Decreases in net primary production and net ecosystem production along a repeated-fires induced forest/grassland gradient

Fire is one of the most important disturbances in ecosystems. Fire rapidly releases stored carbon into atmosphere and also plays critical roles on soil properties, light and moisture regimes, and plant structures and communities. With the interventions of climate change and human activities, fire regimes become more severe and frequent. In many parts of world, forest fire regimes can be further altered by grass invasion because the invasive grasses create a positive feedback cycle through their rapid recovery after fires and their high flammability during dry periods and allow forests to be burned repeatedly in a relatively short time. For such invasive grass-fire cycle, a great change of native vegetation community can occur. In this study, we examined a C4 invasive grass () fire-induced forest/grassland gradient to quantify the changes of net primary production (NPP) and net ecosystem production (NEP) from an unburned forest to repeated fire grassland. Our results demonstrated negative effects of repeated fires on NPP and NEP. Within 4 years of the onset of repeated fires on the unburned forest, NPP declined by 14%, mainly due to the reduction in aboveground NPP but offset by increase of belowground NPP. Subsequent fires cumulatively caused reductions in both aboveground and belowground NPP. A total of 40% reduction in the long-term repeated fire induced grassland was found. Soil respiration rate were not significantly different along the forest/grassland gradient. Thus, a great reduction in NEP were shown in grassland, which shifted from 4.6 Mg C ha-1 yr-1 in unburnt forest to -2.6 Mg C ha-1 yr-1. Such great losses are critical within the context of forest carbon cycling and long-term sustainability. Forest management practices that can effectively reduce the likelihood of repeated fires and consequent likelihood of establishment of the grass fire cycle are essential for protecting the forest.