Predicting the unpredictable: Assessing the invasion risk of bioenergy crops

Background/Question/Methods

Through the 2007 Energy Independence and Security Act the federal government has mandated production of 36 billion gallons of renewable fuel. The most promising crops for co-firing and liquid fuels will likely include a selection of non-native plant species.  Economically and sustainably competitive energy crops will need to be high-yielding, perennial species, which require minimal inputs despite their ability to grow on marginal land. These traits, which are highly desirable for successful cultivation, have caused concern for their potential to become invasive. 

The use of weed risk assessments has become a useful tool for predicting invasive potential in many parts of the world.  To evaluate several proposed bioenergy crops we used the standard Australian Weed Risk Assessment (A-WRA) and the Animal and Plant Health Inspection Service PPQ weed risk assessment (PPQ-WRA), which will serve as the new U.S. standard for evaluation of potentially invasive species.  Our evaluation compared 13 proposed bioenergy crops (eg, Miscanthus × giganteus) against 10 known weedy species specifically introduced for agronomic purposes (johnsongrass), as well as 14 non-weedy crops (eg, Zea mays). This study was designed to identify differences between the two WRAs as well as the invasion risk of candidate bioenergy crops under various assumptions (eg, genetically engineered for herbicide resistance).

Results/Conclusions

Both assessments suggest that many of the leading bioenergy candidates have the potential to become weedy in specific ecoregions of the United States.  Differences in the screening process are evident in the percentage of plants rejected by the Australian model (82%) in contrast to those receiving a high risk rating by the PPQ model (72%). However it should be noted, despite the similar percentage of high risk recommendation by the assessments, the species receiving acceptable or low risk ratings was not identical between the two.  For example, the A-WRA recommends rejecting Panicum virgatum, while the PPQ-WRA suggests further evaluation. Conversely the PPQ-WRA rejects Sorghum bicolor (grain), but is scored acceptable by the A-WRA. While both models benefit from objectivity, the PPQ-WRA includes the ability to include uncertainty and leaves management decisions to policy makers. The PPQ-WRA also allows hypothetical iterations, which can be run for plants not yet in the U.S. or taking into consideration genotypic variability.  Based on the outcomes of these assessments we can be better prepared for widespread commercialization of these crops, while planning ahead for improved stewardship in those that are predicted to be of high risk.