Lifecycle assessments suggest that production of lignocellulosic biofuels, especially from high-yielding biomass crops such as switchgrass can be achieved with net energy production and substantial greenhouse gas reduction (Farrell et al. 2006). For example, the study of Schmer et al. of 10 multi-hectare fields in the northern midwestern U. S., measured the on-farm energy balance for switchgrass production and then used the literature to estimate the energy balance for conversion to ethanol. The average results were a yield of 2800 L per ha, a net energy yield (energy produced-input) of 80 GJ/ha, and a green house gas displacement of ~80% for use of the bioethanol compared with gasoline (Schmer et al. 2008). This study assumed an ethanol yield of 0.38 L per kg of biomass, which is typical of what is commonly found in other assessments and of laboratory saccharification yields (Fu et al. 2011). Based on the percentage of switchgrass biomass that is sugar (Vogel et al. 2010, Fig. 1), this is ~65% of the theoretical yield of conversion of all polysaccharide to ethanol (0.58 L per kg).

As we have described, substantial progress has been made via a panoply of approaches to improve plant biomass to close the gap between the typical and theoretical saccharification yields. These improvements in yield now await translation from the laboratory into the field. Of course, such efforts require substantial time and money, not to mention regulatory approvals if transgenic methods are employed. Furthermore, we can expect different phenotypes in the field since the range of biotic and abiotic stress conditions under which published studies have been conducted is limited. This is especially important for biofuel crops because they need to be produced on degraded or abandoned crop lands that do not displace substantial food production in order to avoid indirect increases in greenhouse gas production due to land clearing (Fargione et al. 2008; Youngs et al. 2012). No doubt, an even more thorough understanding of cell wall biosynthesis and regulation will be necessary to anticipate and mitigate pleiotropic effects of manipulating the major components of plant biomass. In addition, since the vast majority of studies have not been conducted in switchgrass or other biofuel species, there remains substantial work to be done in testing genes in biofuel species. As thus far only single genes have been examined, when selecting genes for testing in a species such as switchgrass one wonders about whether additive or synergistic effects might be achieved from simultaneously manipulating multiple cell wall synthesis or regulatory pathways. Modeling and informatics studies will certainly facilitate the transfer of information from model species and selection of engineering targets in bioenergy crops (e. g., Ruprecht and Persson 2012).

On the bioprocessing side, the baseline 65% of theoretical ethanol yield is typically achieved with harsh, i. e., expensive, dilute acid pretreatment (0.5% at 180°C for 8 min) and cellulase loading of approximately 15 active units/g biomass (Fu et al. 2011). More efficient catalysts that can function with milder pretreatments at lower concentrations would also facilitate attainment of near maximal yields. Indeed, by combining optimized feedstocks with improved enzymes and bioprocessing methods, we may have already achieved complete and efficient saccharification. Again, the bottleneck seems to be in translating current progress to the industrial scale. As for work with plants, translation to industrial-scale microbiology requires substantial additional understanding and process tuning. Effective scale-up will be facilitated by continued accumulation of additional options in terms of enzymes, strains, and organisms, and understanding at a both detailed biochemical and systems-wide levels. Platforms that reduce the capital requirements of biofuel production will be especially helpful for establishing a second-generation biofuel industry.

Acknowledgements

Thanks to Dr. M. Peck and K. Zhao for helpful comments on the manuscript. This work was supported by the National Science Foundation EPSCoR program under Grant No. EPS-0814361. Any opinions, findings, conclusions, or recommendations expressed are those of the authors and do not necessarily reflect the views of the National Science Foundation.