Over the past few centuries, novel human activity has reversed a 50-million-year-long global cooling trend and driven biodiversity losses worldwide. These climate and ecosystem changes began decades to centuries before the advent of long-term monitoring programs – meaning that most of our contemporary records come from systems that have already undergone significant alteration. To robustly evaluate the rate and magnitude of past ecological change, identify likely tipping points, and determine whether modern ecosystems have already moved towards novel ecological states, we need exceptional fossil records that have the temporal resolution to both capture fine-scale biological dynamics and reach many thousands of years into the past. In this seminar, I will showcase recent findings from an “invisible timescale” system, the Southern California Borderland Basins, where sediment cores with annual layers preserve a high-resolution fossil record can be directly tied to long-term modern observations to retroactively extend ecosystem monitoring prior to the onset of anthropogenic climate change. This dataset combines seafloor community composition and individual functional traits with paleoclimatic data, and spans a period of 35,000 years that includes the Last Glacial Maximum and the abrupt warming events of the last deglaciation. By placing the ecological impacts of contemporary anthropogenic change in the context of ecosystem changes since the last Ice Age, I find that the “Anthropocene” is truly unique compared to the geologic past. Beginning in the 19th century, seafloor ecosystems undergo a regime shift towards novel community states, with community composition being distinct from any recorded in the past 35,000 years. At the same time, benthic communities experience  reproductive life history changes, biomass declines, and population losses that accelerate in the mid-20th century. This ecological state shift may represent a “tipping point” triggered by land-use changes initiated during the Spanish and American colonization of California. Lessons from this “invisible timescale” system can be applied more broadly to better understand how to manage ecosystems for long-term resilience in the Anthropocene.