Analysis of twenty-five years of forest floor sampling on Watershed 6 (W6) at the Hubbard Brook Experimental Forest (Woodstock, NH, USA) shows no significant evidence of N accumulation in the forest floor (O horizon). The uncertainty in this estimate is high (2 ± 19 kg N ha^-1 y^-1) due to large spatial variation in forest floor mass, as well as interannual effects that can bias sampling depth. However, the C:N ratio of the forest floor has increased slightly (p = 0.05), which is unexpected under the “N saturation” hypothesis. Forest floor monitoring data from other mature hardwood forest stands in the region yields similar results (-2 kg N ha^-1 y^-1). A steady-state forest floor N stock has different ecosystem-scale implications than previous estimates of N accumulation (8 kg N ha^-1 y^-1) by Covington (1981).

This finding, along with long-term monitoring of other N fluxes and stock changes in W6, allows us to update the pioneering small-watershed ecosystem budget work of Bormann, Likens, and Melillo (1977). Much has changed in W6 since this budget was published. Stream losses of DIN have decreased 60%, while aboveground biomass has stopped aggrading. Previously unmeasured fluxes (e.g. dry deposition, DON input and outputs) help complete the budget. Together, these adjustments make little net change to the 1970’s era budget – there is still a “missing source” of N while the forest was aggrading. However, beginning in the 1990s, the budget instead has a missing sink. Our data suggests this sink is not forest floor accumulation but may instead be mineral soil accumulation or gaseous loss.