Resumen: El reconocimiento de la necesidad de incluir criterios económicos en el proceso de toma de decisiones sobre políticas de conservación ha impulsado el uso de técnicas de valoración económica. Sin embargo, aún se debate si es posible asignar valores económicos precisos a la biodiversidad, así como lo que esos valores realmente representan. Revisamos 60 artículos recientes sobre valoración económica de la biodiversidad y realizamos un meta análisis de estos estudios para determinar los factores que afectan la disposición a pagar por la conservación de la biodiversidad. Analizamos las variables internas del método de valoración contingente (medida de los beneficios, forma de pago, formato de respuesta o frecuencia de pago) y de factores antropomórficos, antropocéntricos y científicos. La asignación de recursos favoreció principalmente a la conservación de especies con características antropomórficas y antropocéntricas en vez de considerar factores científicos. Recomendamos que los investigadores y políticos contemplen cuidadosamente las valoraciones económicas de la biodiversidad, considerando los sesgos inherentes del método de valoración contingente y los factores antropomórficos y antropocéntricos resultantes de las actitudes del público hacia las especies. Debido a la creciente tendencia por incluir consideraciones económicas en las prácticas de conservación, sugerimos que equipos interdisciplinarios de ecólogos, economistas y científicos sociales en el futuro colaboren y dirijan análisis comparativos, tal como hemos hecho aquí. El uso del método de valoración contingente en las políticas de conservación de la biodiversidad puede proporcionar información útil sobre estrategias de conservación alternativas si los cuestionarios son cuidadosamente elaboradas, los encuestados estén suficientemente informados y se identifican los factores que influyen sobre la disposición a pagar.
Abstract: Recognition of the need to include economic criteria in the conservation policy decision-making process has encouraged the use of economic-valuation techniques. Nevertheless, whether it is possible to accurately assign economic values to biodiversity and if so what these values really represent is being debated. We reviewed 60 recent papers on economic valuation of biodiversity and carried out a meta-analysis of these studies to determine what factors affect willingness to pay for biodiversity conservation. We analyzed the internal variables of the contingent-valuation method (measure of benefits, vehicle of payment, elicitation format, or timing of payment) and anthropomorphic, anthropocentric and scientific factors. Funding allocation mostly favored the conservation of species with anthropomorphic and anthropocentric characteristics instead of considering scientific factors. We recommend researchers and policy makers contemplate economic valuations of biodiversity carefully, considering the inherent biases of the contingent-valuation method and the anthropomorphic and anthropocentric factors resulting from the public's attitude toward species. Because of the increasing trend of including economic considerations in conservation practices, we suggest that in the future interdisciplinary teams of ecologists, economists, and social scientists collaborate and conduct comparative analyses, such as we have done here. Use of the contingent-valuation method in biodiversity conservation policies can provide useful information about alternative conservation strategies if questionnaires are carefully constructed, respondents are sufficiently informed, and the underlying factors that influence willingness to pay are identified.
Social sciences need to be incorporated into conservation science and practice because biodiversity conservation is as much about people as it is about other species (Mascia et al. 2003). For instance, environmental economics can inform conservation biologists and policy makers about why species are endangered, the opportunity costs of protection activities, and the economic incentives for conservation (Shogren et al. 1999). Scientists argue that economic criteria need to be a part of the design and implementation of conservation policies (MEA 2005). Similarly, many institutional programs, such as the Convention on Biological Diversity (CBD) or the Natural Resource Management program (OECD 2002), recognize the importance of understanding the economic value of biodiversity for conservation policy making. Economic-valuation techniques have recently moved from scientific forums to management practices in the design of systems that pay landowners for ecosystem services (payments for environmental services [PES]). For example, the World Bank is developing PES schemes based on economic-valuation techniques in several countries of Central and South America and Africa.
Among the economic-valuation techniques, the contingent-valuation method has been used widely to measure the economic value of species. The procedure is based on a hypothetical market in which people are asked through questionnaires to express their maximum willingness to pay (WTP) for the protection of biodiversity (Loomis & White 1996). Although contingent valuation has been commonly used in policy-related research, there are numerous critiques in the literature that concern their content and questionnaire design and the validity and reliability of their results (Mitchell & Carson 1989; Venkatachalam 2004). It is necessary to identify and measure the inherent biases of contingent-valuation studies in order to reduce internal inconsistency (White et al. 2001).
Debate continues on the factors, from anthropomorphic to scientific, that affect WTP for biodiversity conservation (e.g., Tisdell et al. 2007). Other factors, such as species' usefulness to humans, may also play important roles in determining WTP for biodiversity conservation.
Our objectives here were to report and analyze the underlying factors that explain the economic values of species; obtain evidence about the relationship between human attitudes to animals and the WTP for conservation; generate useful criteria to incorporate economic values in conservation policies, and explore the crucial need for interdisciplinary research teams.
To address the controversies surrounding the economic valuation of biodiversity, we performed an extensive meta-analysis of contingent-valuation studies. We used the following criteria to select studies for inclusion in the analysis. First, studies had to be published in peer-reviewed journals to avoid unknown and inaccessible studies. Nevertheless, we included gray literature if it was included in some of the studies published in the peer-reviewed literature (Hageman 1985; King et al. 1988; Duffield, unpublished , 1992; Duffield et al. 1993; Tanguay et al. 1993, 1995; Brown et al. 1994). Second, because some studies reported multiple estimates of WTP, we used the best estimation if it was identified by the authors; otherwise, we averaged their WTP, unless the variation in estimated values was related to our explanatory variables. All WTP estimates were converted to 2005 U.S. dollar values (Consumer Price Index, http://www.bls.gov/cpi/home.htm). Third, the studies included in the analysis determine WTP for single species, instead of biodiversity in general terms.
To examine financial support for biodiversity conservation, we considered variables that economic-valuation theory suggests are important and variables that explain human attitudes toward biodiversity (Table 1).
|benefit measure||nominal||WTP to secure a gain||Kruskal–Wallis|
|WTP to avoid loss|
|WTP to biodiversity plan|
|vehicle payment||dummy||1, WTP by coercive payments||Mann–Whitney|
|0, WTP by voluntary payments|
|dummy||1, continuous; 0, discrete choice||Mann–Whitney|
|timing of payment||dummy||1, lump sum; 0, annual payment||Mann–Whitney|
|length||continuous||mean length of a species (cm)||Spearman correlation|
|weight||continuous||mean weight of a species (kg)|
|eye sizea||continuous||axial length of eye (mm)|
|dummy||1, mammal; 0, otherwise||multivariate regression|
|dummy||1, bird; 0, otherwise|
|dummy||1, reptile; 0, otherwise|
|dummy||1, fish; 0, otherwise|
|dummy||1, marine, anadromous, or marine-coastal species; 0, otherwise||multivariate regression|
|usefulness||dummy||1, useful; 0, otherwise||Mann–Whitney and multivariate regression|
|economically negative||dummy||1, economic negative; 0, otherwise||Mann–Whitney and multivariate regression|
|population sampled||dummy||1, resident; 0, visitor||Mann–Whitney|
|change in species population size||continuous||change in species population size proposed in questionnaire (%)||simple regression|
|IUCN statusb||ordinal||5, CR; 4, EN; 3, VU; 2, NT; 1, LC; 0, nonendangered||Kruskal–Wallis and multivariate regression|
|ecological role||dummy||1, functional role; 0, nonfunctional role in ecosystem||Mann–Whitney and multivariate regression|
|geographical range||dummy||1, endemic species in the region of||Mann-Whitney|
|economic study; 0, otherwise|
Contingent-valuation surveys are context-dependent, that is, the values estimated are subject to various aspects of the questionnaire design. Although some elements of the survey are expected to be neutral (e.g., questions about family size should not influence an individual's response to the WTP question), others have a significant influence on a respondents' valuation (Bateman et al. 2002), such as the measure of benefits, vehicle of payment, how information was gathered (elicitation method), and timing of payment (Table 1).
We classified the types of benefit measures as WTP to avoid a loss, WTP to secure a gain, and WTP to invest in a biodiversity plan. Usually the estimates from contingent-valuation studies are derived from either WTP to avoid loss of a species (e.g., amount one is willing to pay to prevent a species from going extinct) or WTP for a proposed gain in numbers (e.g., amount one is willing to pay to improve the chance of survival of a species by 50–99%). Another common measure of WTP is the amount a person is willing to pay to a conservation plan for a species.
We classified payment vehicles as voluntary or coercive. Coercive payments included taxes, fees, or charges. Voluntary payments were donations or gifts (Bateman et al. 2002).
Stated-preference (elicitation) methods differed in how much information they conveyed to and collected from respondents (Bateman et al. 2002). The open-ended stated-preference format asked respondents what maximum amount they would be willing to pay for species conservation. The payment-card formats contained a range of values from which the individuals chose their maximum WTP. Dichotomous-choice elicitation methods required respondents to answer yes or no when asked if they were willing to pay a given amount for species conservation. In previous approaches, researchers assumed respondents had no uncertainty regarding their preferences. Nevertheless, allowing only yes-or-no answers can result in respondents who are uncertain answering yes in support of biodiversity conservation programs (Brown et al. 1996). To allow for respondents' uncertainty, Welsh and Bishop (1993) developed the multiple-bounded approach, which contains elements of both the payment card and dichotomous-choice formats. As for the payment-card format, respondents were presented with an ordered sequence of WTP amounts, but rather than circling a single value, the respondents were given a multiple-choice response option, including “definitely yes,”“probably yes,”“unsure,”“probably no,” and “definitely no” to each amount presented. Accordingly, we classified the elicitation questions into open-ended, payment-card, dichotomous-choice, and multiple-choice formats.
Comparisons of different contingent-valuation studies indicate there were systematic differences between values elicited with continuous (open-ended and payment-card) and discrete choice (dichotomous and multiple-choice) formats (Brown et al. 1996). In general, values collected with discrete-choice formats exceed values collected with open-ended (Reaves et al. 1999) or payment-card formats (Welsh & Poe 1998). For that reason we also classified the elicitation formats into continuous and discrete-choice formats.
For timing of payment some researchers asked respondents to express their WTP as an annual amount, whereas others used a single lump sum payment. The expected effect was that a one-time payment would be larger than an annual payment stretched over time into the immediate future.
Among variables that can explain social preferences, we studied 3 factors: (1) anthropomorphic, associated with the likeability and similarity of species to humans, (2) anthropocentric, related to the usefulness of species, and (3) scientific, which determine whether scientific knowledge influences estimates of WTP (Table 1).
The scientific literature indicates that conservation support is positively related to the perceived attractiveness of nonhuman species, which usually is an extension of human similarity (i.e., the similarity principle) (Plous 1993; Gunnthorsdottir 2001). Perceived similarity between humans and nonhuman species is related to the phylogenetic level (Eddy et al. 1993) and to physical characteristics such as length, weight, and eye size (Herzog & Burghardt 1988). To study the effect of phylogenetic level, we classified species into mammals, birds, reptiles, fishes, and invertebrates. These categories sometimes cause some species to be perceived as charismatic fauna. Moreover, usually the charismatic species are related to different ecosystems (e.g., Asian elephant [Elephas maximus] with Indian forests, giant panda [Ailuropoda melanoleuca] with Chinese bamboo forests, and giraffes [Giraffa camelopardalis] with African savannahs). For that reason we studied the effect of the ecosystem on WTP. We classified the ecosystems according to the Millennium Ecosystem Assessment (MEA 2003). Because many fish species are anadromous, we also included marine-inland water as an ecosystem type (Table 1).
In addition, individuals can favor conservation of those species with anthropocentric characteristics (DeKay & McClelland 1996; Martín-López et al. 2007). In general, species useful to humans are positively related to WTP, and those that produce economic damages are negatively related to WTP estimates. To study in detail the social role of species, we categorized them as species that (1) generate crop damages, (2) generate economic loss by predation on cattle, (3) are hunted or fished for recreation, and (4) are a nonconsumptive tourism resource.
Nevertheless, many wild species are perceived as having opposing attributes; sometimes they are considered pests and other times they are considered valuable assets. For instance, elephants are widely considered a pest by local people in rural regions (Bandara & Tisdell 2003); on the other hand, elephants attract tourism in protected areas (Wilkie & Carpenter 1999). The dual character of the elephant as both agricultural pest and valuable economic asset reflects the difficulty in classifying it as a useful or economically negative species. Similarly, humans have positive attitudes toward socially controversial animals in the context of abstract existence values, but these attitudes quickly become negative when the presence of the species is associated with economic costs in their immediate surroundings (Kaltenborn et al. 2006). This means visitors have a higher WTP than local people (Loomis & Larson 1994; Loomis & White 1996). For that reason we studied the WTP estimates for visitor and for resident respondents.
Finally, any preservation decision is likely to consider scientific knowledge about the species. We studied the relationship between WTP and degree of endangerment, in terms of 2 variables: change in population size of a species proposed in the contingent-valuation questionnaire and the endangerment categories of the IUCN (World Conservation Union) Red List. To study the effect of IUCN categories, we classified the species in the studies we examined as critically endangered (CR), endangered (EN), vulnerable (VU), near threatened (NR), least concern (LC) (IUCN 2006), or not endangered. Other scientific variables that can influence WTP are knowledge of the ecological role of species and whether the species is endemic.
To test the effect of these variables on WTP values for species conservation, we used different statistical analyses. We used descriptive statistics to summarize the distribution of contingent-valuation studies by allocation of study, taxonomic group, and ecosystem. To test the individual effect of nominal and dummy variables on WTP, we used nonparametric statistics (Mann–Whitney and Kruskal–Wallis). When the Kruskal–Wallis test achieved 90% significance, we used Dunn's multiple comparison posttest to compare WTP estimates of one group with another. We also used correlation and simple regression analyses to test the effect of continuous variables on WTP. To obtain evidence about the relationship between human attitudes toward animals and WTP, we examined the joint effect of anthropomorphic, anthropocentric, and scientific variables through multivariate regression analysis. Here, we used only those variables for which data were available for all the species we examined. To improve the model we expressed the frequency of payment as a dichotomous variable (lump sum). The dependent variable was the natural log of WTP.
Contingent Valuation of Biodiversity
The number of selected studies was 60 (Table 2). The distribution of valuations by place indicated that 65% of the economic-valuation studies were localized in the United States, 15% in Europe, 8% in Australia, 6% in Canada, and 6% in Sri Lanka. The dominance of the United States is due in part to the history of economic valuation there (Gen 2004).
|Taxa||Common name||Scientific name||Mean value (US$2005)a||Reference|
|Rodentia||Eurasian red squirrel||Sciurus vulgaris||2.87||White et al. 2001|
|water vole||Arvicola terrestris||15.24||White et al. 1997|
|Artiodactyla||bighorn sheep||Ovis canadensis||21.94||King et al. 1988Brookshire et al. 1983Bulte & Kooten 1999|
|elk (red deer)||Cervus elaphus||206.93||Hammack & Brown 1974|
|moose||Alces alces||145.49||Horne & Petäjistö 2003|
|woodland caribou||Rangifer tarandus||44.74||Tanguay et al. 1993, 1995Adamowicz et al. 1998|
|Carnivora||coyote||Canis latrans||5.49||Stevens et al. 1991, 1994|
|California sea otter||Enhydra lutris nereis||36.76||Hageman 1985, 1986|
|European otter||Lutra lutra||24.40||White et al. 1997|
|giant panda||Ailuropoda melanoleuca||13.81||Kontoleon & Swanson 2003Hsee & Rottenstreich 2004|
|gray wolf||C. lupus||19.26||Duffield, unpublished data Duffield 1992Duffield et al. 1993USFWS 1994Chambers & Whitehead 2003|
|gray seals||Halichoerus grypus||12.83||Bosetti & Pearce 2003|
|grizzly bear||Ursus arctos horribilis||38.89||Brookshire et al. 1982|
|Hawaiian monk seal||Monachus schauinslandi||93.87||Samples & Hollyer 1990Brown et al. 1994|
|Mediterranean monk seal||M. monachus||17.54||Langford et al. 1998|
|northern elephant seal||Mirounga angustirostris||31.53||Hageman 1986|
|Steller sea lion||Eumetopias jubatus||73.83||Giraud et al. 2002|
|Cetacea||beluga whale||Delphinapterus leucas||14.20||Tkac 1998|
|blue whale||Balaenoptera musculus||44.57||Hageman 1985, 1986Bulte & Kooten 1999|
|bottlenose dolphin||Tursiops truncatus||23.17||Hageman 1986|
|gray whale||Eschrichtius robustus||34.70||Hageman 1985, 1986Loomis & Larson 1994|
|humpback whale||Megaptera novaeangliae||128.34||Samples et al. 1986 Samples & Hollyer 1992 Brown et al. 1994Wilson & Tisdell 2003|
|Lagomorpha||brown hare||Lepus europaeus||0.00||White et al. 2001|
|Perissodactyla||Pentro horse||Equus caballus||33.89||Cicia et al. 2003|
|Proboscidea||Asian elephant||Elephas maximus||1.94||Bandara & Tisdell 2003Bandara 2004Bandara & Tisdell 2005|
|Diprotodontia||mahogany glider||Petaurus gracilis||29.88||Tisdell et al. 2005b|
|tree kangaroos||Dendrolagus bennettianus||53.10||Tisdell & Wilson 2004|
|Marsupial||Leadbeater's possum||Gymnobelideus leadbeateri||25.83||Jakobsson & Dragun 1996|
|Anseriformes||Harlequin Duck||Histrionicus histrionicus||11.15||Tkac et al. 1998|
|wild goose||Anser sp.||11.91||Macmillan et al. 2002|
|Galliformes||Wild Turkey||Meleagris gallopavo||11.59||Stevens et al. 1991|
|Gruiformes||Whooping Crane||Grus americana||53.42||Bowker & Stoll 1988|
|Falconiformes||Peregrine Falcon||Falco peregrinus||29.89||Kotchen & Reiling 1998|
|Bald Eagle||Haliaeetus leucocephalus||114.67||Boyle & Bishop 1987Stevens et al. 1991Swanson 1993Bulte & Kooten 1999|
|Strigiformes||Northern Spotted Owl||Strix occidentalis caurina||59.43||Rubin et al. 1991Hagen et al. 1992Loomis & González-Cabán 1998Bulte & Kooten 1999|
|Mexican Spotted Owl||S. occidentalis lucida||74.38||Loomis & Ekstrand 1997, 1998Giraud et al. 1999|
|Piciformes||Red-cockaded Woodpecker||Picoides borealis||12.10||Bulte & Kooten 1999Reaves et al. 1999|
|White-backed Woodpecker||Dendrocopos leucotos||66.39||Fredman 1995Fredman & Boman 1996|
|Reptile||loggerhead sea turtle||Caretta caretta||16.98||Whitehead 1992Wilson & Tisdell 2003|
|Salmoniformes||Atlantic salmon||Salmo salar||9.45||Stevens et al. 1991Bulte & Kooten 1999|
|arctic grayling||Thymallus arcticus arcticus||22.69||Duffield & Patterson 1992b|
|chinook salmon||Oncorhynchus tshawytscha||126.66||Hanemann et al. 1991Olsen et al. 1991|
|cutthroat trout||O. clarki||17.02||Duffield & Patterson 1992|
|steelhead||O. mykiss||64.47||Olsen et al. 1991|
|Acipensiriformes||shortnose sturgeon||Acipenser brevirostrum||30.86||Kotchen & Reiling 1998|
|Cipriniformes||Colorado squawfish||Ptychocheilus lucius||10.91||Cummings et al. 1994Bulte & Kooten 1999|
|striped shiner||Luxilus chrysocephalus||6.83||Boyle & Bishop 1987Bulte & Kooten 1999|
|Perciformes||kelp bass||Paralabrax clathratus||43.35||Carson et al. 1994|
|white croaker||Genyonemus lineatus||43.35||Carson et al. 1994|
|Crustacean||riverside fairy shrimp||Streptocephalus woottoni||24.85||Stanley 2005|
The final data set included 50 species distributed among mammals (56%), birds (20%), reptiles (2%), fishes (20%), and invertebrates (2%). Nevertheless, the studies dedicated to these groups manifested a taxonomic bias because 64% of studies focused on mammals, 23% on birds, and only 1% on reptiles and crustaceans (Table 2). Similarly, the economic-valuation research showed preference for studying species that live in marine and forest ecosystems; these were 40% and 33% of studies, respectively. In contrast, species that live in dryland and urban-cultivated ecosystems were studied in only 4% and 1% of cases, respectively.
Effect of Study Design on the Economic Value of Species
WTP differed significantly with the choice of benefit measured in the hypothetical market, in which WTP to secure a gain generated the highest values and WTP to biodiversity plan resulted in the lowest values. Willingness to pay was significantly higher for coercive payments than voluntary payments. Similarly, discrete-choice elicitation formats generated significantly higher values than continuous formats. Willingness to pay for biodiversity conservation was significantly greater for one-time payments than for annual payments (Table 3).
|Benefit measure||WTP to secure a gain||55.06||Kruskal–Wallis, χ2= 8.30, df = 2, p= 0.02;|
|WTP to avoid loss||24.75||Dunn's multiple comparison, p < 0.01|
|WTP to biodiversity plan||10.29|
|Payment vehicle||coercive||40.16||Mann–Whitney, U= 753; z=−1.64, p= 0.10|
|Elicitation method||open-ended||34.58||Kruskal–Wallis, χ2= 5.52, df = 3, n.s.|
|continuous format||36.96||Mann–Whitney, U= 649.5; z=−2.23, p= 0.03|
|discrete choice format||48.90|
|Timing of payment||annual payment||40.72||Mann–Whitney, U= 532.5; z=−2.57, p= 0.01|
Effect of Anthropomorphic, Anthropocentric, and Scientific Factors on WTP
Nonparametric analysis demonstrated that the effect of the anthropomorphic and anthropocentric factors on WTP was higher than scientific factors (Table 4). All physical variables (length, weight, and eye size) had a positive and significant effect on WTP. Willingness to pay did not differ among taxonomic groups or ecosystems. Nevertheless, the effect of the ecosystem on WTP differed by taxonomic group (mammals and fishes). The Dunn's multiple comparisons test for WTP of mammal species distinguished 3 statistically different ecosystem groups: marine, forest or mountain, and inland waters and dry lands. In addition, WTP for fish differed significantly between continental and marine or anadromous species (Table 4).
|physical characteristics||length||–||Spearman's rho = 0.31, p= 0.06|
|weight||–||Spearman's rho = 0.33, p= 0.05|
|eye size||–||Spearman's rho = 0.43, p= 0.01|
|phylogeny||mammals||43.39||Kruskal–Wallis, χ2= 0.74, df = 4, n.s.|
|ecosystems||forest||56.37||Kruskal–Wallis, χ2= 8.64, df = 7, n.s.|
|mammals||marine coastal||46.53||Kruskal–Wallis, χ2= 10.96, df = 5, p= 0.05;|
|marine||61.77||Dunn's multiple comparison, p < 0.01|
|birds||forest||60.32||Kruskal–Wallis, χ2= 3.10, df = 2, n.s.|
|fish||marine–inland water||68.85||Kruskal–Wallis, χ2= 6.10, df = 2, p= 0.05;|
|marine||39.19||Dunn's multiple comparison, p < 0.02|
|usefulness||44.33||Mann–Whitney, U= 246; z=−0.13, n.s.|
|economically negative||13.95||Mann–Whitney, U= 55; z=−2.29, p= 0.02|
|resident||43.36||Mann–Whitney, U= 1629.5; z=−0.53, n.s.|
|IUCN||CR||17.54||Kruskal–Wallis, χ2= 0.89, df = 5, n.s.|
|ecological role||yes||36.18||Mann–Whitney, U= 255.5; z=−0.88, n.s.|
|geographical range||endemism||40.36||Mann–Whitney, U= 276; z=−0.48, n.s.|
Although whether a species is useful did not have an effect on WTP, WTP differed significantly between those species that have a negative economic impact and those that do not (Table 4). When we compared the mean WTP for species conservation among the categories of (1) species that generate crop damage ($6.41), (2) species that cause damage to cattle ($21.21), (3) species that are a fishing or hunting resource ($54.60), and (4) species that are a nonconsumptive tourist resource ($44.57), we found significant differences (Kruskal–Wallis, χ2= 5.42, df = 3, p= 0.10). Dunn's test also showed significant differences (p < 0.01) between species that cause economic loss and those that are exploited for recreation.
Although for the global data set the mean WTP was higher for visitors than for residents, there were no significant effects. Nevertheless, visitor and resident WTP for carnivore conservation differed significantly (Mann–Whitney, U= 36; z=−2.68, p= 0.007). Carnivores were usually valued higher by visitors ($147.80) than by residents ($62.21).
The level of endangerment had an effect on the public's decision to invest in biodiversity conservation when it was stated in the questionnaire in terms of change in a species' population size (Fig. 1). There was an exponential relation (y= 10.88e0.0108x, R2= 84.7%, n= 55) between the change in population size and WTP. Nevertheless, when we analyzed the effect of the degree of endangerment on the basis of IUCN categories, there were no significant differences. Ecological role and endemism variables did not significantly influence WTP.
Allocation of Funds for Biodiversity Conservation
Because of the strong correlations between length of an animal and eye size (Spearman's rho = 0.794, p < 0.0001) and between weight and eye size (Spearman's rho = 0.914, p < 0.0001), we used only the eye-size variable as an indicator of the anthropomorphic effect because it was the most significant (Table 4).
Forty-six percent and 40% of the variation in WTP was explained by the explanatory variables in the full and reduced model, respectively (Table 5). The effect of eye size was positive and significant at the 10% level. The taxonomic-class dummies had the expected effect on WTP: higher forms of life were assigned higher economic values by the respondents. Nevertheless, mammals and birds did not have a significant effect, whereas reptiles had a significant negative effect in both expanded and reduced models. Fishes presented only a significant negative effect in the reduced model. In addition, WTP for marine species was statistically greater than for species that live in continental ecosystems.
|Variable||Full model||Reduced model|
|coefficient||t ratio||p||coefficient||t ratio||p|
|Ln (eye size)||0.554||1.895||0.065||0.506||1.750||0.087|
|World Conservation Union||−0.115||−1.036||0.306|
|Akaike's information criteria||11.58||5.35|
In the case of anthropocentric criteria, although a species' usefulness did not have an effect on WTP, the economic impact of species was significantly negatively related to WTP. Scientific criteria such as the degree of endangerment measured by the IUCN categories and a species' ecological role did not have a significant effect on WTP. Finally, one-time payments were considered more favorably than annual payments.
Because there are budget constraints on biodiversity conservation, the contingent-valuation technique is becoming increasingly important as a supplement to biological information in helping to define objectives and priorities in conservation biology (White et al. 2001). Nevertheless, the economic valuation of biodiversity is affected not only by the inherent variables of contingent valuation (measurement of benefits, payment vehicle, elicitation format, or timing of payment) but also by the public's attitudes toward biodiversity. Therefore, for effective conservation management, apart from knowledge of the economic value that people assign to biodiversity conservation, it is also important to determine the underlying factors influencing WTP for biodiversity conservation.
Willingness to pay for species conservation is strongly determined by human attitudes toward these species. People's attitudes toward animals are generally based on 2 distinct motivational considerations: affect, representing people's affective responses to animals, and utility, representing people's perceptions of animals' instrumental value (Serpell 2004). People's affective responses toward species are influenced by anthropomorphic (Kellert & Berry 1980; Eddy et al. 1993; Plous 1993) and anthropocentric variables (Serpell 1986; Herzog & Burghardt 1988). On one hand, species that are phylogenetically close and physically similar to humans are likely to attract more conservation support than dissimilar species (Gunnthorsdottir 2001; White et al. 2001; Martín-López et al. 2007). On the other hand, species perceived as useful or beneficial to humans are regarded more positively than those perceived as useless or detrimental (DeKay & McClelland 1996; Martín-López et al. 2007).
In contrast to previous studies (e.g., DeKay & McClelland 1996), we found phylogeny did not explain well WTP for biodiversity conservation. This might be partly due to the combination of species included in each taxonomic class (Tisdell et al. 2005a). For example, support for turtle species may be almost as strong as for some birds and mammals, although on the whole there is stronger support for the latter (Tisdell et al. 2006). Similarly, Stanley (2005) found considerable support for the conservation of the Riverside fairy shrimp (Streptocephalus wootoni), even though it is an invertebrate (Tisdell et al. 2005a). Nonetheless, because the economic valuation of Riverside fairy shrimp constitutes the only study of a crustacean, Stanley's results may not be representative of WTP for the conservation of other invertebrates. On the other hand, as the similarity principle suggests, our reduced regression analysis showed that fish have a significant negative effect on WTP. Nevertheless, the effect of fish on WTP is ambiguous because salmoniform species have an important cultural and recreational value in the Pacific Northwest (U.S.A.) (Loomis & White 1996). The higher amounts of WTP for anadromous fishes may apply to all recreational fisheries. Thus, species that are hunting or fishing resources were the ones most valued by the public. In addition, among anthropocentric variables, a key variable determining WTP for biodiversity conservation was the economic damage caused by species, which was clearly and negatively related to WTP.
With regard to scientific criteria, our results revealed that respondents' previous knowledge of changes in species population size was the only significant variable in determining economic value. The fact that information regarding endangerment status influences allocation of funding for biodiversity conservation is supported by the results of other experimental studies (e.g., Fredman 1995; Tkac 1998; Bandara & Tisdell 2005; Tisdell & Wilson 2006). Accordingly, endangered species are liable to be greatly disadvantaged in competing for conservation funds when the public is poorly informed about them (Tisdell & Wilson 2006). Accurate information on conservation status of species can be important for improving social decisions regarding biodiversity conservation. One of the most obvious examples of the importance of public information in biodiversity conservation is the environmental campaigns developed to prevent commercial whaling and sealing. Through these campaigns, whales and seals acquired an iconic value for the conservation movement in the 1970s (Corkeron 2004).
To provide useful and reliable information to policy makers about biodiversity conservation, it seems appropriate to pay great attention to the underlying anthropomorphic and anthropocentric factors of species, particularly in those cases when the contingent-valuation surveys do not provide knowledge of the scientific issues concerning species. Understanding human attitudes toward biodiversity is essential to the work of correcting the inherent bias associated with species valuation. To understand the underlying motives behind WTP for biodiversity conservation, contingent-valuation studies should be improved through the incorporation of other scientific disciplines, such as environmental psychology or human ecology. Therefore, conservation decision-making processes call for interdisciplinary knowledge in which conservation biologists and economists collaborate with anthropologists and psychologists (Mascia et al. 2003; Saunders et al. 2006). Implementing contingent valuation for biodiversity is a difficult task because the public has a low level of understanding of what biodiversity is and why it matters (Christie et al. 2006). Providing accurate information about endangerment level, population status, geographical range, and the ecological role of species can increase the reliability of the contingent-valuation method.
Consequently, use of the contingent-valuation method in biodiversity conservation policies can provide useful information about alternative conservation strategies if questionnaires are carefully constructed, respondents are sufficiently informed, and the underlying factors that influence willingness to pay are identified.
We thank 3 anonymous reviewers for helpful suggestions on an early version of the paper and E. Main for careful editing of the manuscript. Funding was provided by the Department of Environment of the Andalusian Regional Government (Project NET413308/1) and by the Spanish Ministry of the Environment (Project 13/2006).