Direct and indirect effects of herbicides on soil biota

Background/Question/Methods

Millions of gallons of herbicides are applied yearly to control invasive weeds, but their impact on non-target soil biota is little known.  We designed an experiment to assess the direct and indirect effects of the broadleaf-specific herbicide Tordon on arbuscular mycorrhizal fungi (AMF), a group of ubiquitous root symbionts that can increase plant nutrient acquisition and pathogen protection.  AMF abundance was quantified using neutral lipid fatty acids (NLFA). Potential effects on the biomass of bacteria, actinomycetes, and general fungi were also measured using phospholipid fatty acids (PLFA).  To assess direct effects of Tordon, we sprayed native bunchgrass communities because Tordon does not eradicate graminoids. To quantify indirect effects mediated by a shift in plant community composition, we sprayed monodominant stands of the broadleafed forb knapweed (Centaurea stoebe).  We predicted that AMF responses to Tordon would be largely indirect and that AMF abundance would decline in treated knapweed plots because knapweed is a good AMF host.  In addition, AMF has been shown to build and maintain water stable aggregates (WSA), and we measured WSA one and two years after herbicide applications to assess cascading effects on ecosystem function. 

Results/Conclusions

As expected, Tordon applications did not affect the composition of the bunchgrass communities, but knapweed was largely replaced by the short-lived exotic bulbous bluegrass (Poa bulbosa) one year after spraying.  This shift from a good host (knapweed) to a poor host (bulbous bluegrass) reduced the abundance of AMF in sprayed knapweed plots, whereas no difference was observed in the bunchgrass communities (P_{Plant community x Tordon} = 0.002). The abundance of bacteria, fungi and actinomycetes did not differ among treatments, suggesting that these groups are either not affected by herbicides or take longer to respond. WSA was 40% higher (p<0.001) in knapweed plots after one year, and Tordon reduced (P=0.03) WSA by 18% two years after spraying.  Thus, functional consequences of the reduced AMF abundance in the sprayed knapweed plots were temporally offset, reinforcing the importance of measuring treatment effects over appropriate (often longer) time scales.  Overall, our findings suggest that soil microbial groups respond differently to herbicides, that responses are largely indirect, and that shifts in biomass of non-target microbial groups can influence ecosystem processes such as water infiltration and soil stability.  A greater consideration of these changes may be key to restoration success.