Effects of plant functional group removal on CO₂ fluxes and belowground C stocks across contrasting ecosystems

Changes in plant communities can have large effects on ecosystem carbon (C) dynamics and long-term C stocks. However, how these effects are mediated by environmental context or vary among ecosystems is not well understood. To study this, we used a long-term plant removal experiment set up across 30 forested lake islands in northern Sweden that collectively represent a strong gradient of soil fertility and ecosystem productivity. We measured forest floor CO2 exchange and aboveground and belowground C stocks for a 22-yr experiment involving factorial removal of the two dominant functional groups of the boreal forest understory, namely ericaceous dwarf shrubs and feather mosses, on each of the 30 islands. We found that long-term shrub and moss removal increased forest floor net CO2 loss and decreased belowground C stocks consistently across the islands irrespective of their productivity or soil fertility. However, we did see context-dependent responses of respiration to shrub removals because removals only increased respiration on islands of intermediate productivity. Both CO2 exchange and C stocks responded more strongly to shrub removal than to moss removal. Shrub removal reduced gross primary productivity of the forest floor consistently across the island gradient, but it had no effect on respiration, which suggests that loss of belowground C caused by the removals was driven by reduced litter inputs. Across the island gradient, shrub removal consistently depleted C stocks in the soil organic horizon by 0.8 kg C/m2 . Our results show that the effect of plant functional group diversity on C dynamics can be relatively consistent across contrasting ecosystems that vary greatly in productivity and soil fertility. These findings underline the key role of understory vegetation in forest C cycling, and suggest that global change leading to changes in the relative abundance of both shrubs and mosses could impact on the capacity of boreal forests to store C.