Perennial forb invasions of annual grassland vary in their enhancement of soil carbon flux and carbon pools

Background/Question/Methods

Soil CO₂ flux (R_soil) plays a major role in carbon cycling. It exceeds all other terrestrial-atmospheric carbon exchanges and is estimated to be an order of magnitude greater than the combined carbon emitted by fossil fuel combustion and deforestation. Soil organic carbon (SOC) pools are predicted to decline if carbon loss through R_soil is not offset, possibly resulting in elevated atmospheric carbon. Plant invasions have been shown to increase ecosystem carbon cycling. Despite the importance of changing atmospheric carbon, only a few studies have addressed potential effects of invasive plants on terrestrial carbon cycling and carbon pools. To our knowledge, this has not been investigated in annual grassland invaded by perennial forbs. We asked: 1) Do invasive forbs increase R_soil and/or reduce SOC pools? 2) Do invasive forbs produce more litter and/or have higher litter decomposition rates compared to a native graminoid? To address these questions, we set up 3-2x2 m² plots with a 5 m buffer zone enclosing areas with invasive forbs (Alternanthera philoxeroides, Solidago canadensis) and native graminoid (Eragrostis pilosa) monocultures. We measured SOC monthly and measured R_soil weekly. We measured litter decomposition rate in situusing litter bags.

Results/Conclusions

As we predicted, both R_soil and SOC pools were increased following invasion by either forb species. The increases in cumulative R_soil were smaller with goldenrod than with alligator weed (~45% vs. ~71%) but SOC increases were larger (~47% vs. ~16%). Compared to the native grass plots, each invasive forb species produced larger amounts (alligator weed ~155%, goldenrod ~360%) of more rapidly decomposing litter (k_{alligator weed} : k_Eragrostis ~3.6:1, k_goldenrod : k_Eragrostis ~1.9:1). Increases in R_soil and SOC with either invader suggest that there are consistent effects of high litter quantity and quality of perennial forbs relative to the native grass. However, the relative magnitudes of R_soil and SOC increases were sensitive to differences in litter quantity and quality that were negatively related for these two invasive forbs. The larger quantity of litter with goldenrod invasion was associated with relatively higher SOC pools, while higher litter quality associated with alligator weed invasions was associated with relatively higher R_soil. Functional groups of invaders and natives appear to be useful predictors of directional changes in R_soil and SOC but the magnitude of effects on soil carbon fluxes and pools seem sensitive to variation in traits of invaders.