The preservation of biodiversity during forest harvesting is a key objective of riparian buffer design, yet the effectiveness of different buffer types at preserving biodiversity is rarely tested. Leaving a strip of forest along streams should preserve habitat, thus protecting organisms on land and in water. Headwater stream macroinvertebrates are particularly vulnerable to the increases in light, increased sediment input, and changes in organic matter input associated with clear cutting, and therefore provide an ideal study organism for testing the effectiveness of riparian buffer management. In this project, we investigated the effectiveness of different buffer alternatives at preserving biodiversity. Using harvest trials, we performed a BACI (before-after-control-impact) study, where we assessed macroinvertebrate community changes 1 year before and 1-3 years after harvest under four different buffer alternatives (business as usual, variable retention, one-sided and gapped), and compared these to two unharvested controls and one harvested site without a buffer (negative reference). All treatment sites had better buffers compared to standard forestry practices, with the mean buffer width > 10-20 m on each side, compared to an industry standard of less than 7 m. We measured univariate diversity and multivariate community responses, and functional feeding group responses. Increases in the number of filter feeders and decreases in the numbers of gatherers three years post-harvest indicates some change in fine particulate organic matter dynamics. Minimal differences in invertebrate diversity, community composition, combined functional diversity and abundance between unharvested controls and buffered sites implied that all buffers mitigated the impact of forestry on aquatic biodiversity in the short term.