| 英文摘要: | To the Editor —
Climate change is, and will continue, altering the supply of ecosystem services1, 2. Cultural ecosystem services provide important societal benefits but are challenging to operationalize2, 3, 4, 5. The impact of warming on these cultural activities, such as ice skating, are likely to be among the most broadly obvious and compelling impacts of climate change6. Here we report that the availability and benefits of skating on the world's largest outdoor ice skating facility: declined from 1972 to 2013, was strongly dependent on weather, and is projected to continue declining with an accelerated rate between 2020–2090.
Ecosystem services, or the “benefits people obtain from ecosystems”, can be categorized as provisioning, regulating, supporting, or cultural services2, 3. Cultural ecosystem services, or “non-material benefits”2, include aesthetics, spirituality, education, and recreation. These are often intangible, subjective, and difficult to quantify2, 3, 4, 5, 7, 8, particularly in the context of human benefits3. Cultural ecosystem services are generally underrepresented within ecosystem service research5, 8, and climate change projections1. Yet these services are amongst the most recognized and acknowledged by the general public3, 4, 5, 6.
We projected local weather-mediated declines in the availability and benefit of a recreational cultural ecosystem service: outdoor ice-skating. Clearly ice9, and by extension, ice-based recreation6, 10, will be impacted by a changing climate11. The Rideau Canal in Ottawa, Canada, is the world's largest outdoor ice-skating surface and a UNESCO heritage site, with up to 1.3 million visitors annually (Supplementary Fig. 1). We used season length, or the days between opening and closing, to represent ecosystem service availability, and user number to represent service benefit3. We evaluated weather as a predictor of availability, and fitted ecological models to the relationship between use and availability12. Treating user numbers as analogous to resource consumption, and skating days as analogous to resource density, we compared three responses: constant increase in use with more skating days (type I); saturating increase in use with more skating days (type II); and accelerating increase in use initially, followed by saturating increase in use with more skating days12 (type III, Supplementary Fig. 2). Combining these models with MarkSim weather projections13 (Supplementary Methods), we projected the availability and use of this cultural ecosystem service to 2090.
Unsurprisingly, service availability was highly weather dependent. From 1972–2013, season length varied substantially year-to-year (35–90 days) with an overall decline (-5.2 ± 2.9 days per decade; 95% CI; Fig. 1; Supplementary Table 1) driven by later opening dates (6.3 ± 2.0 days per decade), not earlier closing dates (1.0 ± 2.6 days per decade). Among top models (maximum variance explained, R2max = 0.53), the most important explanatory variable was mean daily temperatures of the 100 coldest days of the year (Supplementary Methods). Using skating days (season length minus within-season closures) as the response variable had the same result (Supplementary Table 2). |