population as it approaches the space limit, mimicking zoo management. For example, a Program Leader may begin to recommend fewer breeding pairs if available spaces for a population become limited. To determine an appropriate space limitation for the models, the PVA team, in consultation with the AZA Wildlife Conservation and Management Committee (WCMC), developed the approach of using the number of projected spaces in 5 years based on a Taxon Advisory Group’s (TAG) Regional Collection Plan (RCP). If that number is unavailable or unsuitable (i.e. if the population is already close to or larger than that space), the team will use the current population size + 10% or 10 individuals, whichever is greater. In some instances when more information is available, i.e. a more recent survey by the Species Coordinator, we will use this value.
100 Alternate scenario = 115
Modeling team decision Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
2014 Population Viability Analyses for AZA Orangutan Animal Program 31
The 2007 Ape TAG RCP set the desired spaces for Bornean orangutans over 4 years old at 85. However, animals younger than 4 hold spaces in the model, and some more spaces may become available for Sumatran orangutans if the population grows slightly.
Approach to space limit
The model can allow the population to grow/decline immediately to its space limit (if birth rates will allow it), but it is more likely that increases/decreases in space will occur gradually as new institutions join or leave a program.
Approach space limit gradually over 5 years
Modeling team decision
Number of Years How far into the future the model projects. 100 Modeling team decision
Number of Iterations
Since stochastic models have inherent variation, each model run (or iteration) will produce a slightly different population trajectory, and the model is run many times to reflect the full potential variation a population could experience.
500 Modeling team decision
Extinction Threshold
Size at which the population will be considered extinct. 0 Modeling team decision
Importations/Exportations
ZooRisk can model removal or addition of individuals into the population. An importation event will bring a specific number of individuals (of a specified sex and age) that are completely unrelated to the current population (potential founders) into the population in a specified year. These might be individuals coming from the wild or from non-AZA institutions in other regions or in North America. Exportations can be used to model reintroductions into the wild or transfer of individuals outside of the AZA population. ZooRisk can model two types – a simple export that selects a specific number of individuals of the designated sex and age classes in the specified year, or a threshold export that will select the 'extra' individuals above some threshold to export in the specified year.
p(B) calibration = 1:1 ratio of males to females ages 10-30, mimicking amounts from the past decade: Imports: 2009- 1 2010- 1 Exports: 2004- 1 2005- 1 Scenarios = 0 imports and 0 exports
Studbook data, filters = AZA, 1 Jan 2004 1 Dec 2013, UDF: SP = S Modeling team consulted the following sources: Program Leader, Studbook Keeper, Population Advisor
Table A5. Male and female mortality rates used in all hybrid orangutan scenarios of the ZooRisk model are listed below. Each year, the model determines whether each individual lives or dies stochastically based on that individuals age- and sex- specific mortality rate. Male Female
Age(x) Qx Number at Risk Age(x) Qx Number at Risk
0 0.2 118 0 0.11 96
1 0 92.5 1 0.11 83.5
2 0 91 2 0 87
3 0 93 3 0 91
4 0 93 4 0 90
5 0 96 5 0 91
6 0 101 6 0 95
7 0.04 100 7 0 98
8 0.04 97 8 0 100
9 0 94 9 0 100
10 0 97 10 0 102
11 0.04 94 11 0 99
12 0 91 12 0 101
13 0.12 95 13 0 104
14 0.05 91 14 0.04 103
15 0 89 15 0.04 97
16 0.05 90 16 0 102
17 0 92 17 0.04 106
18 0 92 18 0 99
19 0 88 19 0 97
20 0 88 20 0 102
21 0 84 21 0 105
22 0.05 81 22 0 104
2014 Population Viability Analyses for AZA Orangutan Animal Program 32
23 0 78 23 0 101
24 0 78 24 0 106
25 0 76 25 0 105
26 0 72 26 0 104
27 0 66 27 0 102
28 0.08 64 28 0 102
29 0 52 29 0.05 100
30 0 50 30 0 93
31 0 50 31 0 84
32 0 49 32 0 77
33 0 42 33 0.07 70
34 0 38 34 0 59
35 0.12 31 35 0 56
36 0.2 24 36 0 54
37 0 17 37 0 48
38 0 14 38 0.12 43
39 0 11 39 0 39
40 0 8 40 0 35
41 0 8 41 0 33
42 0 6 42 0 31
43 0 3 43 0 27
44 0 2 44 0 24
45 0 1 45 0 20
46 1 1 46 1 15
47 0 11
48 0 11
49 0 9
50 0 9
51 0 8
52 0 7
53 0 4
54 0 3
55 0 2
56 0 2
57 0 1 Table A6. Male and female mortality rates used in all Bornean orangutan scenarios of the ZooRisk model are listed below. Each year, the model determines whether each individual lives or dies stochastically based on that individuals age- and sex- specific mortality rate. Male Female
Age(x) Qx Number at Risk Age(x) Qx Number at Risk
0 0.25 118 0 0.13 96
1 0 92.5 1 0.03 83.5
2 0.03 91 2 0 87
3 0 93 3 0.02 91
4 0.03 93 4 0 90
5 0.03 96 5 0.02 91
6 0 101 6 0 95
7 0 100 7 0.02 98
8 0.03 97 8 0 100
9 0 94 9 0 100
10 0 97 10 0.03 102
11 0.07 94 11 0 99
12 0.04 91 12 0 101
13 0 95 13 0 104
14 0 91 14 0 103
15 0.04 89 15 0 97
16 0.04 90 16 0 102
17 0 92 17 0.03 106
18 0.06 92 18 0 99
2014 Population Viability Analyses for AZA Orangutan Animal Program 33
19 0.07 88 19 0 97
20 0.07 88 20 0 102
21 0.07 84 21 0.03 105
22 0 81 22 0.03 104
23 0 78 23 0.06 101
24 0.03 78 24 0.03 106
25 0.04 76 25 0 105
26 0.04 72 26 0 104
27 0.08 66 27 0.03 102
28 0.05 64 28 0.03 102
29 0.1 52 29 0 100
30 0.11 50 30 0.09 93
31 0.06 50 31 0.07 84
32 0.06 49 32 0 77
33 0.21 42 33 0.04 70
34 0.18 38 34 0 59
35 0.12 31 35 0 56
36 0.16 24 36 0.06 54
37 0.2 17 37 0.06 48
38 0.5 14 38 0.07 43
39 0.5 11 39 0.08 39
40 0 8 40 0.09 35
41 0 8 41 0 33
42 0 6 42 0.2 31
43 1 3 43 0.12 27
44 0 24
45 0.31 20
46 0.25 15
47 0 11
48 0 11
49 0 9
50 0 9
51 0 8
52 1 7 Table A7. Male and female mortality rates used in all Sumatran orangutan scenarios of the ZooRisk model are listed below. Each year, the model determines whether each individual lives or dies stochastically based on that individuals age- and sex- specific mortality rate. Male Female
Age(x) Qx Number at Risk Age(x) Qx Number at Risk
0 0.23 127 0 0.22 101
1 0.05 98.5 1 0 87.5
2 0 97 2 0 91
3 0 99 3 0 95
4 0 99 4 0 94
5 0 102 5 0 95
6 0 107 6 0 99
7 0.05 106 7 0 102
8 0.03 102 8 0.03 104
9 0.03 97 9 0 104
10 0 99 10 0.02 106
11 0.05 95 11 0 103
12 0 92 12 0.02 104
13 0.03 96 13 0.05 107
14 0 92 14 0.07 106
15 0.03 90 15 0 100
16 0 91 16 0 105
17 0.05 94 17 0.05 109
18 0 93 18 0 102
19 0.06 89 19 0 100
2014 Population Viability Analyses for AZA Orangutan Animal Program 34
20 0.12 89 20 0 105
21 0.03 85 21 0.02 108
22 0.04 82 22 0.05 106
23 0.04 79 23 0 103
24 0 79 24 0.04 109
25 0.07 78 25 0.02 108
26 0.07 74 26 0.02 107
27 0.04 68 27 0.02 105
28 0.04 66 28 0 105
29 0 54 29 0.02 102
30 0 52 30 0.05 95
31 0 52 31 0.03 86
32 0.17 51 32 0.09 79
33 0 43 33 0.07 72
34 0.17 39 34 0 61
35 0.15 32 35 0 58
36 0.09 25 36 0 56
37 0.1 18 37 0.04 50
38 0 15 38 0 45
39 0.25 12 39 0.05 41
40 0 9 40 0 37
41 0.33 9 41 0.05 35
42 0 7 42 0 32
43 0.33 4 43 0 28
44 0.5 3 44 0.06 25
45 0 2 45 0.15 21
46 0 2 46 0 16
47 0 1 47 0 12
48 0 1 48 0.25 11
49 0 1 49 0 9
50 1 1 50 0 9
51 0.2 8
52 0 7
53 0.25 4
54 0 3
55 0 2
56 0.5 2
57 1 1
2014 Population Viability Analyses for AZA Orangutan Animal Program 35
APPENDIX B. INCLUDED INDIVIDUALS As of April 26, 2014 (the studbook currentness date):
The Hybrid Orangutan population consisted of 45 individuals (21 males, 24 females, 0 unknown) (Table B1).
The AZA Bornean Orangutan population consisted of 84 individuals (33 males, 51 females, 0 unknown) (Table B2). Five
individuals are excluded from the breeding population. Those animals with “Allowed to Breed = NO” hold space but are never
eligible for reproduction. This leaves a potentially breeding population of 79 (33.46.0) individuals.
The AZA Sumatran Orangutan population consisted of 85 individuals (36 males, 48 females, 1 unknown) (Table B3).
Seventeen individuals are excluded from the breeding population. Those animals with “Allowed to Breed = NO” hold space
but are never eligible for reproduction. This leaves a potentially breeding population of 68 (34.33.1) individuals.
Table B1. Individuals included in the starting Hybrid population
Studbook ID Sex % Known Age Institution Allowed to Breed Reason For Exclusion
1018 Neutered Female
100 44 LITTLEROC NO Hybrid
1182 Male 100 42 TOLEDO NO Hybrid
1196 Female 100 42 SEATTLE NO Hybrid
1235 Neutered Female
100 41 HOUSTON NO Hybrid
1289 Neutered Female
100 41 NZP-WASH NO Hybrid
1335 Neutered Female
100 40 FORTWORTH NO Hybrid
1402 Neutered Female
100 39 BUSCH TAM NO Hybrid
1504 Neutered Male
100 38 DENVER NO Hybrid
1539 Neutered Female
100 37 ST PAUL NO Hybrid
1545 Neutered Female
100 37 NZP-WASH NO Hybrid
1602 Neutered Female
100 36 HONOLULU NO Hybrid
1614 Neutered Male
100 36 HOUSTON NO Hybrid
1616 Neutered Male
100 36 INDIANAPL NO Hybrid
1617 Neutered Male
100 36 ATLANTA NO Hybrid
1723 Neutered Male
100 35 LOSANGELE NO Hybrid
1733 Neutered Female
100 34 INDIANAPL NO Hybrid
1756 Neutered Male
100 34 HONOLULU NO Hybrid
1769 Neutered Female
100 33 NORFOLK NO Hybrid
1781 Neutered Female
100 33 MILWAUKEE NO Hybrid
1832 Female 100 32 SEATTLE NO Hybrid
1872 Neutered Male
100 32 MILWAUKEE NO Hybrid
1883 Neutered Female
100 31 LEON NO Hybrid
2014 Population Viability Analyses for AZA Orangutan Animal Program 36
1886 Neutered Male
100 31 NORFOLK NO Hybrid
1920 Neutered Female
100 31 FRESNO NO Hybrid
1972 Female 100 30 INDIANAPL NO Hybrid
1973 Female 100 28 RACINE NO Hybrid
1981 Female 100 29 INDIANAPL NO Hybrid
1987 Neutered Female
100 29 FT WAYNE NO Hybrid
2007 Neutered Female
100 29 LEON NO Hybrid
2021 Neutered Male
100 28 LITTLEROC NO Hybrid
2064 Neutered Male
100 28 RACINE NO Hybrid
2066 Neutered Male
100 27 WAHPETON NO Hybrid
2121 Female 100 27 NZP-WASH NO Hybrid
2127 Neutered Male
100 26 WACO NO Hybrid
2132 Neutered Male
100 26 NZP-WASH NO Hybrid
2138 Neutered Female
100 26 LOUISVILL NO Hybrid
2139 Neutered Male
100 26 LOUISVILL NO Hybrid
2248 Female 100 25 INDIANAPL NO Hybrid
2255 Neutered Male
100 25 SEATTLE NO Hybrid
2431 Neutered Male
100 23 CHICAGOBR NO Hybrid
2655 Male 100 20 INDIANAPL NO Hybrid
2905 Neutered Male
100 15 BUSCH TAM NO Hybrid
3331 Male 100 9 INDIANAPL NO Hybrid
941 Neutered Male
100 46 SEATTLE NO Hybrid
942 Female 100 46 SEATTLE NO Hybrid
Table B2. Individuals included in the starting Bornean population
Studbook ID Sex % Known Age Institution Allowed to Breed Reason For Exclusion
449 Female 100 52 CHICAGOBR NO Age
660 Female 100 52 SANDIEGOZ NO Age
971 Female 100 45 LOSANGELE NO Age; Health
1387 Female 100 39 COLUMBUS YES
1515 Neutered Female
100 37 KANSASCTY NO Sterile
1621 Female 100 36 ROCHESTER YES
1635 Male 100 36 BUSCH TAM YES
1640 Female 100 36 TOLEDO YES
1671 Male 100 35 CHICAGOBR YES
1683 Female 100 35 JACKSON YES
1704 Female 100 35 PHOENIX YES
1753 Female 100 34 LOWRY YES
1789 Female 100 33 BUSCH TAM YES
1793 Female 100 33 HOUSTON YES
1813 Female 100 33 PITTSBURG YES
1817 Female 100 33 CHICAGOBR YES
1824 Female 100 33 LOSANGELE YES
1865 Neutered 100 32 TOPEKA NO Sterile
2014 Population Viability Analyses for AZA Orangutan Animal Program 37
Female
1887 Female 100 31 LOSANGELE YES
2023 Male 100 28 JACKSON YES
2025 Female 100 29 LOWRY YES
2037 Female 100 28 TOPEKA YES
2048 Female 100 29 KANSASCTY YES
2062 Male 100 28 ERIE YES
2063 Male 100 28 CLEVELAND YES
2083 Female 100 27 CLEVELAND YES
2120 Male 100 27 PHOENIX YES
2175 Male 100 25 KANSASCTY YES
2176 Male 100 25 WACO YES
2201 Male 100 25 TOPEKA YES
2270 Male 100 24 GREENVISC YES
2274 Male 100 24 TOLEDO YES
2354 Male 100 24 HOGLE YES
2374 Female 100 23 GREENVISC YES
2379 Female 100 23 HOGLE YES
2500 Female 100 22 ATLANTA YES
2503 Female 100 22 ERIE YES
2602 Male 100 21 OMAHA YES
2615 Female 100 20 TOLEDO YES
2626 Male 100 20 COLUMBUS YES
2658 Male 100 19 COLO SPRG YES
2755 Female 100 17 COLO SPRG YES
2757 Male 100 17 NZP-WASH YES
2758 Female 100 17 NZP-WASH YES
2801 Female 100 16 BUSCH TAM YES
2851 Male 100 16 OMAHA YES
2852 Female 100 15 WACO YES
2853 Female 100 15 CLEVELAND YES
2902 Male 100 14 KANSASCTY YES
2906 Male 100 14 LOWRY YES
2908 Female 100 14 OMAHA YES
2923 Female 100 14 NORFOLK YES
3000 Female 100 13 MADISON YES
3003 Female 100 13 CLEVELAND YES
3005 Female 100 12 OMAHA YES
3053 Female 100 11 KANSASCTY YES
3054 Male 100 11 ROCHESTER YES
3103 Female 100 11 TOLEDO YES
3104 Female 100 10 OMAHA YES
3105 Male 100 10 HOUSTON YES
3108 Male 100 10 ATLANTA YES
3151 Male 100 10 MADISON YES
3166 Male 100 10 TOLEDO YES
3200 Female 100 9 LOSANGELE YES
3201 Female 100 9 TOPEKA YES
3203 Female 100 8 HOGLE YES
3226 Female 100 8 LOWRY YES
3251 Male 100 8 PITTSBURG YES
3252 Male 100 8 GREENVISC YES
3253 Female 100 8 PHOENIX YES
3254 Male 100 8 OMAHA YES
3255 Male 100 8 PHOENIX YES
3256 Female 100 7 TOLEDO YES
3258 Male 100 7 OMAHA YES
3334 Female 100 5 LOWRY YES
3335 Female 100 5 CHICAGOBR YES
2014 Population Viability Analyses for AZA Orangutan Animal Program 38
3350 Female 100 5 KANSASCTY YES
3370 Male 100 4 ERIE YES
3428 Female 100 3 HOUSTON YES
3436 Female 100 2 LOSANGELE YES
3520 Male 100 1 TOPEKA YES
3536 Female 100 0 ROCHESTER YES
3545 Male 100 0 ATLANTA YES
3562 Male 100 0 TOLEDO YES
Table B3. Individuals included in the starting Sumatran population
Studbook ID Sex % Known Age Institution Allowed to Breed Reason For Exclusion
366 Female 100 57 SEDGWICK NO Age
433 Female 100 54 PORTLAND NO Age
881 Female 100 46 OKLAHOMA NO Age
915 Neutered Female
100 46 DENVER NO Age; Sterile
1020 Female 100 44 ST LOUIS YES
1042 Female 100 44 RIO GRAND NO Age; Sterile
1044 Female 100 50 BROWNSVIL NO Age; Sterile
1100 Female 100 43 FRESNO YES
1106 Female 100 43 ATLANTA NO Age
1145 Male 100 42 ATLANTA NO Age; Sterile
1164 Female 100 47 TORONTO YES
1209 Female 100 42 SACRAMNTO NO Age
1455 Female 100 38 FORTWORTH NO Health
1529 Male 100 37 ROLLING H YES
1576 Female 100 36 MEMPHIS NO Behavior
1587 Female 100 36 BROWNSVIL YES
1703 Male 100 35 ATLANTA YES
1714 Male 100 34 RIO GRAND YES
1765 Female 100 34 ROLLING H NO Sterile
1773 Male 100 33 BIRMINGHM YES
1841 Female 100 32 COLUMBUS NO Health
1850 Female 100 32 SEDGWICK YES
1882 Male 100 31 MEMPHIS YES
1904 Female 100 32 BIRMINGHM YES
1924 Female 100 30 ATLANTA YES
1926 Female 100 30 RIO GRAND YES
1932 Male 100 30 DENVER YES
1971 Male 100 29 FRESNO YES
1980 Female 100 29 LOUISVILL YES
2015 Female 100 29 METROZOO YES
2016 Female 100 29 CINCINNAT NO Management exclusion
2018 Male 100 28 ST PAUL YES
2024 Female 100 28 TORONTO YES
2034 Male 100 28 EL PASO YES
2069 Male 100 27 FT WAYNE YES
2070 Female 100 27 SANDIEGOZ YES
2116 Female 100 27 ST PAUL YES
2131 Female 100 26 COLO SPRG YES
2137 Male 100 26 LOUISVILL YES
2157 Female 100 25 AUDUBON YES
2161 Female 100 25 DENVER YES
2203 Female 100 25 FRESNO YES
2361 Male 100 23 COLO SPRG YES
2400 Male 100 23 CINCINNAT YES
2419 Female 100 22 EL PASO YES
2501 Male 100 22 METROZOO YES
2014 Population Viability Analyses for AZA Orangutan Animal Program 39
2507 Neutered Female
100 21 SANDIEGOZ NO Sterile
2509 Male 100 21 ST LOUIS NO Behavior
2512 Female 100 21 TORONTO YES
2517 Female 100 21 PHILADELP YES
2604 Male 100 21 AUDUBON YES
2621 Neutered Female
100 20 RIO GRAND NO Sterile
2706 Male 100 19 SANDIEGOZ YES
2708 Female 100 19 FT WAYNE YES
2718 Male 100 18 SEDGWICK YES
2726 Male 100 18 PHILADELP YES
2752 Female 100 18 ATLANTA YES
2802 Female 100 16 MEMPHIS YES
2958 Male 100 13 OKLAHOMA YES
3006 Male 100 12 INDIANAPL YES
3052 Male 100 11 ATLANTA YES
3100 Male 100 11 SACRAMNTO YES
3150 Female 100 10 HOUSTON YES
3152 Male 100 10 ST LOUIS YES
3167 Female 100 9 ST LOUIS YES
3202 Male 100 9 BROWNSVIL YES
3257 Male 100 7 TORONTO YES
3259 Male 100 7 ATLANTA YES
3270 Female 100 7 TORONTO YES
3333 Male 100 6 ST PAUL YES
3338 Male 100 5 COLO SPRG YES
3347 Female 100 5 RIO GRAND YES
3351 Female 100 4 AUDUBON YES
3373 Female 100 4 PHILADELP YES
3395 Female 100 3 DENVER YES
3397 Male 100 3 FRESNO YES
3400 Male 100 3 ATLANTA YES
3403 Female 100 3 FRESNO YES
3432 Male 100 2 SEDGWICK YES
3450 Female 100 2 BIRMINGHM YES
3507 Male 100 1 ATLANTA YES
35171 Male 100 1 BROWNSVIL YES
3520 Male 100 1 TOPEKA YES
3535 Male 100 1 RIO GRAND YES
3547 Female 100 0 SANDIEGOZ YES 1Individual with unknown sex that was randomly assigned a sex.
2014 Population Viability Analyses for AZA Orangutan Animal Program 40
APPENDIX C. RISK CATEGORIES AND RESULTS ZooRisk uses five standardized risk tests to evaluate different aspects of a population’s demography, genetics, and
management that might put the population at risk (Table C1). The ZooRisk development team and members of the AZA small
Population Management Advisory Group (SPMAG) worked to develop the cutoff for each test. For a given scenario, the
overall risk level is based on the most severe score received among the tests. This approach standardizes assessments across
species and allows managers to compare species programs using the same framework.
For a given scenario, the overall risk level is based on the most severe score it achieved for any of the five tests. Tests 2-4 are
based on the population’s history, and are the same across all model scenarios. For the Orangutan Animal Programs, the risk
levels showed some variation among scenarios (Table C2). The goal for managers should be to move the population from a
Critical status towards a Low Risk status, utilizing some of the management tactics highlighted in the model results.
Table C1: Standardized Risk Test, reflecting the population’s status in zoos and aquariums. Tests Low Risk (LR) Vulnerable (V) Endangered (E) Critical (C)
Probability of extinction in 100 years, based on PVA
0 - 9% 10 - 15% 20 - 49% 50 - 100%
Number of zoos with breeding-aged mixed-sex groups in current population
>3 Zoos 3 Zoos 2 Zoos 1 Zoo
Number of breeding-aged animals (m.f); based on current population
>10.10 7.7 to 10.10 4.4 to 6.6 0.0 to 3.3
Reproduction in the last generation, based on historical studbook data
>9 pairs reproducing 6-9 pairs reproducing 3-5 pairs reproducing 0-2 pairs reproducing
GD of starting population and/or population in 100 years based on PVA
Starting: > 0.9 100 yrs: > 0.9
Starting: 0.8 – 0.9 100 yrs: 0.75 – 0.9
Starting: 0.75 – 0.8 100 yrs: 0.5 – 0.75
Starting: < 0.75 100 yrs: < 0.5
Table C2: Detailed risk results for the population model scenarios.
Scenario
Individual Risk Score for Each Test
OVERALL RISK SCORE FOR
EACH SCENARIO (HIGHLIGHTED)
Probability of
extinction
Number of zoos
Breeding Groups
Breeding in Last Gen.
GD
B Bornean baseline; p(B) = 11.5% LR LR LR LR V LR V E C
C Bornean open; p(B) = 19% LR LR LR LR LR LR V E C
D Bornean; p(B) = 25%; space = 115
LR LR LR LR LR LR V E C
E Sumatran baseline; p(B) = 14.8% LR LR LR LR LR LR V E C
F Sumatran; p(B) = 18% LR LR LR LR LR LR V E C
G Sumatran; p(B) = 25%; space = 115
LR LR LR LR LR LR V E C