Impacts of nonnative grass invasion and precipitation variability on carbon partitioning in a Hawaiian tropical dry forest

Background/Question/Methods

Few studies have examined the impacts of nonnative invasions on ecosystem carbon (C) cycling, and no study has examined how invasions influence the partitioning of gross primary productivity (GPP), or how changes in partitioning as a result of invasion are impacted by climate variability. We used nonnative grass invasion of Pennisetum setaceum into native Hawaiian dry forests to examine how invasion impacts the partitioning of GPP (flux as a fraction of GPP) to aboveground (aboveground net primary productivity (ANPP) plus aboveground autotrophic respiration (R_A)) vs. belowground (total belowground carbon flux (TBCF)) across two years in which precipitation was normal (Year 1) and 60% higher than normal (Year 2). We hypothesized that grass invasion would shift partitioning from aboveground to belowground due to increased competition for belowground resources. In addition, we hypothesized that the shift in partitioning to belowground in invaded plots would decline in Year 2 as belowground resource availability increased. To address our hypotheses, we constructed complete stand C budgets in 400 m² replicated forest plots (n = 4) with P. setaceum (grass plots) and in areas where P. setaceum had been removed for 5 years (removal plots).  ANPP was estimated for trees and grass with traditional techniques, R_A was estimated using global relationships between ANPP and R_A, and TBCF was estimated using a mass balance approach.

Results/Conclusions

GPP varied by treatment and year (P < 0.05), ranging from 1.3 to 1.7 kg C m^-2 yr^-1 in removal plots in Year 1 and 2, respectively, and from 1.6 to 2.4 kg C m^-2 yr^-1 in grass plots in Year 1 and 2, respectively.  In support of our first hypothesis, invasion shifted partitioning from aboveground to belowground in Year 1.  In removal plots, 36% of GPP was partitioned to aboveground and 64% to belowground, while in grass plots 31% of GPP was partitioned to aboveground and 69% to belowground.  In the wetter Year 2, 38% of GPP in grass plots was partitioned to aboveground and 62% to belowground, in support of our second hypothesis that increased belowground resource availability would shift partitioning from belowground to aboveground.  Changes in GPP and its partitioning with invasion are particularly important in light of the ubiquitous presence of nonnative invasive species and the need for a better understanding of the role that they will play in global C cycling under a changing climate.