Wildlife corridors may help to mitigate the effects of decreasing landscape connectivity resulting from land conversion and habitat degradation, however corridor identification and assessment frequently rely on anecdotal information or complex models built on inaccurate land cover datasets. The resulting landscape connectivity models often lack wildlife movement data to verify their accuracy, leading to concerns about how well they match ecological realities on the ground. To address these concerns, here we use least-cost methods coupled with the most accurate and up-to-date land conversion dataset for East Africa created using high-resolution Google Earth imagery, to model potential locations of wildlife corridors throughout the whole of Tanzania. We compare the resulting 48 connectivity model outputs to interview data on the location of wildlife movements in eastern Tanzania to determine the most accurate corridor model.
Results/Conclusions
We identify a total of 52 structural connections between protected areas that are potentially open to wildlife movement, and in so doing add 24 to those initially identified by other methods in Tanzanian Government reports (28 of the original 34 that likely remain open). We find that the vast majority of corridors noted in earlier reports as “likely to be severed” have not been cut structurally (20 of 24). Nonetheless, a sixth of all the wildlife corridors identified in Tanzania in 2009 have potentially been separated by land conversion, and 38% now pass across lands likely to be converted to human use in the near future. Our study uncovers two corridor hubs (Uvinza Forest Reserve in the west and Wami-Mbiki Wildlife Management Area in the east) linking three or more protected areas, which require far more serious conservation support. Methods used in this study are readily applicable to other nations lacking detailed data on wildlife movements and plagued by inaccurate land cover datasets.