Abstract
Human activities around parks can alter vegetation patterns within them, resulting in edge effects that degrade their ability to sustain ecological processes and support biodiversity. We quantified vegetation patterns and edge effects over a large geographic extent in Murchison Falls Conservation Area, Uganda, using freely available remotely sensed data. Over a 13-yr period, we quantified seasonal patterns in productivity using the Normalized Difference Vegetation Index (NDVI), comparing the park exterior, a boundary zone <= 10 km from the border inside the park, and the park interior. To identify the extent of edge effects, we further fit mixed models by vegetation type within 1-km bands within the boundary zone. Productivity was higher in the park interior than exterior in both wet and dry seasons, and in the wet season, it was also lower in the boundary zone than the interior. NDVI variability differed between seasons; it declined from exterior to interior in the wet season, but was highest in the boundary zone and similar between interior and exterior during the dry season. Within the boundary zone, edge effects varied by land cover type and extended 4-6 km into MFCA. Abrupt differences in vegetation patterns between the park and adjacent unprotected areas indicated a "hard edge" in this system. While hard edges are readily apparent, the subtle changes in productivity that extend into and degrade park systems are harder to detect. We demonstrated a low-cost and novel approach to detect such effects using readily available satellite imagery, which indicated human influence affecting 29-40% of the park. As human populations grow, parks will become further isolated, and measuring and managing edge effects may be crucial to achieving conservation objectives.