£% EPA/600/R-18/195F| June 2018 | www.epa.gov/research
United States
Environmental Protection
Agency
Biofuels and the Environment
Second Triennial Report to Congress
-------
Biofuels and the Environment
Second Triennial Report to Congress
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC
June 29,2018
-------
1. Table of Contents
List of Figures iv
List of Tables vi
Contributors vii
Executive Summary viii
1 Introduction 1
1.1 Organization of this Report 6
2 Drivers of Environmental Impacts 6
2.1 Biofuel Volumes 7
2.1.1 U.S. Biofuel Production 7
2.1.2 Biofuel Imports 8
2.1.3 Biofuel Exports 8
2.2 Feedstocks 9
2.2.1 Acreage 10
2.2.2 Total Production of Biofuel Feedstocks 11
2.2.3 End Use of Biofuel Feedstocks 11
2.2.4 Nutrients Applied 14
2.2.5 Pesticides Applied 16
2.2.6 Conservation Practices 18
2.3 Technologies 19
2.3.1 Biofuel Conversion Technologies 19
2.4 U.S. Land Use Change 20
2.4.1 Overview 20
2.4.2 Observed Land Use Change to Date 24
2.4.3 Economic-Based Projections of U.S. Land Use Change Impacts 43
2.4.4 Conclusions 43
2.5 International Land Use Change 45
2.5.1 Observed International Land Use Change 45
2.5.2 Economic-Model Based Estimates of Biofuel-Induced Land Use Change 49
2.5.3 Conclusions 52
3 Environmental and Resource Conservation Impacts 55
3.1 Air Quality 55
3.1.1 2011 Report Conclusions 55
3.1.2 Drivers of Impacts to Air Quality 56
3.1.3 Impacts to Air Quality 58
3.1.4 Potential for Future Changes in Impacts 62
3.1.5 Conclusions: Air Quality 64
3.1.6 Research Needs: Air Quality 65
3.2 Water Quality 65
3.2.1 2011 Report Conclusions 66
3.2.2 Drivers of Impacts to Water Quality 67
3.2.3 Impacts to Water Quality 68
3.2.4 Potential for Future Changes in Water Quality Impacts 73
li
-------
3.2.5 Conclusions: Water Quality 73
3.2.6 Research Needs: Water Quality 74
3.2.7 Opportunities for Future Environmental Improvements 74
3.3 Water Quantity 74
3.3.1 2011 Report Conclusions 74
3.3.2 Drivers of Impacts to Water Quantity 75
3.3.3 Changes in Relationships between Drivers and Impacts 78
3.3.4 Potential for Future Changes in Impacts to Water Quantity 81
3.3.5 Conclusions: Water Quantity 83
3.3.6 Research Needs: Water Quantity 83
3.3.7 Opportunities for Future Environmental Improvements 83
3.4 Ecosystem Health and Biodiversity 84
3.4.1 2011 Report Conclusions 84
3.4.2 Drivers of Impacts to Ecosystem Health and Biodiversity 85
3.4.3 Impacts to Ecosystem Health and Biodiversity 87
3.4.4 Key Points from Recent Literature 91
3.4.5 Potential for Future Changes in Impacts to Ecosystem Health and Biodiversity 91
3.4.6 Conclusions: Ecosystem Health and Biodiversity 92
3.4.7 Opportunities for Future Environmental Improvements: Ecosystem Health and
Biodiversity 92
3.4.8 Research Needs: Ecosystem Health and Biodiversity 93
3.5 Soil Quality 93
3.5.1 2011 Report Conclusions 93
3.5.2 Drivers of Impacts to Soil Quality 94
3.5.3 Changes in Relationships Between Drivers and Impacts 96
3.5.4 Potential for Future Changes in Impacts to Soil Quality 97
3.5.5 Conclusions: Soil Quality 97
3.5.6 Opportunities for Future Environmental Improvements: Soil Quality 98
3.5.7 Research Needs: Soil Quality 98
3.6 Invasive Species 98
3.6.1 2011 Report Conclusions 98
3.6.2 Drivers of Impacts to Invasive Species 99
3.6.3 Potential Changes in Relationships Between Drivers and Impacts 100
3.6.4 Potential for Future Changes in Impacts to Invasive Species 101
3.6.5 Conclusions: Invasive Species 101
3.6.6 Research Needs: Invasive Species 102
3.7 International Impacts 102
3.7.1 2011 Report Conclusions 102
3.7.2 Drivers of International Impacts 103
3.7.3 Changes in Drivers of International Impacts 107
3.7.4 Potential for Future Changes in International Impacts 108
3.7.5 Conclusions: International Impacts 108
3.7.6 Research Needs: International Impacts 109
in
-------
4 Conclusions and Recommendations 110
4.1 Overarching Conclusions 110
4.2 Specific Conclusions Ill
4.2.1 Land Use Change Ill
4.2.2 Air Quality 112
4.2.3 Water Quality 113
4.2.4 Water Quantity 113
4.2.5 Ecosystem Health and Biodiversity 113
4.2.6 Soil Quality 114
4.2.7 Invasive Species 114
4.2.8 International Impacts 115
4.3 Opportunities for Future Environmental Improvements 115
4.4 Limitations 116
4.5 Research Needs 116
4.6 Recommendations 118
5 References 120
Appendix A: Abbreviations and Glossary 136
Appendix B: Key Terms from Major Land Use Change Studies 140
USD A Major Uses of Land in the US, 2012 (MLU) 140
USDA 2012 Census of Agriculture 141
USDA 2012 National Resources Inventory 142
USGS U.S. Conterminous Wall-to-Wall Anthropogenic Land Use Trends, 1974-2012
(MY ALT) 143
FAOSTAT Land Use Data 143
List of Figures
Figure 1 Annual U.S. biofuel production, 2000-2016 7
Figure 2 Annual biofuel volumes imported into the U.S., 2000-2016 9
Figure 3 Annual biofuel volumes exported from the U.S., 2000-2016 10
Figure 4 Total U.S. annual planted acres of corn and soybeans, 1996-2017 11
Figure 5 Total annual U.S. corn and soybean production volumes, 2000-2016 12
Figure 6 Annual volumes of U.S. corn used for fuel and other purposes, 2000-2016 12
Figure 7 Annual volumes of U.S. soybeans used for fuel and other purposes, 2000-2016 14
Figure 8 Total volumes of fertilizers used for U.S. corn production (A) and rates of application
(B), 2000-2016 15
Figure 9 Total volumes of fertilizers used for U.S. soybean production (A) and rates of
application (B), 2000-2015 16
Figure 10. Changes through time (1945-2012) in cropland used for crops by MLU region 27
iv
-------
Figure 11. Changes through time (1963-2012) in principal crops harvested for the 48 contiguous
States 28
Figure 12. Changes in cultivated, noncultivated, and total cropland from 1982-2012 from the
NRI 30
Figure 13. Net changes in all major land cover/use categories between 2007 and 2012 in the NRI.
31
Figure 14 Difference in cropland (class 43, top) and pasture/hay (class 44, bottom) between 2012
and 2002 by county from the NWALT 33
Figure 15 Net rates (a) of land use change from non-cropland to cropland between 2008 and
2012 36
Figure 16 Most common 'break-out' crop by region from Lark et al. (2015) 37
Figure 17 Crop rotation patterns for the 9-country area of eastern Iowa in Ren et al. (2016) for
2002-2007 (a) and 2007-2012 (b) 42
Figure 18 Global land use change by aggregate region 46
Figure 19 Global total factor productivity 47
Figure 20 Summary land use change GHG emissions estimates for corn ethanol 50
Figure 21 USGS mapper tool showing total nitrogen concentration trends between 2002 and
2012 69
Figure 22 Changes in the measured size of the GoM hypoxic zone as related to the amount of
nitrate-nitrate loading 72
Figure 23 An estimate of the blue, green and grey water footprint associated with corn grain,
stover, wheat straw and soybean during the crop growing phase 77
Figure 24 Acres of irrigated land in 2012, based on the USDA Farm and Ranch Irrigation Survey.
80
Figure 25 Relative conversion rates of arable non-cropland to cropland (2008-2012), including
conversion located along the Ogallala aquifer 81
Figure 26 Relative conversion rates to cropland of (a) grassland, (b) forest, (c) shrubland, and (d)
wetland from 2008 to 2012 86
Figure 27 Map of (A) wild bee status and (B) status of wild bee supply vs. demand for pollination
services across coterminous U.S 88
Figure 28 Trends in annual metric tons of U.S. exports of corn and brewers' and distillers' dregs
and waste (DDGS) 105
Figure 29 Trends in annual metric tons of U.S. exports of soybeans 106
v
-------
Figure 30 Trends in annual U.S. ethanol exports...
Figure 31 Trends in annual U.S. biodiesel exports
106
107
List of Tables
Table 1. Estimates of total applied glyphosate and atrazine (in million pounds) by different data
sources 17
Table 2 Comparison of different simulation studies summarized in Wallander et al. 2011 23
Table 3. Major land uses (in millions of acres) from the MLU (Bigelow et al. 2017) 26
Table 4 Comparison of major national studies on land use change, harmonized to the degree
possible 39
Table 5. Methods and metrics used to assess the water quantity impacts of biofuels 78
Table 6. Annual U.S. ethanol imports by country of origin (million gallons)1 103
Table 7. Annual U.S. biodiesel imports by country of origin (million gallons)1 104
Table 8. Annual U.S. renewable diesel imports by country of origin (million gallons)1 105
VI
-------
Contributors
Report Leads
Paul Argyropoulos1
Julia Burch
Lead Authors
Randy Bruins1
Dallas Burkholder
Christopher Clark
Rich Cook
Rebecca Dodder
Anne Grambsch1
Contributing Authors
Ellen Cooter
Kathleen Fahey
Mark Johnson
Aaron Levy
Anne Grambsch1
C. Andrew Miller
Stephen D. LeDuc
Joe McDonald
Briana Niblick
C. Andrew Miller
Scott Unger2
Dilip Venugopal3
Mike Madden
Robert Sabo
Jay R. Reichman
Joseph Schubauer-Berigan
1 Retired from EPA.
2 Now at Maricopa County, Arizona
3 Now at the University of Maryland
vii
-------
Executive Summary
Background
This report is the second of the U.S. Environmental Protection Agency's (EPA's) triennial
reports to Congress required under the 2007 Energy Independence and Security Act (EISA). EISA
Section 204 calls for EPA to report to Congress on the environmental and resource conservation impacts
of the Renewable Fuel Standard (RFS) program, specifically air and water quality, water quantity,
ecosystem health and biodiversity, soil quality, invasive species, and international environmental impacts
(hereafter referred to as the Section 204 statutory impacts).
Consistent with how EPA conducted the first Section 204 report, EPA has chosen in this
assessment to focus on the Section 204 statutory impacts and not to expand the scope of the report beyond
the factors explicitly enumerated in the law. As a result, some environmental impacts are not discussed in
this report. Lifecycle greenhouse gas emissions impacts of biofuel use, for example, are addressed as part
of the RFS program and are not included in this report. Furthermore, this report does not include a
comparative assessment of the impact of biofuels on the environment relative to the impacts of other
transportation fuels or energy sources, including fossil fuels, for every environmental endpoint. For
example, the environmental impacts of growing corn, refining ethanol from that corn, and burning that
ethanol in an internal combustion engine has a different environmental impact than drilling for oil,
refining gasoline, and burning that in an internal combustion engine. EPA recognizes that a fully
comprehensive assessment of the benefits and impacts of biofuel production and use would be broader
than what is represented by this report, but conducting such an evaluation is beyond the scope of this
study.
This report updates the findings of the first Report to Congress, published in 2011, with respect to
environmental and resource conservation impacts, and, together, the two reports are intended to address
the Section 204 statutory impacts since the passage of the EISA. The primary conclusions of the 2011
Report included the following two findings: (1) the environmental and resource conservation impacts of
biofuel production and use as delineated in Section 204 of EISA were, on balance, negative; and (2)
EISA's goals could be achieved with relatively minimal adverse environmental impacts if existing
conservation and best management practices were widely employed, concurrent with advances in
technologies that facilitate the use of second-generation feedstocks. The 2018 Report reaffirms the
findings of the 2011 Report and reflects the current understanding about biofuel production using data
gathered through May 2017. The 2018 Report also reviews data on U.S. land use and the scientific
literature through April 2017.
viii
-------
Major Findings
• Data from observations made since the 2011 Report indicate that the biofuel production and use
conditions that led to the conclusions of that report have not materially changed.
• Substantial volumes of cellulosic and advanced biofuels have not been produced as anticipated by
EISA. The Section 204 statutory impacts anticipated as a consequence of large-scale use of
feedstocks other than corn and soybeans have therefore not occurred.
• Corn grain and soybeans remain by far the dominant feedstocks for biofuel production. Biofuel
production associated with large-scale cultivation of corn and soybeans contributes to the adverse
environmental and resource conservation impacts of the type listed in EISA Section 204, though
we caution that this report does not evaluate the net effects of displacing petroleum or other
energy sources with biofuels.
• There has been an observed increase in acreage planted with soybeans and corn between the
decade leading up to enactment of EISA and the decade following enactment. Evidence from
observations of land use change suggests that some of this increase in acreage and crop use is a
consequence of increased biofuel production mandates.
• It is likely that the Section 204 impacts associated with land use change are, at least in part, due to
increased biofuel production and use associated with the RFS. However, at this time we cannot
quantify with precision the amount of land with increased intensity of cultivation nor confidently
estimate the portion of crop land expansion that is due to the market for biofuels.
Likely Future Impacts
Section 204 of EISA also requires that the triennial report identify likely future impacts. We
interpret the requirement to address "likely futures" as encompassing near-term future impacts presuming
current technologies and rates of market penetration, and current policy and market dynamics. Thus,
where this report discusses likely future impacts, it is addressing anticipated changes over the next three
to five years. This report finds that there are some indications of growth in cellulosic ethanol production,
primarily from corn stover, but that large-scale production at levels approaching the original EISA targets
is not likely to be reached in the next few years. Available data suggest that current trends using corn
starch and soybeans as primary biofuel feedstocks, with associated environmental and resource
conservation impacts, will continue in the near term.
IX
-------
Overall Conclusions
Reports and data published since the first Report to Congress have increased the confidence in the
conclusions of that report (Table ES-1).
Table ES-1. Comparison of overarching conclusions from first and second reports to Congress.
Conclusions from the first Report to
Congress
Conclusions from second Report to Congress
Evidence to date from the scientific literature
suggests that current enviromnental impacts from
increased biofuels production and use associated with
EISA 2007 are negative but limited in magnitude.
Disregarding any effects that biofuels have on
displacing other sources of transportation energy, evidence
since 2011 indicates the specific enviromnental impacts
listed in EISA Section 204 are negative. The enviromnental
and resource conservation impacts, whether positive or
negative, related to displacement of other transportation
energy sources by biofuels were not assessed.
Published scientific literature suggests a
potential for both positive and negative enviromnental
effects in the future.
Literature published since 2011 supports the
conclusion of the potential for positive and negative effects.
Available information suggests, without accounting for the
enviromnental effects of displacing other sources of
transportation energy, the specific enviromnental impacts
listed in EISA Section 204 are negative in comparison to
the period prior to enactment of EISA.
EISA goals for biofuels production can be
achieved with minimal enviromnental impacts if
existing conservation and best management practices
are widely employed, concurrent with advances in
technologies that facilitate the use of second-
generation feedstocks.
Evidence continues to support the conclusion that biofuel
production and use could be achieved with reduced
enviromnental impacts. The majority of biofuels continue
to be produced from corn grain and soybeans, with
associated impacts that are well understood. Cellulosic and
other feedstocks remain a minimal contributor to total
biofuel production.
Specific Conclusions
Land use change
Evidence since enactment of EISA suggests an increase in acreage planted with soybeans
and corn, with strong indications from observed changes in land use that some of this
increase is a consequence of increased biofuel production.
x
-------
Since the first Report to Congress there have been several advances in our understanding of land
use change in the United States. Land use change has been identified as one of the primary drivers of
environmental impacts from an expanding biofuels industry. However, the connections between land use
change due to biofuels and environmental effects have not been evaluated sufficiently to allow
quantification specifically attributable to biofuel production. There are strong indications that biofuel
feedstock production is responsible for some of the observed changes in land used for agriculture since
enactment of EISA. However, we cannot quantify with precision the amount of land with increased
intensity of cultivation nor confidently estimate the portion of crop land expansion associated with the
market for biofuels.
Air Quality
• The emission impacts of biofuel production and distribution, and offsetting indirect impacts
on petroleum fuel production and distribution, are important to consider along with end-
use impacts for volatile organic compounds (VOCs), particulate matter (PM), and oxides of
nitrogen (NOx); emission and air quality impacts associated with feedstock production and
conversion of feedstock to biofuels are highly localized.
Emissions of NOx, SOx, CO, VOCs, NH3, and particulate matter can be impacted at each stage of
biofuel production, distribution, and usage. These impacts depend on feedstock type, land use change,
and feedstock production practices. Ethanol from corn grain has higher emissions across the life-cycle
than ethanol from other feedstocks, and ethanol facilities relying on coal have higher air pollutant
emissions than facilities relying on natural gas and other energy sources. The magnitude, timing, and
location of emissions changes can have complex effects on the atmospheric concentrations of criteria
pollutants (e.g., O3 and PM2 5) and air toxics, the deposition of these compounds, and subsequent impacts
on human and ecosystem health. Only limited data exist on the impacts of biofuels on the tailpipe and
evaporative emissions of Tier 3 light-duty vehicles and light-duty vehicles using advanced gasoline
engine technologies to meet GHG emissions standards. Comprehensive studies of the impacts of biofuels
on the emissions from advanced light-duty vehicle technologies, similar in scope to previous studies of
such impacts on Tier 2 vehicles, would improve understanding.
Water quality
Demand for biofuel feedstocks may contribute to harmful algal blooms, as recently
observed in western Lake Erie, and to hypoxia, as observed in the northern Gulf of Mexico.
Changes to future nitrogen and phosphorus loadings will depend on feedstock mix and crop
management practices.
xi
-------
The increased intensity of corn production on land already under cultivation and the expansion of
corn and soybean cultivation onto grasslands negatively impact water quality but have not been
consistently quantified to date. Differences in nutrient application, management practices, and runoff
characteristics make direct connections between increased feedstock production and water quality impacts
difficult to quantify and assess. Empirical studies suggest water quality impacts but the magnitude of
these changes is variable across the landscape and may be detectable only in some regions. Recent
modeling studies conclude that row crop agriculture plays an important role in driving downstream
impacts such as harmful algal blooms, particularly in fresh waters, and hypoxia, particularly in coastal
waters, and suggest that biofuel feedstock production is a contributing factor. Continued adoption and
expansion of sustainable conservation practices are expected to decrease nutrient loadings and associated
adverse impacts.
Water quantity
There are some indications of increased water use due to increases in irrigated areas for
corn and elevated land conversion rates in more arid Western states. Adverse water
availability impacts will most likely arise in already-stressed aquifers and surface
watersheds. Irrigation practices are dependent on a number of economic and agronomic
factors that drive land management practices making attribution of increased irrigation
and water quantity to biofuels difficult.
Quantitative evaluations are needed to understand increases in water use through changes in land
use and/or land management change, to understand whether those changes can be directly or indirectly
attributed to feedstock production for biofuels, and to determine whether increases in water demands and
water stress have occurred or are occurring along water-stressed areas or "hot spots" (e.g., Ogallala
aquifer) where high water demands and water stress are coinciding.
Ecosystem health and biodiversity
The conversion of environmentally-sensitive land to cropland consistent with increased
production of current biofuel feedstocks is associated with negative impacts to ecosystem
health and biodiversity
Loss of grasslands and wetlands are occurring in ecologically sensitive areas, including the
Prairie Pothole Region. Row crop expansion is resulting in the loss of habitat and landscape
simplification. Increasing pesticide use for feedstock production is associated with negative impacts to
pollinators, birds, soil-dwelling organisms, and other ecosystem services both in terrestrial and aquatic
Xll
-------
habitats. Increased fertilizer applications of N for corn and of P for corn and soybean have known
negative effects on aquatic biodiversity. Opportunities exist for continued adoption and expansion of
practices and technologies that will enhance ecosystem services and sustainable feedstock production.
Soil quality
Conversion of grasslands to annual production of the dominant biofuel feedstocks typically
adversely affects soil quality, with increases in erosion and the loss of soil nutrients and soil
organic matter, including soil carbon.
Impacts of this conversion can be partially mitigated - though not entirely - through the adoption
of management practices such as conservation tillage. Corn stover, a cellulosic biofuel feedstock, is now
being harvested at the commercial scale in Iowa, and the scientific literature suggests this must be done
carefully to avoid negatively affecting soil quality and crop yields.
Invasive Species
Current biofuel feedstocks pose little risk of becoming invasive species. Cultivation of
herbicide-tolerant feedstock crops (e.g., glyphosate-tolerant soybean) and concomitant
application of the associated herbicide (e.g., glyphosate) has the potential to contribute to
herbicide-resistant weed development, just as herbicide-tolerant crops grown for other
purposes.
International
U.S. ethanol imports have decreased while biodiesel and renewable diesel imports have
increased, leading to potential land use change impacts in countries of origin. Exports of
corn, DDGS, soybeans, and ethanol primarily increased.
Reports indicate that demands for biofuel feedstocks have led to market-mediated land use
impacts (both direct and indirect land use changes) in the past decade. Quantification and causal
attribution of land use change and international environmental impacts due to biofuel production and
renewable fuel standards remains uncertain. Comprehensive causal analysis frameworks and coordinated
frameworks for evaluating land use changes across biofuel trading nations may help our understanding of
international land use change and environmental impacts.
Recommendations
To promote actions to address impacts, EPA recommends the following:
Xlll
-------
Additional research in coordination with other organizations (e.g., federal agencies, states,
trade organizations) is recommended to better characterize land use change due to changes
in biofuel feedstock production.
Efforts at the federal level, as described by the Biomass Research and Development Board,
to improve efficiencies and sustainability of processes across the biofuel supply chain should
be continued and strengthened where possible.
An ecosystem approach is recommended to evaluate environmental and natural resource
impacts of biofuel production. Such an approach provides an integrative perspective that
accounts for complex interactions of multiple stressors across different locations.
Incorporating local information and perspectives will improve understanding of changes at
local scales, which will enhance opportunities for improved information and will enable
targeted responses to prevent and mitigate adverse impacts of biofuel production and use.
Best management practices should be encouraged, incentivized, and otherwise expanded to
promote conservation and sustainability in agricultural systems.
xiv
-------
1 Introduction
In December 2007, Congress enacted Public Law 110-140, the Energy Independence and
Security Act (EISA), with the stated goals of providing "greater energy independence and security [and]
to increase the production of clean renewable fuels." In accordance with these goals, EISA revised the
Renewable Fuel Standard (RFS) program, created under the 2005 Energy Policy Act1 and managed by
the U.S. Environmental Protection Agency (EPA), to increase the volume of renewable fuel required to
be blended into transportation fuel from 9 billion gallons per year in 2008 to 36 billion gallons per year
by 2022.
The revised statutory provisions and implementing regulations (commonly known as the RFS2
program) specify increasing applicable volumes of cellulosic biofuel, biomass-based diesel, advanced
biofuel, and total renewable fuel that EPA is directed to use (unless it establishes lower volume
requirements using specified waiver authorities) in establishing annual percentage standards for these
renewable fuel categories in transportation fuel. The purpose of this report is to examine the
environmental and resource conservation impacts of the RFS2 program, as required under EISA Section
204.
EISA Section 204 calls for EPA to report to Congress every three years on the environmental
and resource conservation impacts of increased biofuel production and use as stated in the relevant text
of the Act:
In General. Not later than 3 years after the enactment of this section and every 3 years
thereafter, the Administrator of the Environmental Protection Agency, in consultation with the Secretary
of Agriculture and the Secretary of Energy, shall assess and report to Congress on the impacts to date
and likely future impacts of the requirements of Section 211(o) of the Clean Air Act2 on the following:
1. Environmental issues, including air quality, effects on hypoxia, pesticides3, sediment,
nutrient and pathogen levels in waters, acreage and function of waters, and soil
environmental quality.
1 The 2005 Energy Policy Act amended the Clean Air Act and established the first national
renewable fuel standards. The statute specifies the total volume of renewable fuel that is to be used based
on the volume of gasoline sold in the United States.
2 EISA 2007 amended Section 21 l(o) of the Clean Air Act to include the definitions and
requirements of RFS2.
3 Pesticides include antimicrobials, fungicides, herbicides, nematicides, insecticides, and
rodenticides.
1
-------
2. Resource conservation issues, including soil conservation, water availability, and ecosystem
health and biodiversity, including impacts on forests, grasslands, and wetlands.
3. The growth and use of cultivated invasive or noxious plants and their impacts on the
environment and agriculture.
4. The report shall include the annual volume of imported renewable fuels and feedstocks for
renewable fuels, and the environmental impacts outside the United States ofproducing such
fuels and feedstocks. The report required by this subsection shall include recommendations
for actions to address any adverse impacts found.
The first report to Congress was completed in 2011 (hereafter the 2011 Report) and provided an
assessment of the environmental and resource conservation impacts associated with increased biofuel
production and use (EPA 2011). Although many impacts had been speculated or anticipated by the July
2010 publication cutoff date for inclusion in the 2011 Report, few had been actually observed in the peer
reviewed literature. Thus, the first report was largely forward-looking and evaluated the potential
impacts of several assumed future scenarios that were common in the literature. The overarching
conclusions of the 2011 Report were: (1) the environmental impacts of increased biofuel production and
use were likely negative but limited in impact; (2) there was a potential for both positive and negative
impacts in the future; and (3) EISA goals for biofuels production could be achieved with minimal
environmental impacts if best practices were used and if technologies advanced to facilitate the use of
second-generation biofuel feedstocks (corn stover, perennial grasses, woody biomass, algae, and waste).
This is the second report on the current and potential future environmental impacts associated
with the requirements of Section 21 l(o) of the Clean Air Act. This report updates the findings of the
2011 Report with respect to EISA Section 204 statutory impacts, provides recommendations to address
adverse impacts, and reflects the current understanding concerning biofuel production using data
gathered through the RFS program and other federal databases through May 2017. We also reviewed
U.S. data on land use and peer-reviewed scientific literature through April 2017, focusing on observed
changes as opposed to projected changes in impacts associated with changes in feedstocks, fuel types,
and volumes. This report focuses on the Section 204 statutory impacts since passage of the EISA in
2007. Where appropriate, the report provides additional information over longer time frames to provide
context for the discussion of impacts.
EPA identified a number of studies during the review process that were published after the April
2017 cut-off date or that were not peer-reviewed. Where their findings were sufficient to require changes
to this report's conclusions or recommendations, such changes were made, if they demonstrated the
2
-------
required data quality. We did not conduct original quantitative analyses for this report. A qualitative
review was considered necessary to meet the publication schedule for this report, and we anticipate that
the recommendations from this report will guide research, including possible quantitative analyses, in
preparation for the next triennial report.
In establishing the scope of this report, EPA chose to adhere closely to the enumerated
requirements of EISA Section 204. Therefore, this report addresses only the Section 204 statutory
impacts associated with implementation of the RFS2 program. This report does not attempt to
quantitatively evaluate either the current or potential future benefits associated with the program. It does,
however, point to possible opportunities for future improvements related to the Section 204 statutory
impacts. It is important to note the distinction between a "likely future" as prescribed in Section 204 of
EISA and a "potential future." We interpret likely futures to encompass near-term future impacts
presuming current technologies and rates of market penetration, and current policy and market dynamics.
Thus, where this report discusses likely future impacts, it is addressing anticipated changes over the next
three to five years. Where this report discusses potential future impacts, it is recognizing the possibilities
for changes over the longer term that can affect the environmental and natural resource impacts
associated with biofuels.
An exploration of the potential longer-term future benefits associated with use of biomass for
energy and other products is found in the 2016 Billion-Ton Report developed by the U.S. Department of
Energy (DOE) (DOE 2016). Operating with greater analytical freedom than indicated by EISA's Section
204, DOE developed scenarios of biomass production and use that far exceed current levels. The DOE
report provides a potentially useful complement to this report's focus on observed changes in recent
years. In addition, the U.S. Department of Agriculture (USDA) supports the sustainable production of
high-quality, nonfood feedstocks for conversion into bioproducts, bioenergy, bioheat, and biopower.
These efforts support the broader federal Bioeconomy Initiative "...to develop and coordinate innovative
approaches to expanding the sustainable use of America's abundant biomass resources, while
maximizing economic, social, and environmental benefits."4
Consistent with the 2011 Report and EISA Section 204, this report does not evaluate emissions
of carbon dioxide (CO2) or other greenhouse gases (GHGs), nor does it review and assess studies that
analyze GHG impacts in its conclusions (see Box 1, "Greenhouse Gas Emissions and Impacts" for
4 The Billion Ton Bioeconomy Initiative: Challenges and Opportunities; Biomass Research &
Development Board, Washington, DC, 2016. Available at:
https://biomassboard.gov/pdfs/the_bioeconomy_initiative.pdf
3
-------
details). Emissions of GHGs over the life cycle of biofuel production, conversion, and use are addressed
under the RFS program. This report focuses on the Section 204 statutory impacts and therefore does not
attempt to make detailed comparisons to the estimated impacts associated with use of other
transportation fuels or energy sources. EPA acknowledges that a lack of comparative assessments for
every environmental endpoint is a limitation on the ability of this report to draw conclusions regarding
the comprehensive environmental impacts of biofuels, but we believe that the information provided
nonetheless provides value by reviewing observed impacts specifically from biofuel production and use.
This report emphasizes U.S. impacts; however, the substantial market created for biofuels by the
U.S., Brazil, and other countries has important global implications. For example, countries that produce
Box 1. Greenhouse Gas Emissions and Impacts
A key feature of EISA is the establishment of mandatory life cycle GHG reduction thresholds for the
renewable fuels that are intended by the law to displace petroleum based fuels. EPA used state-of-the-
art models, data, and other information to project the GHG impacts of biofuels, as described in the
RFS2 Final Regulatory Impact Analysis (RIA) (EPA 2010). EPA conducted a formal, independent
peer review of key components of the analysis. The modeling of GHG emissions in the RFS2 RIA
provides a reasonable and scientifically sound basis for making threshold determinations and
estimating GHG impacts. As EPA conducts lifecycle assessments for new fuel pathways, the most
recent science and data are incorporated wherever possible (see https://www.epa.gov/renewable-fuel-
standard-program/fuel-pathwavs-under-renewable-fuel-standard). For example, EPA has updated its
analyses to reflect new data on fuel conversion efficiencies, forest carbon stocks, projected crop
yields, and agricultural inputs. The GHG impacts associated with biofuel production and use remains
an area of active research, and EPA continues to evaluate the relevant science to inform consideration
of the need for any reevaluation of previous determinations of life cycle GHG emissions. Other
agencies and institutions also evaluate life cycle GHG emissions, providing information for
comparative purposes. The Greenhouse gases, Regulated Emissions, and Energy use in
Transportation (GREET) spreadsheet analysis tool developed by Argonne National Laboratories
(Burnham et al. 2006) is an example of such a tool. As discussed above, this report does not attempt
an evaluation of emissions of carbon dioxide or other GHGs, nor does it attempt to encompass GHG
impacts in its conclusions. Instead, this report provides information on other impacts that is
complementary to the GHG impacts described in the RIA (EPA 2010), which should be consulted for
more information on this topic.
4
-------
(or will produce) feedstocks that are converted to biofuels that qualify for use in the U.S. will experience
direct impacts; other countries (including the U.S.) will have to adapt to changing agricultural
commodity distributions that result from diversion of food exports to biofuel production. While there
may be economic or other benefits to such market changes, this report focuses on the environmental and
natural resource impacts of increased feedstock and biofuel production in other countries as a result of
U.S. policy, as required and defined under EISA Section 204.
The information included in both the first and second biofuels Reports to Congress is considered
foundational for future efforts to quantitatively compare the potential environmental impacts of
alternative scenarios for meeting the goals of the RFS2 program. They serve as a starting point for future
assessments, especially for the next triennial assessment, and for taking action to achieve the goals of
EISA. Future reports will reflect the evolving understanding of biofuel impacts in light of new research
results and data as they become available.
Box 2. Second Generation Biofuels
The requirement in EISA Section 204 to assess "the likely future impacts" of the Renewable
Fuel Standard means that second-generation biofuels and feedstocks must be considered in this report.
It is clear from the fuel volumes specified in EISA that Congress anticipated the production of
substantial volumes of cellulosic ethanol and other second-generation fuels. Although such levels of
production have not yet been reached, the requirement to assess the likely future impacts remains.
Because these fuels have not yet reached large-scale commercial production when compared
to other fuel types, there are limited observational data to illustrate the environmental and resource
conservation impacts related to production of those fuels. Thus, this report must rely upon model-
based projections to provide any assessment of impacts. There is potential that the projected impacts
will be more serious than anticipated, given that large-scale commercial production will likely create
incentives for increased use of production methods and chemical inputs that can reduce the anticipated
environmental benefits of grasses and other second-generation feedstocks.
Although second-generation biofuels have not yet demonstrated an economic benefit over
current feedstock-fuel combinations, the promise of improved environmental performance at
economically acceptable costs continues to encourage development.
5
-------
1.1 Organization of this Report
Chapter 2 provides information on the drivers of environmental issues, including biofuel
volumes, feedstocks, conversion technologies, agricultural practices, and U.S. land use changes. Chapter
3 focuses on the implications of biofuel production for environmental and natural resource issues and
includes a summary of impacts to date. Chapter 4 presents overarching conclusions and provides
recommendations for improving scientific understanding, as well as practices for minimizing
environmental impacts. Chapter 5 describes a path forward for future reports, including options for the
scope of the next triennial report based on the findings of this report and reported advances in the
science.
2 Drivers of Environmental Impacts
Numerous factors influence the markets for biofuels and the associated environmental impacts
of their production and use. These factors, which are "drivers" of the Section 204 statutory impacts,
include: regional considerations; scale and volume of commercial biofuel operations; development of
biofuel conversion processes; changes in vehicle technologies; and changes in agricultural practices due
to biofuel production and implications for environmental impacts. Each of these, whether individually or
in combination, will affect the ultimate environmental impacts associated with biofuel production and
use. Land use change is both a driver of environmental impacts and an environmental impact directly
affected by other market and non-market drivers discussed above.
As noted in the 2011 Report, many potential Section 204 statutory impacts of biofuel production
and use (e.g., water quality, water quantity, biodiversity) are a result of land use conversion and the
subsequent management of that land. Management includes tillage practices, nutrient application, and
other chemical inputs during feedstock production. Air quality impacts depend largely on the volume of
biofuels used, their impact on vehicle air pollutant emission rates, and the emissions associated with their
production and distribution.
As EPA does lifecycle assessments for new fuel pathways, the most recent science and data are
incorporated where possible. For example, EPA has updated its analyses to reflect new data on fuel
conversion efficiencies, forest carbon stocks, projected crop yields, and agricultural inputs. The GHG
impacts associated with biofuel production and use remains an area of active research, and EPA
continues to evaluate relevant science to inform consideration of the need for any reevaluation of
previous GHG determinations.
6
-------
This chapter presents information on these key drivers: biofuel production, feedstock production,
vehicle type and use, conversion technologies, and land use change.
2.1 Biofuel Volumes
2.1.1 U.S. Biofuel Production
Since 2012 the production of biofuels in the U.S. has grown steadily, rising from 14.1 billion
gallons in 2012 to 16.6 billion gallons in 2016 (see Figure 1). As in 2012, ethanol and biodiesel remain
the types of biofuels produced and consumed in the largest quantities in the U.S. However, in recent
years the production of other biofuels, such as renewable diesel and biogas used as transportation fuel,
have increased.
20,
15,
16,
14,
12,
cn
c
O
03
ID 10
c
o
CO
00
00
00
00
00
00
00
00
00
00
I Ethanol
I Biodiesel
I Renewable Diesel
Biogas
I Other
I
<$?
HP3
rCV
r\r
'V
$ $
'V3 rY
s>
4*
.'V
r\r
Jo
Figure 1 Annual U.S. biofuel production. 2000-2016.5
After a rapid rise in U.S. ethanol production from 2007 to 2011, more recent U.S. ethanol
production has increased relatively slowly, from 13.22 billion gallons in 2012 to 15.33 billion gallons in
5 Data for ethanol and biodiesel from USDA ERS US Bioenergy Statistics
(https://www.ers.usda.gov/data-products/us-bioenergy-statistics/); ethanol data available in Table 2, and
biodiesel data available in Table 4. Renewable diesel, biogas, and other data from EPA's public data for
7
-------
2016. This slower rate of growth is likely due to challenges associated with the E10 blendwall.6'7 Since
2013, nearly all gasoline sold in the U.S. has contained at least 10% ethanol. To further expand the
ethanol market in the U.S. would require greater sales of fuel blends that contain higher levels of
ethanol, such as E15 or E85. To date, sales of such fuels have been limited. If transportation fuel
consumption in the U.S. declines in future years, as projected by the U.S. Energy Information
Administration (EIA), demand for ethanol will likely also decline unless sales of E15 and/or E85
increase to offset the lower consumption of E10. In this section we have presented data from 2000-2016
where available, but have focused our discussion on the years since 2007.
U.S. production of biodiesel has increased fairly steadily since 2007, with temporary declines in
2009 and 2010. In 2016, biodiesel production in the U.S. reached a record high of 1.56 billion gallons.
Demand for biodiesel has likely been driven by increasing volume requirements under the RFS as well
as national and state-level incentives and requirements.
2.1.2 Biofuel Imports
Since 2012, the volume of biofuels imported into the U.S. has grown, rising from 666 million
gallons in 2012 to 1.07 billion gallons in 2016 (see Figure 2). Over the same time period, the types of
biofuels imported into the U.S. have changed significantly. Prior to 2012, ethanol was the predominant
biofuel type imported into the U.S. However, ethanol imports decreased significantly starting in 2013.
At the same time rising RFS standards, along with national and state-level incentives and requirements,
have resulted in a significant increase in the volume of biodiesel and renewable diesel8 imported into the
U.S.
2.1.3 Biofuel Exports
After reaching a high of 1.27 billion gallons in 2011, biofuel exports decreased to 870 million
gallons in 2012 before rising steadily to 1.18 billion gallons in 2016 (see Figure 3). Ethanol exports
increased significantly in 2011 and have remained high since, as ethanol production capacity in the U.S.
the Renewable Fuel Standard (https://www.epa.gov/fuels-registration-reporting-and-compliance-
help/public-data-renewable-fuel-standard).
6 E10 is a gasoline blend with 9% to 10% ethanol content; E15 is a gasoline blend with >10% to
15% ethanol content; and E85 is a gasoline blend with 51% to 83% ethanol content.
7 The blendwall for El0 refers to the point at which all gasoline in the US is blended with 10
volume percent ethanol, at which point the ability to consume additional ethanol through blending in
gasoline is challenged by limitations on the existing vehicle fleet and market to go to higher blend
concentrations.
8 Biodiesel is a renewable fuel produced through transesterification of organically derived oils
and fats. Renewable diesel is derived from biomass, generally using a thermal depolymerization process.
8
-------
1.00
0.90
0.80
0.70
to
c
_o
0.60
~ru
ID
0.50
c
o
=
0.40
co
0.30
0.20
0.10
I Ethanol
I Biodiesel
I Renewable Diesel
Other
II
^ # <§>* # ^ <§& $ <§£ $ <$? <$?
Figure 2 Annual biofuel volumes imported into the U.S., 2000-2016.5
has exceeded the ability to consume ethanol due to the E10 blendwall. Biodiesel exports have been low
since 2010, as the RFS program has provided a significant incentive for the U.S. consumption of
biofuels, especially non-ethanol biofuels that are not subject to the challenges associated with the E10
blendwall. From 2012-2016 ethanol exports ranged from a low of 620 million gallons in 2013 to a high
of 1.05 billion gallons in 2016. From 2012-2016 biodiesel exports ranged from a low of 80 million
gallons in 2014 to a high of 200 million gallons in 2013. A small volume of renewable diesel exports (40
million gallons) were reported for the first time in 2016.
2.2 Feedstocks
The primary planted crops used as biofuel feedstocks in the U.S. are corn and soybeans. This
section will therefore focus on the planted acres, total production, end uses, and management practices
for these two crops. In this section we again present data from 2000-2016 where available, but have
focused our discussion on the years since enactment of the EISA.
9
-------
1.40
1.20
1.00
to
c
J 0.80
03
g 0.60
in
0.40
0.20
0.00
I Ethanol
I Biodiesel
I Renewable Diesel
Other
I
^ <$> & QV> ^
Figure 3 Annual biofuel volumes exported from the U.S., 2000-2016.5
2.2.1 Acreage
The number of planted corn acres fluctuated considerably between 2006 and 2016. After an average of
roughly 80 million acres between 2000 and 2007, planted acres of corn increased to roughly 90 million
acres between 2007-2016 (see Figure 4), with higher variability in the 2007-2016 period than before
2007. A modest general increase in soybean acreage is also evident over the period 2000-2016,
averaging between 70-75 million acres between 2000 and 2006 and increasing to 82-83 million acres
from 2014-2016. Total corn and soybean acres planted in 2007 were significantly different than in
preceding years, with high corn and low soybean acreages planted because of U.S. and international
market and climatic factors.
10
-------
120
M Corn
Soybeans
100
80
ISl
1—
u
<
o
40
20
0
o^rMm'^-Lnior^.ooaio^rMm'^-LniD
O O O O O O O O O C) tH T—I T—I T—I T—I T-I T—I
ooooooooooooooooo
rvJrsirNirsirNi(NrNirsirsirsirsirsirsirsirsirsi(N
Figure 4 Total U.S. annual planted acres of corn and soybeans, 1996-20179
2.2.2 Total Production of Biofuel Feedstocks
The total production of corn and soybeans has increased over time since enactment of EISA in
2007. From 2007-2016 corn production ranged from a low of 10.8 billion bushels in 2012 to a high of
15.1 billion bushels in 2016 (see Figure 5). From 2007-2016 soybean production ranged from a low of
2.7 billion bushels in 2007 to a high of 4.3 billion bushels in 2016. Productivity for both corn and
soybeans was unusually low in 2012 due to drought conditions in many areas of the U.S.
2.2.3 End Use of Biofuel Feedstocks
Corn used for ethanol production has increased since enactment of EISA, but remained
relatively steady from 2010 through 2016, accounting for a low of 4.64 billion bushels of corn in 2012
and a high of 5.21 billion bushels of corn in 2016 (see Figure 6). Corn used for ethanol production as a
9 https://quickstats.nass.usda.gov/
11
-------
16.00
14.00
12.00
10.00
& 8.00
c
o
m 6.00
I Corn
l Soybeans
4.00
2.00
0.00
^ ^ <$5° q\° qV* ^ ^
VTTTTTTTTTVTTTVTt
Figure 5 Total animal U.S. corn and soybean production volumes, 2000-2016.s
16.00
14.00
12.00
10.00
_£=
CO
8.00
c
o
co 6.00
4.00
2.00
0.00
i Ethanol
I Feed
i Exports
Other (Food, Industrial, etc/
OrHfNm^LniDr^oocTiO<-ifNrr)'^Lnta
OOOOOOOOOOrHrH tH t—I tH t—I t—I
ooooooooooooooooo
(NrsirsirNjfNrMrNJrNirMrNirsirsirNjrsifNrNirNi
Figure 6 Annual volumes of U.S. corn used forfuel and other purposes, 2000-2016.
10
10 Data from USD A ERS US Bioenergy Statistics (https://www.ers.usda.gov/data-products/us-
bioenergy-statistics/); U.S. com use data available in Bioenergy Statistics Table 5.
12
-------
percentage of overall corn production increased from 19% in 2007 (the first commodity market year
prior to enactment of EISA in December 2007) to between 38% and 42% between 2011 and 2016. Corn
used for ethanol production as a percentage of overall corn production was relatively stable from 2012-
2016, with a high of 42% in 2013 and a low of 38% from 2014-2016. Corn used for feed generally
decreased from 2007 through 2011 and then generally increased from 2012-2016, with a low of 4.52
billion bushels of corn used for feed in 2012 and a high of 5.28 billion bushels used for feed in 2015.
These numbers do not account for feed sourced from distillers grains, an important co-product of corn
ethanol production. Approximately 32% of each bushel of corn used for ethanol production
(approximately 12% of the total corn production from 2014-2016) is returned to the feed market in the
form of distillers grains.11 Therefore, these numbers may overstate the quantity of corn used for ethanol
production and under-represent the quantity of corn used for feed by approximately 32%. Corn exports
generally decreased from 2007-2011 but increased from 2012-2016. From 2012-2016 corn exports
ranged from 1.54 billion bushels in 2012 to 1.90 billion bushels in 2016. All uses of corn decreased in
2013 following the relatively low corn production in 2012 caused by a drought.
The use of soybeans in all sectors increased from 2007 to 2016. Soybeans used for biodiesel
production increased from 0.40 billion bushels in 2012 to 0.53 billion bushels in 2016 (see Figure 7).
Soybeans used for biodiesel production as a percentage of overall soybean production increased from
9% in 2007 to 13% in 2011. Soybeans used for biodiesel production as a percentage of overall soybean
production was relatively stable from 2012-2016, with a high of 13% in 2012 and 2016, and a low of
11% in 2014. These percentages are estimated based on the total soybean crush and the percentage of
U.S. soy oil used to produce biodiesel in each year.12 These percentages likely overstate the percentage
of the soybean crop used for biodiesel production by approximately 80%, as only the soybean oil (which
is approximately 20% of the soybean by weight) is used for biodiesel production, while the non-oil
components of the soybean are generally used in the feed market. Soybean crush (for non-biodiesel uses)
increased from 1.69 billion bushels in 2012 to 1.94 billion bushels in 2016. Soybean exports increased
from 1.33 billion bushels in 2012 to 2.03 billion bushels in 2016. Soybeans used for seed and feed
increased from 0.09 billion bushels in 2012 to 0.13 billion bushels in 2016.
11 A bushel of corn weighs approximately 56 pounds. On average, 18 pounds of distillers grains
are produced for every bushel of corn used to produce ethanol.
12 U.S. soybean use for biodiesel production is estimated by multiplying the total soybean crush
by the percentage of U.S. soybean oil used for biodiesel production in each year.
13
-------
4.50
4.00
3.50
3.00
00
"§ 2.50
CO
o 2.00
I Crush (Biodiesel Use)
I Crush (Other Uses)
I Exports
Seed/Feed
co
1.50
1.00
0.50
0.00
cr
rV
->
nV
'& n<$> J? ^
f\V rsV
Figure 7 Annual volumes of U.S. soybeans used for fuel and other purposes, 2000-2016.13
2.2.4 Nutrients Applied
Nutrient usage for corn production has generally increased since 2000 in total amount applied
per year and the rate of application per unit area, albeit with a notable decline in 2002 (see Figure 8).
Total nutrients in the form of nitrogen applied increased from 9.75 billion pounds in 2000 to 12.2 billion
pounds in 2016 (Figure 8A). This was mainly due to increased acreage and the increase in average rate
applied per year from 136 pounds per acre to 145 pounds per acre over the same period (Figure 8B).
Phosphate usage increased from 3.1 to 4.2 billion pounds, and potash usage increased from 3.8 to 4.5
billion pounds between 2000 and 2016 (Figure 8A). Sulfur use increased from 0.13 to 0.5 billion pounds
during 2005-2016 (Figure 8A).14
Nutrient usage for soybean production has increased since 2000 in total amount applied per year but
remained somewhat stable in the rate of application per unit area (see Figure 9). Total nitrogen supplied
increased from 0.32 to 0.38 billion pounds from 2000 to 2015. Potash applied increased from 1.4 billion
13 Data for Soybean use from USDA ERS Oil Crops Yearbook (https://www.ers.usda.gov/data-
products/oil-crops-vearbook/oil-crops-vearbook/). Soybean use for biodiesel was estimated using data on
total soybean crush and the share of soybean oil seed for biodiesel from USDA ERS US Bioenergy
Statistics Table 6 (https://www.ers.usda.gov/data-products/us-bioenergy-statistics/).
14 USDA, National Agricultural Statistics Service Quick Stats, https://quickstats.nass.usda.gov/.
14
-------
pounds in 2000 to 2.5 billion pounds in 2015 mainly due to increased acreage and the increase in average
rate applied per year from 76 to 83 pounds per acre. Similarly, phosphate usage increased from 0.82 to
1.56 billion pounds from 2000 to 2015 along with the average applied per year from 48 to 51 pounds per
acre.
25
J| 20
tz
0
1 15
"O
aj
CL
Q.
nj
ID
£
0
A
350
d 300
£ 250
B 200
ro
r> IjU
ru
-------
application of nitrogen fell by about 28% in terms of pounds per bushel. Application of phosphate and
potash per bushel both increased, by 22% for phosphate and 15% for potash.13
Soy: Nutrients applied
Nitrogen ¦ Phosphate ¦ Potash ¦ Sulfur
IA ^
A
c
.2 3
-a 2
"o.
Q.
"S
u
nj
B
I
250
2000 2001 2002 2004 2006 2012 2015
Year
Nitrogen ¦ Phosphate ¦ Potash ¦ Sulfur
.200
t/>
-Q
.1150
TO
U
^100
a
to
cc
0
2000 2001 2002 2004
Year
2006 2012 2015
Figure 9 Total volumes of fertilizers used for U.S. soybean production (A) and rates
of application (B), 2000-2015.
2.2.5 Pesticides Applied
Pesticide usage data is relevant for understanding potential risks of environmental impacts,
including water quality and biodiversity, and to a lesser extent the estimation of water usage. Pesticides
16
-------
for biofuel feedstocks (corn and soybeans) include herbicides, fungicides, insecticides, and nematicides.
The pesticides are applied as foliar sprays, direct applications to soil (pre-plant or post-plant), seed
treatments, or in the case of Bt,15 incorporated through genetic engineering. Pesticides are also used on
stored grains, including fumigants and insecticides applied to grain bins; such applications are not likely
to have the same potential for environmental impacts as other approaches.
Yearly data on pesticide usage in corn and soybeans since 2007 are not available for all
pesticides. For those pesticides for which data are available, the estimated usage for production of corn
and soybeans varies among different sources (see Table 1 for example). All these data sources, however,
documented the increasing trend in usage of the herbicide glyphosate over the past 10 years (Benbrook
2016). Based on data from the USDA's Agricultural Chemical Use Program, a total of 82.3 million
pounds of glyphosate (in different forms) was applied to corn during 2016, and 106.9 million pounds to
soybeans during 2015. For insecticides, the usage of neonicotinoids, which are applied as seed
treatments, is not captured by USDA (Douglas et al. 2015b). Approximately 90% of corn (Douglas et al.
2015b) and 30% of soybean fields planted during 2008-2012 contained neonicotinoid seed treatments.16
Table 1. Estimates of total applied glyphosate and atrazine (in million pounds) by different data sources.
Herbicide Crop USDA Agricultural USGS Pesticide Benbrook (2016)
Chemical Use National Synthesis
Survey3 Project13
Glyphosate
Corn (2014)
61.4
92.2
68.9
(different forms)
Soybeans
109.3
115.9
113.9
(2012)
Atrazine
Corn (2014)
45.2
61.3
NA
a. USDA National Agricultural Statistics Service: Agricultural Chemical Use - Corn 2016.
https://quickstats.nass.usda.gov/.
b. USGS National Water Quality Assessment Program: Pesticide National Synthesis Project. State-
level pesticide use estimates by major crop and crop groups.
https://water.usgs.gov/nawqa/pnsp/usage/maps/countv-level/.
15 Bacillus thuringiensis (Bt) is a naturally occurring soil bacterium that produces proteins active
against certain insects. Beginning in the mid-1990s, crop plants expressing Bt genes were
commercialized in the United States.
16 Benefits of Neonicotinoid Seed Treatments to Soybean Production (2014). U.S. EPA
memorandum, Office of Chemical Safety and Pollution Prevention.
https://www.epa.gov/sites/production/files/2Q14-
10/documents/benefits of neonicotinoid seed treatments to soybean production 2.pdf.
17
-------
During 2007-2011 neonicotinoid usage increased from 1.2 million to 2.1 million pounds for corn, and
0.26 million to 1.3 million pounds for soybeans (Douglas et al. 2015b).
Through genetic engineering, herbicide-tolerant (HT) corn and soybeans were developed to
survive application of specific herbicides targeting weeds. Similarly, insect-resistant Bt corn containing
the gene from the soil bacterium Bacillus thuringiensis expresses insecticidal Cry proteins. During 2007-
2016, the percentage of planted acres of genetically engineered corn steadily increased from 73% to 92%
for Bt only, HT only, and "stacked" Bt/HT varieties ("stacked" varieties have both types of traits, and in
some cases, multiple Bt and HT traits). The percentage of planted acres of corn with HT or stacked traits
increased from 52% to 89%, and from 49% to 79% for Bt varieties, including stacked traits. The
percentage of planted HT soybean varieties has remained around 94% since 2007.17 Reports indicate that
herbicide usage in HT corn and soybean increased relative to non-HT, whereas less insecticide (in kg/ha)
was applied in Bt corn relative to non-Bt corn (NAS 2016; Perry et al. 2016).
2.2.6 Conservation Practices
Agricultural conservation practices can reduce the impacts of feedstock production and appear to
be increasing in prevalence. USDA data show increased use of conservation buffers on 6.2% of planted
corn acres in 2001 to 11% in 2010, although soil erosion controls remained roughly the same, with
17.8% of planted corn acres using such controls in 2001 and 18.0% in 2010. Precision agriculture18 and
variable rate technology (VRT), both of which can improve the efficiency of chemical treatments,
increased considerably over the same time period. Precision agriculture was applied on 37% of planted
corn acres in 2001 and on 72% in 2010. Likewise, VRT for fertilizer application increased from use on
8% of planted corn acres in 2001 to 19% in 2010.19 The most recent USDA data on these practices is
from 2010, so it is uncertain how the extent of these practices has changed since then.
17 U.S. Department of Agriculture: Adoption of genetically engineered crops in the United
States, by trait, 2000-2017
https://www.ers.usda.gov/webdocs/charts/55237/biotechcorn_d.html?v=42565
18 Precision agriculture is a set of technologies, methods, and information that are applied at a
local scale to improve production efficiencies related to outcomes including yield, application of
chemical treatments, and irrigation. See USDA's 2007 publication, Precision Agriculture: NRCS
Support for Emerging Technologies,
https://www.nrcs.usda.gov/Internet/FSE DOCUMENTS/stelprdb 1043474.pdf
19 U.S. Department of Agriculture, Economic Research Service: Agricultural Resource
Management Survey Farm Financial and Crop Production Practices, https://www.ers .usda. gov/data-
products/arms-farm-financial-and-crop-production-practices/.
18
-------
2.3 Technologies
2.3.1 Biofuel Conversion Technologies
The primary technologies used to produce biofuels remain much the same as in 2011: (1) the
fermentation of corn starch to produce ethanol; and (2) the conversion of virgin vegetable and other
biogenic oils (including waste fats, oils, and greases) to produce fatty acid methyl esters (biodiesel).
Since 2011, the use of feedstocks other than virgin vegetable oils, such as waste oils and corn oil
produced at ethanol production facilities, has increased significantly. The production of renewable
diesel, produced by hydrotreating vegetable or other biogenic oils, has increased since 2011, with several
new large-scale production facilities coming online. In 2014, EPA determined that compressed natural
gas (CNG) and liquefied natural gas (LNG) derived from biogas qualified as a cellulosic biofuel.20 Since
this determination, the use of CNG/LNG derived from biogas as transportation fuel has increased
significantly, and these fuels now account for the majority of the cellulosic biofuel in EPA's RFS
program.21
Significant investments have been made by government, universities, and private parties to
develop the technologies necessary to economically convert cellulosic biomass to transportation fuel at
commercial scale, including both ethanol and hydrocarbon fuels. Several large-scale cellulosic ethanol
plants have been constructed, although production from these facilities remains very limited.22 Recently
new technologies have been developed that enable the conversion of the cellulosic portions of the corn
kernel (corn kernel fiber) to ethanol at existing corn ethanol production facilities. While the expected
production volume of cellulosic ethanol from corn kernel fiber at any individual facility is relatively
small (generally less than 5% of the volume of ethanol produced from starch), if widely adopted, this
technology could be used to produce significant volumes of cellulosic ethanol. Other technologies being
developed to convert cellulosic biomass to hydrocarbon fuels have faced, and continue to face,
challenges associated with relatively high capital costs of these facilities coupled with other market and
20 79 FR 42128
21 Production of cellulosic biofuel RINs for CNG/LNG derived from biogas increased from
approximately 32.6 million RINs in 2014 to 188.6 million RINs in 2016. In 2016 CNG/LNG derived
from biogas accounted for 97.8% of all cellulosic RINs (D3 and D7) generated. All data are from EPA's
public website: https://www.epa.gov/fuels-registration-reporting-and-compliance-help/public-data-
renewable-fuel-standard
22 While the production of cellulosic biofuel has increased significantly in recent years, the vast
majority of this fuel has been CNG/LNG derived from biogas. According to EPA data, total production
of cellulosic ethanol in 2017 was approximately 10 million gallons. See https://www.epa.gov/fuels-
rcgistration-rcporting-and-compliancc-hclp/2017-rcncwablc-fucl-standard-data.
19
-------
policy uncertainties. Several companies are focusing on the production of bio-crudes (a synthetic liquid
produced from cellulosic biomass that generally has a high oxygen content relative to petroleum based
crudes) that may be able to be processed in traditional refineries. Such technologies, if successfully
commercialized, could require lower capital investment if they are able to be utilized in traditional
refineries to upgrade these bio-crudes to finished drop-in transportation fuel. These drop-in fuels could
lower the need and cost currently required to deploy new expensive supply and distribution
infrastructure, and they also have the added benefit of being compatible with existing engines and
vehicles.
2.4 U.S. Land Use Change
2.4.1 Overview
Land use change has been identified as one of the primary drivers of potential environmental
impacts from an expanding biofuels industry (EPA 2011). Land use is commonly distinguished from
land cover, in that land cover strictly describes the physical cover of the land surface (e.g., grassland),
while land use involves human activity and reflects human decisions about how land will be used (e.g.,
grassland used for grazing livestock) (Campbell 1996; Nickerson et al. 2015). The term land use is often
also used to describe how the land is used for a particular purpose (e.g., agriculture), which can include
many land management practices (e.g., fertilizer application, irrigation). Here we use the term land use
change generally to describe changes in any of these processes (i.e., land cover, land use, land
management) that can affect how land is used and managed.23
Land use affects most environmental end points considered in this report, including runoff from
agricultural lands, emissions of criteria air pollutants, and habitat acreage and quality for various plant
and animal species. Increased agricultural production can come from two distinct processes,
23 Several publications and organizations use slightly different terms and acronyms to describe
generally similar processes (e.g., LULCC for land use/land cover-change in the Intergovernmental Panel
on Climate Change Fifth Assessment Report (Ciais et al. 2013); LULUC for land-use-land-use-change in
the EPA's Greenhouse Gas Inventory). The USDA National Resources Inventory defines land cover as
"the vegetation or other kind of material that covers the land surface," and land use as "the purpose of
human activity on the land; it is usually, but not always, related to land cover." DOE's Billion Ton Study
defines land use change as "Modification of the human actions of using land, or human purposes of land
(e.g., zoning), or human management of natural resources, or benefits derived from natural resources.
Note: Almost anything humans do, or dictate, or refrain from doing, that impacts land and related natural
resources, could be considered LUC" (DOE 2016). Our use of the term land use change is intended to be
general to encompass all these processes of how the land is used and the physical cover is affected to
meet that use. It is not the purpose of this report to reconcile varying definitions in the literature.
20
-------
extensiflcation (i.e., the expansion of agricultural land onto previously uncultivated land) and
intensification (i.e., increased production from the land without an increase in acreage). Intensification
often occurs from changes in land management and agronomic practices, including double cropping,
irrigation, seed improvements, and changes in fertilizer or other chemical inputs. Although land use
change is commonly associated with agricultural extensiflcation and intensification, it can also lead to a
reduction in total crop acreage (e.g., through cropland abandonment) or a decrease in management
intensity (e.g., through replacement of chemical fertilizer with manure inputs).
A second common distinction of land use change types is direct and indirect land use change. In
the context of biofuels, direct land use change is the land use change that occurs to support the
cultivation of feedstocks specifically for biofuel production. Indirect land use change occurs from the
diversion of crops to the biofuel market, which results in an unmet market demand for agricultural
products that then induces land use change to meet that demand (Searchinger et al. 2008; EPA 2010).
These terms are often used in the context of greenhouse gas emissions from biofuels production and use
but are relevant for all environmental end points.
There are many methods used to assess or forecast changes in land use, including empirical
observations (e.g., based on remote sensing or plot sampling), surveys (e.g., the USDA Census of
Agriculture), and dynamic models (e.g., FASOM-GHG, FAPRI-CARD, GTAP, POLYSYS24). This
section focuses on domestic U.S. land use change that has been assessed either through empirical
observations or surveys of respondents. These are often conducted using comprehensive land use
categories or representative statistical samples, rather than focusing on one particular economic sector or
region. For example, the corn and soy trends from the USDA NASS data in section 2.2 describe
increases in both of these crops, but without a comprehensive land classification assessment it is
impossible to know whether these increases came from existing agricultural lands or new lands that were
not recently in cultivation. A strength of this approach is greater confidence in the amounts and types of
land use change actually occurring. However, there are still uncertainties and challenges with comparing
different empirical observations, including differences in definitions, methods, and scope (Nickerson et
al. 2015). A weakness of empirical approaches is the difficulty of confidently attributing the causes of
land use change. There are many potentially contributing market (e.g., crop prices, transportation costs)
24 FASOM-GHG is the Forest and Agricultural Sector Optimization Model Greenhouse Gas
Version developed by Texas A&M and Oregon State University. FAPRI-CARD is the agricultural
model developed by the Food and Agricultural Policy Research Institute (FAPRI) and the Center for
Agricultural and Rural Development (CARD). GTAP refers to the international trade model developed
by the Global Trade Analysis Project at Purdue University. POLYSYS is the Policy Analysis System
developed by the University of Tennessee and the USDA ERS to simulate the U.S. agricultural sector.
21
-------
and nonmarket (e.g., climate, pests) factors that influence land use changes that could be coincident with
the passage of EISA and therefore correlated in an empirical analysis. Attributing all of the observed
trends to biofuels is not appropriate, and thus methods of causal analysis for biofuels has emerged as an
active area of research [e.g., Efroymson et al. (2016)]. Some empirical studies have attempted to assess
attribution to biofuels through proxy data, such as the proximity to a biorefinery (Brown et al. 2014;
Motamed et al. 2016; Wright et al. 2017) or surveys of farmers (Wallander et al. 2011; Gray et al. 2013).
Challenges with quantifying attribution are summarized in Box 3.
Dynamic agricultural models are simplified representations of complex agronomic, economic,
social, and biophysical systems. These often include a reference scenario (e.g., without the RFS) and a
focal scenario (e.g., with the RFS) to isolate the simulated effect of a given policy (Koponen et al. 2018).
A strength of this approach is the ability to look at "what if scenarios" to isolate the effects from a given
policy or scenario. Weaknesses of this approach include: (1) many models make different projections
about the same subject; (2) it is difficult to objectively assess model skill at projecting future
unobservable states; and (3) because the models are simplifications of real systems, they often lack many
details known to influence the system. As an example for biofuels, in 2011 USDA summarized six major
modeling efforts available at the time (Wallander et al. 2011) and found wide ranges in predicted
increases in cropland (0.7-8.1 million acres) and corn acres (1.8-19.4 million acres) (see Table 2). These
limitations are not unique to modeling biofuels or agricultural systems more broadly and are addressed in
other areas of study (e.g., climate change research) by using averages or other statistics derived from
multiple models, using historical data to assess and improve on performance, or other measures. These
approaches are still in their relative infancy in the area of biofuel simulation modeling. Nevertheless,
dynamic models are useful tools for assessing system behavior to better understand sensitivities to
changes in key parameters and scenarios of interest.
Section 204 of EISA requires that the triennial report assess "impacts to date and likely future
impacts." For impacts to date we rely on the empirical record. For likely future impacts, we assume that
the trends to date are a reasonable estimate of trends over the short-term future (e.g., less than 3 to 5
years or the interval between Section 204 Reports). As discussed in the introduction, we minimize our
discussion of likely impacts further into the future because of inherent uncertainties in such projections,
but we include some examples for illustrations.
There is a large body of research that uses dynamic models mentioned above and other methods
to assess potential future impacts from bioenergy and biofuel production [e.g., (Souza et al. (2015); Dale
et al. (2016); Emery et al. (2016); DOE (2017))]. It is important to note the distinction between a "likely
22
-------
future" as prescribed in Section 204 of EISA, and a "potential future." We interpret likely futures to
encompass future effects under conditions of current policy and market dynamics. Potential futures are
much more broad and may implicitly assume potential changes in current policy, technological advances
not yet observed, changes in land ownership and decision making processes, and/or market dynamics
that do not reflect current conditions. Such research, although less relevant for assessing "likely futures"
under the requirements of Section 204, are still valuable tools for understanding the complex agro-
economic system and are helpful for decision makers who are designing public policy. One recent
notable example is the 2016 DOE Billion Ton Study (DOE 2016; DOE 2017), which highlighted these
limitations in the disclaimer25, and used POLYSYS to generate several potential future scenarios of land
Table 2 Comparison of different simulation studies summarized in Wallander et al. 2011 (source
material from Searchinger et al. 2008, Malcolm et al. 2009, and EPA 2010).
Study
Searchinger et al.
Searchinger et al.
Malcolm et
EPA (2010) RFS2
EPA (2010) RFS2
(2008)
(2008)
al. (2009)
RIA(FASOM)
RIA (FAPRI-CARD)
Year modeled
2016/2017
2016/2017
2015
2022
2022
Billion gallons
2.7 corn-based
Increase in
ethanol
(from 12.3 to
14.77
8.08
1.7
2.7 corn-based
15.00)
(from 14.75 to
(from 14.75 to
(from 13.30
(from 12.3 to 15.00)
plus small change in
29.52)
22.84)
to 15.00)
plus 13.5 cellulosic
imported ethanol
Predicted change in land-use/cropping selection
Million acres
Predicted
increase in
corn acres
19.4
10.0
3.2
3.6
1.8
Predicted
increase in
cropland
5.5
2.9
4.9
8.1
0.7
Other major
predicted
increases
Soybeans
(1.9)
Switchgrass (12.5)
Wheat (-2.9)
Soybeans (-1.4)
Major predicted
decreases
Soybeans (-9.6)
Wheat (-4.8)
Soybeans (-4.1)
Wheat (-3.3)
Rice and
sorghum
(each -0.1)
Barley (-1.2)
Rice and hay
(each -0.8)
Soybeans (-0.7)
Oats and cotton
(each -0.2)
25 The disclaimer in volume 2 of the 2016 Billion-Ton Report states: "BT16 volume 2 is not a
prediction of environmental effects of growing the bioeconomy, but rather, it evaluates specifically
defined biomass-production scenarios to help researchers, industry, and other decision makers identify
possible benefits, challenges, and research needs related to increasing biomass production. Users should
refer to the chapters and associated information on the Bioenergy Knowledge Discovery Framework
(bioenergykdf.net/billionton) to understand the assumptions and uncertainties of the analyses presented."
(DOE 2017)
23
-------
use change and biomass/biofuel production based on assumptions of minimal cropland extensification,26
which were then run through a suite of models or other approaches to assess environmental effects to air,
land, and water resources.
The following sections detail the land use change observed to date, briefly discuss future
projections of land use change, and the conclusions of this work. We focus on extensification because
bringing new lands into cultivation can have a larger environmental impact per unit area than
intensification (EPA 2011). However, some forms of intensification can also have significant effects and
will be discussed. Other land use changes focused on land management operations (e.g., tillage practices,
tile drainage, fertilizer) are addressed elsewhere in this report.27
2.4.2 Observed Land Use Change to Date
The 2011 Report found that quantifiable land use change had not been reported as of the
Report's publication and thus reviewed potential environmental impacts of different land use change
scenarios common in the literature (EPA 2011). It concluded that land use change would likely drive
most environmental effects aside from air quality and that the most plausible land use change scenario
for corn and soybeans was for conventionally managed corn to replace no-till soybean or other row crops
and soybeans to maintain a stable acreage (EPA 2011).
2.4.2.1 National trends in major land uses and cropland extensification
Since the 2011 Report there have been many important studies on land use change trends in the
U.S. Five major national efforts have been published: (1) the USDA's Major Uses of Land in the United
States, 2012 (termed "Major Land Uses" series, MLU) (Bigelow et al. 2017); (2) the USDA 2012
Census of Agriculture ("Census") (USDA 2014); (3) the USDA 2012 National Resources Inventory
(NRI) (USDA 2015); (4) the USGS U.S. Conterminous Wall-to-Wall Anthropogenic Land Use Trends
(NWALT), 1974-2012 (Falcone 2015); and (5) a pair of studies from the University of Wisconsin (Lark
et al. 2015) and the University of Minnesota (Wright et al. 2017). As mentioned above, these efforts vary
26 The DOE's 2016 Billion Ton Report "explicitly assumes no changes in the size for each major
land class (forest, agriculture, etc.) and also keeps both plantation and natural commercial forest areas
fixed; there are minimal changes in management on pasture and cropland within the agricultural land
base, and other projected demands for goods and services are met in addition to biomass for energy to
produce over one billion dry tons of biomass with minimum environmental effects by 2040" (DOE
2016).
27 Trends in fertilizer and chemical inputs are discussed in section 2.2.4 and 2.2.5, respectively,
while trends in tillage are discussed briefly in section 3.5.
24
-------
in their definitions, scope, and approach, influencing their comparability. We clarify these differences
below and provide key definitions of terms in Appendix B. There have also been several regional studies
documenting land use change in different parts of the country, including the Prairie Pothole Region
(Johnston 2013; Johnston 2014; Reitsma et al. 2016), around the Great Lakes (Mladenoff et al. 2016),
for the western corn belt (Shao et al. 2016), for lands in the Conservations Reserve Program (CRP)
(Morefield et al. 2016), and for corn/soybean farms (Wallander et al. 2011). We focus primarily on the
major national efforts covering many crops and land uses but mention the more specific studies where
appropriate.
The USDA Major Land Uses (MLU) report (Bigelow et al. 2017) is produced by the USDA's
Economic Research Service (ERS) and is one of the most comprehensive land use assessments available
in the United States. The MLU is constructed using information from several sources, including USDA
(Census, ERS, NASS, NRI), the US Census Bureau, and the U.S. Forest Service (USFS, Forest
Inventory and Analysis, or FIA). The MLU has been produced since 1949 and reports on five-year
intervals coincident with the Census. Total cropland as defined in the MLU has five components. The
first three (harvested cropland, failed crops, and summer fallow crops) make up "cropland used for
crops" and describe the acreage devoted to crop production. The last two (cropland pasture and idle
cropland) are not directly used for crop production in a given year but may rotate into production (see
Appendix B for full definitions). The MLU includes set-asides such as the Conservation Reserve
Program, the Acreage Reduction Program, and other Federal acreage-reduction programs into the "Idle"
category. The MLU was released in August of 2017 and was not available at the time of the External
Review Draft of this Report (i.e., after the May 2017 cutoff), thus the ERD was updated with
information from the MLU for the Final Report.
The MLU found that total cropland decreased by 16 million acres between 2007 and 2012,
continuing a long decline in total cropland that began in the 1970s. We focus on changes since 2007
because of the intended focus of this Report, and we refer readers to the MLU and elsewhere for longer
term discussions of land use in the U.S. The decrease in total cropland between 2007 and 2012 reported
in the MLU came primarily from a 23-million-acre decline in cropland pasture offsetting a 5-million-
acre increase in cropland used for crops (see Table 3). Thus, land devoted to crop production increased
by 5 million acres between 2007 and 2012. The increase in cropland used for crops was mostly in the
Northern Plains region (ND, SD, NB, KS) and was from corn and soybean increases, with the largest
decreases from hay (see Figure 10 and Figure 11). The MLU explicitly noted biofuels as a potential
25
-------
Table 3. Major land uses (in millions of acres) from the MLU (Bigelow et al. 2017).
Land Use
1945
1949
1959
1964
1969
1974
1978
1982
1987
1992
1997
2002
2007
2012
Million acres
Cropland
451
478
458
444
472
465
471
469
464
460
455
442
408
392
Cropland used for
363
383
359
335
333
361
369
383
331
338
349
340
335
340
crops
Idle cropland
40
26
34
52
51
21
26
21
68
56
39
40
37
39
Cropland pasture
47
69
66
57
88
83
76
65
65
67
68
62
36
13
Grassland pasture
and range
659
632
633
640
604
598
587
597
591
591
580
587
614
655
Forest-use land
602
760
728
732
723
718
703
655
648
648
642
651
671
632
Grazed forest-
use land
345
320
245
225
198
179
172
158
155
145
140
134
127
130
Other forest-use
land
257
440
483
507
525
539
531
497
493
503
501
517
544
502
Special-use areas
85
87
123
144
141
147
158
270
279
281
286
297
313
316
Urban areas
15
18
27
29
31
35
45
50
57
59
66
60
61
70
Miscellaneous
other land
93
298
293
277
291
301
301
224
227
224
236
228
197
196
Total land area
1305
2273
2271
2266
2264
2254
2254
2265
2265
2263
2263
2264
2264
2260
contributing source for the reported land use changes.28 The MLU attributed the large decrease in
cropland pasture as largely attributable to a methodological shift in the Census that occurred in 2007 and
2012.29 Because of the coincidence of the methodological change and the passage of EISA, it is not
possible with this dataset to attribute changes in cropland pasture as reported in the Census to any one
28 The MLU states: "Another trend that has affected U.S. crop plantings over the past 30 years is
the use of crops as a biofuel input source. Over the past decade, the use of com for biofuel increased
sharply due to the mandate in the Energy Policy Act of 2005 to increase the amount of renewable fuels
in the U.S. fuel supply. This law, coupled with an expansion of required amounts of renewable fuels in
2007, boosted production of corn ethanol."
29 From the MLU: "Cropland pasture estimates, one of two nonpermanent grazing uses tracked
in MLU, declined nearly 80 percent in the past 10 years (2002-12) after exhibiting relative stability for
more than 50 years. This decline is largely attributable to methodological changes [i.e., change in
wording and location of the question in the Census, emphasis added] in the collection of cropland
pasture data in the Census of Agriculture, the data source of the cropland pasture category.. .While there
is no way to definitively determine the extent of the effects of changes in the placement and wording of
the cropland pasture question, it seems likely, given the relatively stable cropland pasture acreage trend
from 1949 to 2002, that the changes contributed to the large decrease between 2002 and 2012." (Bigelow
et al. 2017). The changes are described further in Bigelow et al. (2017).
26
-------
source. The MLU also reported a large increase in grassland between 2007 and 2012 (+41 million acres).
However, this also was attributed to methodological changes in the Census and the USFS Forest
Inventory and Analysis (FIA) causing a corresponding decrease in forest over the same period (-39
million acres).3" It is important to consider these methodological changes through time, as trends can be
mischaracterized if taken out of context. Historical estimates from the MLU are not updated as methods
change, making trends analysis difficult to conduct if based solely on this resource. It is also important to
note that the MLU reports net changes in agricultural land use at the county scale, making it impossible
to track conversion of land from one cover/use type to another at the field scale. Only a dataset that
explicitly tracks land use change for individual land units [e.g., (Lark et al. (2015); USDA (2015);
Wright et al. (2017))] can quantify the amount and type of land use conversion across the U.S.
100
90
80
70
£ 60
0
ro
|50
1 40
30
20
10
0
¦Northeast
¦Lake States
¦Corn Belt
¦Northern Plains
¦Appalachian
¦Southeast
¦ Delta States
¦Southern Plains
¦ Mountain
¦ Pacific
Lncr>stcn^'cn<3-corMr-»rMr'-.(Nr'-.(N
^t-^t-LnLnix>i£>r^r^cococncr)00'-i
cncno^cncncncncncncncncnooo
T—I T—I T—I T—1 T—1 T—1 T—1 T-1 T-1 T-1 T-1 ("Nj OJ (~Nj
Figure 10. Changes through time (1945-2012) in cropland used for crops by MLU region (Bigelow et al. 2017).
30 The MLU attributed the increase in grassland to a methodological change in the USFS FIA
and the Census. For the FIA change, large areas of chaparral and shrubland which were originally
classified as forests because of the presence of tree cover, were reclassified as woodland or grasslands
because the relatively sparse tree cover meant the lands were more likely used as grassland and
rangeland than for timber production (Bigelow et al. 2017). Changes to the Census that likely
contributed to increases in grassland are from the same change to the cropland pasture question
described in footnote 26.
27
-------
100
90
80
70
60
50
I 40
30
20
10
0
1963 1982
Corn, all -
Cotton —
1987 1992
— Soybeans —*
— Sorghum, all
1997
•Hay
•Barley
2002 2007
>« Wheat
Rice
2012
Figure 11. Changes through time (1963-2012) in principal crops harvested for the 48 contiguous States (Bigelow et
al. 2017).
Data from the Census of Agriculture is used directly but not without adjustment by the MLU,
thus it is not surprising that the 2012 Census reported total cropland decreasing by a similar amount over
the same period (16 million acres from 2007-2012) (USDA 2014). However, as with the MLU, the
Census includes different types of land that are managed quite differently in their definition of total
cropland. Total cropland in the Census includes: (1) harvested cropland; (2) other pasture and grazing
land that could have been used for crops without additional improvements31; and (3) other cropland
(which includes three subcategories: cropland on which all crops failed or were abandoned, cropland in
cultivated summer fallow, and cropland idle or used for cover crops or soil improvement but not
harvested and not pastured or grazed). Thus, according to the definitions, on an annual basis potential
cropland is used more like a pasture than a field of row crops.
Examining the individual land types that make up total cropland in the Census demonstrates an
increase in harvested cropland from 2007 to 2012 by 5.4 million acres and a large decrease in potential
cropland by 23 million acres (USDA 2014), similar to numbers reported in the MLU. Other cropland did
31 For convenience, we use "potential cropland" for the Census category "other pasture and
grazing land that could have been used for crops without additional improvements" due to its length.
28
-------
not change much in aggregate from 2007-2012 (Rippey 2015).32 Because the Census and the MLU
include lands that are predominantly used as pasture in their definition of total cropland, focusing on
changes in total cropland can mask conversions from pasture to rowcrops that can have significant
environmental effects (EPA 2011). The most comparable term in the Census for croplands used as crops
in the MLU is a combination of three terms (harvested cropland, failed/abandoned cropland, summer
fallow cropland), which together increased by 7.8 million acres between 2007 and 2012. This
discrepancy is likely due to methodological differences between the Census and the MLU.33 Regardless,
it is clear that both sources report an increase in actively managed croplands.
The 2012 USDA National Resources Inventory (NRI) is an independent data source from the
Census and the MLU that is produced by the USDA Natural Resources Conservation Service (NRCS).
The NRI uses a permanent statistical sampling frame that is used to obtain scientifically credible
information on conditions and trends of soil, water, and related resources (USDA 2015). Instead of based
on survey responses (e.g., the Census and consequently parts of the MLU), the NRI is a representative
statistical sample of all non-Federal lands over a 30-year period (1982-2012 for the most recent NRI).
The NRI reviews and revises historical estimates as necessary with each new Report as methods are
updated (the Census and MLU do not); thus, changes due to methodology are removed from the NRI so
long as historical comparisons are made within the same year's Report. A consistent methodology is a
significant advantage when trying to examine trends through time. Thus, different reports have different
strengths and weaknesses - estimates of trends may be better assessed with reports such as the NRI
where methods through time are internally consistent, whereas estimates of acreages at a point in time
are probably better reflected with more comprehensive assessments such as the MLU and NWALT.
The 2015 NRI reported that after a 25-year decrease from 1982 to 2007, total cropland increased by 3.9
million acres between 2007 and 2012 primarily from an increase in cultivated cropland of 4.3 million
acres, as shown in Figure 12 (USDA 2015).34 Uncultivated cropland (e.g., hay) was relatively steady at
32 Other cropland increased from 61 to 62 million acres mostly from an increase in cropland
failed or abandoned (+4 million acres), offsetting decreases in idle cropland (-1.6 million acres) and
summer fallow (-1.5 million acres). The increase in failed/abandoned cropland was likely due to the
2012 drought in the Midwest (Rippey et al. 2015).
33 In the MLU, annual estimates of cropland harvested are based on both Census data and NASS
data on principal crops. Annual estimates of crop failure are based on differences in planted and
harvested acreage of principal crops from the NASS data series. Annual estimates of cultivated summer
fallow historically have been based on fragmentary data from a variety of sources. (Bigelow et al. 2017)
34 Methodologically, the NRI separates total cropland into two types: cultivated (e.g., rowcrops
and land in rotation with rowcrops) and noncultivated (e.g., permanent hay). See Appendix B for full
definitions.
29
-------
52.9 and 52.4 million acres in 2007 and 2012, respectively. The NRI does not include classifications for
whether the crop failed or was abandoned. The increase in total cropland in the NRI came primarily from
lands formerly in the CRP (50%) and pasture (41%). Net changes in land cover/use between 2007 and
2012 included large decreases in CRP (-8.2 million acres) and increases in total cropland and developed
land (+3 million acres) as seen in Figure 13. Morefield et al. (2016) also reported conversion of CRP
lands to row crops from 2010 to 2013, with almost 30% of the 1.3 million acres coming out of the
program in the Midwest going to five row and grain crops (corn, soy, winter and spring wheat, and
sorghum).
450
400
350
300
250
200
150
100
50
0
44.3
43.3
46.6
49.6
53
52.9
52.4
376.2
362.9
335.4
326.7
314.7
306
310.3
1982 1987 1992 1997 2002 2007 2012
Year
¦ Cultivated ¦ Noncultivated
Figure 12. Changes in cultivated, noncultivated, and total cropland from 1982-2012 from the NRI (USDA 2015).
The USGS U.S. Conterminous Wall-to-Wall Anthropogenic Land Use Trends (NWALT), 1974-
2012 report (Falcone 2015) is, along with the MLU, the most comprehensive land use dataset available
for assessment of land use trends in the US. The main purpose of the NWALT is to provide a
comprehensive land use dataset that is consistent with the high resolution (60-m pixel) USGS National
Land Cover Dataset (NLCD) schema and that can be hindcast to the 1970's as part of the USGS
National Water Quality Assessment (NAWQA) Program. The 2015 NWALT is primarily based on
satellite data from the 2011 NLCD (lin et al. 2013), but it is supplemented and cross-validated with
30
-------
5000
5000
-10000
Figure 13. Net changes in all major land cover/use categories between 2007 and 2012 in the NRI (USDA 2015).
The net change estimate for forest land is not reliable (nr) as the margin of error is greater than the estimate.
many other datasets [e.g., USDA Census of Agriculture, USDA NRI, USDA CDL, etc.; Falcone (2015)].
Because of the primary basis in the NLCD, changes in agriculture from the NWALT are muted
compared to that in the Census.35 The years covered in the NWALT are 1974, 1982, 1992, 2002, and
2012. Like the NRI, historical estimates in the NWALT are updated as methods change. Agriculture in
the NWALT36, although not directly based on the Census for agricultural data, is adjusted to match
changes in total cropland at the county and state level from the Census.
35 From Falcone (2015): "The NLCD typically shows smaller agriculture changes than would be
suggested by the CoA. For example, in the CoA, for any 10-year period, approximately 60 percent of
counties had a TC increase or decrease of more than 1 percent of county area, and about 40 percent had
more than a 2 percent increase or decrease. For the NLCD 2001-2011, only 12 percent of counties show
an Agriculture change of more than 1 percent of county area, and only 3 percent of counties had more
than a 2 percent change. The magnitudes of agriculture changes in this product are typically somewhat
more than what the NLCD indicates, but less than the CoA."
36 "Agriculture" in the NWALT is the sum of classes 43 (Production, Crops) and 44 (Production,
Pasture/Hay), and these two together are most comparable to the category of total cropland in the Census
(Falcone et al. 2015). The NWALT also notes that class 45 (Production, Grazing Potential) is a
""swing"' category that could go into Agriculture or not depending on the user's goals.
31
-------
The NWALT reported that between 2002 and 2012, crop area increased by 3.9 million acres,
pasture/hay decreased by 5.7 million acres, and developed lands increased by 5.5 million acres. No
intermediate estimates are available for 2007. Geographically, the distribution of increased crop area was
similar to other studies, with hotspots (i.e., greater than 5,000 acre increase in a county) in the eastern
plains from Texas to North Dakota, along with pockets in already agricultural areas of the Midwest and
elsewhere (see Figure 14). Decreases in pasture/hay were more universal across much of the country east
of the Rockies (see Figure 14).
The NWALT undergoes an extensive quality assurance process, and Falcone (2015) reports
general agreement with the trends and magnitudes with other products (e.g., the Census, NRI, etc.). For
example, the 2015 NWALT compared the number of counties that gained or lost >1% total cropland
with the Census for all 1992-2002, 1982-1992, and 1974-1982, and found >93% agreement.37 It is
important to note that the NWALT did not perform comparisons with the Census with the 2002-2012
interval for agricultural lands because of "less certainty in the validation data" from the Census owing to
the methodological changes that occurred in 2007 as mentioned above33 (Falcone 2015).
The final national assessment of land use change since 2007 was a pair of studies led by
researchers at the University of Wisconsin (Lark et al. 2015) and the University of Minnesota (Wright et
al. 2017). They used the 2012 USDA Cropland Datalayer (CDL) along with several other datasets to
assess land use change from 2008 to 2012. The CDL is a satellite-derived land cover data product (30-m
resolution) produced by the USDA's NASS based on several satellite retrievals (MODIS, IRS-P6
Re source sat-1, Landsat). Detailed accuracy assessments of the CDL are produced by NASS by
comparing crop pixels with ground based samples from the FSA Common Land Unit (CLU) Program
and with NLCD for non-crop pixels. This is an important distinction, because although comparison with
the FSA CLU is considered very robust, there is no robust "noncrop" national datalayer with which to
compare, with the NLCD as a reasonable substitute. Nevertheless, CDL accuracies vary by state and
crop, are fairly high for corn and soy (>90%), and are lower for grassland (<50%) (Reitsma et al.
2016).38 The Lark et al. (2015) and Wright et al. (2017) studies differ from many others that use the
CDL, in that they went through an extensive screening process to make sure the lands they identified as
37 The NWALT reported good agreement with the Census for counties that lost >1 percent total
cropland (3,058 of 3,108 correct, 98.4 percent), and for counties that gained >1 percent total cropland
(2,434 of 2,593 correct, 93.9 percent). This is overall for all time periods except 2002-2012. It is difficult
to interpret the meaning of this agreement since the NWALT is partially calibrated with Census data.
38 See also USDA National Agricultural Statistics Service, CropScape and Cropland Data Layer
- Metadata at https://www.nass.usda.gov/Research_and_Science/Cropland/metadata/meta.php.
32
-------
Absolute Difference in Cropland Coverage (2012-2002)
Units - Acres
-57000 - -5000
-5000--100
-100-0
0.0-100
100-1000
LHI
1000-5000
5000 - 27000
Absolute Difference in Pasture/Hay Coverage (2012-2002)
Units - Acres
-57000 - -5000
-5000--100
-100-0
•
I
1000-5000
5000 - 27000
Figure 14 Difference in cropland (class 43, top) and pasture/hay (class 44. bottom) between 2012 and 2002 by
county from the NWALT (Falcone 2015).
33
-------
converted from non-crop to crop had no evidence of cultivation for 20 years or more.39 Thus, their focus
was on one-time conversion between 2008 and 2012 of areas with no evidence of cultivation
(termed "conversion", and vice versa, termed "abandonment") and does not include intermittent
pasture/cropland rotations as conversions.
Lark et al. (2015) found that total cropland from 2008 to 2012 increased nationally by 3.0
million acres, with gross land conversion40 nearly four times greater than net land conversion.
Grasslands made up the bulk of the source of land converted to new croplands (77%), with much lower
percentages from shrublands (8%), idle (8%), forest (3%), wetlands (2%), or other land covers (2%).
Roughly 50% of this expansion occurred on marginal lands as defined by the USDA's Natural Resource
Conservation Service and an additional 15% on lands deemed unsuitable for agriculture.41 The first crop
planted on converted land was dominated by corn (27%), wheat (25%), soybeans (20%), and then
alfalfa (7%).
The follow up study by Wright et al. (2017) focused on land use changes within 25, 50, 75, and
100 miles of the nearest biorefinery in order to try and isolate land use changes that may be attributable
to biofuels. Furthermore, Wright et al. (2017) improved on the methodology in Lark et al. (2015) by
including a validation step using aerial photography from the USDA's National Aerial Imagery Program
and found that the estimates of net conversion in Lark et al. (2015) were biased low because of
overestimation of agricultural abandonment. Instead of 3.0 million acres nationally converted from the
39 Lark et al. (2015) and Wright et al. (2017) first combined the many individual CDL land cover
types into "superclasses" (i.e., cropland and non-cropland) because although some individual crop
accuracies may be low, the accuracy of aggregated classes is higher. Second, they removed any lands
that "flip-flopped" between crop and non-crop within the period of the CDL they were examining (2008-
2012). Thus, remaining lands either did not change superclass at all, or they changed once and remained
in the new superclass (e.g., non-cropland to cropland, or vice versa). Third, they compared converted
pixels with the NLCD from 1992, 2001 and 2006 to make sure that none of the lands identified as
converted between 2008 and 2012 had been agricultural in any of those prior three years. Fourth, they
compared their land cover superclasses with plot data from the USGS Land Cover Trends Project that
identified areas that had been cultivated or not from 1973-2002. Lark also used NASS metadata for the
CDL (https://www.nass.usda.gov/Research_and_Science/Cropland/metadata/meta.php) and determined
that the accuracy of their crop and noncrop superclasses was fairly high, ranging from 97.0-97.5% and
79.8-87.2%, respectively, across years from 2008-2012 (Lark et al. 2017).
40 Total cropland increases (extensification) is the net effect of two processes: gross conversion
of land from non-cropland to cropland (expansion) and gross conversion from cropland to non-cropland
(abandonment).
41 Lark et al. (2015) used the USDA's Natural Resource Conservation Service's (NRCS) land
capability classification (LCC) system to quantify the quality of converted land as "prime" (LCC 1-2;
prime farmland), "marginal" (LCC 3-4; land characterized by severe to very severe limitations), and
"unsuitable" (LCC 5-8; land with limitations that restrict use to non-crop purposes).
34
-------
earlier study, Wright et al. (2017) reported roughly 2.7 and 4.2 million acres of noncropland converted to
cropland within a 50- and 100-mile radius of biorefineries, respectively, across the nation. Fifty-miles is
commonly cited as the economic "break-even point" for transporting feedstock to a biorefinery (Mueller
2010a; Mueller 2010b). Furthermore, they reported higher rates of conversion closer to the biorefineries.
The finding of higher rates of conversion closer to the biorefineries is important and suggests a causal
link, a finding that has been found in other regional studies of Kansas (Brown et al. 2014) and in a nine-
state area in the Midwest (Motamed et al. 2016). There were hotspots of conversion all over the country
(see Figure 15), but the bulk of the expansion was of soybeans, corn, and wheat in North and South
Dakota, of soybeans in the steeper areas of southern Iowa and northern Missouri normally used for
grazing, and of wheat in western areas of Kansas, Oklahoma, and Texas over the Ogallala aquifer (see
Figure 16). Wright et al. (2017) estimated that expansion within 50 miles of biorefineries could generate
roughly 0.37 billion gallons of ethanol per year.
Earlier estimates of cropland extensification from Johnston (2014) and Wright et al. (2013),
although received with much attention, have significant limitations. Johnston (2014) used aerial imagery
from the 1970s and 1980s to identify wetland areas that had been converted to agriculture by 2010-2011.
This land conversion could have occurred long before EISA. Both the Johnston (2014) and Wright et al.
(2013) studies (along with several others) used the USDA Crop Data Layer (CDL) without adjustment
with the NLCD or other sources [unlike Lark et al. (2015) and Wright et al. (2017)], which can lead to an
overestimate of land use change, particularly for grasslands (Dunn et al. 2017). In particular, Dunn et al.
(2017) compared land use change estimates for 20 counties in the Prairie Pothole Region using three
methods: unadjusted CDL, adjusted CDL [per Wright et al. (2017)], and the NAIP. They also reported
that unadjusted CDL data could overestimate land use change, and found, consistent with Wright et al.
(2017), that adjustments led to much lower estimates of land use than either unadjusted CDL and the
NAIP for almost all counties examined. Nevertheless, these earlier studies qualitatively agree with
patterns reported in more recent national studies.
35
-------
Sources of new croplands,
2008-2012
.SJrubSnd
F-_ Fores*
¦j. Wm lands
\ 2*
votncr
Net cropland change (%)
0-25 25-50 50-75 75-100 0-25 25-50 50-75 75-100
Nearest refinery (mi.) Nearest refinery (mi.)
Figure 15 Net rates (a) of land use change from non-cropland to cropland between 2008 and 2012 (purple outlines
represent 100 mile radius around biorefineries) from Wright et al. (2017). Positive numbers (green to red) denote
net cropland expansion while negative numbers (blue) denote net cropland abandonment. Rates of land use change
(b) and total acreage convered (c) from non-cropland to cropland (black bars and left axes) and from cropland to
non-cropland (white bars and right axes) for different distances from nearest biorefineries for 2008-2012 [figure
modified from Wright et al. (2017)]. Pixel size 3.5 mile. Also shown (d) is the source of new croplands [modified
from Lark et al. (2015)]. ©2017 IOP Publishing Ltd for Wright et al. (2017).
36
-------
MOST COMMON "BREAK-OUT" CROP BY REGION
Corn
Wheal
Soybean
AJIalfa
Co11on
Sorghjm
Other
6.9%
4.4%
3.7%
13.9%
20,2%
Oals T.45S
Sunftowfe 1.2%
Dctitte winter wheat with say t.2%
Bartey T. 1%
26.6%
24.5%
Peanuts 0,9%
Apple, 0.6%
©ropes 0.7%
Hiceftra
Figure 16 Most common 'break-out' crop by region from Lark et al. (2015). Map represents the most common first
crop to be planted after conversion to cropland 2008-2012. Corn and soybeans dominated much of the Midwest
and periphery of the Appalachians, while wheat becomes more common moving westward across the plains, with
spring wheat in the north and winter wheat in the south. Note that the map depicts only the predominant type of
breakout crop grown In an area and does not necessarily reflect the amount of each breakout crop grown there.
Nationwide prevalence of each breakout crop is indicated in the legend bar graphs. © 2015 IOP Publishing Ltd for
Lark et al. (2015).
Synthesizing all of these major national efforts (see Table 4), there is a consistent signal
emerging that demonstrates an increase in actively managed cropland by roughly 4-7.8 million acres,
whether from the MLU (+5 million acres, cropland used for crops), the Census (+7.8 million acres of
37
-------
harvested cropland, failed/abandoned crops, and summer fallow crops 42), the NWALT (+3.9 million
acres, production of crops 43), or Lark et al. (2015) and Wright et al. (2017) (+4.2 million acres within
100 miles of a biorefinery44). Comparison of acreage amounts across studies is difficult due to
aforementioned differences in definitions and scope, but we have tried to harmonize those to the degree
possible. Comparisons of percent changes are more robust because differences among studies are
normalized. These also show a consistent increase in actively managed croplands across all studies
(1.2-2.4%, Table 4). This increase of actively managed croplands may be coincident with a decrease in
total cropland, as lightly managed pasture has either been reclassified into grassland, or converted to
actively managed cropland or urban areas. However, the reported decrease in total cropland in the MLU
and the Census are not found in the NRI, and both could be from the same methodological change in the
Census. The only estimate that is longitudinal and does not suffer from potential methodological
changes in the Census is the NRI, which reported comparable estimates of 4.3 million acres. This
acreage (4-5 million acres) is a small increase relative to the large agricultural land base,45 but is a large
increase in absolute terms, being almost the size of the land area of New Jersey. These changes are
reported to be coming mostly from lands that were formerly in grassland for 20 or more years, and
going to corn, soy, and wheat. These trends are likely occurring throughout the country but especially in
the Northern Plains, the western margin of the corn belt, and with infilling of the central corn belt. It is
unknown whether these trends have continued after approximately 2012.
42 If one defines "active cropland" as the sum of harvested cropland, summer fallow cropland,
and failed cropland, the increase according to the Census was 7.8 million acres between 2007 and 2012.
This would exclude "idle cropland" and "potential cropland." Potential cropland has been defined earlier
(footnote 30), and idle cropland is defined as "Cropland idle includes any other acreage which could
have been used for crops without any additional improvement and which was not reported as cropland
harvested, cropland on which all crops failed, cropland in summer fallow, or other pasture or grazing
land that could have been used for crops without additional improvements." Idle cropland includes land
used for cover crops or soil improvement but not harvested or grazed, land in Federal or State
conservation programs, and a few other minor categories.
43 Note that the NWALT estimates changes between 2002 and 2012, while the other studies
estimate changes roughly between 2007 and 2012.
44 A national estimate that includes the NAIP correction from Wright et al. (2017) has not been
published.
45 The Census estimates roughly 315 million acres of harvested cropland and 380 million acres
of total cropland in 2012.
38
-------
Table 4 Comparison of major national studies on land use change, harmonized to the degree possible.
Shown are the source publication, the comparable term(s) and definition(s), years assessed, and the
change in acreage in millions of acres (and % from study- specific reference)
Study
Comparable
Definition
Years
Change in
term(s)
reported
million acres
(%)
USDAMLU
Cropland used for
Three of the cropland acreage components-
(2017)
crops
cropland harvested, crop failure, and cultivated
summer fallow—are collectively termed cropland
2007 - 2012
+5 (1.5%)
used for crops, or the land used as an input to crop
production.
USDA Census
(2017)
Harvested
cropland +
failed/abandoned
+ summer fallow
Harvested cropland - This category includes land
from which crops were harvested and hay was cut,
land used to grow short-rotation woody crops,
Christmas trees, and land in orchards, groves,
vineyards, berries, nurseries, and greenhouses. No
separate definition for failed/abandoned, or
summer fallow cropland
2007 - 2012
+7.8 (2.4%)
(a)
USDA NRI
(2015)
Cultivated
cropland
Cultivated cropland comprises land in row crops or
close-grown crops and also other cultivated
cropland, for example, hayland or pastureland that
is in a rotation with row or close-grown crops.
2007-2012
+4.3 (1.4%)
USGS NWALT
Production, Crops
Areas used for the production of crops, such as
corn, soybeans, wheat, vegetables, or cotton, as
well as perennial woody crops such as orchards
2002 - 2012
3.9 (1.2%)
and vineyards. Includes cultivated crops, row crops,
small grains, and fallow fields.
Lark et al.
2015
Net cropland
Net cropland increases (gross expansion - gross
abandonment) of lands in the lower 48 states that
have no evidence of cultivation since 1992.
2008-2012
3 (1%) (b)
Wright et al.
Net cropland
Net cropland increases (gross expansion - gross
2017
abandonment) of lands within 100 miles of a
biorefinery that have no evidence of cultivation
since 1992.
2008-2012
4.2 (NA) (c)
a. Harvested cropland, failed/abandoned cropland, and summer fallow cropland changed by +5.4,
+4.0, and -1.5 million acres, respectively between 2007 and 2012 according to the Census.
b. Estimates from Lark or Wright are likely to be lower because they focus on a subset of lands that
had no evidence of cultivation for 20 years or more, rather than all land. We include these in the
table for convenience and completeness.
c. We could not calculate the percent increase from Wright et al. (2017) because the 2008 baseline
acreage within 100 miles of a biorefinery was not reported.
39
-------
2.4.2.2 Trends in intensification: Double cropping and changes in crop plantings and rotations
There has been less information published on trends in U.S. agricultural intensification since the
2011 Report. As mentioned above, intensification can take many forms (e.g., double cropping, changes
in fertilizer, chemical inputs, etc.). This section focuses on trends in double cropping and crop plantings
and rotations, with other forms covered elsewhere in the report.
In the most recent national study to date on double cropping, Borchers et al. (2014) used NASS
data from the June Area Survey (JAS) to assess the prevalence of double cropping across the contiguous
U.S. Borchers et al. (2014) use the term double cropping broadly to mean two crops planted (not
necessarily harvested) in the same field or two uses of the same crop. Thus, this could include cropland-
livestock systems and cover crops, in addition to two crops planted and harvested for the market. Thus,
to assess whether JAS-based estimates reflect actual harvesting of multiple crops, JAS-estimates were
compared with those from the Census and were found to roughly agree.46 The authors report that double
cropping only occurred on roughly 2% of total cropland for most years between 1999 and 2012 and did
not show a consistent trend for any of the seven regions examined. Thus, Borchers et al. (2014) suggest
that increased double cropping does not contribute to intensification.
As for changes in crop plantings (e.g., wheat to corn) or changes in crop rotation patterns (e.g.,
corn-soy-corn to corn-corn-soy), much less has been published to date. Wallander et al. (2011) used the
Agricultural Resource Management Survey to focus on land use change for corn and soybean farmers
nationally, with an emphasis on 2006-2008. They found that corn acreage increased mostly on farms that
previously grew soybeans, but other farms (primarily cotton) offset these shifts by shifting to soybean
production (Wallander et al. 2011). However, the short time window for this study that centered on the
relatively anomalous year of 2007 (see Figure 4) suggest that different conclusions may be reached if the
time window were moved to subsequent years or for a longer period. Thus, whether these short-term
changes in crop plantings have been sustained is unclear. More recently, Beckman et al. (2013) reported
that increases in corn acreage from 2001-2012 resulted in a net decrease in barley, oats, and sorghum.
Regional studies on changes in crop plantings and rotations focused mainly on the central
Midwestern areas that are already highly agricultural. An analysis across a nine-state area in the Midwest
reported the area of continuous corn increased by 2.5-5 million acres from 2006-2010, with a smaller
decrease in continuous soybean and little change in corn-soy rotations (Plourde et al. 2013). In contrast,
a detailed study in eastern Iowa examining changes in corn and soybean rotations found that the most
46 The Census-based estimate double cropping was 7% higher than the JAS-based estimate.
40
-------
common rotation between 2002 and 2007 (corn-soy) was absent between 2007 and 2012, with 59%
replaced by two or more years of continuous corn and 41% replaced by two or more years of continuous
soybean (Figure 17) (Ren et al. 2016). The authors reported that corn tended to be planted on higher
quality lands, while soybeans were pushed to lower quality lands. A study in Kansas compared corn
extensification and intensification, focusing on land use changes to corn (Brown et al. 2014).47 They
reported that corn intensification far outweighed corn extensification (79% and 16% of corn land use
changes, respectively), with more extensification in the arid west where there was already less corn and
more intensification in the rest of the state where corn was already grown. As with Wright et al. (2017),
Brown et al. (2014) also found an influence of biorefinery proximity, with an 8% increase in conversion
to corn from already cropped land, and a 10% increase in extensification, when one moved 1% closer to
a biorefinery that was 50 miles away
These region- and state-specific studies are not inconsistent with the national studies. For
Kansas, Wright et al. (2017) also reported most of the conversions from non-crop to crop occurred in the
west of the state. For Iowa, the areas of extensification in southern Iowa reported in Wright et al. (2017)
were not included in the nine-county area of eastern Iowa examined in Ren et al. (2016). Thus, a
consistent picture from multiple sources appears to be emerging, with both cropland extensification and
crop switching towards more intensively managed crops occurring throughout the country.
Intensification appears to be dominating in already agricultural areas, while extensification dominates
along the large agricultural margins and within formerly uncultivated areas in the central Midwest.
47 Brown et al. (2014) define extensification as a conversion from noncropland in 2007 to corn
cultivation in 2008 and 2009, and intensification as a conversion from non-corn crop cultivation in 2007
to corn cultivation in 2008 and 2009.
41
-------
- m- -
kr 1
:b- Ifc
Mwv !¦' tk
I
FSf 1T^
w?p,
V - i—
1 ?. v.;
i ".w —
ji - •••¦
fej-1 1- JJV-
J / EEH r' &¦.' '!U
. „¦ rjy
If j' §3
^ i;
IS;; .*S
Historical 02-07 Rotations
I I Cwty bwrtfary
CRP tand*
C CM n n ¦ Sot-Ccf^
B Cw>r»uou» Gyp
3+ Y#*r* df C-3rtt*°uOvil COffl
| Cttw ftefiaiicni
QHwr cropland
H UrtCuliialftl Urtd
JV'-t
Us.
m
v*> -*•* ,
l* ?4i
¦ Mrffci
H
A
ii
Ss
*r i *k
peg
a™ -"> I
• i -a • 95^4**
H T-
JEi ';-l
• W"¦ ,'' I ,i"'-13
•• :"•• ' -V.-
¦'.; : '¦¦¦ : v %
m • - '\ -- ¦
-¦'- V . • •••.
a
—
iW..
^ r
i •?.
¦war ¦- vl. ^"ii.V^'f5.'
V|Jm?«£s *•« '•: F 1
?W*?3K.» . -r,•,;• v^-
- « . v - i' >'
¦V. 5'.>
Historical 07-12 Rotations
I I CsM-Jy tatfdWV
(¦I CRP l*xf*
C wn 42«f iv So yfc*ar<>3ai'Man ¦ Qcv n-Cw n
H Ge«*l*SW* Com
: 3+"Voarti cf Coi-ti'ijiOirt Cwi
IB Qthe' K^lixts
Qtii*' crxx&nd
3 Uf'seul/rifrMti Liix'l
—
jaE? ' '^&v
.
iSSM
••¦! .. ' .
jh^-SSSVi**
vLv-.
. 1 ¦ .
- •
?9 iS?e
Figure 17 Crop rotation patterns for the 9-country area of eastern Iowa in Ren et al. (2016) for 2002-2007 (a) and
2007-2012 (b), with a blowup in Benton county for clarity. Note the disappearance of the corn-soybean rotation
(yellow) between 2002-2007 and 2007-2012.
42
-------
2.4.3 Economic-Based Projections of U.S. Land Use Change Impacts
The 2011 Report highlighted several studies using agro-economic models to simulate the
potential impacts of land use change compared to different future biofuel scenarios (Searchinger et al.
2008; Malcolm et al. 2009; EPA 2010; Tyner et al. 2010). These were summarized by Wallander et al.
(2011) and show a wide range of projections (see Table 2). Comparing the predicted changes
summarized in Wallander et al. (2011) with the observed changes in sections 2.2 and 2.4,48 suggest that
the observed increase in corn and cropland, by roughly 10 million and 3.9-7.8 million acres respectively,
are generally well approximated in the models. However, the observed increase in soybean by roughly 8
million acres was not well represented, and was only directionally consistent in Malcolm et al. (2009),
with the other studies projecting decreases in soybean acreage. A comprehensive assessment of the
performance of these and other models is beyond the scope of this assessment. Since 2011, multiple
studies have continued to evaluate potential land use change impacts associated with increased biofuel
use. Estimates of potential land use change impacts from increased biofuel demand continue to vary
significantly, particularly when evaluating indirect and international land use change impacts (Dunn et
al. 2013; Taheripour et al. 2013b; Macedo et al. 2015; Plevin et al. 2015; Valin et al. 2015). Because
many of these studies are global in nature in order to incorporate global commodity trading, they are
discussed below in Section 2.5 (International Land Use Change).
Regarding likely future land use changes in the U.S., the USDA reports no major changes in
total cropland nor in the eight major crops reported (corn, soybeans, wheat, upland cotton, sorghum, rice,
barley, and oats) in their long-term projections to 2026. USDA expects CRP acreage to hold near the
maximum levels legislated by the 2014 Farm Bill at 24 million acres (USDA 2017). It is likely that these
more recent efforts will have similar difficulty in matching observed changes, because of the complexity
of these agro-economic systems as well as the inherent challenges of comparing models with
observations.
2.4.4 Conclusions
Biofuel feedstock production is responsible for some of the observed changes in land used for
agriculture, but we cannot quantify with precision the amount of land with increased intensity of
cultivation nor confidently estimate the portion of crop land expansion that is due to the market
for biofuels.
48 This exercise is only relevant for Malcolm et al. (2009) which projected to 2015 and
Searchinger et al. (2008) which projected to 2016/2017. EPA (2010) projected to 2022.
43
-------
Recent research and anticipated updates to data are expected to improve our ability over the next
three years to quantify the fraction of land use change attributed to biofuel feedstock production
in the U.S.
Evidence from multiple sources demonstrates an increase in actively managed cropland in the
U.S. since the passage of EISA by roughly 4-7.8 million acres, depending upon the source.
Much of this increase is likely occurring in the western and northern edges of the corn belt with
reductions of pasture and grassland, but also through infilling of already agricultural areas.
Thus, intensification likely dominates in already agricultural areas and extensification dominates
in less agricultural areas.
Research is needed to quantify changes in the intensity of cultivation on existing agricultural
land.
Research is also needed to more effectively connect changes in land use to the environmental
impacts of concern.
There are five major national-scale studies that suggest that cropland has increased in total
acreage in the U.S. by 4-7.8 million acres between 2007-2008 and 2012. For context, 4.7 million acres is
approximately the land area of the state of New Jersey. This has been primarily a conversion of
grassland or pasture to corn, soybeans, and wheat, along the extensive agricultural margin, and through
infilling of previously uncultivated areas in the central Midwest. There are no national updates since
2012, but several are forthcoming, including the 2017 Census of Agriculture, 2017 MLU, and the 2017
Nri 49 xhus, it is not known whether these national trends have continued to the present. There is also
substantial evidence of crop shifting on existing agricultural lands from other row crops towards more
corn and more sequential rotations of corn. The total U.S. acreage experiencing these shifts among
croplands is unknown, but regional studies suggest that the magnitude may be larger than that of
extensification. There is strong correlational evidence that biofiiels are responsible for some of this
observed land use change, but exactly how much remains unclear (see Box 3). Both of these trends in
land use change have direct and indirect effects on many of the environmental end points listed in
Section 204 of EISA and are elaborated further in Chapter 3 below. Additional research is needed to
quantify changes in the intensity of cultivation on existing agricultural land and to more effectively
connect changes in land use to the environmental impacts of concern.
49 A follow up study from Wright et al. (2017) to examine trends from 2008-2016 is also in
preparation (Tyler Lark personal communication).
44
-------
2.5 International Land Use Change
This section discusses land use change drivers at the global scale. First, we provide an overview
of observed land use change globally, including brief discussions of trends in agricultural intensification
and land use changes in regions that have been major exporters of biofuels to the United States. Second,
we discuss economic modelling studies that have attempted to estimate the global land use change
impacts attributable to crop-based biofuels. Overall, we find that the conclusions from the 2011 Report
on international land use change still apply.
2.5.1 Observed International Land Use Change
Land use changes that occur outside of the U.S. are also drivers of the environmental impacts
associated with biofuel production. Such land use changes may be directly or indirectly linked with the
production of biofuel feedstocks, and there are many other direct and indirect drivers for land use
change, such as urbanization, economic development, and climate (UNCCD 2017).
While U.S. biofuel production is accountable for only a fraction of global crop land area, it is
instructive to review global trends in land use that coincide with the recent ramp up in biofuel
production. For context, the figure below shows net land use changes from 2000 to 2007 and from 2007
to 2014, as reported by the Food and Agriculture Organization of the United Nations (FAO).50 The FAO
land use data are an annual time series that can be used to evaluate any number of time period
combinations. Figure 18 shows global land use change during the seven years preceding and following
the enactment of EISA 2007.
For both time periods, Figure 18 shows gains in area harvested and arable land and permanent
crops, and losses in the area of forests and permanent meadows and pastures.51 For area harvested, crops
that are planted and harvested more than once on the same field during the year are counted as many
times as harvested. For the category called arable land and permanent crops (hereafter "arable land"), the
same field would be counted only once per year. Arable land includes cropland currently in production
as well as potential cropland (similar to the total cropland category reported in the USDA Census,
discussed above in Section 2.4). Area harvested and arable land overlap and are therefore not mutually
exclusive or additive. From 2000 to 2014, harvested area increased by 504 million acres, while arable
50 FAOSTAT, available at http://www.fao.Org/faostat/en/#home. accessed January 2018. The
most recent year reported for land use data was 2014 at the time the data were accessed.
51 Definitions for these land use categories are provided in the Appendix B table of Key Terms
for Major Land Use Change Studies.
45
-------
Land Use Change from 2007 to 2014 (Million Acres)
400
300
200
100
(100)
(200)
Land Use Change from 2000 to 2007 (Million Acres)
I
(300)
Africa Brazil China European Rest of Asia Rest of Rest of USA Grand Total
Union 27 Latin World
America
¦ Harvested Area ¦ Forested Area ¦ Permanent Meadows & Pastures
Figure 18 Global land use change by aggregate region (data from FAOSTAT50),
land increased by 116 million acres. The ratio of area harvested to arable land increased during this
period, implying an increase in harvests per planted acre, an increase in the share of potential cropland
planted and harvested, or likely some combination of both. The FAO data do not allow us to separate
these effects, and understanding the extent and details of these intensification channels is an area for
ongoing research.
46
-------
With increased population, per capita income, and biofuel production during this time period,
crop extensification would likely have been larger without concurrent intensification through increased
crop yields, rates of harvesting planted areas and harvests per year (Ray et al. 2013; Langeveld et al.
2014; Babcock 2015). For example, based on data reported by FAO, from 2000 to 2014 crop production
(total mass) increased by 42 percent, while harvested area increased by only 17 percent, accompanied by
a 21 percent increase in yield (tons per acre). Total factor productivity (TFP) is a measure that provides
a more comprehensive accounting of productivity gains than yield per acre (Fuglie et al. 2013). Data
from USDA-ERS suggests that since the year 2000 TFP (growth due to getting more output from
existing inputs) has been the main factor driving global agricultural output growth (see Figure 19). It is
unclear whether or to what extent U.S. biofuel policies have contributed to such gains in TFP.
Although the use of agricultural land has intensified, cropland extensification and deforestation
has continued. Cropland expansion that results in forest loss is a particularly acute driver of
ro
Cl)
>-
&_
CD
Q_
CD
DJD
C
ro
c
CD
U
s_
CD
Q_
O
&_
CD
-M
D
Q_
+-»
D
o
"ro
+-»
o
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Expansion of agricultural land
More inputs per acre
Fv+ <=»ncirm r\f aoriri ilti iro tn rrnnlanrl
1961-70
1971-80
1981-90
1991-2000
2001-10
Figure 19 Global total factor productivity. From Fuglie et al. (2013), derived from FAO and other agricultural data
using method described in Fuglie et al. (2012).
47
-------
environmental impacts. Although forest loss is caused by many factors, it is instructive to look at recent
trends. The 2015 Global Forest Resources Assessment (FAO 2016)52 reported atotal loss in forest area
of 83 million acres from 2005-2015, with per year forest area losses being roughly equal between 2005-
2010 and 2010-2015 [8.15 million acres/year during 2010-2015 (Keenan et al. 2015)]. Overall, the net
annual rate of forest loss has slowed from 0.18 percent in the early 1990s to 0.08 percent during the
period from 2010-2015.
In addition to looking at global trends, it is helpful to consider individual regions. Here we
touch briefly on trends in recent land use changes in countries that are major exporters of biofuels to the
U.S. In recent years, the largest sources of biofuel imports to the U.S. have been sugarcane ethanol from
Brazil, soy biodiesel from Argentina and palm oil biodiesel from Indonesia (see Section 3.7). Planted
and harvested area of sugarcane in Brazil increased by about 9.9 million acres between 2005 and 2010
(Adami et al. 2012; Marin et al. 2016), with an additional 3.3 million acres added between 2010 and
2014.53 The sugarcane expansion occurred mainly through conversion of pasture land and has been
linked to conversion of other natural vegetation including forests (Adami et al. 2012; Filoso et al. 2015).
Soybean harvested area increased by 29.1 million acres from 2004 to 2017.55 From 2004 to 2016, the
annual rate of deforestation in the Amazon decreased from 6.9 to 2.0 million acres but has been
increasing in recent years from a low of 1.1 million acres in 2012.54 The harvested area of soybeans in
Argentina increased by 1.6 million acres from 2004 to 2017,55 largely at the expense of native grasslands
in the Pampas region (Modernel et al. 2016). In Indonesia, harvested palm oil area has increased by 12.2
million acres from 2004 to 2017,55 while forest loss has been around 1.7 million acres per year between
2010 and 2015 (Keenan et al. 2015).
Cropland expansion and natural habitat loss (including forests) have been observed
internationally during the implementation of the RFS program. It is likely that increased biofuel
production has contributed to these land use changes, but significant uncertainty remains about the
amount and type of land use changes that can be quantitatively attributed to U.S. biofuel consumption
(see Box 3 on Attribution).
52 FRA 2015 data was developed from responses to surveys by individual countries. The survey
has a common reporting framework, agreed definitions and reporting standards.
53 FAOSTAT. http://www.fao.org/faostat/en/#home.
54 Brazil National Institute for Space Research (INPE). http://www.inpe.br/ingles/index.php.
55 USDA Foreign Agricultural Service, Production, Supply, and Distribution, PSD Online.
https ://apps .fas .usda. go v/p sdo n 1 i n c/ap p/i n dc x. htm 1 #/app/h o m c.
48
-------
2.5.2 Economic-Model Based Estimates of Biofuel-Induced Land Use Change
The 2011 Report reviewed modeled estimates of biofuel-induced land use changes and their
impacts. It summarized the land use change results from USEPA 2010, which estimated land use change
GHG impacts using two partial-equilibrium models (FASOM and FAPRI-CARD). EPA (2010)
produced a range of results based on quantitative sensitivity analysis of the satellite data and land use
change emissions factors used in the modeling framework. Quantitative sensitivity analysis was not
performed for the economic parameters within the FASOM and FAPRI-CARD models, but a "high-
yield" scenario was run for comparison. The 2011 Report compared EPA (2010) with other modeling
projections available at the time and found "the results of modeling projected impacts are diverse and it
not possible at this time to predict with any certainty what type of land use change in other countries will
result from increased U.S. demand for biofuel or what its environmental consequences will be (p. 5-7)."
This section reviews additional modeling studies since the 2011 Report.
Figure 20 below summarizes results from studies since the 2011 Report that estimated the land
use change associated with corn ethanol. The figure and the discussion in this section focuses on corn
ethanol since it is the most intensively studied biofuel in the U.S. and accounts for the largest volume of
biofuel. Although GHG impacts are outside the scope of the current report, this figure presents GHG
emissions per unit of corn ethanol produced as a proxy for the overall scale of land use change. This was
the only single readily available common metric across the studies reviewed that summarizes the scale
and nature of the land use changes projected. Displaying GHG results has the benefit of synthesizing
multi-dimensional results into one comparable metric. All else equal, results with higher land use change
GHG emissions are associated with greater areas of land use change and greater clearing of high-carbon
stock lands such as primary forests. The figure is not meant to be comprehensive, and the results
presented are limited to studies that did original modeling in a peer-reviewed publication or as part of a
regulatory analyses performed for a governmental body.
The studies reviewed can be categorized by the type of model or analytical methods used. A
number of studies used partial equilibrium models representing the agricultural sector (e.g., FAPRI) or
the agricultural and forestry sectors (e.g., FASOM and GLOBIOM). Another group of studies used
computable general equilibrium models (e.g., GTAP-BIO or MIRAGE). While CGE models have the
advantage of representing the entire global economy, they often lack detail in the agricultural and
forestry sectors compared to partial equilibrium models. Another group of studies developed reduced
form models for the express purpose of evaluating the uncertainty flowing from certain aspects of
biofuel-induced land use change modeling.
49
-------
Taheripour et al.
(2017) -
Valin et al.
(2015)-
Plevin et al.
(2015)-
CARB
(2014)-
Bento and Klotz
(2014)-
Taheripour et al.
(2013) -
Taheripour and Tyner
(2013)-
Dunn et al.
(2013)-
Plevin et al.
(2010)-
Hertel et al.
(2010)-
EPA
(2010)-
,
-------
It is also important to note that many of the recent studies with lower estimates do not include
quantitative sensitivity analysis. Many of the key parameters in land use change models remain highly
uncertain [e.g., Zilberman et al. (2013) and Tokgoz et al. (2014)], which reduces the weight that can be
given to any individual model run. Studies that have included robust sensitivity analysis have reported
wide ranges of results stemming from parametric uncertainty alone [e.g., Plevin et al. (2015) and Valin
et al. (2015)]. Other sources of uncertainty that are difficult to quantify flow from variability in model
structures, scenario design, and other methodological choices (Plevin et al. 2017b).
Another set of studies (not shown in the figure above because they did not include original
model estimates) reviewed biofuel modeling and proposed alternative methodologies for consideration.
For example, Kim et al. (2012) suggested that some of the land use change that modelers have attributed
to biofuels should instead be apportioned to consumers' dietary preferences. On the other hand,
Searchinger et al. (2015) argued that current models project benefits for biofuels only because they
assume global food consumption will be reduced as biofuel production increases. These studies highlight
just a couple of the methodological questions surrounding biofuel modeling that have not been fully
resolved.
The discussion in this section has focused on corn ethanol, but many of the same general
observations apply to soybean oil biodiesel, sugarcane ethanol, and other biofuels derived from planted
crops or trees. It is worth noting, however, that, indirect land use change "factors for biodiesel crops are
considerably higher and subjected to higher uncertainty levels than ethanol crops" (Souza et al. 2015).
Modeling results for dedicated energy crops tend to be lower than comparable studies for food-based
crops [e.g., EPA (2010), Valin et al. (2015), and Dunn et al. (2013)], but since energy crop production
has been limited in scale these studies must rely on assumptions about how the industry will develop.
As predicted by NAS (2011), "scientists will undoubtedly continue to refine their models to
improve estimates of GHG emissions as a result of land use changes. However, uncertainty of GHG
emissions from land use and land cover changes can be expected to remain large because actual land
changes and their relation to increasing biofuels production in the United States will only be observed as
markets adjust to increased biofuel production. Even with long-term empirical data on land use and
landcover changes, measurement of associated GHG emissions, and data on agricultural markets,
estimating the global GHG benefits or emissions from U.S. biofuel production will require a comparison
to reference scenario, which inevitably is a simulation of what would have happened absent biofuels" (p.
192). Since a reference scenario cannot be measured, indirect land use change impacts are by definition
uncertain. A recent study conducted by Woltjer et al. (2017)Woltjer et al. (2017) that reviewed biofuel
51
-------
land use change modeling since 2012 for the European Commission concluded that, "progress in the
calculation of [indirect land use change] effects from biofuel production, and reduction of uncertainties,
has been limited" (p. 95). Significant room for improvement remains in basic areas such as model
comparison, data standardization, and empirical support for economic parameters (Souza et al. 2015).
Many authors have also highlighted the inherent uncertainty associated with biofuel-induced land use
changes and proposed various ways to address and factor such uncertainties into decision making
(Kocoloski et al. 2013; DeCicco et al. 2016; Plevin et al. 2017a). New research since the 2011 Report
has improved our understanding of biofuel-induced land use change modeling, but the overall
conclusions we can draw from this body of modeling have not changed.
2.5.3 Conclusions
Conclusions for observed international land use change:
• Global cropland area has expanded since the year 2000, coinciding with the increase in U.S.
biofuel production. During this period, the ratio of area harvested to arable land increased and
crop yields increased significantly, due in large part to gains in total factor productivity.
• Agricultural extensification and deforestation have been documented in countries that are major
exporters of biofiiels to the U.S., including Brazil, Argentina, and Indonesia.
• Cropland expansion and natural habitat loss (including forests) have been observed
internationally during the implementation of the RFS program. It is likely that increased biofuel
production has contributed to these land use changes, but significant uncertainty remains about
the amount and type of land use changes that can be quantitatively attributed to U.S. biofuel
consumption (see Box 3 on Attribution).
Conclusions for economic-model based estimates of biofuel induced land use change:
• Researchers have continued to update and refine economic models to estimate biofuel-induced
land use changes.
• Due to inherent challenges, uncertainties are large and progress in reducing the sources of
uncertainty has been limited.
52
-------
Box 3. Attribution of the Environmental Effects of Biofuels
Most environmental effects of biofuel production are associated with the feedstock production
stage (EPA 2011; Hellwinckel etal. 2016). At the feedstock production stage, land use change has
been identified as one of the primary drivers affecting environmental impacts. Farmers' decisions
regarding land use and management are influenced in part by market prices (e.g., future price of corn),
which are in turn affected by myriad antecedent factors, such as weather and policies (Roberts et al.
2013; Hellwinckel et al. (2016); Carter et al. 2017). The dominant biofuel feedstocks in the U.S.
currently are corn and soybeans (see Section 2.2); thus, the environmental effects of biofuels at this
time are due to some portion of the land use and management of growing corn and soybeans.
However, these feedstocks are also produced for other purposes, such as animal feed, many food and
industrial products, and export. Therefore, only a percentage of the environmental consequences of
growing corn and soybeans can be attributed to biofuel feedstock production. The question is what
percentage of the environmental effects of producing corn and soybeans are attributable to corn-grain
ethanol and soy biodiesel, respectively? And, from this follows - what percentage of these
environmental effects are attributable to the Renewable Fuel Standard Program specifically?
Understanding the type and location of land use attributable to biofuel feedstock production is
a first step towards attribution of environmental effects. Changes in crop types or domestic land use,
such as conversion of land to agriculture, can be caused by variety of factors, and allocating
proportional causation to these factors, including biofuels, can be difficult (Efroymson et al. 2016).
One simple method is to apply the percentage of corn grain and soybeans used for biofuels directly to
land use, but there are limitations with this approach. Currently, approximately 40% of corn grain and
12% of soybeans produced nationally go to biofuels (see Section 2.2). But it would be inaccurate to
assume that changes in these percentages are equivalent to changes in land use due to biofuel
production. Improved production efficiency can result in more volume produced on the same land
area. A constant level of biofuel feedstock production and associated land use combined with lower
total production and land use would result in a higher percentage of production attributed to biofuels
but no actual land use change.
A co-product of corn ethanol production is distillers dried grains with solubles, which can
displace corn grown for animal feed and reduce the percentage of land (and environmental effects)
attributable to corn grown for ethanol. A recent study examined this displacement and found that
accounting for distillers grains reduced corn acreage attributable to ethanol from 40% to 25%
nationally in 2011 (Mumm et al. 2014). Additionally, only soybean oil (which is approximately 20%
of the soybean by weight; Section 2.2) is used for biodiesel production, so this means approximately
2.5% of the soybean harvest by mass is attributable to biodiesel. If feedstock production for biofuels
were evenly distributed across the country, then 25% and 2.5% of corn and soybean acreage,
respectively, are reasonable first estimates attributable to biofuels.
(Continued)
53
-------
Box 3 (Continued)
However, biofuel-induced land use changes may not be evenly distributed, with feedstock
production potentially concentrated in certain areas. Land use changes in areas around biorefineries or
those with new plantings since 2007 may indicate the effects of biofuel production. This can be
accounted for using another approach for assessing attribution: statistical correlative analysis.
Motamed et al. (2016), for instance, estimated that for every 1% increase in an area's ethanol refining
capacity, its corn acreage and total agricultural acreage increased by 1.5% and 1.7%, respectively.
This finding suggests that corn ethanol production has been responsible for increasing corn production
and land conversion around biorefineries.
Agro-economic models (e.g., FASOM, POLYSYS, and REAP) have also been employed,
comparing land use and crop production with and without a given amount of biofuel production [e.g.,
Malcolm et al. (2009)]. As an example, Malcolm et al. (2009) estimated that biofuel volume targets
would lead to an increase of approximately five million acres of cropland by 2015, with most due to
corn ethanol. There are tradeoffs between direct measurements versus economic modeling approaches.
It can difficult to assign cause using direct observation, while modeling studies may be overly
simplistic, failing to account for key drivers or complex interactions, and it is difficult to validate
model projections with historical data.
Besides land use, the environmental effects on air, soil, water quality, and other end-points
depend not just on the crop and preceding land cover, but how the land is managed (e.g., no-till
management, pesticide usage, riparian buffer strips, etc.). For example, if additional biofuel-induced
corn is grown on marginal lands at higher rates than average, it could require more fertilizer and lead
to higher nutrient runoff or leaching to waterways. Conversely, if additional biofuel-induced corn is
grown on prime agricultural land with riparian buffers, less fertilizer and lower nutrient and sediment
runoff could result. Empirical or dynamic ecosystem models (such as SPARROW, EPIC or SWAT)
can help simulate these effects once land use is assigned. These estimates can then be used in life
cycle analyses to determine environmental effects across all stages of production and use. Such
combined analyses remain an area of emerging research.
Currently, we can state that biofuels are responsible for a percentage of domestic land used
for—and the environmental effects from—corn and soybean production, including newly converted
land. However, using peer reviewed information that forms the backbone of this Report, we cannot
quantify these percentages with confidence at this time based solely on that information without new
analyses. Since 2011, a clearer picture of U.S. land use change has emerged (see Section 2.4)
potentially allowing a quantitative attribution to biofuels and estimation of environmental effects in
the future. Moreover, the general relationship between the two current U.S. biofuel feedstocks (corn
and soybeans) and environmental impacts is well known from decades of agricultural research. Each
individual section in the Impacts Chapter below discusses what is currently known about the effects of
corn and soybean production in general, and biofuel production specifically, on their respective
environmental end-points.
54
-------
3 Environmental and Resource Conservation Impacts
3.1 Air Quality
3.1.1 2011 Report Conclusions
According to the 2011 Report, the most negative air quality impacts from biofuel production
were associated with production facilities using coal as the main energy source.56 The report added that
air quality impacts could be mitigated through use of cleaner fuels, such as natural gas, and more
efficient processes and energy-generation equipment. In addition, energy-saving technologies such as
those used by combined heat and power (CHP) facilities are also an effective means to reduce air
emissions associated with biofuel production. The report concluded that the impacts from transport of
biofiiels are not expected to be significant, although air quality will be affected to a small degree locally
by emissions from biofuel transport via rail, barge, and tank truck and by evaporative, spillage, and
permeation emissions from transfer and storage activities.
The 2011 Report also concluded that, for ethanol blends, end-use emission rates were expected
to be higher for nitrogen oxides (NOx) relative to gasoline. The effect of ethanol on NOx occurs because
addition of ethanol to gasoline adds oxygen to the fuel and modifies the air/fuel mixture in a way that
leads to higher NOx emissions. The 2011 Report found that end-use emissions for ethanol blends were
independent of feedstock. In 2011, the National Academy of Sciences also released its report,
"Renewable Fuel Standard: Potential Economic and Environmental Effects of U.S. Biofuel Policy"
(NAS 2011), which reached conclusions generally consistent with EPA's 2011 Report. It concluded that
air quality modeling suggests that production and use of ethanol as fuel to displace gasoline is likely to
increase such air pollutants as PM2 5, ozone, and SOx in some locations. The NAS report was a synthesis
of available research by a team of experts and emphasizes the spatial component of impacts with some
effects being local (air quality) and others regional or global (greenhouse gases). Discussion of air
quality impacts in the NAS report focused on ethanol.
The 2011 Report concluded that, relative to petroleum-based diesel fuel, biodiesel increases NOx
emissions and decreases particulate matter (PM), hydrocarbon, and CO emissions. It also found that
56 EPA's RFS2 regulatory impact analysis (EPA, 2010) identified significantly higher CO, NOx,
PM10, PM2 5, and SOxemissions for coal plants than plants using natural gas. However, the majority of
ethanol is produced by plants using natural gas.
55
-------
emissions of some pollutants (PM, nitrous oxide, CO) are higher with plant-based rather than animal-
based biodiesel feedstocks.
The 2011 Report also summarized results of air quality modeling done for EPA's RFS2 rule
(EPA 2010; Cook et al. 2011). This modeling suggested that the increased biofuel use mandated by
EISA would increase ambient PM2.5" in some areas and decrease PM2 5 in others, with small ozone
increases over much of the country along with improvements in a few areas. Ozone increases occur in
NOx-limited areas of the country (VOC levels are high relative to NOx). However, in a few VOC-limited
areas, such as Southern California, NOx increases may decrease ozone. The RFS2 modeling also found
little impact on ambient concentrations of most air toxics. In reaching these conclusions, the 2010
assessment of the RFS2 rule (EPA 2010) took into consideration offsetting emissions impacts associated
with reductions in fossil fuel volumes due to replacement with biofuels. Uncertainties with that analysis
included limited vehicle emissions data for advanced technology vehicles, uncertainty in the assumed
fuel types and blend concentrations, uncertainty in emissions from (cellulosic) ethanol production plants,
uncertainty in transport/fuel storage, and uncertainties in the model itself (e.g., in the chemical
mechanism). However, since that study there has been some limited additional research on emissions
impacts associated with production of different feedstocks [e.g., Zhang et al. (2016)], and especially
biofuel end-use emissions, as discussed below. The relatively limited research that has been published
over the last six years continues to support the findings discussed above.
In the following sections, we revisit these 2011 air quality conclusions. First, we provide a brief
overview of the major changes in the drivers of air quality and their impacts since 2011. We address
these by life cycle stage. Second, we highlight changes in our understanding of the connections between
drivers and impacts since 2011. Third, we focus on likely future changes, and, finally, we provide
bulleted conclusions.
3.1.2 Drivers of Impacts to Air Quality
Air quality, as measured by the concentration of air pollutants in the ambient atmosphere, can be
directly affected by increased production and use of biofuels through changes in emissions of air
pollutants during: (1) feedstock production; (2) conversion of feedstocks to biofuels; (3) transport of
biofuels and feedstocks; and (4) combustion of biofuels in vehicles. Air quality can also be impacted
indirectly, through price-induced impacts associated with increased production and use of biofuels, such
as changes in petroleum fuel consumption and changes in agricultural production and land use. Direct
57 Particulate matter with an aerodynamic diameter of 2.5 (jm or less.
56
-------
impacts on emissions occur due to changes in biofuel volumes produced and consumed and changes in
technologies and practices in each of the previous four processes. For example, as farmers replace older
equipment with new clean equipment with modern emission controls, emissions associated with
feedstock production decrease. Indirectly, petroleum production displacement from increased use of
biofuels impacts emissions, as do changes in fuel properties due to the addition of biofuels to petroleum
fuels. Emissions of NOx, sulfur oxides (SOx), CO, volatile organic compounds (VOCs), ammonia
(NH3), and PM can be impacted at each stage of biofuel production, distribution, and usage and depend
on feedstock type, land use change, and land management/cultivation practices. As a result, the emission
impacts of biofuel production and distribution and the offsetting impacts on petroleum fuel production
and distribution are substantial and must be considered along with end-use impacts for VOC, PM, and
NOx (EPA 2010). In addition, emission and air quality impacts associated with feedstock production and
conversion of feedstock to biofuels are highly localized. The magnitude, timing, and location of all
these emissions changes can have complex effects on atmospheric concentrations of criteria pollutants
(e.g., 03 and PM2.5) and air toxics, the deposition of those compounds, and subsequent impacts on
human and ecosystem health. In this review, we focus primarily on changes in emissions as a surrogate
for changes in air quality.
As discussed in Chapter 2, current renewable fuel volumes are much lower than the applicable
volumes specified in EISA. The vast majority of renewable fuel sold is ethanol, primarily produced
from corn, and biodiesel, primarily produced from soybean but also other plant- and animal-based oils.
There has been very little market penetration of fuels derived from cellulosic and other advanced
feedstocks. As a result, research on biofuel impacts on air quality has focused on corn ethanol and soy
biodiesel more than on biofuels from other feedstocks. The next section will focus on drivers impacting
air quality from ethanol production and use, while the following section will focus on biodiesel.
Discussion will focus on research published since the last report; thus, the discussion will be limited to
those drivers where significant new information is available. Key drivers from ethanol use include
production of feedstock, production of the biofuel itself, transport of the fuel, and end use of the fuel in
vehicles.
Separately, recall that the current report does not address GHG emissions or associated impacts
from biofuels. EISA established mandatory life cycle GHG reduction thresholds for qualifying
renewable fuels that would replace petroleum-based fuels under the program.58 In a previous analysis,
58 The Act exempts fuel from facilities that commenced construction prior to EISA enactment,
and ethanol from facilities fired by natural gas or biomass that commenced construction prior to
December 31, 2009, from the minimum 20% lifecycle greenhouse gas reduction requirement that
generally applies to non-advanced renewable fuels.
57
-------
EPA used state-of-the-art models, data, and other information to assess the GHG emissions from
biofiiels (EPA 2010). The modeling of GHG emissions conducted for the RFS2 Regulatory Impact
Analysis (RIA) provided a reasonable and scientifically sound basis for making determinations of
whether various biofuel production pathways meet thresholds established in EISA. As discussed in
Chapter 1, this report does not evaluate emissions of carbon dioxide or other GHGs from biofuel
production and use, nor does it attempt to encompass GHG impacts in its conclusions. Instead, this
report provides complementary information to the GHG impacts described in the RIA (EPA 2010),
which should be consulted for more information on this topic.
3.1.3 Impacts to Air Quality
3.1.3.1 Impacts from Ethanol Emissions
3.1.3.1.1 Ethanol Feedstock Production and Transport
Recent research to characterize and/or quantify air quality impacts resulting from biofuel
feedstocks has several unifying characteristics. A number of publications since the 2011 Report have
developed spatially and/or temporally explicit life cycle inventories (LCI) of U.S. biofuel feedstock
production systems' air pollutants (Tessum et al. 2012; Heath et al. 2013; Yu et al. 2013; Zhang et al.
2016).
Zhang et al. (2016) and the U.S. Department of Energy (DOE) (2017) also conducted extensive
inventory analysis at the county level for various feedstocks. They concluded that switchgrass and
miscanthus generate lower emissions than corn grain on a per unit biomass basis due to greater yield.
They also concluded that among various cellulosic feedstocks, emission differences associated with
production are offset by differences in emissions associated with transport due to differences in transport
distance.
Tessum et al. (2012) described spatially and temporally explicit LCIs of air pollutants from
gasoline, ethanol derived from corn grain, and ethanol from corn stover. Their results indicated that life-
cycle air emissions of ethanol were concentrated in the Midwestern "Corn Belt," and that ethanol's life
cycle emissions exhibit different temporal patterns when compared to gasoline. Their study also
concluded that life cycle fine PM emissions were higher for ethanol from corn grain than ethanol from
corn stover. They estimated that the production and consumption of ethanol from corn stover would
increase Midwestern NOx, NH3, and PM2 5 emissions but decrease Midwestern SOx emissions.
Yu et al. (2013) estimated the emissions associated with hauling switchgrass and energy
sorghum feedstocks for biofuel production facilities in Tennessee. Their study generated the least-cost
58
-------
solutions between the feedstock supply systems and biorefineries and estimated resulting emissions from
hauling feedstock using EPA's MOtor Vehicle Emission Simulator (MOVES) model.59 Their results
indicated that the degree of feedstock draw area dispersion and the topography of the draw area around a
biorefinery site are critical factors pertaining to the emissions associated with hauling feedstock to a
biorefinery. On a more local scale, they determined that switchgrass was more suitable than energy
sorghum for biofuel production in Tennessee, primarily due to the higher cost and hauling emissions
associated with sorghum.
An overarching conclusion of these publications was that ethanol from corn grain had the
highest overall air pollutant emission levels and that the magnitude of change in air pollutant emissions
was directly connected to the spatial and temporal characteristics of the feedstock production site. These
conclusions align with air quality impacts described in the 2011 Report.
3.1.3.1.2 Ethanol Production
As of mid-2017, there are approximately 200 ethanol production facilities in the U.S.60 Over
90% of these facilities are dry mill facilities processing corn. Facilities producing ethanol from corn and
cellulosic feedstocks tend to have greater air pollutant emissions relative to petroleum refineries on a
per-BTU of fuel produced basis, but emission rates vary widely among facilities (EPA 2010). Emissions
for the vast majority of biofuel plants are included in EPA's National Emission Inventory (NEI). Current
NEI data support the conclusion of the 2011 Report that ethanol plants relying on coal have the highest
air pollutant emissions.61 However, a 2015 study based on airborne measurements suggests that
emissions of hydrocarbons may be substantially underestimated in the NEI for one of the largest coal-
fired biofuel production plants in the country (de Gouw et al. 2015), which could indicate more
systematic underestimation if confirmed at other facilities. However, industry characterization data
indicate that the number of plants relying on coal as an energy source is relatively small (less than 10%
of all ethanol production facilities, accounting for less than 15% of production) and has slowly decreased
over time. The changing nature of ethanol production facilities indicates that additional research on
emissions from biofuel plants and factors that impact these emissions is desirable.
59 MOVES and Related Models, U.S. Environmental Protection Agency,
https://www.epa.gov/moves/previous-moves-versions-and-documentation.
60 EPA: Public Data for the Renewable Fuel Standard, at https://www.epa.gov/air-emissions-
inventories.
61 EPA: Air Emissions Inventories, at https://www.epa.gov/air-emissions-inventories.
59
-------
3.1.3.1.3 Ethanol Distribution and Storage
While the 2011 Report concluded that emissions from biofiiel distribution and storage, including
emissions from loading and unloading, are not significant, the regulatory impact analysis for the RFS2
rule indicated that EISA-mandated volumes of ethanol (36 billion gallons in 2022) could result in
additional annual U.S. emissions of 7,600 tons ofNOx from combustion processes during storage and
transport and 19,000 tons of VOCs, primarily from storage and transport losses for ethanol and
ethanol/gasoline blends. Strogen et al. (2012) concluded that, although transport of ethanol has a small
impact on the overall transportation sector, suboptimal transportation (i.e., supply chain inefficiencies) of
ethanol during the 2000-2009 timeframe resulted in unnecessary emissions. Some of this suboptimal
transport could be reduced by direct blending of E85 at ethanol plants, a practice that is becoming more
prevalent.
3.1.3.1.4 End Use
Light-duty Vehicle Fleet Emissions Testing Since 2011. Federal Tier 2 light-duty vehicle emission
standards regulating NOx, non-methane organic gases (NMOG), CO, PM, formaldehyde, and fuel sulfur
began phasing in starting in 2004. These standards were fully implemented at the time of the preparation
of the 2011 Report, but at that time only limited data on air emissions were available on the biofuel-
related tailpipe and evaporative emissions of Tier 2 light-duty vehicles, within the peer reviewed
literature or from direct vehicle and engine testing conducted by EPA. From 2009 to 2013, EPA
conducted a joint study with DOE and the Coordinating Research Council, known as the EPAct/V2/E-89
Phase 3 Study. The study assessed the effects of five gasoline properties,62 including ethanol volume, on
exhaust emissions from light-duty vehicles certified to Federal Tier 2 Standards (EPA 2013a; EPA
2013b). This study continued to find that ethanol increased NOx emissions, even though modern
technology vehicles have near instantaneous control of the air/fuel ratio, as most emissions occur in
these systems during times when the vehicle catalyst is not yet warmed up or air/fuel ratio is not
perfectly controlled. The relationships between fuel properties and emissions developed from the
EPAct/v2/e-89 Phase 3 study have been incorporated into the EPA MOVES2014a model to develop
emission inventories that account for the geographic variation of in-use gasoline properties. This updated
information on the effects of ethanol and other fuel properties on Tier 2 vehicle emissions addresses the
significant uncertainty in EPA's RFS2 air quality modeling analysis about the effects of ethanol on Tier
62 Ethanol volume, aromatic content, Reid Vapor Pressure (RVP), T50 distillation point, and T90
distillation point.
60
-------
2 vehicle emissions (EPA 2010). The EPA and the states use the updated MOVES2014a model for
emissions analysis that informs regulations and transportation planning.
In 2016, EPA analyzed data from four different test programs63 to determine exhaust emissions
differences between light-duty flexible fuel vehicles (FFVs) fueled with E85 relative to E10 (EPA
2016a). Only non-methane hydrocarbons (NMHC) and CH4 emissions showed statistically significant
differences between E10 and E85, with reductions in NMHC and increased CH4 for E85 relative to E10.
These results for FFVs fueled with E85 have been incorporated into the EPA MOVES2014a model.
Mid-level Ethanol Blends. The use of mid-level (20% and 30% ethanol content) ethanol blends
specifically formulated to increase octane to between 96 and 101 research octane number (RON) has
been investigated as a means to reduce GHG emissions by allowing powertrain design changes such as
additional engine downspeeding, increased compression ratio, and/or further downsizing of boosted
engines (i.e., higher boost and maximum brake mean effective pressure levels) with improvements in
protection against abnormal combustion phenomena such as preignition and knocking combustion (Stein
et al. 2013; Leone et al. 2014; Theiss et al. 2016). Vehicles with specific design attributes to take
advantage of higher RON mid-level ethanol blends have yet to be introduced. While current FFVs are
capable of operation on such high RON mid-level ethanol blends, they are currently specifically
designed to allow operation on lower RON E10 fuels and do not have the design attributes necessary to
take full advantage of increased RON fuels. Therefore, potential air quality improvements from broad
adoption of these technologies has not been seen or studied.
3.1.3.2 Impacts from Biodiesel Emissions
3.1.3.2.1. Biodiesel Production
As of mid-2017, there are approximately 119 biodiesel production facilities in operation in the
U.S.63 Emissions are associated with extraction, flaring, boiler operation, and cooling processes. The
most recent emission estimates are found in the 2014 NEI.
3.1.3.2.2. End Use
Renewable diesel and biodiesel blends up to 5% (B5) are fully fungible with petroleum diesel
fuels and meet the ASTM D975 specifications for summer and winter grades of light distillate diesel
63 (1) EPAct/v2/e-89 Phase 3 with 4 FFVs; (2) National Renewable Energy Laboratory, E40
with 9 FFVs (Yanowitz et al.); (3) Coordinating Research Council, E-80 with 7 FFVs (Haskew et al.
2011); and (4) EPA's Office of Research and Development with 2 FFVs (Hays et al. 2013; George et al.
2014).
61
-------
fuel. Heavy-duty diesel engines without catalysts and certified to the 1994 - 2004 heavy-duty emissions
standards were found to have slightly lower PM emissions, but slightly higher NOx emissions with the
use of B5 biodiesel blends (EPA 2010). However, engines equipped with exhaust catalysts (MY2007
and newer for PM; MY2010 and newer for NOx) are not anticipated to experience any significant
impact on criteria pollutant emissions due to use of these fuels, compared to petroleum diesel fuel.64
Diesel engine manufacturers normally consider compliance with in-use emissions requirements
and emissions control system durability prior to approving specific biodiesel blend levels for use in
engines. Use of fuels that contain higher levels of biodiesel than approved by the engine manufacturer's
recommendations could adversely affect the durability of diesel exhaust catalyst systems and result in
significantly higher pollution emissions. Therefore, use of the correct biodiesel blend and emissions
control systems in heavy-duty diesel vehicles is critical to ensuring low emissions needed to meet and
maintain local air quality goals. For example, deviation from recommended biofuel content may result
in increased NOx and secondary nitrate particulate matter emissions due to potassium or sodium ash
impacts on selective catalytic reduction (SCR) systems used for diesel NOx control (Williams et al.
2011). Such deviation may also result in higher ash accumulation within the catalyzed diesel particulate
filter (CDPF). This can result in shorter CDPF ash maintenance intervals in heavy-duty applications or
CDPF plugging in light-duty/light-heavy-duty applications that may not include CDPF cleaning as part
of regularly scheduled maintenance (Brookshear et al. 2013).
3.1.4 Potential for Future Changes in Impacts
As of this report, only limited data are available on the impacts of biofuels on the exhaust and
evaporative emissions from vehicles using advanced gasoline engine technologies (e.g., turbocharging/
downsizing, GDI, and Atkinson/Miller Cycle) to meet current and future light-duty GHG emission
standards. However, major impacts of biofuels on inventories of criteria pollutant emissions from
vehicles with advanced gasoline engines are not anticipated since vehicles complying with the
2017-2025 light-duty GHG standards also must comply with Tier 3 emissions standards.65 Tier 3
64MOVES and Related Models, U.S. Environmental Protection Agency,
https://www.epa.gov/moves/previous-moves-versions-and-documentation.
65 In 2014, EPA finalized Tier 3 light-duty vehicle emissions and fuel standards. Implementation
of Tier 3 began in 2017, with gradual phase-in of more stringent emissions standards from 2017 to
2025 65 The new standards require light-duty vehicles to meet a lower fleet-average tailpipe emissions
standard and per vehicle emissions standards, which represent reductions of 60% and 70%, respectively,
from Tier 2. Exhaust and evaporative emissions under Tier 3 are projected to result in U.S. fleet-average
emissions at approximately the same levels as California partial-zero-emission vehicle requirements.
Tier 3 is expected to reduce emissions of NOx, VOCs, PM2 5 and sulfur dioxide (SO2) emissions by 60 to
62
-------
Box 4. Toxicology Research Related to Biofuels
Since 2011, EPA's Office of Research and Development (ORD) has conducted a series of
studies to examine the potential for adverse biological responses associated with inhalation exposure
to biofuel vapors or emissions from engines using biofuels. The effects of vapors from ethanol-
gasoline mixtures (up to E85) with repeated exposures were tested in several animal models. These
studies observed only mild effects in the exposed rodents at exposure concentrations estimated to be
four- to six-fold greater than those experienced by the general population during fueling operations
(Beasley et al. 2014; Boyes et al. 2014; Oshiro et al. 2014; Bushnell et al. 2015; Oshiro et al. 2015).
Additional research by ORD since 2011 demonstrated that biodiesel combustion emission
exposure - to either 100% biodiesel or a blend in petroleum diesel - can induce biological effects
(Madden 2015; Madden 2016). In order to minimize emissions variability, ORD researchers
conducted multiple exposure studies using the same fuel lot across assays ranging from bacterial
mutagenicity to rodent models of human sensitivity (EPA (2010); Bass et al. 2015; Farraj et al. 2015;
Hazari et al. 2015). The evidence from this work suggests biodiesel emissions can have some similar
effects to petroleum diesel emissions on inflammatory, vascular, mutagenic, and other responses.
There are few findings to date in the available literature on whether repeat-exposure scenarios to
biodiesel emissions can induce human effects or even a weaker response compared to emissions from
petroleum diesel.
Additional research and analyses are needed to adequately understand the potential health
effects of exposure to biofuels and emissions from vehicles using biofuels under real-world
conditions, concentrations, and exposures including to susceptible human populations. It would be
appropriate to study health effects in populations exposed to biodiesel and ethanol blends in
"hotspots," such as fuel production sites, and those exposed to combustion products of biodiesel and
ethanol blends, especially at high blend levels. Such studies could include drivers of vehicles utilizing
those fuels.
70 % and emissions of air toxics, including benzene, 1,3-butadiene, acetaldehyde, formaldehyde,
acrolein, and ethanol, by approximately 10% to 30% relative to Tier 2.
63
-------
standards include specific provisions for emissions compliance using E10 test fuels for non-FFVs and
E85 test fuels for FFVs. Studies similar in scope to the EPAct/V2/E-89 study have yet to be conducted
for light-duty Tier 3-compliant vehicles or for vehicles using advanced gasoline engine technologies to
comply with current and future GHG standards. Such studies would improve understanding of emissions
impacts of biofuels. For example, if advanced engine technologies change the speciation profile of VOC
and PM, the same mass may have a different potential for forming ozone or secondary PM.
Impacts of changes in criteria pollutant levels due to increased biofuel use have the potential to
adversely impact human health (see Box 4). Any alterations in the criteria pollutant concentrations, such
as ozone and PM2 5, that have impacts on expected health outcomes are more fully addressed in the
appropriate Integrated Science Assessment for each pollutant, which summarize the substantial body of
literature on the respective topics.66
3.1.5 Conclusions: Air Quality
• There is no new evidence that contradicts the conclusions of the 2011 Report concerning
air quality. Those conclusions emphasized that life cycle emissions of NOx, SOx, CO,
VOCs, NH3, and particulate matter can be impacted at each stage of biofuel production,
distribution, and usage. These impacts depend on feedstock type, land use change, and
land management/cultivation practices and are therefore highly localized. The impacts
associated with feedstock and fuel production and distribution are important to consider
when evaluating the air quality impacts of biofuel production and use, along with those
associated with fuel usage.
• Ethanol from corn grain has higher emissions across the life-cycle than ethanol from
other feedstocks.
• Ethanol plants relying on coal have higher air pollutant emissions than plants relying on
natural gas and other energy sources.
• The magnitude, timing, and location of all these emissions changes can have complex
effects on the atmospheric concentrations of criteria pollutants (e.g., O3 and PM2 5) and
air toxics, the deposition of these compounds, and subsequent impacts on human and
ecosystem health.
66 Since 2008, EPA's Integrated Science Assessments (ISAs) have formed the scientific
foundation for the review of the National Ambient Air Quality Standards by providing the primary
(human health-based) and secondary (welfare-based, e.g. ecology, visibility, materials) criteria
assessments. See https://www.epa.gov/isa.
64
-------
• Ethanol increased NOx emissions from light-duty vehicles certified to Federal Tier 2
Standards, likely occurring during times when the vehicle catalyst is not yet warmed up
or air/fuel ratio is not perfectly controlled. However, only limited data exist on the
impacts of biofuels on the tailpipe and evaporative emissions of light-duty Tier 3
vehicles and light-duty vehicles using advanced gasoline engine technologies to meet
GHG emissions standards.
3.1.6 Research Needs: Air Quality
• Comprehensive studies of the impacts of biofuels on the emissions from advanced light-
duty vehicle technologies (Tier 3), similar in scope to studies cited in this report for
light-duty Tier 2 vehicles, would improve the understanding of the potential for biofuel-
specific pollutants and associated health impacts as new technologies enter the vehicle
fleet. These studies should consider engine technologies being phased into use for
compliance with current and future light-duty GHG standards, with a focus on vehicles
compliant with the Federal Tier 3 or California LEV III criteria pollutant emissions
standards currently under implementation. Such technologies would include engine
downsizing with addition of turbocharging, gasoline direct injection, and non-traditional
thermodynamic cycles such as Miller or Atkinson.
• Additional research and analyses are needed to adequately understand the potential
health effects of exposure to biofuels and emissions from vehicles using biofuels under
real-world conditions, concentrations, and exposures including to susceptible human
populations. It would be appropriate to study health effects in populations exposed to
biodiesel and ethanol blends in "hotspots," such as fuel production sites, and those
exposed to combustion products of biodiesel and ethanol blends, especially at high blend
levels. Such studies could include drivers of vehicles utilizing those fuels.
• Updated modeling is needed to incorporate improved emissions estimates as laboratory,
field, and other studies lead to a better understanding of biofuel-related emissions
changes and associated changes in the magnitude and composition of pollutants on air
quality, health, and attainment of ambient air quality standards.
3.2 Water Quality
Water quality is adversely affected by the production of biofuel feedstocks, primarily due to the
sediment, nutrients, pesticides, and pathogens directly or indirectly released during different biofuel
65
-------
production phases (e.g., upstream feedstock production, biofuel production, and transportation) (EPA
2003). These releases are dependent on the biofuel feedstock source, the feedstock production site's
management practices, and direct or indirect land use changes associated with feedstock production.
Water quality impacts, in the context of this report, are characterized as either proximal (i.e.,
geographically close to the water body's emission source) or as downstream water quality impacts (with
more distant emission sources). Chemical (e.g., nitrogen, phosphorus) and sediment loadings are the
most significant proximal effects related to biofuel production. Hypoxia and harmful algal blooms are
the most significant downstream water quality impact related to biofuels, which can be found in coastal
and non-coastal waters.
3.2.1 2011 Report Conclusions
The 2011 Report concluded that water quality impacts from biofuels are primarily driven by
chemical inputs at the feedstock production stage (EPA 2011). The Report concluded that effluent
discharge and other already-regulated factors associated with processing biomass into biofuel would
likely have a lesser impact on water quality. At the time of the Report's publication, water quality
impacts from EISA were characterized as negative, particularly due to corn and soybean production
intensification, which was associated with higher levels of erosion and agricultural chemical inputs (e.g.,
nitrogen fertilizer, pesticides). The 2011 Report linked increased fertilizer runoff to eutrophication and
coastal hypoxia, but it also argued that these impacts can be mitigated through conservation practices.
Mitigation efforts, particularly in the Mississippi River Basin, have included the establishment of
loading reduction goals and research on improved drainage strategies and the use of created and restored
wetlands and vegetated buffers. The 2011 Report also suggested that water quality benefits could be
achieved through perennial grass cultivation (e.g., switchgrass, giant miscanthus) on land designated for
row crops. While commercial-scale use of those feedstocks was only a potential at that time, perennial
grass cultivation was argued to have lower chemical inputs and higher utilization efficiencies when
compared to traditional feedstocks like corn and soybeans. Lower chemical inputs and less soil
disturbance may ultimately lead to lower sediment and nutrient losses to the surrounding environment.
The 2011 Report also concluded that water quality, including acreage and function of waters,
was affected by pollutants discharged from biofuel production processes. Different pollutants were
attributed to different biofuel production processes, where biological oxygen demand (BOD), brine,
ammonia-nitrogen, and phosphorus were characterized as primary pollutants of concern from ethanol
facilities, and BOD, total suspended solids, and glycerin were primary pollutants of concern from
biodiesel facility effluent. The 2011 Report noted that explicit impacts resulting from biofuel
production-
66
-------
related pollutants were dependent on a range of factors, including the type of feedstock processed,
biorefinery technology, effluent controls, water re-use/recycling practices, facility location, source water,
and receiving water.
The 2011 Report also pointed to leaks and spills of biofuel from above-ground, underground,
and transport tanks as potential contaminant sources to ground, surface, and drinking water.
Additionally, the Report noted that leaking tanks present increased risk potential for fires and explosions.
The 2011 Report suggested that water contamination via spills and leaks can be minimized by enforcing
existing regulations concerning corrosion protection, leak detection, spill prevention, and overfill
prevention. Additionally, the 2011 Report suggested that biofuel leaks could be prevented by using
appropriate materials, material standards, and/or manufacturer recommendations.
Lastly, the 2011 Report concluded that impacts on surface waters from algal cultivation for
biofuel would depend on the configuration of any eventual production at commercial scales; either
releases of nutrient-rich waters and/or removal of nutrients from existing polluted waters by algae were
considered feasible, which could lead to either more severe or less severe water quality impacts,
respectively.
3.2.2 Drivers of Impacts to Water Quality
The drivers discussed in Chapter 2 (i.e., biofuel volumes, land use, conversion technologies,
agricultural practices) are inherently connected to water quality, including to the acreage and functions
of waters. Direct and indirect water quality impacts attributed to biofuel volumes are dependent on
several biofuel life cycle processes, including but not necessarily limited to: upstream feedstock
production, biofuel production, and transportation. Land use for biofuel feedstock production has direct
water quality impacts, which can include effluents and/or discharges occurring at a feedstock production
site. The application of nutrients, pesticides, and/or other chemical additives for feedstock production
can also ultimately affect the water quality of a feedstock production site or the surrounding area of a
feedstock production site (EPA 2003; EPA 2011).67
As noted in Chapter 2, since the 2011 Report, corn production has intensified on land already
under cultivation, and corn, soybeans, and wheat have expanded to land that was previously
uncultivated. Strong correlational evidence exists that suggests biofuel production contributes to these
changes (Brown et al. 2014; Wright et al. 2017), but we cannot yet quantify how much (see Attribution
67 U.S. Environmental Protection Agency, Water Quality Assessment and TMDL Information,
https://ofmpub.cpa.gov/watcrs 10/attains index.home.
67
-------
Box 3 in Section 2). Despite varying nutrient application and runoff characteristics of these different
practices, direct connections between increased feedstock production and water quality impacts are
beginning to be assessed. Research to evaluate the impacts of increased biofuel production and use on
water quality has largely been based on modeling rather than observed changes. Models enable
evaluation of the change in water quality attributable to biofuel feedstock production, which is an
exceptionally difficult problem to examine by field measurements.
3.2.3 Impacts to Water Quality
In the following sections we examine the proximal water quality impacts (those near the sources
of emissions into water bodies) as well as the impacts to water quality more distant from emission
sources.
3.2.3.1 Proximal effects: Pesticides, sediment, nutrient, and pathogen levels in waters
Corn ethanol and soy biodiesel are currently associated with the highest national production
levels. Due to their high national output, studies since the 2011 Report have evaluated water quality
impacts associated with existing, projected, and/or hypothetical national biofuel production levels for
corn ethanol and soy biodiesel. Several drivers can impact water quality, including the type of feedstock,
management practices at a feedstock production site, and direct or indirect land use changes associated
with feedstock production. Demissie et al. (2012) simulated water quality impacts in the year 2022 for
the Upper Mississippi River Basin based on projected national feedstock production characteristics,
which included: increased corn production, increased wheat production, increased idle land, decreased
soybean production, decreased pasture-hay land, decreased use of conventional and reduced soybean
tillage, and no change in soybean no-till area. While it is not possible to comprehensively evaluate the
accuracy of these assumptions based on the empirical record, short-term trends (2008-2012, see land use
change discussion in Section 2.4) suggest that these assumptions are consistent with observations,
although soybean production may be increasing in this area. Demissie et al. (2012) concluded that
projected feedstock production has mixed effects on water quality, projecting a 12% increase in annual
suspended sediment and a 45% increase in total phosphorus loadings, but a 3% decrease in total nitrogen
loading.
Similarly, Wu et al. (2012a) developed future scenarios of biofuel feedstock production to assess
potential water quality and quantity changes associated with an increase in biofuel production and
converting land to switchgrass production. Garcia et al. (2017) simulated groundwater nitrate
contamination responses associated with nitrogen (N) fertilizer application and increased corn
68
-------
production at a national level (with an emphasis on agricultural areas throughout the U.S.). They
concluded that increased corn production between 2002 and 2022 could result in approximately a 56% to
79% increase in nitrate-N groundwater concentrations.
These studies were based on projected impacts; future work with a focus on observable and
attributable water quality impacts resulting from biofiiels is needed to evaluate the accuracy of those
projections. One instance of such work is the U.S. Geological Survey (USGS) interactive online mapper
that provides results from the largest-ever assessment of water-quality changes in the U.S. (USGS 2017).
The mapper illustrates and provides data for surface water chemistry trends (i.e., nutrients, pesticides,
sediment, carbon, salinity) and aquatic ecology from 1972 to 2012. An example from the mapper is
shown in Figure 21, which presents total nitrogen concentration trends between 2002 and 2012.
This resource unfortunately has little data from many of the hotspots of land use change
identified in Section 2.4 (e.g., South Dakota, North Dakota). However, it does show in the central
agricultural areas that total nitrogen concentrations appear to be declining in Iowa and increasing in
Oklahoma between 2002-2012. Total phosphorus concentrations appear to be decreasing in Iowa and
increasing in Kansas, Oklahoma, and parts of western South Dakota. Future reports could use the USGS
•^Vancouver
Seattle ~
o © °
%
Ottawa Montreal
o . o
Sap
I i ii
^ncisco
Los
Angeles
©
5
-V
GREAT
J* "»«»»**
V..
-1 uYaqo^rfDelrcnl
jdelphia
~
Monterrey
Trend results
~ Likely up
Somewhat likely up
® About as likely as not
',' Somewhat likely down
~ Likely down
Figure 21 USGS mapper tool showing total nitrogen concentration trends between 2002 and 2012.6
m U.S. Geological Survey, Water-Quality Changes in the Nation's Streams and Rivers,
https://nawqatrends.wim.usgs. gov/swtrends/.
69
-------
mapper and other related tools to evaluate the water quality impacts attributable specifically to biofuel
feedstock production.
Estimates of fertilizer increase from biofuel cropland expansion can be deduced from existing
related studies. For example, according to the USDA Economic Research Service, the average nitrogen
fertilizer input rate and the average phosphate fertilizer input rate for corn are approximately 140
pounds/acre and 60 pounds/acre, respectively.69 Lark et al. (2015) estimated that approximately 1.28
million acres of extensification in the U.S. is due to corn. There is also an unknown amount of net
conversion to corn from other crops at a national level, as well as changes in crop rotations to more
continuous corn. Regional studies suggest these unknowns could be significant (Plourde et al. 2013; Ren
et al. 2016). Using the national extensification estimate and nitrogen fertilizer input rates of Lark et al.
(2015), these studies suggest an approximate increase of 170 million pounds of nitrogen fertilizer usage,
with the potential for some of this to eventually reach waterways.
It is important to recognize that there are many factors that affect the fraction of nitrogen, or any
other nutrient or chemical, applied that might reach water bodies. Higher crop yields (bushels per acre)
can take up additional nutrients and conservation measures such as no-till production can reduce loss of
nutrients or chemicals that run off into water bodies (Wade et al. 2015).
Since 2011, studies have quantified and confirmed the findings related to cellulosic biofuels
suggested in the 2011 Report. For example, there have been several studies that have quantified nitrate-
runoff reductions from croplands (VanLoocke et al. 2012; VanLoocke et al. 2017). One model found
that certain scenarios of increased miscanthus production (in favor of 40% corn production devoted to
ethanol) would result in a 6% reduction in dissolved inorganic nitrogen to runoff and streamflow
throughout the drier portions of the Mississippi-Atchafalaya River Basin (VanLoocke et al. 2017). The
collection of corn stover in places with high rates of production (e.g., Iowa) allows implementation of
no-till agriculture, which is known to reduce runoff and improve water quality compared to the
alternative (Dale et al. 2017). Furthermore, switchgrass uses less fertilizer than corn and thus can reduce
adverse water quality effects relative to corn (Parish et al. 2012).
3.2.3.2 Downstream Effects
The 2011 Report noted that biofuel demand-related increases in corn and soybean cultivation
would likely increase nutrient loadings to streams, rivers, and lakes, adding to existing high levels of
69 U.S. Department of Agriculture, Fertilizer Use and Price: https://www.ers.usda.gov/data-
products/fertilizer-use-and-price/.
70
-------
impairment due to eutrophication and affecting the function of the waters (EPA 2011). Eutrophication
impacts to surface waters have included harmful algal blooms (HABs), particularly in fresh waters, and
hypoxia, particularly in coastal waters. Recent modeling studies have continued to conclude that row
crop agriculture plays an important role in driving these downstream impacts, and they continue to
suggest that biofuel feedstock production is a contributing factor. Downstream effects are also driven by
weather patterns, including temperature rises, as well as the timing, amount, and form of precipitation.
Harmful Algal Blooms in Freshwater Systems. A major harmful algal bloom (HAB)
observed in western Lake Erie in 2011 was attributed to unusual weather patterns coupled with long-
term trends in agricultural practices that increase runoff of dissolved reactive phosphorus (DRP)
(Michalak et al. 2013). A modeling study by Michalak et al. (2013) concluded that, if corn acreages
continued to be at recent high levels, along with projected future increases in spring precipitation,
similar events could be more likely in the future.
The main driver of HABs in western Lake Erie is phosphorus (P), particularly from the Maumee
River watershed. Two recent studies indicated that biofuel production could contribute to increased P
loadings to surface waters (LaBeau et al. 2014) and aquatic systems (Jarvie et al. 2015). Modeling
scenarios using the Soil and Water Assessment Tool (SWAT)70 suggest that conservation practices (e.g.,
filter strips, cover crops, riparian buffers) can help achieve total P targets, whereas DRP is much more
responsive to reductions of P application to fields (especially inorganic P). Modeling also suggested that
conversion to perennial grasses such as switchgrass and Miscanthus, even with manure application,
would significantly reduce P runoff into water bodies (Muenich et al. 2016).
While P loadings determine the physical volume of a HAB, N loading appears to play a critical
role in determining bloom composition. The cyanobactcrium Microcystis, which produces the
hepatotoxin microcystin, lacks the N-fixing capability of other cyanobacteria and therefore is favored by
the presence of excess N. The detection of microcystin led to a temporary shutdown of the Toledo, Ohio,
water supply during a Lake Erie HAB in 2014 (Levy 2017). Analyses by Taranu et al. (2017) confirm
that total N concentration in lake water is a much stronger predictor than total P of the probability of
detecting Microcystis in U.S. lakes; the percent of land cover that was agriculture within the ecoregion
of a given lake was also a strong predictor (Taranu et al. 2017). Therefore, while it appears likely that
demand for biofuel feedstocks increases agriculture-related nutrient loadings to surface waters, the
70 http: //s wat .tamu. edu/
71
-------
appearance of HABs, and in particular the prevalence of algal toxins in HAB events, will depend on a
complex interplay of land use, conservation practices, and weather events.
Downstream Effects on Coastal Waters. The size of the Gulf of Mexico hypoxic zone (i.e.,
area with bottom dissolved oxygen < 2.0 mg/L) is a function of climate, weather, basin morphology,
circulation patterns, water retention time, freshwater inflows, stratification, mixing, and nutrient loadings
(Dale et al. 2010). The hypoxic zone size is also a function of loading of nitrate-plus-nitrite from the
Mississippi and Atchafalaya River system during May, as well as the periodic action of tropical storms
to re-aerate the bottom layer (Turner et al. 2016). However, the nature of this relationship is changing - a
given nitrate/nitrite load is causing a larger hypoxic zone in recent years than in earlier years (Figure 22).
Assumptions about future nitrogen loadings from agricultural areas, and the influence of biofuel
feedstock cultivation on those loadings, are critical to the estimation of future impacts. Future scenarios
of increased biofuel production for Europe, simulated through the year 2050 using the Global Nutrient
Export from Watersheds (Global NEWS) model (van Wijnen et al. 2015), suggested that riverine
loadings of N and P would increase as a consequence, resulting in increased risks of HABs and hypoxia
in vulnerable coastal areas. This modeling exercise assumed constant nutrient use efficiencies by crops.
By contrast, in modeling scenarios of future agriculture response to biofuel demand in the
Mississippi River Basin (MRB) using a similar NEWS-derived model, McCrackin et al. (2017) assumed
a 24% improvement in nutrient recovery efficiency over the period 2002-2022 and further assumed that
fertilizer application was matched to crop requirements. In spite of projected 28% increase in corn
plantings over the period, these researchers estimated that dissolved inorganic nitrogen export from the
MRB would decrease by 8%. It should be noted that the assumptions" values used by McCrackin et al.
(2017) may differ from observed values moving forward.
~ Km2 per 1000 mt N as NO
*2016 predic!«l *
100,000 200,000
May Nitrate load (mt N as N03+2) year
Figure 22 Changes in the measured size of the GoM hypoxic zone as related to the amount of nitrate-nitrate loading.
(Turner and Rabalais 2016).
72
-------
3.2.4 Potential for Future Changes in Water Quality Impacts
Recent research has shown that changing precipitation patterns influence water quality. Loecke
et al. (2017) statistically connected drought-to-flood transitions (termed "weather whiplash") to increases
in riverine nitrogen loads and concentrations, and pointed out that these "whiplash" events are projected
to increase in the future. Given that recent studies have connected cellulosic biofuel feedstock production
to relatively lower nitrogen loadings in surface waters, there is potential to decrease the water quality
impact of "weather whiplash" events under specific biofuel feedstock production scenarios.
In addition, cellulosic-based biofuel production could increase in the future, which may impact
water quality. Corn stover is already being used at the POET bioreflnery in Iowa, and studies have
shown that as a perennial, native plant, switchgrass offers several advantageous qualities, including:
drought and flood tolerance; high yield capacity with little to no fertilizer application; the ability to
stabilize soils and sequester carbon with long root systems; and the potential to improve water quality
(McLaughlin et al. 1998; Tolbert et al. 2002; Dale et al. 2014).
3.2.5 Conclusions: Water Quality
• The 2011 Report found that corn production intensification was associated with higher levels of
erosion, chemical loadings to surface waters, and eutrophication.
• Modeling studies since the 2011 Report suggest that demand for biofuel feedstocks, particularly
corn grain, may contribute to harmful algal blooms, as recently observed in western Lake Erie,
and to hypoxia, as observed in the northern Gulf of Mexico.
• Empirical studies documenting cropland extensiflcation and crop switching to more corn suggest
water quality impacts, but the magnitude of these changes is variable across the landscape and so
may be detectable only in some regions.
• Implementation of conservation practices has been observed to result in a decrease of nitrogen,
phosphorus, and soil erosion.
• Changes to future nitrogen and phosphorus loadings will depend on feedstock mix and crop
management practices. Decreases in nitrogen and phosphorus loadings are possible should
perennial feedstocks become dominant.
• Specific biofuel production scenarios expected to improve water quality may help decrease the
water quality impact of predicted future extreme weather events.
73
-------
3.2.6 Research Needs: Water Quality
• Studies are needed of water quality impacts associated with leaks and/or spills from biofuel
production facilities and storage tanks. Such work would address the effectiveness of existing
leak detection and cleanup approaches to address releases to the environment and resulting
contamination plumes.
3.2.7 Opportunities for Future Environmental Improvements
• A decrease of N and/or P loadings is possible should perennial feedstocks become dominant.
3.3 Water Quantity
3.3.1 2011 Report Conclusions
The production of biomass feedstocks and the conversion of those feedstocks to biofuel requires
water resources. The 2011 Report generally interpreted water availability in EISA Section 204 as
referring to water quantity. The report concluded that water use for feedstock production would "likely
not change appreciably if production takes place, as the majority does now, in regions where irrigation is
not needed" (EPA 2011). However, the 2011 Report also noted that water use for feedstock production
could increase under certain conditions. Some of those conditions included expansion of feedstock
production into regions where irrigation is required and cultivation of row crops instead of perennial
grasses with lower irrigation requirements. The 2011 Report also suggested that the water use for
irrigation of feedstocks greatly exceeds the water required for conversion of feedstocks to biofuels.
Water use for biofuel conversion could have localized impacts, depending on facility size and water
reuse, whereas feedstock production covers a larger regional area. Finally, the 2011 Report highlighted
the difficultly in generalizing the impacts of water use on water availability, suggesting that "impacts are
most likely to be adverse in already stressed aquifers or surface watersheds." (EPA 2011).
Since the 2011 Report, several studies have advanced our understanding of the water footprint of
biofuels [see Wu et al. (2014) as a review]. We will first discuss research that has taken a life cycle
assessment (LCA) perspective, starting from feedstock production through conversion to end-use.71
These studies have also examined water use looking at different aspects of water use, where withdrawals
71 We note that life cycle assessment (LCA) in this section focuses only on water use for
biofuel production supply chains. The system boundaries may differ from other LCA studies for other
environmental impacts. For example, studies of the life cycle water use for ethanol production may not
fully account for co-products such as distillers' grains for livestock operations.
74
-------
represent the total water removed and consumptive water use the part of water withdrawn that is
evaporated, transpired, incorporated into products or crops, and not returned to the same watershed.72 A
number of studies also further differentiate the consumptive water use between blue water use (irrigation
water sourced from surface and groundwater and consumed through evapotranspiration [ET]) and green
water use (water from precipitation and soil moisture consumed through ET) in the feedstock growing
stage.73 We will then describe how research has also moved toward more refined spatial analysis of
watersheds when accounting for feedstock-production water use, recognizing the differences across
regions within the U.S. as well as the influence of agricultural management practices. Finally, we will
look at potential future water use impacts related to cellulosic feedstock production and provide bullets
for conclusions and areas for future research.
3.3.2 Drivers of Impacts to Water Quantity
As noted above, the primary driver of impacts to water quantity is the water used for irrigation
of biofuel feedstocks. To the extent that feedstock production expands into regions where irrigation is
required, the demand for water will increase, whether the expansion is a direct consequence of
production specifically for biofuel feedstocks or an indirect result of increased production for all uses.
The question of attribution to biofuel feedstock production was addressed in more detail in the land use
section in Chapter 2. Water demand for biofuel conversion processes can also drive impacts to water
quantity. Although water quantity impacts may be much smaller at a national scale than those related to
feedstock production, they may be locally consequential in areas that are already experiencing stress on
water availability.
3.3.2.1 Feedstock Production
Several highly cited and visible articles compared the life cycle water use of biofuels relative to
petroleum-based fuels on the basis of "gallons of water per mile" or "gallons of water per gallon of fuel."
These early studies characterized this issue as biofuel's water intensity (King et al. 2008), embodied
water (Chiu et al. 2009), and water footprint (Dominguez-Faus et al. 2009; Scown et al. 2011).74 Scown
72 These definitions are consistent with the (U.S. Geological Survey's (USGS) compilation of
data on the nation's water use (see www.usgs.gov/watuse).
73 There is also a grey water category that accounts for the "virtual quantity of water required
to assimilate the pollutant load from the permissible standards down to the natural background
concentration" (Chiu et al. 2012). However, for this report, water quality issues are addressed separately
in Section 3.2.
74 Dominguez-Faus et al. (2009) also characterized the water demands of transportation biofuels
as a "drink or drive" issue, i.e., the water is available for either drinking or for fuel production.
75
-------
et al. (2011) compared different transportation energy sources and found ethanol from corn-based
feedstocks to be one of the most significant uses of freshwater. Calculating the gallons of water
consumed per mile of travel, they found the full life cycle water footprint of corn grain and stover to
ethanol (using average irrigation rates) would require almost seven times as much surface water
consumption as any other transportation power source and an order of magnitude more groundwater
consumption when compared to other transportation energy sources.
Researchers have continued to refine the LCA-based water footprint of biofuels with a focus on
feedstock production for both current biofuels crops and future feedstocks. Because more than 90% of
corn is located in rain-fed areas where corn production is non-irrigated, Wu et al. (2014) suggested that,
at the highly aggregated level, the "national water footprint of corn is consistently low to modest."
However, water quantity demands depend on the crops grown, where they are grown, and how they are
grown. In terms of differences among feedstocks, Dominguez-Faus et al. (2009) calculated the irrigation
water required for corn-based ethanol at an average of approximately 600 liters of water per liter of
ethanol-equivalent (liter/liter) and soybean biodiesel at 1300 liter/liter. Sorghum, used in some primarily
corn-based ethanol facilities, was estimated to have irrigation water requirements of roughly 1500
liter/liter [see supporting information in Dominguez-Faus et al. (2009)].
Where and how crops are grown also matter because irrigation rates for the same crops can vary
enormously, from no irrigation in rain-fed acres in the Midwest to high irrigation rates in more arid
regions in the West. Dominguez-Faus et al. (2013) calculated a range of irrigation water use for corn
ethanol between 350 and 1400 gal/gal. They estimated that if 20% of corn production was used to
produce 12 billion gallons per year of ethanol in 2011 (irrigated at a weighted average of 800 gal/gal),
that would amount to 1.8 trillion gallons (7 trillion liters) of irrigation water withdrawals per year. While
not an insignificant amount, it represents only 4.4% of all irrigation withdrawals (Dominguez-Faus et al.
2013). Other researchers have similarly focused on the wide range of water intensity estimates between
rain-fed and irrigated acreage and among a variety of crops (see Figure 23). Gerbens-Leenes et al. (2012)
76
-------
7,000
a Regional range
• County max
-National average
5.000
j us;
1.000
IMP
B ue Water
Green Water
Grey Water
Figure 23 An estimate of the blue, green and grey water footprint associated with corn grain stover, wheat straw
and soybean during the crop growing phase. The national production-weighted average is represented by the
horizontal bar, while the regional ranges (USDA regions including the Corn Belt, Southern Plains, etc.) are
represented by the shaded bars. County-level variation in feedstock water footprints, shown in dashed lines, are
driven by differences in irrigation and evapotranspiration (ET). [Source: Cliiu and Wu (2012)].
estimated Nebraska's blue water (irrigation) footprint at three times higher than the U.S. weighted
average blue water footprint. Many other corn producing states have minimal irrigation demands relative
to Nebraska. Yet, it should be noted that after Iowa, Nebraska is the second largest producer of corn-
based ethanol in the U.S., with 25 active ethanol facilities, many concentrated in southern Nebraska (EIA
2017). Moreover, higher irrigation demands may coincide with areas of already-stressed surface and
groundwater resources, such as the Ogallala Aquifer. A report by the National Academy of Sciences
(NAS 2011) highlighted the groundwater drawdown in the Ogallala Aquifer, noting that Nebraska is
"among the states with the largest water withdrawals for irrigation, and its usage has continued to
increase in recent years, largely driven by the need to irrigate corn for ethanol/' This suggests that the
majority of groundwater consumption would come from areas like Nebraska that are already impacted
by over-pumping due to their high blue water footprint for corn production (Gerbens-Leenes et al. 2012).
3.3.2.2 Biofiiel Conversion
Studies of water use for biofuels conversion facilities have generally quantified water
consumption as gallons of water per gallon of biofiiel produced, with most of the focus on ethanol,
especially dry mill facilities. Process level engineering studies and surveys of biofuel facilities (Mueller
2010a) have shown declines in water requirements from 5.8 gallons of water per gallon of ethanol
77
-------
(gal/gal) in 1998 to 2.7 gal/gal in 2012 (Wu et al. 2012b). Anecdotal evidence75 also points to decreases
in the water intensity of ethanol facilities through more efficient water use, water recovery, and use of
treated wastewater for processes such as fermentation or possibly cooling towers. There are no recently
published surveys of water consumption representing all current ethanol facilities, and there are no
comprehensive data on the type of water sources utilized (e.g., groundwater, surface freshwater, public
supply, etc.).
3.3.3 Changes in Relationships between Drivers and Impacts
Because of the need to better understand the variation in feedstock energy use and the actual
impacts on local water resources, since the 2011 Report, researchers have moved toward watershed
hydrological modeling (see Table 5). A number of studies have projected future water use for scenarios
of higher cellulosic feedstock production, discussed below in Section 3.3.5.
Because the majority of the growth in biofuels production has come from corn grain-based
biofuels, the water consumption impacts to date would have come from additional water use for corn and
soybean acreage. To our knowledge, there have been no studies of the changes in irrigated acres, rates of
Table 5. Methods and metrics used to assess the water quantity impacts of biofuels.
Method / metric Definition Example studies
LCA / Water
footprint
(e.g., gallons of
water consumed per
gallon of biofuel)
Volume of water used in a biofuel production pathway. It
can include blue (irrigation) and green (rainfall) water.
Calculating gallons of water per gallon of biofuel requires
data or assumptions regarding feedstock water
consumption, feedstock yields, and biofuel conversion
rates. This water footprint can also be compared to other
fuel/energy pathways.
Chiu et al.
(2012); Scown et
al. (2012)
LCA / Water stress
index
(e.g., index from
0.01 to 1.00)
The share of water consumed that is considered to be no
longer available for downstream users. This is used in LCA
to show the impact of water consumption on water
resources, usually at a more local level, focusing on areas
such as drought-prone regions.
Pfister et al.
(2014)
Watershed modeling
/ Streamflow
(total flow as m3 s"1
or % change)
The rate of water flow measured or modeled at a watershed Cibin et al.
outlet. This can be reported as predicted stream flow under (2016); Housh et
biofuel production scenarios or as percent changes in al. (2015)
stream flow.
75 http://www.ethanolproducer.com/articles/8860/dropping-water-use
78
-------
irrigation, or changes in surface and groundwater supplies associated specifically with the increased
production of corn grain-based ethanol or soybean-based biodiesel. The land use section (Chapter 2)
highlights analyses in Lark et al. (2015) and Wright et al. (2017) that show changes in land use,
including cropland expansion in the western Dakotas and Kansas, which are areas unlikely to have
sufficient precipitation for corn growth. However, there are no similar analyses that explicitly attribute
recent changes in management practices, such as irrigation, to increased biofuels production and
feedstock demand.
USDA Farm and Ranch Irrigation Surveys provide a general indication of the changes in water
demands. From 2007 to 2012, there was a decrease in total irrigated acres of nearly 0.8 million acres in
the U.S. The USDA notes that "most of the area decline occurred in the Western U.S. where drought
conditions contributed to water-supply scarcity across the region" and that irrigation area is not static,
but dynamic, across the U.S.76 Over the same time period, irrigated acres of corn for grain and seed
increased from 12.0 million acres to 13.3 million acres harvested, along with a higher irrigation rate of
1.1 acre-feet applied in 2012 compared to 1.0 acre-feet applied in 2007 (USDA 2013). Irrigated corn
grain/seed acres are heavily concentrated in Nebraska (5.4 million acres) followed by Kansas (1.5
million acres) (see Figure 24), up by 6% and 10% respectively from 2007 levels.
Changes in irrigation practices are dependent on a number of economic and agronomic factors
that affect how land is managed, making it difficult to attribute expanded irrigation to biofuels
production and use without more detailed analysis. That said, studies of land use change rates have noted
that "along the Ogallala Aquifer, elevated rates of land use change to corn production in Western
Kansas, Oklahoma and Texas coincided with areas experiencing groundwater depletion rates ranging
from 5-20% per decade" (Wright et al. 2017) (see Figure 25). Because of the potential impact on surface
and groundwater resources, further studies of both land use change and land management practices
should examine the linkages between increased biofuel feedstock production and changes in irrigation
demands. Moreover, this work should have a particular focus on water stressed areas such as the
Ogallala Aquifer.
76 https://www.ers.usda.gov/topics/farm-practices-management/irrigation-water-use/background.aspx
79
-------
Acres of irrigated land, 2012
¦ 1 dot= 10,000 acres
Source: USDA, National Agricultural Statistics Service, Map Atlases for the
2012 Census of Agriculture.
Figure 24 Acres of irrigated land in 2012, based on the USDA Farm and Ranch Irrigation Survey.
Source: https://www.ers.usda.gov/topics/farm-practices-management/irrigation-water-
use/baekground.aspx
There also have been advancements in understanding the drivers of water use for cellulosic
biofuel feedstocks. For cellulosic feedstocks, given the small amounts of crops such as switchgrass or
miscanthus actually in production, any assessed water use impacts will be based more on modeling
studies or research and experimental scale production, rather than on widespread commercial production
levels. We also caution that results from these studies depend on aggregate feedstock scenarios and
simulations, compared to what can be observed empirically for corn-based ethanol production and
changes in water demand and stress. Impacts will depend on which cellulosic feedstocks are grown,
where and how they are grown, which best management practices are followed, technological change in
irrigation practices, and potential changes in rainfall and air temperatures due to a changing climate (Le
et al. 2011; Dominguez-Faus et al. 2013; Ha et al. 2017).
80
-------
+• y
. * Shio
Indiana .
¦:* vv,
A*,#
, \ . * -; , 4-''^
W\r -. > * / -
-?A L^iL-rfi ".
v/rg/n/t)
TerwM.see •..* J-T rto/tfi CtotoflW
' Oklahoma
Arkansas
ipPlAlabama
- mESs~
1 J.v^lfcS
I I Ogallala Aquifer
Non-crop to crop (%)
I
1 T
200 400
,4
.8
1.2 2.0 2.7 3.9 5,5 7.8 12.1 35
Figure 25 Relative conversion rates of arable non-cropland to cropland (2008-2012), including
conversion located along the Ogallala aquifer. Stars denote biofuel production facilities.
(Source: Wright et al. 2017)
3.3.4 Potential for Future Changes in Impacts to Water Quantity
Studies have examined the water-use implications of removal of corn stover and for future
scenarios of perennial feedstocks (Demissie et al. 2012). Switchgrass, Miscanthus, and forest wood have
high evapotranspiration (ET) rates, longer growing seasons, and therefore higher green water
requirements, which can be important if looking at how these feedstocks could affect the broader
hydrologic cycle when produced at a large scale. However, because of lower irrigation requirements,
they are anticipated to have a smaller blue water footprint. Some studies (Demissie et al. 2012; Wu et al.
2012b; Cibin et al. 2016) have coupled multiple scenarios of feedstock production with watershed
models, such as SWAT, to translate projected changes in land use/management - driven by demand for
81
-------
biofuel feedstocks - into changes in water demands. These watershed models often focus on changes in
streamflow (see Table 5), along with indicators of water and soil quality (see Sections 3.2 and 3.5). Cibin
et al. (2016) used a modified SWAT model (Trybula et al. 2015) that included improved representation
of perennial bioenergy crops like Miscanthus and switchgrass and data from research plots. The model
was used to assess the impacts of 13 biofuel scenarios for two watersheds in the Midwest. Cibin et al.
(2016) found slight reductions in stream flow under biofuel production scenarios, ranging from 0.2% to
4.5%, with somewhat greater reductions for Miscanthus. They highlight that water use and water quality
(e.g., nutrient removal) trade-offs need to be assessed carefully and that even with some reductions in
stream flow "Miscanthus and switchgrass production may be a strong candidate for implementation in
watersheds that would generally benefit from sediment and nutrient load reduction and can sustain base
flows during drought conditions" (Cibin et al. 2016). These and other studies (Housh et al. 2015)
indicate that while energy crops can reduce nitrate run-off they can reduce sub-surface water flows and
streamflow. During low flow periods or drought, these reductions can have negative impacts on aquatic
and riparian ecosystems. Thus, water quantity and water quality effects have potential trade-offs that
should be carefully assessed in any cellulosic scenario.
Advances have also been made in development of publicly available tools for assessing future
feedstock scenarios. An online web-based model WATER (Water Analysis Tool for Energy
Resources)77 characterizes county level water footprint for biofuel produced from corn, soybean, wheat,
perennial grasses, and forest wood residue via various conversion processes for the U.S. The model
presents a geospatial distribution of water consumptions (blue, green, and grey water) under historical
and future land use scenarios. A number of studies have been based on this model. Most recently, DOE
(2017) has underscored the importance of appropriate land management planning and choice of
feedstock mix, including use of non-agricultural based feedstocks. Under a highly optimistic scenario of
high-yields and production for both agricultural and wood-based feedstocks, and shifts toward non-
irrigated perennial crops, they suggest that the states in Ogallala Aquifer region could actually reduce
irrigation water consumption if planned and managed carefully (DOE 2017).
In terms of biofuel-processing water use, cellulosic ethanol facilities are anticipated to be more
water intensive at first, ranging from 6 to 10 gallons of water per gallon of ethanol, primarily based on
process engineering studies (Davis et al. 2015). Looking ahead, data collection could better quantify
water use efficiency for existing and future cellulosic biofuel conversion facilities, along with their water
77 Access to the WATER model is available at: http://water.es.anl.gov/
82
-------
source and water demands relative to local water availability, particularly for potential hot spots of high
water demands in water stressed areas.
3.3.5 Conclusions: Water Quantity
• As discussed in the 2011 Report, the irrigation of corn and soybeans grown for biofuels is the
predominant water quantity impact. Water use for feedstock production is significantly greater
than water use in the biofuel conversion process.
• There are indications of increased water consumption in irrigated areas for corn between 2007
and 2012 and elevated rates of land use change to corn production in more arid Western states
including the Ogallala region. Adverse water availability impacts will most likely arise in
already stressed aquifers and surface watersheds.
• Irrigation practices are dependent on a number of economic and agronomic factors that drive
land management practices making attribution of increased irrigation and water quantity to
biofuels difficult.
3.3.6 Research Needs: Water Quantity
• Studies are needed to determine the extent to which increases in water consumption and
withdrawals - due to changes in land use/management change - can be attributed to feedstock
production for biofuels.
• In particular, studies should continue to explore increases in water demands that have occurred
or are occurring along water-stressed areas, both for surface and groundwater.
• Research, both modeling and field work to verify modeling parameters, is needed to better
understand future cellulosic feedstock water demands while assessing water quantity, water
quality, and soil quality in an integrated manner.
3.3.7 Opportunities for Future Environmental Improvements
• Priority should be placed on identifying effective strategies to manage withdrawals in "hot
spots" (e.g., Ogallala aquifer) where high water demands and water stress are coinciding.
• While cellulosic feedstocks such as perennial grasses can provide environmental benefits for
biodiversity and ecosystem services, their potential impact on streamflow within a watershed
should be carefully considered.
83
-------
3.4 Ecosystem Health and Biodiversity
Diverse biological communities are crucial to establishing and maintaining healthy ecosystems,
as each species fulfills unique and necessary roles for maintaining ecosystem function. Ecosystem health
can also be viewed in terms of resilience - the ability to resist external stressors over relevant temporal
and spatial scales. The 2011 Report concluded that a variety of environmental factors related to biofuel
production can affect ecosystem health and biodiversity, including changes to land use and land
management, especially with regard to crop management and runoff from nutrients, pesticides, and
sediment. Furthermore, the 2011 Report noted that overuse or misuse of management techniques can
impact biodiversity and ecosystem health far beyond the confines of the farm field, potentially causing
lasting impacts up and down the production chain (EPA 2011).
3.4.1 2011 Report Conclusions
The 2011 Report addressed terrestrial and aquatic biodiversity and ecosystem health with respect
to grasslands, forests, wetlands, and impacts to aquatic systems (EPA 2011). In general, biofuel
feedstock production was found to negatively impact biodiversity through loss of habitat, often in
sensitive areas, and especially if idled lands in the Conservation Reserve Program (CRP) (with
established conservation covers) were to be returned to crop production. Quantitative data linking
biofuel production and ecosystem health remain sparse. Most of this information is qualitative in nature
and often regional, and as such, merits broader research. The following paragraphs detail some of the
most supported conclusions from the 2011 Report, including topics such as grasslands, forests, and
feedstock management.
Grasslands are at the forefront of conversations surrounding biofuel production landscapes.
Conversion of grasslands to row crops has been found to displace species reliant on grassland habitats
whereas retaining some grasslands for perennial grass feedstocks could mitigate this loss of habitat (EPA
2011). Furthermore, using grasslands for buffer zones could reduce erosion and runoff, thereby reducing
the likelihood of exposure to nutrients, pesticides, or other chemicals at levels above those determined to
be protective.78 The 2011 Report also noted that, while some feedstock management practices exacerbate
the release of sediment, nutrients, pesticides, and pathogens into downstream waters, other more
conservation-based practices (e.g., constructed or restored wetlands) can increase habitat availability for
certain freshwater species. Releases or discharges with high concentrations of nutrients, total suspended
78 Conservation Buffers to Reduce Pesticide Losses, USDA Natural Resources Conservation
Service, March 2000 http://www.nrcs.usda.gov/Internet/FSE_DOCUMENTS/nrcsl43_023819.pdf
84
-------
solids (TSS), and other contaminants can decrease ecosystem health and lead to fewer sensitive species
in affected bodies of water, especially in areas where streamflow is already low (EPA 2011).
In terms of terrestrial impacts linked to biofuel production, the 2011 Report focused on how
changes to forest harvests could affect biodiversity. The 2011 Report noted that shortening the harvest
interval for short rotation woody crops and residue harvesting could decrease habitat availability and
biodiversity, while moderate thinning could increase species diversity and abundance for certain species.
However, given that the United States does not yet have woody biomass-based feedstock in production
at the commercial scale, this topic is left for future research.
3.4.2 Drivers of Impacts to Ecosystem Health and Biodiversity
As concluded in the 2011 Report, ecosystem health and biodiversity are impacted by
environmental factors, such as changes in land use and land management, including cropland
extensification, cropland conversions and intensification, and nutrient, pesticide, and sediment runoff
(EPA 2011). In addition to these site-specific factors, some environmental health indicators are
associated with the spatial and structural arrangement of different habitat types across the landscape. The
location of biofuel production and types of management practices employed affect the ecosystem
impacts and potential mitigation opportunities.
Wright et al. (2017) reported that approximately 2 million acres of grassland within the standard
draw of a biorefinery plant (50 miles) were converted to row crops between 2008 and 2012. Smaller
acreages were reported as converted to row crops for forests (60,000 acres), shrublands (52,000 acres),
and wetlands (14,000 acres). Figure 26, modified from Wright et al. (2017), illustrates the geospatial
distribution of these conversions. The bulk of the grassland conversions occurred in South Dakota
(348,000 acres), Iowa (297,000 acres), Kansas (256,000 acres), Missouri (239,000 acres), Nebraska
(213,000 acres), and North Dakota (176,000 acres). The conversion reported by Wright et al. (2017)
explicitly included only lands that had not been in cropland for at least 20 years, so although they may
not represent pristine habitats they are expected to represent habitats in a relatively natural state.
85
-------
A
B
FcraM In crop (M
0 04 1,2 2T 78 100
Wrtljnd tpocf (M
c
5 hr ubijud *> Cf op {%)
0 \Z 3.5 8.2 21.2 1M
Figure 26 Relative conversion rates to cropland of (a) grassland, (b) forest, (c) shrubland, and (d) wetland from
2008 to 2012. Each rate is relativized by type of ecosystem witliin a 3.5-mile spatial grid (modified from Wright et
al. 2017).
The row crop expansion and intensification is correlated with higher chemical inputs, including
fertilizer and pesticides (Meehan et al. 2011; Meehan et al. 2015). Each of these classes of chemicals has
the potential to impact ecosystem health and biodiversity (Malaj et al. 2014; Chagnon et al. 2015). Corn
receives substantial levels of fertilizer, while soybeans typically are not fertilized. Corn and soybean, the
dominant feedstocks for biofuels currently, are both treated with herbicides, fungicides, and insecticides.
Neonicotinoid seed treatments (including thiamethoxam, clothianidin, and imidacloprid) are widely
applied to biofiiel feedstock crops, including com and soybeans.79 Approximately 90% of corn (Douglas
et al. 2016) and 30% of soybean fields planted during 2008-2012 contained neonicotinoid seed
treatments.86 Detection of neonicotinoids in aquatic systems in regions of high corn and soybean
79 Pesticides can be sold or distributed only after EPA approval and users must follow directions
specified on the pesticides label to ensure safe use. Prior to being approved for use, EPA assesses a wide
variety of potential human health and environmental effects associated with use of each pesticide, based
on scientific data pertaining to the composition, potential adverse effects, and environmental fate of each
pesticide. See https://www.epa.gov/pesticide-registration/about-pesticide-registration.
m Benefits of Neonicotinoid Seed Treatments to Soybean Production (2014). U.S. EPA
memorandum, Office of Chemical Safety and Pollution Prevention.
https://www.epa.gov/sites/production/files/2Q14-
10/documents/benefits of neonicotinoid seed treatments to soybean production 2.pdf.
86
-------
production has raised concerns about the effects of neonicotinoids on aquatic communities and
ecosystems (Hladik et al. 2014; Hladik et al. 2015; Miles et al. 2017). Proper application can reduce the
risks of adverse environmental effects. For example, EPA's risk assessment for imidacloprid found the
lowest overall aquatic risk profile for aquatic invertebrates when using seed treatments, although risks
were still identified with some use scenarios. Alternatively, soil application was found to exceed the
acute risk level of concern for over half of the agricultural and non-agricultural use scenarios modeled
(EPA 2016b).
3.4.3 Impacts to Ecosystem Health and Biodiversity
3.4.3.1 Grassland Birds and Ducks
Widespread changes in land use for biofuel production (e.g., the conversion of environmentally
sensitive land to cropland) have negative impacts to ecosystem health and biodiversity [see review by
Immerzeel et al. (2014)]. The production of other forms of energy also have negative impacts on
ecosystem health and biodiversity. As noted above, assessment of the environmental impacts of other
energy sources is beyond the scope of this report. The type and severity of impacts depend on factors
such as crop type, geographic location, and management practices. For example, degradation and loss of
grasslands has been found to adversely affect grassland bird populations (Meehan et al. 2010; Fletcher et
al. 2011; Robertson et al. 2011; Robertson et al. 2012; Blank et al. 2014; Werling et al. 2014; Evans et
al. 2015). Studies of the effects of bioenergy feedstock production suggest that grassland bird species of
conservation concern are more adversely affected by increased corn production than are more common
species of birds (Fletcher et al. 2011; Blank et al. 2014). Similarly, the loss of wetlands to row crops and
related production practices is associated with reduced duck habitat and productivity of duck food
sources, including aquatic plants and invertebrates (Gleason et al. 2011; Wright et al. 2013). Increasing
grassland cover by planting perennial grasses, including biofuel feedstocks, and replacing marginal
croplands can also enhance ecosystem services, including pollination and biological control (Bennett et
al. 2014a; Bennett et al. 2014b; Werling et al. 2014; Landis et al. 2017).
3.4.3.2 Pollinators
Pollinators, such as wild and commercial bees, are also affected by land use changes (e.g., forage loss
from grassland conversion to corn and soybeans), among other pressures (National Research Council
2007). Commercial bees in the Midwestern and Great Plains states experience implications for honey
production and colony health (Koh et al. 2016; Otto et al. 2016; Smart et al. 2016). Based on model
87
-------
estimates, these states include counties with large areas of pollinator-dependent crops and low levels of
bee abundance (see Figure 27) (Koh et al. 2016). The expansion of corn and soybeans results in
landscape simplification that exacerbates insect pest pressure and is linked to increased use of
insecticides, such as neonicotinoids (Meehan et al. 2011; Meehan et al. 2015). Exposures of pollinators
to neonicotinoids has been evaluated for the potential to cause adverse impacts on pollinators (Krupke et
al. 2012; Krupke et al. 2015; EPA 2016b; EPA 2017) and other non-target organisms (Bonmatin et al.
2015; Pisa et al. 2015; van der Sluijs et al. 2015). Neonicotinoids also travel through the soil food chain
and detrimentally affect beneficial arthropods, disrupt biological control of crop pests, and reduce
soybean yields (Seagraves et al. 2012; Douglas et al. 2015a; Douglas et al. 2016).
The concerns about possible impacts to pollinators due to neonicotinoid exposure have been
widely discussed in recent years. While it is beyond the scope of this report to review those studies,
Godfray et al. (2014) summarize the issues well.
"There is clear evidence of the great value of neonicotinoids in agriculture as
well as the importance of the ecosystem services provided to agriculture by managed
and wild pollinators. Pollinators also have intrinsic importance as components of natural
biodiversity that cannot, or can only inexactly, be accorded economic value. In some
cases, intelligent regulation of insecticide use can provide 'win-wins' that improve both
agricultural and biodiversity outcomes but in other cases there will be trade-offs, both
within and between different agricultural and environmental objectives. Different
stakeholders will quite naturally differ in the weightings they attach to the variety of
>
» Status
f - so
fifB
Ik n
" abundance
Ort oopfefc-xj
Figure 27 Map of (A) wild bee status and (B) status of wild bee supply vs. demand for pollination services
(summed area of animal-pollinated crops, weighted by respective pollinator dependence) across coterminous U.S.
(modified from Koh et al. 2016). The yellow border lines in B identify counties that have high acreage of pollinator
dependent crops (y axis on the legend figure) but low bee abundance in crop land (x axis on legend figure).
Stslus
88
-------
objectives affected by insecticide use, and there is no unique answer to the question of
how best to regulate neonicotinoids, an issue that inevitably has both economic and
political dimensions."
In their review, Godfray et al. (2015) concluded that there is evidence of adverse impacts to
pollinators due to neonicotinoid exposure but that the evidence is mixed, with several studies reporting
no effects of exposure. They note that major gaps remain in our understanding of "how pollinator
colony-level (for social bees) and population processes may dampen or amplify the lethal or sublethal
effects of neonicotinoid exposure and their effects on pollination services" (Godfray et al. 2015).
EPA's preliminary assessment of the risk to bees from imidacloprid, clothianidin, and
thiamethoxam found on-field risk to be low for these pesticides applied to corn, which is the dominant
use pattern for this crop (EPA 2016b; EPA 2017). For other biofuel crops (e.g., soybeans), risks were
considered uncertain at the time and are currently undergoing re-evaluation by EPA with the submission
of additional exposure and effects data. Neonicotinoids, like all pesticides, are approved for use under
specific conditions that are designed to protect ecosystems and human health. EPA has expanded its
pesticide risk assessment process specifically for bees to quantify or measure exposures and relate them
to effects at the individual and colony level.81
Crop intensification also influences the population dynamics of pollinators and pest organisms.
Loss of milkweed in agricultural fields from the increased adoption of herbicide-tolerant corn and
soybeans and related usage of herbicides negatively affect monarch butterfly populations in Midwest
landscapes (Pleasants et al. 2013). Extensive adoption of transgenic corn and intensification are reported
as primary drivers of the resistance of western corn rootworm, Diabrotica virgifera virgifera LeConte,
and corn earworm, Helicoverpa zea Boddie, to multiple Cry proteins expressed in Bt corn (Dively et al.
2016; Gassmann et al. 2016; Jakka et al. 2016). Fausti (2015) noted that U.S. biofuel energy policy was a
key contributor for rapid Bt corn adoption in U.S. corn production system. Through analysis of causal
relationships, Fausti (2015) reported that 7-9% of the increase in Bt corn acres was induced by biofuel
policies. The evolution of insect resistance to transgenic corn varieties, given the increased adoption of
transgenic corn and intensification, hence, is an ecological impact that may be partially associated with
increased biofuel feedstock production. To help manage western corn rootworm resistance, EPA recently
announced enhancements to its long-standing requirements for companies that supply Bt corn to
81 How We Assess Risks to Pollinators, U.S. EPA, https://www.epa.gov/pollinator-
protection/how-we-assess-risks-pollinators; accessed March 11, 2018.
89
-------
implement integrated pest management programs in the Corn Belt, including measures such as crop
rotation to an alternate non-corn rootworm host crop (typically soybean).82
3.4.3.3 Aquatic Ecosystems
The effects of biofuel crop production on aquatic ecosystems are understudied in comparison to
terrestrial ecosystems, partly due to a lack of monitoring data on aquatic species (Immerzeel et al. 2014).
Crop expansion in the ecologically-sensitive Prairie Pothole region resulted in significant loss of
wetlands and their associated biodiversity (aquatic plants and invertebrates) and ecosystem services,
such as surface water flow, groundwater recharge and reduction in sedimentation (Gleason et al. 2011;
Wright et al. 2013; Johnston 2014). Models predict that changes in hydrologic and sediment generation
dynamics through land use change - mainly conversion to row crops - may extirpate native mussel
populations due to shifts in river ecology in the Minnesota River Basin (southern Minnesota, parts of
South Dakota, Iowa, and North Dakota) (Foufoula-Georgiou et al. 2015).
Increased applications of the pesticides imidacloprid and atrazine resulting from corn and
soybean expansion/intensification have also been shown to have aquatic ecological effects. EPA (2016b)
generated acute and chronic risk quotients (RQs) for aquatic organisms by modeling risks to aquatic
organisms from agricultural uses of imidacloprid.83 Aquatic invertebrates were correlated with the
greatest risk from imidacloprid, where foliar spray and combination applications resulted in acute RQs
ranging from 1.6 to 44, with the chronic RQs ranged from 39 to 2130, which are above the level of
concern for acute and chronic risk (0.5 and 1.0, respectively). An aquatic exposure assessment for
atrazine combined modeling approaches and monitoring data to estimate atrazine occurrence in surface
water at different spatial scales (EPA 2016c). The report found that, on an acute exposure basis, atrazine
is moderately toxic to freshwater and estuarine/marine fish, highly toxic to freshwater aquatic
invertebrates, and even more toxic to estuarine/marine aquatic invertebrates. Effects on survival, growth,
and/or reproduction were also shown from chronic exposure studies for freshwater and estuarine/marine
fish, aquatic phase amphibians, and aquatic invertebrates. The risks from atrazine application and
82 US EPA: Regulation of Biotechnology under TSCA and FIFRA. Framework to delay corn
rootworm resistance, https://www.epa.gov/regulation-biotechnology-under-tsca-and-fifra/framework-
delav-corn-rootworm-resistance.
83 EPA uses a deterministic approach or the quotient method to compare toxicity to
environmental exposure. In the deterministic approach, a risk quotient (RQ) is calculated by dividing a
point estimate of exposure by a point estimate of effects. This ratio is a simple, screening-level estimate
that identifies high- or low-risk situations. See https://www.epa.gov/pesticide-science-and-assessing-
pesticide-risks/technical-overview-ecological-risk-assessment-risk.
90
-------
exposure to biota are not limited to corn and soybeans used for biofuels. Attributing these risks to
specific crops faces the same attributional challenges as other endpoints (see Attribution Box 3, Section
2.4). Nevertheless, since over 80% of atrazine use is for corn, according to USGS,84 and some fraction of
corn production is attributable to biofuels, a direct link is present even if not explicitly quantified.85
3.4.4 Key Points from Recent Literature
Recent literature has emphasized: (1) impacts to biodiversity and ecosystem health due to the
conversion of environmentally-sensitive lands; (2) the loss of ecosystem services, such as groundwater
recharge, reduction in sedimentation, nutrient cycling, biological control of crop pests, and pollination;
and (3) the need for better environmental data collection and monitoring (Newbold et al. 2015; Landis et
al. 2017). Field studies and simulation models report that increased corn and soybean production often
leads to the loss of grasslands (native mixed and tallgrass prairie), which negatively impacts ecosystem
services, including pollination, biological control of crop pests, and nutrient cycling. Increasing
grassland cover by planting perennial grasslands as biofuel feedstock and by replacing marginal
croplands can enhance biodiversity and ecosystem services (Bennett et al. 2014a; Bennett et al. 2014b;
Werling et al. 2014; Koh et al. 2016; Landis et al. 2017). Modeling approaches that incorporate both
economic factors and ecosystem services frameworks, such as those popularized by the Millennium
Ecosystem Assessment,86 may be of benefit to better understand how impacts to ecosystem health and
biodiversity can affect the broader economy.
3.4.5 Potential for Future Changes in Impacts to Ecosystem Health and Biodiversity
Given the trends in biofuel feedstock production and technology development discussed in
Chapter 2, relatively few near-term changes in direction are expected for biodiversity and ecosystem
health. A decrease in CRP lands could lead to decreases in biodiversity and terrestrial ecosystem health.
This affects habitat availability as well as species diversity and abundance. More effective agricultural
management practices would reduce incidences of sedimentation and eutrophication in streams and
rivers, and increase streamflow levels. Reduced use of insecticides and genetically engineered species
could lead to an increase in biodiversity. Increased adoption of pollinator conservation practices could
84 USGS National Water Quality Assessment Program: Pesticide National Synthesis Project.
State-level pesticide use estimates by major crop and crop groups.
https://water.usgs.gov/nawqa/pnsp/usage/maps/countv-level/.
85 US EPA. 2017. Refined Ecological Risk Assessment for Atrazine. External Review Draft.
Environmental Fate & Effects Division, Office of Pesticide Programs, Washington. DC.
86 http://www.millenniumassessment.org/en/Framework.html.
91
-------
benefit pollinators and other beneficial insects.87 Such changes, however, would require widespread
coordination and adoption to be effective at regional and/or national scales.
Future improvements to and expansion of conservation practices related to biofuel feedstock
production could play a major role in reducing the magnitude and severity of impacts to ecosystem
health and biodiversity. For instance, the use of precision agriculture, as well as guidance systems, has
become more prevalent over the past 20 years.88 Increased adoption of fertilizer technologies such as
time-release and other enhanced-efficiency fertilizers, alternative fertilizer placement methods, and
variable-rate application could further improve nutrient-related conservation practices in biofuel
feedstock production.
3.4.6 Conclusions: Ecosystem Health and Biodiversity
• Loss of grasslands and wetlands is occurring in ecologically sensitive areas, including the Prairie
Pothole Region.
• Loss of habitat and landscape simplification are associated with negative impacts to pollinators,
birds, soil-dwelling organisms, and other ecosystem services in both terrestrial and aquatic
habitats.
• Increased fertilizer applications of nitrogen and phosphorus have negative effects on aquatic
biodiversity.
3.4.7 Opportunities for Future Environmental Improvements: Ecosystem Health and Biodiversity
• Planting perennial grasslands and replacing marginal croplands with perennial grasslands can
enhance ecosystem services.
• Increased use of effective conservation practices can have multiple benefits, from reduced
stream sedimentation and nutrient runoff to protection of pollinator habitat.
• Increased adoption of other technologies such as time-release and other enhanced-efficiency
fertilizers, alternative fertilizer placement methods, and precision agriculture could further
improve conservation practices in biofuel feedstock production.
87 Using 2014 Farm Bill Programs for Pollinator Conservation (2015). USDA Biology Technical
Note No. 78 (2nd Ed.)
https://directives.sc.egov.usda.gov/OpenNonWebContent.aspx?content=3737Q.wba.
88 USDA Economic Research Service (2017). Tailored Reports: Crop Production Practices;
https://data.crs.usda.gov/rcports.aspx'.)ID= 17SS3
92
-------
3.4.8 Research Needs: Ecosystem Health and Biodiversity
• Studies that target the interactive effects of land use change and feedstock production could help
identify impacts to specific organisms.
• Research on the efficacy of methods to expand pollinator habitat in agricultural systems can
improve understanding of appropriate methods and their potential tradeoffs for different
agricultural areas.
3.5 Soil Quality
The production of biofuel feedstocks can also affect soil quality, which is the capacity of a soil
to function.89 The EPA's 2011 Report focused on soil erosion, soil organic matter (SOM), and soil
nutrients as general indicators of soil quality (EPA 2011). Soil erosion can impact soil quality by
preferentially removing the finest soil particles at the soil surface that are generally higher in organic
matter, plant nutrients, and water-holding capacity than the remaining soil. Soil organic matter is critical
to soil quality because it provides plant nutrients and water, promotes soil structure, and reduces erosion,
while also sequestering carbon from the atmosphere (Sparks 2003).90 Lastly, soil nutrients (e.g.,
nitrogen, phosphorus) are necessary for plant growth. Too little of these nutrients can reduce crop yields;
too much can lead to eutrophication of waterways via runoff or leaching.
3.5.1 2011 Report Conclusions
Overall, the 2011 Report concluded biofuel feedstock production could either negatively or
positively affect soil quality depending upon the feedstock used, the particular land converted, and the
management of the feedstock. For corn and soybeans, environmental effects were estimated to be most
negative if these crops were produced on former CRP land or other relatively unmanaged grasslands.
Conversely, the soil quality effects were estimated to be minimal if the feedstocks were produced on
89The USDA Natural Resources Conservation Service defines soil quality as "The capacity of a
specific kind of soil to function, within natural or managed ecosystem boundaries, to sustain plant and
animal productivity, maintain or enhance water and air quality, and support human health and habitation.
In short, the capacity of the soil to function" (USDA-NRCS 2017). Here, soil conservation and soil
environmental quality, as listed in Section 204 of the 2007 Energy Independence and Security Act
(EISA), are subsumed under this broader heading of soil quality. The term soil quality in this section is
used as a general term, independent of area—it is used both to describe effects on single soil types and
cumulative effects across large areas and multiple soil types.
90Soil organic matter is defined by Brady et al. (2000) as "The organic fraction of the soil that
includes plant and animal residues at various stages of decomposition, cells and tissues of soil
organisms, and substances synthesized by the soil population." The main elemental constituents of SOM
are carbon (52-58%), oxygen (34-39%), hydrogen (3.3-4.8%) and nitrogen (3.7-4.1%) (Sparks 2003).
93
-------
land currently in corn and soybeans. Effects could be moderated by the use of conservation practices
(e.g., no-till management). For corn stover, high removal rates were found to increase erosion and loss of
SOM. Soil quality impacts of perennial grasses and woody biomass were estimated to be largely
determined by the type of land converted, as in the case for annual feedstocks, whereas algae production
was considered to have minimal-to-no-effects on soil quality.
In the following sections, we revisit these 2011 soil quality conclusions. First, we provide a brief
overview of the major changes in the drivers of soil quality (feedstocks, type of land-converted, and
production practices) and their impacts since 2011. We address these by feedstock type. Second, we
highlight a few changes in our understanding of the connections between drivers and impacts since 2011.
Third, we focus on potential future changes, and, finally, we provide a bulleted list of conclusions. As in
2011, we discuss effects on soil erosion, SOM, and soil nutrients, while also acknowledging it may be
advantageous to add other soil quality indicators in future reports.
3.5.2 Drivers of Impacts to Soil Quality
Corn-grain ethanol and soy biodiesel account for most of the biofuel volumes produced to date.
As a result, almost all the soil quality impacts from biofuels thus far are from the production of corn and
soybeans. Since the 2011 Report, new evidence adds support to the understanding that grasslands,
including CRP grasslands, have been converted to corn and soybeans (see Section 2.4). Biofuels are
responsible for a proportion of this change, although, as noted in Box 3, the percentage attributable to
biofuels cannot now be quantified with confidence, nor can the resulting effects on soil quality be
quantitatively attributed to biofuel production. In general, however, grassland-to-annual-crop conversion
negatively impacts soil quality because it increases erosion and the loss of soil nutrients and SOM,
including soil carbon (Gregorich et al. 1985; Gelfand et al. 2011; Qin et al. 2016; Yasarer et al. 2016).
These in turn increase sediment and nutrient loadings to waterways and carbon loss to the atmosphere
(Lai 2003; Yasarer et al. 2016).
A couple of factors can mitigate—to an extent, but not entirely—the negative soil quality
impacts of this land use conversion. First, the type of CRP land, conservation lands, or other grasslands
converted can affect soil quality. In a modeling study, LeDuc et al. (2017) simulated that more erosion
and loss of soil carbon and nitrogen occurs from converting low productivity, highly sloped CRP
grasslands compared to those with higher productivity soils and lower slopes. Second, the effects will
also depend upon production practices. Most corn and soybeans are grown using conservation tillage,
94
-------
with a smaller percent grown using no-till management (USDA-NRCS 2010; Wade et al. 2015).91
Conservation tillage, including no-till, reduces soil erosion and increases SOM content relative to
conventional tillage (Cassel et al. 1995; West et al. 2002). Use of conservation tillage practices can
partially mitigate the effects of converting CRP areas or other grasslands to corn or soybeans (Follett et
al. 2009; Gelfandetal. 2011).
Since 2011, evidence has become stronger that corn and soybeans are replacing other cropland,
not just CRP land and other grasslands (see Section 2.4). The soil quality impacts of converting to corn
or soybeans from other crops, such as wheat, are generally less than those of the conversion of
grasslands (Zuber et al. 2015; Qin et al. 2016; Yasarer et al. 2016). Zuber et al. (2015) observed similar
soil effects of no-till, continuous corn rotations, and corn-soybean-wheat rotations on high organic
matter, fine textured soils. From this evidence, they suggest a movement from wheat to corn may not
materially affect soil quality, provided a shift from no-till to conventional tillage does not occur
concomitantly. Qin et al. (2016), in a meta-analysis, found that corn replacing other cropland (e.g.,
soybean, wheat) increased soil organic carbon, whereas the opposite occurred when corn replaced
grassland or forest land. Notably, the percent increase in soil organic carbon of other-cropland moving to
corn was exceeded in magnitude by the percent decrease in soil organic carbon by the conversion of
grassland-to-corn (Qin et al. 2016).
In contrast to corn grain and soybeans, the use of other feedstocks for biofuels has been much
more limited. Since 2011, at least two commercial-scale corn stover ethanol plants have started
operations - the DuPont and POET-DSM corn stover plants in Iowa (EPA 2016d). Partial stover
removal can increase corn yields in some locations, in part by reducing nitrogen uptake from the soil by
microorganisms and potentially by increasing soil temperatures in no-till systems (Coulter et al. 2008;
Karlen et al. 2014). Yet too much stover removal can increase soil erosion, decrease SOM and soil
nutrients, and ultimately decrease corn yields as noted in the 2011 Report. Whether corn stover can be
harvested sustainably, and at what removal rate, depends on many site-specific factors, including yields,
topography, soil characteristics, climate, and tillage practices (Karlen et al. 2014). In a study across
multiple locations in seven states, stover harvesting slightly increased corn grain yields, although the
91Conservation tillage is defined as any tillage practice leaving at least 30% of the soil surface
covered by crop residues; whereas conventional tillage leaves less than 15% of the ground covered by
crop residues (Lai 1997). No-till management, a subset of conservation tillage, disturbs the soil
marginally by cutting a narrow planting strip. Nationally, approximately 30% and 45% of the area
planted to corn and soybeans, respectively, are under no-till (Wade et al. 2015). Since 2000, there has
been a general trend toward greater percent residue remaining after planting for both crops (USDA-ERS
2018 https://data.ers.usda.gov/reports.aspx?ID= 17883; data accessed 2/15/2018)
95
-------
authors cautioned against extrapolating these results to other sites and noted the need to conduct site-
specific planning with soil testing (Karlen et al. 2014). DuPont limits harvesting stover to corn fields in a
no-till or conservation tillage system, with yields of 180 bushels per acre or higher, and on relatively flat
land with a slope of four percent or less (DuPont 2017). POET-DSM recommends harvesting only 20-
25% of stover on slopes less than four percent, coupled with soil testing to monitor soil nutrients and
SOM (POET-DSM 2017). These criteria are designed to encourage stover harvest at sustainable rates
and locations, although additional research is needed to understand effects on soil quality if these criteria
are followed.
In contrast, perennial grasses, woody biomass, and algae generally have not been used yet as
biofuel feedstocks at the commercial scale, with a few exceptions (e.g., algal biofuels for the U.S. Navy;
Ziolkowska et al. 2014). Therefore, there have not been major changes in the drivers and soil quality
impacts from these feedstocks since the 2011 Report.
3.5.3 Changes in Relationships Between Drivers and Impacts
Since 2011, research has improved our understanding of the relationship between drivers and
soil quality impacts. On the negative side, the scientific literature suggests there may be a relationship
between no-till management and the loss of nutrients such as phosphorus to waterways (e.g., Jarvie et al.
2017; see Water Quality Section). On the positive side, a recent study suggests the carbon benefits of no-
till corn may have been previously underestimated due to a failure to account for carbon accrual at
greater soil depths (Follett et al. 2012). For corn stover, recent research has focused on the use of cover
crops, manure, or biochar to add organic matter to the soil to compensate—at least partially—for the
organic matter removed (Blanco-Canqui 2013).92 The scientific literature continues to emphasize that
perennial grasses or woody biomass grown on marginal lands (e.g., abandoned agricultural land) can
help restore soil quality [e.g., Blanco-Canqui (2016)]. Notably, however, effects of these perennial
feedstocks can depend upon the plant species grown and the type of land converted (Robertson et al.
2017), and literature definitions of what constitutes marginal land and estimates of its extent vary widely
(Emery et al. 2016). Finally, the 2011 Report concluded algae production would have minimal-to-no-
effects on soil quality. It is possible, however, that some of the algal residues following oil extraction
could be used as a soil amendment, increasing soil carbon content (Rothlisberger-Lewis et al. 2016).
92Biochar is the product of heating biomass in the absence of or with limited air, with the
resulting material rich in organic carbon (Lehmann et al. 2015). This material can be used as a soil
amendment.
96
-------
3.5.4 Potential for Future Changes in Impacts to Soil Quality
It is likely that corn and soybeans will be the predominant biofuel feedstocks grown in the near
future, which is expected to continue to put pressure on soil resources. Leaps forward in biotechnology
and/or increasing yields may ameliorate some of these concerns (e.g., Brusamarello-Santos et al. 2017).
The two new, commercial-scale, corn stover plants now in operation could signal a beginning of a corn
stover industry, and the use of soil amendments, such as biochar, may be expanded to counterbalance
organic matter removal in both agricultural and forest soils (Blanco-Canqui 2013; Scott et al. 2016).
Should the large-scale production of perennial grasses or woody biomass become economically viable as
feedstocks, they would fundamentally alter effects on soil quality, most likely positively if grown on
marginal lands or lands with soils otherwise limited by physical or chemical problems (Blanco-Canqui
2016).
3.5.5 Conclusions: Soil Quality
• Corn-grain ethanol and soy biodiesel account for most of the biofuel volumes produced to date.
As a result, almost all the soil quality impacts from biofuels, thus far, are from the production of
the dominant conventional feedstocks.
• Conversion of grasslands to annual cropland typically negatively affects soil quality, with
increases in erosion, and the loss of soil nutrients and soil organic matter, including soil carbon.
Impacts of this conversion can be partially mitigated - though not entirely - through the
adoption of management practices such as conservation tillage.
• The soil quality impacts of converting other crops to corn or soybeans are generally less than
those of the conversion of grasslands. The production of corn on existing cropland can provide
soil carbon benefits, although these benefits are outweighed on a per area basis by the negative
effects of grassland conversion.
• Overall, these land use trends suggest that negative impacts to soil quality from biofuel
feedstocks have increased since 2011, but this has not been quantified and the magnitude of
effects depends predominantly on the relative areas of grasslands converted versus existing
croplands attributable to biofuels.
• Corn stover is now being harvested at the commercial-scale in Iowa, and the scientific literature
indicates this must be done carefully to avoid negatively affecting soil quality and crop yields.
97
-------
3.5.6 Opportunities for Future Environmental Improvements: Soil Quality
• Alternative biofuel feedstocks, such as perennial grasses and woody biomass, are not yet used at
commercial scales. Studies have shown that these feedstocks can improve soil quality relative to
current conditions, contingent on species grown and type of land converted (e.g., marginal,
abandoned, or degraded lands).
3.5.7 Research Needs: Soil Quality
• Quantitative estimates of the cumulative soil quality effects are needed for the land use changes
described in Section 2.4 and the proportion attributable to biofuel feedstock production (this
includes both the conversion of land to corn and soybeans and the management practices
implemented).
3.6 Invasive Species
The National Invasive Species Council defines an invasive species as "with regard to a particular
ecosystem, a non-native organism whose introduction causes or is likely to cause economic or
environmental harm or harm to human, animal, or plant health."93 In the context of biofiiels, and similar
to the 2011 Report, this report also includes additional characteristics of species for evaluating
environmental impacts and invasiveness.
3.6.1 2011 Report Conclusions
The 2011 Report noted that biological traits of some plant species and perennial grasses favored
as biofuel feedstocks overlap with those of high invasion potential (fast growing species that form dense
stands, efficiently use resources, tolerate broad environmental conditions and perturbations, are disease
and pest resistant, and are able to disperse and establish widely) (EPA 2011). The 2011 Report listed
several mitigation options for reducing the potentially negative environmental impacts from perennial
grass production. Prominent options included conducting a weed risk assessment (WRA) and rejecting
planting species or varieties that are predicted to be invasive. Under the RFS requirements and in
collaboration with USDA, EPA examines invasion risk WRA and includes further regulatory
requirements (e.g., a Risk Mitigation Plan) as needed to reduce the invasion potential and other negative
93 Executive Order 13751, "Safeguarding the Nation from the Impacts of Invasive Species,"
December 5, 2016.
98
-------
environmental impacts. The 2011 Report also concluded that corn, soybean, and perennial grasses, such
as Giant Miscanthus, pose little invasive species risk (EPA 2011).
3.6.2 Drivers of Impacts to Invasive Species
The renewable volume obligations through 2016 have principally been met with renewable fuels
from corn grain (ethanol) and soybean (biodiesel) as feedstocks, and these crops do not pose a risk of
invasion in the U.S. The invasiveness of these crop species has not altered since the 2011 Report. To
date, no cases of invasive corn or soybeans have been reported in natural areas in the U.S.
Recent studies have linked the increased adoption and extensive cultivation of corn and soybean
that are genetically engineered to resist glyphosate and the widespread application of this herbicide (see
Chapter 2) to development of glyphosate resistance in 15 weed species in total (Benbrook 2012; Heap
2014; Benbrook 2016; Myers et al. 2016). This results in increased alternative herbicidal treatments and
higher active ingredient application per unit area, which further increases evolutionary pressure for
resistance (Benbrook 2016; Myers et al. 2016) and other environmental impacts. Resistance to particular
herbicides provides a fitness advantage to weeds in areas where direct or indirect exposure to those
herbicides occurs, which may effectively enhance the potential invasiveness of weed species in certain
habitats.
For other feedstocks, reports highlight the invasion potential of Panicum virgatum L.
(switchgrass) in areas where the species are non-native within the U.S. For example, reports predict that
in California, where it is potentially invasive, switchgrass could establish successfully in disturbed
riparian areas (Barney et al. 2012). Comparison of different switchgrass cultivars with the wildtypes in
Ohio and Iowa showed some cultivars performing better, supporting the need for further assessments
prior to large-scale planting for biofuels (Palik et al. 2016). Previous reports also noted that giant
miscanthus (Miscanthus x giganteus) posed little risk of invasion because it is sterile and propagated by
cutting (Heaton et al. 2010; Gordon et al. 2011). However, not allM it giganteus cultivars are sterile.
Spatial demographic models indicate that sterile and fertile cultivars of M x giganteus have substantially
different invasive potential. Whereas frequent and severe habitat disturbances are predicted to raise
invasion risk for feral populations of sterile M. x giganteus, fertile cultivars would likely be difficult to
contain (Matlaga et al. 2013). Comparison of many noninvasive and invasive species of the genus
Miscanthus in Virginia and Georgia showed thatM x giganteus is less likely to be invasive in
conventional agricultural fields subject to tillage or herbicide applications (Smith et al. 2014). Recent
results suggest that potentially invasi\q Miscanthus species could become established outside of
99
-------
cultivated areas, but a lag in any impacts on receiving communities presents a widow of time for
management (West et al. 2017).
3.6.3 Potential Changes in Relationships Between Drivers and Impacts
Since the 2011 Report, EPA has approved pathways for feedstocks using Camelina sativa
(Camelina), Saccharum spp. (energy cane), Arundo donax (giant reed), and Pennisetum purpureum
(napier grass) (EPA 2013d; EPA 2013c). For the highly invasive giant reed and napier grass (Gordon et
al. 2011; USDA 2012), approval requires a risk mitigation plan demonstrating these species will not pose
a significant likelihood of spread beyond the intended planting area.94 Additional registration, reporting,
and record keeping requirements to address potential invasiveness are also required (EPA 2013c).
Energy cane is a hybrid of different Saccharum spp. As S. spontaneum is on the Federal Noxious Weed
List,95 it is excluded as a potential feedstock although hybrids derived from S. spontaneum, developed
and publicly released by USDA (Bischoff et al. 2008) are included in this definition of the energy cane
feedstock. Among other approved species, the risk of invasion by Camelina sativa in the northern Great
Plains region is low (Davis et al. 2011) based on a two-tiered approach (incorporating demographic
models to field-estimated parameters in addition to weed risk assessment). These feedstocks are not yet
used for commercial scale production of biofuels, which precludes monitoring and assessment of any
additional invasion impacts. Other potentially invasive feedstocks that have been analyzed by EPA with
respect to lifecycle greenhouse gas emissions include Thlaspi arvense (pennycress), Jatropha curcas
(Jatropha), and Brassica carinata (Carinata),96 but these have not yet been approved for RIN-generating
renewable fuel production. Weed risk assessments of other potential biofuel species conclude that
Eucalyptus camaldulensis and Eucalyptus grandis have high potential to become invasive in the U.S.
(Gordon et al. 2011).
Because these advanced-generation biofuel feedstocks are not now used for commercial scale
production of biofuels, the invasive impacts remain a potential, rather than current, risk. Thus, the full
94 U.S. Environmental Protection Agency, Office of Transportation Air Quality. Approved
Pathways for Renewable Fuel. Policies and Guidance, https://www. epa.gov/renewable-fuel-standard-
program/approved-pathwavs-renewable-fuel.
95 U.S. Department of Agriculture, Natural Resources Conservation Service. Introduced,
Invasive, and Noxious Plants, https ://plants.usda.gov/iava/noxious
96U.S. Environmental Protection Agency, Office of Transportation Air Quality. Other actions for
the renewable fuel standard program. Policies and Guidance, https://www.epa.gov/renewable-fuel-
standard-program/other-actions-renewable-fuel-standard-program.
100
-------
invasive impacts of newly approved feedstocks and others with a completed lifecycle analysis remain
unclear and unknown.
3.6.4 Potential for Future Changes in Impacts to Invasive Species
Recent genetic engineering of biofuels feedstock trees such as Populus spp. (poplar) and grasses,
such as P. virgatum (switchgrass) andMiscanthus spp., have focused on improving the conversion of
cellulosic feedstocks to ethanol. There have also been efforts to engineer improvements of both oil yield
and oil quality for biodiesel from crops including Glycine max (soybean), Brassica napus (canola), C.
sativa, and J. curcas (NAS 2016). So far, there are no data to indicate that these modifications are
changing the invasiveness of the engineered crops. Nevertheless, there is well documented potential for
gene flow to native populations or indigenous species from trees, grasses, and crucifers used to produce
biofuels. Thus, crop protection genes that are engineered or bred into such feedstocks, along with those
genes for improvement of other crop qualities, may also be introduced into recipient populations,
depending on the ecological and management context of the species involved (DiFazio et al. 2012;
Gressel 2015; NAS 2016; Chang et al. 2018). Future impacts from invasive species remain to be
determined.
Studies suggest methodological advancements and strategies, including modifications to weed
risk assessments, for improved evaluation of the invasion risk of biofuel crops (Davis et al. 2011; Hulme
2012; Lewis et al. 2014; Quinn et al. 2015a). This includes a 'white-list' approach for policy decisions
on incentivizing the cultivation of promising new feedstocks without increasing the probability of non-
native plant invasions in natural systems (Quinn et al. 2015a). Studies also point to shortcomings in the
regulatory framework for weed management and stress the need to incorporate insights from other
commercial industries (horticulture, forestry, agroforestry) to inform strategies to reduce environmental
impacts due to invasive biofuel feedstocks (Richardson et al. 2011; Quinn et al. 2013; Quinn et al.
2015b).
3.6.5 Conclusions: Invasive Species
• Biofuels are primarily produced in the forms of bioethanol and biodiesel derived from food
crops (i.e., non-invasive first generation biofuels - corn and soy). Hence, current production of
biofuel feedstocks poses little risk of invasion, consistent with findings in the 2011 Report.
• Weed risk assessments, part of the formal biofuel regulatory process, provide information on
invasion risk and are designed to inform protective management of species and varieties that are
predicted to be invasive.
101
-------
• Increased cultivation of crops engineered for herbicide tolerance (e.g., glyphosate) and
concomitant application of the herbicide has led to a widespread increase in the number of
glyphosate-resistant weed species.
• Potentially invasive species approved as feedstocks require risk management actions under
current RFS requirements. However, invasive species are not presently being used for
commercial scale production of biofuels.
3.6.6 Research Needs: Invasive Species
• Methodological advancements for weed risk assessments and lessons from other industries (e.g.,
horticulture) should be incorporated to inform on potential invasiveness of biofuel feedstocks.
• Modeling and field work are needed to investigate the impacts of gene flow between novel
feedstock varieties (genetically engineered, selectively bred, or a combination) and local natives.
3.7 International Impacts
3.7.1 2011 Report Conclusions
The 2011 Report showed that in the global context, biofuel demands have direct and indirect
impacts for biofuel-producing countries as well as those importing agricultural commodities. Potential
environmental impacts included direct and indirect effects from land use change and impacts on air
quality, water quality, and biodiversity. This section focuses on the potential environmental impacts in
foreign countries from implementation of the RFS2 standards in the United States. Simulations prepared
for the RFS2 projected that the EISA biofuel targets could alter U.S. and international trade patterns and
commodity prices (EPA 2010). The manner in which countries respond to U.S. market conditions,
including influences from deforestation, and biofuel feedstock crop expansion and intensification could
affect net GHG savings derived from biofuels.
The 2011 Report anticipated import volumes to be very low in years preceding 2015, followed
by a significant increase in import volumes between 2015 and 2022. Similarly, it anticipated a decrease
in exports of corn and soybeans for agricultural or other uses, probably resulting in land use change
through conversion to agriculture in other countries, and other environmental impacts. As with biofuel
production in the U.S., these impacts depend largely on where the crops are grown, forest and
agricultural management practices and technologies used, and the efficacy of environmental policies.
Therefore, land use changes in other countries and other environmental impacts could not be quantified.
102
-------
However, the 2011 Report noted that, if natural ecosystems are converted to cropland, the environmental
impacts may be more severe.
3.7.2 Drivers of International Impacts
The volume and location of U.S. imports and exports, both of biofuel and displaced agricultural
goods through international trade, affect the severity of the direct and indirect land use impacts.
3.7.2.1 Trends in Annual U. S. Imports
Ethanol imports have decreased significantly since 2012 (see Table 6), as predicted in the 2011
Report, likely due to increased U.S. ethanol production, limitations on U.S. demand, and other economic
and policy factors. Actual import volumes were much lower than the estimates from the 2011 Report
(compare with Figure 5-2 in 2011 Report). Brazil has been the dominant source of ethanol between 2011
and 2016, and the overall decrease in imports is largely due to the decrease in imports from Brazil.
Before 2015, significant volumes of ethanol were reported as being imported from other countries in
South America and the Caribbean; however, these volumes were likely produced in Brazil and imported
through the Caribbean Basin Initiative. Although the EIA does not publish ethanol import data by
feedstock, the vast majority of the ethanol imported from Brazil was likely produced from sugarcane.97
Table 6. Annual U.S. ethanol imports by country of origin (million gallons)1
Year Brazil2 Other Latin Canada EU Total
America3
2011
101
69
2
-
172
2012
404
82
4
4
494
2013
322
50
5
-
377
2014
56
11
5
2
74
2015
88
-
3
-
92
2016
36
-
1
-
36
1. Source: https://www.eia.gov/dnav/pet/pet move impcus a2 nus epooxe imO mbbl a.htm. 4/28/17
2. Volumes of ethanol imported from Brazil have demonstrated substantial variability since 2011; it is unclear
whether the decreasing trend since 2011 will be maintained or will return to the higher values observed in
earlier years.
3. Other Latin America includes: Ecuador, Argentina, Costa Rica, El Salvador, Guatemala, Jamaica, Nicaragua
and Trinidad and Tobago
97 USDA, Brazil Biofuels Annual 2017,
https://gain.fas.usda.gov/Recent%20GAIN%20Publications/Biofuels%20Annual Sao%20Paulo%20AT
O Brazil 9-15-2017.pdf.
103
-------
In contrast, biodiesel imports have increased in recent years (see Table 7), reaching almost 700
million gallons in 2016. Imports from Argentina, which are likely soybean oil biodiesel, more than
doubled from 2015 to 2016, reaching almost 450 million gallons. The second largest import country of
origin in 2016 was Indonesia, where palm oil is the dominant feedstock. Although EPA has not approved
a pathway for the production of palm oil biodiesel that would meet the minimum 20% lifecycle
greenhouse gas reduction requirement, Indonesian biodiesel imports may include grandfathered volumes
that are nevertheless eligible as conventional biofuel under the RFS program. Imports from Canada,
which reached almost 100 million gallons in 2016, are likely from a combination of canola/rapeseed oil,
soybean oil, and waste oils, such as used cooking oil and inedible tallow.98
Table 7. Annual U.S. biodiesel imports by country of origin (million gallons)1
Argentina
Canada
EU
Indonesia
Other2
Total
2011
0
11
5
0
4
20
2012
0
18
10
0
8
36
2013
132
45
88
52
25
342
2014
52
71
8
59
3
192
2015
196
61
3
72
21
353
2016
444
98
25
102
24
693
1. https://www.eia.gov/dnav/pet/pet move impcus a2 nus EPOORDB imO mbbl m.htm. 4/28/17.
2. Other: Australia, Korea, Panama, Singapore and Taiwan.
Over 200 million gallons of renewable diesel were imported from Singapore in both 2015 and 2016, as
shown in Table 8. These volumes were likely drop-in renewable diesel produced through hydrotreating
of fats and oils, including waste and vegetable oils. The increase in biodiesel and renewable diesel
imports" could have resulted in direct and indirect land use changes and other associated environmental
impacts in some of the trading nations.
98 USDA, Canada Biofuels Annual 2017,
https://gain.fas.usda.gov/Recent%20GAIN%20Publications/Biofuels%20Annual Ottawa Canada 8-9-
2016.pdf.
99 Renewable diesel and biodiesel, which differ chemically, are both included as 'Biomass-based
diesel.' Non-ester renewable diesel is produced through hydrotreating, thermal conversion or biomass-
to-liquid, and can be used in its pure form, or as an additive. Biodiesel (mono-alkyl esters) is produced
using a transesterification process. For the RFS program implementation, EPA utilizes 'Equivalence
Values' based on energy content (renewable diesel -1.7 & biodiesel - 1.5) for determining RIN
generation.
104
-------
Table 8. Annual U.S. renewable diesel imports by country of origin (million gallons)1
Aruba Finland Singapore
2012
2
14
9
2013
6
36
164
2014
0
9
111
2015
0
0
205
2016
0
0
223
1. https://www.eia.gOv/dnav/pet/pet_move_impcus_a2_nus_EPOORDO_imO_mbbl_m.htm, 4/28/17.
3. 7.2.2 Trends in Annual U.S. Exports
Corn exports reduced from about 61,000 metric tons in 2007 to 20,000 metric tons in 2012.
However, since 2012 exports steadily increased to about 56,500 metric tons in 2016. Exports of brewers
and distillers dregs and waste, sometimes known as distillers dried grains with solubles (DDGS), have
also been on the rise since 2012 (see Figure 28). Soybean oilseed exports were similar, ranging between
35,000 to 41,000 metric tons during 2007-2012, but have increased since 2012 (see Figure 29).
U.S. Corn & DDGS Exports (1000 MT)
80000
60000
40000
20000
0
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
¦ Corn
Figure 28 Trends in annual metric tons of U.S. exports of corn and brewers' and distillers' dregs and waste
(DDGS).92
105
-------
80000
60000
40000
20000
US. Soybean Exports (1000 MT)
I Oilseed, Soybean ¦ Meal, Soybean
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Figure 29 Trends in annual metric tons of U.S. exports of soybeans
100
U.S. ethanol exports varied during 2010-2013, with a peak in 2011. Since 2013, exports have
increased from 15 million barrels to 25 million barrels (see Figure 30). For biodiesel, exports increased
until 2013 to 4.6 million barrels, then decreased to 2 million barrels in 2014, holding a similar trend until
2016 (see Figure 31).
30000
25000
o 20000
O
*—I
— 15000
10000
5000
0
I
2010
2011 2012 2013 2014
Year
2015
2016
Figure 30 Trends in annual U.S. ethanol exports1
100 Source USDA
h ttps://apps.fas, usda.gov/psdonl inc/app/indc\.html#/app/do\\nloads?tabNamc=dcfault 6/2/17
101 Source EIA
http://tonto.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=M EPOOXE EEX NUS-
Z00 MBBL&f=A 6/2/17
106
-------
5000
4500
4000
o 3500
S 3000 h
l l
I II I I I
2011 2012 2013 2014 2015 2016
Year
1 02
Figure 31 Trends in annual U.S. biodiesel exports
Broadly, corn, soybean, ethanol, and biodiesel exports were the same or higher compared to
2007 or 2011, correlating with the increase in US ethanol and biodiesel production. A probable
exception was the year 2012, which could be attributed to the drought conditions and yield decreases
(see Chapter 2). These data differ from the projections in the 2011 Report, which expected declines in
ethanol and biodiesel exports.
3.7.3 Changes in Drivers of International Impacts
Reports indicate that in the past decade market-mediated land use impacts (both direct and
indirect land use changes) occurred due to demands for biofuel stocks. These changes probably resulted
in decreased forest and pasture lands, crop intensification and multiple cropping, and depletion of global
phosphorous reserves (Hein et al. 2012; Timilsina et al. 2012; Hertel et al. 2013; Langeveld et al. 2014;
Tokgoz et al. 2014; Babcock 2015). As reported earlier in Section 2.5, forest loss is reported in countries
exporting biofuels to the United States. Expansion and intensification of soybeans are observed in
Argentina in relation to increased biofuel production, coinciding with loss of native grasslands (Solomon
et al. 2015). The use of soybean as livestock feed is also reported as driving the expansion of soybeans
into native grasslands (Modernel et al. 2016). In Indonesia, forest loss (driven in part by demand for
biodiesel) and increased multiple cropping for palm oil are reported (Wicke et al. 2011; Langeveld et al.
2014). Market-mediated land use changes are reported, and it is likely that increased biofuel production
102 Source http://www.eia.gov/dnav/pet/pet move expc a epoordb eex mbbl a.htm 6/2/17
107
-------
has contributed to these land use changes (as noted in Section 2.5). Quantification and causal attribution
of these forest losses and other land use changes and environmental impacts due to biofuel production
and renewable fuel standards remain uncertain and are an area for further research (see Section 2.5.2).
3.7.4 Potential for Future Changes in International Impacts
Biofuels made from more sustainable grasses or woody crops using higher-yield cellulosic
technologies, or from waste biomass or biomass grown on degraded and abandoned agricultural lands,
have been promoted as causing less environmental damage and having less impact on agricultural lands.
If fuels from these feedstocks reach production at scales large enough to meet a substantial fraction of
global fuel demand, the international environmental and natural resource impacts would be considerably
less than those from current technologies. Global supply of such feedstocks remains only a small fraction
of first-generation biofuels.
Efroymson et al. (2016) challenge the current practice of economic simulation models that
incorporate commodity trade and coarse land cover data as causal pathways for land use change due to
biofuel production, and they propose a comprehensive causal analysis framework. Beginning with the
definition of the change that occurred, the causal analysis framework put forth by Efroymson et al.
(2016) utilizes a strength-of-evidence approach that incorporates mechanistic plausibility of relationship,
completeness of causal pathway, spatial co-occurrence, time order, analogous agents, simulation model
results, and quantitative agent-response relationships. Other reports also indicate that complex analyses
combining economic simulation models, place-based empirical studies, value chain analyses, and
biophysical accounting could help infer causal mechanisms and quantification of such land use impacts
(Hertel et al. 2013; Meyfroidt et al. 2013). Beyond causal analysis, Sanchez et al. (2012) argue that
establishing a comparable and coordinated framework for estimating land use changes across the major
biofuel trading nations could better inform policy outcomes.
3.7.5 Conclusions: International Impacts
• Since the 2011 Report, U.S. ethanol imports decreased, while biodiesel and renewable diesel
imports increased, leading to potential land use change impacts in countries of origin. Exports of
corn, DDGS, soybeans, and ethanol primarily increased or are similar in comparison with 2007
levels.
• Reports suggest that demands for biofuel feedstocks have led to market-mediated land use
impacts (both direct and indirect land use changes) in the past decade.
108
-------
• Cropland expansion and natural habitat loss (including forests) have been observed
internationally, and it is likely that increased biofuel production has contributed to these land use
changes.
• Quantification and causal attribution of land use change and international environmental impacts
due to biofuel production remain uncertain and undetermined.
3.7.6 Research Needs: International Impacts
• Comprehensive causal analysis frameworks and coordinated frameworks for evaluating land use
changes across biofuel trading nations may help our understanding of international land use
change and environmental impacts.
109
-------
4
Conclusions and Recommendations
4.1 Overarching Conclusions
The 2011 Report presented three overarching conclusions:
Evidence to date from the scientific literature suggests that current environmental impacts from
increased biofuels production and use associated with EISA 2007 are negative but limited in
magnitude.
Published scientific literature suggests a potential for both positive and negative environmental
effects in the future.
EISA goals for biofuels production can be achieved with minimal environmental impacts if
existing conservation and best management practices are widely employed, concurrent with
advances in technologies that facilitate the use of second-generation feedstocks.
Reports and data published since the 2011 Report have increased the confidence in the
conclusions of that report. Research also generally confirms the expected environmental and resource
conservation impacts of increased biofuel production and use, given the increased production of biofuels
from corn grain and soybeans observed since the 2011 Report was published. There has been an increase
in U.S. acreage planted with soybeans and a modest increase in U.S. acreage planted with corn since
enactment of the EISA (see Figure 4), with strong indications that some of this increase is a consequence
of increased biofuel production. There has not been a significant increase in cellulosic feedstocks (e.g.,
corn stover, perennial grasses, and woody biomass) since the 2011 Report was published. As a result, the
environmental impacts continue to be primarily those associated with increased production of corn and
soybeans, the associated conversion to fuels, and end use.
Since the 2011 Report, findings from the scientific literature and data from observations allows
the conclusions of the 2011 Report to be reaffirmed, with qualification:
Disregarding any effects that biofuels have on displacing other sources of transportation energy,
evidence since 2011 indicates the specific environmental impacts listed in EISA Section 204 are
negative. However, without assessing biofuels' displacement of other sources of transportation
energy, there is insufficient evidence to support a conclusion on the overall direction or
magnitude of effect.
110
-------
Literature published since 2011 supports the conclusion of the potential for positive and negative
effects. Available information suggests, without accounting for the environmental effects of
displacing other sources of transportation energy, the specific environmental impacts listed in
EISA Section 204 are negative in comparison to the period prior to enactment of EISA.
Evidence continues to support the conclusion that biofuel production and use could be achieved
with reduced environmental impacts. The majority of biofuels continue to be produced from
corn grain and soybeans, with associated impacts that are well understood. Cellulosic and other
feedstocks remain a minimal contributor to total biofuel production.
4.2 Specific Conclusions
Conclusions regarding the environmental and resource conservation impacts for each of the
sections are summarized below.
4.2.1 Land Use Change
Biofuel feedstock production is responsible for some of the observed changes in land used for
agriculture, but we cannot quantify with precision the amount of land with increased intensity of
cultivation nor confidently estimate the portion of crop land expansion that is due to the market
for biofuels.
Recent research and anticipated updates to data are expected to improve our ability over the next
three years to quantify the fraction of land use change attributed to biofuel feedstock production
in the U.S.
Evidence from multiple sources demonstrates an increase in actively managed cropland in the
U.S. since the passage of EISA by roughly 4-7.8 million acres, depending upon the source.
Much of this increase is likely occurring in the western and northern edges of the corn belt with
reductions of pasture and grassland, but also through infilling of already agricultural areas.
Thus, intensification likely dominates in already agricultural areas and extensification dominates
in less agricultural areas.
• Global cropland area has expanded since the year 2000, coinciding with the increase in U.S.
biofuel production. During this period, the ratio of area harvested to arable land increased and
crop yields increased significantly, due in large part to gains in total factor productivity.
• Agricultural extensification and deforestation have been documented in countries that are major
exporters of biofuels to the U.S., including Brazil, Argentina, and Indonesia.
Ill
-------
• Cropland expansion and natural habitat loss (including forests) have been observed
internationally during the implementation of the RFS program. It is likely that increased biofuel
production has contributed to these land use changes, but significant uncertainty remains about
the amount and type of land use changes that can be quantitatively attributed to U.S. biofuel
consumption (see Box 3 on Attribution).
• Researchers have continued to update and refine economic models to estimate biofiiel-induced
land use changes.
• Due to inherent challenges, uncertainties are large, and progress reducing the sources of
uncertainty has been limited.
4.2.2 Air Quality
• There is no new evidence that contradicts the conclusions of the 2011 Report concerning air
quality. Those conclusions emphasized that life cycle emissions of NOx, SOx, CO, VOCs, NH3,
and particulate matter can be impacted at each stage of biofuel production, distribution, and
usage. These impacts depend on feedstock type, land use change, and land
management/cultivation practices and are therefore highly localized. The impacts associated
with feedstock and fuel production and distribution are important to consider when evaluating
the air quality impacts of biofuel production and use, along with those associated with fuel
usage.
• Ethanol from corn grain has higher emissions across the life-cycle than ethanol from other
feedstocks.
• Ethanol plants relying on coal have higher air pollutant emissions than plants relying on natural
gas and other energy sources.
• The magnitude, timing, and location of all these emissions changes can have complex effects on
the atmospheric concentrations of criteria pollutants (e.g., O3 and PM2 5) and air toxics, the
deposition of these compounds, and subsequent impacts on human and ecosystem health.
• Ethanol increased NOx emissions from light-duty vehicles certified to Federal Tier 2 Standards,
likely occurring during times when the vehicle catalyst is not yet warmed up or air/fuel ratio is
not perfectly controlled. However, only limited data exist on the impacts of biofuels on the
tailpipe and evaporative emissions of light-duty Tier 3 vehicles and light-duty vehicles using
advanced gasoline engine technologies to meet GHG emissions standards.
112
-------
4.2.3 Water Quality
• The 2011 Report found that corn production intensification was associated with higher levels of
erosion, chemical loadings to surface waters, and eutrophication.
• Modeling studies since the 2011 Report suggest that demand for biofuel feedstocks may
contribute to harmful algal blooms, as recently observed in western Lake Erie, and to hypoxia,
as observed in the northern Gulf of Mexico.
• Empirical studies documenting cropland extensification and crop switching to more corn suggest
water quality impacts, but the magnitude of these changes is variable across the landscape and so
may be detectable only in some regions.
• Implementation of conservation practices has been observed to result in a decrease of nitrogen,
phosphorus, and soil erosion.
• Changes to future nitrogen and phosphorus loadings will depend on feedstock mix and crop
management practices. Decreases in nitrogen and phosphorus loadings are possible should
perennial feedstocks become dominant.
• Specific biofuel production scenarios expected to improve water quality may help decrease the
water quality impact of predicted future extreme weather events.
4.2.4 Water Quantity
• As discussed in the 2011 Report, the irrigation of corn and soybeans grown for biofuels is the
predominant water quantity impact. Water use for feedstock production is significantly greater
than water use in the biofuel conversion process.
• There are some indications of increased water use in irrigated areas for corn between 2007 and
2012 and elevated rates of land use change to corn production in more arid Western states,
including the Ogallala region. Adverse water availability impacts will most likely arise in
already stressed aquifers and surface watersheds.
• Irrigation practices are dependent on a number of economic and agronomic factors that drive
land management practices, making attribution of increased irrigation and water quantity to
biofuels difficult.
4.2.5 Ecosystem Health and Biodiversity
• Loss of grasslands and wetlands is occurring in ecologically sensitive areas, including the Prairie
Pothole Region.
113
-------
• Loss of habitat and landscape simplification are associated with negative impacts to pollinators,
birds, soil-dwelling organisms, and other ecosystem services, both in terrestrial and aquatic
habitats.
• Increased fertilizer applications of nitrogen and phosphorus have known negative effects on
aquatic biodiversity.
4.2.6 Soil Quality
• Corn-grain ethanol and soy biodiesel account for most of the biofuel volumes produced to date.
As a result, almost all the soil quality impacts from biofuels, thus far, are from the production of
the dominant conventional feedstocks.
• Conversion of grasslands to annual cropland typically negatively affects soil quality, with
increases in erosion and the loss of soil nutrients and soil organic matter, including soil carbon.
Impacts of this conversion can be partially mitigated - though not entirely - through the
adoption of management practices such as conservation tillage.
• The soil quality impacts of converting from other crops to corn or soybeans are generally less
than those of the conversion of grasslands. The production of corn on existing cropland can
provide soil carbon benefits, although these benefits are outweighed on a per area basis by the
negative effects of grassland conversion.
• Overall, these land use trends suggest that negative impacts to soil quality from biofuel
feedstocks have increased since 2011, but this has not been quantified and the magnitude of
effects depends predominantly on the relative areas of grasslands converted versus existing
croplands attributable to biofuels.
• Corn stover is now being harvested at the commercial-scale in Iowa, and the scientific literature
indicates this must be done carefully to avoid negatively affecting soil quality and crop yields.
4.2.7 Invasive Species
• Biofuels are primarily produced in the forms of bioethanol and biodiesel derived from food
crops (i.e., non-invasive first generation biofuels - corn and soybeans). Hence current
production of biofuel feedstocks poses little risk of invasion, consistent with findings in the 2011
Report.
• Weed risk assessments, part of the formal biofuel regulatory process, provide information on
invasion risk and are designed to inform protective management of species and varieties that are
predicted to be invasive.
114
-------
• Increased cultivation of crops engineered for herbicide tolerance (e.g., glyphosate) and
concomitant application of the herbicide has led to a widespread increase in the number of
glyphosate-resistant weed species.
• Potentially invasive species approved as feedstocks require risk management actions under
current RFS requirements. However, invasive species are not presently being used for
commercial scale production of biofuels.
4.2.8 International Impacts
• Since the 2011 Report, U.S. ethanol imports decreased, while biodiesel and renewable diesel
imports increased, leading to potential land use change impacts in countries of origin. Exports of
corn, DDGS, soybeans, and ethanol primarily increased or are similar in comparison with 2007
levels.
• Reports suggest that demands for biofuel feedstocks have led to market-mediated land use
impacts (both direct and indirect land use changes) in the past decade.
• Cropland expansion and natural habitat loss (including forests) have been observed
internationally, and it is likely that increased biofuel production has contributed to these land use
changes.
• Quantification and causal attribution of land use change and international environmental impacts
due to biofuel production remain uncertain and undetermined.
4.3 Opportunities for Future Environmental Improvements
Some cellulosic feedstock production scenarios are expected to reduce surface water nitrogen
loadings, particularly following extreme weather events. (Water Quality)
Priority should be placed on identifying effective strategies to manage withdrawals in "hot
spots" (e.g., Ogallala aquifer) where high water demands and water stress are coinciding. (Water
Quantity)
While cellulosic feedstocks, such as perennial grasses, can provide environmental benefits for
biodiversity and ecosystem services, their potential impact on streamflow within a watershed
should be carefully considered. (Water Quantity)
Planting perennial grasslands and replacing marginal croplands with perennial grasslands can
enhance ecosystem services. (Ecosystems and Biodiversity)
115
-------
Increased use of effective conservation practices can have multiple benefits, from reduced
stream sedimentation and nutrient runoff to protection of pollinator habitat. (Ecosystems and
Biodiversity)
Increased adoption of other technologies, such as time-release and other enhanced-efficiency
fertilizers, alternative fertilizer placement methods, and precision agriculture, could further
improve conservation practices in biofuel feedstock production. (Ecosystems and Biodiversity)
Alternative biofuel feedstocks, such as perennial grasses and woody biomass, are not yet used at
commercial scales. Studies have shown that these feedstocks can improve soil quality relative to
current conditions, contingent on species grown and type of land converted (e.g., marginal,
abandoned, or degraded lands). (Soil Quality)
4.4 Limitations
This report does not include a comparative assessment of the impact of biofuels on the
environment relative to the impacts of other transportation fuels or energy sources, including
fossil fuels, for every environmental endpoint, limiting the ability of this report to draw
conclusions regarding the comprehensive environmental impacts of biofuels.
The environmental impacts discussed in this report are not constant across all locations due to
local factors, such as the extent of land use change and local relationships between land use
change and their associated direct and indirect impacts.
We cannot now confidently quantify the fraction of increased land use change and associated
environmental impacts due to changes in biofuel production.
Numerous factors influence the markets for biofuels and thus the associated environmental
impacts, including: regional considerations; scale and volume of future commercial biofuel
operations; development of hybrid biofuel conversion processes; changes in vehicle
technologies; and changes in agricultural practices due to biofuel production and implications
for environmental impacts. Each of these, whether individually or in combination, will affect the
ultimate environmental impacts associated with biofuel production and use.
4.5 Research Needs
• Research is needed to quantify changes in the intensity of cultivation on existing agricultural
land. (Land Use Change)
116
-------
Research is also needed to more effectively connect changes in land use to the environmental
impacts of concern. (Land Use Change)
Comprehensive studies of the impacts of biofuels on the emissions from advanced light-duty
vehicle technologies (Tier 3), similar in scope to studies cited in this report for light-duty Tier 2
vehicles, would improve the understanding of the potential for biofiiel-specific pollutants and
associated health impacts as new technologies enter the vehicle fleet. These studies should
consider engine technologies phasing into use for compliance with current and future light-duty
GHG standards, with a focus on vehicles compliant with the Federal Tier 3 or California LEV III
criteria pollutant emissions standards currently under implementation. Such technologies would
include engine downsizing with addition of turbocharging, gasoline direct injection, and non-
traditional thermodynamic cycles such as Miller or Atkinson. (Air Quality)
Additional research and analyses are needed to adequately understand the potential health effects
of exposure to biofuels and emissions from vehicles using biofuels under real-world conditions,
concentrations, and exposures including to susceptible human populations. It would be
appropriate to study health effects in populations exposed to biodiesel and ethanol blends in
"hotspots," such as fuel production sites, and those exposed to combustion products of biodiesel
and ethanol blends, especially at high blend levels. Such studies could include drivers of
vehicles utilizing those fuels.
Updated modeling is needed to incorporate improved emissions estimates as laboratory, field,
and other studies lead to a better understanding of biofuel-related emissions changes and
associated changes in the magnitude and composition of pollutants on air quality, health, and
attainment with ambient air quality standards. (Air Quality)
Studies are needed of water quality impacts associated with leaks and/or spills from biofuel
production facilities and storage tanks. Such work would address the effectiveness of existing
leak detection and cleanup approaches to address releases to the environment and resulting
contamination plumes. (Water Quality)
Studies are needed to determine the extent to which increases in water consumption and
withdrawals - due to changes in land use/management change - can be attributed to feedstock
production for biofuels. In particular, studies should continue to explore increases in water
demands that have occurred or are occurring along water-stressed areas, both for surface and
groundwater. (Water Quantity)
117
-------
Research, both modeling and field work to verify modeling parameters, is needed to better
understand future cellulosic feedstock water demands while assessing water quantity, water
quality, and soil quality in an integrated manner. (Water Quantity)
Studies that target the interactive effects of land use change and feedstock production could help
identify impacts to specific organisms. (Ecosystem Health and Biodiversity)
Research on the efficacy of methods to expand pollinator habitat in agricultural systems can
improve understanding of appropriate methods and their potential tradeoffs for different
agricultural areas. (Ecosystem Health and Biodiversity)
Quantitative estimates of the cumulative soil quality effects are needed for the land use changes
described in Section 2.4 and for the proportion attributable to biofuel feedstock production (this
includes both the conversion of land to corn and soybeans and the management practices
implemented). (Soil Quality)
Methodological advancements for weed risk assessments and lessons from other industries (e.g.,
horticulture) should be incorporated to inform on potential invasiveness of biofuel feedstocks.
(Invasive Species)
Modeling and field work are needed to investigate the impacts of gene flow between novel
feedstock varieties (genetically engineered, selectively bred, or a combination) and local natives.
(Invasive Species)
Comprehensive causal analysis frameworks and coordinated frameworks for evaluating land use
changes across biofuel trading nations may help our understanding of international land use
change and environmental impacts. (International Impacts)
Recommendations
Additional research in coordination with other organizations (e.g., federal agencies, states, trade
organizations) is recommended to better characterize land use change due to changes in biofuel
feedstock production.
Efforts at the federal level, as described by the Biomass Research and Development Board, to
improve efficiencies and sustainability of processes across the biofuel supply chain should be
continued and strengthened where possible.
An ecosystem approach is recommended to evaluate environmental and natural resource impacts
of biofuel production. Such an approach provides an integrative perspective that accounts for
complex interactions of multiple stressors across different locations.
118
-------
• Incorporating local information and perspectives will improve understanding of changes at local
scales, which will enhance opportunities for improved information and will enable targeted
responses to prevent and mitigate adverse impacts of biofuel production and use.
• Best management practices should be encouraged, incentivized, and otherwise expanded to
promote conservation and sustainability in agricultural systems.
119
-------
5 References
Acquaye AA, Sherwen T, Genovese A, Kuylenstierna J, Koh SCL and McQueen-Mason S (2012).
Biofuels and their potential to aid the UK towards achieving emissions reduction policy targets.
Renewable & Sustainable Energy Reviews 16(7): 5414-5422: 10.1016/j.rser.2012.04.046.
Adami M, Rudorff BFT, Freitas RM, Aguiar DA, Sugawara LM and Mello MP (2012). Remote sensing
time series to evaluate direct land use change of recent expanded sugarcane crop in Brazil.
Sustainability 4(4): 574-585:
Al-Riffai P, Dimaranan B and Laborde D (2010). Global trade and environmental impact study of the
EU biofuels mandate, IFPRI Washington, DC
Babcock BA (2015). Extensive and intensive agricultural supply response. Annu. Rev. Resour. Econ.
7(1): 333-348:
Baldos U (2017). Development of GTAP version 9 Land Use and Land Cover database for years 2004,
2007 and 2011. Center for Global Trade Analysis, GTAP Research Memorandum No. 30: 40 pp.
Wet Lafayette, IN.
Barney JN, Mann JJ, Kyser GB and DiTomaso JM (2012). Assessing habitat susceptibility and
resistance to invasion by the bioenergy crops switchgrass and Miscanthus x giganteus in
California. Biomass & Bioenergy 40: 143-154: 10.1016/j.biombioe.2012.02.013.
Bass VL, Schladweiler MC, Nyska A, Thomas RF, Miller DB, Krantz T, King C, Ian Gilmour M,
Ledbetter AD, Richards JE and Kodavanti UP (2015). Comparative cardiopulmonary toxicity of
exhausts from soy-based biofuels and diesel in healthy and hypertensive rats. Inhalation
Toxicology 27(11): 545-556: 10.3109/08958378.2015.1060279.
Beasley T, Evansky P, Martin S, McDaniel K, Moser V, Luebke R, Norwood J, J.M.Rogers, Copeland
CB and Bushnell PJ (2014). Toxicological outcomes in rats exposed to inhaled ethanol during
gestation Neurotoxicology and Teratology 45: 59-69:
Beckman J, Borchers A and Jones C (2013). Agriculture's Supply and Demand for Energy and
Electricity Products. U.S. Department of Agriculture, Economic Research Service. Washington,
DC.
Benbrook CM (2012). Impacts of genetically engineered crops on pesticide use in the U.S. ~ the first
sixteen years. Environmental Sciences Europe 24: 24:
Benbrook CM (2016). Trends in glyphosate herbicide use in the United States and globally.
Environmental Sciences Europe 28:
Bennett AB and Isaacs R (2014a). Landscape composition influences pollinators and pollination services
in perennial biofiiel plantings. Agriculture, Ecosystems & Environment 193: 1-8:
https://doi.Org/10.1016/j.agee.2014.04.016.
Bennett AB, Meehan TD, Gratton C and Isaacs R (2014b). Modeling Pollinator Community Response to
Contrasting Bioenergy Scenarios. PLOS ONE 9(11): el 10676: 10.1371/journal.pone.0110676.
Bento AM and Klotz R (2014). Climate Policy Decisions Require Policy-Based Lifecycle Analysis.
Environmental Science & Technology 48(10): 5379-5387: 10.1021/es405164g.
Bigelow DP and Borchers A (2017). Major Uses of Land in the United States, 2012. U.S. Department of
Agriculture, Economic Research Service. August 2017.
Bischoff KP, Gravois KA, Reagan TE, Hoy JW, Kimbeng CA, LaBorde CM and Hawkins GL (2008).
Registration of 'L 79-1002' Sugarcane. Journal of Plant Registrations 2: 211-217:
Blanco-Canqui H (2013). Crop Residue Removal for Bioenergy Reduces Soil Carbon Pools: How Can
We Offset Carbon Losses? Bioenergy Research 6(1): 358-371: 10.1007/s 12155-012-9221 -3.
Blanco-Canqui H (2016). Growing Dedicated Energy Crops on Marginal Lands and Ecosystem Services.
Soil Science Society of America Journal 80(4): 845-858: 10.2136/sssaj2016.03.0080.
Blank PJ, Sample DW, Williams CL and Turner MG (2014). Bird Communities and Biomass Yields in
Potential Bioenergy Grasslands. PLOS ONE 9(10): el09989: 10.1371/journal.pone.0109989.
120
-------
Bonmatin J-M, Giorio C, Girolami V, Goulson D, Kreutzweiser DP, Krupke C, Liess M, Long E,
Marzaro M, Mitchell EAD, Noome DA, Simon-Delso N and Tapparo A (2015). Environmental
fate and exposure; neonicotinoids and fipronil. Environmental Science and Pollution Research
22(1): 35-67: 10.1007/sl 1356-014-3332-7.
Borchers A, Truex-Powell E, Wallander S and Nickerson C (2014). Multi-Cropping Practices: Recent
Trends in Double Cropping. U.S. Department of Agriculture, Economic Research Service, EIB-
125.
Boyes WK, Degn LL, Martin SA, Lyke DF, Hamm CW and Herr DW (2014). Neurophysiological
assessment of auditory, peripheral nerve, somatosensory, and visual system functions after
developmental exposure to ethanol vapors. Neurotoxicology and Teratology 43: 1-10:
http://dx.doi.Org/10.1016/i.ntt.2014.02.006.
Brady N and Weil R (2000). Elements of the Nature and Properties of Soils. Upper Saddle River, NJ,
USA, Prentice-Hall, Inc.
Brookshear D, Nguyen K, Toops T, Bunting B and Rohr W (2013). Impact of Biodiesel-Based Na on the
Selective Catalytic Reduction of NOx by NH3 Over Cu-Zeolite Catalysts. Topics in Catalysis
56: 62-67:
Brown JC, Hanley E, Bergtold J, Caldas M, Barve V, Peterson D, Callihan R, Gibson J, Gray B,
Hendricks N, Brunsell N, Dobbs K, Kastens J and Earnhart D (2014). Ethanol plant location and
intensification vs. extensification of corn cropping in Kansas. Applied Geography 53: 141-148:
10.1016/j.apgeog.2014.05.021.
Brusamarello-Santos LC, Gilard F, Brule L, Quillere I, Gourion B, Ratet P, de Souza EM, Lea PJ and
Hirel B (2017). Metabolic profiling of two maize (Zea mays L.) inbred lines inoculated with the
nitrogen fixing plant-interacting bacteria Herbaspirillum seropedicae and Azospirillum
brasilense. Plos One 12(3): 10.1371/journal.pone.0174576.
Burnham A, Wang M and Wu Y (2006). Development and Applications of GREET 2.7 — The
Transportation Vehicle-Cycle Model. Argonne National Laboratory: 124 pp. Argonne, IL,
November.
Bushnell PJ, Beasley T, Evansky P, Martin S, McDaniel K, Moser V, Luebke R, Norwood J and
Copeland CB (2015). Toxicological Assessments of Rats Exposed Prenatally to Inhaled Vapors
of Gasoline and Gasoline-Ethanol Blends. Neurotoxicology and Teratology 49: 19-30:
Campbell JB (1996). Introduction to remote sensing. New York, NY, The Guildford Press
CARB (2014). Detailed Analysis for Indirect Land Use Change. C. A. R. Board. Sacramento, CA: 113.
Carter CA, Rausser GC and Smith A (2017). Commodity Storage and the Market Effects of Biofuel
Policies. American Journal of Agricultural Economics 99(4): 1027-1055: 10.1093/ajae/aaw010.
Cassel DK, Raczkowski CW and Denton HP (1995). Tillage effects on corn production and soil physical
conditions. Soil Science Society of America Journal 59(5): 1436-1443:
Chagnon M, Kreutzweiser D, Mitchell EAD, Morrissey CA, Noome DA and Van der Sluijs JP (2015).
Risks of large-scale use of systemic insecticides to ecosystem functioning and services.
Environmental Science and Pollution Research 22(1): 119-134: 10.1007/sl 1356-014-3277-x.
Chang H, Snow AA, Mutegi E, Lewis EM, Heaton EA and Miriti MN (2018). Extent of pollen-mediated
gene flow and seed longevity in switchgrass (Panicum virgatum L.): Implications for biosafety
procedures. Biomass andBioenergy 109: 114-124:
https://doi.Org/10.1016/j.biombioe.2017.12.016.
Chiu Y-W, Walseth B and Suh S (2009). Water Embodied in Bioethanol in the United States.
Environmental Science & Technology 43(8): 2688-2692: 10.1021/es8031067.
Chiu Y-W and Wu M (2012). Assessing County-Level Water Footprints of Different Cellulosic-Biofuel
Feedstock Pathways. Environmental Science & Technology 46(16): 9155-9162:
10.1021/es3002162.
Ciais P, Sabine C, Bala G, Bopp L, Brovkin V, Canadell J, Chhabra A, DeFries R, Galloway J, Heimann
M, Jones C, Le Quere C, Myneni RB, Piao S and Thornton P (2013). Carbon and Other
121
-------
Biogeochemical Cycles. In: . Climate Change 2013: The Physical Science Basis. Contribution of
Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate
Change. T. F. Stacker, D. Qin, G.-K. Plattneret al (eds). Cambridge, United Kingdom and New
York, NY, USA, Cambridge University Press.
Cibin R, Trybula E, Chaubey I, Brouder SM and Volenec JJ (2016). Watershed-scale impacts of
bioenergy crops on hydrology and water quality using improved SWAT model. GCB Bioenergy
8(4): 837-848: 10.1111/gcbb. 12307.
Cook R, Phillips S, Houyoux M, Dolwick P, Mason R, Yanca C, Zawacki M, Davidson K, Harvey C and
Somers J (2011). Air Quality Impacts of Increased Use of Ethanol Under the United States'
Energy Independence and Security Act. Atmospheric Environment 45: 7714-7724:
Coulter JA and Naftiger ED (2008). Continuous Corn Response to Residue Management and Nitrogen
Fertilization. Agronomy Journal 100(6): 1774-1780: 10.2134/agronj2008.0170.
Dale BE, Anderson JE, Brown RC, Csonka S, Dale VH, Herwick G, Jackson RD, Jordan N, Kaffka S,
Kline KL, Lynd LR, Malmstrom C, Ong RG, Richard TL, Taylor C and Wang MQ (2014). Take
a Closer Look: Biofuels Can Support Environmental, Economic and Social Goals.
Environmental Science & Technology 48(13): 7200-7203: 10.1021/es5025433.
Dale VH, Kline KL, Buford MA, Volk TA, Tattersall Smith C and Stupak I (2016). Incorporating
bioenergy into sustainable landscape designs. Renewable and Sustainable Energy Reviews 56:
1158-1171: https://doi.Org/10.1016/j.rser.2015.12.038.
Dale VH, Kline KL, Richard TL, Karlen DL and Belden WW (2017). Bridging biofuel sustainability
indicators and ecosystem services through stakeholder engagement. Biomass and Bioenergy.
https://doi.Org/10.1016/j.biombioe.2017.09.016.
Dale VH, Kling C, Meyer JL, Sanders J, Stallworth H, Armitage T, Wangsness D, Bianchi TS,
Blumberg A, Boynton W, Conley DJ, Crumpton W, David MB, Gilbert D, Howarth RW,
Lowrance R, Mankin K, Opaluch J, Paerl H, Reckhow K, Sharpley AN, Simpson TW, Snyder C
and Wright. D (2010). Hypoxia in the Northern Gulf of Mexico. New York, Springer
Davis PB, Menalled FD, Peterson RKD and Maxwell BD (2011). Refinement of weed risk assessments
for biofuels using Camelina sativa as a model species. Journal of Applied Ecology(48): 989-
997:
Davis R, Tao L, Scarlata C, Tan ECD, Ross J, Lukas J and Sexton D (2015). Process Design and
Economics for the Conversion of Lignocellulosic Biomass to Hydrocarbons: Dilute-Acid and
Enzymatic Deconstruction of Biomass to Sugars and Catalytic Conversion of Sugars to
Hydrocarbons. National Renewable Energy Laboratory, NREL/TP-5100-62498. Golden, CO.
de Gouw JA, McKeen SA, Aikin KC, Brock CA, Brown SS, Gilman JB, Graus M, Hanisco T, Holloway
JS, Kaiser J, Keutsch FN, Lerner BM, Liao J, Markovic MZ, Middlebrook AM, Min KE,
Neuman JA, Nowak JB, Peischl J, Pollack IB, Roberts JM, Ryerson TB, Trainer M, Veres PR,
Warneke C, Welti A and Wolfe GM (2015). Airborne measurements of the atmospheric
emissions from a fuel ethanol refinery. Journal of Geophysical Research: Atmospheres 120(9):
4385-4397: 10.1002/2015JD023138.
DeCicco JM, Liu DY, Heo J, Krishnan R, Kurthen A and Wang L (2016). Carbon balance effects of U.S.
biofuel production and use. Climatic Change 138(3): 667-680: 10.1007/sl0584-016-1764-4.
Demissie Y, Yan E and Wu M (2012). Assessing regional hydrology and water quality implications of
large-scale biofuel feedstock production in the Upper Mississippi river basin. Environmental
Science & Technology 46(16): 9174-9182:
DiFazio SP, Leonardi S, Slavov GT, Garman SL, Adams WT and Strauss SH (2012). Gene flow and
simulation of transgene dispersal from hybrid poplar plantations. New Phytologist 193(4): 903-
915: 10.1111/j.1469-8137.2011.04012.x.
Dively GP, Venugopal PD and Finkenbinder C (2016). Field-Evolved Resistance in Corn Earworm to
Cry Proteins Expressed by Transgenic Sweet Corn. PLOS ONE 11(12): eO 169115:
10.1371/journal.pone.0169115.
122
-------
DOE (2016). 2016 Billion-Ton Report: Advancing Domestic Resources for a Thriving Bioeconomy,
Volume 1: Economic Availability of Feedstocks. Oak Ridge National Laboratory, ORNL/TM-
2016/160: 448 pp. Oak Ridge, TN, July.
DOE (2017). 2016 Billion-Ton Report: Advancing Domestic Resources for a Thriving Bioeconomy,
Volume 2: Environmental Sustainability Effects of Select Scenarios from Volume 1. Oak Ridge
National Laboratory: 640 pp. Oak Ridge, TN.
Dominguez-Faus R, Folberth C, Liu J, Jaffe AM and Alvarez PJJ (2013). Climate Change Would
Increase the Water Intensity of Irrigated Corn Ethanol. Environmental Science & Technology
47(11): 6030-6037: 10.1021/es400435n.
Dominguez-Faus R, Powers SE, Burken JG and Alvarez PJ (2009). The Water Footprint of Biofuels: A
Drink or Drive Issue? Environmental Science & Technology 43(9): 3005-3010:
10.1021/es802162x.
Douglas MR, Rohr JR and Tooker JF (2015a). Neonicotinoid insecticide travels through a soil food
chain, disrupting biological control of non-target pests and decreasing soya bean yield. Journal
of Applied Ecology 52: 250-260:
Douglas MR and Tooker JF (2015b). Large-Scale Deployment of Seed Treatments Has Driven Rapid
Increase in Use of Neonicotinoid Insecticides and Preemptive Pest Management in U.S. Field
Crops. Environmental Science & Technology 49(8): 5088-5097: 10.1021/es506141g.
Douglas MR and Tooker JF (2016). Meta-analysis reveals that seed-applied neonicotinoids and
pyrethroids have similar negative effects on abundance of arthropod natural enemies. PeerJ 4:
e2776: 10.7717/peeij.2776.
Dunn JB, Merz D, Copenhaver KL and Mueller S (2017). Measured extent of agricultural expansion
depends on analysis technique. Biofuels Bioproducts & Biorefining-Biofpr 11(2): 247-257:
10.1002/bbb.l750.
Dunn JB, Mueller S, Kwon H-y and Wang MQ (2013). Land-use change and greenhouse gas emissions
from corn and cellulosic ethanol. Biotechnology for Biofuels 6(1): 51: 10.1186/1754-6834-6-51.
DuPont. (2017). "DuPont cellulosic ethanol corn stover harvest program." Retrieved 5/10, 2017, from
http://www.dupont.com/products-and-services/industrial-biotechnology/advanced-
biofuels/articles/nevada-corn-stover-harvest-program.html.
Efroymson RA, Kline KL, Angelsen A, Verburg PH, Dale VH, Langeveld JWA and McBride A (2016).
A causal analysis framework for land-use change and the potential role of bioenergy policy.
Land Use Policy 59(516-527):
EIA. (2017, February 16). "Nebraska State Profile and Energy Estimates: Profile Analysis." Retrieved
June 2, 2017, from https://www.eia.gov/state/analvsis.php?sid=NE.
Emery I, Mueller S, Qin Z and Dunn JB (2016). Evaluating the potential of marginal land for cellulosic
feedstock production and carbon sequestration in the United States. Environmental Science &
Technology 51: 733-741:
EPA (2003). National Management Measures to Control Nonpoint Pollution from Agriculture. U.S.
Environmental Protection Agency Office of Water, EPA-841-B-03-004. Washington, DC, July.
EPA (2010). Renewable fuel standard program (RFS2) regulatory impact analysis. U.S. Environmental
Protection Agency Office of Transportation Air Quality, EPA-420-R-10-006. Washington, DC.
EPA (2011). Biofuels and the Environment: First Triennial Report to Congress. U.S. Environmental
Protection Agency, EPA/600/R-10/183F: 220 pp. Washington, DC, December.
EPA (2013a). Assessing the Effect of Five Gasoline Properties on Exhaust Emissions from Light-Duty
Vehicles Certified to Tier 2 Standards: Analysis of Data from EPAct Phase 3 (EPAct/V2/E-89).
U.S. Environmental Protection Agency, EPA-420-R-13-002. April.
EPA (2013b). EPAct/V2/E-89: Assessing the Effect of Five Gasoline Properties on Exhaust Emissions
from Light-Duty Vehicles Certified to Tier 2 Standards - Final Report on Program Design and
Data Collection. U.S. Environmental Protection Agency, EPA-420-R-13-004. April.
123
-------
EPA (2013c). Regulation of fuels and fuel additives: Additional qualifying renewable fuel pathways
under the renewable fuel standard program; Final rule approving renewable fuel pathways for
giant reed (Arundo donax) and napier grass (Pennisetum purpureum). Federal Register 78:
EPA (2013d). Regulation of fuels and fuel additives: Identification of additional qualifying renewable
fuel pathways under the renewable fuel standard program. Federal Register 78:
EPA (2016a). Draft Technical Assessment Report: Midterm Evaluation of Light-Duty Vehicle
Greenhouse Gas Emission Standards and Corporate Average Fuel Economy Standards for
Model Years 2022-2025. U.S. Environmental Protection Agency, EPA-420-D-16-900.
Washington, DC, July.
EPA (2016b). Preliminary Aquatic Risk Assessment to Support the Registration Review of Imidacloprid.
U.S. Environmental Protection Agency Office of Chemical Safety and Pollution Prevention,
EPA-HQ-OPP-2008-0844-1086: 219 pp. Washington, DC, December 22.
EPA (2016c). Refined ecological risk assessment for atrazine. U.S. Environmental Protection Agency
Office of Chemical Safety and Pollution Prevention, EPA-HQ-OPP-2013-0266-0315.
Washington, DC.
EPA (2016d). Renewable Fuel Standard Program: Standards for 2017 and Biomass-Based Diesel
Volume for 2018. Federal Register Vol 81 No. 238, US Environmental Protection Agency.
EPA (2017). Preliminary Bee Risk Assessment to Support the Registration Review of Clothianidin and
Thiamethoxam. Office of Pesticide Programs, EPA-HQ-OPP-2011-0865-0173: 414 pp.
Washington, DC.
Evans S and Potts M (2015). Effect of Agricultural Commodity Prices on Species Abundance of US
Grassland Birds. Environmental & Resource Economics 62(3): 549-565:
Falcone JA (2015). U.S. conterminous wall-to-wall anthropogenic land use trends (NWALT), 1974-
2012. U.S. Geological Survey: 33 pp. Washington, DC.
FAO (2016). Global Forest Resources Assessment 2015. United Nations Food and Agriculture
Organization. Rome.
Farraj AK, Haykal-Coates N, Winsett DW, Gilmour MI, King C, Krantz QT, Richards J and Hazari MS
(2015). Comparative electrocardiographic, autonomic and systemic inflammatory responses to
soy biodiesel and petroleum diesel emissions in rats. Inhalation Toxicology 27(11): 564-575:
10.3109/08958378.2015.1057884.
Fausti SW (2015). The causes and unintended consequences of a paradigm shift in corn production
practices. Environmental Science & Policy 52: 41-50:
Filoso S, do Carmo JB, Mardegan SF, Lins SRM, Gomes TF and Martinelli LA (2015). Reassessing the
environmental impacts of sugarcane ethanol production in Brazil to help meet sustainability
goals. Renewable and Sustainable Energy Reviews 52: 1847-1856:
Fletcher RJ, Robertson BA, Evans J, Doran PJ, Alavalapati JRR and Schemske DW (2011). Biodiversity
conservation in the era of biofuels: risks and opportunities. Frontiers in Ecology and the
Environment 9(3): 161-168: 10.1890/090091.
Follett RF, Varvel GE, Kimble JM and Vogel KP (2009). No-Till Corn after Bromegrass: Effect on Soil
Carbon and Soil Aggregates. Agronomy Journal 101(2): 261-268: 10.2134/agronj2008.0107.
Follett RF, Vogel KP, Varvel GE, Mitchell RB and Kimble J (2012). Soil Carbon Sequestration by
Switchgrass and No-Till Maize Grown for Bioenergy. Bioenergy Research 5(4): 866-875:
10.1007/s 12155-012-9198-y.
Foufoula-Georgiou E, Takbiri Z, Czuba JA and Schwenk J (2015). The change of nature and the nature
of change in agricultural landscapes: Hydrologic regime shifts modulate ecological transitions.
Water Resources Research 51: 6649-6671:
Fuglie KO and RadaN. (2013). "USDA Economic Research Service: Growth in Global Agricultural
Productivity: An Update." from https://www.ers.usda.gov/amber-waves/2013/november/growth-
in-global-agricultural-productivitv-an-update/
124
-------
Fuglie KO, Wang SL and Ball VE, Eds. (2012). Productivity Growth in Agriculture: An International
Perspective. Oxfordshire, UK, CAB International.
Garcia V, Cooter E, Crooks J, Hinckley B, Murphy M and Xing X (2017). Examining the impacts of
increased corn production on groundwater quality using a coupled modeling system. Science of
The Total Environment 586: 16-24: https://doi.Org/10.1016/j.scitotenv.2017.02.009.
Gassmann AJ, Shrestha RB, Jakka SRK, Dunbar MW, Clifton EH, Paolino AR, Ingber DA, French BW,
Masloski KE, Dounda JW and Clair CRS (2016). Evidence of resistance to Cry34/35Abl corn
by western corn rootworm (Coleoptera: Chrysomelidae): Root injury in the field and larval
survival in plant-based bioassays. Journal of Economic Entomology 109(4): 1872-1880:
Gelfand I, Zenone T, Jasrotia P, Chen JQ, Hamilton SK and Robertson GP (2011). Carbon debt of
Conservation Reserve Program (CRP) grasslands converted to bioenergy production.
Proceedings of the National Academy of Sciences of the United States of America 108(33):
13864-13869: 10.1073/pnas. 1017277108.
George I, Hays M, Snow R, Faircloth J, George B, Long T and Baldauf R (2014). Cold Temperature and
Biodiesel Fuel Effects on Speciated Emissions of Volatile Organic Compounds from Diesel
Trucks. Environmental science & technology 48: 14782-14789:
Gerbens-Leenes W and Hoekstra AY (2012). The water footprint of sweeteners and bio-ethanol. Environ
Int 40: 202-211: 10.1016/j.envint.2011.06.006.
Gibbs HK, Sheehan, J., Nelson, R. (2010). Land cover types subgroup. Low carbon fuel standard
(LCFS) Indirect land use change expert workgroup A Report to the California Air Resources
Board:
Gleason RA, Euliss NH, Tangen BA, Laubhan MK and Browne BA (2011). USDA conservation
program and practice effects on wetland ecosystem services in the Prairie Pothole Region.
Ecological Applications 21: S65-S81:
Gordon DR, Tancig KJ, Onderdonk DA and Gantz CA (2011). Assessing the invasive potential of
biofuel species proposed for Florida and the United States using the Australian Weed Risk
Assessment. Biomass & Bioenergy 35(1): 74-79: 10.1016/j.biombioe.2010.08.029.
Gray BJ and Gibson JW (2013). Actor-Networks, Farmer Decisions, and Identity. Culture, Agriculture,
Food and Environment 35(2): 82-101: 10.1111/cuag.l2013.
Gregorich EG and Anderson DW (1985). Effects of cultivation and erosion on soils of four
toposequences in the Canadian prairies. Geoderma 36(3-4): 343-354: 10.1016/0016-
7061(85)90012-6.
Gressel J (2015). Dealing with transgene flow of crop protection traits from crops to their relatives. Pest
Management Science 71(5): 658-667: 10.1002/ps.3850.
Ha M and Wu M (2017). Land management strategies for improving water quality in biomass
production under changing climate. Environmental Research Letters 12(3): 034015:
Haskew HM and Liberty TF (2011). Exhaust and Evaporative Emissions Testing of Flexible Fuel
Vehicles. Coordinating Research Council Report No. E-80. Alpharetta, GA.
Hays M, D. WP, George BJ, Schmid J, Baldauf R, Snow R, Robinson JR, Long T and Faircloth J (2013).
Carbonaceous aerosols emitted from light-duty vehicles operating on gasoline and ethanol fuel
blends. Environmental Science & Technology 47(24): 14502-14509:
Hazari MS, Haykal-Coates N, Winsett DW, King C, Krantz QT, Gilmour MI and Farraj AK (2015). The
effects of B0, B20, and B100 soy biodiesel exhaust on aconitine-induced cardiac arrhythmia in
spontaneously hypertensive rats. Inhalation Toxicology 27(11): 557-563:
10.3109/08958378.2015.1054967.
Heap I (2014). Global perspective of herbicide-resistant weeds. Pest Management Science 70: 1306—
1315:
Heath G, Carpenter A and Fisher N (2013). Comparative Air Pollutant Emissions of Selected Biofuel
Feedstock Production in 2022. International Congress on Sustainabilitv Science Engineering.
Cincinnati, OH.
125
-------
Heaton E, Dohleman F, Miguez F, Juvik J, Lozovaya V, Widholm J, Zabotina O, Mclsaac G, David M
and Voigt T (2010). Miscanthus: a promising biomass crop. Advances in Botanical Research 56:
76-137:
Hein L and Leemans R (2012). The Impact of First-Generation Biofuels on the Depletion of the Global
Phosphorus Reserve. AMBIO 41: 341-349:
Hellwinckel C, Clark C, Langholtz M and Eaton L (2016). Simulated impact of the renewable fuels
standard on US Conservation Reserve Program enrollment and conversion. Global Change
Biology Bioenergy 8(1): 245-256: 10. Ill 1/gcbb. 12281.
Hertel TW, Golub AA, Jones AD, O'Hare M, Plevin RJ and Kammen DM (2010). Effects of US Maize
Ethanol on Global Land Use and Greenhouse Gas Emissions: Estimating Market-mediated
Responses. Bioscience 60(3): 223-231: 10.1525/bio.2010.60.3.8.
Hertel TW and Tyner WE (2013). Market-mediated environmental impacts of biofuels. Global Food
Security 2: 131-137:
Hladik ML, Kolpin DW and Kuivila KM (2014). Widespread occurrence of neonicotinoid insecticides in
streams in a high corn and soybean producing region, USA. Environmental Pollution 193: 189-
196: https://doi.Org/10.1016/j.envpol.2014.06.033.
Hladik ML and Kolpin. DW (2015). First national-scale reconnaissance of neonicotinoid insecticides in
streams across the USA. Environmental Chemistry 13(1): 12-20:
Housh M, Yaeger MA, Cai X, Mclsaac GF, Khanna M, Sivapalan M, Ouyang Y, Al-Qadi I and Jain AK
(2015). Managing multiple mandates: a system of systems model to analyze strategies for
producing cellulosic ethanol and reducing riverine nitrate loads in the upper Mississippi river
basin. Environmental Science & Technology 49(19): 11932-11940:
Hulme PE (2012). Weed risk assessment: a way forward or a waste of time? Journal of Applied Ecology
49(1): 10-19: 10.1111/j. 1365-2664.2011 02069.x.
Immerzeel DJ, Verweij PA, Hilst Fvd and Faaij APC (2014). Biodiversity impacts of bioenergy crop
production: a state-of-the-art review. GCB Bioenergy 6: 183-209:
Jakka SR, Shrestha RB and Gassmann AJ (2016). Broad-spectrum resistance to Bacillus thuringiensis
toxins by western corn rootworm (Diabrotica virgifera virgifera). Scientific Reports 6: 27860:
Jarvie HP, Johnson LT, Sharpley AN, Smith DR, Baker DB, Bruulsema TW and Confesor R (2017).
Increased Soluble Phosphorus Loads to Lake Erie: Unintended Consequences of Conservation
Practices 1 Journal of Environmental Quality 46(1): 123-132: 10.2134/jeq2016.07.0248.
Jarvie HP, Sharpley AN, Flaten D, Kleinman PJA, Jenkins A and Simmons T (2015). The pivotal role of
phosphorus in a resilient water-energy-food security nexus. Journal of Environmental Quality
44(4): 1049-1062: 10.2134/jeq2015.01.0030.
Jin S, Yang L, Danielson P, Homer CG, Fry JA and Xian G (2013). A comprehensive change detection
method for updating the National Land Cover Database to circa 2011. Remote Sensing of
Environment 132: 159-175:
Johnston CA (2013). Wetland losses due to row crop expansion in the Dakota Prairie Pothole Region.
Wetlands 33(1): 175-182:
Johnston CA (2014). Agricultural expansion: land use shell game in the U.S. Northern Plains. Landscape
Ecology 29(1): 81-95: 10.1007/sl0980-013-9947-0.
Karlen DL, Birrell SJ, Johnson JMF, Osborne SL, Schumacher TE, Varvel GE, Ferguson RB, Novak
JM, Fredrick JR, Baker JM, Lamb JA, Adler PR, Roth GW and Nafziger ED (2014).
Multilocation Corn Stover Harvest Effects on Crop Yields and Nutrient Removal. Bioenergy
Research 7(2): 528-539: 10.1007/sl2155-014-9419-7.
Keenan RJ, Reams GA, Achard F, de Freitas JV, Grainger A and Lindquist E (2015). Dynamics of
global forest area: results from the FAO Global Forest Resources Assessment 2015. Forest
Ecology and Management 352: 9-20:
126
-------
Kim S, Dale BE and Ong RG (2012). An alternative approach to indirect land use change: Allocating
greenhouse gas effects among different uses of land. Biomass and Bioenergy 46: 447-452:
https://doi.Org/10.1016/j.biombioe.2012.07.015.
King CW and Webber ME (2008). Water Intensity of Transportation. Environmental Science &
Technology 42(21): 7866-7872: 10.1021/es800367m.
Kloverpris JH and Mueller S (2013). Baseline time accounting: Considering global land use dynamics
when estimating the climate impact of indirect land use change caused by biofuels. The
International Journal of Life Cycle Assessment 18(2): 319-330: 10.1007/sl 1367-012-0488-6.
Kocoloski M, Mullins KA, Venkatesh A and Michael Griffin W (2013). Addressing uncertainty in life-
cycle carbon intensity in a national low-carbon fuel standard. Energy Policy 56: 41-50:
https://doi.Org/10.1016/j.enpol.2012.08.012.
Koh I, Lonsdorf EV, Williams NM, Brittain C, Isaacs R, Gibbs J and Ricketts TH (2016). Modeling the
status, trends, and impacts of wild bee abundance in the United States. Proceedings of the
National Academy of Sciences 113(1): 140-145: 10.1073/pnas.l517685113.
Koponen K, Soimakallio S, Kline KL, Cowie A and Brandao M (2018). Quantifying the climate effects
of bioenergy - Choice of reference system. Renewable and Sustainable Energy Reviews 81:
2271-2280: https://doi.Org/10.1016/j.rser.2017.05.292.
Krupke CH, Hunt GJ, Eitzer BD, Andino G and Given K (2012). Multiple Routes of Pesticide Exposure
for Honey Bees Living Near Agricultural Fields. PLOS ONE 7(1): e29268:
10.13 71 /j ournal .pone .0029268.
Krupke CH and Long EY (2015). Intersections between neonicotinoid seed treatments and honey bees.
Current Opinion in Insect Science 10: 8-13: https://doi.Org/10.1016/j.cois.2015.04.005.
LaBeau MB, Robertson DM, Mayer AS, Pijanowski BC and Saad DA (2014). Effects of future urban
and biofuel crop expansions on the riverine export of phosphorus to the Laurentian Great Lakes.
Ecological Modelling 277: 27-37: https://doi.Org/10.1016/j.ecolmodel.2014.01.016.
Lai R (1997). Residue management, conservation tillage and soil restoration for mitigating greenhouse
effect by C02-enrichment. Soil & Tillage Research 43(1-2): 81-107:
Lai R (2003). Soil erosion and the global carbon budget. Environment International 29(4): 437-450:
10.1016/sO 160-4120(02)00192-7.
Landis DA, Gratton C, Jackson RD, Gross KL, Duncan DS, Liang C, Meehan TD, Robertson BA,
Schmidt TM, Stahlheber KA, Tiedje JM and Werling BP (2017). Biomass and biofuel crop
effects on biodiversity and ecosystem services in the North Central US. Biomass and Bioenergy.
1-12: https://doi.Org/10.1016/j.biombioe.2017.02.003.
Langeveld JWA, Dixon J, Keulen Hv and Quist-Wessel PMF (2014). Analyzing the effect of biofuel
expansion on land use in major producing countries: evidence of increased multiple cropping.
Biofuels, Bioproducts and Biorefining 8: 49-58:
Lark TJ, Salmon JM and Gibbs HK (2015). Cropland expansion outpaces agricultural and biofuel
policies in the United States. Environmental Research Letters 10(4): 10.1088/1748-
9326/10/4/044003.
Le PV, Kumar P and Drewry DT (2011). Implications for the hydrologic cycle under climate change due
to the expansion of bioenergy crops in the Midwestern United States. Proceedings of the
National Academy of Sciences 108(37): 15085-15090:
LeDuc SD, Zhang XS, Clark CM and Izaurralde RC (2017). Cellulosic feedstock production on
Conservation Reserve Program land: potential yields and environmental effects. Global Change
Biology Bioenergy 9(2): 460-468: 10.1111/gcbb. 12352.
Lehmann J and Joseph S (2015). Biochar for environmental management: science, technology and
implementation. New York, NY, USA, Routledge
Leone T, Olin E, Anderson J, Jung H, Shelby M and Stein R (2014). Effects of Fuel Octane Rating and
Ethanol Content on Knock, Fuel Economy, and C02 for a Turbocharged DI Engine. SAE
International Journal of Fuels and Lubricants 7: 9-28:
127
-------
Levy S (2017). Microcystis rising: why phosphorus reduction isn't enough to stop cyanoHABs.
Environmental health perspectives 125(2): A34:
Lewis KC and Porter RD (2014). Global approaches to addressing biofiiel-related invasive species risks
and incorporation into US laws and policies. Ecological Monographs 84 IS 2: 171-201:
10.1890/13-1625.1.
Loecke TD, Burgin AJ, Riveros-Iregui DA, Ward AS, Thomas SA, Davis CA and Clair MAS (2017).
Weather whiplash in agricultural regions drives deterioration of water quality. Biogeochemistry
133(1): 7-15:
Macedo IdC, Nassar AM, Cowie AL, Seabra JEA, Marelli L, Otto M, Wang MQ and Tyner WE (2015).
Greenhouse Gas Emissions from Bioenergy. Bioenergy and Sustainabilitv. G. M. Souza, R. L.
Victoria, C. A. Joly and L. M. Verdade (eds). Paris, France, Scientific Committee on Problems
of the Environment.
Madden M (2015). Comparative Toxicity and Mutagenicity of Soy-Biodiesel and Petroleum-Diesel
Emissions: Overview of Studies from the U.S. EPA, Research Triangle Park, NC. Inhalation
Toxicology 27: 511-514:
Madden M (2016). A Paler Shade of Green? The Toxicology of Biodiesel Emissions: Recent Findings
from Studies with this Alternative Fue. Biochimica et Biophysica Acta 1860: 2856-2862:
Malaj E, von der Ohe PC, Grote M, Kiihne R, Mondy CP, Usseglio-Polatera P, Brack W and Schafer RB
(2014). Organic chemicals jeopardize the health of freshwater ecosystems on the continental
scale. Proceedings of the National Academy of Sciences 111(26): 9549:
Malcolm SA, Aillery M and Weinberg. M (2009). Ethanol and a Changing Agricultural Landscape,
Economic Research Report 86, U.S. Dept. of Agriculture, Economic Research Service.:
Marin FR, Martha GB, Cassman KG and Grassini P (2016). Prospects for increasing sugarcane and
bioethanol production on existing crop area in Brazil. BioScience 66(4): 307-316:
Matlaga DP and Davis AS (2013). Minimizing invasive potential of Miscanthusx giganteus grown for
bioenergy: identifying demographic thresholds for population growth and spread. Journal of
Applied Ecology 50(2): 479-487:
McCrackin ML, Cooter EJ, Dennis RL, Harrison JA and Compton JE (2017). Alternative futures of
dissolved inorganic nitrogen export from the Mississippi River Basin: influence of crop
management, atmospheric deposition, and population growth. Biogeochemistry. 1-15:
10.1007/s 10533-017-0331-z.
McLaughlin SB and Walsh ME (1998). Evaluating environmental consequences of producing
herbaceous crops for bioenergy. Biomass and Bioenergy 14(4): 317-324:
https://doi.org/10.1016/S0961-9534(97) 10066-6.
Meehan TD and Gratton C (2015). A consistent positive association between landscape simplification
and insecticide use across the Midwestern US from 1997 through 2012. Environmental Research
Letters 10: 114001:
Meehan TD, Hurlbert AH and Gratton C (2010). Bird communities in future bioenergy landscapes of the
Upper Midwest. Proceedings of the National Academy of Sciences 107: 18533-18538:
Meehan TD, Werling BP, Landis DA and Gratton C (2011). Agricultural landscape simplification and
insecticide use in the Midwestern United States. Proceedings of the National Academy of
Sciences(\Wy. 11500-11505:
Meyfroidt P, Lambin EF, Erb K-H and Hertel TW (2013). Globalization of land use: distant drivers of
land change and geographic displacement of land use. Current Opinion in Environmental
Sustainability 5: 438-444:
Michalak AM, Anderson EJ, Beletsky D, Boland S, Bosch NS, Bridgeman TB, Chaffin JD, Cho K,
Confesor R, Daloglu I, DePinto JV, Evans MA, Fahnenstiel GL, He L, Ho JC, Jenkins L,
Johengen TH, Kuo KC, LaPorte E, Liu X, McWilliams MR, Moore MR, Posselt DJ, Richards
RP, Scavia D, Steiner AL, Verhamme E, Wright DM and Zagorski MA (2013). Record-setting
algal bloom in Lake Erie caused by agricultural and meteorological trends consistent with
128
-------
expected future conditions. Proceedings of the National Academy of Sciences 110(16): 6448-
6452: 10.1073/pnas.l216006110.
Miles JC, Hua J, Sepulveda MS, Krupke CH and Hoverman JT (2017). Effects of clothianidin on aquatic
communities: Evaluating the impacts of lethal and sublethal exposure to neonicotinoids. PLOS
ONE 12(3): e0174171: 10.1371/journal.pone.0174171.
Mladenoff DJ, Sahajpal R, Johnson CP and Rothstein DE (2016). Recent Land Use Change to
Agriculture in the US Lake States: Impacts on Cellulosic Biomass Potential and Natural Lands.
Plos One 11(2): 10.1371/journal.pone.0148566.
Modernel P, Rossing WA, Corbeels M, Dogliotti S, Picasso V and Tittonell P (2016). Land use change
and ecosystem service provision in Pampas and Campos grasslands of southern South America.
Environmental Research Letters 11(11): 113002:
Morefield PE, LeDuc SD, Clark CM and Iovanna R (2016). Grasslands, wetlands, and agriculture: the
fate of land expiring from the Conservation Reserve Program in the Midwestern United States.
Environmental Research Letters 11(9): 10.1088/1748-9326/11/9/094005.
Mosnier A, Havlik P, Valin H, Baker J, Murray B, Feng S, Obersteiner M, McCarl BA, Rose SK and
Schneider UA (2013). Alternative U.S. biofuel mandates and global GHG emissions: The role of
land use change, crop management and yield growth. Energy Policy 57: 602-614:
https://doi.Org/10.1016/j.enpol.2013.02.035.
Motamed M, L. M and Williams R (2016). Corn area response to local ethanol markets in the United
States: a grid cell level analysis. American Journal of Agricultural Economics 98: 726-743:
Mueller S (2010a). 2008 National dry mill corn ethanol survey. Biotechnology Letters 32(9): 1261-1264:
10.1007/sl0529-010-0296-7.
Mueller S (2010b). Detailed Report: 2008 National Dry Mill Corn Ethanol Survey. University of Illinois
at Chicago, Energy Resources Center: 21 pp. Chicago, May 4.
Muenich RL, Kalcic M and Scavia D (2016). Evaluating the Impact of Legacy P and Agricultural
Conservation Practices on Nutrient Loads from the Maumee River Watershed. Environmental
Science & Technology 50(15): 8146-8154:
Mullins KA, Griffin WM and Matthews HS (2011). Policy Implications of Uncertainty in Modeled Life-
Cycle Greenhouse Gas Emissions of Biofuels. Environmental Science & Technology 45(1): 132-
138: 10.1021/esl024993.
Mumm RH, Goldsmith PD, Rausch KD and Stein HH (2014). Land usage attributed to corn ethanol
production in the United States: Sensitivity to technological advances in corn grain yield,
ethanol conversion, and co-product utilization. Biotechnology for Biofuels 7: 17:
Myers JP, Antoniou MN, Blumberg B, Carroll L, Colborn T, Everett LG, Hansen M, Landrigan PJ,
Lanphear BP, Mesnage R, Vandenberg LN, Saal FSv, Welshons WV and Benbrook CM (2016).
Concerns over use of glyphosate-based herbicides and risks associated with exposures: A
consensus statement. Environmental Health 15: 19:
NAS (2011). Renewable Fuel Standard: Potential Economic and Environmental Effects ofU. S. Biofuel
Policy. National Academy of Sciences. Washington, DC.
NAS (2016). Genetically Engineered Crops: Experiences and Prospects. Washington, DC, The National
Academies Press
Newbold T, Hudson LN, Hill SLL, Contu S, Lysenko I, Senior RA, Borger L, Bennett DJ, Choimes A,
Collen B, Day J, De Palma A, Diaz S, Echeverria-Londono S, Edgar MJ, Feldman A, Garon M,
Harrison MLK, Alhusseini T, Ingram DJ, Itescu Y, Kattge J, Kemp V, Kirkpatrick L, Kleyer M,
Correia DLP, Martin CD, Meiri S, Novosolov M, Pan Y, Phillips HRP, Purves DW, Robinson
A, Simpson J, Tuck SL, Weiher E, White HJ, Ewers RM, Mace GM, Scharlemann JPW and
Purvis A (2015). Global effects of land use on local terrestrial biodiversity. Nature 520: 45:
10.1038/nature 14324
https://www.nature.eom/articles/naturel4324#supplementarv-information.
129
-------
Nickerson C, Harper M, Henrie C, Mayberry R, Shimmin S, Smith B and Smith J (2015). Land Use and
Land Cover Estimates for the United States, prepared for the Interagency Council on
Agricultural and Rural Statistics, subcommittee of the Interagency Council on Statistical Policy.
U.S. Department of Agriculture, Economic Research Service. Washington, DC.
Oshiro WM, Beasley TE, McDaniel KL, Evansky PA, Martin SA, Moser VC, Gilbert ME and Bushnell
PJ (2015). Prenatal exposure to vapors of gasoline-ethanol blends causes few cognitive deficits
in adult rats. Neurotoxicology and Teratology 49: 59-73:
https: //doi. org/10.1016/j .ntt.2015.04.001.
Oshiro WM, Beasley TE, McDaniel KL, Taylor MM, Evansky P, Moser VC, Gilbert ME and Bushnell
PJ (2014). Selective cognitive deficits in adult rats after prenatal exposure to inhaled ethanol.
Neurotoxicology and Teratology 45: 44-58: http://dx.doi.Org/10.1016/i.ntt.2014.07.001.
Otto CRV, Roth CL, Carlson BL and Smart MD (2016). Land-use change reduces habitat suitability for
supporting managed honey bee colonies in the Northern Great Plains. Proceedings of the
National Academy of Sciences 113(37): 10430-10435: 10.1073/pnas.1603481113.
Palik DJ, Snow AA, Stottlemyer AL, Miriti MN and Heaton EA (2016). Relative performance of non-
local cultivars and local, wild populations of switchgrass (Panicum virgatum) in competition
experiments. Plos One: ll:e0154444.
Parish ES, Hilliard MR, Baskaran LM, Dale VH, Griffiths NA, Mulholland PJ, Sorokine A, Thomas NA,
Downing ME and Middleton RS (2012). Multimetric spatial optimization of switchgrass
plantings across a watershed. Biofuels, Bioproducts and Biorefining 6(1): 58-72:
10.1002/bbb.342.
Perry ED, Ciliberto F, Hennessy DA and Moschini G (2016). Genetically engineered crops and pesticide
use in US maize and soybeans. Science Advances 2(8): el600850:
Pfister S and Bayer P (2014). Monthly water stress: spatially and temporally explicit consumptive water
footprint of global crop production. Journal of Cleaner Production 73: 52-62:
http://dx.doi.Org/10.1016/i.iclepro.2013.ll.031.
Pisa LW, Amaral-Rogers V, Belzunces LP, Bonmatin JM, Downs CA, Goulson D, Kreutzweiser DP,
Krupke C, Liess M, McField M, Morrissey CA, Noome DA, Settele J, Simon-Delso N, Stark JD,
Van der Sluijs JP, Van Dyck H and Wiemers M (2015). Effects of neonicotinoids and fipronil on
non-target invertebrates. Environmental Science and Pollution Research 22(1): 68-102:
10.1007/sl 1356-014-3471-x.
Pleasants JM and Oberhauser KS (2013). Milkweed loss in agricultural fields because of herbicide use:
effect on the monarch butterfly population. Insect Conservation and Diversity 6(2): 135-144:
10.1111/j .1752-4598.2012.00196.X.
Plevin R, Delucchi MA and O'Hare M (2017a). Fuel carbon intensity standards may not mitigate climate
change. Energy Policy 105(Supplement C): 93-97: https://doi.Org/10.1016/j.enpol.2017.02.037.
Plevin RJ, Beckman J, Golub AA, Witcover J and O'Hare M (2015). Carbon accounting and economic
model uncertainty of emissions from biofuels-induced land use change. Environmental science
& technology 49(5): 2656-2664:
Plevin RJ, Delucchi MA and O'Hare M (2017b). Fuel carbon intensity standards may not mitigate
climate change. Energy Policy 105: 93-97: https://doi.Org/10.1016/j.enpol.2017.02.037.
Plevin RJ, O'Hare M, Jones AD, Torn MS and Gibbs HK (2010). Greenhouse Gas Emissions from
Biofuels' Indirect Land Use Change Are Uncertain but May Be Much Greater than Previously
Estimated. Environmental Science & Technology 44(21): 8015-8021: 10.1021/es 101946t.
Plourde JD, Pijanowski BC and Pekin BK (2013). Evidence for increased monoculture cropping in the
Central United States. Agriculture, ecosystems & environment 165: 50-59:
POET-DSM. (2017). "Responsible stover harvest: the benefits and detriments of corn stover harvest."
Retrieved 5/17, 2017.
130
-------
Qin ZC, Dunn JB, Kwon HY, Mueller S and Wander MM (2016). Soil carbon sequestration and land use
change associated with biofuel production: empirical evidence. Global Change Biology
Bioenergy 8(1): 66-80: 10.1111/gcbb. 12237.
Quinn LD, Barney JN, McCubbins JSN and Endres AB (2013). Navigating the "Noxious" and
"Invasive" Regulatory Landscape: Suggestions for Improved Regulation. Bioscience 63(2): 124-
131: 10.1525/bio.2013.63.2.8.
Quinn LD, Gordon DR, Glaser A, Lieurance D and Flory SL (2015a). Bioenergy Feedstocks at Low
Risk for Invasion in the USA: a "White List" Approach. Bioenergy Research 8(2): 471-481:
10.1007/sl2155-014-9503-z.
Quinn LD, Scott EC, Endres AB, Barney JN, Voigt TB and McCubbins J (2015b). Resolving regulatory
uncertainty: Legislative language for potentially invasive bioenergy feedstocks. GCB Bioenergy
7:909-915:
Ray DK and Foley JA (2013). Increasing global crop harvest frequency: recent trends and future
directions. Environmental Research Letters 8(4): 044041:
Reitsma KD, Clay DE, Clay SA, Dunn BH and Reese C (2016). Does the US Cropland Data Layer
Provide an Accurate Benchmark for Land-Use Change Estimates? Agronomy Journal 108(1):
266-272: 10.2134/agronj2015.0288.
Ren J, Campbell JB and Shao Y (2016). Spatial and temporal dimensions of agricultural land use
changes, 2001-2012, East-Central Iowa. Agricultural Systems 148: 149-158:
10.1016/j.agsy.2016.07.007.
Richardson DM and Blanchard R (2011). Learning from our mistakes: minimizing problems with
invasive biofuel plants. Current Opinion in Environmental Sustainability 3(1-2): 36-42:
10.1016/j.cosust.2010.11.006.
Rippey BR (2015). The U.S. drought of 2012. Weather and Climate Extremes 10: 57-64:
https://doi.Org/10.1016/j.wace.2015.10.004.
Roberts MJ and Schlenker W (2013). Identifying Supply and Demand Elasticities of Agricultural
Commodities: Implications for the US Ethanol Mandate. American Economic Review 103(6):
2265-2295: doi: 10.1257/aer.l03.6.2265.
Robertson BA, Doran PJ, Loomis LR, Robertson J and Schemske DW (2011). Perennial biomass
feedstocks enhance avian diversity. GCB Bioenergy 3(3): 235-246:
Robertson BA, Rice RA, Sillett TS, Ribic CA, Babcock BA, Landis DA, Herkert JR, Fletcher RJ,
Fontaine JJ, Doran PJ and Schemske DW (2012). Are Agrofuels a Conservation Threat or
Opportunity for Grassland Birds in the United States? The Condor 114(4): 679-688:
10.1525/cond.2012.110136.
Robertson GP, Hamilton SK, Barham BL, Dale BE, Izaurralde RC, Jackson RD, Landis DA, Swinton
SM, Thelen KD and Tiedje JM (2017). Cellulosic biofuel contributions to a sustainable energy
future: Choices and outcomes. Science 356(6345): 10.1126/science.aal2324.
Rothlisberger-Lewis KL, Foster JL and Hons FM (2016). Soil carbon and nitrogen dynamics as affected
by lipid-extracted algae application. Geoderma 262: 140-146:
http://dx.doi.Org/10.1016/i.geoderma.2015.08.018.
Sanchez ST, Woods J, Akhurst M, Brander M, O'Hare M, Dawson TP, Edwards R, Liska AJ and
Malpas R (2012). Accounting for indirect land-use change in the life cycle assessment of biofuel
supply chains. Journal of The Royal Society Interface 9: 1105-1119:
Scott DA and Page-Dumroese DS (2016). Wood Bioenergy and Soil Productivity Research. Bioenergy
Research 9(2): 507-517: 10.1007/sl2155-016-9730-6.
Scown CD, Horvath A and McKone TE (2011). Water Footprint of U.S. Transportation Fuels.
Environmental Science & Technology 45(7): 2541-2553: 10.1021/esl02633h.
Scown CD, Nazaroff WW, Mishra U, Strogen B, Lobscheid AB, Masanet E, Santero NJ, Horvath A and
McKone TE (2012). Lifecycle greenhouse gas implications of US national scenarios for
cellulosic ethanol production. Environ Res Lett 7: 10.1088/1748-9326/7/1/014011.
131
-------
Seagraves MP and Lundgren JG (2012). Effects of neonicitinoid seed treatments on soybean aphid and
its natural enemies. Journal of Pest Science 85: 125-132:
Searchinger T, Edwards R, Mulligan D, Heimlich R and Plevin R (2015). Do biofiiel policies seek to cut
emissions by cutting food? Science 347(6229): 1420-1422: 10.1126/science.1261221.
Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J, Tokgoz S, Hayes D and Yu T-
H (2008). Use of US croplands for biofuels increases greenhouse gases through emissions from
land-use change. Science 319(5867): 1238-1240: 10.1126/science. 1151861.
Shao Y, Taff GN, Ren J and Campbell JB (2016). Characterizing major agricultural land change trends
in the Western Corn Belt. Isprs Journal of Photogrammetry and Remote Sensing 122: 116-125:
10.1016/j.isprsjprs.2016.10.009.
Smart MD, Pettis JS, Euliss N and Spivak MS (2016). Land use in the Northern Great Plains region of
the U.S. influences the survival and productivity of honey bee colonies. Agriculture, Ecosystems
& Environment 230: 139-149:
Smith LL and Barney JN (2014). The Relative Risk of Invasion: Evaluation of Miscanthus X giganteus
Seed Establishment. Invasive Plant Science and Management 7(1): 93-106: 10.1614/ipsm-d-13-
00051.1.
Solomon BD, Baneijee A, Acevedo A, Halvorsen KE and Eastmond A (2015). Policies for the
sustainable development of biofuels in the pan American region: A review and synthesis of five
countries. Environmental Management 56: 1276-1294:
Souza GM, Victoria R, Joly C and Verdade L, Eds. (2015). Bioenergy & Sustainability: Bridging the
gaps. Paris, SCOPE.
Sparks DL (2003). Environmental soil chemistry. San Diego, CA, USA, Academic Press
Stein RA, Anderson JE and Wallington TJ (2013). An overview of the effects of ethanol-gasoline blends
on SI engine performance, fuel efficiency, and emissions. SAE International Journal of Engines
6: 470-487:
Strogen B, Horvath A and McKone T (2012). Fuel Miles and the Blend Wall: Costs and Emissions from
Ethanol Distribution in the United States. Environmental Science & Technology 46: 5285-5293:
Taheripour F, Hertel TW and Liu J (2013a). The role of irrigation in determining the global land use
impacts of biofuels. Energy, Sustainability and Society 3(1): 4:
Taheripour F and Tyner WE (2013b). Induced Land Use Emissions due to First and Second Generation
Biofuels and Uncertainty in Land Use Emission Factors. Economics Research International
2013: 12: 10.1155/2013/315787.
Taranu ZE, Gregory - Eaves I, Steele RJ, Beaulieu M and Legendre P (2017). Predicting microcystin
concentrations in lakes and reservoirs at a continental scale: A new framework for modelling an
important health risk factor. Global Ecology and Biogeography:
Tessum C, Marshall J and Hill J (2012). A spatially and temporally explicit life cycle inventory of air
pollutants from gasoline and ethanol in the United States. Environmental Science & Technology
46: 11408-11417:
Theiss TJ, Alleman T, Brooker A, Elgowainy A, Fioroni G, Han J, Huff SP, Johnson C, Kass MD, Leiby
PN, Martinez RU, McCormick R, Moriarty K, Newes E, Oladosu GA, Szybist JP, Thomas JF,
Wang M and West BH (2016). Summary of High-Octane Mid-Level Ethanol Blends Study. Oak
Ridge National Lab.; National Renewable Energy Lab.; Argonne National Lab., ORNL/TM-
2016/42: Medium: ED; Size: 29 p. pp.
Timilsina GR, Beghin JC, Mensbrugghe Dvd and Mevel S (2012). The impacts of biofuels targets on
land-use change and food supply: A global CGE assessment. Agricultural Economics(43): 315—
332:
Tokgoz S and Laborde D (2014). Indirect land use change debate: What did we learn? Current
Sustainable/Renewable Energy Reports(l): 104-110:
132
-------
Tolbert VR, Todd DE, Mann LK, Jawdy CM, Mays DA, Malik R, Bandaranayake W, Houston A, Tyler
D and Pettry DE (2002). Changes in soil quality and below-ground carbon storage with
conversion of traditional agricultural crop lands to bioenergy crop production. Environmental
Pollution 116: S97-S106: https://doi.org/10.1016/S0269-7491(01)00262-7.
Trybula EM, Cibin R, Burks JL, Chaubey I, Brouder SM and Volenec JJ (2015). Perennial rhizomatous
grasses as bioenergy feedstock in SWAT: parameter development and model improvement.
GCB Bioenergy 7(6): 1185-1202: 10.1111/gcbb. 12210.
Turner RE and Rabalais NN (2016). 2016forecast: Summer hypoxic zone size Northern Gulf of Mexico.
Louisiana Universities Marine Consortium: 14 pp.
Tyner WE, Taheripour F, Zhuang Q, Birur D and Baldos U (2010). Land use changes and consequent
CO2 emissions due to US corn ethanolproduction. A comprehensive analysis. Purdue
University, Department of Agricultural Economics
UNCCD (2017). The Global Land Outlook. Bonn, Germany, United Nations
USDA-NRCS (2010). Assessment of the effects of conservation practices on cultivated cropland in the
Upper Mississippi River Basin. N. R. C. S. United States Department of Agriculture.
https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/pub/?cid=nrcs 143
014161
USDA-NRCS. (2017). "Soil health glossary." Retrieved 5/1, 2017, from
https://www.nrcs.usda.gov/wps/portal/nrcs/detailfull/soils/health/?cid=nrcsl42p2 053848.
USDA (2012). Weed Risk Assessment for Arundo donax L. (Poaceae) - Giant reed Animal Plant and
Health Inspection Service, U.S. Department of Agriculture: 18 pp. Raleigh, NC, June 14.
USDA (2013). Farm and Ranch Irrigation Survey. US Department of Agriculture, AC-12-SS-1: 266 pp.
Washington, DC, November.
USDA (2014). 2012 Census of Agriculture. U. S. Department of Agriculture National Agricultural
Statistics Service, AC-12-A-51. Washington, DC, May.
USDA (2015). Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
Service, Washington, DC, and Center for Survey Statistics and Methodology, Iowa State
University, Ames, Iowa, http://www.nrcs.usda.gov/technical/nri/12summary.
USDA (2017). USDA Agricultural Projections to 2026. Office of the Chief Economist, U.S. Department
of Agriculture, OCE-2017-1: 106 pp.
USGS (2017). Water-Quality Changes in the Nation's Streams and Rivers, United States Geological
Survey.
Valin H, Peters D, van den Berg M, Frank S, Havlik P, Forsell N, Hamelinck C, Pirker J, Mosnier A and
Balkovic J (2015). The land use change impact of biofuels consumed in the EU: Quantification
of area and greenhouse gas impacts.
van der Sluijs JP, Amaral-Rogers V, Belzunces LP, Bijleveld van Lexmond MFIJ, Bonmatin J-M,
Chagnon M, Downs CA, Furlan L, Gibbons DW, Giorio C, Girolami V, Goulson D,
Kreutzweiser DP, Krupke C, Liess M, Long E, McField M, Mineau P, Mitchell EAD, Morrissey
CA, Noome DA, Pisa L, Settele J, Simon-Delso N, Stark JD, Tapparo A, Van Dyck H, van
Praagh J, Whitehorn PR and Wiemers M (2015). Conclusions of the Worldwide Integrated
Assessment on the risks of neonicotinoids and fipronil to biodiversity and ecosystem
functioning. Environmental Science and Pollution Research 22(1): 148-154: 10.1007/sl 1356-
014-3229-5.
van Wijnen J, Ivens WPMF, Kroeze C and Lohr AJ (2015). Coastal eutrophication in Europe caused by
production of energy crops. Science of The Total Environment 511: 101-111:
https://doi.Org/10.1016/j.scitotenv.2014.12.032.
VanLoocke A, Twine TE, Kucharik CJ and Bernacchi CJ (2017). Assessing the potential to decrease the
Gulf of Mexico hypoxic zone with Midwest US perennial cellulosic feedstock production. GCB
Bioenergy 9(5): 858-875: 10.1111/gcbb. 12385.
133
-------
VanLoocke A, Twine TE, Zeri M and Bernacchi CJ (2012). A regional comparison of water use
efficiency for miscanthus, switchgrass and maize. Agricultural and Forest Meteorology 164: 82-
95: https://doi.Org/10.1016/j.agrformet.2012.05.016.
Wade T, Claassen R and Wallander S (2015). Conservation-practice adoption rates vary widely by crop
and region, EIB-147, US Department of Agriculture Economic Research Service.
Wallander S, Claassen R and Nickerson C (2011). The Ethanol Decade: An Expansion of U.S. Corn
Production, 2000-09. U.S. Department of Agriculture, Economic Research Service, EIB-79.
August.
Werling BP, Dickson TL, Isaacs R, Gaines H, Gratton C, Gross KL, Liere H, Malmstrom CM, Meehan
TD, Ruan L, Robertson BA, Robertson GP, Schmidt TM, Schrotenboer AC, Teal TK, Wilson JK
and Landis DA (2014). Perennial grasslands enhance biodiversity and multiple ecosystem
services in bioenergy landscapes. Proceedings of the National Academy of Sciences 111(4):
1652-1657: 10.1073/pnas. 1309492111.
West N, M„ Matlaga DP, Muthukrishnan R, Spyreas G, Jordan NR, Forester JD and Davis AS (2017).
Lack of impacts during early establishment highlights a short-term management window for
minimizing invasions from perennial biomass crops. Frontiers in Plant Science 8:
10.3389/fpls.2017.00767.
West TO and Post WM (2002). Soil organic carbon sequestration rates by tillage and crop rotation: A
global data analysis. Soil Science Society of America Journal 66(6): 1930-1946:
Wicke B, Sikkema R, Dornburg V and Faaij A (2011). Exploring land use changes and the role of palm
oil production in Indonesia and Malaysia. Land Use Policy 28(193-206):
Williams A, McCormick R, Luecke J, Brezny R, Geisselmann A, Voss K, Hallstrom K, Leustek M,
Parsons J and Abi-Akar H (2011). Impact of Biodiesel Impurities on the Performance and
Durability of DOC, DPF and SCR Technologies. SAE International Journal of Fuels and
Lubricants4: 110-124:
Woltjer G, Daioglou V, Elbersen B, Ibanez GB, Smeets EMW, Gonzalez DS and Barno JG (2017).
Study Report on Reporting Requirements on Biofuels and Bioliquids Stemming from the
Directive (EU) 2015/1513. Brussels, European Commission
Woltjer G, Daioglou, V., Elbersen, B.,Ibanez, G.B., Smeets, E., Gonzalez, D.S., Barno, J.G. (2017).
STUDY REPORT ON REPORTING REQUIREMENTS ON BIOFUELS AND BIOLIQUIDS
STEMMING FROM THE DIRECTIVE (EU) 2015/1513. European Comission: 124:
Wright CK, Larson B, Lark TJ and Gibbs HK (2017). Recent grassland losses are concentrated around
US ethanol refineries. Environmental Research Letters 12(4): 10.1088/1748-9326/aa6446.
Wright CK and Wimberly MC (2013). Recent land use change in the Western Corn Belt threatens
grasslands and wetlands. Proceedings of the National Academy of Sciences 110(10): 4134-4139:
10.1073/pnas. 1215404110.
Wu M, Demissie Y and Yan E (2012a). Simulated impact of future biofiiel production on water quality
and water cycle dynamics in the Upper Mississippi river basin. Biomass and Bioenergy 41: 44-
56:
Wu M, Zhang Z and Chiu Y-w (2014). Life-cycle Water Quantity and Water Quality Implications of
Biofuels. Current Sustainable/Renewable Energy Reports 1(1): 3-10: 10.1007/s40518-013-0001 -
2.
Wu Y and Liu S (2012b). Impacts of biofuels production alternatives on water quantity and quality in the
Iowa River Basin. Biomass and Bioenergy 36: 182-191: 10.1016/j.biombioe.2011.10.030.
Yanowitz J, Knoll K, Kemper J, Luecke J and McCormick R Impact of adaptation on flex-fuel vehicle
emissions when fueled with E40. Environmental Science & Technology 47: 2990-2997:
Yasarer LMW, Sinnathamby S and Sturm BSM (2016). Impacts of biofuel-based land-use change on
water quality and sustainability in a Kansas watershed. Agricultural Water Management 175: 4-
14: 10.1016/j.agwat.2016.05.002.
134
-------
Yu T, English B, Larson J, Fu J, Ugarte DDLT, Yun J, Calcagno J and Wilson B (2013). Modeling the
Air Quality Impacts of Feedstocks Transportation for Cellulosic Biofuel Production in
Tennessee. TRB 2013 Annual Meeting. Washington, DC, Transportation Research Board.
Zhang Y, Heath G, Carpenter A and Fisher N (2016). Air pollutant emissions inventory of large - scale
production of selected biofuels feedstocks in 2022. Biofuels, Bioproducts and Biorefming 10:
56-69:
Zilberman D, Barrows G, Hochman G and Rajagopal D (2013). On the Indirect Effect of Biofuel.
American Journal of Agricultural Economics 95(5): 1332-1337: 10.1093/ajae/aat038.
Ziolkowska JR and Simon L (2014). Recent developments and prospects for algae-based fuels in the US.
Renewable & Sustainable Energy Reviews 29: 847-853: 10.1016/j.rser.2013.09.021.
Zuber SM, Behnke GD, Nafziger ED and Villamil MB (2015). Crop Rotation and Tillage Effects on Soil
Physical and Chemical Properties in Illinois. Agronomy Journal 107(3): 971-978:
10.2134/agronj 14.0465.
135
-------
Appendix A: Abbreviations and Glossary
Abbreviations
AMOX ammonia oxidation catalyst
bbl barrel
BOD biological oxygen demand
Bt Bacillus thuringiensis
CAFE corporate average fuel economy
CDL crop data layer
CDPF catalyzed diesel particulate filter
CH4 methane
CNG compressed natural gas
CO carbon monoxide
C02 carbon dioxide
CRP Conservation Reserve Program
DDGS distillers dregs and waste, or distiller's dried grain and solubles
DOC diesel oxidation catalyst
DOT U.S. Department of Transportation
DRP dissolved reactive phosphorus
EXX fuel blend of XX volume % ethanol and remainder gasoline
EISA Energy Independence and Security Act of 2007
EPIC Environmental Policy Integrated Climate (model)
FAO Food and Agriculture Organization (United Nations)
FAPRI Food and Agricultural Policy Research Institute (model)
FASOM Forest and Agricultural Sector Optimization Model
FFV Flexible-fuel vehicle
gal U.S. gallon
GHG(s) greenhouse gas(es)
HAB(s) harmful algal bloom(s)
HC hydrocarbon
HHDDE heavy-heavy-duty diesel engine
HT herbicide tolerant
LCA life cycle assessment
LCI life cycle inventory
LHDDE light-heavy-duty diesel engine
LNG liquefied natural gas
LUC land use change
MHDDE medium-heavy duty diesel engine
MOVES MOtor Vehicle Emissions Simulator (model)
MRB Mississippi River Basin
N nitrogen
NAIP National Agriculture Imagery Program (USDA)
NASS National Agricultural Statistics Service (USDA)
136
-------
NEI National Emission Inventory
NEWS Nutrient Export from Watersheds (model)
NH3 ammonia
NMHC non-methane hydrocarbons
NMOG non-methane organic gas
NOx oxides of nitrogen
NREL National Renewable Energy Research Laboratory (DOE)
NRI National Resources Inventory (USDA)
NTE not to exceed (emissions)
P phosphorus
POLYSYS Policy Analysis System (model)
PM particulate matter
PM2.5 particulate matter with aerodynamic diameter less than 2.5 (j,m
RFS Renewable Fuel Standard
RFS2 revised Renewable Fuel Standard
RPA Resource Planning Act
RIA regulatory impact analysis
RIN Renewable Identification Number
RON research octane number
RQ risk quotient
SCR selective catalytic reduction
SOM soil organic matter
SOx sulfur oxides
SPARROW SPAtially Referenced Regressions On Watershed model
SWAT Soil and Water Assessment Tool
TP total phosphorus
USDA U.S. Department of Agriculture
USDA ERS U.S. Department of Agriculture Economic Research Service
US DOE (or DOE) U.S. Department of Energy
US EPA (or EPA) U.S. Environmental Protection Agency
USGS U.S. Geological Survey
VMT vehicle-miles traveled
VOC volatile organic compounds
WRA weed risk assessment
Glossary
advanced biofuel: A renewable fuel, other than ethanol derived from corn starch, that has life cycle
greenhouse gas emissions that are at least 50 percent less than life cycle GHG emissions from petroleum
fuel. Cellulosic biofuels must achieve a 60 percent reduction in GHG to get credit for being "advanced."
biochar: the product of heating biomass in the absence of- or with limited air, with the resulting material
rich in organic carbon (Lehmann et al. 2015). This material can be used as a soil amendment.
137
-------
biodiesel (also known as "biomass-based diesel"): A renewable fuel produced through transesterification
of organically derived oils and fats. May be used as a replacement for, or component of, diesel fuel.
According to 40 CFR 80.1401, biodiesel means "amono-alkyl ester that meets ASTM D6751 ('Standard
Specification for Biodiesel Fuel Blend Stock (B100) for Middle Distillate Fuels') "
biodiversity: The variety and variability among living organisms and the ecological complexes in which
they occur. Biodiversity can be defined as the number and relative frequency of different items, from
complete ecosystems to the biochemical structures that are the molecular basis of heredity. Thus, the
term encompasses ecosystems, species, and genes.
biofuel: Any fuel made from organic materials or their processing and conversion derivatives.
biofuel production: The process or processes involved in converting a feedstock into a consumer-ready
biofuel.
biorefinery: A facility that converts biomass into fuels, heat, chemicals and other products using a
variety of processes and equipment.
blendwall: The amount of ethanol that gasoline companies are permitted to blend into petroleum based
fuel, current 10 percent (E10).
conservation tillage: Any tillage practice leaving at least 30% of the soil surface covered by crop
residues; whereas conventional tillage leaves less than 15% of the ground covered by crop residues (Lai
1997). No-till management, a subset of conservation tillage, disturbs the soil only marginally by cutting
a narrow planting strip.
corn stover: The stalks, leaves, husks, and cobs that are not removed from the fields when corn is
harvested.
direct land use change: Land conversion that is directly related to the biofuel supply chain. An example
of direct land use change would be the planting of biofuel feedstock on land that was previously native
forest, to increase the supply of ethanol to export to the U.S.
double cropping: A form of agricultural intensification. Practice of growing two crops on the same
piece of land during a single growing season.
ecosystem health: The ability of an ecosystem to maintain its metabolic activity level and internal
structure and organization, and to resist external stress over time and space scales relevant to the
ecosystem.
ethanol (also known as "bioethanol"): A colorless, flammable liquid produced by fermentation of
sugars. Ethanol is generally blended with gasoline and used as a fuel oxygenate.
extensification: The expansion of agricultural land, like row crops, onto previously uncultivated land.
feedstock: In the context of biofuel, "feedstock" refers to a biomass-based material that is converted for
use as a fuel or energy product.
hypoxia: The state of an aquatic ecosystem characterized by low dissolved oxygen levels (less than 2 to
3 parts per million) due to accelerated algal growth, decay of excess vegetation and algae, and reduced
light penetration because of excessive nutrient levels (eutrophication). Low dissolved oxygen can reduce
fish populations and species diversity in the affected area.
indirect land use change: land conversion that is a market-oriented response to changes in the supply
and demand of goods that arise from increased production of biofuel feedstocks. An example of indirect
land use change would be the clearing of foreign land to plant corn in response to an increase in global
commodity prices caused by a decrease in U.S. corn exports.
138
-------
intensification: Increased intensity of cultivation with no change in total agricultural land acreage
land use change: Conversion of land from one use or cover-type to another. Often human induced.
life cycle assessment: A comprehensive systems approach for measuring the inputs, outputs, and
potential environmental impacts of a product or service over its life cycle, including resource
extraction/generation, manufacturing/production, use, and end-of-life management.
renewable biofuel: A fuel produced from renewable biomass that is used to replace or reduce the use of
fossil fuel.
renewable diesel: Diesel fuel derived from biomass, generally using a thermal depolymerization
process, which meets the requirements of the American Society of Testing and Materials D975 or D396
standards.
row crop: A crop planted in rows wide enough to allow cultivators between the rows. Examples include
corn, soybeans, peanuts, potatoes, sorghum, sugar beets, sunflowers, tobacco, vegetables, and cotton.
soil organic matter (SOM): "The organic fraction of the soil that includes plant and animal residues at
various stages of decomposition, cells and tissues of soil organisms, and substances synthesized by the
soil population" (Brady et al. 2000).
soil quality: "The capacity of a specific kind of soil to function, within natural or managed ecosystem
boundaries, to sustain plant and animal productivity, maintain or enhance water and air quality, and
support human health and habitation. In short, the capacity of the soil to function" (USDA-NRCS 2017).
water availability: In the context of this report, water availability refers to the amount of water that can
be appropriated from surface water sources (e.g., rivers, streams, lakes) or groundwater sources (e.g.,
aquifers) for consumptive uses.
water quality: Water quality is a measure of the suitability of water for a particular use based on
selected physical, chemical, and biological characteristics. It is most frequently measured by
characteristics of the water such as temperature, dissolved oxygen, and pollutant levels, which are
compared to numeric standards and guidelines to determine if the water is suitable for a particular use.
weed risk assessments: Formalized procedures for determining invasion risk. They are designed to
predict invasive and non-invasive species/varieties and distinguish between them based on a set of
questions about their history of invasiveness in other places, biological traits, and suitability for the
environment into which they will be introduced.
139
-------
Appendix B: Key Terms from Major Land Use Change Studies
This Appendix provides the abbreviated definitions of key terms from the four major federal land use
studies discussed in Section 2.4. Full definitions are available in the source literature. Definitions are
indented if they are a subset of a larger category.
USD A Major Uses of Land in the US, 2012 (MLU)
Cropland. Total cropland includes five components: cropland harvested, crop failure, cultivated summer
fallow, cropland used only for pasture, and idle cropland.
Cropland used for crops. Three of the cropland acreage components—cropland harvested, crop
failure, and cultivated summer fallow—are collectively termed cropland used for crops, or the
land used as an input to crop production.
Cropland harvested. Includes row crops and closely sown crops; hay and silage crops;
tree fruits, small fruits, berries, and tree nuts; vegetables and melons; and miscellaneous
other minor crops.
Crop failure. Consists mainly of the acreage on which crops failed because of weather,
insects, and diseases but does include some land not harvested due to lack of labor, low
market prices, or other factors.
Cultivated summer fallow. Refers to cropland in subhumid regions of the West that are
cultivated for one or more seasons to control weeds and accumulate moisture before
small grains are planted.
Cropland pasture. Generally is considered to be in long-term crop rotation. This category
includes acres of crops hogged or grazed but not harvested and some land used for pasture that
could have been cropped without additional improvement.
Idle cropland. Includes land in cover and soil-improvement crops and cropland on which no
crops were planted. Some cropland is idle each year for various physical and economic reasons.
Grassland pasture and range. Grassland pasture and range encompass all open land used primarily for
pasture and grazing, including shrub and brush land types of pasture; grazing land with sagebrush and
scattered mesquite; and all tame and native grasses, legumes, and other forage used for pasture or
grazing—regardless of ownership.
140
-------
Forested land. As defined by the Forest Service, the 766 million acres of forested land in 2012 consist
of "land at least 120 feet (37 meters) wide and at least 1 acre (0.4 hectare) in size with at least 10 percent
cover (or equivalent stocking) by live trees, including land that formerly had such tree cover and that
will be naturally or artificially regenerated."
Timberland. Forestland that produces or is capable of producing crops (in excess of 20 cubic
feet per acre per year) of industrial wood and not withdrawn from timber use by statute or
administrative regulation.
Reserved forestland. Forestland withdrawn from timber use through statute, administrative
regulation, or designation without regard to productive status. Forested wilderness areas and
parks are included in this category.
Other forestland. Forestland other than timberland and productive reserved forestland. It
includes available forestland, which is incapable of annually producing 20 cubic feet (1.4 cubic
meters) per acre (0.4 hectare) of industrial wood under natural conditions because of adverse site
conditions, such as sterile soils, dry climate, poor drainage, high elevation, steepness, or
rockiness.
USDA 2012 Census of Agriculture
Total cropland. This category includes cropland harvested, other pasture and grazing land that could
have been used for crops without additional improvements, cropland on which all crops failed or were
abandoned, cropland in cultivated summer fallow, and cropland idle or used for cover crops or soil
improvement but not harvested and not pastured or grazed.
Harvested cropland. This category includes land from which crops were harvested and hay was
cut, land used to grow short-rotation woody crops, Christmas trees, and land in orchards, groves,
vineyards, berries, nurseries, and greenhouses.
Other pasture and grazing land that could have been used for crops without additional
improvements. This category includes land used only for pasture or grazing that could have
been used for crops without additional improvement. Also included are acres of crops hogged or
grazed but not harvested prior to grazing.
Other cropland. This includes all cropland other than harvested cropland or other pasture and
grazing land that could have been used for crops without additional improvements. It includes
141
-------
cropland idle, used for cover crops or soil improvement, cropland which all crops failed or were
abandoned, and cropland in cultivated summer fallow.
Cropland idle or used for cover crops or soil improvement, but not harvested and
not pastured or grazed. Cropland idle includes any other acreage which could have
been used for crops without any additional improvement and which was not reported as
cropland harvested, cropland on which all crops failed, cropland in summer fallow, or
other pasture or grazing land that could have been used for crops without additional
improvements.
Cropland on which all crops failed or were abandoned. No separate definition.
Cropland in cultivated summer fallow. No separate definition.
USDA 2012 National Resources Inventory
Cropland. A land cover/use category that includes areas used for the production of adapted crops for
harvest. Two subcategories of cropland are recognized: cultivated and noncultivated.
Cultivated cropland comprises land in row crops or close-grown crops and also other cultivated
cropland, for example, hayland or pastureland that is in a rotation with row or close-grown
crops.
Noncultivated cropland includes permanent hayland and horticultural cropland.
Hayland. A subcategory of cropland managed for the production of forage crops that
are machine harvested. The crop may be grasses, legumes, or a combination of both.
Hayland also includes land in set-aside or other short-term agricultural programs.
Horticultural cropland. A subcategory of cropland used for growing fruit, nut, berry,
vineyard, and other bush fruit and similar crops. Nurseries and other ornamental
plantings are included.
Land cover/use. A term that includes categories of land cover and categories of land use. Land cover is
the vegetation or other kind of material that covers the land surface. Land use is the purpose of human
activity on the land; it is usually, but not always, related to land cover.
Pastureland. A land cover/use category of land managed primarily for the production of introduced
forage plants for livestock grazing. Pastureland cover may consist of a single species in a pure stand, a
grass mixture, or a grass-legume mixture. Management usually consists of cultural treatments:
142
-------
fertilization, weed control, reseeding or renovation, and control of grazing. For the NRI, includes land
that has a vegetative cover of grasses, legumes, and/or forbs, regardless of whether or not it is being
grazed by livestock.
Rangeland. A land cover/use category on which the climax or potential plant cover is composed
principally of native grasses, grasslike plants, forbs or shrubs suitable for grazing and browsing, and
introduced forage species that are managed like rangeland. This would include areas where introduced
hardy and persistent grasses, such as crested wheatgrass, are planted and such practices as deferred
grazing, burning, chaining, and rotational grazing are used, with little or no chemicals or fertilizer being
applied. Grasslands, savannas, many wetlands, some deserts, and tundra are considered to be rangeland.
Certain communities of low forbs and shrubs, such as mesquite, chaparral, mountain shrub, and pinyon-
juniper, are also included as rangeland.
Row crops. A subset of the land cover/use category cropland (subcategory, cultivated) comprising land
in row crops, such as corn, soybeans, peanuts, potatoes, sorghum, sugar beets, sunflowers, tobacco,
vegetables, and cotton.
USGS U.S. Conterminous Wall-to-Wall Anthropogenic Land Use Trends,
1974-2012 (NWALT)
Production, Crops. Areas used for the production of crops, such as corn, soybeans, wheat, vegetables,
or cotton, as well as perennial woody crops such as orchards and vineyards. Includes cultivated crops,
row crops, small grains, and fallow fields. Identical definition to NLCD 2011 class 82.
Production, Pasture/Hay. Areas of grasses, legumes, or grass-legume mixtures planted for livestock
grazing or the production of seed or hay crops, typically on a perennial cycle. Identical definition to
NLCD 2011 class 81.
Production, Grazing Potential. Areas of good grazing potential beyond what is indicated by the
NLCD. Information suggests the land could and has been used at least on a seasonal or occasional basis
for animal grazing, including woodland pasture.
FAOSTAT Land Use Data
Arable and Permanent Crops. Arable land is the land under temporary agricultural crops (multiple-
cropped areas are counted only once), temporary meadows for mowing or pasture, land under market
and kitchen gardens and land temporarily fallow (less than five years). The abandoned land resulting
143
-------
from shifting cultivation is not included in this category. Data for "Arable land" are not meant to indicate
the amount of land that is potentially cultivable. Permanent crops is the land cultivated with long-term
crops which do not have to be replanted for several years (such as cocoa and coffee); land under trees
and shrubs producing flowers, such as roses and jasmine; and nurseries (except those for forest trees,
which should be classified under "forest"). Permanent meadows and pastures are excluded from land
under permanent crops.
Area harvested. Data refer to the area from which a crop is gathered. Area harvested, therefore,
excludes the area from which, although sown or planted, there was no harvest due to damage, failure,
etc. It is usually net for temporary crops and some times gross for permanent crops. Net area differs from
gross area insofar as the latter includes uncultivated patches, footpaths, ditches, headlands, shoulders,
shelterbelts, etc. If the crop under consideration is harvested more than once during the year as a
consequence of successive cropping (i.e., the same crop is sown or planted more than once in the same
field during the year), the area is counted as many times as harvested. On the contrary, area harvested
will be recorded only once in the case of successive gathering of the crop during the year from the same
standing crops. With regard to mixed and associated crops, the area sown relating to each crop should be
reported separately. When the mixture refers to particular crops, generally grains, it is recommended to
treat the mixture as if it were a single crop; therefore, area sown is recorded only for the crop reported.
Source: FAO Statistics Division
Forest area is the land spanning more than 0.5 hectares with trees higher than 5 metres and a canopy
cover of more than 10 percent, or trees able to reach these thresholds in situ. It does not include land that
is predominantly under agricultural or urban land use. Forest is determined both by the presence of trees
and the absence of other predominant land uses. The trees should be able to reach a minimum height of 5
metres (m) in situ. Areas under reforestation that have not yet reached but are expected to reach a canopy
cover of 10 percent and a tree height of 5 m are included, as are temporarily unstocked areas, resulting
from human intervention or natural causes, which are expected to regenerate. Includes: areas with
bamboo and palms provided that height and canopy cover criteria are met; forest roads, firebreaks and
other small open areas; forest in national parks, nature reserves and other protected areas such as those of
specific scientific, historical, cultural or spiritual interest; windbreaks, shelterbelts and corridors of trees
with an area of more than 0.5 ha and width of more than 20 m; plantations primarily used for forestry or
protective purposes, such as: rubber-wood plantations and cork, oak stands. Excludes: tree stands in
agricultural production systems, for example in fruit plantations and agroforestry systems. The term also
excludes trees in urban parks and gardens."
144
-------
Permanent meadows and pastures is the land used permanently (for a period of five years or more) for
herbaceous forage crops, either cultivated or naturally growing. A period of five years or more is used to
differentiate between permanent and temporary meadows.
145
-------
® EPA
United States
Environmental Protection
Agency
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC
Recycled/Recyclable
Printed with vegetable-based ink on paper that
contains a minimum of 50% post-consumer
fiber and is processed chlorine free.
------- |