LAUSR

Human‐induced environmental change is occurring globally and alters ecosystem properties both directly, by changing abiotic conditions, and indirectly, by modifying community composition. The effects of changes in plant composition in determining ecosystem properties may be determined by the identity of the plants, their respective biomass in the community and also by the ability of remaining vegetation to compensate for their loss. We used the graminoid removal treatment from a long‐term (12 year) removal experiment to examine the role of graminoids in determining ecosystem properties and also the potential for biomass compensation by remaining species in a grassland in northern Canada. We conducted removals in both fertilized and non‐fertilized environments to examine context‐dependency of graminoid effects. There was full biomass compensation for graminoid removal by the remaining functional groups (i.e., total biomass was equal to no‐removal controls) after 5 years, and after 12 years fertilized plots showed overcompensation for removals. After 12 years of graminoid removals, most ecosystem properties were not affected by removals in either fertilization environment. Similarly, there were few effects on microbial biomass or function with plant removal or fertilization treatments. Comparison to earlier published responses in this experiment shows a strengthening of the responses of the plant community to graminoid removals over time while effects on ecosystem properties diminished. Full biomass compensation for removal of graminoids occurred after 5 years, and after 12 years of removals, fertilized plots had overcompensated for the loss of graminoids. In contrast, the shorter term (4 years) responses of soil available nutrients and soil moisture to removals have mainly disappeared after eight further years, likely because of the biomass compensation by the plant community. Synthesis. Comparing the effects of graminoid removals on ecosystem structure and function after 4, 7, and now 10 and 12 years of removals, we now argue that, after over a decade of graminoid removal, recovery of plant biomass is required for the maintenance of most ecosystem properties and not functional group identity, as we concluded earlier. This highlights the importance of maintaining field experiments in the long term, particularly in northern ecosystems.