Protection of the Ethiopian Wolf: What are tourists

willing to pay for?

Tafesse Estifanosa,c,* , Maksym Polyakova,b, Ram Pandita, Atakelty Hailua ,

Michael Burtona

aUWA School of Agriculture and Environment, The University of Western Australia, 35

Stirling Highway, Crawley, WA 6009, Australia

bCentre for Environmental Economics and Policy, UWA School of Agriculture and

Environment, The University of Western Australia, Crawley, WA 6009, Australia

cDepartment of Biology, Hawassa University, P.O. Box 05, Hawassa, Ethiopia

*E-mail address: tafesse.estifanos@research.uwa.edu.au/tafekefe@gmail.com

7 May 2018 Working Paper 1803

UWA Agricultural and Resource Economics

http://www.are.uwa.edu.au

UWA Agricultural and Resource Economics

Citation: Estifanos, T., Polyakov, M., Pandit, R., Hailu A., Burton, M. (2018) Protection of the Ethiopian Wolf:

What are tourists willing to pay for? Working Paper 1803, Agricultural and Resource Economics, The

University of Western Australia, Crawley, Australia.

© Copyright remains with the authors of this document.

2

Protection of the Ethiopian Wolf: What are tourists willing to pay for?

Estifanos, T., Polyakov, M., Pandit, R., Hailu A. and Burton, M.

Abstract

Ethiopian wolf, Canis simensis, is among the most threatened carnivore species in Africa.

Habitat loss and disease transmission threaten its survival. Our understanding of the wolf’s

contributions to ecosystem services and economic benefits of its preservation is limited

because there has been insufficient research on economic valuation. This study uses choice

experiments to investigate preferences and willingness to pay (WTP) for conservation

programs among tourists travelling to the Bale Mountains National Park, one of the major

habitats for the wolf. Program attributes valued include size of the protected area, size of the

wolf population, tourist access to wolf habitat and recreational facilities. A random parameter

logit model is fitted to the data to account for heterogeneity in preferences. The results show

that tourists WTP is up to Ethiopian Birr 130.81 (US$ 5.82)/day/trip for enhancement of the

wolf population from 200 to 250 wolves, but very little beyond that level. The visitors are

willing to pay for the increase of protected area and the ease of access to the wolf habitat.

Visitors WTP is significantly influenced by prior visitor experience in other protected areas

in Ethiopia and interest in viewing other unique species in the park. The findings suggest that

non-use values must be taken into account in decisions on the nature and magnitude of more

appropriate Ethiopian wolf conservation programs.

Key words: Afroalpine habitat; choice experiments; endangered species; Ethiopian wolf;

protected area; random parameter logit

JEL classifications

Q26 Recreational Aspects of Natural Resources, Q51 Valuation of Environmental Effects, Q57

Ecological economics: Ecosystem Services; Biodiversity Conservation; Bioeconomics; Industrial

Ecology Government Policy

3

1 Introduction

Large carnivores, top predators, are vitally important for ecological balance and maintaining

biodiversity through trophic cascades (Beschta & Ripple 2009; Ripple et al. 2014). For

example, wolves (Canis lupus) in Banff National Park, Canada may reduce their prey elk

density, and changes in the abundance and diversity of plant species and songbird community

(Hebblewhite et al. 2005). Conservation of large carnivores particularly wolves (canis spp)

poses real challenge (Boman & Bostedt 1999; Williams et al. 2002) because of increased

competition for area, game (Morrison et al. 2007) and human-wildlife conflicts associated

with real or perceived threats to livestock and human safety (Treves & Karanth 2003).

Consequently, wolves have been experiencing extirpation, loss of home range, and

subsequent population decline threatening them with extinction (Marino et al., 2013;

Morrison et al, 2007). Among species of such concern and the least studied wolf species is

the Ethiopian wolf, Canis simensis Ruppell 1840 (Mills et al. 2001; Ray et al. 2005).

The endemic Ethiopian wolf, one of the world’s rarest canid species, is an important flagship

species for conservation of the afrolpine biodiversity (International Union for Conservation

of Nature/Species Survival Commission, UCN/SSC, Canid Specialist Group 2011) but faces

multiple serious threats to its survival (Marino & Sillero-Zubiri 2013; Ripple et al. 2014).

These include 98% of habitat loss, the spread of rabies diseases, declining population and is

classified as an endangered on the IUCN Red List of Threatened Species (Marino et al. 2017;

Marino & Sillero-Zubiri 2013). These threats along with its specialization on threatened prey

may put the Ethiopian wolf populations at greater risk of extinction (Marino & Sillero-Zubiri

2013; Wolf & Ripple 2016). For example, in recent decades the Ethiopian wolf has gone

extinct from two highland areas in northern Ethiopia (Mount Guna) and Gosh Meda, North

Shoa (IUCN/SSC Canid Specialist Group 2011). The remaining wolf populations are

restricted to six small and isolated Ethiopian highland areas: Bale Mountains National Park

4

(BMNP), Seimen Mountains National Park (SMNP), North Wollo, South Wollo, Guassa-

Menz in North Shoa, and Arsi Mountains (Marino 2003). Although there are conservation

efforts in these areas, formal protection to the wolf is provided primarily through two national

parks, BMNP and SMNP. The BMNP, which offers the largest and most suitable afroalpine

habitat area (Marino 2003), hosts more than half of the wolf’s global population, i.e., 250 out

of 400-450 wolves (IUCN/SSC Canid Specialist Group 2011). However, Ethiopian wolf

conservation efforts are being challenged by lack of adequate funding (OARDB 2007).

Conservation programs designed to protect the Ethiopian wolf, over the last two decades

include the development of the Ethiopian Wolf Recovery Programme (EWRP) in 1994

(Sillero-Zubiri & Macdonald 1997), a revised version of the program in 2011 and a National

Action Plan (NAP) developed to protect the species (IUCN/SSC Canid Specialist Group

2011). The revised strategy and NAP promote the Ethiopian wolf as a flagship species for the

benefit of broader biodiversity and the local people. Flagship species, such as the Ethiopian

wolf, have the potential to support biodiversity conservation priorities development through

wildlife tourism (Walpole & Leader-Williams 2002) due to their charisma, endangered status

and ecological significance (Ripple et al. 2014; Macdonald et al. 2017), which in turn has a

potential of generating significant economic incentives for both conservation and local

development (Balmford et al. 2009; Hunt et al. 2015) and create environmental awareness

(Caro & Girling 2010). Cognisant to this, the government of Ethiopia aims to implement

effective management strategies to improve biodiversity conservation and promotes

charismatic and unique species based wildlife tourism (Conservation Capital 2015).

However, empirical evidence required to guide species management options and conservation

programs and their potential to wildlife-based tourism in the Ethiopian protected areas is

lacking (United Nations Development Program-Global Environment Facility, UNDP-GEF,

2008).

5

For wildlife tourism to support predator conservation both public and political support and

effective management are crucial (Macdonald et al. 2017). Although species protection is

preferred by the public (Morse-Jones et al. 2012), willingness to pay for endangered species

are influenced by the change in the size of the species population, type of species, type of

respondents, type of values either use or nonuse or both values (Richardson & Loomis 2009).

The purpose of this study is to address the existing research gap by investigating tourists’

preference and willingness to pay (WTP) for the protection of the Ethiopian wolf in the

BMNP and the factors that influence the tourist’s WTP using a non-market valuation method

– the choice experiment (CE). We focus on international tourists because they form the

majority and growing segment of the park visitors (Conservation Capital 2015), whose

activities may be impacted by the management practices in the BMNP. To this end, we

propose management options defined by a set of attributes (size of the area to be protected,

wolf population, access to wolf habitat, recreational facilities, and park entry fee) where

respondents choose either the existing management (status quo) or alternative management

options to preserve the Ethiopian wolf. A CE, widely used stated preference method, is used

to analyse the trade-off between the attributes and their levels. CE enables us to estimate the

economic values of the proposed programs and evaluate the policy options (Hanley et al.

2001). Further, CE is also applicable to the valuation of threatened or endangered species

with passive use values (Adamowicz et al. 1998; Bateman et al. 2002).

Past valuation studies conducted elsewhere have shown wolves provide recreational services

by attracting tourists (Wilson & Heberlein 1996; Buckley 2012). Such studies include

valuation of the Gray wolf, Canis lupus, in the US by Jorgensen et al. (2001), Chambers and

Whitehead (2003) and Heberlein et al. (2005), and in Sweden by Ericsson et al. (2008) and

Boman and Bostedt (1999). Our CE study on the Ethiopian wolf makes three contributions to

the existing literature. Firstly, it is the first study on the valuation of the endemic wolf in

6

Ethiopia, where such economic studies are non-existent. Secondly, unlike most previous

studies on canid species, our study accounts for the heterogeneity of preferences for

conservation programs among respondents. Thirdly, it considers protection of afroalpine

habitat, a biodiversity-rich unique ecosystem, as a separate attribute giving prominence to a

crucial aspect of Ethiopian wolf conservation program. By incorporating afroalpine area

protection in the study, we complement a few existing studies that have included area

attribute in their CE studies of endangered species (Jacobsen et al. 2012). We used the

random parameter logit model to allow for preference heterogeneity among tourists and

estimate the WTP for the wolf conservation programs. We show that enhancement of area

protection, and, to a limited level, the wolf population, generate substantial economic values

of conservation while recreational facilities are valued negatively by the respondents.

The remaining sections of the paper are organised as follows. Section 2 provides background

on the description of the study area and the Ethiopian wolf. Then, in the methods part, we

present the theoretical framework used and describe how CE data are analysed using a

random utility model (RUM) framework in section 3. Section 3 also describes the study

design, attribute selection, and survey implementation. In section 4 we present the results

from the random parameter logit models and discuss these results in section 5. Finally, we

conclude the paper by providing some policy implications for the conservation of the

Ethiopian wolf.

2 Background

Study Area: Bale Mountains National Park

The BMNP (Fig.1.), situated at 6º29’ – 7º10’N and 39º28’ – 39º57’E, is part of Eastern

Afromontane biodiversity hotspot within the 34 global biodiversity hotspot identified by

Conservation International (Williams et al. 2004). Located 400 km southeast of Addis Ababa

7

in the Oromiya regional state, BMNP covers 2150 km2 area representing one of the most

extensive high altitude plateau and mountain areas in Africa with an altitude ranging from

1500 to 4377 meters above sea level, asl (Hillman 1988). Two distinct climatic seasons: dry

and wet with annual rainfall varying from 520 mm to 2370 mm characterises the park

(Hillman 1986). Regarding its biodiversity, BMNP is well known for its species richness and

high-level endemism (Williams et al. 2004). It hosts 78 mammals (20 of which are endemic),

180 birds (15 endemics), and 17 amphibians (4 endemics) (Hillman 1988; Williams et al.

2004). There are 19 threatened species, including five endangered or critically endangered

species, which exist in the Bale Mountains only (BirdLife International 2012).

BMNP was established in 1971 and formally gazetted in 2015 as a national park. The park’s

primary aim was to conserve the endemic and endangered species, particularly the Ethiopian

wolf (Canis simiensis), and the mountain nyala (Tragelaphus buxtoni) and their unique

habitat (Hillman 1986). The park hosts the largest global population of the two charismatic

species (Sillero-Zubiri & Macdonald 1997). Thus, it has become an essential destination for

rising number of tourists interested in viewing these unique species, bird watching, and

mountain trekking. More importantly, the Ethiopian wolf is regarded as an ‘exceptional

resource value’ for both conservation and wildlife tourism (OARDB 2007). For this reason,

the BMNP is a suitable case study site for CE experiments to value the Ethiopian wolf.

8

Fig. 1. Map of the Bale Mountains National Park BMNP in the Ethiopian highlands

The Ethiopian wolf

The Ethiopian wolf is the most threatened carnivore species in Africa (Marino et al. 2011). It

is known by different names, such as Semien fox, red fox, Semien jackal, and Abyssinian

wolf. However, phylogenetic analysis shows Ethiopian wolf is closely related to the gray

wolf, C. lupus (Gottelli et al. 1994). While they live in social packs, Ethiopian wolves are

solitary hunters of rodents, and most importantly the endemic giant molerat, Tachyoryctes

microcephalus forms the dominant prey in the Bale Mountains (Marino & Sillero-Zubiri

2013). Thus, the wolf density is positively correlated with rodent density but negatively with

9

the height of vegetation in the grassland habitat (Gottelli & Sillero-Zubiri 1992). The density

(the number of wolves/km2) for the three categories of suitable habitat was estimated to be 1-

1.2, 0.2-0.3, and 0.1, for optimal habitat, good habitat, and marginal habitats respectively.

3 Methods

Theoretical background: Choice experiment and random parameter logit

CE, a stated preference survey-based method, is widely applied in environmental valuation

studies (Adamowicz et al, 1998; Hanley et al, 2001). CE evaluates alternative management

options (in our case management options for the preservation of Ethiopian wolf at BMNP)

defined by a set of selected attributes and attribute levels. The theoretical framework for the

analysis of the CE is McFadden’s (1974) random utility model (RUM). In the RUM,

respondents, who face choices, are assumed to choose the alternative providing the maximum

utility. Formally, the utility of individual n choosing alternative i from a set of J alternatives

in choice situation t, for attributes X can be expressed as:

Unit = β'nXnit + εnit, (1)

where Xnit represents a vector of observed attributes, βn is the vector of coefficients for

attributes, and εnjt is an unobserved and random error. As a result, the utility is modelled as

probabilistic. Different assumptions on the distributions of the random term give rise to

various statistical models, including the standard multinomial logit (MNL) and mixed logit

models (Louviere et al. 2000). The mixed logit (random parameter logit, RPL) relaxes the

independence of irrelevant alternatives (IIA) assumption and is superior to the MNL model as

it accounts for the heterogeneity in preferences (Hensher & Greene 2003; Train 2009). Thus,

for respondent in choice situation t, the RPL model probabilities become:

Pnjt =∫(𝑒(𝛽𝑛

′ 𝑋𝑛𝑗𝑡)

∑ 𝑒(𝛽𝑛′ 𝑋𝑛𝑖𝑡)𝑖∈J

) f (β|θ) dβ (2)

10

The vector β is distributed as a normal random variable with density f (β|θ), where θ is a

vector of parameters. Maximum likelihood estimation of the model parameters was carried

out using Stata 14 Econometrics Software.

The estimated models allow us to evaluate the trade-offs respondents make between the

attributes, based on the marginal utilities of the different attributes (Bateman et al. 2002). If

one of the attributes is a cost variable, model coefficients can be used to estimate implicit

prices (WTP) for a discrete change in attribute levels of the non-marketed attributes:

Implicit price (WTP) = -βk/βc (3)

where the βk and βc are the estimated parameters for the non-monetary attribute k and cost

variable c, respectively.

Design of Choice Experiment

3.2.1 Attribute selection

An essential first step in CE design is to characterise the good (environmental or otherwise)

to be evaluated as a set of attributes and attribute levels that are relevant and realistic to

respondents and policymakers (Bennett & Blamey 2001). In this study, the good to be

evaluated is the Ethiopian wolf preservation. The attributes that describe aspects of the

Ethiopian wolf preservation were identified and chosen based on extensive review of the

literature, park documents, expert reviews, and focus group discussioni1. The selection

1The CE questionnaire was refined in two focus group meetings. The first meeting involved with experts

(ecologists’, natural resource experts, and biologists) opinion on the wording of the survey questions and CE

scenarios for clarity. Before the discussion, the draft version of the questionnaire was distributed to participants

so that they could fill it out. This was followed by an open discussion on the questionnaire content and its

clarity. At the end, participants were asked to provide their final written comments on the survey questionnaire.

This approach was also implemented by Carlsson et al. (2003) on their wetland evaluation using CE. The

second meeting with BMNP warden and experts (ecologists, and experts of wildlife, tourism, and community

11

included environmental quality and development of ecotourism access and attributes of the

wolf preservation at the BMNP.

The final CE exercise comprises five attributes: area protection, Ethiopian wolf population,

access to the wolf habitat, recreational facilities (interpretive signs and improved trails) and

entry fee. Table 1 summarises the full description of the attributes and levels. Area (hereafter

‘AREA’) was chosen because habitat protection is the key aspect of biodiversity Ethiopian

wolf conservation for maintaining the population of a species. The four levels of area

represent the existing home range of the wolf in the afroalpine area based on habitat quality,

which determines the wolf density (Sillero-Zubiri et al. 1997). Ethiopian wolf population

(hereafter ‘ETW’) has four levels including the status quo, 200 or fewer wolves. Currently,

the wolf is at risk with a declining population (Marino et al. 2011). Thus, improving the

population of the Ethiopian wolf, a key ecological attribute of BMNP, is one of the

management strategies to avoid the risk of extinction and ensure future survival. To this end,

the improved levels include 250 wolves, the population over several years period; the 300

and 400 wolves is an improvement from immediate risk of extinction. Enhancement of the

wolf population may indicate the effectiveness of wolf preservation.

Two further attributes related to the recreational values included were access to the wolf

habitat (hereafter ‘ACCESS’) with two levels, and development of recreational facilities

(hereafter ‘SIGNSTRAILS’) having four levels of interpretive signs and or improved trails in

the wolf habitat. These two attributes help to examine the trade-offs users make between the

environmental quality and the recreational development in wolf preservation programs.

development) focused on the relevance of attributes to the preservation of the Ethiopian wolf. Finally, park

experts tested the survey before it was pilot tested among tourists to gain further clarity.

12

The last attribute was payment vehicle, entry fee (‘FEE’) aimed at increasing from the current

level to support the Ethiopian wolf protection in the BMNP. The fee attribute has four levels.

Based on its practicability to inform policy options, the park staff evaluated the entry fee and

levels. They suggested the maximum entry fee level at Ethiopian Birr (ETB2) 350 (i.e.,

reduced by ETB 50 from the initial proposal). A park entry fee is a commonly used payment

vehicle in many environmental valuation studies (e.g., Juutinen et al. 2011; Di Minin et al.

2013; Pandit et al. 2015).

2ETB refers to Ethiopian Birr, the currency in Ethiopia, which has an exchange rate of 1 $US = ETB 22.471 at

the time of data collection in December 2016

13

Table 1.

Ethiopian wolf management attributes and levels used in the choice experiment. Bold

italicised levels represent the status quo.

Attribute Description provided in the survey Attribute levels

and (status quo)

Short

name

AREA (in

km2)

The Area protected to preserve the Ethiopian wolf

BMNP comprises the largest afroalpine area and

most of the suitable habitat (ca 1079 km2) for the

Ethiopian wolf. This habitat is categorised into three:

optimal (253 km2), good (329 km2) and marginal

areas (497 km2) based on vegetation quality and

presence of prey rodent population. Currently, all

habitat range in BMNP is a de facto open access

resource for livestock grazing and natural resource

extraction. Four levels of the area were proposed for

strict protection.

de facto open

access

253 km2 (optimal

habitat)

582 km2 (optimal

and good habitat)

1079 km2

(suitable habitat)

AREA0

AREA253

AREA582

AREA1079

ETW Number of the Ethiopian wolves in the park (adult

and sub-adult, >1 year)

BMNP hosts more than 50% the global population of

the Ethiopian wolf. However, in 2015, the wolves’

number had crushed 196 individuals due to rabies

disease outbreak since 2014.

<200

250

300

400

ETW200

ETW250

ETW300

ETW400

ACCESS The extent of visitors access to the wolf habitat

Currently, visitors have unrestricted access

everywhere. Additionally, restricted access such as

walking to only designated trails in core conservation

areas is proposed.

Unrestricted

Restricted to only

designated trails

UNRESTR

ICT

RESTRICT

14

Attribute Description provided in the survey Attribute levels

and (status quo)

Short

name

SIGNSTRAI

LS

The presence of interpretive signs and walking trails

Under the current situation, there are no interpretive

signs or improved trails, which are strictly designated

to use in the wolf habitat. We propose the

establishment of signs and or improved trails (four

levels). Based on this, four levels are proposed.

No interpretative

signs & no

improved trails

Interpretive signs

and trails

Interpretive signs

Improved trails

NO SIGNS

& TRAILS

SIGNS &

TRAILS

SIGNS

TRAILS

FEE (in

ETB)

This is a one-time park entry fee payment per person

per day. Thus, aiming at increasing income four feel

levels are presented here.

This income will be put in a special fund and can be

used only for improvement of the park services and

protection of the Ethiopian wolf

90

150

250

350

FEE

a In Ethiopia, visitors of all age group and residence status are expected to pay park entry fee

set by the Ethiopian Wildlife Conservation Authority. The current park entry fee profile for

international tourists (older than 18 years) is 90 Ethiopian Birr (ETB).

3.2.2 Experimental design and structure of the survey

In total, a full factorial design of the five attributes produced 512 possible combinations (i.e.,

4x4x2x4x4). An orthogonal fractional factorial design using SAS software (Kuhfeld 2004) to

produce 16 choices sets with three alternatives, one of which was the status quo in each case.

Then, the choice sets were blocked into two subsets, each with eight choice sets. A

questionnaire containing one block of eight choice sets was randomly allocated for each

respondent. A sample of choice situation is presented in Appendix Figure.A.1.

15

The final survey questionnaire comprises three main sections other than the introductory

page. The introduction includes the purpose of the study and ethical integrity information.

The first section presented questions regarding environmental attitudes, trip characteristics,

trip purpose, visit activities in the park and park management issues. The second section

included the CE scenario, a reminder of their household income, choice situations, and follow

up questions. Background information about the BMNP, the Ethiopian wolf conservation,

and description of attributes was given in a separate information pamphlet3. Further, colour

map of BMNP and photographs of the Ethiopian wolf were included. Lastly, respondents’

socio-demographic questions were included in the third section of the questionnaire.

Data and implementation of the survey

The survey was conducted in BMNP and administered to international tourists4 (hereafter

‘tourists’), who visited BMNP between November 2016 and February 2017. The period

corresponds to the high –visitation season of the park to achieve adequate sample size for CE

analysis. Before implementation of the survey, formal permission was sought from the park

administration followed by approval on Human Research Ethics from the University of

Western Australia (RA/4/1/8558). The questionnaire was designed in English.

3 The pamphlet and full CE questionnaire is presented at Appendix C.

4Our survey included only international tourists for two reasons. Firstly, they form the majority of visitors of

protected areas in Ethiopia and BMNP in particular. For example, during our data collection period in

2016/2017, the composition of BMNP visitors’ was 56, 33, and 11% for international tourists, foreign resident

visitors, and national visitors, respectively (BMNP unpublished data). Secondly, coupled with low rate of

visitation by national visitors, the social instability and subsequent state of emergency imposed in the region by

the government during the data collection did not allow us to collect sufficient data from domestic visitors.

Overall, the data collection was challenged by the security situation that limited the park visitation. BMNP

tourists’ data for the year 2016/2017 showed that the international visitors’ number was 1709 only, with a

significant reduction by 40% compared to previous year.

16

The survey was conducted on-site (at recreation sites, accommodation places inside and

outside the park) and administered to the tourists after they visited the park. Two well-trained

university graduate enumerators and the first author conducted the in-person interview during

mid-weeks and weekends. As all park visitors should report at the park’s main entrance on

arrival, and be accompanied by local park guides, survey participants were easily identified

for the interview. In case of tourists’ time limitation, the questionnaires were given for later

collection.

4 Results

Descriptive statistics

Of 270 questionnaires distributed to tourists, 248 were completed and returned. Twelve

responses were excluded due to incomplete information, resulting in a full sample of 236

respondents. Among these, 68 (28.8%) respondents who always choose the status quo on all

choice occasions were identified to be protesting5 the payment mechanism and were excluded

from the statistical analysis.

The tourists come from 30 different countries, with the majority (55.1%) being from Europe

followed by USA and Canada (23.7%). This appears to be consistent with the general trend

for tourist profile from the park archive (Conservation Capital 2015). A large proportion of

respondents had attained tertiary education (59%), were male (55%), and the average age was

37 years ± 10.95 standard deviation. Close to half or 47% had monthly incomes in the range

US$ 2001-4500, while just over a fifth or 23.2% had incomes in the US$ 4501-6500 range.

5 Protest responses were identified based on agreement of respondents’ responses (1 “strongly disagree” to 5

“strongly agree) to the following debriefing statement on a five-point Likert scale: “the government should pay

for protection of the wildlife and its habitat; Such form of debriefing statement for the payment mechanism was

also used by Jorgensen et al (2001) in their valuation of wolves

17

19.1% and 10.7% were in the lowest (less than US$ 2000) and the highest (above US$ 6500)

income ranges, respectively. Fifty-eight percent of respondents were first-time visitors to

Ethiopian protected areas. About 64.1% had either prior information or seen the Ethiopian

wolf before this survey, and 44.3% stated that viewing the Ethiopian wolf was one of the

purposes of their visit to the park.

Random parameter logit model results

4.2.1 Random parameter logit model specifications

The CE indirect utility function was assumed to be linear and additive in parameters for the

attributes. An alternative specific constant, ASC (status quo =1) was included to capture the

mean effect of unobserved factors not explained by the attributes and interaction terms, and

also possible status quo bias (Adamowicz et al. 1998). All the attributes were dummy coded

except the entrance fee, which was a continuous variable.

We carried out an extensive modelling of the various interactions of the ASC and the

attributes with respondent’s socio-demographic6 variables and trip characteristics (summary

presented in Appendix Table A.1). The following interactions improved the model fit.

Interactions of the ASC with whether the respondent was older than 40 years (‘AGE40’), if

respondents were interested in trekking (‘PURP_TREKK’), if they had visited another

national park that hosts small population of Ethiopian wolf (or the SMNP) during the current

trip (‘VISITED_SMNP’), whether respondents intended to visit BMNP in the future

(‘VISIT_BMNP’), if the wolf habitat condition at BMNP is best (‘HABITAT_BEST’), and

whether the respondent belonged to an environmental group (‘MEMBER’). The attributes,

AREA variables were interacted with whether respondents considered the park scenery

6 Interactions of the ASC and the attributes with other socio-demographic characteristics (e.g. gender, income,

and education) were modelled but they were insignificant. Summary

18

attractive (‘HIGH_ATTRACT’). The ETW population variables were interacted with

whether respondents had visited protected areas in Ethiopia previously (‘PREV_VISIT’), and

if respondents had viewed other unique wildlife species during the current park visit

(‘VIEWUNQWL’). The attribute interactions help to examine the presence of observed

preference heterogeneity of respondents (Hensher et al. 2005b).

In our RPL modelling, all the attribute variables coefficients except for the FEE variable were

assumed to be normally distributed (Carlsson et al. 2003; Hensher et al. 2005a; Birol et al.

2006). That means respondents may have positive or negative utilities for the attributes of

environmental quality (area protection and Ethiopian wolf population) and recreational

attributes (access and facilities) depending on their socio-demographic characteristics and

environmental affiliations (Ericsson et al. 2008; Williams et al. 2002). However, the cost

attribute (FEE) is considered fixed in parameter to ease estimation of WTP and reduce

econometric complexity7 (Hensher et al. 2005a).

4.2.2 Parameter estimates and relative importance of the attributes

Table 2 presents the RPL model estimated using Halton draws with 1000 replications (Train

1999). Results presented here excluded the “protest” respondents8. The coefficient for ASC

was negative and significant, indicating that there is a preference for change from the status

quo independent of the attribute levels. Considering first the attributes without interactions,

AREA582, AREA1079, ETW250, ETW300 and ETW400, and UNRESTRICT were positive

and significant. For the AREA variables, respondents had higher and positive preference for

7 The parameter for fee attribute was modelled using a lognormal distribution but this did not converge. In

analyses not shown here, we also estimated a conditional logit model and the results are available in the

Appendix Table A.2.

8 Results from the full sample model (including the 68 protest responses) varied little from what is presented

here. Those results can be found at the Appendix Table A.3.

19

larger areas (AREA582 and AREA1079), than for the smaller AREA253 protection. The

coefficients for AREA582 and AREA1079 were not statistically different (Wald test, Chi2

(2) = 1.83, p=0.1759). Similarly, respondents had a positive preference for having wolf

populations higher than 200, but there was no scope effect for such population increase

(Wald test, Chi2 (2) = 0.16, p=0.9219).

The interaction effects of socio-demographic on the mean of the distributions indicate a

degree of observable heterogeneity in preferences. The negative and significant coefficients

for the interactions of ETW levels with PREV_VISIT and VIEWUNQWL, except for the

ETW250 with VIEWUNQWL, indicates that respondents with previous experience in

Ethiopian protected areas and had viewed other unique wildlife species in the park tended to

have lower levels of utility associated for an increase in the wolf population.

The FEE and recreational development attributes were also significant. The FEE was

negative and highly significant, meaning respondents’ prefer lower entry fees, ceteris

paribus. By contrast, UNRESTRICT access attribute is positive and significant, suggesting a

recreational use value of the BMNP. However, the coefficients for the interpretation SIGNS

only, and SIGNS AND TRAILS were negative and significant, implying that respondents’

disutility for the establishment of these facilities in the wolf habitat.

Furthermore, the standard deviation of the random parameters, except for AREA253, was

significant suggesting an unobserved heterogeneity in tastes for all other attributes. In

particular, the highly significant and large standard deviation for AREA582, AREA1079,

SIGNS, and TRAILS is reflecting a greater variability in preferences for the respective

attributes levels.

20

Table 2

Random parameter logit model estimation results

Mean Standard deviations

Attributes Coefficients Std. Errors Coefficients Std. Errors

ASC -3.413*** 0.6586

AREA253 0.0767 0.3294 -0.319 0.2856

AREA582 0.9097* 0.4391 2.037*** 0.3188

AREA1079 1.7829** 0.6108 3.777*** 0.5591

ETW250 2.0115** 0.6566 1.055*** 0.3360

ETW300 1.7596** 0.5352 0.904*** 0.2797

ETW400 1.8787** 0.6228 0.867† 0.4523

UNRESTRICT 0.6550*** 0.1606 0.941*** 0.2131

SIGNS &TRAILS -1.1109** 0.3413 1.088*** 0.3548

SIGNS -1.2606*** 0.3379 1.686*** 0.2988

TRAILS -0.3606 0.2837 1.847*** 0.3067

FEE -0.0068*** 0.0010

Interactions

AREA253*HIGH _ATTRACT -0.6793* 0.3451

AREA582*HIGH _ATTRACT 0.0106 0.4951

AREA1079*HIGH _ATTRACT 1.0030 0.7087

ETW250*VIEWUNQWL -0.9377 0.6410

ETW300*VIEWUNQWL -1.5075** 0.5384

ETW400*VIEWUNQWL -1.6163** 0.6149

ETW250*PREV_VISIT -0.7123† 0.4122

ETW300*PREV_VISIT -0.7708* 0.3818

ETW400*PREV_VISIT -1.5056** 0.4662

ASC*AGE40 0.3909 0.3817

ASC*VISITED_SMNP -1.4054* 0.6151

ASC*PURP_TREKK 1.1812** 0.4544

ASC*VISIT_BMNP -0.2507 0.3752

ASC*HABITAT_BEST -0.9336* 0.4668

ASC*MEMBER -1.4235* 0.5856

Model statistics

AIC 1816.0005

BIC 2047.051

Log-Likelihood -871.0003

No. of observations 3807

No. of respondents 168 b Asterisks denote significance level: *** p<.001, ** p<.01; * p<.05; † p<.1

21

Considering the interactions of the ASC with respondent characteristics, except for

VISIT_BMNP and AGE40, three of them (VISITED_SMNP, HABITAT_BEST, and

MEMBER) appeared significant and negative. The negative signs imply those respondents

who visited SMNP, considered BMNP as the best habitat for the wolves, or were members of

an environmental group were more likely to choose alternative management options.

However, the positive and significant coefficient for PURP_TREKK shows that tourists

interested in trekking were more likely to prefer the status quo to management options.

4.2.3 The willingness to pay estimates of management attributes

Table 3 reports the WTP estimates for the management attributes evaluated at the mean of the

random distribution and at the mean values for the socio-demographic variables, i.e., for an

‘average respondent’. Compared to the de facto open access situation and except for the

smallest area (253 km2), the WTP values for the area had increases with the size of area

protection (ETB 134.85 for 582 km2) with the highest value (ETB 349.49) for 1079 km2.

However, the WTP for the wolf population was significant and positive for 250 wolves (ETB

130.81) but was much lower or negative and statistically non-significant for the population

levels beyond 250 wolves in comparison with the 200 wolves (baseline). The mean WTP

estimates for the wolf populations varied for different groups of respondents, depending on

their prior visit experience and visit activity inside BMNP (Fig. 2). The WTP value of

respondents who had viewed other unique species and previously visited Ethiopian protected

areas was lower and negative for higher wolf populations. Contrary to this, all other groups,

especially those with no prior visit experience and had not viewed other unique species had

positive WTP values for all enhanced wolf population levels.

Figure 3a-c shows the RPL model simulations, considering the socio-demographics, of the

individual level WTP distributions for the wolf population and the area levels, and the

22

programs that involved combinations of the two attributes levels. For the wolf populations, a

vast majority of respondents had positive WTP values for 250 wolves with positive mean

WTP value; but a substantial proportion of respondents had negative WTP for the 300 and

the 400 wolves, having their mean WTP estimates shifted towards zero or negative values

(Fig.3a). The disutility for the 300 and the 400 wolves is more likely attributed to the

combined effect of several sets of individual specific characteristics. Conversely, for the area

attribute levels, a large proportion of respondents’ WTP was positive for increased area

protection of 582 km2 and 1079 km2, with a wide range of variation and higher WTP values

for the extensive area coverage, 1079 km2 (Fig 3b).

Based on the premise that area protection would enhance the wolf population, the WTP

distributions for the combination of the attribute levels of AREA and ETW was simulated

(Fig. 3c). Overall, large proportion of respondents had positive WTP values for the programs

that target simultaneous protection of the area and the Ethiopian wolf population, with an

increase in the WTP for the higher levels. Respondents WTP for the protection of the entire

suitable habitat range (ca 1079km2) and 400 wolves (ETW400 & AREA1079) was higher

(mean WTP ETB 308, ~ US$13.7), having much more variability than for the combinations

of lower levels, ETW300 &AREA582 (ETB, 158, ~ US$ 7) and ETW250 &AREA253 (ETB

83, ~ US$ 3.7). Detail analysis showed that the proportion of respondents with positive WTP

for ETW400 & AREA1079 was lower than for the ETW300 &AREA582, with respondents

having much higher WTP for the former, suggesting a high level of commitment.

Regarding the recreational development attributes, the WTP values with no interactions were

estimated. Respondents had positive and significant WTP value for unrestricted access (ETB

96.43). By comparison, WTP for interpretive SIGNS AND TRAILS, or just SIGNS were

negative (ETB -163.54 and -185.59, respectively), implying a disutility for the presence of

these recreational facilities inside the wolf habitat.

23

Table 3

Mean Marginal WTP (per person per day) of attribute levels and their confidence interval

Attribute level WTP (in ETB) [95% Conf. Interval]

Area for strict protection

253 km2 (optimal habitat) -47.64 [-118.63, 23.36]

582 km2 (good and optimal habitat) 134.85 [32.53, 237.1708]*

1079 km2(all suitable habitat) 349.49 [191.30, 507.67]***

Ethiopian Wolf population

250 wolves 130.81 [32.29, 229.32]**

300 wolves 16.25 [-56.78, 89.29]

400 wolves -25.39 [-98.07, 47.28]

Access to wolf habitat

Unrestricted access 96.43 [51.95, 140.91]***

Recreational facilities

Signs and Trails -163.54 [-251.15, -75.92]***

Signs only -185.59 [-272.18, -98.98]***

Improved trails NSc NS

NS denotes the coefficients were not significant to estimate the WTP

Fig. 2. Tourists’ WTP for different wolf population levels by respondents’ type

Average (of all

respondents)

Visitors with

previous visit

expereince

Visitors viewed

other unique

species

Visitors with

prior visit

experience and

viewed other

unique species

First time

visitors and

viewed

Ethiopian wolf

only

250 wolves 130.81 191.25 158.07 53.2 296.12

300 wolves 16.25 145.56 37.11 -76.36 259.04

400 wolves -25.39 54.93 38.64 -183.01 276.58

-300

-200

-100

0

100

200

300

400

WT

P (

in E

TB

)

24

Finally, the estimate of WTP for a change in a policy attribute can be translated into an

aggregate value generated by such a change. Suppose the policy goal is increasing the

endangered Ethiopian wolf population, and respondents were WTP for the wolf population to

rise to 250 wolves. This estimate can be translated into an annual benefit equal to ETB

130.81 per tourist per day multiplied by the total number of annual visits to the BMNP.

According to the annual tourists’ visitation data for the last five years (BMNP unpublished

data, 2017), there were, on average, 2746 tourists who visited BMNP annually, and the

average park visit days based on our survey was 4.1. This resulted in an annual benefit

estimate of ETB 1.473 million (ranging ETB 0.364–2.582 million), which is higher than the

annual government budget (ETB 1.16 million) allocated for the management of BMNP for

the year 2015/16, excluding salaries. This estimated value can be considered as benefits,

given the cost of livestock depredation or other socioeconomic costs to local community due

to wolves presence at BMNP is almost non-existent (Atickem et al. 2010).

25

1

Fig. 3. Individual specific WTP distributions for the Ethiopian wolf population levels, area protection, and the combinations of programs2

26

5. Discussion

5.1. Willingness to pay for Ethiopian wolf conservation

Consistent with previous studies that have shown that protection of species is valued by the

public (Morse-Jones et al. 2012; Jacobsen et al. 2012), our research focusing on the Ethiopian

wolf provides evidence that tourists valued conservation programs that targeted an increase, to

a limited level, of the wolf population and size of habitat protection. This suggests the

fundraising potential of such conservation practices relative to recreational development

programs.

The significant differences in WTP for the afroalpine area levels under protection, other than

the smallest area (253km2), suggest that larger protected area improve the economic value of

conservation programs. The WTP for the protection of all suitable habitat for the wolf (1079

km2) was estimated to be ETB 349.5 (~ US$ 15.55), which is about 2.6 times higher than the

WTP (ETB 135, ~ US$ 6.00) for the optimal and good habitat area only (582 km2). The first

WTP estimate is comparable to the WTP ($14.86) estimated for the protection of endangered

species habitat in the Grand Canyon National Park (Mueller et al. 2017). Our higher WTP

estimate for area protection compared with all the other attributes might be attributed to the

park’s spectacular scenery and landscape feature, besides its rich biodiversity value (UNESCO,

2008; Williams et al. 2004). This can also be supported by the positive and substantial WTP

for unrestricted access to the afroalpine wolf habitat, suggesting the importance of other nature-

based features of the park.

On the other hand, there is a significant difference in the WTP for the different wolf

populations valued in the study (positively significant for 250 wolves versus non-significant for

300 and 400 wolves). Higher Ethiopian wolf population numbers did not attract higher WTP

27

values. Though such result was unexpected, it could be explained by two possible reasons.

First, respondents were informed that the wolf population at BMNP had been 250 wolves’ over

the past several years but had declined to less than 200 (our baseline) in 2015. It is possible that

such background information has influenced what respondents consider is a viable or stable

population level (Adamowicz et al. 1998). Second, a significantly large proportion (about 75%)

of the respondents9 were from Europe and the US, where the public is less likely to have strong

positive attitudes towards wolves or had previous experience with wolves (Williams et al.

2002). Thus, respondents might consider increasing the predator population beyond a certain

level as undesirable for the public (Heberlein et al. 2005).

Although increasing wildlife population may support both use and non-use values (Jacobsen et

al. 2012), the latter, mainly due to the existence value, seems to be the predominant motivation

of the WTP for the Ethiopian wolf. This appeared to be consistent with studies of the gray wolf

in Sweden and the US that considered wolves controversial (Boman & Bostedt 1999;

Chambers & Whitehead 2003). Boman and Bostedt (1999) suggested that people might have

valued the perceived ‘secured survival’ of the species, not the size of the population, and

possibly people might be interested in saving the endangered species only but not in higher

populations (Jacobsen et al. 2012).

The preservation of endangered species also involves protection of their habitat. Our finding

also suggested that preservation programs that target both the wolf population and its habitat

have more value than programs targeting just wolf numbers; such programs are also likely to

gain more financial support. Thus, our study has implications for the national action plan

(NAP) for the preservation of the Ethiopian wolf, which aims “to secure viable and

9 Respondents origin (Europe and/or US=1, 0 otherwise) was modelled by interacting with the attribute ETW

levels. The coefficients for interactions, not presented here, was negative and insignificant for all ETW levels.

This might suggest that respondent originated from such countries did not have any particular difference in

preferences for the wolf population.

28

ecologically functioning Ethiopian wolf populations and protect suitable range, and to

emphasise its role as a flagship for the conservation” (IUCN/SSC Canid Specialist Group

2011)”. The results can also be compared with estimates from a different but related

conservation context obtained by Chambers and Whitehead (2003) for the gray wolf in the US.

They find that there were substantial economic values (US$ 22.64 per household) for

maintaining a stable population of 1600 gray wolves and protection of its habitat. Our WTP

estimate for preserving 400 Ethiopian wolves and its suitable habitat is lower at ETB 308 (US$

14.7). Of course, a comparison of values across studies is not straightforward. Further, our

sample of respondents was dominated by relatively low-income groups. Despite this, our study

also provides evidence that people attach significant positive economic value to the wolves’

protection.

Previous findings on how access to wildlife habitat is valued are mixed, i.e., people may have

positive or negative WTP for different levels of restriction (Jacobsen et al. 2012). In our case,

respondents are WTP (ETB 96.43 or 4.3 US$) for unrestricted access to the wolf habitat,

suggesting the importance of passive use values. Additionally, analysis of tourists’ responses to

survey questions shows that a fair proportion of respondents (44.3%) stated that the purpose of

their trip was to view the Ethiopian wolf and the majority (75.1%) said they would recommend

a visit to BMNP to view the wolf.

Concerning values for recreational facilities, the significantly negative WTP for the interpretive

signs and trails, and signs only may be attributed to the profile of tourists’ visiting BMNP. It is

known that the majority of tourists are wildlife specialists and mountain trekkers (Kinahan

2011). For a wildlife tourism site with increasing number of visitors, like the BMNP, Duffus

and Dearden (1990) argued that initial users tend to be dominated by wildlife specialists, who

pay little attention for infrastructure and interpretational facilities, but consider social and

29

ecological characteristics most important. Wildlife tourism in BMNP could exemplify this

because it is in its infancy and is characterised by poor infrastructure.

5.2. Tourists’ wildlife viewing preference heterogeneity

The RPL model results provide evidence of preference heterogeneity for all management

attributes and levels except the value of the smallest preservation area considered (AREA253).

To this end, our findings show that previous visit experience to protected areas in Ethiopia and

experience of having viewed other unique species in the park influenced respondents’

preference as well as their WTP for increasing the wolf population. That is, first-time visitors

and those who had not viewed other unique species are more likely to have a positive and

higher WTP for the higher levels, though scope insensitivity exists (Heberlein et al. 2005). On

the contrary, return visitors or those who viewed other unique species have less WTP and those

with both features have negative WTP for the 300 and 400 wolves. This result conforms

previous arguments that visitors experience is an important factor that affects preferences for

wildlife species (Lindsey et al. 2007; Di Minin et al. 2013).

According to Lindsey et al. (2007), inexperienced overseas visitors to South African protected

areas were interested primarily in large predators whereas experienced wildlife viewers tend to

show viewing preferences for a diverse and non-charismatic species including bird, plant

diversity, and scenery. The latter is more likely to support conservation that targeted the

broader biodiversity than the charismatic species alone (Hausmann et al. 2017). Analysis of

such heterogeneity in preferences regarding preservation of species is vital for the

understanding of who will be affected by a policy change (Boxall & Adamowicz 2002). Thus,

park management strategies and conservation policies need to address both preferences,

conservation of the charismatic species, such as the wolves, and also maintenance of broader

biodiversity to achieve conservation goals and promote ecotourism.

30

6. Conclusion

This study uses a CE survey of tourists to estimate economic benefits of preserving the

Ethiopian wolf in the BMNP, Ethiopia. Increases in the size of the protected area and, to a

limited extent, of the wolf population are found to generate significant economic value.

Tourists also valued unrestricted access to the wolf habitat. This economic value is influenced

by respondents’ characteristics, previous visit experience to Ethiopian protected areas and

interest in viewing other unique wildlife species. The results suggest that preservation of the

wolf generates both use and non-use values, the latter possibly dominating due to the existence

value of securing the wolf’s survival.

Our results have implications for the conservation of the Ethiopian wolf in the BMNP and

Ethiopia in general. Firstly, it highlights the role the BMNP itself has in the biodiversity

protection. Currently, the critical issue of the wolf conservation is securing the afroalpine

habitat. Indeed, the positive economic benefit generated by the protected area and that of the

combined program of largest habitat supporting higher wolf population could help justify the

need for enforcing effective habitat protection in the BMNP, and possibly in other wolf

inhabited Ethiopian highland areas, to ensure the future survival of the species.

Secondly, for a policy in which conservation supports ecotourism, the Ethiopian wolf could

deliver substantial economic benefit, especially from new visitors. Attractions of more

experienced overseas tourists, who are interested in viewing other unique species, require

protection of the broader biodiversity including the unique species in the park. Creating a better

understanding of what such a broader conservation strategy requires further research to

evaluate public preferences for the broader biodiversity and using data for both domestic and

international visitors. Despite this caveat, the overall evidence from our study is that

conservation of large carnivore species such as the Ethiopian wolf can generate significant

economic benefits and that these benefits need to be considered in designing future programs.

31

Acknowledgment

This research was supported by funding from the Australian Government Research Training

Program, Graduate Research School Travel Award, University of Western Australia (UWA)

and, the Australian Research Council Centre of Excellence for Environmental Decisions

(CEED).

References

Adamowicz, W., Boxall, P., Williams, M., Louviere, J., 1998. Stated Preference Approaches

for Measuring Passive Use Values: Choice Experiments and Contingent Valuation.

American Journal of Agricultural Economics 80, 64-75

Ashenafi, Z.T., Coulson, T., Sillero-Zubiri, C., Leader-Williams, N., 2005. Behaviour and

ecology of the Ethiopian wolf (Canis simensis) in a human-dominated landscape outside

protected areas. Animal Conservation 8, 113-121

Balmford, A., Beresford, J., Green, J., Naidoo, R., Walpole, M., Manica, A., 2009. A Global

Perspective on Trends in Nature-Based Tourism (Global Trends in Nature-Based

Tourism). PLoS Biology 7, e1000144

Bateman, I.J., Carson, R.T., Day, B., Hanemann, M., Hanley, N., Großbritannien, 2002.

Economic valuation with stated preference techniques: a manual. Edward Elgar,

Cheltenham, UK.

Bedin, E., Marino Jorgelina , Sillero-Zubiri, C., & Team, E. (2015). EWCP Annual Report:

Ethiopian wolf conservation Programme.

Bennett, J., Blamey, R., 2001. The Choice Modelling Approach to Environmental Valuation.

Edward Elgar Publishing Limited, Cheltenham.

Beschta, R.L., Ripple, W.J., 2009. Large predators and trophic cascades in terrestrial

ecosystems of the western United States. Biological Conservation 142, 2401-2414

BirdLife International, 2012. Ecosystem profile: Eastern Afromontane Biodiversity Hotspot

Critical Ecosystem Partnership Fund (CEPF). Conservation International, Arlington, USA.

Birol, E., Karousakis, K., Koundouri, P., 2006. Using a choice experiment to account for

preference heterogeneity in wetland attributes: The case of Cheimaditida wetland in

Greece. Ecological Economics 60, 145-156

Boman, M., Bostedt, G., 1999. Valuing the Wolf in Sweden: are Benefits Contingent on the

Supply? Topics in Environmental Economics. Kluwer Academic Publishers, Dordrecht,

pp. 157–174.

32

Boxall, P.C., Adamowicz, W.L., 2002. Understanding Heterogeneous Preferences in Random

Utility Models: A Latent Class Approach. Environmental and Resource Economics 23,

421-446

Buckley, R., 2012. Sustainable tourism: Research and reality. Annals of Tourism Research 39,

528-546

Carlsson, F., Frykblom, P., Liljenstolpe, C., 2003. Valuing wetland attributes: an application of

choice experiments. Ecological Economics 47, 95-103

Caro, T.M., Girling, S., 2010. Conservation by proxy: indicator, umbrella, keystone, flagship,

and other surrogate species. Island Press, Washington.

Chambers, C.M., Whitehead, J.C., 2003. A Contingent Valuation Estimate of the Benefits of

Wolves in Minnesota. Environmental and Resource Economics 26, 249-267

Conservation Capital, 2015. Bale Mountains National Park Tourism Development Plan

February 2015, Addis Ababa, Ethiopia.

Di Minin, E., Fraser, I., Slotow, R., MacMillan, D.C., 2013. Understanding heterogeneous

preference of tourists for big game species: implications for conservation and

management. Animal Conservation 16, 249-258

Duffus, D.A., Dearden, P., 1990. Non-consumptive wildlife-oriented recreation: A conceptual

framework. Biological Conservation 53, 213-231

Ericsson, G., Bostedt, G., Kindberg, J., 2008. Wolves as a Symbol of People's Willingness to

Pay for Large Carnivore Conservation. Society & Natural Resources 21, 294-309

Gottelli, D., Sillero-Zubiri, C., 1992. The Ethiopian wolf – an endangered endemic canid. Oryx

26, 205-214

Gottelli, D., Sillero-Zubiri, C., Applebaum, G.D., Roy, M.S., Girman, D.J., Garcia-Moreno, J.,

Ostrander, E.A., Wayne, R.K., 1994. Molecular genetics of the most endangered canid: the

Ethiopian wolf Canis simensis. Molecular Ecology 3, 301-312

Hanley, N., Mourato, S., Wright, R.E., 2001. Choice Modelling Approaches: A Superior

Alternative for Environmental Valuatioin? Journal of Economic Surveys 15, 435-462

Hausmann, A., Slotow, R., Fraser, I., Di Minin, E., 2017. Ecotourism marketing alternative to

charismatic megafauna can also support biodiversity conservation. Animal Conservation

20, 91-100

Hebblewhite, M., White Clifford, A., Nietvelt Clifford, G., McKenzie John, A., Hurd Tomas,

E., Fryxell John, M., Bayley Suzanne, E., Paquet Paul, C., 2005. HUMAN ACTIVITY

MEDIATES A TROPHIC CASCADE CAUSED BY WOLVES. Ecology 86, 2135-2144

33

Heberlein, T.A., Wilson, M.A., Bishop, R.C., Schaeffer, N.C., 2005. Rethinking the scope test

as a criterion for validity in contingent valuation. Journal of Environmental Economics and

Management 50, 1-22

Hensher, D., Shore, N., Train, K., 2005a. Households’ Willingness to Pay for Water Service

Attributes. Environmental and Resource Economics 32, 509-531

Hensher, D.A., Greene, W.H., 2003. The Mixed Logit model: The state of practice.

Transportation 30, 133-176

Hensher, D.A., Rose, J.M., Greene, W.H., 2005b. Applied choice analysis : a primer,

Cambridge.

Hillman, J.C., 1986. Conservation in Bale Mountains National Park, Ethiopia. Oryx 20, 89-94

Hillman, J.C., 1988. The Bale Mountains National Park Area, Southeast Ethiopia, and Its

Management. Mountain Research and Development 8, 253-258

Hunt, C.A., Durham, W.H., Driscoll, L., Honey, M., 2015. Can ecotourism deliver real

economic, social, and environmental benefits? A study of the Osa Peninsula, Costa Rica.

Journal of Sustainable Tourism 23, 339-357

IUCN/SSC Canid Specialist Group, 2011. Strategic plan for Ethiopian wolf Conservation.

IUCN/SSC Canid Specialist Group. Oxford, United Kingdom.

Jacobsen, J.B., Lundhede, T.H., Thorsen, B.J., 2012. Valuation of wildlife populations above

survival. Biodiversity and Conservation 21, 543-563

Jorgensen, B.S., Wilson, M.A., Heberlein, T.A., 2001. Fairness in the contingent valuation of

environmental public goods: attitude toward paying for environmental improvements at

two levels of scope. Ecological Economics 36, 133-148

Juutinen, A., Mitani, Y., Mäntymaa, E., Shoji, Y., Siikamäki, P., Svento, R., 2011. Combining

ecological and recreational aspects in national park management: A choice experiment

application. Ecological Economics 70, 1231-1239

Kinahan, A., 2011. Bale Mountains National Park Business and Sustainable Finance Plan

2011‐2016. FZS/BMNP publication. pp. 1-30, Addis Ababa

Kuhfeld, W., F. , 2004. Marketing Research Methods in SAS: Experimental Design, Choice,

Conjoint, and Graphical Techniques, NC, SAS Inc. Institute.

Lindsey, P.A., Alexander, R., Mills, M.G.L., Romañach, S., Woodroffe, R., 2007. Wildlife

Viewing Preferences of Visitors to Protected Areas in South Africa: Implications for the

Role of Ecotourism in Conservation. Journal of Ecotourism 6, 19-33

Louviere, J.J., Swait, J., Hensher, D.A., 2000. Stated choice methods : analysis and application,

Cambridge.

34

Macdonald, C., Gallagher, A.J., Barnett, A., Brunnschweiler, J., Shiffman, D.S.,

Hammerschlag, N., 2017. Conservation potential of apex predator tourism. Biological

Conservation 215, 132-141

Marino, J., 2003. Threatened Ethiopian wolves persist in small isolated Afroalpine enclaves.

Oryx 37, 62-71

Marino, J., Sillero-Zubiri, C., 2013. Canis simensis. The IUCN Red List of Threatened Species

2013.

Marino, J., Sillero-Zubiri, C., Deressa, A., Bedin, E., Bitewa, A., Lema, F., Rskay, G.,

Banyard, A., Fooks, A.R., 2017. Rabies and Distemper Outbreaks in Smallest Ethiopian

Wolf Population. Emerging Infectious Diseases 23, 2102-2104

Marino, J., Stewart, A., Gordon, C., Gottelli, D., Tefera, Z., Laurenson, M., Sillero-Zubiri, C.,

2011. Ethiopian wolf status review. pp. 1-50: Strategic planning for Ethiopian wolf

conservation. IUCN/SSC Canid Specialist Group (eds.). Oxford, United Kingdom.

McFadden, D., 1974. Conditional logit analysis of qualitative choice behavior. Academic

Press, New York, NY.

Mills, M.G.L., Freitag, S., Van Jaarsveld, A.S., 2001. Geographic priorities for carnivore

conservation in Africa. In: Gittleman JL, Funk SM, Macdonald DW & Wayne RK (eds.)

Carnivore conservation. Cambridge University Press, Cambridge, UK., pp. 467-483.

Morrison, J.C., Sechrest, W., Dinerstein, E., Wilcove, D.S., Lamoreux, J.F., 2007. Persistence

of Large Mammal Faunas as Indicators of Global Human Impacts. Journal of Mammalogy

88, 1363-1380

Morse-Jones, S., Bateman, I.J., Kontoleon, A., Ferrini, S., Burgess, N.D., Turner, R.K., 2012.

Stated preferences for tropical wildlife conservation amongst distant beneficiaries:

Charisma, endemism, scope and substitution effects. Ecological Economics 78, 9-18

Mueller, J.M., Lima, R.E., Springer, A.E., 2017. Can environmental attributes influence

protected area designation? A case study valuing preferences for springs in Grand Canyon

National Park. Land Use Policy 63, 196-205

OARDB, 2007. Bale Mountains National Park General Management Plan. Oromia Agriculture

and Rural Development Bureau, Oromia National Regional State, Ethiopia.

Pandit, R., Dhakal, M., Polyakov, M., 2015. Valuing access to protected areas in Nepal: The

case of Chitwan National Park. Tourism Management 50, 1-12

Ray, J.C., Hunter, L., Zigouris, J., 2005. Setting Conservation and Research Priorities for

Larger African Carnivores. WCS Working Paper 24: Wildlife Conservation Society, New

York. 1–203.

35

Richardson, L., Loomis, J., 2009. The total economic value of threatened, endangered and rare

species: An updated meta-analysis. Ecological Economics 68, 1535-1548

Ripple, W.J., Estes, J.A., Beschta, R.L., Wilmers, C.C., Ritchie, E.G., Hebblewhite, M.,

Berger, J., Elmhagen, B., Letnic, M., Nelson, M.P., Schmitz, O.J., Smith, D.W., Wallach,

A.D., Wirsing, A.J., 2014. Status and Ecological Effects of the World’s Largest

Carnivores. Science 343

Sillero-Zubiri, C., Macdonald, D., 1997. The Ethiopian wolf: status survey and conservation

action plan. IUCN/SSC Canid Specialist Group, Gland, Switzerland, and Cambridge, UK.

Train, K., 1999. Halton sequences for mixed logit. Working Paper Department of Economics,

University of California, Berkeley.

Train, K., 2009. Discrete Choice Methods with Simulation. Cambridge university press,

Cambridge, UK.

Treves, A., Karanth, K.U., 2003. Human-Carnivore Conflict and Perspectives on Carnivore

Management Worldwide. Conservation Biology 17, 1491-1499

UNDP-GEF, 2008. Full Project: Sustainable Development of the Protected Area System of

Ethiopia (SDPASE). United Nations Development Program- Global Environment Facility

PIMS 494, Resubmission Ethiopia PAS PIMS 494 Prodoc, Washington, DC 20433 USA.

UNESCO, 2008. Bale Mountains National Park in UNESCO Tentative World Heritage List

URL http://whc.unesco.org/en/tentativelists/5315

Walpole, M.J., Leader-Williams, N., 2002. Tourism and flagship species in conservation.

Biodiversity & Conservation 11, 543-547

Williams, C.K., Ericsson, G., Heberlein, T.A., 2002. A Quantitative Summary of Attitudes

toward Wolves and Their Reintroduction (1972-2000). Wildlife Society Bulletin (1973-

2006) 30, 575-584

Williams, S., Vivero Pol, J.L., Spawls, S., Shimelis, A., Kelbessa, E., 2004. Ethiopian

Highlands. In: Hotspots revisited Mittermeier, R.A., Gill, P.R., Hoffmann, M., Pilgrim, J.,

Brooks, T., Mittermeier, C.G., Lamoreux, J. and Da Fonseca, G.A.B. (eds.) CEMEX

Publisher, Washington.

Wilson, M.A., Heberlein, T.A., 1996. The wolf, the tourist, and the recreational context: New

opportunity or uncommon circumstance? Human Dimensions of Wildlife 1, 38-53

Wolf, C., Ripple, W.J., 2016. Prey depletion as a threat to the world's;s large carnivores. Royal

Society Open Science 3

36

Appendix A

Table A.1.

Description and summary of individual specific characteristics used in the model (N=168).

Variable Description Proportion

AGE40 Respondents older than 40 years (=1) 0.3452

VISITED_SMNP

Respondents visited Siemen Mountains National Park

(SMNP) during the current trip park (=1).

SMNP is a national park and UNESCO Natural Heritage

Site hosting the Ethiopian wolf

0.1547

PURP_TREKK Respondents included trekking as one of their purpose of

visiting BMNP (=1)

0.2142

VISIT_BMNP Respondents who intend for future visit to BMNP (=1) 0.4702

HABITAT_BEST Respondents who believe that the wolf habitat in BMNP is

very good or excellent (=1)

Opinion on the current status of Ethiopian wolf habitat,

based on 6 point Likert scale (1=very poor to 6=

excellent)

0.3155

MEMBER If respondents are members of environmental or

conservation organization (=1)

0.1964

HIGH_ATTRACT Respondents considered the park scenery as very good or

excellent (=1)

Opinion on the attractiveness of the park scenery, based on

6 point Likert scale (1=very poor to 6= excellent)

0.5893

VIEWUNQWL Respondents who viewed unique species other than the

Ethiopian wolf during their BMNP visit (=1)

0.8809

PREV_VISIT Respondents who have previous visit experience to

Ethiopian protected area(s) (=1)

0.4167

37

Table A.2.

Conditional logit results: coefficient and standard error values, reduced model N=168

Variable Coefficient Std. error

ASC -2.4525*** 0.3576

AREA253 0.0936 0.1962

AREA582 0.4597* 0.2029

AREA1079 0.7302*** 0.1806

ETW250 1.0472** 0.3482

ETW300 1.0358*** 0.2954

ETW400 1.0676** 0.3246

UNRESTRICT 0.3269*** 0.0730

SIGNS &TRAILS -0.8474*** 0.1546

SIGNS -0.8086*** 0.1440

TRAILS -0.2876* 0.1303

FEE -0.0036*** 0.0005

Interactions

AREA253*HIGH _ATTRACT -0.3631† 0.2106

AREA582*HIGH _ATTRACT 0.0553 0.2210

AREA1079*HIGH _ATTRACT 0.2131 0.2029

ETW250*VIEWUNQWL -0.6662† 0.3487

ETW300*VIEWUNQWL -0.9756** 0.2982

ETW400*VIEWUNQWL -0.9292** 0.32044

ETW250*PREV_VISIT -0.4929* 0.2259

ETW300*PREV_VISIT -0.4927* 0.2141

ETW400*PREV_VISIT -0.7179** 0.2249

ASC*AGE40 0.3503 0.2461

ASC*VISITED_SMNP -1.1209** 0.4297

ASC*PURP_TREKK 1.1685*** 0.2494

ASC*VISIT_BMNP -0.2405 0.2412

ASC*HABITAT_BEST -0.7402* 0.2994

ASC*MEMBER -1.0118* 0.4141

Model statistics

Log likelihood -965.29637

Pseudo R2 0.3076

Number of Observations 3807

d Asterisks denote significance level: *** p<.001, ** p<.01; * p<.05; † p<.1

38

Table A.3

RPL model results for the full model (include protest responses, N=236)

Mean Standard deviations

Variable Coefficients Std. Errors

ASC -0.5695 0.6208

AREA253 -0.3187 0.3557 1.095*** 0.2985

AREA582 0.3387 0.6491 3.814*** 0.4153

AREA1079 -0.6286 0.6971 8.046*** 0.8954

ETW250 1.5675* 0.7050 2.106*** 0.3659

ETW300 1.8118** 0.6327 2.106*** 0.3055

ETW400 2.7109*** 0.6651 2.169*** 0.3755

UNRESTRICT 1.0485*** 0.2621 2.348*** 0.3054

SIGNS &TRAILS -1.0609** 0.3374 0.532 0.4306

SIGNS -1.7062*** 0.3866 2.903*** 0.3432

TRAILS -0.7702* 0.3882 3.344*** 0.3713

FEE -0.0079*** 0.0011

Interactions

AREA253*HIGH _ATTRACT 0.1354 0.3735

AREA582*HIGH _ATTRACT 1.6685* 0.7009

AREA1079*HIGH _ATTRACT 3.4654*** 0.8427

ETW250*VIEWUNQWL 0.1128 0.7019

ETW300*VIEWUNQWL -1.2347* 0.6135

ETW400*VIEWUNQWL -2.3348*** 0.6999

ETW250*PREV_VISIT -.9967* 0.4937

ETW300*PREV_VISIT -1.1237* 0.4938

ETW400*PREV_VISIT -1.6518** 0.5366

ASC*AGE40 0.7559* 0.3608

ASC*VISITED_SMNP -2.3914*** 0.5675

ASC*PURP_TREKK 1.3917*** 0.4128

ASC*VISIT_BMNP -0.6659† 0.3464

ASC*HABITAT_BEST -0.3294 0.4275

ASC*MEMBER -1.0958* 0.4463

Model statistics

AIC 2873.8045

BIC 3119.90

Log-Likelihood -1399.902

No of observations 5646

No. of respondents 236 e Asterisks denote significance level: *** p<.001, ** p<.01; * p<.05; † p<.1

39

Appendix B

Fig. B.1.An example of choice set used for the choice experiment

40

Appendix C

Information Pamphlet included in the tourists survey questionnaire

Tourists’ preferences for protection the Ethiopian Wolf in Bale Mountains National Park

This part of the survey is designed to gather information on visitors/tourists preference about possible

management options for protecting the Ethiopian wolf in BMNP.

This separate pamphlet has two sections (A & B). Section A describes background information about

description of Bale Mountains National Park (BMNP), afro-alpine habitat, biodiversity and current

situation followed by possible management scenarios is presented.

In section B EIGHT choice set questions were presented. In each choice set, you are asked to choose

between the current situation (the status quo) and two new possible management options to protect

Ethiopian wolf in BMNP followed by some follow-up questions.

SECTION A: Background information

Bale Mountains National Park

Bale Mountains National Park is located 400 km southeast of Addis Ababa in the Oromia Regional State,

Ethiopia. BMNP was established in 1971 to conserve the two unique wildlife species, Ethiopian wolf

(Canis simensis) and the mountain nyala (Tragelaphus buxtoni) and the afro-alpine habitat. The park

covers an area of 2150 km, altitude ranging from 1500 to 4377 meters above sea level stratified by diverse

vegetation types (Fig C.1).

Internationally, the park is one of the tourist destinations for wildlife tourism and mountains trekking in

Ethiopia.

Fig C.1. Bale Mountains National Park (Bale Guide, 2013)

Management information The park management plan (2007 -2017) has been setup in 2007

Key biodiversity in the BMNP include:

78 mammal species, of which 22 are

endemic

278 bird species, of which 6 are

endemic to Ethiopia.

Endemic giant mole rat ((Tachyoryctes

macrocephalus), the major prey to

Ethiopian wolf

19 threatened and 5 endangered and

critically endangered species

The largest population of the endemic

and endangered mountain nyala and

Ethiopian wolf

More than 340 medicinal plants.

(Willams et al.., 2011; Hillman, 1986)

41

The park is administered by Ethiopian Wildlife Conservation Authority (EWCA)

Management plan includes regular monitoring of Ethiopian wolf

Re-demarcated and formally gazetted in 2015

Major problems to the park and its biodiversity

Habitat degradation and loss due to human activities (agricultural expansion, grazing and

settlement) including the afro-alpine habitat

Frequent disease outbreak seriously affecting Ethiopian wolf survival

Possible new management actions

Strict protection of the Afro-alpine habitat to maintain healthy Ethiopian wolf

Install interpretative facilities and establish well developed trails to improve wildlife tourism

Increase in park entrance fees to cover cost of Ethiopian wolf protection and ecotourism facilities

Features/characteristics of the BMNP

(1) Afro-alpine habitat

The BMNP encompasses the largest afroalpine habitat on altitudes above 3000 m in Africa (Fig.C.2, 3).

Afro-alpine refers to ‘the highest mountains of Africa with areas of shrublands and grasslands’. The

BMNP afroalpine habitat forms the largest area (1079 km2) of a suitable habitat for the endemic Ethiopian

wolf. This habitat is categorised into three: optimal (253 km2), good (329 km2) and marginal areas (497

km2) based on vegetation quality and prey rodent population presence to Ethiopian wolf (Fig 2B).

Fig. C.3. The Sanetti plateau Afro-alpine habitat at BMNP

However, an increasing number of livestock population in this habitat caused overgrazing in some areas

leading to habitat fragmentation and habitat loss including core Ethiopian wolf habitat (Fig. C.6).

Establishment of a strict core conservation zone would protect the afroalpine habitat and its Ethiopian wolf

population.

We propose new management options of strict habitat protection, which preclude any human activities

including grazing. The options are protection of either: (1) the optimal habitat; (2) the optimal and good

habitat; or (3) the entire range of suitable Ethiopian wolf habitat to maintain healthy habitat and increase

the wolf population.

(2)The endemic and endangered Ethiopian wolf

Ethiopian wolf is one of the world’s rarest endangered species of dog family found only in Ethiopia. It has

a global population of 400-450 individuals localised in six highland areas with the largest population (~

250 ind.) existing in the BMNP (Fig. C.4). However, habitat degradation and infectious disease caused a

dramatic decline in the population posing a greater risk of extinction to this unique species. For example,

Fig. C.2. The afro-alpine habitat of Ethiopian wolf

in the BMNP (Sillero-Zubiri & Macdonald, 1997)

42

the assessment of rabies disease in recent year, since 2014, the Ethiopian wolf population in the BMNP has

crushed to only 196 (131 adults and 65 pups) individuals due to rabies disease transmitted from domestic

dogs (Bedin et al., 2015). In this regard, Proper wildlife management could improve the wolf population

and ensure for long term biodiversity conservation.

Fig. C.4. Distribution of the populations of Ethiopian wolf in the Ethiopian highlands (IUCN/SSC Canid

Specialist Group 2011).

Fig.C.5. Ethiopian Wolf maintained and suitable habitat Fig.C.6. Ethiopian wolf, overgrazed habitat

(3) Visitors’ access to the Ethiopian wolf habitat

Currently, park visitors can freely roam everywhere in the afroalpine habitat to view Ethiopian wolf once

they pay gate fee. In this regard, access restriction, for example, walking to only designate trails within

core conservation areas can reduce the number of wildlife encounters and excessive disturbance in core

areas of the Ethiopian wolf habitat.

(4) Recreation facilities: Interpretive signs and walking trails Interpretive signs and walking trails are non-existent in the Ethiopian wolf habitat. Tourists and visitors,

who would view Ethiopian wolf, rely on guide either from the park or tour operators for interpretational

services. In this regard, interpretive signs can provide useful information about the wildlife including

warnings getting close to them. Therefore, installing signs along with improved trails can improve visitors’

ecotourism experience.

43

(5) Park entry fee (cost of visiting the park)

The park’s revenue is mainly generated from the gate entry fees, which is not sufficient to cover park

management expenses. Thus, a reasonable increment in the entry fee will help to get fund that would go

towards improving the ecotourism experience and ensure visitors’ satisfaction. The extra entry fee will be

placed in a special fund to be used only for improvement of the park services and protection of the

Ethiopian wolf.

Table C.1. Summary of the attributes and their levels and coding; bold italicised levels represent status quo or current

situation.

Characteristics Attribute levels and current situation

Afro-alpine Area (km2)

1= de facto open access 2= 253 km2 (11.5%)

3= 582 km2 (26.5%)

4= 1079 km2 (49%)

Ethiopian Wolf population 1= <200 individuals

2= 250

3= 300

4= 400

Access 1= Unrestricted access

2= Access restricted to only designated trails

Recreation facilities: Interpretive signs and trails 1= No trails and no interpretative signs

2= Interpretive sign

3= Well maintained trails

4= Interpretive sign and trails

Cost in ETB 1= 90,

2= 150

3= 250

4= 350

Section B: Choice Sets

In the following questions, we ask you to make choices among alternative management options to improve

wildlife tourism and protection of Ethiopian wolves in the BMNP. The possible management actions are

described in Section A. Please, consider the following issues carefully before you go to the choice sets.

Management options

Three management scenarios are given for conservation of the Ethiopian wolf in each choice set.

Option 1 (the status quo) is the same in each choice set. Status quo refers to continuation of the

current situation and human activities in the Ethiopian wolf habitat causing habitat fragmentation

and habitat loss. Moreover, disease transmitted from domestic dogs will cause a significant

reduction in number of wolf that will undermine Ethiopian wolf viewing in the BMNP area in the

future.

Options 2 and 3 involve new management scenarios that are likely to improve the Ethiopian wolf

protection in their habitat. These management scenarios involve strict protection of the core wolf

habitat with restrictions on human activities such as grazing, agriculture, and settlements.

Making a choice

We ask you to choose your most preferred option on each choice set. When deciding please consider:

The future outcomes (changes) presented under the column for each option.

Consider your income and other expenses before you make your choices.

Assume that the management options given are the only available options.

Each choice set is independent other choices and need to be answered independently