Recreational boaters value biodiversity: The case of the California Channel Islands National Marine Sanctuary

Marine Policy 81 (2017) 91–97

Contents lists available at ScienceDirect

Marine Policy journal homepage: www.elsevier.com/locate/marpol

Recreational boaters value biodiversity: The case of the California Channel Islands National Marine Sanctuary

MARK

⁎

Daniel Vianaa, , Kiya Gornika,b, Ching-Cheng Lina, Gavin McDonalda, Nathaniel S.R. Nga,c, Christine Quigleya, Matthew Potoskia a b c

Bren School of Environmental Science and Management, University of California Santa Barbara, Santa Barbara, CA 93106-5131, USA Marine Science Institute, University of California Santa Barbara, Santa Barbara, CA 93106-5131, USA Department of Biological Sciences, National University of Singapore, Singapore

A R T I C L E I N F O

A B S T R A C T

Keywords: Random Utility Model Non-market valuation Site-choice model Marine recreation

Marine ecosystems provide a range of valuable services, some of which come with market prices to quantify value and others for which markets have not set prices. Lacking perfect information, policy makers are at risk of undercounting non-priced values and services, leading to biases in policy decisions in favor of services valued through markets. Furthermore, understanding users’ valuation of specific site attributes, such as marine biodiversity, can contribute to effective policy decisions. This paper presents a non-market valuation of private recreational boaters (PRBs) in the Channel Islands National Marine Sanctuary located in California, USA, using data from an intercept survey conducted in 2006 and 2007. A Random Utility Model is used to estimate PRBs’ daily trip values and the importance of specific site attributes. The average consumer surplus was estimated at $48.62 per trip, with a total non-market value of non-consumptive private recreational boating of $86,325 annually. PRBs show a preference for visiting locations with lower exposure to prevailing winds and greater species richness and abundance, which to the authors’ knowledge is the first time that PRBs have been found to value biological diversity in site choices. Furthermore, this suggests that improved biodiversity and productivity of marine ecosystems contribute to better recreational experiences. The results from this study reveal the importance of including non-market services and stakeholder's preferences into policy decisions.

1. Introduction Marine ecosystem services provide environmental, economic, and social value to a variety of users and activities [1] including commercial activities such as fisheries and mining; recreational activities such as snorkeling, SCUBA diving, sport fishing and wildlife viewing [2,3]; and environmental values such as coastal protection and welfare derived from healthy ecosystems [4]. Ecosystem services reflecting the value of environmental amenities are particularly important to measure because they are often directly affected by resource management practices [5–7]. For example, private recreational boaters engage in activities, such as sport fishing, snorkeling, SCUBA diving, and whale watching, that bring them into close contact with an ecosystem's environmental amenities. Policy makers require detailed information about specific ecosystems services that generate value in order to make informed and balanced resources management decisions [8]. For example, scuba divers may value the diversity of species as much or more as the total biomass of species at their diving site [2]. Understanding such

⁎

Corresponding author. E-mail address: [email protected] (D. Viana).

http://dx.doi.org/10.1016/j.marpol.2017.03.017 Received 11 April 2016; Received in revised form 7 March 2017; Accepted 13 March 2017 0308-597X/ © 2017 Elsevier Ltd. All rights reserved.

preferences can have important implications for marine resource management. Effective ecosystem management requires taking into account value across social groups, their activities and the type of value they receive [7]. In many cases, natural resource access can be restricted at low cost, allowing well-functioning markets to allocate ecosystem resources and prices to signal economic activity. In such circumstances, market prices combined with costs and quantities demanded and supplied provide policy makers with quantified information of ecosystem service value for specific user groups [9,10]. Restricting access can be a useful policy tool when one group's activity interferes with another groups’ use of the resource. For example, overfishing through large scale commercial operations can reduce a fishery's productivity [11], and commercial fishing rights can be restricted to those holding permits. In other cases, consumption of ecosystem resources is left unrestricted so that users enjoy unfettered access to the resource. Unrestricted access can be socially desirable when resource consumption does not diminish others’ use, which generally occurs with non-consumptive recreational activ-

Marine Policy 81 (2017) 91–97

D. Viana et al.

ities like boating, wildlife viewing, and SCUBA diving [8,12,13].1 Such activities are often undercounted in policy making, partially because of the absence of price and value information from well-functioning markets. This can undermine the perceived legitimacy and acceptance of the resource management regime among important stakeholder groups [8]. Management actions such as establishing marine protected areas (MPAs), adjusting shipping lanes, and increasing large scale commercial fishing can change the value of ecosystem services for recreational users. MPAs can enhance fish abundance and diversity, which is associated with better SCUBA diving and snorkeling experiences [2,14,15]. Conversely, increased commercial fishing can reduce fish abundance and diversity and thereby negatively affect the recreational experiences of divers and recreational fishers [15]. An increase in commercial shipping activities in important whale habitat can raise the risk of whale strikes and therefore have negative impacts on whale watching [16]. Non-market valuation can contribute to accounting for ecosystems services’ overall values while also providing estimates of the value for the system's specific features, both of which can be crucial for effective policy decisions. Non-market valuation techniques estimate consumer values for ecosystem services in the absence of price signals. Site choice methods estimate consumers’ values by assuming the sum of the opportunity cost of time and travel cost to a location equals the price they would otherwise be willing to pay for the resource [17]. Differences in consumers’ willingness to travel to sites with different attributes indicate their valuation of those attributes [18]. The Random Utility Model (RUM) used in this study is a multiple-site choice model that estimates recreational demand for ecosystem services and quantifies the value of site characteristics (e.g. biodiversity, water quality). The model assumes that individuals reveal their relative values of site attributes in the sites they choose to recreate, where each visit is assumed to be a function of site attributes and the trip cost of reaching the site [19]. Such models are often used to inform policy interventions since they can be utilized to calculate monetary costs or benefits of changes in site attributes [20,21]. This paper addresses the problem of undervalued non-market ecosystem services. A site choice revealed preference model was used to estimate ecosystem service values among private recreational boaters in the Channel Islands National Marine Sanctuary (CINMS or Sanctuary), located in California – USA. Data was obtained from an intercept survey conducted by CINMS staff in 2006 and 2007 in the Sanctuary that gathered information on individuals’ characteristics and their recreational experiences [22]. This study aims to (1) quantify the value the Sanctuary provides to private recreational boaters (PRBs), and (2) determine the effect of biological and physical attributes on PRBs’ site choices. Identifying ecosystem characteristics that are important to recreational users can improve resources management in these locations. The results show that PRBs experience positive value from marine resources, around $50 per trip. Moreover, the results show, for the first time to the author's knowledge, that PRBs value biological richness and abundance in determining site choices. Such results provide managers with information on how this stakeholder group values their recreational experience and how their decisions are affected by characteristics within the Sanctuary, thus offering insight on a stakeholder group that is underrepresented by previous research.

California, USA: Anacapa, Santa Cruz, Santa Rosa, San Miguel, and Santa Barbara. The Sanctuary lies within the Southern California Bight, an area of the coastline that stretches from Point Conception to San Diego. It is an attractive location for many commercial endeavors, research activities, and recreational pastimes [23]. Within the Sanctuary is a network of 10 no-take State Marine Reserves and two limitedtake marine conservation areas [24]. With nationally significant cultural and ecological resources, the Sanctuary is responsible for maintaining local biological communities and, where appropriate, restoring and enhancing natural habitats, populations, and ecological processes [25]. On the western side of the Channel Islands, the California Current travels south year round, bringing colder, nutrient-rich water to the region. As this current reaches the U.S.-Mexico border, it turns east and begins to flow northward along the coastline, bringing warmer water into the Santa Barbara Channel and along the eastern and southern sides of the Sanctuary. This ultimately creates a temperature gradient along the island chain, which supports an array of habitats and species [26]. These habitats include kelp forests, seagrass beds, intertidal and subtidal zones, as well as benthic and pelagic habitats. High levels of productivity from nutrient rich waters sustain numerous species of invertebrates, fish and marine mammals. This productive and unique ecosystem is significant to residents and visitors along the Southern California coast because it provides many opportunities for economic and recreational engagement [27]. Each year, over 30,000 visitors visit the CINMS, with another 60,000 recreating in their surrounding waters [28]. In 1999, an estimated total visitation of 437,908 person-day (number of visitors times number of days) was recorded within the CINMS [27]. Private recreational boaters represent a significant portion of the Sanctuary's stakeholders. In 2007, about 1621 private boat trips were observed [23], with boaters engaging in a range of recreational activities, both consumptive (e.g. recreational fishing) and non-consumptive (e.g. snorkeling, whale watching). A survey conducted in 2006–2007 by postcard found that 47% of boaters engaged in only non-consumptive activities and 51% of boaters participated in both non-consumptive and consumptive activities [22]. Additionally, about 16% of users described consumptive activities as the most important factor when choosing an anchorage site, while the remaining users (84%) named environmental factors or non-consumptive activities [22]. The CINMS also supports a broad range of commercial activities, including private and charter boat recreational fishing, commercial fishing, and whale watching. Commercial fishing in the CINMS had an average harvest yearly value of $27 million and generated 659 jobs from 2010 to 2012 [29]. The whale watching industry operates daily trips to the CINMS with an estimated value of $1.5 million per year and 119 jobs [30]. The average annual economic value from private and commercial recreational fishing trips to the CINMS from 2010 to 2012 is estimated to be $31.4 million, generating around 200 full time jobs [31].2

3. Materials and methods 3.1. Data As part of a series of studies on PRBs, an Intercept Survey was conducted in 2006 and 2007 by CINMS staff and contracted researchers [22]. The intercept survey took place in waters surrounding Santa Cruz Island, although respondents reported visits at three islands: Anacapa, Santa Cruz, and Santa Rosa. The survey was conducted between the months of May and October to reflect the greatest frequency of

2. Study site The Channel Islands National Marine Sanctuary encompasses a 1470 square mile area around five of the eight Channel Islands in 1 In some cases, crowding can diminish the value of non-consumptive activities. For example, while scuba divers can generally spread out to maximally enjoy their experiences, sometimes enough divers may concentrate in an area that harms their experience [50].

2 Of course, market size is not the same as the social value derived from consumer surplus. Estimating consumer surplus for activities like commercial fishing requires information such as prices, costs, and quantities demanded and supplied.

92

Marine Policy 81 (2017) 91–97

D. Viana et al.

Table 1 Characterized recreational activities in the Channel Island National Marine Sanctuary, California, USA. Recreational activities are grouped into underwater non-consumptive, surface non-consumptive, consumptive and land based categories. Underwater nonconsumptive

Surface nonconsumptive

Consumptive

Land based

Snorkeling

Exploring by dinghy Kayaking Marine mammal watching Bird watching

Hook and line fishing Spearfishing Lobster diving

Beach going/ exploring Tide-pooling Hiking

SCUBA diving

Table 2 Summary statistics of site and individual attributes utilized in the Random Utility Model on site characteristics preferences of private recreational boaters in the Channel Islands National Marine Sanctuary, California – USA, in 2006.

Hoop netting

recreational visitation, which was determined from the Sanctuary Aerial Monitoring Spatial Analysis Program (SAMSAP) [22]. A survey of 43 questions was distributed to the operator of each approached vessel that was willing to participate. Respondents were invited aboard the Sanctuary Research Vessel in order to complete the survey. Survey questions included respondents’ ports of departure, trip activities, incomes, demographic information, and vessel types. In addition to the survey questions, respondents were asked to draw polygons of the areas where they participated in recreational activities. This was done using an interactive GIS-based survey tool called OceanMap [32]. For each polygon drawn, an activity was assigned from a list of 15 activity choices. After accounting for missed responses or collection errors, the full dataset comprised of 110 respondents who participated in 15 activities (Table 1) and drew a total of 323 polygons. The majority of survey respondents participated in more than one activity at multiple sites. These repeated decisions complicate the analysis because they require an unlikely assumption: that each respondent's choice of activity and site is independent of any previous choices. Due to the complications of modeling multiple decisions made by the same person [12], this analysis considered only one activity-site for each boater. Selecting single activity-sites required creating a hierarchy of activities [33]. As the large majority (81%) of users cited environmental factors or non-consumptive activities as their main reason for visiting the islands [22], non-consumptive activities were prioritized as the top two activity groups in the hierarchy. Additionally, non-consumptive users have been a historically underrepresented stakeholder group. Therefore, the first level in the hierarchy included non-consumptive underwater activities (N=49), followed by nonconsumptive surface activities (N=42), consumptive activities (N=5), and land-based activities (N=14). In order to create a discrete set of choices available to each individual, the coastlines of Anacapa, Santa Cruz, and Santa Rosa islands were divided into 31 unique sites. Each site was characterized with attributes thought to influence boaters’ visitation decisions, including kelp cover, exposure to winds, and biological quality. Kelp percent cover was collected from California Department of Fish and Wildlife (CDFW) aerial surveys. Due to the dynamic nature of kelp, the analysis included values for 2006 only, as this sample best represents the conditions present during the time of the survey [22]. To determine wind exposure, the direction of the coastline was used as a proxy since wind in the Channel Islands tends to come predominantly from one direction. Therefore, coastline exposure was expressed continuously from 0 to 180, where 0 represents a site with a coastline that is completely protected from the predominant winds (coastline facing SE) and 180 represents a site where the coastline has no wind protection (coastline facing NW). Table 2 provides summary statistics for our data. Fig. 1 presents boxplots of key variables, showing their variability within sites. The model measures four biological features: fish richness, fish abundance, invertebrate richness, and invertebrate abundance. Data for these measures are from the Partnership for Interdisciplinary Studies of Coastal Oceans (PISCO) [34] subtidal community surveys [12]. These

Parameter

Bio Index

Kelp Cover (%)

Exposure (degrees)

Opportunity Cost (land)

Total Travel Cost

N Mean Medium Std. dev. Max Min

31 2.03 1.93 0.34 1.45 3.12

31 4.37 1.13 6.78 0 26.63

31 88.45 99 49.50 0 165

111 $421.20 43.43 $1695.52 $14,800 $1.25

111 $517.48 120.91 $1745.56 $15,044 $25.43

Fig. 1. – Distribution of site attributes among sites in the Channel Islands National Marine Sanctuary, California – USA, in 2006.

surveys take place at various sites around the islands. A number of transects are run at each site and pass through important kelp and rocky reef habitats. In order to reach a single representative value for each site, species richness was calculated by summing the total number of species represented across all site transects. A representative abundance value was taken from the transect with the maximum reported number of fish within the site. This transect represents the best possible viewing or fishing experience for users within the site. An index was calculated to represent the biological quality of sites due to the high correlation observed between all biological attributes, which would reduce the robustness and accuracy of the Random Utility Model if we had chosen to keep them separate. The following equation was used to combine fish abundance, FAj, fish richness, FRj, invertebrate abundance, IAj, and invertebrate richness, IAj, into a single site-specific biological index, Bj:

⎛ FAj FRj IAj IRj ⎞ Bj = ⎜ + + + ⎟ FRmax IAmax IRmax ⎠ ⎝ FAmax

(1)

3.2. Travel cost analysis A travel cost analysis was conducted in order to determine how much each individual valued a particular trip to the Sanctuary. Total travel cost to recreation sites within the Sanctuary consists of two components: the actual monetary cost of traveling to and from the site (including on-land and on-water expenses such as fuel), and the opportunity cost of time associated with on-land travel only. On-water travel time was not included as part of the cost in the analysis because it was considered part of the recreational experience [35]. Some studies exclude all time travel costs on the grounds that all travel is part of the recreational experience [36,37]. In light of this debate, we also investigated analyses with both on-water and on-land travel time 93

Marine Policy 81 (2017) 91–97

D. Viana et al.

excluded from the travel cost calculation. Since the results are essentially the same as presented below, we present the model including on-land opportunity cost and discuss analysis without all travel time in the discussion section. The program PC-Miler [38] was used to calculate on‐land round-trip distance traveled between respondents’ homes and the harbor. Distance traveled was multiplied by an average fuel and maintenance cost of $0.21 per mile from the Automobile Association of America (AAA) [39] and the US Department of Energy that reflects costs corresponding to 2006/2007. To calculate the cost of traveling on water, the distances each of the boaters traveled was multiplied by the fuel cost they incurred with each mile traveled. On‐water fuel costs per mile vary according to boat type (sail, motor‐sail, power boats) and boat size. Figures for the fuel consumption of boats of different sizes (gallons per mile) were derived from Thomas and Stratis (2002) [35]. Distance from harbor to the recreation site for each boater was measured using the software ArcGIS [40]. Opportunity cost of time was only considered during on-land travel and was calculated as a function of the individual's income and distance traveled, according to data from the intercept survey (N=110) [35]. Hourly wage of each respondent was calculated by dividing their annual income values by the expected number of hours worked during the year, assuming each individual works 40 h per week. Opportunity cost of time was then computed using 50% of the individual's hourly wages [17]. Opportunity cost of time represent, on average, 40% of total individual travel cost, with a maximum of 98% and a minimum of 3%.

Pij =

exp(Vij ) m n ∑i ∑ j exp(vij )

(3)

Where m is the total number of individuals and n is the total number of activity-sites. Daily trip value is calculated based on the estimated coefficients. The total consumer surplus (CS) of a single trip is calculated using the following equation [41]:

CS = −

1 dU /dTC

(4)

Where dU/dTC is the travel cost coefficient for a particular activity group. The individual value of the Sanctuary described above was extrapolated to the entire user group of non-consumptive private recreational boaters visiting the Sanctuary. To extrapolate the total value of non-consumptive PRB use of the Sanctuary, the average value of an individual trip for non-consumptive activities was multiplied by 1,621, the total estimated number of private non-consumptive boat trips to the Sanctuary in 2007 [23]. Additionally, the monetary value of specific attributes can also be calculated [42,43]. The marginal willingness to pay for a 1-unit change in a particular attribute is given by the following equation:

MWTP = −

dU / dZ dU /dTC

(5)

Where dU/dZ is the attribute coefficient for a particular activity group and dU/dTC is the travel cost coefficient for a particular activity group.

3.3. The Random Utility Model 4. Results

The Random Utility Model is a commonly used non-market valuation technique that is applied in travel cost recreational demand analyses to quantify the value of recreation site characteristics, along with the value of recreation activities as a whole [12,33]. This type of model is appropriate when individuals are choosing between a discrete choice set of alternative recreation sites and activities, where each site has a number of unique and quantifiable attributes [33]. Choice behavior is modeled such that individuals choose sites to visit based on attributes of that site and the total travel cost to access it [12]. In this study of recreational boater use of the Sanctuary, two models were created to assess recreation activities and site characteristics: a model differentiating among activity groups and a pooled model that grouped together all non-consumptive activities into one choice category and excludes consumptive activities due to low number of observations (N=5). In the pooled model non-consumptive boaters choose among 31 possible site choices (around Anacapa, Santa Cruz, and Santa Rosa islands). In the model where activities are modeled separately (nonpooled model) each boater chooses among 31 possible sites and 4 possible activity group choices at each site. The activity groups are: underwater non-consumptive, surface non-consumptive, consumptive, and land-based, as reported in Table 1. The location and activity pairings result in a discrete choice set of 124 unique site-activity pair alternatives (henceforth referred to as “activity-sites”). It is assumed that each individual is equally willing to choose among the 124 activity-sites (or 31 sites in the pooled model), no single activity-site serves as a substitute for another, and activity-sites are chosen based on observable attributes of that option [19]. Each activity-site is assumed to give a specific individual utility, which is a function of the travel cost and site attributes. This model has the following utility function [35]:

Uij = Vij (TCij , Zj )+eij

4.1. Output coefficients from the Random Utility Model Tables 3, 4 provides the RUM results (for the non-pooled and pooled model, respectively), listing coefficients, standard errors and statistical significance. For all activity groups and models, travel cost has a significantly negative effect on respondents’ choices, indicating that boaters tended to visit sites with lower costs of travel. Travel cost have a significantly stronger effect on consumptive users’ site choice compared to underwater, surface and land users. Boaters tended to visit sites with a greater value of the biological index, which had a significantly positive effect on all activity categories, with the highest impacts observed in underwater and surface non-consumptive users. Exposure had a significantly negative effect on respondents’ choices for all activity categories, as boaters tended to choose sites with lower wind exposure. Kelp did not have a significant effect on boaters’ decisions. Across all attributes in the non-pooled model, coefficients from consumptive users were less significant than attribute coefficients of other non-consumptive activity categories. Results from pooled and non-pooled model were very similar, suggesting that the results are not sensitive to the differentiation among activity-groups. 4.2. Daily trip value Daily value of a trip to the Channel Islands was determined by using the travel cost model coefficients. The results from the non-pooled model are reported in Table 5 and indicate trip values ranging from $34.72 to $53.69 across activity categories. Trip value is highest for surface non-consumptive and underwater non-consumptive users ($53.69 and $53.21 respectively), and lowest for consumptive users ($34.72). However, these differences should be interpreted with caution because they are not different within a 95% confidence interval with a mean trip value of $48.62. The pooled model of non-consumptive activities presented a daily trip value of $50.42, with a 95% confidence interval between 40.33 and 67.24, which did not differ significantly between the two models.

(2)

Where Uij is the utility of individual i selecting activity-site choice j; TCij is the total travel cost for individual i to the activity-site j; Zj is a vector of attributes at site j; Vij is a vector of coefficients and eij is the random error term. Based on this utility function, the probability that individual i will choose activity-site j is given as: 94

Marine Policy 81 (2017) 91–97

D. Viana et al.

Table 3 Output coefficients from the non-pooled Random Utility Model on site characteristics preferences of private recreational boaters in the Channel Islands National Marine Sanctuary, California – USA, in 2006. Underwater Coefficient TRAVELCOST BIOINDEX KELPCOVER EXPOSURE Log-Likelihood: AIC Sample Size

−0.019 0.315*** −0.011 −0.024*** −400.85 833.7094 110 ***

Surface

Consumptive

Std. error

Coefficient

0.0049 0.0362 0.0214 0.0041

−0.019 0.291*** −0.012 −0.017*** ***

Std. error

Coefficient

0.0049 0.0358 0.0253 0.0040

−0.029 ** 0.291*** −0.11 −0.03*

Land Std. error

Coefficient

Std. error

0.0100 0.0567 0.1458 0.0131

−0.019 0.283*** −0.155 −0.025**

0.0050 0.0421 0.1290 0.0078

***

* α < 0.05 ** α < 0.01 *** α < 0.001

between $13.10 and $16.62. The marginal value of exposure for the pooled model was $−1.05, with a confidence interval between $−0.92 and $1.19. Both significant attributes (biological index and exposure) from the pooled model did not differ significantly from the non-pooled model. While the model does suggest that boaters who engage in underwater non-consumptive activities were willing to pay the most for biological attributes ($16.78 per unit), the differences in willingness to pay between this group, the surface non-consumptive, and land-based user groups were not statistically significant at the 95% confidence level ($15.62 and $14.98 per unit, respectively). Only consumptive PRBs, with a willingness to pay of $10.11, were statistically different from the other three activity categories. However, given the small number of respondents (N=5) in this activity group, this result should be interpreted with high degree of caution. The other three activity groups had larger sample sizes ranging from 15 to 49.

Table 4 Output coefficients from the Random Utility Model with pooled non-consumptive activities on site characteristics preferences of private recreational boaters in the Channel Islands National Marine Sanctuary, California – USA, in 2006.

TRAVELCOST BIOINDEX KELPCOVER EXPOSURE Log-Likelihood: AIC Sample Size

Coefficient

Std. error

−0.019*** 0.294*** −0.019 −0.021*** −419.03 846.0534 110

0.0049 0.0349 0.0162 0.0027

*α < 0.05, **α < 0.01 *** α < 0.001 Table 5 Non-pooled model daily trip value for private recreational boaters in the Channel Islands National Marine Sanctuary, California – USA, in 2006. Activity group

Value of 1 day trip

95% Confidence intervals

Underwater Surface Consumptive Land

$53.21 $53.69 $34.72 $52.86

[42.14, [42.44, [25.79, [41.89,

4.4. Total annual consumer surplus The total annual consumer surplus of private recreational nonconsumptive boaters estimated in the Sanctuary for 2006/2007 is $86,642. As SAMSAP aerial survey data could only differentiate nonconsumptive PRBs, this is the only user group that could be aggregated. Consumptive PRBs, therefore, are not included in this aggregate value, meaning that this value is an underestimate of the value the entire community of PRBs associates with the Sanctuary.

72.17] 73.06] 53.10] 71.61]

4.3. Marginal value of significant attributes Table 6 reports the marginal value of significant attributes for each activity category for the non-pooled model. The marginal value of one biological index unit ranged from $10.11 to $16.78, indicating that PRBs placed more value on more biologically rich locations. Underwater non-consumptive, surface non-consumptive, and land-based groups showed statistically similar biological valuations, while the consumptive group retained a statistically significant lower value. The marginal value of exposure to predominant wind and swell ranged from -$0.91 to -$1.32, with no significant difference among user groups. The pooled model of non-consumptive activities presented a marginal value of one biological index unit of $14.86, with a confidence interval

5. Discussion The results demonstrate the influence of travel cost and sanctuary site attributes on boater behavior. The negative correlation between travel cost and site choice in the model validates the assumption that users consider cost and distance when choosing places to recreate. A positive biological index coefficient suggests that private recreational boaters consider some measure of underwater biological quality in their site choice. Land-based non-consumptive users are not directly affected by underwater biological quality. However, it is possible that PRBs are making their choices based on attributes omitted from this analysis that

Table 6 Marginal willingness to pay for one unit of the biological index and wind exposure for private recreational boaters in the Channel Islands National Marine Sanctuary, California – USA, in 2006. Underwater

BIOINDEX EXPOSURE

Surface

Consumptive

Land

Marginal value

95% Confidence Interval

Marginal value

95% Confidence Interval

Marginal value

95% Confidence Interval

Marginal value

95% Confidence Interval

$16.78 -$1.28

[14.85, 18.70] [−1.49, −1.06]

$15.62 -$0.91

[13.70, 17.54] [−1.13, −0.69]

$10.11 −$1.03

[8.14, 12.08] [−1.49, 0.58]

$14.98 −$1.32

[12.75, 17.21] [−1.73, −0.91]

95

Marine Policy 81 (2017) 91–97

D. Viana et al.

actual costs of owning and maintaining a boat. The true amount a PRB pays to visit the CINMS is likely to be substantially greater than the cost of fuel, vehicle maintenance, and opportunity cost alone. Based on the results of the model, managers may incorporate the benefits PRBs receive from a single trip to the Sanctuary (i.e. about $50) in quantitative or qualitative analyses for policy decisions. To improve the value that recreational boaters gain from access to the CINMS, managers may also consider actions that protect or restore the biological quality of the Sanctuary. The research suggests that these types of actions are especially relevant for non-consumptive PRBs. Water quality improvements, limits on oil and gas development, fishing restrictions and marine protected areas are examples of management actions that are linked to proportional increases in fish and invertebrate biomass for species targeted by fishing [24,45]. To the authors’ knowledge this is the first study to monetize the value of biological quality to private recreational boaters. This result can be used elsewhere with intense boating activity to justify management actions that protect the environment. Furthermore, a value of $35 to $54 per trip represents a reasonable estimate for PRBs according to the literature. Using data from a 1984 study on private boaters, Leeworthy et al. [27] found a per trip value of $43.25 for private recreational boaters and $45.6 for non-consumptive SCUBA divers within the Channel Islands (adjusted for 2007 values). Several reviews [14,46–48] and databases [49] show a wide range of non-market values for marine recreation, with most values similar to what was found in this study. As the analyses in this paper indicate, distance from port is a substantial deterrent for PRBs in choosing sites as recreational boating is most often concentrated in islands closer to port such as Anacapa and Santa Cruz. This finding is consistent with other research, which report a spatial concentration of boaters within the eastern portion of the Sanctuary [22]. The trend suggests that management efforts that impact these highly frequented regions are likely to disproportionately affect PRB activities. Directing policy to maintain the quality and value of eastern sites may therefore be an effective use of limited management resources. The estimated total value of the Sanctuary to PRBs ($86,259) is only a small proportion of the total potential recreational non-market value of the sanctuary. Recreational boaters comprise about 4% of the total number of non-consumptive recreational users of the Sanctuary, most of whom engage in non-market activities such as island sightseeing, kayaking and whale watching [27]. Furthermore, recreational users are an important stakeholder group with high social importance for the Sanctuary [23]. Unlike many commercial activities, recreational users’ enjoyment of the sanctuary often does not interfere with other market and non-market uses, while commercial uses can interfere with noncommercial uses. Thus, the finding that PRBs have a positive consumer surplus suggests that the total value for all non-market users is likely to be substantial. Optimizing tradeoffs among competing market and nonmarket uses is beyond the scope of this project. However, the results of this study provide new quantitative information describing the value of ecosystem services, thus contributing to more balanced management of marine resources.

are correlated with underwater biodiversity; for example, seabird diversity and abundance may be dependent upon fish abundance as a prey resource. Given that the index is a compilation of fish and invertebrate species richness and abundance, the model suggests that improved biodiversity and productivity of marine ecosystems contribute to better recreational experiences [2]. The results also suggest that PRBs consider wind exposure of the coastline when choosing sites, with increased exposure serving as a deterrent. This is logical because improved protection from the predominant wind and swell in sites with lower exposure is likely to result in a safer and more pleasant experience of the activities considered in the study. The statistical insignificance of percent kelp cover in the model suggests that PRBs do not consider the presence of kelp at the surface in their decision process. However, this result may also be due to the fact that kelp cover is highly dynamic [44], and therefore the aerial survey of kelp data taken in October of 2006 might not accurately reflect kelp coverage patterns at the time of the survey (May-Oct 2006/2007). To investigate whether our results were sensitive to pooling activities and to how we measured travel costs, we conducted two additional analyses as specification checks. First, we compared analyses that differentiated among consumptive activities (Table 3) to analyses that combined consumptive activities (Table 4). The results from the pooled model were very similar to the model that differentiated activity groups, demonstrating that actual activities are not as important. The pooled model was tested because of high similarity of results among the different activity groups, especially the non-consumptive groups (underwater, surface and land). This high similarity might be explained by the fact that these activities can occur concomitantly in the same areas. Thus, the trip to the reserve itself is what is actually being modeled and travel costs to each individual site dominate travel cost estimates. Second, we explored whether our results are sensitive to measurement assumptions when calculating time travel costs. The analyses above include land travel time in the cost calculation and excluded on-water travel time, on the assumption that the latter is part of the recreational experience while the former is not. The use of opportunity cost of time in travel cost models is a source of some debate since the trip to and from the area can be considered as part of the recreational activity [17,36]. In light of these issues, we also estimated a model excluding all travel time costs. Although the model with opportunity cost of time had slightly higher average daily trip value than the model without opportunity cost of time, the coefficients between the two were almost indistinguishable, suggesting that our conclusions are not sensitive to the use of opportunity cost of travel time. This might be explained by the fact that most respondents did not travel great distances on-land so that on-water fuel costs dominated travel cost estimates. Additionally, only the opportunity cost of time on-land was considered in the original model, with on-water travel costs being considered a part of the recreational experience in both models [35]. Data limitations and inherent assumptions of the random utility model [19] suggest that the results may be underestimates, and therefore the total CINMS recreational value of $86,259 should be considered a lower bound. Each boat was assumed to carry a single individual since the survey provided demographic data on only the boat's captain. In reality, it is highly likely that most boats carried multiple passengers. Assuming a single travel cost per boat undervalues trips with multiple individuals because it ignores the on-land travel and opportunity cost of all of the passengers excluding the captain. Furthermore, the RUM model only captures the utility PRBs receive from participating in a single activity in one location. In reality, most boaters participated in many activities in multiple sites. Any additional utility PRBs may derive from these extra activities and/or sites was not included because of the model assumptions. It is expected that boaters are receiving utility from these additional activities, so by not including them, total utility is underestimated. Additionally, extra expenditures were not accounted for, including but not restricted to food/beverage costs, gear, slip fees (for those boats that are kept in harbors), and the

6. Conclusion Quantifying the value of recreational activities in areas where conventional economic markets do not operate is challenging and complex. Non-market valuation methods are important tools for demonstrating that resources have important value to a variety of stakeholders and users. Managers otherwise risk mismanaging natural resources by undervaluing ecosystem services and their uses at different locations. Random utility models can be an effective methodology for assessing the value of non-market ecosystem services. The total value of PRBs to the CINMS found in this study can be used in future cost-benefit analysis for policy and resource management decisions. Furthermore, the monetary value that PRBs place on the biological quality of a given 96

Marine Policy 81 (2017) 91–97

D. Viana et al.

2873ac54e51f9e04d3e3e4d4d848bca4〉, 2002. [19] J. Walker, M. Ben-Akiva, Generalized random utility model, Math. Soc. Sci. 43 (2002) 303–343, http://dx.doi.org/10.1016/S0165-4896(02)00023-9. [20] Y. Kaoru, Measuring marine recreation benefits of water quality improvements by the nested random utility model, Resour. Energy Econ. 17 (1995) 119–136. [21] N.E. Bockstael, K.E. McConnell, I.E. Strand, A random utility model for sporfishing: some preliminary results for Florida, Mar. Resour. Econ. 6 (1989) 245–246. [22] C. LaFranchi, L. Pendleton, Private Boating and Boater Activities in the Channel Islands: A Spatial Analysis and Assessment. 〈http://sanctuaries.noaa.gov/science/ socioeconomic/channelislands/pdfs/privboat1.pdf〉, 2008. [23] V.R. Leeworthy, Estimates of Non-consumptive Recreation Private Household Boats in the Channel Islands National Marine Sanctuary 2007 19. 〈http://sanctuaries. noaa.gov/science/socioeconomic/channelislands/pdfs/cinms_use_estimates.pdf〉, 2013. [24] S. Airamé, J.E. Dugan, K.D. Lafferty, H.M. Leslie, D. McArdle, R. Warner, Applying ecological criteria to marine reserve design: a case study from the California Channel Islands, Ecol. Appl. 13 (2003) 170–184, http://dx.doi.org/10.1890/10510761(2003)013[0170:aectmr]2.0.co;2. [25] National Marine Sanctuary Act, 〈http://sanctuaries.noaa.gov/library/national/ nmsa.pdf〉, 2000. [26] S.L. Hamilton, J.E. Caselle, D.P. Malone, M.H. Carr, Incorporating biogeography into evaluations of the Channel Islands marine reserve network, Proc. Natl. Acad. Sci. U.S.A. 107 (2010) 18272–18277, http://dx.doi.org/10.1073/pnas. 0908091107. [27] V.R. Leeworthy, P.C. Wiley, Socioeconomic Impact Analysis of Marine Reserve Alternatives for the Channel Islands National Marine Sanctuary, 2003, p. 118. [28] D.L. Engle, Assessment of coastal water resources and watershed conditions at Channel Islands National Park. 〈http://www.nps.gov/chis/learn/management/ loader.cfm*CsModule=security/getfile&PageID=128403〉, 2006. [29] V.R. Leeworthy, D. Jerome, K. Schueler, Economic Impact of the Commercial Fisheries on Local County Economies from Catch in the Channel Islands National Marine Sanctuary 2010, 2011 and 2012, Mar. Sanctuaries Conserv. Ser. ONMS-1404, U.S. Dep. Commer. Natl. Ocean. Atmos. Adm. Off. Natl. Mar. Sanctuaries, Silver Spring, MD, 2014, p. 33. [30] NOAA, Channel Islands National Marine Sanctuary Socioeconomics, 〈http:// sanctuaries.noaa.gov/science/socioeconomic/pdfs/ci_final.pdf〉 n.d. [31] V.R. Leeworthy, D. Schwarzmann, Economic impact of the recreational fisheries on local county economies in the channel Islands National Marine Sanctuary 2010, 2011 and 2012, Mar. Sanctuaries Conserv. Ser. 3 (2015) 23. [32] M.S. Merrifield, W. McClintock, C. Burt, E. Fox, P. Serpa, C. Steinback, et al., MarineMap: a web-based platform for collaborative marine protected area planning, Ocean Coast. Manag. 74 (2013) 67–76, http://dx.doi.org/10.1016/j. ocecoaman.2012.06.011. [33] C.F. Manski, The structure of random utility models, Theory Decis. 8 (1977) 229–254, http://dx.doi.org/10.1007/BF00133443. [34] M.A. Janssen, R. Holahan, A. Lee, E. Ostrom, Partnership for Interdisciplinary Studies of Coastal Oceans (PISCO). 〈http://www.piscoweb.org/research〉, 2009. [35] M. Thomas, N. Stratis, Compensating variation for recreational policy: a Random Utility Approach to boating in Florida, Mar. Resour. Econ. 17 (2002) 23–33. [36] V.K. Smith, W.H. Desvousges, M.P. McGivney, The opportunity cost of travel time in recreation demand models, Land Econ. 59 (1983) 259–278. [37] K.E. Mcconnell, I. Strand, Measuring the cost of time in recreation demand analysis: an application to sportfishing, Am. J. Agric. Econ. 63 (1981) 153–156. [38] Alk Technologies, PC-Miler Version 23.1, 2012. [39] AAA, Your Driving Costs. 〈www.AAA.com〉, 2011. [40] Environmental Systems Rresearch Institute, ArcGIS Desktop, 2011. [41] H.S. Rosen, K. a. Small, Applied welfare economics with discrete choice models, Econometrica 49 (1981) 105–130〈http://www.nber.org/papers/w0319/ npapers3://publication/uuid/1D54B6DE-3E46-4B55-B494-A08A98FF2368〉. [42] M. Sillano, J.D.D. Ortúzar, Willingness-to-pay estimation with mixed logit models: some new evidence, Environ. Plan. A. 37 (2005) 525–550, http://dx.doi.org/10. 1068/a36137. [43] S. Banfi, M. Farsi, M. Filippini, M. Jakob, Willingness to pay for energy-saving measures in residential buildings, Energy Econ. 30 (2008) 503–516, http://dx.doi. org/10.1016/j.eneco.2006.06.001. [44] P.K. Dayton, M.J. Tegner, P.E. Parnell, P.E. Edwards, Temporal and spatial patterns of disturbance and recovery in a kelp forest community, Ecol. Monogr. 62 (1992) 421–445, http://dx.doi.org/10.2307/2937118. [45] S. Lester, B. Halpern, K. Grorud-Colvert, J. Lubchenco, B. Ruttenberg, S. Gaines, et al., Biological effects within no-take marine reserves: a global synthesis, Mar. Ecol. Prog. Ser. 384 (2009) 33–46, http://dx.doi.org/10.3354/meps08029. [46] L.M. Brander, P. Van Beukering, H.S.J. Cesar, The recreational value of coral reefs: a meta-analysis, Ecol. Econ. 63 (2007) 209–218, http://dx.doi.org/10.1016/j. ecolecon.2006.11.002. [47] A. Ghermandi, P. a.L.D. Nunes, A global map of coastal recreation values: results from a spatially explicit meta-analysis, Ecol. Econ. 86 (2013) 1–15, http://dx.doi. org/10.1016/j.ecolecon.2012.11.006. [48] L. Pendleton, P. Atiyah, A. Moorthy, Is the non-market literature adequate to support coastal and marine management*, Ocean Coast. Manag. 50 (2007) 363–378, http://dx.doi.org/10.1016/j.ocecoaman.2006.11.004. [49] C. Plantier-Santos, C. Carollo, D.W. Yoskowitz, Gulf of Mexico Ecosystem Service Valuation Database (GecoServ): gathering ecosystem services valuation studies to promote their inclusion in the decision-making process, Mar. Policy 36 (2012) 214–217, http://dx.doi.org/10.1016/j.marpol.2011.05.006. [50] A. Coghlan, Linking natural resource management to tourist satisfaction: a study of Australia's Great Barrier Reef, J. Sustain. Tour. 20 (2012) 41–58, http://dx.doi.org/ 10.1080/09669582.2011.614351.

site can be used in the Sanctuary and other locations elsewhere in the world where recreational boaters use marine resources. These results can contribute to a more effective management of the Channel Islands National Marine Sanctuary and other marine ecosystems around the globe. Future research can focus on other types of non-market activities using the RUM model and how better measurement of non-market services can improve resource management. Acknowledgements This study was made possible by contributions from the Environmental Defense Center and the Channel Islands National Marine Sanctuary. This work would not have been possible without the help of Sean Hastings, Bob Leeworthy, Jenn Caselle, Chris Cohen, Katie Davis, Rod Ehler, James Frew, Steve Katz, Charles Kolstad, Todd Matson, Will McClintock, Steve Miller, Craig Mohn, Corey Olfe, and others who generously provided data, feedback, and technical assistance. Earlier drafts received helpful comments from Dan Ovando, Sean Hastings and Bob Leeworthy. Lastly, this work is dedicated to ChingCheng Lin, one of the authors of this project who is no longer with us. References [1] B.S. Halpern, C. Longo, D. Hardy, K.L. McLeod, J.F. Samhouri, S.K. Katona, et al., An index to assess the health and benefits of the global ocean, Nature 488 (2012) 615–620, http://dx.doi.org/10.1038/nature11397. [2] E. Sala, C. Costello, D. Dougherty, G. Heal, K. Kelleher, J.H. Murray, et al., A general business model for marine reserves, PLoS One 8 (2013) 1–9, http://dx.doi.org/10. 1371/journal.pone.0058799. [3] S. Pascoe, A. Doshi, Q. Dell, M. Tonks, R. Kenyon, Economic value of recreational fishing in Moreton Bay and the potential impact of the marine park rezoning, Tour. Manag. 41 (2014) 53–63, http://dx.doi.org/10.1016/j.tourman.2013.08.015. [4] B.S. Halpern, S. Walbridge, K. a Selkoe, C.V. Kappel, F. Micheli, C. D’Agrosa, et al., A global map of human impact on marine ecosystems, Science 319 (2008) 948–952, http://dx.doi.org/10.1126/science.1149345. [5] S.E. Rees, L.D. Rodwell, M.J. Attrill, M.C. Austen, S.C. Mangi, The value of marine biodiversity to the leisure and recreation industry and its application to marine spatial planning, Mar. Policy 34 (2010) 868–875, http://dx.doi.org/10.1016/j. marpol.2010.01.009. [6] G.R. Parsons, S.M. Thur, Valuing changes in the quality of coral reef ecosystems: a stated preference study of SCUBA diving in the Bonaire National Marine Park, Environ. Resour. Econ. 40 (2008) 593–608, http://dx.doi.org/10.1007/s10640007-9171-y. [7] A. Ruiz-Frau, M.J. Kaiser, G. Edwards-Jones, C.J. Klein, D. Segan, H.P. Possingham, Balancing extractive and non-extractive uses in marine conservation plans, Mar. Policy 52 (2014) 11–18, http://dx.doi.org/10.1016/j.marpol.2014.10.017. [8] N. Raheem, S. Colt, E. Fleishman, J. Talberth, P. Swedeen, K.J. Boyle, et al., Application of non-market valuation to California's coastal policy decisions, Mar. Policy 36 (2012) 1166–1171, http://dx.doi.org/10.1016/j.marpol.2012.01.005. [9] C. White, B.S. Halpern, C.V. Kappel, Ecosystem service tradeoff analysis reveals the value of marine spatial planning for multiple ocean uses, Proc. Natl. Acad. Sci. 109 (2012) 4696–4701, http://dx.doi.org/10.1073/pnas.1114215109/-/ DCSupplemental.www.pnas.org/cgi//10.1073/pnas.1114215109. [10] F.W. Bell, Food from the SEA: the Economics and Politics of Ocean Fisheries, Westview Press Inc., Boulder, Colorado, 1978. [11] J.B. Jackson, M.X. Kirby, W.H. Berger, K. a Bjorndal, L.W. Botsford, B.J. Bourque, et al., Historical overfishing and the recent collapse of coastal ecosystems, Science 293 (2001) 629–637, http://dx.doi.org/10.1126/science.1059199. [12] T. Haab, K. McConnell, Valuing environmental and natural resources: the econometrics of non-market valuation, Edward Elgar Publishing, 2002http://dx.doi.org/ 10.1111/j.0002-9092.2005.740_2.x. [13] S. Pascoe, A. Doshi, O. Thébaud, C.R. Thomas, H.Z. Schuttenberg, S.F. Heron, et al., Estimating the potential impact of entry fees for marine parks on dive tourism in South East Asia, Mar. Policy 47 (2014) 147–152, http://dx.doi.org/10.1016/j. marpol.2014.02.017. [14] M. Farr, N. Stoeckl, R. Alam Beg, The non-consumptive (tourism) “value” of marine species in the Northern section of the Great barrier reef, Mar. Policy 43 (2014) 89–103, http://dx.doi.org/10.1016/j.marpol.2013.05.002. [15] I.D. Williams, N.V.C. Polunin, Differences between protected and unprotected reefs of the western Caribbean in attributes preferred by dive tourists, Environ. Conserv. 27 (2000) 382–391, http://dx.doi.org/10.1017/S0376892900000436. [16] M. Berman-Kowalewski, F.M.D. Gulland, S. Wilkin, J. Calambokidis, B. Mate, J. Cordaro, et al., Association Between Blue Whale (Balaenoptera musculus) Mortality and Ship Strikes Along the California Coast, Aquat. Mamm. 36 (2010) 59–66, http://dx.doi.org/10.1578/AM.36.1.2010.59. [17] F.A. Ward, D. Beal, Valuing Nature With Travel Cost Models: A Manual, Cheltenham, UK; Northhampton, Edward Elgar, MA, USA, 2000. [18] T.C. Haab, K.E. McConnell, Valuing Environmental and Natural Resources. 〈http:// www.sciencedirect.com/science/article/B6VP6-4FPJBC8-1/2/

97