Integrating VGI and Authoritative Data for Wildlife Habitat Mapping

Wildlife habitat mapping is widely used to support decision making in conservation. It requires environmental data and wildlife data as inputs. With the rapid advancement of geospatial technologies such as remote sensing, environmental data can be easily obtained. Collecting wildlife data requires much effort, especially in remote mountainous areas. Under such circumstances VGI (volunteered geographic information) can help provide wildlife data. Local villagers such as farmers, shepherds and hunters encounter wildlife on their daily routines and thus sightings can be elicited from villagers for habitat mapping. On the other hand, nature reserve administrations conduct regular patrols to monitor wildlife in protected areas, which can provide authoritative data for habitat mapping. However, wildlife sightings from villagers or patrols suffer certain data quality issues such as spatial bias and positional uncertainty.

This research develops a framework for integrating wildlife sightings from villagers and patrols for habitat mapping, hypothesizing that data integration would overcome the shortcomings of individual data sources and maximize their combined strengths to improve the accuracy of habitat mapping. A case study of mapping habitat suitability for the black-and-white snub-nosed monkey (Rhinopithecus bieti) based on sightings obtained from local villagers and nature reserve patrol records is conducted in Yunnan, China. Data integration was explored at three levels (data-level, model-level and map-level) for habitat mapping.