Mangrove ecosystem services discrete choice experiment in Tobago
Mangrove provide many benefits to society through various goods and services such as coastal protection, water flow regulation, carbon sequestration and recreation. However, the coastal zone is an area that is subject to multiple development scenarios in a limited land space. Given the issues surrounding the trade-offs between economic development and mangrove conservation where there is considerable land use challenges due to coastal development it seems timely to explore the preferences of local residents in particular for mangrove ecosystem services over economic development and vice versa, using the island of Tobago as a case study. We achieve this in a few ways, by applying methodological treatments to a discrete choice experiment. This was undertaken by 1) estimating the willingness-to-pay (WTP) of Tobagonian households for a mangrove conservation programme; 2) use various methodological treatments in payment horizons and individual characteristics to determine if length of payments, uncertainty and socio-demographic characteristics influence respondent preferences in the case; 3) conduct follow-up surveys to explore effects on WTP of the attributes and respondents’ thinking; 4) consider potential implications of our findings on mangrove conservation policy. The data shows the code, the derived model output and the dataset for all Hierarchical Bayesian models analysed with Hamiltonian Monte-Carlo Markov Chains in Stan and could be used in either PyStan or RStan. Stan is an open-source software available from https://mc-stan.org/. Details can be found in the documentation from https://mc-stan.org/users/documentation/. The output is expressed in willingness-to-pay Space in Trinidad and Tobago Dollars (TT) and are directly interpretable in monetary values from the mean, median, standard deviation and confidence intervals. The data was gathered from a questionnaire using convenience sampling targeted key decision makers' of households on their preferences for mangrove ecosystem services at the Bon Accord/Buccoo Bay site. Study participants (n = 292) were given 12 choice tasks (6 at 5 year payment horizon and 6 at 25 year payment horizon) and asked to choose one option amongst 3 alternatives (status quo, option A and option B) with the 7 attributes: fisheries; coastal erosion; flood frequency; mangrove cover; number of species; tourism revenue; and price. The data is interpreted as the WTP for an increase to mangrove ES for instance, using the merged model- with certainty scaling there is a mean WTP by respondents of approximately $29 TT in tax annually to avoid 4 floods. Notable findings include a high WTP to avoid 50% mangrove cover removal at $102 TT per year and a WTP of $30 TT to avoid a 30% decrease in tourism revenue. The data is used to give monetary values to non-market goods expressing the trade-offs across ecosystem services which can have policy implications.