Public support for restoration: Does including ecosystem services as a goal engage a different set of values and attitudes than biodiversity protection alone?

ES benefits, with no benefit to flora or fauna). The scenarios also necessarily differed in length and cognitive burden because we needed to explain only the biodiversity benefits in one sce- nario, but both biodiversity and ES benefits in the other. We also gauged agreement with statements about the potential drawbacks of restoration (S3 Appendix), using a 5-point Likert scale from “strongly disagree” to “strongly agree.” Scores were combined into a Disservices Index by assigning the value 3 for “neither agree nor dis- agree,” 4 for “agree” and 5 for “strongly agree” with statements about potential disservices, and 2 for “disagree” and 1 for “strongly disagree” statements. We used ordinary least squares regression models to understand the relationships between variables. To compare the impact of different attitudes and values on WTP and on the Disser- vices Index (DI), we used multiple regression and compared standardized partial regression coefficients among variables predicting WTP and DI. We used logistic regression to predict, from values and attitudes, the probability of making binary choices, e.g., preferring an ecosys- tem service over a biodiversity benefit. Biospheric and Altruistic value scores were square- transformed, and WTP amounts log-transformed, to reduce heteroskedasticity. All analyses were performed in R [48]. This research was performed according to standards designed to protect human subjects and approved by the Institutional Review Board of Santa Clara University, permit number 14-

06-536, with Recognition University of Queensland HREC Approval 20140000747. Written consent was obtained from anonymous respondents by their acceptance of the consent form in the first page of the online survey. Results In the BO scenario, Biospheric was the only significant value orientation; by contrast, in the BES scenario, the only significant value orientation was Egoistic (Table 1). An Altruistic value orientation did not significantly predict WTP in either case. If we do not adhere strictly to a standard of α = 0.05 for significance, we could note that all value orientations have p-values below 0.10 in the BO scenario; however, the Egoistic value orientation emerges as the only strong value predictor in the BES scenario. Regardless, in both scenarios, values were less influ- ential than attitudes, as WTP was most strongly predicted by scores on the Preservation axis. The Utilization axis was negatively associated with WTP in the BES scenario. Values and atti- tudes explained relatively little of the variance in WTP: adjusted R2 values were 0.134 and 0.117 for the BES and BO scenarios, respectively. We also asked how values and attitudes influenced respondents’ choice of an ES rather than a biodiversity attribute as the preferred outcome of restoration in the BES scenario (Table 2). The Egoistic value orientation and Utilization attitude were positively associated with choosing Table 1. Regression statistics for the dependence of willingness to pay on values and attitudes in each restoration scenario. BO scenarioa BES scenariob Standard β t p Standard β t p

Altruistic -0.063 -1.66 0.098 -0.027 -0.73 0.467 Biospheric 0.100 2.30 0.022 0.041 0.97 0.332 Egoistic 0.061 1.88 0.060 0.069 2.23 0.026 Preservationist 0.269 7.23 < .001 0.277 7.67 < .001 Utilizationist -0.066 -1.67 0.095 -0.123 -3.30 0.001 a Biodiversity Only scenario. b Biodiversity plus Ecosystem Services scenario. https://doi.org/10.1371/journal.pone.0245074.t001 PLOS ONE Values and attitudes towards biodiversity and ecosystem service goals of restoration PLOS ONE | https://doi.org/10.1371/journal.pone.0245074 January 19, 2021 5 / 15