My dissertation weaves together paleo-ecological data, archival records, historical evidence, and modern inventories to understand the long-term impacts of fire disturbance on forest carbon dynamics in Six Rivers National Forest, California. Over three chapters, I use my results to evaluate the consequences of management activities, namely anthropogenic manipulation of fire, on aboveground live forest biomass. In the first chapter, I present a reconstruction of forest structure and composition during the 1880s and compare it to modern forests conditions. I found that modern forests are considerably denser and increasingly favor shade-tolerant taxa, likely due to twentieth century fire suppression. In the second chapter, I focus on developing an emerging fossil pollen analysis technique as the first step in quantitative reconstruction of past vegetation biomass. By modeling the relationship between modern pollen influx and modern biomass, I demonstrated that calibrated pollen influx-biomass relationships provide a robust means to infer changes in past plant biomass. My third chapter concludes with the application of the models I developed in the previous chapter to 3000 years of pollen influx data. These models allowed me to estimate aboveground biomass in a mixed conifer forest over the late Holocene. Taking a trans-disciplinary approach in this chapter, I integrate empirical datasets about biomass and fire with Native oral history to illustrate the significant and important role that Indigenous peoples played pre-contact in shaping the forest ecosystems of northwestern California. With this dissertation, I document forest dynamics at different spatial scales over three millennia and provide a practical benchmark for land managers by indicating the scale of intervention needed to move California forests closer to their long-term historical conditions.