COS 9-3
Quantitative scenarios for land cover change in New Hampshire: What is the potential impact on ecosystem services?
Ecosystem services depend on land cover, which is shaped by human activities and ecological and physical processes. Decadal change in land cover is difficult to predict because of the challenges to predicting human behavior over long time scales. Therefore, a scenarios approach is suitable for modeling the long term effects of land cover. We developed four scenarios for New Hampshire land cover, based on key informant interviews with stakeholders and a review of existing plans, visions and surveys. We identified 13 stakeholder groups representing five key sectors within the state: environmental nonprofits, business and industry, timber interests, public sector agencies, and academics and natural resource management consultants. In interviews, stakeholders were prompted to describe: 1) what they would like New Hampshire to look like, and 2) what they expect New Hampshire to look like, two to four decades in the future. We combined interview notes with published documents to develop scenario narratives, and for each scenario we simulated land change based on regression tree (RT) models of land cover change between 1996 and 2010. Although all our models were based on RT analysis, we permitted other structural differences among the models used for the quantitative simulation of each scenario.
Results/Conclusions
Our scenarios vary from concentrated development prioritizing ecosystem services (Community Amenities = CA) to spatially dispersed development not prioritizing ecosystem services (Backyard Amenities = BA). Between our CA and BA scenarios is our Current Trends (CT) scenario, which extends current rates of growth, development, and conservation. We also include a variant on CA, based on a New England Food Vision scenario in which agricultural land area expands such that by the year 2060, half of the calories consumed by New England residents are produced locally (Community Amenities with Agriculture = CAwA). By 2100, substantial forest area is lost to development in BA and CT (21% and 10% respectively), and to agriculture in CAwA (13%). Impervious surface is lowest in CA and CAwA, and highest in BA. In the CA scenarios, runoff is reduced due to low impact development. Habitat fragmentation is lowest for CA, intermediate for CT, and high for CAwA and BA. Statewide agricultural productivity increases dramatically CAwA compared to little change in other scenarios. Our scenarios provide outer bounds on possible future change, and will serve as input for aquatic and ecosystem models quantifying impacts on productivity, hydrology, water quality, flooding, and other ecosystem services.