ENDANGERED SPECIES RESEARCHEndang Species Res

Vol. 26: 147–159, 2014doi: 10.3354/esr00619

Published online December 10

INTRODUCTION

The Indo-Pacific humpback dolphin Sousa chinen-sis inhabits coastal waters of the Indian and westernPacific Oceans (Jefferson & Karczmarski 2001, Ree -ves et al. 2008) and is found predominantly closeinshore in waters less than 20 m deep (Karczmarski1999, 2000, Karczmarski et al. 2000, Jefferson &Karcz marski 2001, Reeves et al. 2008, Ross et al.2010, Mendez et al. 2013). This restricted inshoredistri bution exposes humpback dolphins to variousanthropogenic impacts, such as incidental mortalityin fishing gear (bycatch), vessel collisions, resourcede pletion, bioaccumulation of harmful pollutants and

habitat destruction through direct human activities(Jefferson & Karczmarski 2001, Reeves et al. 2008,Jefferson et al. 2009, Ross et al. 2010, Huang et al.2013, Slooten et al. 2013). The long-term survival ofhumpback dolphin populations has received remark-able conservation attention in recent years, espe-cially in areas neighboring intense human urbaniza-tion, industrialization and coastal land exploitation(Reeves et al. 2008, Jefferson et al. 2009, Ross et al.2010, Huang et al. 2012b, 2013). The current IUCNRed List of Threatened Species classifies humpbackdolphins globally as Near Threatened (NT) (Reeveset al. 2008). However, adequate evidence of the rateof population decline and risk of extinction relevant

*Corresponding author: leszek@hku.hk

Population trends and vulnerability of humpback dolphins Sousa chinensis off the west coast of Taiwan

Shiang-Lin Huang, Wei-Lun Chang, Leszek Karczmarski*

The Swire Institute of Marine Science, School of Biological Sciences, The University of Hong Kong, Cape d’Aguilar, Shek O, Hong Kong, SAR

ABSTRACT: Predictive modeling of population trends can indicate the rate of population declineand risk of extinction, providing quantitative means of assessing conservation status and threats.Our study tests the rate of population change and risk of extinction of the Indo-Pacific humpbackdolphin Sousa chinensis off the west coast of Taiwan, the only humpback dolphin population clas-sified as Critically Endangered (CR) by the IUCN Red List of Threatened Species. Under the mostoptimistic assumptions, almost 60% of simulations (out of 250 replications × 5000 iterations) pre-dicted that population decline will exceed 80% within 3 generations, while the mean estimate ofpopulation decline within 1 generation was >50% of the current population numbers. Status clas-sification performed using IUCN Red List Categories and Criteria Version 3.1 supported previousCR classification, while risk assessment models that factored in anthropogenic impacts furtherincreased the estimated extinction risk. At an adult survival rate of 0.95, a modeled increment ofannual bycatch rate by 1% of population size increased the probability of extinction within 100 yrby 7.5%; this increase was lower at a higher adult survival rate. The estimated extinction risk wasgreatest under the impact of habitat loss, reaching a hazardous level when habitat carrying capac-ity dropped to less than 50%, indicating that habitat fragmentation and alteration of coastal envi-ronments pose the greatest threats to this population, even if the cumulative sum of fragmentedpatches of habitat may superficially appear to be large.

KEY WORDS: Sousa chinensis · Demographic analyses · Individual-based model · Status assessment · Bycatch · Habitat degradation

© The authors 2014. Open Access under Creative Commons byAttribution Licence. Use, distribution and reproduction are un -restricted. Authors and original publication must be credited.

Publisher: Inter-Research · www.int-res.com

OPENPEN ACCESSCCESS

Endang Species Res 26: 147–159, 2014

to population status assessment under Criteria A, Cand E of the IUCN Red List Categories and Criteriaversion 3.1 (IUCN 2001) remains rare across the spe-cies range (Huang et al. 2012b, Huang & Karcz-marski 2014), leaving relevant impact assessmentsunder various anthropogenic threats to qualitativerather than quantitative means. The effectiveness ofmanagement policies based on qualitative assess-ments can be easily compromised or, even worse, thepolicies themselves may be misguided and focus onthe mitigation of the apparent threats but overlookother, indirect but sometime powerful impacts, asindicated by the recent history of the extinction of thebaiji Lipotes vexillifer (Turvey et al. 2007, 2013) andthe currently accelerating decline of the Yangtze fin-less porpoise Neophocaena asiaeorientalis asiaeori-entalis (Mei et al. 2012, 2014).

A common management approach for ensuringpopulation viability in cetacean species is to define atolerable threshold of incidental removal, termedpotential biological removal (PBR) (Wade 1998,Slooten et al. 2006), to mitigate or minimize anthro-pogenic threats. In practice, the causes of incidentalmortality are often generalized to fishery bycatch(Caswell et al. 1998, Read & Wade 2000, Dans et al.2003, Lennert-Cody et al. 2004, Moore & Read 2008).The PBR calculation for cetaceans is based on anintrinsic population growth rate that is usually set as0.04 when demographic data are not available (Wade1998). However, for species and/or populations clas-sified under the IUCN Red List of Threatened Spe-cies as NT or higher, such as Vulnerable (VU),Endangered (EN) or Critically Endangered (CR), thathave been declining (as is the case with humpbackdolphins) the PBR calculation based on an assumedvalue of the rate of increase, 0.04, may need to beviewed with caution. The strategy of keeping by -catch under the hypothetically calculated PBR maynot facilitate long-term survival of endangered anddeclining populations. Such risk has not yet beenquantitatively addressed for humpback dolphins.

Impact of fishery bycatch is only one of manyanthropogenic threats that endanger the long-termviability of coastal cetaceans. Other impacts, such asresource depletion through overharvesting of localfish stocks (Bearzi et al. 2006, 2008, 2010, Piroddi etal. 2011), harmful levels of pollutant and pathogenaccumulation (Tanaka 2003, Parsons 2004, Reeves etal. 2008), and habitat degradation due to land recla-mation and coastal development (Huang et al. 2013),can severely affect population viability through slowbut irreversible processes. Such processes usuallydeteriorate habitat quality and, in turn, decrease the

habitat carrying capacity (Thomas et al. 2001,Clausen & York 2008, Griffen & Drake 2008) thatdefines the upper limit of fluctuation in populationnumbers (Lacy 1993, Lande 1993, Huang et al.2012a). Consequently, the decline of habitat carryingcapacity fundamentally reduces the capability of apopulation to resist environmental and anthropo -genic impacts (Lacy 1993, Doak 1995, Hilderbrand2003, Griffen & Drake 2008). As a considerable pro-portion of the coastal waters inhabited by humpbackdolphins off southeast Asia is increasingly degradedas a result of industrial and economic developmentand human population growth (Mackinnon et al.2012, Huang et al. 2013, Huang & Karczmarski 2014),an assessment of such impacts on dolphin survivalbecomes increasingly important and urgent.

Of the known humpback dolphin populations, theanimals off the Taiwan west coast (TWC) representthe only population currently classified as CR underCriteria C2 and D (Reeves et al. 2008). In the pastdecades, the west coast of Taiwan, including bothterrestrial and aqueous zones, has experienced rapidhuman population growth and urbanization that haseliminated native coastal and estuarine landscapewith no adequate conservation consideration or envi-ronmental mitigation measures (Huang et al. 2013).This large-scale degradation or even destruction ofcoastal and estuarine ecosystems has likely affectedthe quality of the habitat vital to the long-term viabil-ity of TWC humpback dolphins. In this study, wemodel the rate of decline and extinction probabilityof TWC humpback dolphins within 1 to 5 generations(Criteria A, C and E; IUCN 2001) using demographicparameters estimated from the available data. Subse-quently, we quantify the risk of extinction contri bu -ted by mortality through bycatch and habitat degra-dation. The results of this study provide a preliminaryquantitative risk assessment of the population viabil-ity of TWC humpback dolphins and offer insightsinto population processes that can function as a use-ful case study for other populations or subpopulationsof this species where baseline demographic data arenot available and precautionary measures have to beapplied.

MATERIALS AND METHODS

Demographic rates

Demographic rates of TWC humpback dolphins,including instantaneous rate of increase (r) and gen-eration time (T0), were calculated using the tradi-

148

Huang et al.: Humpback dolphins off Taiwan — trends and vulnerability 149

tional method with the following equations (summa-rized in Krebs 1989):

(1)

and (2)

where l(x) and m(x) are age-specific survivorship andmortality rate at age x, respectively. For humpbackdolphins, as well as many other cetacean species, thedetailed age-specific m(x) is not available and there-fore was defined by:

(3)

where RI, Am and Ax represent reproductive interval(in years), age at maturity (in years) and expectedlifespan (in years) of females, respectively (Huang etal. 2012a,b, Mei et al. 2012). The ratio of daughters, ρ,was assumed to be 0.50.

For most cetaceans, l(x), is usually estimated usingdata collected from stranded or bycaught animals(Stolen & Barlow 2003, Moore & Read 2008, Huang etal. 2012b, Mei et al. 2012). Given the small popula-tion size (Wang et al. 2007, 2012, Yu et al. 2010) andshort time span of previous research in Taiwan, thecurrently available number of collected specimenswas insufficient to build a representative life-historytable for TWC humpback dolphins that could facili-tate the construction of the l(x) model. Therefore, weapplied an alternative approach where the l(x) modelis built as follows (Huang et al. 2012a):

(4)

where Sc and Sa are the survival rates of calf (x ≥ 1)and non-calf dolphins, respectively. The value of Scwas defined as 0.62 (± 0.15 SD) based on recentphoto-ID analysis (Chang 2011). A precise estimate

of Sa, however, is not yet available for the TWChumpback dolphins, despite recent work by Wang etal. (2012) where the Sa estimate (0.985) is accompa-nied by an extremely wide variation (CI = 0.832 −0.998). Therefore we use here the range of Sa be -tween 0.832 and 0.998 reported by Wang et al.(2012), rather than a fixed value (0.985). To includeuncertainty as a precautionary measure when calcu-lating demographic parameters and running popula-tion trend projection, Sa was randomly and repeat-edly re-sampled between 0.832 and 0.998.

The 3 life history parameters that relate to demo-graphic parameter calculation and population trendprojection, Am, RI and AX, were re-sampled withintheir plausible lower (LHPl) and upper (LHPu) bounds(Table 1) by the following equation (Huang et al.2012a,b, Mei et al. 2012):

(5)

where σ is a random number between 0 and 1 gener-ated by MATLAB function ‘rand’ (MathWorks). Byre-sampling these parameters, the effect of para -meter uncertainty could be included in both the de -mographic parameter estimates and predictions ofpopulation trends. The calculation of r and T0 was re -peated using 5000 iterations, and mean (±SD, CI)estimates were calculated.

Population trend projection

The population change of TWC humpback dol-phins, N(t), was simulated using an individual-basedmodel that factors in the effects of parameter uncer-tainty and demographic stochasticity (Slooten et al.2000, Currey et al. 2009a, Huang et al. 2012a,b, Meiet al. 2012). In this model, the population change wassimulated by the following processes:

(1) An individual survived from age x at year t toage x + 1 at year t + 1 whenever the random number

Tx l x m x

l x m x0( ) ( )

( ) ( )=

× ××

∑∑

rl x m x

T

ln( ( ) ( ))=

×∑0

( )0, 0

RI,

m

m

m xx A

A x Ax=

≤ <ρ ≤ <

⎧⎨⎪

⎩⎪

l x S S x( ) c a1= × −

iLHP( ) LHP (LHP LHP )l u l= + − × σ

Parameter Range Reference

Am (yr) 9−11a Jefferson (2000), Jefferson & Hung (2004), Jefferson et al. (2012)RI (yr) 2.19−2.48 Chang (2011)Ax (yr) 25b−43a Jefferson (2000), Jefferson et al. (2012), Huang et al. (2012b)

aValues from the Pearl River Estuary populationbThe largest number of growth layer groups counted in specimens held at the National Museum of Nature and Science,Taiwan (provided by C. J. Yao). GLGs indicate the age and are obtained through sectioning of teeth and counting annuallayers of incremental growth of cementum and dentine

Table 1. Range of life-history parameters, including age at maturity (Am), reproductive interval (RI) and expected lifespan (Ax), of female humpback dolphins Sousa chinensis

Endang Species Res 26: 147–159, 2014150

σ (between 0 and 1) exceeded the mortality rate (q) atage x, as per the following equation: q(x) = 1 − S,where S is the survival rate, either Sc (for x ≤ 1) or Sa(for x > 1). In cases when σ ≤ q(x), the individual dol-phin was considered a victim of mortality, otherwiseit survived.

(2) A female that survived to the next year wasdetermined to give birth by comparing σ with

, with the presence of newborn when ,

where i represents the ith iteration in simulation.

(3) The sex of the newborn was male when σ ex -ceeded the sex ratio ρ (default = 0.50), otherwise thecalf was a female.

(4) For each of the above simulations, a new ran-dom number σ was generated.

The initial abundance (N0) was defined as between54 and 74, according to Wang et al. (2012). Thisrange, however, may be an underestimate (compa -red with Yu et al. 2010), as Wang et al. (2012) appar-ently overlooked the humpback dolphins in habitingthe southern TWC waters (Chou & Lee 2010, Ross etal. 2010). At least 20 individually distinctive animalsinhabit this region and have never been reported further north in the area studied by Wang and col-leagues (Chang 2011, Yeh 2011).

The length of population trend projection in ourstudy was defined at 5 generations or 100 yr,whichever is longer (IUCN, 2001). Each simulationran 250 replications and each replication repeatedlyran 5000 iterations. As the population trend mightfluctuate upward at high Sa value (for example, Sa =0.985), we introduced a density-dependent effect onpopulation change using

(6)

(Lacy 1993, Huang et al. 2012a), where r ’ is a den-sity-adjusted r when N(t) exceeds the ceiling K0. Thevalue of K0 was defined as 99 animals because it isthe earliest abundance estimate (Wang et al. 2007)and so far the highest; none of the subsequent esti-mates (Yu et al. 2010, Wang et al. 2012) exceeded thisvalue. Furthermore, recent analyses of satelliteimagery of the Taiwan west coast indicate a substan-tial loss of humpback dolphin habitat to land recla-mation (Huang et al. 2013), suggesting that popula-tion numbers above this previously estimated ceilingare unlikely.

For each iteration, we estimated the percentage ofabundance decline, within 1 and 3 genera-

tions according to Criteria A3b and C1 in the IUCNRed List Categories and Criteria Version 3.1 (IUCN

2001, Currey et al. 2009a, Wilson et al. 2011, Huanget al. 2012b, Mei et al. 2012). For each replication, weestimated the probabilities of extinction (PE) within100 yr (PE100), 3 generations (PE3T0) and 5 generations(PE5T0) based on Criterion E (IUCN 2001). The condi-tion of population extinction was defined as havingoccurred whenever only one sex, either male orfemale, remained in the population.

Status classification

We calculated the probabilities of the percentageabundance decline within 1 generation and 3 gener-ations meeting the standards for classification as VU,EN or CR under Criteria A3b (in 3 generations) andC1 (in 1 generation). Under Criterion B1b (iii, v)(IUCN 2001), the population status is classified by itsgeographic range and population decline (r). As cur-rent habitat area is obviously larger than 100 km2 butsmaller than 5000 km2 (Wang et al. 2007, 2012, Rosset al. 2010), which are the size thresholds for classifi-cation as CR (≤100 km2) and EN (≤5000 km2), weestimated the probability for the condition when r < 0as the probability of meeting the classification of ENunder Criterion B1bv. The PE3T0, PE5T0 and PE100 esti-mates were used to classify the population as VU, ENor CR under Criterion E (IUCN 2001).

Sensitivity test for adult survival rate

Besides projecting r and the population trendbased on a dynamic range of Sa (0.832−0.998), weestimated r2 and PE100 values of the TWC humpbackdolphin at different fixed Sa. In this process, the valueof Sa increased from 0.832 to 0.998. Means (±SD) of r2and PE100 at each Sa value were calculated.

Vulnerability test for anthropogenic impacts

The influence of anthropogenic impacts on popula-tion trends and the probability of extinction wereassessed as 2 categories: those directly causing inci-dental mortality (e.g. bycatch and vessel collisions)were categorized as ‘bycatch impact’, while indirectimpacts that could potentially decrease the availabil-ity and quality of habitat (e.g. land reclamation,resource depletion and habitat degradation) werecategorized as ‘habitat loss impact’. The range ofbycatch impact was tested from 0 to 5% N(t) perannum. The extent of habitat loss impact, in contrast,

r rN tK

' (( )

)= × −10

1RIi

σ ≤1RIi

ΔN tN

( )

0

Huang et al.: Humpback dolphins off Taiwan — trends and vulnerability

was defined between 0 and 90% K0, where K0 wasdefined as 99 animals (Wang et al. 2007) by introduc-ing a previously described density-dependent effect.This ceiling was defined because the availability andquality of potential habitat for humpback dolphins offthe Taiwan west coast is unlikely to increase in thefuture due to substantial recent alteration of habitatand environmental change (Huang et al. 2013). Asthe Sa initially used has a very wide range that couldlikely interfere with the assessment of impacts, weadopted fixed Sa values in this test. The Sa weredefined as 0.985 (Wang et al. 2012) and 0.95 (Karcz-marski 2000, Jefferson & Karczmarski 2001, Taylor etal. 2007) and tested separately. The latter value wasselected because it appears close to the Sa estimatefor an ‘undisturbed’ population of cetaceans (Tayloret al. 2007), such as the bottlenose dolphin (Tursiopstruncatus) (Sa2 = 0.951: Small & Demaster 1995).Means (±SD) of PE100 under these 2 impact scenarioswere estimated. Table 2 summarizes the simulationenvironment of trend projections for status assess-ment and sensitivity of decline at different Sa valuesand under different anthropogenic impacts.

RESULTS

Demographic rates and population trend

Demographic rate estimates for the TWC hump-back dolphins are summarized in Table 3. The r2 esti-mate of −0.0278 indicates a moderate populationdecline at 2.74% (1 − exp[−0.0278]) per annum. How-ever, the CI of the estimate is large, ranging from−0.113 (rapidly downward) to 0.0317 (moderatelyupward). The generation time T30 is estimated at21.28 yr (SD = 3.76 yr, CI = 16.20 − 28.80 yr). Thus,the temporal length for status classification under

Criteria A3b, C1 and E was chosen as 22, 64 and100 yr for 1, 3 and 5 generations, respectively.

The population change N(t) projected by the indi-vidual-based model is shown in Fig. 1. The indeter-ministic plot shows the stochasticity of simulationsand indicates a wide uncertainty and the difficultyof reaching a determinate conclusion (Fig. 1A). Al -though 43% (SD = 1.54%) of simulations (250 re -plications × 5000 iterations) suggested a ‘stable’ pop-ulation, almost 54% (SD = 1.52%) of simulationspre dicted population extinction within 100 yr. Thedirection of the deterministic population trend wasmost affected by the value of Sa (Fig. 1B). With Sa =0.985, the dolphin population was projected to slowlyincrease, while with Sa = 0.95 it was projected todecline at a slow but continuous rate (Fig. 1B).

Mean probability of extinction estimates, using Sa =0.832 − 0.998, were 47.30% (SD = 1.51%), 64.13%(SD = 1.62%) and 53.93% (SD = 1.52%) of simula-tions for PE3T0, PE5T0 and PE100, respectively (Fig. 2).A total of 69.4% of simulations (SD = 8.71%) pre-dicted a population decline greater than 25% within1 generation (22 yr), with the mean estimate suggest-ing that population numbers would diminish by anaverage of 50.48% (SD = 50.79%) (Fig. 3A). The pop-ulation loss reached an average of 66.09% (SD =58.23%) decline in 3 generations (64 yr), while 59%(SD = 3.23%) of simulations indicated that the popu-lation decline over 3 generations could exceed 80%(Fig. 3B).

Status classification

The status classification of the TWChumpback dolphins under CriteriaA−E is summarized in Table 4. For allpopulation trend projections (250 re -plications × 5000 iterations), 59%(SD = 3.23%) of simulations meet theconditions for classification as CR un -der Criterion A3b (Fig. 3B), al thoughthe mean population decline esti-mates meet the conditions for classifi-cation as EN, defined as the percent-

151

Parameter Value or range Reference

Initial abundance, N0 54−74 Wang et al. (2012)Survival rate (yr–1)Calf (age ≤ 1) 0.62 Chang (2011)Non-calf (age > 1) 0.832−0.998 Wang et al. (2012)Ratio of daughters 0.5 Present studyLength of projection 100 yr or 5 generations, IUCN (2001)

whichever is longerAnthropogenic impactsBycatch 0−5% N(t) Present studyHabitat degradation 0−90% K0 Present study

Table 2. Definition of parameters used for projecting population trends and risk of extinction. K0: initial habitat carrying capacity

Parameter Mean SD CI

T0 21.28 3.76 16.20 to 28.80r −0.0278 0.0434 −0.113 to 0.0317

Table 3. Estimates of the demographic parameters genera-tion time (T0) and instantaneous rate of increase (r) for

humpback dolphins off the west coast of Taiwan

Endang Species Res 26: 147–159, 2014

age decline higher than 50% but lower than 80%within 3 generations. Applying the precautionaryprinciple, the CR classification under Criterion A3bseems appropriate, although EN classification is alsoplausible. Under Criterion C1, as the current popula-tion has fewer than 250 adults and the estimated per-

centage decline within 1 generation is higher than25%, the TWC humpback dolphins can be classifiedas CR. The PE estimates meet the conditions for clas-sification as EN under Criterion E, defined as PE >20% within 5 generations but

Huang et al.: Humpback dolphins off Taiwan — trends and vulnerability 153

0

10

20

30

40

50

<10

10–1

9

20–2

9

30–3

9

40–4

9

50–5

9

60–6

9

70–7

9

80–8

9

90–9

9

<10

10–1

9

20–2

9

30–3

9

40–4

9

50–5

9

60–6

9

70–7

9

80–8

9

90–9

9

Decline within 1 generation (%)

% s

imul

atio

ns

A

0

10

20

30

40

50

60

Decline within 3 generations (%)

% s

imul

atio

ns

B

Criterion no. Details of criterion Results of population viability analyses Status

A3b 50% ≤ ΔN < 80% within 3T0 59.00% (SD = 3.23%) simulations predicted ΔN ≥ 80% (Fig. 3B) EN ΔN32 66.09% (SD 58.23%) (CR)B1b (iii,v) 100 km2 ≤ extent of Current extent of occurrence

Endang Species Res 26: 147–159, 2014

rate for most toothed cetaceans; Taylor et al. 2007), a1% increment of bycatch per annum raised PE100 by7.5% (Fig. 5B).

The impact of habitat loss, represented by the per-centage K0 loss, affected population survival in anonlinear pattern. At a low level of habitat loss, PE100slowly increased for both S22a = 0.985 (Fig. 6A) andS22a = 0.95 (Fig. 6B). When the extent of habitat lossexceeded 50% K0 loss, PE100 escalated rapidly andreached a hazardous level even at S22a = 0.985(Fig. 6A).

DISCUSSION

Factors affecting the results of trend projection and status assessment

The trajectory of population trend projections and,consequently, the predictions of extinction risk relyon the accuracy of demographic parameter estimatesand the number and magnitude of different impactsincluded in a population model (Lacy 1993). Theaccuracy of demographic parameter estimates, inturn, depends on the accuracy of other estimatessuch as life-history parameters, age-specific survi -vorship and reproductive rates. Some of these para -

154

0

1

2

3

4

5

0 1 2 3 4 5

0 1 2 3 4 5

Bycatch rate (% N(t))

ASa: 0.985

20

30

40

50

60

70 BSa: 0.95

PE

100

(%)

PE

100

(%)

Fig. 5. Sensitivity test of Sousa chinensis population survival(measured using the probability of extinction in 100 yr,PE100) impacted by varied intensities of bycatch (repre-sented by % N(t), where N(t) is initial population size) at (A)Sa = 0.985 (Wang et al. 2012) and (B) Sa = 0.95 (undisturbedvalue assumed for most toothed cetaceans; Taylor et

al. 2007)

0

5

10

15

20

25

30

35

40

10 30 50 70 90

ASa: 0.985

20

30

40

50

60

70

80

90

100

10 30 50 70 90

Habitat loss (% K0)

BSa: 0.95

PE

100

(%)

PE

100

(%)

Fig. 6. Sensitivity test of population survival (measured withthe probabilities of extinction in 100 yr, PE100) at dif ferentextents of habitat-loss impact, presented as % of the initialcarrying capacity (K0); (A) Sa = 0.985 and (B) Sa = 0.95, where

Sa is a non-calf survival rate

Huang et al.: Humpback dolphins off Taiwan — trends and vulnerability

meters, however, such as the expected lifespan (Ax),age at maturity (Am) and age-specific survivorship(l(x)), are currently not available for TWC humpbackdolphins. To calculate these parameters, one wouldrequire access to a substantial number of stranded orbycaught specimens, the accumulation of whichtakes a long time even in substantially larger popula-tions (Jefferson 2000, Jefferson et al. 2012, Mei et al.2012). Such a representative number of specimens isunlikely to ever be available for the TWC humpbackdolphins due to the very small population size.

Wang et al. (2012) claim that the non-calf survivalrate (S22a) of the TWC humpback dolphin approxi-mates 0.985 with a very wide variation, from 0.832 to0.998, which was the range applied in our studyreported here. This S22a estimate, however, contradictssome earlier findings and should therefore be treatedwith caution. The TWC humpback dolphin popula-tion has been classified as CR under Criteria C2 andD of the IUCN Red List of Threatened Species(Reeves et al. 2008). This classification stipulates ‘acontinuing decline, observed, projected, or inferred,in numbers of mature individuals’ (IUCN 2001, p. 18).In the case of S22a = 0.985, however, as suggested byWang et al. (2012), the TWC humpback dolphin pop-ulation would actually be increasing with a moderaterate, r2 = 0.0256 (SD = 0.006), equivalent to a 2.59%N(t) increase per annum or a 75.44% N(t) increaseper generation (~22 yr). If the S22a estimate (0.985)indeed truly represents the non-calf survival rate, theTWC humpback dolphins should be viewed ashighly resilient to anthropogenic impacts, especiallyfishery bycatch (Wade 1998). This inference contra-dicts the currently gathered evidence on the status ofTWC humpback dolphins (Wang et al. 2007, Reeveset al. 2008, Ross et al. 2010, Slooten et al. 2013).

Recent demographic analysis for the Pearl RiverEstuary humpback dolphin population based on age-structure data reports a declining trend (r2 = –0.0245;Huang et al. 2012b), indicating an Sa estimate farlower than 0.94 (W. C. Lam, unpubl. data) correspon-ding to the Sc estimate (0.61; Jefferson et al. 2012),which is comparable to many other cetacean popula-tions under high anthropogenic pressure (Slooten etal. 1992, Stolen & Barlow 2003, Currey et al.2009a,b). The real non-calf survival rate (or lowerbound of the Sa estimate) for the TWC humpbackdolphins is likely lower, possibly substantially lowerthan the 0.985 suggested by Wang et al. (2012).

Considering the extensive environmental degrada-tion of the shallow-water coastal habitats off theTWC in recent decades due to land reclamation andexploitation of adjacent terrestrial environments

(Huang et al. 2013), the actual rate of decline, risk ofextinction and susceptibility to anthropogenic im -pacts such as bycatch and habitat loss are likely to behigher than the estimates presented here. As ouranalyses had to incorporate recently published esti-mates, the resulting PE estimates are affected by theevident imprecision of the earlier studies (Wang et al.2012) and the real risk of extinction is likely to bemore severe than our current models predict.

The accuracy of the trend projection and riskassessment presented here may be further affectedby impacts of inbreeding, genetic drift and loss ofgenetic diversity in a small population (Souléa & Sim-berloff 1986, Lacy 1993, Mills & Smouse 1994, Sato &Harada 2008). A minimum population size, at least250 adults (Shaffer 1981, Nunney & Campbell 1993)but usually up to thousands of adults (Harcourt 2002,Brito & Figueiredo 2003, Reed et al. 2003, Traill et al.2007), is needed in mammals to resist the minimallevel of stochastic genetic diversity loss. The currentpopulation size of TWC humpback dolphins (fewerthan 100 animals; Wang et al. 2007, 2012, Yu et al.2010), is substantially lower than the lowest pre-dicted limit of at least 250 adults, and geneticexchange with neighboring populations, e.g. thoseoff Xiamen and in the Pearl River Estuary, is unlikely(Wang et al. 2008). Therefore, applications of modelsthat factor in genetic impacts, such as the VORTEXmodel (Lacy 1993), although currently not possibledue to the lack of genetic data, would very likelygenerate predictions of higher extinction risks thanthose presented here.

Most demographic models assume constant envi-ronmental conditions where the extent of anthro-pogenic impacts and environmental stochasticitydoes not change over time (Lacy 1993, Caswell et al.1999, Winship & Trites 2006). This, however, canhardly be the case under ever-increasing humanpressures, such as off the TWC. Recent analyses indi-cate that the habitat quality of coastal waters inhab-ited by the TWC humpback dolphin has been sub-stantially degraded in past decades (Huang et al.2013), and there are no signs that this process islikely to come to a halt, let alone be reversed in theforeseeable future (Ross et al. 2010). The cumulativeimpacts of habitat destruction and alteration ofcoastal environments on dolphin ecology and sur-vival across temporal scales remain little understood(e.g. Currey et al. 2009b, Mei et al. 2012). This is yetanother reason for caution when considering thepopulation trend predictions presented in this study,as the current analytical limitations have likely gen-erated an overly optimistic projection.

155

Endang Species Res 26: 147–159, 2014

Status and threats to the persistence of TWC humpback dolphins

Our study indicates that the status of humpbackdolphins off the TWC can be classified as CR under3 of the 5 IUCN criteria, A3b, C1 and D, although theaverage percentage of decline appears to be lowerthan the threshold: 80% of the original abundancewithin 3 ge nerations (IUCN 2001). This classificationdoes not contradict the current status under IUCNclassification (Reeves et al. 2008); in fact, it providesa quantitative basis for such a classification andemphasizes the actual risk that the TWC humpbackdolphins are currently facing. Under the assumedcondition (Sc = 0.62, Sa = 0.832 − 0.998), more than60% of the current population numbers may be lostin the next 3 generations (approximately 64 yr). Thisdecline might be even higher with a non-calf survivalrate lower than recently published Sa estimates. Asall current and recent population estimates indicate avery small population size (Wang et al. 2007, 2012,Yu et al. 2010), such a drastic decline in numbers willplace the TWC humpback dolphins under an ever-increasing risk of the effects of demographic andgenetic stochasticity, with diminished resilience toanthropogenic impacts.

Direct mortality caused by fishery bycatch andvessel collisions can present real threats to thelong-term persistence of a dolphin population(Wade 1998, D’agrosa et al. 2000, Dans et al. 2003,Read et al. 2006, Slooten et al. 2006, Cramer et al.2008, Moore & Read 2008, Ross et al. 2010).Although Wade (1998) proposes the PBR calcula-tion for marine mammals that defines the upperlimit of tolerable incidental mortality, the calcula-tion of PBR (2−4% abundance per annum) shouldnot be directly applied to the TWC humpback dol-phin population to define the upper tolerablebycatch removal. According to our simulations,there is a high probability that the TWC humpbackdolphin population will decline even under theapparently most optimistic scenario and withoutfactoring in the impacts of bycatch. With by catchmortality included, even as low as 1% per annum,the decline of the population will escalate substan-tially. If the 2% PBR per annum was directlyapplied in the management plan of TWC hump-back dolphins, not accounting for the ongoingpopulation decline, we can anticipate a far greaterdrop in population numbers on time scales shorterthan the projections presented in this study.Therefore, we postulate that a direct application ofthe PBR calculation in population management

strategies without projecting the trend of the pop-ulation concerned should be done with caution,even if this calculation appears to be numericallytolerable.

Unlike fishery bycatch or vessel collisions, whichcause immediate mortality of impacted animals, thedemographic consequences of habitat degradationare usually not obvious but are long-lasting and canbe irreversible (Huang et al. 2012a). Impacts ofcoastal development activities, especially land recla-mation (Huang et al. 2013), sand dredging and bot-tom trawling, have immediate effects by reducingthe extent, integrity and quality of suitable habitats,and effectively reducing the habitat carrying capa -city. Other indirect impacts, such as overfishing,which causes resource depletion (Bearzi et al. 2006,2008, 2010, Piroddi et al. 2011), and pollutant accu-mulation, which functionally alters the planktonfauna and breaks down the energetic function oflocal ecosystems, may not cause immediate threatsbut cumulative impacts that may, over time, be irre-versible (Huang et al. 2012a).

Although the scale of habitat loss simulated in ourstudy, up to 90% of initial carrying capacity, mayseem unreasonably high, it is likely to occur in anarrow strip of shallow-water coastal habitats. Offthe TWC, as in various other locations inhabited byhumpback dolphins, suitable habitat does notextend far offshore but along the shore and at timesfor hundreds of kilometers (Corkeron et al. 1997,Karczmarski et al. 1999, 2000, Parra et al. 2004, Rosset al. 2010, Yeh 2011). Alteration of such coastalenvironments through urban and industrial coastaldevelopments, large-scale land reclamation or sanddredging can disrupt habitat continuity and in -tegrity, fragmenting it to ever-smaller discontinuouspatches (Huang et al. 2013). The carrying capacityof such a fragmented mosaic of patches is drasticallyreduced and can drop below 10% K0. Consequently,the risk of local extinction in the fragmented habi-tats is high and increases with further decreases inpatch size (e.g. Fig. 6). Severe habitat discontinuitywith large-scale habitat loss, as off the Taiwan cen-tral west coast, may lead to population fragmenta-tion (Chang 2011), likely escalating further the rateof population decline and increasing the risk ofextinction.

Acknowledgements. This study was funded by the NationalScience Council of Taiwan (grant NSC 101-2311-B-019-002). The participation of L.K. was supported by the GeneralResearch Fund (GRF) of the Research Grants Council (RGC)of Hong Kong (RGC grant HKU768913M). All authorsdeclare that they have no conflict of interest.

156

Huang et al.: Humpback dolphins off Taiwan — trends and vulnerability

LITERATURE CITED

Bearzi G, Politi E, Agazzi S, Azzellino A (2006) Prey deple-tion caused by overfishing and the decline of marinemegafauna in eastern Ionian Sea coastal waters (centralMediterranean). Biol Conserv 127: 373−382

Bearzi G, Agazzi S, Gonzalvo J, Costa M and others (2008)Overfishing and the disappearance of short-beakedcommon dolphins from western Greece. Endang SpeciesRes 5: 1−12

Bearzi G, Agazzi S, Gonzalvo J, Bonizzoni S, Costa M, Pet-roselli A (2010) Biomass removal by dolphins and fish-eries in a Mediterranean Sea coastal area: Do dolphinshave an ecological impact on fisheries? Aquat Conserv: Mar Freshw Ecosyst 20: 549−559

Brito D, Figueiredo MSL (2003) Minimum viable populationand conservation status of the Atlantic forest spiny ratTrinomys eliasi. Biol Conserv 113: 153−158

Caswell H, Brault S, Read AJ, Smith TD (1998) Harbor por-poise and fisheries: an uncertainty analysis of incidentalmortality. Ecol Appl 8: 1226−1238

Caswell H, Fujiwara M, Brault S (1999) Declining survivalprobability threatens the North Atlantic right whale.Proc Natl Acad Sci USA 96: 3308−3313

Chang WL (2011) Social structure and reproductive para -meters of Indo-Pacific humpback dolphins (Sousa chi-nensis) off the west coast of Taiwan. Master’s thesis,National Taiwan University, Taipei

Chou LS, Lee JD (2010) Habitat hotspot of humpback dol-phin, Sousa chinensis, and master planning for conserva-tion management (in Chinese). Forestry Bureau, Councilof Agriculture, Executive Yuan, Taiwan

Clausen R, York R (2008) Economic growth and marine bio-diversity: influence of human social structure on declineof marine trophic levels. Conserv Biol 22: 458−466

Corkeron PJ, Morissette NM, Porter L, Marsh H (1997) Dis-tribution and status of humpback dolphins, Sousa chi-nensis, in Australian waters. Asian Mar Biol 14: 49−59

Cramer KL, Perryman WL, Gerrodette T (2008) Declines inreproductive output in two dolphin populations depletedby the yellowfin tuna purse-seine fishery. Mar Ecol ProgSer 369: 273−285

Currey RJC, Dawson SM, Slooten E (2009a) An approach forregional threat assessment under IUCN Red List criteriathat is robust to uncertainty: the Fiordland bottlenosedolphins are critically endangered. Biol Conserv 142: 1570−1579

Currey RJC, Dawson SM, Slooten E, Schneider K and others(2009b) Survival rates for a declining population of bot-tlenose dolphins in Doubtful Sound, New Zealand: aninformation theoretic approach to assessing the role ofhuman impacts. Aquat Conserv: Mar Freshw Ecosyst 19: 658−670

D’agrosa C, Lennert-Cody CE, Vidal O (2000) Vaquitabycatch in Mexico’s artisanal gillnet fisheries: driving asmall population to extinction. Conserv Biol 14: 1110−1119

Dans SL, Alonso MK, Pedraza SN, Crespo EA (2003) Inci-dental catch of dolphins in trawling fisheries off Patago-nia, Argentina: Can populations persist? Ecol Appl 13: 754−762

Doak DF (1995) Source-sink models and the problem ofhabitat degradation: general models and applications tothe Yellowstone grizzly. Conserv Biol 9: 1370−1379

Griffen BD, Drake JM (2008) Effects of habitat quality and

size on extinction in experimental populations. Proc RSoc B Biol Sci 275: 2251−2256

Harcourt AH (2002) Empirical estimates of minimum viablepopulation sizes for primates: tens to tens of thousands?Anim Conserv 5: 237−244

Hilderbrand RH (2003) The roles of carrying capacity, immi-gration, and population synchrony on persistence ofstream-resident cutthroat trout. Biol Conserv 110: 257−266

Huang SL, Karczmarski L (2014) Indo-Pacific humpbackdolphins: a demographic perspective of a threatenedspecies. In: Yamagiwa J, Karczmarski L (eds) Primatesand cetaceans: field research and conservation of com-plex mammalian societies. Primatology Monographs.Springer, Tokyo, p 249−272

Huang SL, Hao Y, Mei Z, Turvey ST, Wang D (2012a) Com-mon pattern of population decline for freshwater ceta -cean species in deteriorating habitats. Freshw Biol 57: 1266−1276

Huang SL, Karczmarski L, Chen J, Zhou R and others(2012b) Demography and population trends of thelargest population of Indo-Pacific humpback dolphins.Biol Conserv 147: 234−242

Huang SL, Karczmarski L, Keith M, Chang WL (2013) Long-term habitat degradation from land reclamation: its con-sequence on coastal cetacean distribution. Proceedingsof The International Conference on Challenges inAquatic Sciences. National Taiwan Ocean University,Keelug

IUCN (2001) IUCN Red List Categories and Criteria: Version3.1. IUCN Species Survival Commission. IUCN, Gland

Jefferson TA (2000) Population biology of the Indo-Pacifichump-backed dolphin in Hong Kong waters. WildlMonogr 144: 1−65

Jefferson TA, Hung SK (2004) A review of the status of theIndo-Pacific humpback dolphin (Sousa chinensis) in Chi-nese waters. Aquat Mamm 30: 149−158

Jefferson TA, Karczmarski L (2001) Sousa chinensis. MammSpecies 655: 1−9

Jefferson TA, Hung SK, Würsig B (2009) Protecting smallcetaceans from coastal development: impact assessmentand mitigation experience in Hong Kong. Mar Policy 33: 305−311

Jefferson TA, Hung SK, Robertson KM, Archer FI (2012) Lifehistory of the Indo-Pacific humpback dolphin in the PearlRiver Estuary, southern China. Mar Mamm Sci 28: 84−104

Karczmarski L (1999) Group dynamics of humpback dol-phins (Sousa chinensis) in the Algoa Bay region. S Afr JZool 249: 283−293

Karczmarski L (2000) Conservation and management ofhumpback dolphins: the South African perspective. Oryx34: 207−216

Karczmarski L, Winter PED, Cockroft VG, Mclachlan A(1999) Population analysis of Indo-Pacific humpback dol-phins, Sousa chinensis, in Algoa Bay, Eastern Cape,South Africa. Mar Mamm Sci 15: 1115−1123

Karczmarski L, Cockcroft VG, Mclachlan A (2000) Habitatuse and preferences of Indo-Pacific humpback dolphins,Sousa chinensis, in Algoa Bay, South Africa. Mar MammSci 16: 65−79

Krebs CJ (1989) Ecological methodology. Harper Collins,New York, NY

Lacy RC (1993) VORTEX: a computer simulation model forpopulation viability analysis. Wildl Res 20: 45−65

157

http://dx.doi.org/10.1071/WR9930045http://dx.doi.org/10.1111/j.1748-7692.2000.tb00904.xhttp://dx.doi.org/10.1111/j.1748-7692.1999.tb00880.xhttp://dx.doi.org/10.1111/j.1748-7692.2010.00462.xhttp://dx.doi.org/10.1016/j.marpol.2008.07.011http://dx.doi.org/10.1644/1545-1410(2001)655%3C0001%3ASC%3E2.0.CO%3B2http://dx.doi.org/10.1578/AM.30.1.2004.149http://dx.doi.org/10.1016/j.biocon.2012.01.004http://dx.doi.org/10.1111/j.1365-2427.2012.02772.xhttp://dx.doi.org/10.1016/S0006-3207(02)00224-0http://dx.doi.org/10.1017/S1367943002002287http://dx.doi.org/10.1578/AM.30.1.2004.94http://dx.doi.org/10.1098/rspb.2008.0518http://dx.doi.org/10.1046/j.1523-1739.1995.09061370.xhttp://dx.doi.org/10.1890/1051-0761(2003)013[0754%3AICODIT]2.0.CO%3B2http://dx.doi.org/10.1046/j.1523-1739.2000.98191.xhttp://dx.doi.org/10.1002/aqc.1015http://dx.doi.org/10.1016/j.biocon.2009.02.036http://dx.doi.org/10.3354/meps07606http://dx.doi.org/10.1111/j.1523-1739.2007.00851.xhttp://dx.doi.org/10.1073/pnas.96.6.3308http://dx.doi.org/10.1890/1051-0761(1998)008[1226%3AHPAFAU]2.0.CO%3B2http://dx.doi.org/10.1016/S0006-3207(02)00344-0http://dx.doi.org/10.1002/aqc.1123http://dx.doi.org/10.3354/esr00103http://dx.doi.org/10.1016/j.biocon.2005.08.017

Endang Species Res 26: 147–159, 2014

Lande R (1993) Risks of population extinction from demo-graphic and environmental stochasticity and randomcatastrophes. Am Nat 142: 911−927

Lennert-Cody CE, Minami M, Hall MA (2004) Incidentalmortality of dolphins in the eastern Pacific Ocean purse-seine fishery: correlates and their spatial association.J Cetacean Res Manag 6: 151−163

Mackinnon J, Verkuil YI, Murray N (2012) IUCN situationanalysis on East and Southeast Asian intertidal habitats,with particular reference to the Yellow Sea (includingthe Bohai Sea). Occasional Paper of the IUCN SpeciesSurvival Commission No. 47. IUCN, Gland

Mei Z, Huang SL, Hao Y, Turvey ST, Gong W, Wang D(2012) Accelerating population decline of Yangtze finlessporpoise (Neophocaena asiaeorientalis asiaeorientalis).Biol Conserv 153: 192−200

Mei Z, Zhang X, Huang SL, Zhao X and others (2014) TheYangtze finless porpoise: on an accelerating path toextinction? Biol Conserv 172: 117−123

Mendez M, Jefferson TA, Kolokotronis SO, Krützen M andothers (2013) Integrating multiple lines of evidence tobetter understand the evolutionary divergence of hump-back dolphins along their entire distribution range: anew dolphin species in Australian waters? Mol Ecol 22: 5936−5948

Mills LS, Smouse PE (1994) Demographic consequences ofinbreeding in remnant populations. Am Nat 144: 412−431

Moore JE, Read AJ (2008) A Bayesian uncertainty analysisof cetacean demography and bycatch mortality usingage-at-death data. Ecol Appl 18: 1914−1931

Nunney L, Campbell KA (1993) Assessing minimum viablepopulation size: demography meets population genetics.Trends Ecol Evol 8: 234−239

Parra GJ, Corkeron PJ, Marsh H (2004) The Indo-Pacifichumpback dolphin, Sousa chinensis (Osbeck, 1765), inAustralian waters: a summary of current knowledge.Aquat Mamm 30: 197−206

Parsons ECM (2004) The potential impacts of pollution onhumpback dolphins, with a case study on the Hong Kongpopulation. Aquat Mamm 30: 18−37

Piroddi C, Bearzi G, Gonzalvo J, Christensen V (2011) Fromcommon to rare: the case of the Mediterranean commondolphin. Biol Conserv 144: 2490−2498

Read AJ, Wade PR (2000) Status of marine mammals in theUnited States. Conserv Biol 14: 929−940

Read AJ, Drinker P, Northridge S (2006) Bycatch of marinemammals in US and global fisheries. Conserv Biol 20: 163−169

Reed DH, O’grady JJ, Brook BW, Ballou JD, Frankham R(2003) Estimates of minimum viable population sizes forvertebrates and factors influencing those estimates. BiolConserv 113: 23−34

Reeves RR, Dalebout ML, Jefferson TA, Karczmarski L andothers (2008) Sousa chinensis. IUCN Red List of Threat-ened Species. Version 2011.2. Available at www. iucnredlist.org (accessed on 10 April 2012)

Ross PS, Dungan SZ, Hung SK, Jefferson TA and others(2010) Averting the baiji syndrome: conserving habitatfor critically endangered dolphins in Eastern TaiwanStrait. Aquat Conserv: Mar Freshw Ecosyst 20: 685−694

Sato T, Harada Y (2008) Loss of genetic variation and effec-tive population size of Kirikuchi charr: implications forthe management of small, isolated salmonid populations.Anim Conserv 11: 153−159

Shaffer ML (1981) Minimum population sizes for species

conservation. Bioscience 31: 131−134Slooten E, Dawson SM, Lad F (1992) Survival rates of photo-

graphically identified Hector’s dolphins from 1984 to1988. Mar Mamm Sci 8: 327−343

Slooten E, Fletcher D, Taylor BL (2000) Accounting foruncertainty in risk assessment: case study of Hector’sdolphin mortality due to gillnet entanglement. ConservBiol 14: 1264−1270

Slooten E, Dawson SM, Rayment WJ, Childerhouse SJ(2006) A new abundance estimate for Maui’s dolphin: What does it mean for managing this critically endan-gered species? Biol Conserv 128: 576−581

Slooten E, Wang JY, Dungan SZ, Forney KA and others(2013) Impacts of fisheries on the Critically Endangeredhumpback dolphin Sousa chinensis population in theeastern Taiwan Strait. Endang Species Res 22: 99−114

Small RJ, Demaster DP (1995) Survival of five species of cap-tive marine mammals. Mar Mamm Sci 11: 209−226

Souléa ME, Simberloff D (1986) What do genetics and ecol-ogy tell us about the design of nature reserves? Biol Con-serv 35: 19−40

Stolen MK, Barlow J (2003) A model life table for bottlenosedolphins (Tursiops truncatus) from the Indian River La -goon System, Florida, USA Mar Mamm Sci 19: 630−649

Tanaka Y (2003) Ecological risk assessment of pollutantchemicals: extinction risk based on population-leveleffects. Chemosphere 53: 421−425

Taylor BL, Chivers SJ, Larese J, Perrin WF (2007) Genera-tion length and percent mature estimates for IUCNassessments of cetaceans. Administrative Report LJ-07-01, National Marine Fisheries Service, Southwest Fish-eries Science Center, La Jolla, CA

Thomas JA, Bourn NAD, Clarke RT, Stewart KE and others(2001) The quality and isolation of habitat patches bothdetermine where butterflies persist in fragmented land-scapes. Proc R Soc B Biol Sci 268: 1791−1796

Thompson PM, Wilson B, Grellier K, Hammond PS (2000)Combining power analysis and population viabilityanalysis to compare traditional and precautionaryapproaches to conservation of coastal cetaceans. Con-serv Biol 14: 1253−1263

Traill LW, Bradshaw CJA, Brook BW (2007) Minimum viablepopulation size: a meta-analysis of 30 years of publishedestimates. Biol Conserv 139: 159−166

Turvey ST, Pitman RL, Taylor BL, Barlow J and others (2007)First human-caused extinction of a cetacean species?Biol Lett 3: 537−540

Turvey ST, Risley CL, Moore JE, Barrett LA and others(2013) Can local ecological knowledge be used to assessstatus and extinction drivers in a threatened freshwatercetacean? Biol Conserv 157: 352−360

Wade PR (1998) Calculating limits to the allowable human-caused mortality of cetaceans and pinnipeds. MarMamm Sci 14: 1−37

Wang JY, Yang SC, Hung SK, Jefferson TA (2007) Distribu-tion, abundance and conservation status of the easternTaiwan Strait population of Indo-Pacific humpback dol-phins, Sousa chinensis. Mammalia 71: 157−165

Wang JY, Hung SK, Yang SC, Jefferson TA, Secchi ER (2008)Population differences in the pigmentation of Indo-Pacific humpback dolphins, Sousa chinensis, in Chinesewaters. Mammalia 72: 302−308

Wang JY, Yang SC, Fruet PF, Daura-Jorge FG, Secchi ER(2012) Mark-recapture analysis of the critically endan-gered Eastern Taiwan Strait population of Indo-Pacific

158

http://dx.doi.org/10.5343/bms.2010.1097http://dx.doi.org/10.1515/MAMM.2008.030http://dx.doi.org/10.1515/MAMM.2007.032http://dx.doi.org/10.1111/j.1748-7692.1998.tb00688.xhttp://dx.doi.org/10.1016/j.biocon.2012.07.016http://dx.doi.org/10.1098/rsbl.2007.0292http://dx.doi.org/10.1016/j.biocon.2007.06.011http://dx.doi.org/10.1046/j.1523-1739.2000.00099-410.xhttp://dx.doi.org/10.1098/rspb.2001.1693http://dx.doi.org/10.1046/j.1523-1739.2000.99409.xhttp://dx.doi.org/10.1016/S0045-6535(03)00016-Xhttp://dx.doi.org/10.1111/j.1748-7692.2003.tb01121.xhttp://dx.doi.org/10.1016/0006-3207(86)90025-Xhttp://dx.doi.org/10.1111/j.1748-7692.1995.tb00519.xhttp://dx.doi.org/10.3354/esr00518http://dx.doi.org/10.1016/j.biocon.2005.10.013http://dx.doi.org/10.1046/j.1523-1739.2000.00099-411.xhttp://dx.doi.org/10.1111/j.1748-7692.1992.tb00049.xhttp://dx.doi.org/10.2307/1308256http://dx.doi.org/10.1111/j.1469-1795.2008.00165.xhttp://dx.doi.org/10.1002/aqc.1141http://dx.doi.org/10.1016/S0006-3207(02)00346-4http://dx.doi.org/10.1111/j.1523-1739.2006.00338.xhttp://dx.doi.org/10.1046/j.1523-1739.2000.99107.xhttp://dx.doi.org/10.1016/j.biocon.2011.07.003http://dx.doi.org/10.1578/AM.30.1.2004.18http://dx.doi.org/10.1578/AM.30.1.2004.197http://dx.doi.org/10.1016/0169-5347(93)90197-Whttp://dx.doi.org/10.1890/07-0862.1http://dx.doi.org/10.1086/285684http://dx.doi.org/10.1111/mec.12535http://dx.doi.org/10.1016/j.biocon.2014.02.033http://dx.doi.org/10.1016/j.biocon.2012.04.029http://dx.doi.org/10.1086/285580

Huang et al.: Humpback dolphins off Taiwan — trends and vulnerability

humpback dolphins (Sousa chinensis): implications forconservation. Bull Mar Sci 88: 885−902

Wilson HB, Kendall BE, Possingham HP (2011) Variability inpopulation abundance and the classification of extinctionrisk. Conserv Biol 25: 747−757

Winship AJ, Trites AW (2006) Risk of extirpation of Stellersea lions in the gulf of Alaska and Aleutian Islands: apopulation viability analysis based on alternative hypo -theses for why sea lions declined in western Alaska. MarMamm Sci 22: 124−155

Yeh CH (2011) Distribution prediction and rangingpattern of Indo-Pacific humpback dolphins (Sousa chi-nensis) in Taiwan. Master’s thesis, National TaiwanUniversity, Taipei

Yu HY, Lin TH, Chang WL, Chou LS (2010) Using the mark-recapture method to estimate the population size of Sousachinensis in Taiwan. Proceedings of Workshop on Popula-tion Connectivity and Conservation of Sousa chinensis offChinese Coast, Nanjing, China. Nanjing Normal Univer-sity, Nanjing

159

Editorial responsibility: Simon Goldsworthy, West Beach, Australia

Submitted: December 27, 2012; Accepted: May 23, 2014Proofs received from author(s): October 24, 2014

http://dx.doi.org/10.1111/j.1748-7692.2006.00009.xhttp://dx.doi.org/10.1111/j.1523-1739.2011.01671.x

cite10: cite12: cite14: cite21: cite23: cite30: cite25: cite32: cite27: cite41: cite4: cite43: cite50: cite34: cite52: cite45: cite29: cite54: cite47: cite61: cite56: cite49: cite63: cite58: cite65: cite70: cite67: cite72: cite69: cite1: cite5: cite9: cite11: cite13: cite20: cite22: cite15: cite24: cite17: cite2: cite26: cite40: cite19: cite28: cite42: cite35: cite37: cite51: cite31: cite39: cite53: cite46: cite55: cite48: cite62: cite66: cite73: cite68: cite3: cite7: