Abstract
Ethnographic accounts document widespread use of low-intensity surface fires by California's Native Americans to manage terrestrial resources, yet the effects of such practices on forest composition and structure remain largely unknown. Although numerous paleoenvironmental studies debate whether proxy interpretations indicate climatic or anthropogenic drivers of landscape change, available data sources (e.g., pollen, charcoal) are generally insufficient to resolve anthropogenic impacts and do not allow for hypothesis testing. We use a modeling approach with LANDIS-II, a spatially explicit forest succession and disturbance model, to test whether the addition of Native American-set surface fires was necessary to approximate vegetation change as reconstructed from fossil pollen. We use an existing 1,600-year pollen and charcoal record from Holey Meadow, Sequoia National Forest, California, as the empirical data set to which we compared modeled results of climatic and anthropogenic fire regimes. We found that the addition of anthropogenic burning best approximated fossil pollen-reconstructed vegetation change, particularly during periods of prolonged cooler, wetter periods coinciding with greater regional Native American activity (1550-1050 and 750-100 cal yr BP). For lightning-caused wildfires to statistically approximate the pollen record required at least twenty times more ignitions and 870 percent more area burned annually during the Little Ice Age (750-100 cal yr BP) than observed during the modern period (AD 1985-2006), a level of natural fire increase we consider highly improbable. These results demonstrate that (1) anthropogenic burning was likely an important cause of pre-Columbian forest structure at the site and (2) dynamic landscape models provide a valuable method for testing hypotheses of paleoenvironmental change.