Movement of Certain Genetically Engineered Organisms

DEPARTMENT OF AGRICULTURE

Animal and Plant Health Inspection Service

7 CFR Parts 330, 340, and 372

[Docket No. APHIS-2018-0034]

RIN 0579-AE47

Movement of Certain Genetically Engineered Organisms

AGENCY:

Animal and Plant Health Inspection Service, USDA.

ACTION:

Final rule.

SUMMARY:

We are amending the regulations regarding the movement (importation, interstate movement, and environmental release) of certain genetically engineered organisms in response to advances in genetic engineering and our understanding of the plant pest risk posed by genetically engineered organisms, thereby reducing the regulatory burden for developers of organisms that are unlikely to pose plant pest risks. This final rule, which marks the first comprehensive revision of the regulations since they were established in 1987, provides a clear, predictable, and efficient regulatory pathway for innovators, facilitating the development of genetically engineered organisms that are unlikely to pose plant pest risks.

DATES:

Effective August 17, 2020. Sections 340.4 and 340.5 are applicable beginning April 5, 2021.

FOR FURTHER INFORMATION CONTACT:

Dr. Alan Pearson, Assistant Deputy Administrator, Biotechnology Regulatory Services, APHIS, 4700 River Road, Unit 98, Riverdale, MD 20737-1238; (301) 851-3944.

SUPPLEMENTARY INFORMATION:

Background

The Animal and Plant Health Inspection Service (APHIS) of the United States Department of Agriculture (USDA) administers the regulations in 7 CFR part 340, “Introduction of Organisms and Products Altered or Produced Through Genetic Engineering Which are Plant Pests or Which There is Reason to Believe are Plant Pests” (referred to below as “the regulations”).

These regulations govern the introduction (importation, interstate movement, or release into the environment) of certain genetically engineered (GE) organisms.

Along with the Environmental Protection Agency (EPA), and the Food and Drug Administration (FDA), APHIS is responsible for the oversight and review of GE organisms. In 1986, the Coordinated Framework for Regulation of Biotechnology (Coordinated Framework)

1

was published by the Office of Science and Technology Policy. It describes the comprehensive Federal regulatory policy for ensuring the safety of biotechnology research and products and explains how Federal agencies use existing Federal statutes to ensure public health and environmental safety while maintaining regulatory flexibility to avoid impeding the growth of the biotechnology industry. The Coordinated Framework explains the regulatory roles and authorities for APHIS, EPA, and the FDA. The Coordinated Framework was updated in 2017 in light of advances that had occurred since 1986 in the field of biotechnology.

1

To view the 1986 framework, go to

https://www.aphis.usda.gov/brs/fedregister/coordinated_framework.pdf.

To view the 2017 revision to the framework, go to

https://www.aphis.usda.gov/biotechnology/downloads/2017_coordinated_framework_update.pdf.

APHIS first issued these regulations in 1987 under the authority of the Federal Plant Pest Act of 1957 and the Plant Quarantine Act of 1912, two acts that were subsumed into the Plant Protection Act (PPA, 7 U.S.C. 7701

et seq.

) in 2000, along with other provisions. Since 1987, APHIS has amended the regulations six times, in 1988, 1990, 1993, 1994, 1997, and 2005, to institute exemptions from the requirement for permits to conduct activities for certain microorganisms and

Arabidopsis,

to institute the current notification process and petition procedure, and to exclude plants engineered to produce industrial compounds from the notification process.

While the regulations have been effective in ensuring the safe introduction of GE organisms during the past 30 years, they do not reflect the findings from APHIS' three decades of experience in evaluating GE organisms for plant pest risk or account for developments in genetic engineering over that period. APHIS' evaluations to date have provided evidence that genetically engineering a plant with a plant pest as a vector, vector agent, or donor does not result in a GE plant that presents a plant pest risk. Further, genetic engineering techniques have been developed that do not employ plant pests as donor organisms, recipient organisms, vectors, or vector agents, yet may result in organisms that do pose a plant pest risk. Given these developments, as well as legal and policy issues discussed below, it has become necessary, in our view, to update our regulations accordingly.

On January 19, 2017, we published in the

Federal Register

(82 FR 7008-7039, Docket No. APHIS-2015-0057) a proposed rule

2

intended to revise our regulatory approach from “regulate first before analyzing risks” to “analyze plant pest and noxious weed risks of GE organisms prior to imposing regulatory restrictions.”

2

To view the 2017 proposed rule, the subsequent withdrawal, all supporting documents, and comments APHIS received, go to

http://www.regulations.gov/#!docketDetail;D=APHIS-2015-0057.

Under the January 2017 proposed rule, a stakeholder could request that we conduct a risk assessment to determine whether a GE organism would pose plant pest or noxious weed risks and thus need to be regulated. Regulated GE organisms could be imported, moved interstate, or released into the environment under a flexible, risk-based permitting procedure.

APHIS received 203 comments on the proposal during the comment period. Commenters expressed concerns about many provisions of the proposed rule. Many stated that the proposed requirements would be too burdensome and had the potential to stifle innovation.

After reviewing the comments, APHIS subsequently withdrew the proposed rule. Following the withdrawal, APHIS conducted extensive outreach. Our outreach efforts took place in all regions of the United States and encompassed all sectors of the agriculture supply chain, as well as academic researchers, growers of various crops, and advocacy groups. Organizations ranged in size from small laboratories to larger scale businesses. APHIS also took proactive steps to meet with organizations both supportive and skeptical of agricultural biotechnology. In total, APHIS met with more than 80 organizations, including 17 universities, State departments of agriculture, and farmer organizations.

Much of the feedback received during this process centered on the need to focus regulatory efforts and oversight upon risk, rather than the method used to develop GE organisms. Stakeholders also expressed a desire for flexible and adaptable regulations so that future innovations do not invalidate the regulations. We also received feedback urging us to keep international trade objectives in mind when proposing new regulations and ensuring that new regulatory requirements are transparent and clearly articulated.

The feedback we received led us to update APHIS' regulatory framework, in a manner that further focuses our regulatory efforts on the properties of the GE organism itself rather than on the

method used to produce it. We believed that this regulatory approach would better reflect our current knowledge of the field of biotechnology and would therefore enable us to evaluate GE organisms for plant pest risk with greater precision than the existing framework allowed. The regulatory framework was also intended to enable APHIS to avoid conducting repetitive analyses, to utilize its staff time more efficiently than before, and to provide better stewardship of taxpayer dollars.

On June 6, 2019, we published in the

Federal Register

(84 FR 26514-26541, Docket No. APHIS-2018-0034) a proposal

3

to amend the regulations in accordance with the Secretary of Agriculture's March 28, 2018, statement on plant breeding innovations. The Secretary's statement and the accompanying explanatory details provided clarification on the USDA's oversight over plants produced through innovative, new breeding techniques, including genome editing techniques. (The statement and further details are available at:

https://www.aphis.usda.gov/aphis/ourfocus/biotechnology/brs-news-and-information/2018_brs_news/plant_breeding.

)

3

To view the proposed rule, the comments we received, and supporting documents, go to

http://www.regulations.gov/#!docketDetail;vD=APHIS-2018-0034.

Additionally, please note that within the body of this document, that rule and this final rule are referred to at times as the Sustainable, Ecological, Consistent, Uniform, Responsible, Efficient (SECURE) rule. The SECURE rule is the nomenclature used by USDA to discuss the rule with stakeholders.

We would note also that the June 2019 proposed rule and this final rule are consistent with the President's “Executive Order on Modernizing the Regulatory Framework for Agricultural Biotechnology Products” (June 11, 2019, Executive Order 13874). Executive Order 13874 directs the Federal Government to adopt regulatory approaches for the products of agricultural biotechnology that are proportionate to the risks such products pose, and that avoid arbitrary or unjustifiable distinctions across like products developed through different technologies. Among other things, Executive Order 13874 states that regulatory decisions should be science- and evidence-based, taking economic factors into account as appropriate and consistent with applicable law; that regulatory reviews should be conducted in a timely and efficient manner; and that biotechnology regulations should be transparent, predictable, and consistent.

We solicited comments on our proposed rule and its supporting analyses until August 6, 2019. We received 6,150 comments by that date. They were from developers of GE organisms; growers of GE plants for food crops and other uses; trade associations representing both of those groups and sellers of such commodities as corn, soybeans, and grain; scientists representing academic institutions; organic farmers and trade associations representing their interests; consumer and public interest groups; and individuals. Most of the comments, while not form letters, expressed a generalized, similarly themed opposition to GE products. Of the comments that specifically addressed the provisions of the rule, approximately 25 expressed some support for the rule. The comments are discussed below by topic.

Applicability of the Regulations

Exemptions

The June 2019 proposed rule exempted from the regulations certain categories of plants that have been modified. Specifically, § 340.1(b)(1) through (4) proposed to exempt such plants if:

• The genetic modification is solely a deletion of any size; or

• The genetic modification is a single base pair substitution; or

• The genetic modification is solely introducing nucleic acid sequences from within the plant's natural gene pool or from editing nucleic acid sequences in a plant to correspond to a sequence known to occur in that plant's natural gene pool; or

• The plant is an offspring of a GE plant and does not retain the genetic modification in the GE plant parent.

In addition to above-listed categories, proposed § 340.1(c) stated that modified plants would not be subject to the regulations if they have plant-trait-mechanism of action (MOA) combinations that are the same as those of modified plants for which APHIS has conducted a regulatory status review (RSR) and found not to be subject to the regulations under part 340.

The above-listed exemptions elicited a broad spectrum of comments. Some commenters welcomed the regulatory relief offered by the exemptions as written, while others viewed them as too broad and still others as excessively restrictive.

Among the commenters who viewed the exemptions as excessively broad, several commenters stated that APHIS did not provide the “necessary scientific justifications” for the exemptions from regulation listed in proposed § 340.1(b)(1) through (3).

The exemptions in § 340.1(b)(1) through (3) are based on the principles listed below. (For reasons discussed later in this document, we are removing from this final rule the exemption contained in § 340.1(b)(4) of the proposed rule, which would have pertained to “null segregants,” or the offspring of a GE plant that does not retain the genetic modification in the GE plant parent; while there is still a paragraph (b)(4) in this final rule, it serves a different purpose which we discuss later in this document.)

1. Plants created through conventional breeding have a history of safe use related to plant pest risk;

2. The types of plants that qualify for these exemptions can also be created through conventional breeding; and

3. There is no evidence that use of recombinant deoxyribonucleic acid (DNA) or genome editing techniques necessarily and in and of itself introduces plant pest risk, irrespective of the technique employed.

When a plant meets one of the above-listed exemptions, therefore, it is not expected to pose plant pest risks greater than the plant pest risks posed by plants modified by conventional breeding methods and thus should rightly not be subjected to regulation under part 340. (The term “conventional breeding” may generally be used interchangeably with “traditional breeding.” In the June 2019 proposed rule, APHIS used both terms, with “traditional breeding” appearing more frequently in the text. Based in part on dialogue with other agencies involved in regulating biotechnology, we have elected to use the term “conventional breeding” throughout this final rule and its supporting documents, except when the need to quote directly indicates otherwise. For purposes of this rule and its supporting documents, “conventional breeding” has the meaning it is understood to have within the context of part 340, based on the examples provided immediately below. Other Federal or State regulations may use the term “conventional breeding” in the context of their regulations and attribute slightly different meanings.)

We noted in the preamble to the June 2019 proposed rule that conventionally bred crops have a long history of safe use with respect to plant pest risk and that the long history of conventional plant breeding gives us extensive experience in safely managing any associated plant pest risks. Conventional breeding techniques generally involve the deliberate selection of plants with desirable traits from existing population genetic variation or from new genetic variation

created through artificial hybridization or induced mutagenesis. As we noted in the June 2019 proposed rule, such techniques include marker-assisted breeding, tissue culture, protoplast, cell, or embryo fusion, and chemical or radiation-based mutagenesis. Products generated solely using such techniques have never been regulated under the part 340 regulations. Although conventional breeding is not risk free, the risks associated with it are, according to a 1989 National Research Council (NRC) report,

4

“manageable by accepted standards.” In other words, the types of traits that can be introduced through conventional breeding have not led to plant pest risk concerns.

4

National Research Council (NRC) 1989. Field Testing Genetically Modified Organisms: Framework for Decisions. Washington DC. National Academy Press. 185 pp. Retrieved from

http://www.nap.edu/catalog/1431.html.

The types of DNA modifications that occur through conventional breeding by mutagenesis are well characterized (Oladosu,

et al.,

2016; Kharkwal,

et al.,

2012). Among the common outcomes that result from mutagenesis are deletions, insertions, inversions, or translocations of DNA and base pair substitutions (Oladosu,

et al.,

2016) which often result from double strand breaks in the DNA followed by natural DNA repair. Base-pair substitution also results from chemical modification of a base followed by natural DNA repair. These types of modifications occur at a low rate from naturally occurring environmental exposure to ionizing radiation, radical oxygen, chemical compounds, or biological agents such as viruses, or at an elevated rate in response to radiation and chemical-induced mutagenesis. In conventional breeding, these types of DNA modifications are introduced randomly. Individual plants possessing a mutation conferring a useful phenotype are isolated by screening, and random mutations that are introduced and do not convey a useful phenotype are addressed during backcrossing. New plant breeding technologies, such as those used in genome editing, can be used to create targeted double strand breaks in specific parts of the genome that when repaired result in deletions and small insertions, just as from natural environmental exposure or radiation mutagenesis (Chen,

et al.,

2019). Likewise, new plant breeding technologies can also be used, in a specific, targeted manner, to create base pair substitutions that are similar to the modifications that can be created by random chemical mutagenesis. In other words, the same types of DNA modifications that occur in conventional breeding can also be constructed precisely using new plant breeding technologies (Custers,

et al.,

2019). We are exempting plants generated using plant breeding technologies that have non-templated insertions and deletions and that have a single base pair substitution, because they could otherwise be created by conventional breeding and pose no increased plant pest risk relative to their conventionally bred counterparts.

The exemption in proposed § 340.1(b)(3) applies to the use of new plant breeding technologies to recreate the introduction of a gene, allele of a gene, or structural variation that could otherwise be introduced by crosses. APHIS notes that conventional methods of plant breeding and new plant breeding technologies often share the same goals with similar results. Human selection of plants has been used for thousands of years; and crossing has been used to introduce alleles into breeding populations since at least the early 18th century (Goulet,

et al.,

2017). More recently, plant breeders have expanded the source of genetic material that can be used to introduce genetic changes into breeding populations through wide crosses, embryo rescue, and protoplast fusion (Bravo,

et al.,

2011; De Filippis, 2014; Singh, 1990), as well as the rate of introduction of genetic material through marker-assisted and genomic selection; all of these approaches are considered conventional breeding methods and are used to expand and guide changes in the gene pool available within a population. Genetic engineering can be used to introduce a genetic sequence from any donor source into plants, which cannot be accomplished through conventional breeding. To limit the exemption in paragraph (b)(3) to what is possible in conventional breeding, the third exemption applies only to the introduction of a gene, allele, or structural variant known to occur from a donor source (1) in the same species as the recipient, or (2) in a species compatible via wide crosses, embryo rescue, or protoplast fusion with the recipient species.

The NRC has concluded in multiple studies

5

that there was no evidence of unique hazards inherent in the use of recombinant DNA techniques with respect to plants, and that crops modified by molecular and cellular methods should pose risks no different from those modified by conventional breeding methods for similar traits. Moreover, new molecular methods for editing genomes have been developed since the NRC studies that can be more specific and precise than those evaluated by the NRC studies, and plants modified by these new methods should also pose plant pest risks that are no different from plants that are modified for similar traits by conventional breeding methods. For all of the foregoing reasons, we consider the exemptions to be based on the best available science.

5

National Academies of Sciences, Engineering, and Medicine (NAS) 1987. Introduction of Recombinant DNA-engineered Organisms into the Environment: Key Issues. Washington, DC National Academy Press. 24 pp. Retrieved from

https://www.nap.edu/read/18907/chapter/1.

National Research Council (NRC) 1989. Field Testing Genetically Modified Organisms: Framework for Decisions. Washington DC. National Academy Press. 185 pp. Retrieved from

http://www.nap.edu/catalog/1431.html.

National Academies of Sciences, Engineering, and Medicine (NAS) 2016. Genetically Engineered Crops: Experiences and Prospects. Washington, DC National Academy Press. 420 pp. doi:10.17226/23395. Retrieved from

http://www.nap.edu/23395.

Some commenters stated that APHIS did not adequately consider risk when developing the exemptions. It was stated that the proposed exemptions do not consider potential pest risks or human, environmental, or agricultural impacts on nontarget organisms. A commenter claimed that APHIS regulates risks other than plant pest risks, such as inadvertent introduction to the food supply and economic impacts from gene flow, so there should be scientific evidence that plants exempted from regulations do not pose any of the full range of risks.

We do not agree with these comments. With regard to the commenters who stated that the exemptions failed to consider impacts on non-target organisms, APHIS considers impacts on non-target organisms that are beneficial to plants to be indirect plant pest impacts. It is not accurate to say that APHIS has previously regulated risks other than plant pest risks. Under the current regulations prior to the effective date of this final rule (referred to below as “the current regulations”), APHIS has imposed measures to limit gene flow from GE plants that already met the definition of a

regulated article.

(Please see the “Implementation Table” on

Regulations.gov

regarding the dates when various provisions of this rule become applicable.) In these cases, APHIS considered the GE plants to be regulated articles because they had used a plant pest as the donor organism, recipient organism, or vector or vector agent, and therefore could pose a plant pest risk. As noted in the proposed rule, APHIS' evaluations to date have provided evidence that genetically

engineering a plant with a plant pest does not in and of itself result in a plant that presents a plant pest risk, however. In cases where GE crops were not subject to regulation, no “other risks” such as inadvertent introduction to the food supply or economic impacts from gene flow have been regulated by APHIS insofar as they were outside the scope of the regulations.

A commenter opposed the exemptions listed in proposed § 340.1(b)(1) through (3) on the basis that plants produced through most methods that would be used for genome editing are regenerated from single cells in tissue culture, resulting in somaclonal variation with unpredictable consequences, and that off-target mutations caused by genome editing are more likely than chemical and radiation mutagenesis to be non-random. A second commenter asked that the exemptions be limited so that they apply only to plants produced using techniques that minimize off-target mutations. A third commenter asked whether off-target mutations are considered when determining eligibility for an exemption.

Somaclonal variation has been utilized extensively for breeding purposes, and the resultant new plant variety is not subject to the APHIS regulations in part 340 that we are replacing with this final rule (Krishna,

et al.,

2016; Neelakandan and Wang, 2012). APHIS is not aware of a reason to mandate government oversight over new plant varieties resulting from somaclonal variation.

Background mutation occurs naturally in plants and does not raise plant pest risk concerns in conventional breeding programs. APHIS does not believe it is necessary to regulate off-target effects of genome editing in plants because (1) the off-target mutation rate from genome editing is low relative to the background mutation rate that occurs in conventional breeding, and (2) whatever changes do occur are likely to be segregated away from the target mutation during the breeding process. Comprehensive CRISPR/Cas off-target analysis on a genome-wide scale has been performed in rice, maize, tomato, and Arabidopsis (Feng,

et al.,

2014; Feng,

et al.,

2018; Peterson,

et al.,

2016; Nekrasov,

et al.,

2017; Lee,

et al.,

2018; Tang,

et al.,

2018). In these cases where the frequency of off-target mutation was measured in CRISPR/Cas expressing lines and their progeny, the authors concluded that the rate of off-target mutation was below the level of background mutation induced during seed amplification or tissue culture (Hahn and Nekrasov, 2019). Although there can be variation in off-target mutation rates due to the nature of the technique used and the biological system to which it is applied, the mutation rates in such conventional breeding techniques as chemical and irradiation-based mutagenesis dwarf the rate associated with such methods.

Due to the nature of plant breeding—in which populations are created and evaluated, and individual plants are selected for the intended modifications—off-target changes are likely to be lost unless they are genetically linked to the targeted modification that is introduced. APHIS wishes to clarify that, for these reasons, off-target mutations are not considered when determining eligibility for an exemption. This is also consistent with APHIS' approach regarding conventional breeding techniques. As noted above, these techniques often have a high mutation rate, but have a history of safe use with respect to plant pest risk. APHIS has modified the regulatory text in § 340.1(b) to indicate that we are considering only targeted modifications when determining eligibility for an exemption.

Some commenters stated that the scope of the exemptions listed in proposed § 340.1(b)(1) through (3) should be broadened to encompass the range of genetic modifications that are accessible to plant breeders through conventional breeding methods, and proposed alternative language that would allow an unlimited number of genetic modifications to be made and exempt from the regulations.

The commenters appear to have interpreted our references in the June 2019 proposed rule and its preamble to plants that could otherwise have been developed through “traditional breeding methods” to mean any type and extent of genetic change that is theoretically possible through conventional breeding methods. There are many biological and practical factors that affect a plant breeder's ability to develop a new crop variety by introducing genetic variation and intentionally selecting for desired traits. These include the number of targeted loci and type of desired genetic changes, the genetic distance between the desired changes, generation time, breeding system (sexual or asexual, self-compatibility), ploidy level and genomic complexity, resource availability (time, money, labor, and genomic resources), and other factors. These factors, and thus the extent of intentionally selected genetic variation that can be introduced, vary widely among plant species. Moreover, new plant breeding techniques can make possible more complex combinations of genetic modifications than can practically be achieved through conventional breeding methods (Custers,

et al.,

2019; Wolter,

et al.,

2019; Najera,

et al.,

2019). Currently, APHIS lacks sufficient familiarity to develop a risk-based exemption for products containing complex combinations that might be produced in the future. APHIS is clarifying that the exemptions listed in § 340.1(b)(1) through (3) are based on types of modifications that are easily recognizable to the developers of the organism and on genetic changes that could be practically achieved by conventional breeding methods in any plant species. However, over time, APHIS expects to gain more familiarity with the products of these new plant breeding innovations. Accordingly, we are revising § 340.1(b) to establish a process for listing additional modifications that plants can contain while still being exempted from the regulations. This process is specified in paragraph (b)(4) of § 340.1 in this final rule.

Some commenters inquired how the exemptions in proposed § 340.1(b)(1) through (3) pertain to combinations of genetic modifications or to sequential edits. For example, would a deletion and a single base substitution made at the same time in a plant qualify for exemption? If a single change is made to a plant, when could another change be made that qualified for an exemption? Some commenters argued that there is no valid scientific reason that the exemptions should not allow multiple simultaneous genomic changes to be made. Other commenters asked us to reaffirm that the exemptions are limited to only a single genome editing change, and that a plant containing multiple changes made at the same or different times would not be exempt, or that we delete the exemptions altogether, since genome edits could be made sequentially such that each intermediate organisms would be exempt, cumulatively resulting in a final organism with many targeted changes that would also be exempt. Several commenters requested that APHIS include a process for adding new categories of exemptions and revising exemptions in order to ensure that the regulatory system stays up to date and keeps pace with advances in scientific knowledge, evidence, and experience.

APHIS seeks to clarify that exemptions listed in § 340.1(b)(1) through (3) apply to plants containing single targeted modifications. The exemptions were formulated to apply to what could otherwise be achieved through conventional plant breeding

techniques in any species. As discussed above, the plants that are eligible for exemption would have no increased plant pest risk than conventionally bred plants. APHIS realizes that in some species, a single targeted modification is often less than what could otherwise be developed through conventional breeding. However, as noted above, the extent of intentionally selected variation that could otherwise be introduced through conventional breeding varies depending on the plant species. To establish clear and unambiguous exemptions that could apply to any plant species while enabling for variation in what can be achieved through conventional breeding, APHIS has revised the regulatory text in § 340.1(b).

Initially, the exemptions will apply only to plants containing a single targeted modification in one of the categories listed. APHIS anticipates scientific information and/or experience may, over time, allow APHIS to list additional modifications that plants can contain and still be exempted from the regulations so that the regulatory system stays up to date and keeps pace with advances in scientific knowledge, evidence, and experience. This may include multiple simultaneous genomic changes. If the Administrator determines that it is appropriate to list additional modifications, APHIS will notify the public in the

Federal Register

and will take public comment. After reviewing the comments, APHIS will issue a subsequent notice announcing its determination. This process is provided in new paragraph (b)(4) in § 340.1.

One commenter requested that APHIS document examples of deletions of any size that could be made by conventional breeding.

The first exemption allows a single deletion of any size because radiation can create any size deletion. As mutations are typically detrimental to the organism, what is achievable in practice is limited by the viability and fertility of the organism. Large mutations can be maintained in a heterozygous state but do not tend to undergo homozygous inheritance (Naito, 2005). For example, in Arabidopsis, which has a genome size of 135 Mb (Arabidopsis Genome Initiative, 2000), a radiation-induced deletion of 3.1 Mb was obtained that disrupted 852 genes and was maintainable only as a heterozygote presumably because genes essential for survival are present in the deleted region (Kazama,

et al.,

2017). Polyploid plants and those with large genomes are better able to accommodate even larger deletions (Men

et al.,

2002). For example, in hexaploid wheat, X-ray mutagenesis was used to create a mutant, Ph1-, widely used in breeding programs, that has a 70 Mb deletion (Sears, 1977). To put the size of this deletion in perspective, it is larger than half of the entire genome of Arabidopsis.

Some commenters recommended that the exemption in § 340.1(b)(1) be broadened to allow for insertions that occur during the natural DNA repair mechanism after double-strand break of the DNA. In the proposed rule, the exemption in paragraph (b)(1) mentions only deletions.

APHIS agrees with the comment. Deletions, small insertions, and combinations of deletions and insertions are all possible outcomes resulting from the cellular mechanisms used to repair DNA breaks that occur naturally or that are induced during conventional plant breeding, and all have been used in conventional plant breeding (Manova and Gruszka, 2015; Wang,

et al.,

2016). The exemption in § 340.1(b)(1) has been revised to reflect all of the possible outcomes of natural DNA repair mechanisms that occur in the absence of a deliberately provided repair template.

A commenter asked that APHIS eliminate the exemptions for deletions and single base pair substitutions, arguing that any type of change in a gene sequence can potentially cause phenotypic changes that have significant consequences.

APHIS disagrees with this argument. Naturally occurring single base pair substitutions and deletions are commonly induced and are widely used to generate new crop varieties in conventional mutation breeding, which includes both chemically induced and irradiation-based mutagenesis (Oladosu,

et al.,

2016; Kharkwal, 2012; Ahloowalia and Maluszynski, 2001). The targeted single base pair substitutions or deletions covered by these exemptions are the same in kind as, and do not pose any increased plant pest risks than, the substitutions or deletions introduced through conventional breeding. Thus, they should not be subject to the regulations.

Many commenters argued that limiting the exemption in proposed § 340.1(b)(1) to a single deletion and the exemption in § 340.1(b)(2) to a single base pair substitution does not take into account that multiple base pair substitutions and/or deletions are routinely and safely introduced into plants using conventional breeding methods, including mutagenesis.

The argument that multiple substitutions or deletions can occur through conventional breeding methods, including mutagenesis, seems to be conflating the specific targeted changes that can be made via genome editing techniques with the multiple random changes that occur during conventional breeding, only one or few of which might contribute to the desired phenotype. In the case of random chemical or radiation mutagenesis, thousands of mutations are introduced into the plant but most are detrimental, or neutral at best. The fact that multiple mutations exist in the plant is a negative feature that needs to be overcome by laboriously self-fertilizing or backcrossing the mutated plant for multiple generations. Even then, a developer may not find an agronomically suitable phenotype. By applying selection, it is possible, though at a very low frequency, to get two desirable mutations in a single mutated line if the mutations are unlinked. It is improbable to get two linked mutations in a single line, particularly if the mutations are sought within the same gene. In contrast, genome editing can easily introduce multiple beneficial changes in one generation, leading to phenotypes that we have not seen by conventional breeding.

The exemptions listed in § 340.1(b) are based on measures that are easily defined, are based on familiarity, and thus are meant to be limited to genetic changes that could practically be achieved by conventional breeding methods in any plant. It is not possible to define a number of such changes greater than one which could practically be achieved by conventional breeding methods in all plant species. The number of changes that can practically be achieved through conventional breeding methods can vary widely from one species to another. For this reason, APHIS is retaining the limitation of a single modification, as this approach ensures that we can identify those plants that pose a plant pest risk. We anticipate that most plants that are not eligible for the exemption and do not pose a plant pest risk will pass through the RSR process quickly.

In addition, as noted above, we are revising § 340.1(b) by adding a new paragraph (b)(4) that establishes a process for listing additional modifications that plants can contain while being exempted from the regulations, based on what could be achieved through conventional plant breeding. Thus, while the exemptions in § 340.1(b)(1) through (3) will initially apply only to plants containing a single modification in one of the categories listed, APHIS anticipates that scientific information and/or experience will,

over time, allow multiple and sequential changes in some species after public notice and comment.

The introductory text of § 340.1(b)(4) provides that the Administrator may propose to exempt plants with additional modifications, based on what could be achieved through conventional breeding. Such proposals may be APHIS-initiated, or in response to a request.

Paragraph (b)(4)(i) sets forth the process for APHIS-initiated proposals. APHIS will publish a notice in the

Federal Register

of the proposal by the Administrator to exempt plants with additional modifications. The notice will make available any supporting documentation, and will request public comment. After reviewing the comments, APHIS will publish a subsequent notice in the

Federal Register

announcing its final determination and responding to the comments received.

Under paragraph (b)(4)(ii), any person may request that APHIS exempt plants developed with additional modifications that could be achieved through conventional breeding. The request will have to include the following supporting information, in writing:

• A description of the modification(s);

• The factual grounds demonstrating that the proposed modification(s) could be achieved through conventional plant breeding;

• Copies of scientific literature, unpublished studies, or other data that support the request; and

• Any information known to the requestor that would be unfavorable to the request.

Paragraph (b)(4)(iii) provides the timeframe for Agency review of such requests. It provides that, after APHIS receives all the information required for a request, APHIS will complete its review of the request and render a final determination within 12 months, except in circumstances that could not reasonably have been anticipated.

Under paragraph (b)(4)(iv) if, after review of the request, APHIS disagrees with the conclusions of the request or determines that there is insufficient evidence that the modification could be achieved through conventional breeding methods, APHIS will deny the request and notify the requestor in writing regarding this denial.

Paragraph (b)(4)(v) provides for Agency actions when we agree with a request. It states that, if APHIS initially determines that the modification could be achieved through conventional breeding methods, APHIS will publish a notice in the

Federal Register

in accordance with the process set forth in § 340.1(b)(4)(i).

Under paragraph (b)(4)(vi), a list specifying the additional modifications allowed will be posted on the APHIS website at

https://www.aphis.usda.gov/aphis/ourfocus/biotechnology.

This list would include both those additional modifications originally proposed by the Administrator and those that originate with a request.

Some commenters suggested a change to the exemption in proposed § 340.1(b)(2) so that it would allow a limitless number of synonymous base pair changes. Synonymous base pair changes, it was stated, do not alter the amino acid composition of the encoded protein. One commenter suggested changing the exemption to allow however many specific and known base pair changes are needed to achieve the intended MOA.

APHIS rejects the first suggestion because synonymous changes can lead, and indeed have been made, to generate significant phenotypic changes,

e.g.,

by altering mRNA splice sites, promoters, and regulatory RNAs. APHIS acknowledges that these types of phenotypic changes could, in principle, also occur through a single deletion, insertion, or base pair change in conventional breeding. However, these types of phenotypic changes are unlikely to be possible in all or perhaps even most genes through deletion or single base pair changes. Moreover, multiple targeted changes within a single gene are generally not likely to be achieved in conventional breeding. Therefore, the exemption will not be broadened to include multiple synonymous base pair changes. However, as discussed below under this same subheading of comment responses, we have revised the exemption in § 340.1(b)(3) to clarify that if multiple sequence changes are needed to generate an allele that will result in the intended phenotype and if those changes are known to occur in the plant's gene pool, the GE plant would qualify for the exemption.

One commenter stated that APHIS should eliminate the exemption in paragraph (b)(3), regarding introducing variation known to occur in the gene pool, because sequences found naturally in closely related, sexually compatible organisms do not necessarily have acceptable risks when introduced into other species. The commenter offered an example, stating that “the introduced nucleic acids can direct the synthesis of toxins, change metabolism in harmful ways, turn on or off genes and metabolic pathways in the genetically engineered host, and make the genetically engineered organism more susceptible to pests and pathogens, or more fit in the wild and more weedy.”

APHIS disagrees with the comment. The commenter is pointing out harms that potentially could occur, and are no less likely to occur, in conventional breeding programs. However, such harms have not materialized in conventional breeding programs because they rarely occur and are intentionally eliminated during the evaluation and selection process (NRC, 1989).

One commenter wished to know whether the exemption in proposed § 340.1(b)(3) supersedes the exemption in § 340.1(b)(1) and (b)(2). Another commenter felt that the exemptions in paragraphs (b)(1) and (b)(2) were too narrow because polymorphisms, insertions, inversions, and multiple megabase deletions and translocations are abundant in nature and frequently induced in breeding programs through mutagenesis.

APHIS seeks to clarify that § 340.1(b)(3) does supersede § 340.1(b)(1) and (b)(2) in the number of changes that can be made under the exemption. APHIS also seeks to clarify that paragraphs (b)(1) and (b)(2) pertain to products of mutagenesis which have not been observed in the gene pool, whereas paragraph (b)(3) applies only to variation already known to occur in the gene pool. Therefore, the exemption in paragraph (b)(3) allows the introduction of a gene,

i.e.,

a functional unit of DNA that encodes an RNA or protein, or of an allele (a variant form of a gene or, for the purposes of this regulation, a genetic sequence) containing multiple sequence changes as long as the allele is known to occur in the gene pool of the plant. With regard to the comment that the exemptions in paragraphs (b)(1) and (b)(2) are unnecessarily restrictive because there are changes abundant in nature not covered by these exemptions, APHIS wishes to clarify that the duplications, inversions, translocations, and transpositions already known to occur in the gene pool would qualify under the exemption in paragraph (b)(3).

Some commenters suggested deleting “natural” from § 340.1(b)(3) because the gene pool of a plant may include variation that has been previously induced through chemical or radiation mutagenesis or that could be introduced via human-assisted wide crosses. Further comments on the exemption in paragraph (b)(3) recommended substituting the phrase “known to occur” with some variation of

“otherwise accessible through traditional plant breeding methods.”

APHIS agrees with the first comment and disagrees with the second. APHIS considers the known and accessible gene pool of a plant to include not only genetic sequences that can be introduced to a plant via crosses that can take place without human assistance, but also genetic sequences that can be introduced to a plant via human-assisted wide crosses between distantly related species. In systems for which breeding techniques such as bridging and embryo rescue have been developed to enable wide crosses, distantly related plants are also considered part of the gene pool. However, these categories may not be considered “natural,” so APHIS is in favor of deleting this term. APHIS is retaining the phrase “known to occur,” however. As discussed above, when we refer to GE plants that could otherwise have been developed through conventional breeding methods, we do not mean any genetic changes that are theoretically possible. Almost any genetic change is theoretically possible, given enough time. APHIS' intention in § 340.1(b)(3) is to exempt from regulation a product that could be practically expected to be pursued and achieved in a conventional breeding program. To qualify for an exemption based on occurrence in the gene pool, the genetic change must be known to occur. We do not intend the exemption to apply to limitless possibilities that are theoretically possible but not currently known to occur in the gene pool. Consequently, the exemption in paragraph (b)(3) has been slightly modified for accuracy and clarity.

Some commenters asked that the exemption in paragraph (b)(3) be expanded to include plants in which an allele has been modified to align with a similar known allele found in a close relative, or in a more distant relative beyond the family level of taxonomy, or that we exempt plants containing any sequence from a plant that is known not to be a plant pest and is routinely used for food.

APHIS considers the known and accessible gene pool of a plant to include not only genetic sequences that can be introduced to a plant via crosses that can take place without human assistance, but also human-assisted wide crosses between more distantly related species. In systems for which breeding techniques such as bridging and embryo rescue have been developed to enable wide crosses, more distantly related plants are also considered part of the known gene pool. APHIS agrees in principle that exchange of genetic information between unrelated species is likely to be safe in most cases. However, APHIS does not have the experience to definitively state that exempting all exchange of DNA between plants will not lead to increased plant pest risk. In cases where genetic material from a more distantly related plant species is introduced into the plant, developers can request an RSR.

A commenter stated that their understanding is that the exemption in § 340.1(b)(3) would include any insertion or other sequence modification of less than 20 base pairs. APHIS disagrees and seeks to clarify that even an insertion or sequence modification smaller than 20 base pairs that does not otherwise qualify for exemptions § 340.1(b)(1) or (b)(2) still has to meet the criteria of paragraph (b)(3) to qualify for exemption under paragraph (b)(3). The exemption does not apply to what is theoretically possible. The genetic variation must be known to occur in the plant's gene pool in order to qualify for the exemption.

A commenter stated that the regulation could clarify that exemption under paragraph (b)(3) covers the introduction of natural or chemically synthesized copies of nucleic acid sequences from one plant species into the same or a crossable plant species, including (a) the targeted insertion or replacement of sequences exceeding 20 base pairs in length (

e.g.,

the insertion or replacement or a promoter, terminator, exon, intron, or small open reading frame, excluding complete genes), (b) the targeted replacement of a cisgenic allele (

i.e.,

perfect allelic replacement), (c) the targeted insertion of a cisgenic sequence at the same or a different location in the genome of the recipient species, and (d) the targeted insertion of a cisgene with a new combination of genetic elements, as plants containing such changes could have occurred naturally or could result from conventional breeding since they fall under exemption under paragraph (b)(3). A second commenter stated that some genetic engineering experiments will replace promoters, altering gene expression patterns in ways that are not attainable by today's breeders.

APHIS does not intend to modify the regulation text per the commenter's suggestion. Exemption under paragraph (b)(3) will exempt from regulation plants that have been modified to introduce a gene known to occur in the plant's gene pool, or that make changes in a targeted sequence to correspond to a known allele of such a gene or to a known structural variation present in the gene pool. Some of the examples provided by the first commenter may thus not be eligible for exemption under paragraph (b)(3). For instance, (b)(3) will not exempt from regulation a plant containing an insertion of a gene that is known to occur in the gene pool if the insertion results in the creation of a gene not known to occur in the gene pool,

e.g.,

a gene that results in the production of a protein or RNA, or a loss or gain of function, that is not known to be produced by plants within the gene pool. However, if a specific modification can be demonstrated to be present in the plant's gene pool, then it can be exempted under paragraph (b)(3). If a developer has a question about whether its plant is exempt from the regulation, the developer can contact APHIS for a consultation.

Some commenters asked how the deletion exemption in § 340.1(b)(1) pertains to diploid and polyploid plants. For example, if a deletion is made to both alleles of a diploid or all four or six alleles in tetraploid and hexaploid plants, respectively, would those plants qualify for the exemption?

APHIS seeks to clarify that exemptions in § 340.1(b)(1) through (3) apply to modifications made to one pair of homologous chromosomes. It is very straightforward in conventional breeding to identify a single allele in a diploid line and then convert the heterozygote to a homozygote in the next generation. However, it is very difficult through conventional breeding to create the same allele in all homoeologous genomes in polyploid plants. Therefore, for polyploid plants, the exemptions would initially apply only to modifications made to one pair of homologous chromosomes. As an example, consider a change to a gene in common wheat (bread wheat). Common wheat has three sets (AA BB DD) of homoeologous chromosomes. A developer can qualify for the exemption if modifying the A genome through a change that qualifies for exemption (b)(1), (b)(2), or (b)(3). If the developer wanted to make the same corresponding changes to the B and D genomes, the developer would go through the RSR process (as described below). Once APHIS determines that this A/B/D plant is unlikely to pose an increased plant pest risk, it will go on the list of plant-trait-MOAs that do not require regulation (

i.e.,

the § 340.1(c) exemption list). At that point, this developer, and any others, would be able to make the same plant-trait-MOA combination and be exempt from regulation under part 340.

Some commenters noted that the exemption in proposed § 340.1(b)(4),

i.e.,

the exemption of null segregants derived from GE plants, is superfluous

because the definition of

genetic engineering

applies only to organisms whose DNA sequence has been modified.

APHIS agrees with these commenters. According to our definition of

genetic engineering,

the genome of null segregants has not been created or modified. Therefore, null segregants do not need an exemption from regulation, and APHIS is removing this exemption from the final rule.

Some commenters stated that the exemption in proposed § 340.1(c) for a GE plant with a plant-trait-MOA combination that has previously undergone an analysis in accordance with § 340.4 and has been found by the Administrator to be unlikely to pose a plant pest risk should be eliminated. One commenter stated that the impact of releasing new GE plants into the environment cannot be accurately predicted or assessed without case-by-case analysis and controlled field experiments. Another commenter stated that every transformation event is unique, and thus potentially has a novel phenotype that must be assessed to determine appropriate regulation. The commenter further stated that the National Academy of Sciences (NAS) has also advocated the use of genetic engineering [

i.e.,

transformation] as “both a useful and scientifically justifiable regulatory trigger” because “there is no scientific basis” on which to exclude GE organisms from regulatory review prior to evaluation of data on the interactions between “trait, organism and environment.”

APHIS disagrees with these points. Based on the risk assessments we have performed in accordance with the petition process over 30 years, we have determined that, in many cases, we would have been able to evaluate the plant pest risks associated with a GE organism without field-test data. Rather, APHIS has discovered that the introduced trait of the GE organism provides the most reliable indicator of the organism's potential for deleterious effects on plants and plant products. These observations are expected and are consistent with the findings of reports of NAS (NRC, 1989; NAS, 2016). APHIS will seek additional information, potentially including data from controlled field experiments, in cases where APHIS identifies a plausible pathway to increased plant pest risk.

The same NAS study (NRC, 2002) cited by the commenter stated the following: “Transgenic organisms have potential environmental risks, but the committee expects that most of them will not produce significant actual environmental risks. Consequently, the committee also suggests that for environmental risk regulatory oversight should be designed to winnow the potentially riskier transgenic crops from the less risky ones before a substantial regulatory burden is imposed on the less risky ones.” APHIS has designed a system where organisms that pose a plausible plant pest risk are rapidly distinguished from those that do not, based on the RSR process described below under the subheading “Regulatory Status Review,” focusing regulation on the former. The exemption that we proposed in § 340.1(c) will apply only to those GE plants that have undergone a risk assessment in the RSR process. The revised regulations are proportionate to risk and are therefore consistent with the recommendation of NAS's study.

Several comments were received on the definition and application of the term MOA as it relates to the exemption in § 340.1(c). The issues raised by the commenters are discussed in detail below.

Two commenters stated that the categories of trait (defined in the June 2019 proposed rule as “an observable (able to be seen or otherwise identified) characteristic of an organism”) and MOA (defined as “the biochemical process(es) through which genetic material determines a trait”) could be interpreted so broadly that new GE plants that have a plant-trait-MOA combination similar to that of a nonregulated plant, yet contain unique features with unknown impacts on non-target organisms and the surrounding ecosystem, would not require review by APHIS. They stated that, for example, the “Cry

6

protein MOA” could include dozens of possibilities with unknown effects, and that it could even be the case that APHIS review would not be required when any gene encoding a Cry protein that targets broad orders of insect pests is inserted into a plant that had previously been engineered with any other trait and had been found by APHIS not to pose a plant pest risk.

6

A Cry protein is a crystalline protein toxic to certain species of insects primarily produced by the bacterium

Bacillus thuringiensis

(Bt). Genes for Cry proteins have been widely used to confer resistance to insect pests in several types of crop plants.

APHIS disagrees with the suggestion that the proposed definition of MOA is too broad. The suggestion is based on a misreading of the definitions and the preamble of the June 2019 proposed rule. As described in the preamble, the MOA refers to the specific manner by which the genetic modification confers the intended trait on the plant. We noted that the same trait can be obtained by different MOAs that would thus be subject to distinct RSRs. In the example cited, the preamble was clear that non-target impacts related to Cry proteins depend on whether the non-target insect has the correct receptor in its gut to bind the Cry protein; thus, for each new Cry protein it will be important to evaluate the potential for non-target impacts. Similarly, the preamble provided an example of RNA interference-based resistance, where it would be important to consider the specific target RNA and its corresponding protein in order to determine whether there could be non-target effects. Moreover, the regulatory text and preamble were clear that it is the specific plant-trait-MOA combination that is the subject of the RSR and decision. Developers could not qualify for exemption under § 340.1(c) by inserting any

cry

gene that encodes a protein targeting a broad order or orders of insects into a plant with any other trait and MOA that was previously reviewed by APHIS.

Another commenter stated that reasonably broad MOA categories should be established that would cover broad protein functional classes, account for all normal polymorphisms found in nature at the DNA and protein levels at the genus level, and account for the normal wide variation in expression seen among transgenic events and backgrounds. An additional commenter recommended that the definition of MOA refer to the biochemical process(es) through which the gene, rather than the genetic material, determines a trait, stating that it is a gene product and not the genetic material that determines the resulting biochemical process. Finally, a commenter requested that the final rule clarify which products would qualify for the exemption in § 340.1(c), noting that APHIS alternately used the terms “same” and “similar” to describe products that could qualify based on their use of a crop-trait-MOA combination that has already been assessed by APHIS and determined unlikely to pose a plant pest risk than the appropriate comparator(s).

APHIS agrees that in most cases, the MOA could cover all normal polymorphisms of a gene found in nature, even at levels broader than the genus. For example, the outcome of an RSR would apply to genetic material encoding an enzyme that catalyzes a specific biochemical reaction regardless of whether the genetic material is sourced from a plant or a microbe, as long as the enzyme catalyzes the same biochemical reaction regardless of the organism from which the genetic material encoding the enzyme is obtained, and does not catalyze any

additional biochemical reactions that differ among the source organisms. APHIS does not agree that the MOA would be so broad as to cover broad functional classes, since broad functional classes could encompass many different proteins that have multiple differences in the biochemical processes in which they participate. Typically, an RSR would be conducted at the level of the MOA of individual genes. If those genes when stacked produce a new phenotype, such as a new biochemical pathway, APHIS will consider the interaction of the gene products in the RSR. Regarding variation in expression, in most cases APHIS anticipates that variation in expression should not affect the outcome of an RSR. However, as we noted in the preamble to the June 2019 proposed rule, there may be cases where it is important to consider where, when, or at what level the genetic material is expressed in the plant. In those cases, APHIS will specify whether and in what way variation in expression limits the outcome of the review.

APHIS will not revise the definition of MOA in response to these additional comments, because some MOAs may not involve changes in gene products but rather changes in genetic material that affect the expression of gene products. As this discussion makes clear, a plant-trait-MOA combination may qualify for the exemption only if the combination is the same as a previously reviewed plant-trait-MOA combination that has been found to be unlikely to pose a plant pest risk. To be clear, a merely “similar” combination does not qualify as a “same” combination, but a “similar” product may qualify for the exemption if it has the same combination as a previously reviewed combination.

One commenter urged that in addition to mutated products of genome editing, the concept of exemptions due to familiarity should be broadened to include plants with transgenic traits that are familiar in type and inherently unlikely to give a significant advantage to wild plants. Examples would be sterility traits, stature reduction traits, and quality traits relevant to industrial processing (

e.g.,

modified lignin in alfalfa and trees). According to the commenter, another class of strong candidates for plant kingdom-wide exemption are the widely used marker genes, such as

nptII

for kanamycin resistance, T-DNA borders, and widely used promoters such as 35S and NOS.

APHIS appreciates these comments. The commenter did not provide any scientific evidence or explanation that would make the comments actionable at this time, however.

Several commenters asked that APHIS clarify the regulation of plants containing stacked traits. One commenter requested that APHIS codify in the regulations that plants developed through conventional breeding that are derived from products determined to not be regulated (either because of an exemption or as a result of an RSR) would themselves be unlikely to pose increased plant pest risk and therefore would not be subject to regulation. Other commenters argued that APHIS should assess the risks of stacked traits, particularly plants containing multiple herbicide resistance traits, using the noxious weed authority.

A discussion of our noxious weed authority in the context of these regulations is presented later in this document.

APHIS notes that in accordance with § 340.1(c), the regulations under part 340 do not apply to a GE plant with a plant-trait-MOA combination that has previously undergone an analysis in accordance with § 340.4 and is not subject to the regulations. APHIS notes that the word “combination” used in the regulation text is deliberately enumerated as singular and not plural in order to denote that the exemption applies to a single plant-trait-MOA combination and not a molecular stack of multiple plant-trait-MOA combinations. Plant-trait-MOA combinations that have undergone an analysis in accordance with § 340.4 and are not subject to the regulations may be stacked by conventional breeding methods and would still qualify for the exemption. However, this is not the case for plant-trait-MOAs stacked molecularly; today stacked traits typically have independent MOAs. In the future, we anticipate seeing more interactions between or among the products of genes in molecular stacks, potentially including new MOAs that were not evident in the review of individual traits. For this reason, APHIS anticipates that plants that are the genetically engineered product of more than one previously evaluated combination will be subject to evaluation under § 340.4. In cases where there is no interaction between trait-MOA combinations, we expect to be able to use the results of previous reviews to quickly reach a regulatory status determination.

Finally, several commenters requested clarity on the regulatory status of plant-trait-MOA combinations that were previously deregulated under part 340 or deemed to be not regulated under the “Am I Regulated” (AIR) process.

To provide the clarity the commenters requested, we are amending paragraph (c) to exempt from these regulations a GE plant that has a plant-trait-MOA combination contained in a GE plant determined by APHIS to be deregulated under a petition submitted prior to October 1, 2021 pursuant to § 340.6 of the current regulations in part 340. We are also adding a new paragraph (d) to § 340.1, stating that all GE plants determined not to require regulation pursuant to the AIR process will retain their nonregulated status under these regulations.

As we have noted, APHIS will publish a list (referred to earlier in this document as the § 340.1(c) exemption list) of plant-trait-MOA combinations that have been evaluated under our new RSR process and found not to require regulation under part 340. That list may be used by a developer to determine whether its novel GE plant would qualify for exemption under § 340.1(c). GE plants previously evaluated under the petition process will be included on the § 340.1(c) exemption list because such plants will have effectively been evaluated at the MOA level and determined not to pose a plant pest risk.

Plants that have been determined not to require regulation pursuant to the previous AIR process will not be included on the § 340.1(c) exemption list because they will not have been evaluated at the MOA level or by analogous criteria. Such plants will be identified at a separate list, at

https://www.aphis.usda.gov/aphis/ourfocus/biotechnology.

Because the plants to be identified on this separate AIR list were not evaluated under the petition process or under the RSR process, developers will not be able to use the AIR list in determining whether new GE plants they develop should be subject to or exempt from the regulations. At the same time, we have multiple reasons for concluding that the specific plants on the AIR list should retain their nonregulated status under these regulations. Not only do we lack a basis for overturning our prior individualized determinations reached pursuant to the AIR process, we also believe that it is appropriate for us to take into account the importance of preventing potential market disruptions, including potential trade disruptions, and providing regulatory certainty for developers, third parties, and the general public.

Self Determination

Under the June 2019 proposed rule, developers would have the option to determine whether their plants belong to one of the categories listed under § 340.1(b) or (c) and are therefore

exempt from the regulations. As stated in the preamble to that proposed rule, allowing for such “self-determinations” would provide developers with regulatory relief and would open more efficient and predictable pathways for innovators to get new modified plants that do not require regulation to market, in turn supporting further innovation. Eliminating the need for redundant evaluations of products would allow APHIS to devote more attention to assessing and regulating GE organisms that are likely to be associated with potential plant pest risks.

While many commenters agreed with the rationale discussed above and welcomed the regulatory relief that allowing for developer “self-determination” would provide, others either opposed the concept entirely or expressed reservations. Many in the latter category cited what they believed to be potential risks that could result from allowing developers to determine whether their products are eligible for exemption from the regulations. Some industry commenters questioned whether allowing developers to make such determinations would actually relieve regulatory burden and incentivize innovation to the extent that we anticipated. The comments are discussed in detail in the paragraphs that follow.

Many commenters opposed “self-determination” on the ground that allowing developers to regulate themselves could result in conflicts of interest. It was stated that developers of GE products with a financial stake in the outcome should not be allowed to determine which products should be subject to regulatory review. According to these commenters, such an approach would fatally undermine the integrity, rigor, and credibility of what must be an independent regulatory process, weakening Agency ability to protect the public interest, and furthering mistrust in the U.S. Federal regulatory system in the public's eye and among key trading partners. By avoiding the RSR or permitting process, these commenters believed, the developer could get its new product to market without its ever having undergone an objective, third-party review. In allowing developers to determine whether their products are eligible for exemption, according to these commenters, we are effectively abdicating our regulatory authority and not carrying out our mission to protect U.S. agriculture.

We do not agree with these comments. The revised regulations in part 340 recognize that plant products that are the result of modifications that coincide with conventional plant breeding do not pose additional plant pest risk and should not be regulated under these regulations. Products that do not fall within the regulatory scope of part 340 have not been subject to compulsory regulation in the past, and developers have always been able to act accordingly to determine whether their products are subject to the regulations.

It was further argued that allowing developers to determine the regulatory status of their products will result in less transparency and greater risk of commingling with organic and other non-GE crops and will damage consumer confidence. Allowing developers to determine the regulatory status of their products, it was claimed, will result in an overall loss of transparency in that the public would not have access to the data used by developers to make their determinations. Organic farmers would have less information about modified crops grown near their fields than they do now, because the information that informed developers' determinations would remain proprietary, and their ability to take preventive measures would be hindered. Some commenters cited the recent finding in Washington of unapproved GE glyphosate-resistant wheat

7

as an example of risks posed by allowing developers to determine whether their products are eligible for exemption and by reducing our regulatory oversight over GE products more broadly.

7

On June 7, 2019, APHIS confirmed the discovery of GE wheat plants growing in an unplanted agricultural field in Washington State. The GE wheat in question was resistant to glyphosate, commonly referred to as Round Up. On July 12, 2019, APHIS announced that the GE wheat plants in question were developed by Monsanto (now owned by Bayer CropScience (BCS)) and referred to as MON 71300 and MON 71800. APHIS also announced that there is no evidence that any GE wheat entered commerce or is in the food supply.

https://www.aphis.usda.gov/aphis/ourfocus/biotechnology/brs-news-andinformation/2019_brs_news/wheat_update_jul2019.

We do not agree with these comments. With regard to transparency, we anticipate that many developers whose products fall within an exemption will request confirmation letters because the letters will help them market their products domestically and overseas. Those letters will be posted on the APHIS website and will be available to the general public, including organic and other growers of non-GE crops. Information from previous RSRs will also be available to the public. We do not agree that self-determinations will limit organic growers from learning whether their neighbors are growing GE crops. This information principally comes from conversation with neighbors and from other voluntary interactions and arrangements, and is not based on USDA decisions on regulatory status. We also do not agree that the finding of GE wheat in Washington fields is relevant to the regulatory changes made in this final rule. Under the new regulations set forth in this final rule, the GE wheat involved in the incident would not be eligible for an exemption and would need to go through the RSR process. The commenters are generally confusing a fact-specific compliance issue, which could arise under any number of regulatory schemes, with broader questions about the appropriate regulatory approach. If APHIS were to find that a plant was unlikely to pose an increased plant pest risk, APHIS would make information publicly available regarding the plant, trait, and a general description of the MOA. In cases where GE crops are not subject to regulation because they are unlikely to pose a plant pest risk, no other risks are regulated by APHIS insofar as they are outside the scope of the regulations.

In the preamble to the June 2019 proposed rule, we stated that a developer who made a determination of regulatory status that APHIS found not to be valid would be subject to remedial measures or penalties in accordance with the compliance and enforcement provisions contained in § 340.6 of the June 2019 proposed rule.

Some commenters stated that there is a need for a plan for detection and enforcement in cases where developers incorrectly determine their products to be non-regulated, or where changes in evidence may call a developer's determination into question. Without a record of what plants are being released, according to these commenters, it will be impossible to conduct any kind of periodic surveillance or audit to ensure compliance. These commenters believe that this difficulty may be partly addressed by having a compulsory reporting mechanism whereby a responsible party fills out a form to declare its modification and assert its exempt status. This would create a searchable record. According to such commenters, a database compiled from self-reported data would not offer complete protection against bad actors, but when combined with penalties that are proportional to the degree of harm done by a developer incorrectly making a determination, such a database may aid in correcting incorrect determinations by developers.

APHIS disagrees with the proposal for a mandatory process and the data base proposals associated with it and has instead included provisions in part 340 for a voluntary confirmation process for

products exempted from the regulation. Voluntary confirmation will be public information, however, and interested parties could search for it of their own volition.

Under APHIS's long-standing regulations, APHIS regulates articles based only upon a narrow and limited plant pest mechanism. The products that commenters are concerned will be “missed” or “overlooked” in the “future” have no current regulatory trigger. Under this rule, APHIS' focus will be on plant pest risk associated with the product, consistent with our legal authority. Consistent with long-standing practices, we will continue to offer voluntary confirmation of regulatory status to those who seek it. APHIS agrees with comments expressing concern that a mandatory process may trigger confusion among both consumers and the international trading partners, by unnecessarily hindering global acceptance of products of biotechnology. That said, if the market demands confirmation of regulatory status, APHIS has created a mechanism for developers to request such confirmation, and for us to provide it.

APHIS also notes that a large number of commenters supported the kind of voluntary confirmation process contained in this final rule for regulatory exemptions, noting public access to the confirmation letters. Those comments noted that a voluntary process would provide domestic and international transparency, be beneficial for marketing of new products, support deregulation processes in other countries, facilitate exports, facilitate the development of new genome edited plant varieties, encourage the continued domestic and global adoption of new traits, and enhance harmonization of global trait approvals.

If a plant pest issue arises from a plant that is exempt from these regulations, APHIS has mechanisms to address such risks subsequently and has a wealth of experience in dealing with such instances. As under the current regulations, a developer could knowingly or unknowingly violate APHIS regulations by transporting, importing, or releasing into the environment a regulated plant without APHIS authorization. The PPA contains authority for the Administrator of APHIS at any point to place such articles under regulation. If a determination made by a developer should be found to be invalid, however, APHIS does have the authority to enforce sanctions. As noted in the preamble to the June 2019 proposed rule, pursuant to sections 7714 and 7731 of the PPA, APHIS may seize, quarantine, treat, destroy, or apply other remedial measures to an organism covered under the regulations that is new to or not widely prevalent or distributed in the United States to prevent dissemination of the organism. Enforcement provisions are also included in § 340.6 of this rule. APHIS has many years of experience in initiating and coordinating enforcement action as appropriate, in cases where compliance issues exist.

Even in cases where we would impose penalties for invalid determinations by developers, some commenters expressed skepticism that those penalties would be efficacious in remediating harm or preventing further harm. In the view of these commenters, if the movement or release of a GE product that had already reached the market based on a faulty determination by a developer resulted in commingling with other crops or the dissemination of plant pests, whatever penalties or remedial actions APHIS would impose would likely neither prove adequate to address injuries to innocent parties nor provide sufficient disincentives to discourage bad actors from making invalid determinations. Elaborating on the latter point, one commenter stated that penalties imposed by APHIS after the fact may not even be legally defensible if we have allowed a developer to determine whether its product is eligible for exemption. Another commenter stated that APHIS, lacking a post-commercialization monitoring program, has little capacity to recall the products of invalid determinations by developers.

We do not agree with these comments. In the event that APHIS discovers that a developer makes an invalid determination, the specific penalties and/or remedial action will be applied case by case, as appropriate. Similarly, whether the discovery of an invalid determination is too late will also be decided on a case-by-case basis. In regard to legal defensibility, the PPA provides ample flexibility and broad civil penalty authority to deter violations of the PPA. For example, the PPA provides statutory maximum penalties of $1,000,000 per violation for any person who willfully violates the PPA.

Other commenters feared that the penalties could be excessive. It was stated that any such penalty applied to a developer must be based on a demonstration of significant economic harm to another entity from the error, and not on technical or minor errors in interpretation. The commenters further stated that in such situations, the penalties must be proportional to that harm.

We agree that penalties must be proportional to the severity of violations and the harms that may result from them, and we will enforce the regulations accordingly. Furthermore, the harms must fall within the harms considered under the PPA. Congress has outlined the factors for consideration in assessing penalties under the PPA. These factors include “the nature, circumstance, extent, and gravity of the violation or violations,” as well as the violator's ability to pay, the effect of the penalties on the violator's ability to continue to do business, and any history of prior violations. (See 7 U.S.C. 7734.)

In the preamble to the June 2019 proposed rule, we stated that one of the benefits of “self-determination” is that it would enable APHIS to focus its regulatory resources and risk analyses on unfamiliar products and thereby to avoid conducting repetitive analyses on GE products that are very similar to those that we have already evaluated for regulatory status. APHIS would thus be able to utilize its staff time more efficiently, and provide better stewardship of taxpayer dollars than it could under the existing regulations.

One commenter viewed allowing developer-made determinations as evading APHIS' regulatory responsibilities rather than enabling APHIS to use its resources more efficiently. The commenter stated that if GE developers are concerned about delays in getting their products to market because, in their view, APHIS does not have sufficient resources to conduct all reviews in a timely manner, then those developers should lobby Congress to provide more funding to enable APHIS to perform its duties in a more timely manner, as opposed to having APHIS reduce its oversight role.

APHIS disagrees with this comment. The plants that qualify for exemption under part 340 fall into three categories: (1) Those that could otherwise have been developed through conventional breeding methods and have a history of safe use related to plant pest risk that does not require regulation (§ 340.1(b)(1) through (3)); (2) those that have the same plant-trait-MOA combination as other plants that have already been evaluated by APHIS and have been found to be not subject to the regulations (§ 340.1(c)); or (3) those determined to be not subject to the regulations under the AIR process. It should be noted that plants that qualify for exemption under § 340.1(c) are very similar to plants that have been evaluated previously by APHIS. APHIS can utilize its resources most efficiently

by evaluating GE plants that do not fall into these categories and therefore may pose a level of plant pest risk that requires regulation.

Many other commenters expressed skepticism from an opposing perspective about the efficacy of allowing developers to determine whether their products are eligible for exemption. These commenters doubted that such “self-determination” would provide the regulatory relief that we claimed in the preamble to the June 2019 proposed rule. One reason given was that most developers would seek certification or confirmation from APHIS that their determinations were valid, given the possible liabilities associated with making incorrect determinations. Such certification would therefore become a

de facto

requirement. One commenter expressed the concern that in order to receive such confirmation, developers would need to provide the information described in proposed § 340.4, which contains information requirements for RSRs. It was further suggested that while academics, startups, and small developers could see some benefit from “self-determination,” companies with existing portfolios of GE crops will be in a better position to benefit.

We do not agree with these comments. If innovators choose to forgo the regulatory relief provisions offered by our revision of the regulations in part 340 for any reason, they are welcome to do so. In this final rule, APHIS focuses on plant protection, while also easing regulatory burdens. Accordingly, we also aim to be responsive to repeated concerns raised by small businesses, academic-based researchers, and other innovators who have reported past difficulty successfully seeing products through to commercialization. The approach APHIS has taken is fully consistent with the priorities and direction provided by Executive Order 13874, which we have discussed earlier.

In § 340.1(d)

8

of the June 2019 proposed rule, we indicated that developers may request confirmation from APHIS that the plant is not within the scope of the regulations in part 340. A developer may find a confirmation letter useful in marketing its products domestically or overseas because the letter would serve as verification to an importing country or other party that APHIS concurs with the developer's determination. Confirmation is not required, however, and for developers not seeking confirmation letters, no submission of information to APHIS is required, nor is any response from APHIS. Guidelines for the information that would need to be submitted to enable APHIS to respond to a request for confirmation are discussed below under this same subheading of comment responses.

8

Due to the addition of a new paragraph (d) in § 340.1, as described earlier, provisions related to confirmation letters are contained in § 340.1(e) of this final rule.

Some commenters expressed doubt that developers would even be able to employ the “self-determination” option due to what they perceived as a lack of clarity surrounding it. It was stated that decisions on a product's regulatory status would be based on APHIS' assessment of plant pest risk, but that because APHIS would define plant pest risk and because APHIS did not provide a list of traits for identification of a plant pest in the proposed rule, a developer would lack the guidance to make a determination safely.

APHIS disagrees with this comment. This rule clearly outlines the kinds of information needed to successfully navigate the APHIS regulatory system, as well as the protection goals and criteria that APHIS will consider as part of this process. Plants that meet the exemptions listed under § 340.1 will not require regulatory oversight under the regulations in part 340. The exemptions in § 340.1(b) are based not on the trait, but on whether the plant could have otherwise been produced through conventional plant breeding techniques. The exemption in § 340.1(c) is based on whether the plant-trait-MOA combination is the same as one that APHIS has previously determined to be nonregulated. APHIS will publish a list of such combinations, which developers may use in determining whether their GE plants qualify for exemption under § 340.1(c). As more GE plants undergo RSRs to determine their regulatory status, that list will grow. A list of traits for identification of a plant pest is not needed in order for developers to determine whether their products meet one of these exemptions in § 340.1(b) or (c).

Several commenters recommended that we provide more certainty about the process by issuing guidance documents to aid developers in making their determinations. Such documents, it was stated, could include, among other things, information requirements and timelines, including timelines for APHIS responses to requests for confirmation. Many commenters stated that, in general, defined timeframes for APHIS regulatory actions are important to improve predictability and to support the planning needed to conduct seasonally based field research, and therefore should be included in the regulations. Most commenters who provided specific timeframes for confirmation requests suggested that APHIS should respond to such requests within 60 days. It was further suggested that to provide developers with additional guidance for making determinations, APHIS should maintain a database of products that have undergone RSRs and been found not to be subject to the regulations.

APHIS has had a longstanding practice of providing guidance to aid the regulated community in complying with the regulations. APHIS will provide guidance to developers regarding the confirmation process. We will also maintain on our website requests for and results of RSRs. That information will aid developers in making their determinations.

Regarding timeframes, in the preamble to the proposed rule, APHIS noted that we anticipate a timely turnaround time in providing confirmation letters. APHIS agrees that providing a more specific timeframe for responses to confirmation requests would improve predictability. Based on our experience with the current AIR process, which is functionally similar to the confirmation process, APHIS has amended § 340.1(e) by adding a sentence indicating that, except in unforeseen circumstances, written responses will be provided within 120 days of receiving a confirmation request containing sufficient detail to determine whether the plant meets one of the exemptions in § 340.1.

One commenter stated that the type of information provided to APHIS by developers should be a description of the crop and the justification for meeting the exclusion, which would be similar to the information submitted for the “Am I Regulated” Process.

APHIS agrees with the sentiment expressed in this comment and is therefore setting out guidelines for parties requesting confirmations to submit to APHIS in support of their requests. The guidelines are listed below and will also be posted on the APHIS website at

https://www.aphis.usda.gov/aphis/ourfocus/biotechnology.

In addition, developers who have specific concerns may consult with APHIS.

In communications with APHIS requesting confirmation of exemption from the regulations, requestors will be expected to submit the following:

1. A description of the plant, trait(s), and modification(s).

2. A clear statement of which regulatory exemption the biotechnology developer is claiming for the plant and

why the plant qualifies for that exemption.

3. Details about the scientific method used to validate that the plant met the exemption criterion.

APHIS expects that the description of the plant will include both the scientific and common names. The trait information should include a description of the intended and any observed phenotype(s) of the plant. Details about the modification(s) must provide APHIS with a clear understanding of the genetic change in the plant. In the case of § 340.1(c) exemptions, requestors must submit the MOA.

Many commenters advocated that we establish a mandatory process for developers to notify APHIS of their determinations and for APHIS to issue confirmations. (We would note here, however, that there was considerable divergence of opinion on this issue, with 25 commenters expressing support for maintaining a voluntary confirmation process.) Some commenters requested that confirmation be mandatory for all determinations made by developers, while others stated that confirmation should be mandatory only for developer-made determinations of products that will be commercialized. Many requested that the process be streamlined and include information and self-reporting requirements and timelines. It was recommended by some commenters that developers be required to provide notice to APHIS 90 days before putting a product on the market.

We will not be making any changes to this final rule in response to these comments. The confirmation process laid out in the June 2019 proposed rule was voluntary, and switching over to a “mandatory” confirmation and/or notification process in this final rule would run counter to the spirit of regulatory relief underlying our new regulatory framework. A voluntary confirmation process allows the market to drive the demand for new plants, avoids codifying a process that may grow antiquated as technology develops, provides developers with a method to obtain confirmation that their products are in fact exempt from the regulations, and avoids differential treatment for genome-edited products that are otherwise equivalent to conventionally bred and/or developed products.

Commenters did not persuasively explain how developers of products that are not subject to the regulations could be compelled to comply with a requirement for mandatory participation in a confirmation process. APHIS notes that even if the commenters had provided a sufficient regulatory mechanism to impose such a requirement, a mandatory process would likely trigger the emergence of trade concerns, as products that are scientifically justified to be exempt would also appear on lists of GE organisms-essentially creating a third category of products that are required to be listed but are otherwise exempt from regulation (in addition to two other categories: (1) Organisms that were subject to RSR and determined not to be regulated by APHIS, and (2) regulated organisms). APHIS further notes that a mandatory process would likely disadvantage the very small-scale, mid-size, and university researchers and innovators that the rule was intended to aid. Lastly, APHIS notes that the proposal for a mandatory confirmation provides no added benefit in plant protection.

Some of the commenters who favored a formal or mandatory confirmation process did so because they questioned the utility of a voluntary process. It was stated that an APHIS confirmation that a determination made by a developer is valid, as provided for in the June 2019 proposed rule, will be a formulaic letter without an accompanying risk assessment. Some trading partners may not view such confirmation letters as sufficient to meet their own requirements for admission of U.S. GE products. It was stated that to keep export markets running smoothly, industry needs an official U.S. attestation that the new traits do not pose a plant pest risk.

We do not agree with these comments. The confirmation letters will state that the product in question meets a regulatory exemption or has a plant-trait-MOA combination that has already been reviewed by APHIS. APHIS currently works with, and is committed to continuing to work with, international trading partners and exporters to resolve trade concerns. International trade issues are discussed in greater detail later in this document.

Some commenters addressed the issue of whether, or how much, information pertaining to determinations made by developers and APHIS confirmations should be made public. Some commenters, citing the need for transparency and certainty, recommended that we post confirmation inquiries and confirmation letters on our website. Others, however, thought that such information should be treated as confidential business information (CBI) and therefore not be made publicly available. One commenter suggested that we use a process similar to that of the existing “Am I Regulated” process, under which CBI exemptions could be claimed in the request for confirmation submitted to APHIS, and a non-CBI version of the submission could be made publicly available.

In the interest of transparency, APHIS will post the confirmation letters online. APHIS notes, however, that confirmation letters are subject to claims of CBI and will proceed in implementation of such posting in accordance with all applicable laws and procedures. In accordance with USDA regulations, 7 CFR 1.8(a) through (c), a submitter of confidential commercial information must use good-faith efforts to designate, at the time of submission, any portion of its submission that it considers to be protected from disclosure under Exemption 4 of the Freedom of Information Act (FOIA) (5 U.S.C. 552). When making discretionary releases of records, as is the case with the posting of the confirmation letters online, APHIS follows the FOIA, USDA, and APHIS implementing regulations (7 CFR subpart A and 7 CFR 370.5, respectively), and guidance from the U.S. Department of Justice's Office of Information Policy relating to the handling of confidential business information.

Finally, there were a few comments on proposed § 340.1 that did not fall into any of the categories discussed above.

One commenter suggested that the exemptions should focus on plant species, not variety, as well as the purpose and type of application of genome editing. The commenter stated that genome editing can be used both to produce or improve on a specific characteristic or phenotype, such as by silencing a disease sensitive gene, and to improve existing breeding processes themselves, such as by using gene editing to more efficiently induce double haploids.

The “purpose and type of application of genome editing” is just another way of describing the plant-trait-MOA combination. In the example given above where genome editing is used to improve an existing breeding process by more efficiently inducing double haploids, genomic modifications will be made to a specific plant, with a specific trait, having a specific MOA. Recently a widely used haploid inducer in corn was identified to be a defective allele (

matL

) of the gene named Matrilineal (Kelleher, 2017). A haploid induction trait was shown to work in rice by genome editing the

matL

allele (Yao, 2018). APHIS considers this new process to be an example of a plant (rice), trait (haploid induction), MOA (defective pollen specific phospholipase) combination. Upon

completion of an RSR for this plant trait MOA combination, the § 340.1(c) exemption would apply to all varieties of rice, not just the variety it was introduced into.

Another commenter thought that there was a possible conflict between §§ 340.1(c) and 340.2(a). The latter paragraph of the proposed rule stated that a plant with a plant-trait-MOA combination that has not been evaluated by APHIS for regulatory status in accordance with § 340.4 would have to move under permit. According to the commenter, the conflict arises because products we would allow to move without permits based on developers' determinations would not have been evaluated by APHIS.

We do not see such a conflict. When a developer determines that a GE plant falls under § 340.1(c), it is not subject to the regulations in part 340 and therefore does not require a permit for movement. We are making an editorial change to § 340.2(a), however, to clarify that a GE plant will be subject to the regulations: (1) If it has not undergone an RSR in accordance with § 340.4; or (2) if it has undergone an RSR and, as a result of the evaluation, is subject to the regulations. Such GE plants will require permits for movement.

One commenter stated that by allowing developers to determine whether their products are eligible for exemption, we would not be in compliance with the requirement of the Cartagena Protocol on Biosafety that countries list all GE organisms released into the environment in the Biosafety Clearing House.

APHIS notes this comment, and wishes to clarify that the United States is not a signatory to the Cartagena Protocol on Biosafety. APHIS also notes that Article 3 of the Cartagena Protocol on Biosafety does not reference “GE organisms.” Instead, Article 3 (g) states that “living modified organism means any living organism that possesses a novel combination of genetic material obtained through the use of modern biotechnology.” Many international efforts are underway to align regulatory approaches and to seek compatibility for emerging technologies that were not in existence when existing policies were developed.

Two commenters requested that APHIS develop and issue guidance for developers of non-plant GE organisms to give them an opportunity to determine for themselves whether their products are subject to the regulations and to apply to APHIS for confirmation of regulatory status.

APHIS does not agree that such a new process needs to be developed. Currently, the Agency responds to the developers' questions about whether a specific GE organism, including a non-plant organism, is subject to the regulations. APHIS will continue that practice after this final rule becomes effective.

Scope of the Regulations

Section 340.2 of the June 2019 proposed rule delineated the scope of the regulations. We proposed to regulate,

i.e.,

require a permit for the movement of, any GE organism that:

1. Is a plant that has a plant-trait-MOA combination that has not been subject to RSR; or

2. Meets our proposed definition of a

plant pest;

or

3. Is not a plant but has received deoxyribonucleic acid (DNA) from a plant pest, and the DNA from the donor organism either is capable of producing an infectious agent that causes plant disease or encodes a compound that is capable of causing plant disease; or

4. Is a microorganism used to control plant pests or an invertebrate predator or parasite (parasitoid) used to control invertebrate plant pests and could pose a plant pest risk.

As was the case with the proposed exemptions, commenters expressed a wide range of views regarding the scope of the proposed regulations. While some supported our overall approach, others expressed the view that the proposed rule would either narrow or broaden our regulatory oversight excessively.

Some commenters who favored a broader scope stated that a regulatory approach that provides for regulations of only those GE organisms that are plant pests or pose a plant pest risk is too narrow. Such an approach, it was stated, isolates the GE organism from the environment in which it is used and the process by which it is developed, thereby impeding science-based risk assessment. According to these commenters, other hazards potentially associated with GE organisms and not accounted for in the June 2019 proposed rule need to be addressed. Some concepts discussed in these submissions included the increased potential for commingling with non-GE crops; the potential for contributing to the creation of herbicide-resistant weeds; pesticide overuse; habitat destruction; reductions in insect populations; and increased herbicide use, which, according to the commenters, has been associated with GE crops and may have additional deleterious effects on the environment and on human health.

While we recognize commenters' interests in addressing these concerns, many of these comments are outside the scope of this rulemaking and APHIS's statutory authority under the PPA. Commingling between GE and non-GE crops is generally a market issue unrelated to plant pest risk. Herbicide use is regulated by EPA, not USDA, so is not within the scope of this regulation. The basis for the commenter's claim that GE crops result in habitat destruction is not clear; however, we note that APHIS does not regulate farming practices. USDA's National Resources Conservation Service does have incentive programs to promote more sustainable farming. The current rule includes an RSR process that considers, as appropriate, impacts (if any) of a GE crop on populations of beneficial insects and other non-target organisms beneficial to agriculture.

Some commenters questioned the scientific justifications for the above listed categories of GE organisms that would fall under the regulations. It was stated that APHIS needs to re-cast its entire proposal and frame it around the identification of the characteristics of the organism or phenotypes of concern for which a plausible case can be made, based not on speculation but data and experience, that they present an unreasonable risk to American agriculture. It was further argued that there is no scientific justification for regulating by plant-trait-MOA instead of phenotype associated with the trait.

In order for the regulations under part 340 to enable future innovation while simultaneously protecting American agriculture from potential risks to plant health, it is vital that the regulations be prospective rather than retrospective, while being appropriately tailored to risk. A regulation that enumerated specific phenotypes that APHIS is concerned with would not only be impractical, since a phenotype may be of concern in one plant species but not in another (including depending on whether the plant has sexually compatible relatives, an attribute important for considering the distribution of a phenotype introduced into a plant), but would become immediately obsolete upon issuance. As articulated clearly in numerous studies, including those by the National Academy of Sciences, no entity has the foresight to identify only those phenotypes that present concerns decades into the future. Moreover, the MOA utilized by the developer matters when determining if there is a plant pest risk. The same intended phenotype can result from multiple different MOAs, but each MOA may differ in other phenotypes and thus may differ in their ability to present a plant pest risk and

in the types of plant pest risk they may present.

APHIS thus does not consider the approach of regulating solely by phenotype to be feasible. Instead, APHIS has articulated a regulatory approach that is adaptable to future innovation and continues to protect against risk, even in cases where it is not possible to envision the kinds of products being developed in the future. In particular, we have developed the RSR process in order to determine, based on scientific knowledge and information, if a GE plant contains a plant-trait-MOA combination that could plausibly present an increased plant pest risk than the appropriate comparator plant(s). We will regulate a GE plant only when we identify and are unable to rule out a plausible pathway to increased plant pest risk. In this way, when sufficient data and experience are lacking to rule out a plausible risk identified by APHIS, we have a mechanism to acquire more information to test the specific plausible risk hypothesis before decision making.

The risk-based system APHIS has developed in part 340 appropriately provides entrance for genetically engineered organisms into the regulatory framework and provides appropriate off-ramps from regulation for those products that do not pose plant pest risks. Conversely, a narrowly focused characterization of an intended phenotype, regardless of the plant species or MOA by which the phenotype is conferred, would not provide a sound scientific basis for an entire regulatory program. Many commenters expressed support for our scientific and risk-based regulatory process that evaluates plants based on their plant-trait-MOA combination.

A commenter stated that the restriction in § 340.2(c) covering a non-plant GE organism that has received DNA from a plant pest is unclear and lacking in scientific justification. The commenter questioned whether receiving DNA from a plant pest would likely make the recipient into a plant pest.

The commenter misconstrues § 340.2(c), which states that non-plant GE organisms that receive DNA from a plant pest will be regulated if that DNA is capable of producing an infectious agent that causes plant disease or if the DNA encodes a compound that is capable of causing plant disease. Such non-plant GE organisms could pose a plant pest risk, justifying their regulation under part 340.

Some commenters stated that organisms and microorganisms used to control plant pests should not require regulation if they are not plant pests themselves or do not pose a plant pest risk. One commenter stated that there appears to be a conflict between § 340.2(d) and EPA's regulatory authority under the Federal Insecticide, Fungicide, and Rodenticide Act for microbial pesticides. The commenter further stated that the intent of the PPA for biological control organisms is to facilitate their development, but that APHIS is proposing to require additional regulatory requirements without indicating a need for these extra requirements in terms of protecting against plant pests.

We agree with the first comment (

i.e.,

that organisms and microorganisms used to control plant pests should not require regulation if they are not plant pests themselves or do not pose a plant pest risk), and this rulemaking does not provide for the regulation of biological control organisms if they are not plant pests themselves or do not pose a plant pest risk. As we noted in the preamble to the June 2019 proposed rule, “GE non-plant organisms that do not pose a plant pest risk would not fall under the scope of the regulations and therefore would not require permits for movement.” We disagree with the remaining comments. As we noted in the preamble to the proposed rule, while biological control organisms are generally not plant pests, some biological control organisms could be plant pests because their potential effects on organisms beneficial to agriculture could indirectly affect plant health. The PPA provides the authority to regulate such biological control organisms used to control plant pests to ensure that they do not pose a plant pest risk. As with non-GE biological control organisms, the types of GE biological control organisms that APHIS would regulate include organisms that could pose a plant pest risk by lacking sufficient specificity for the target pest and thereby harming beneficial non-target organisms, such as other invertebrate predators or parasites (parasitoids), pollinators, or microbes that promote plant health. Because biological control organisms are almost always intended for eventual release into the environment, it is not sufficient for us to consider only their use in controlling their target plant pest. We must also take into consideration the indirect plant pest risks that the organism may pose due to harmful impacts on non-target organisms that are beneficial to agriculture (

e.g.,

harm to natural enemies of plant pests). If the GE organism is known to have harmful impacts on beneficial non-target organisms, it is consistent with APHIS' authority under the PPA to prohibit or restrict its release. To the extent that we do not know whether a GE biological control organism is sufficiently specific to avoid harming beneficial non-target organisms, it is also prudent for us to place regulatory controls on the movement and release of the GE biological control organism until the impacts on beneficial non-target organisms and any resulting direct or indirect plant pest effects are better understood. In addition, we will exempt biological control organism-containing microbial pesticide products that are currently registered with EPA as microbial pesticide products that are not plant pests.

Definitions

In this final rule, we have revised the definition of

article

to provide greater clarity. The definition in the June 2019 proposed rule was drawn from that provided in the PPA. However, while the PPA indicates that an article may be an object that could harbor noxious weeds, upon review of the provisions of the proposed rule, we have determined that it is not appropriate to consider such an object an article under these revised part 340 regulations. The proposed definition could have been interpreted to suggest that APHIS intends to regulate GE organisms, and require permits for their movement, under the revised regulations based solely on their noxious weed potential. As discussed elsewhere in this document, however, this is inconsistent with APHIS' intent. The revised definition reads as follows: “[a]ny material or tangible object that could harbor plant pests.”

A commenter stated that we need to define

environment,

because movement under permit includes release into the environment.

Environment

was defined in the proposed rule, however, and we are retaining that definition in this final rule.

In the June 2019 proposed rule, we defined environment as “[a]ll the land, air, and water; and all living organisms in association with land, air, and water.” We are retaining that proposed definition without modification in this final rule.

Numerous commenters stated that the proposed definition of

genetic engineering

requires greater clarity. Several commenters asked APHIS to clarify that “synthetic” nucleic acids, for the purposes of this regulation, are those that are non-naturally occurring. Some commenters requested that APHIS clarify what is meant by both “recombinant” and “synthetic” nucleic

acids and cited the definitions and exemptions in the National Institutes of Health (NIH) Guidelines for Research Involving Recombinant or Synthetic Nucleic Acid Molecules” (

https://osp.od.nih.gov/wp-content/uploads/NIH_Guidelines.pdf

). One commenter stated that they understood the term “synthetic nucleic acid” to refer to a sequence that was created “new from scratch,” and not to a plant's nucleic acid sequence that was modified.

APHIS does not agree that the term “recombinant” requires further definition in these regulations. After nearly half a century of research and development involving recombinant nucleic acids, the term “recombinant nucleic acids” is well understood. The definition that APHIS proposed was based on the definition of “recombinant and synthetic nucleic acids” contained in Section I-B of the NIH Guidelines. Accordingly, by “synthetic” nucleic acids we mean nucleic acids that are chemically or by other means synthesized or amplified, including those that are chemically or otherwise modified but can base pair with naturally occurring nucleic acid molecules. Such nucleic acids are not limited to those that are non-naturally occurring. They could also include nucleic acids with sequences identical to those that are naturally occurring, but which have been synthesized or amplified, rather than constructed by joining nucleic acid molecules (nucleic acids that have been so constructed are recombinant nucleic acids). APHIS agrees that greater clarity regarding the term “synthetic” would provide developers and other stakeholders with a clearer picture of the products that are included within the scope of the regulations. Therefore, we are changing the definition of “genetic engineering” to “techniques that use recombinant, synthesized, or amplified nucleic acids to modify or create a genome.” This change is consistent with the objectives of the Coordinated Framework, in that it aligns our usage of the term “synthetic” with that of the NIH.

One commenter believes that the definition for

genetic engineering

should include changes to the epigenome.

APHIS does not agree. Epigenetic changes are caused by endogenous regulatory processes, such as DNA methylation and histone modifications through naturally occurring enzymes. Epigenetic changes are also caused by small naturally occurring RNA molecules. Epigenetic changes reflect an interaction of the genome with the environment that leads to changes in gene expression without changing the sequence of DNA. Epigenetic engineering differs from genetic engineering in that the former merely adjusts the innate potential of the genome of an existing organism, whereas genetic engineering has the potential to create organisms that could not exist but for the technology.

Some commenters recommended that we add a definition of

genetically engineered organism

to provide greater clarity relating to which organisms would be regulated. The following language was a suggested definition: “An organism developed using genetic engineering, excluding those offspring that do not retain the genetic modification of the parent. For the purposes of this part, a plant will not be considered a genetically engineered organism if it meets any of the criteria outlined in § 340.1(b)(1)(3).”

We do not agree with this comment. At the forefront, the SECURE rule establishes clear exemptions for products that are not subject to regulatory oversight under part 340, and, thereafter, sets forth definitions for

genetic engineering

and for

organism.

Although we are able to offer regulatory relief in part 340 by excluding those products of biotechnology that mimic what can be achieved though plant breeding, APHIS has not, in this rulemaking or prior rulemakings involving part 340, taken the position that genome editing does not constitute genetic engineering. Taking such a position would be inconsistent with the generally accepted scientific characterization of genome editing technology (Knott and Doudna, 2018). While some commenters have asked APHIS to revisit its proposed definition of “genetically engineered organism” from the 2017 proposed rule involving part 340, even in that rulemaking APHIS did not take the position that genome editing was outside the scope of genetic engineering. Instead, APHIS explained it was defining “genetically engineered organism” for the purpose of establishing regulatory exemptions from part 340, including exemptions for certain organisms created using techniques that fall within the scope of genetic engineering, as follows: APHIS “would also exclude, from its definition of GE organism, certain organisms that are

created using techniques that fall within the scope of genetic engineering,

but that could otherwise have been produced using traditional breeding techniques . . . .” (82 FR pp.7008 and 7015, January 19, 2017). As discussed above, the SECURE rule establishes regulatory exemptions at the forefront, which promotes clarity regarding the scope of part 340, and avoids adopting a confusing characterization of techniques of biotechnology.

A couple of commenters stated that the proposed rule lacked a definition of

natural gene pool

and a discussion of its relevance in terms of safety.

The term was used in the regulatory text in § 340.1(b)(3). As discussed above, we have removed “natural” from that paragraph. We discussed the relevance of exemption under paragraph (b)(3) to plant pest risk above. We are, however, adding a definition of the term

gene pool

to the regulations in this final rule in response to these comments.

Gene pool

is defined as germplasm within which sexual recombination is possible as a result of hybridization, including via methods such as embryo culture or bridging crosses.

One commenter viewed our proposed definition of

person

as potentially problematic in that it could open APHIS to legal challenges. The commenter expressed concern that because the definition includes not only individuals, business entities, and associations but also any other “organized group,” the argument could be made that APHIS falls under the definition. If so, according to the commenter, there might be the possibility of a conflict if decisions under these regulations are taken by the Administrator of APHIS. The commenter requested clarification on this issue.

The definition of

person

would apply to individuals or entities regulated by APHIS, including APHIS. Under the law, a company is an entity that is recognized as a legal person that exists independently, with rights and liabilities. APHIS has, in the past, issued itself permits in conjunction with enforcement of the regulations so that plant products could move legally across state lines. This practice is not inconsistent with the PPA or with the prior or new regulations. Therefore, regulation by APHIS under part 340 will not create conflict or otherwise be adversely impacted.

A commenter stated that the proposed definition of

plant pest

is too broad and could be construed to cover model organisms, such as

Drosophila melanogaster,

that do not have significant negative effects on agriculture. The commenter stated that an overly broad definition is of concern to biomedical researchers because some invertebrates they use could be classified as plant pests. Noting the lack of a mechanism to acknowledge that an organism that consumes plant material is not detrimental to agriculture, the commenter recommended that APHIS establish a mechanism for classifying an

organism as “agriculturally unimportant within the plant pest category” and that such a classification have influence on APHIS' regulatory processes.

APHIS appreciates the comment, but does not believe that it is necessary for APHIS to establish such a mechanism. The definition of

plant pest

is based directly on, and does not exceed, the definition of the term in the PPA. The proposed regulations contained an exemption from the requirement for permit for interstate movement for

Arabidopsis thaliana.

In this final rule, we are adding an exemption from some permitting requirements for GE

Drosophila melanogaster,

which we will discuss in more detail below, under the subheading “Permits.”

Another commenter stated that by adopting a definition of

plant pest

that aligns with the definition provided in the PPA, APHIS would regulate a broad range of GE animals, including those used in medical research, thereby imposing large, new, and unwarranted regulatory burdens on researchers in medical research and other fields.

APHIS disagrees with the comment. As we stated in the preamble to the proposed rule, while the PPA gives APHIS authority to regulate any nonhuman animal as a plant pest, it is longstanding APHIS policy not to regulate vertebrate animals as plant pests. In the absence of such a policy, all herbivores and omnivores could be considered plant pests, and thus subject to regulation, an untenable position since this would require APHIS to consider livestock, such as cows, sheep, and horses, as well as many laboratory research animals, to be plant pests.

In the June 2019 proposed rule, we defined

plant pest risk

as “[t]he possibility of harm to plants resulting from introducing or disseminating a plant pest or exacerbating the impact of a plant pest.” Many commenters viewed the proposed definition as vague and potentially problematic due to the terminology we used.

Commenters expressed concern that the words “possibility of” in the proposed definition are vague and uncharacteristic of standard risk assessment terminology and methodology, which characterizes risk as either a likely or probable adverse outcome. Some commenters requested that the definition of

plant pest risk

be defined in terms of the likelihood and magnitude of harm. Commenters also expressed concern that the word “harm” in the proposed definition is inconsistent with the PPA, and that the regulatory end-point should be risk of causing injury to, damage to, or disease in any plant or plant product. It was stated that the inconsistency and lack of precision in the terminology used in the proposed definition could leave risk-based decisions made by APHIS open to criticism or challenge for not addressing all possibilities for harm, no matter how unlikely.

APHIS agrees with the commenters that greater clarity and consistency in the definition of

plant pest risk

would be useful. APHIS is revising the definition accordingly. We agree that the words “possibility of” could be construed in a manner that is inappropriate. Numerous scenarios could be put forward as the basis for events that represent the “possibility” of harm without any plausible basis for concluding that such scenarios have any likelihood of occurring. The glossary of the Society for Risk Analysis (SRA), which is available at

https://www.sra.org/sites/default/files/pdf/SRA_glossary_20150622.pdf,

defines

risk

as, among other things, “the potential for realization of unwanted, negative consequences of an event.” The SRA glossary makes clear the distinction between the qualitative definition of risk and the metrics that are used to measure or characterize risk, which are framed in terms of likelihood and magnitude of an adverse outcome. We view a qualitative definition as more appropriate for defining risk, and use likelihood and consequence to evaluate scientifically plausible risks identified in the RSR process discussed below under the subheading “Regulatory Status Review.” We also find the SRA terminology to be more useful than “possibility of” and are revising our definition of

plant pest risk

accordingly. We are also revising the definition to refer to injury to, damage to, or disease in any plant or plant product. Accordingly, this final rule defines

plant pest risk

as “[t]he potential for direct or indirect injury to, damage to, or disease in any plant or plant product resulting from introducing or disseminating a plant pest, or the potential for exacerbating the impact of a plant pest.”

Importantly, while APHIS defines

plant pest risk

in this rule in reference to the potential for direct or indirect injury, damage, or disease, the RSR process itself is based on standard risk assessment practices and uses a methodology that focuses on a likelihood and magnitude assessment of

plausible

risks. Since the RSR process will require that a plausible risk be identified in order to proceed with further risk assessment, it will not be an open-ended evaluation of any conceivably “possible” scenario that could be imagined.

One commenter stated that the term

plant-trait-MOA is

not defined as a combination, though the individual terms are defined in the proposed rule, and that if the combination has its own meaning, APHIS should clarify that.

The term

plant-trait-MOA

refers to three individual terms/factors for analyzing whether certain GE organisms may present a plausible pathway to plant pest risk and by which we determine whether a product actually poses a plant pest risk.

Under the definition of

responsible person

in the June 2019 proposed rule, responsibility for maintaining control over a GE organism under permit during its movement and assuring compliance with all permitting conditions could be given to an individual or an institution. A commenter stated that individuals should not be included under the definition. According to the commenter, responsibility should reside only with the institution with which the signatory or any other individual bearing such responsibility is affiliated. The commenter pointed out that staff often move among jobs well before permit conditions are fulfilled.

As discussed in the preamble to the June 2019 proposed rule, attributing responsibility for a GE organism moved under permit to only an institution may be problematic for enforcement of the regulations, because such responsibility can be diffused, resulting in no individual's being held responsible for compliance with the permit conditions, the regulations in part 340, and the PPA. Our definition ensures that for each permit, there is a single individual who is responsible for ensuring an institution's compliance with permit conditions, regulatory requirements, and the PPA. If this individual moves to a different job or otherwise leaves an institution, responsibility for any permits can be officially transferred, subject to APHIS' approval, to another qualified individual, as described in § 340.5(i)(10) of this final rule (“permit conditions”).

A commenter stated that there is no justification for the requirement, contained in the proposed definitions of both

agent

and

responsible person,

that they be legal U.S. residents, and that there is no means of verifying such a requirement.

We are retaining the requirement, as it would be a stronger mechanism for ensuring accountability in the regulatory program than the existing definition. We have learned through administration of the program that the existing definition is not adequate, and has not provided the necessary framework to hold noncompliant

developers responsible (

e.g.,

academic researchers who returned to their native countries without taking steps to destroy their GE-test material prior to departure).

Finally, we have revised the definition of

State

to read as follows: “[a]ny of the several States of the United States, the Commonwealth of the Northern Mariana Islands, the Commonwealth of Puerto Rico, the District of Columbia, Guam, the Virgin Islands of the United States, or any other territories or possessions of the United States.” This definition aligns with that contained in the PPA.

Regulatory Status Review

Section 340.4 of the June 2019 proposed rule set out the RSR process, under which developers may request that APHIS evaluate their novel plants and determine whether or not they fall within the scope of the regulations,

i.e.,

under one or more of the categories in § 340.2. The section contained requirements for submitting requests for reviews and re-reviews, including supporting information; listed the factors that APHIS would consider in the course of its reviews; described the review process; and provided for public notice of RSR determinations. Commenters addressed all these topics.

As noted in the preamble to the June 2019 proposed rule, the RSR process applies only to GE plants. APHIS specifically solicited comments on whether the scope of the RSR should be expanded to include non-plant GE organisms as well as GE plants, whether some equivalent process for evaluating such organisms for regulatory status should be developed instead, and, if so, what factors APHIS should consider in its analyses.

Several commenters did request that APHIS develop a process to evaluate the regulatory status of non-plant GE organisms, based on the subject organism's potential plant pest risk; however, the commenters did not provide specifics on what factors APHIS should consider in its analyses. APHIS believes that further discussion and outreach with impacted developers and other stakeholders on this issue is required before pursuing rulemaking.

We received several comments pertaining to the re-review process. Some commenters stressed the need to consider whether our requirements adequately address the risk of requests for spurious reviews. Noting that we proposed to require that any request for a re-review be supported by “new, scientifically valid evidence bearing on plant pest risk,” commenters urged us to clarify what we mean by “scientifically valid evidence” in order to ensure that trivial evidence or conjecture, or publications in non-credible online “scientific” journals, cannot form the basis of a request. Clarification was also requested as to whether re-reviews can be initiated for all products for which RSRs have been completed or only for those found after an initial RSR to be subject to the part 340 regulations. One commenter stated that in cases of re-reviews initiated by APHIS, APHIS needed to provide for due process by allowing developers adequate time to respond.

APHIS agrees that requests for re-review must be based on “scientifically valid evidence” that relates to plant pest risk. APHIS has experience dealing with such requests and will conduct an objective analysis of re-review requests to determine whether re-reviews are warranted. A valid re-review request would apply only to those GE plants or plant products that were previously found to be subject to the regulations after an initial RSR was conducted.

In the June 2019 proposed rule, § 340.4(a)(4) specified information requirements for persons submitting a request for APHIS to conduct an RSR of a GE plant and stated that additional guidance on how to meet the requirements would be found on the APHIS website. A few commenters requested that APHIS either (1) incorporate the additional guidance into the regulations; (2) commit not to change the guidance without public notice and comment procedures; or (3) make clear that the additional guidance is non-binding because any changes made to it would not otherwise be subject to formal notice and comment.

After reviewing these comments, APHIS has decided to pursue the second of the three recommended options. When APHIS seeks to make a substantive change to the information provided on our website, we will indicate the proposed change, provide an explanation for it, and take public comment on it. We will then review the comments and make a determination as to whether to implement the change. In this final rule, we are revising § 340.4 to incorporate the notice-and-comment process. The revised § 340.4 also uses the term “detailed information” rather than “guidance,” which was used in the proposed rule. We are making this change, which we have placed in a new paragraph (a)(4)(iv), to clarify that in order to satisfy the broad requirements contained in the regulations for information on the comparator plant(s), the genotype of the modified plant, and the new trait(s) of the modified plant, the developer must provide the detailed information indicated on the website. We anticipate that this change will provide more consistency and predictability regarding information requirements than would have been afforded by the June 2019 proposed rule. Such predictability is important for ensuring that developers can adequately comply with the regulations and can plan their product development activities accordingly.

A number of commenters expressed concerns about specific details of how to meet the detailed information requirements for the RSR process that will be maintained on APHIS website. Some commenters were concerned that the requirement for information on the genotype of the modified plant was unclear and could be interpreted as requiring sequence information comparing the entire genome of the modified plant with that of the unmodified plant. Commenters stated that sequence information should be limited to sequence information for the specific genetic modification(s) in the plant. One commenter noted that some gene-edited products could have had genetic material inserted during development that was subsequently segregated away, and that we could clarify that the whole genome sequence information is not required by specifying that the required sequence information pertains to the targeted modified sequence.

APHIS agrees with these comments. It was not our intent to request whole genome sequence information. Rather, we are requesting sequence information on the specific targeted genetic modification(s) in the plant. We have revised the information that will be published on the APHIS website to clarify the sequence information that must be provided.

Some commenters stated that sequence information is not needed to determine whether a GE plant poses a plant pest risk, as long as developers provide the type of modification and describe the genotype by providing information on the insertion, deletion, and/or expressed gene product, and that if sequence information is required, it should be limited only to sequences that confer the trait(s) and should exclude vector sequences that are not in the final plant.

APHIS largely disagrees with these comments. The specified sequence information is needed by APHIS in order to confirm the intended trait(s) at the molecular/genetic level; to understand the MOA for purposes of assessing the plant pest impact(s), if any, of the modification(s); and to assess similarity with previously reviewed GE

plants. For inserted genetic material, APHIS requires the sequence of the entire insert for molecular characterization. All genetic elements integrated into the plant genome need to be described; therefore, vector sequence information is not required if vector sequences are not inserted. For genome editing, the sequence of the entire edited gene or functional motif of a regulatory region (

e.g.,

a transcription factor binding site in a promoter region) is required to understand the targeted sequence modification(s). The characteristics imparted by inserted or edited regulatory sequences (such as expression levels, patterns, and timing) are necessary to verify the full extent of the engineered genetic changes as part of understanding the plant pest risk associated with the modification(s).

Commenters raised concerns about how to meet the information requirements concerning the MOA. One commenter stated that while there may be information on a specific gene product, the precise mechanism of action may not be elucidated.

APHIS recognizes that the MOA may not always be well characterized. As we indicated in the preamble to the June 2019 proposed rule, we are requiring information on the MOA to the extent that it is known. We have revised the detailed information provided on the APHIS website to clarify this point.

Other commenters stated that certain information categories appear to exceed what APHIS has historically asked for when reviewing petitions for nonregulated status under the current regulations, and that RSR information requirements should align with the information APHIS has required previously, should not increase a developer's data submission burden, and should be sufficiently flexible to accommodate the nature of the particular product being evaluated. A commenter stated that gene expression data are unnecessary in many cases and that APHIS should clarify when such data would be required, such as when the intent is to change the expression pattern of a gene. Another commenter stated that information on the production, creation, or enhancement of a reservoir for a plant pest goes beyond the type of information currently submitted by developers in support of petitions for nonregulated status.

APHIS largely disagrees with these comments but recognizes that the preamble to the June 2019 proposed rule lacked sufficient clarity regarding information requirements that apply at various stages of the RSR process. The information developers must submit, as specified in § 340.4(a) of this final rule and on the APHIS website, generally aligns with information APHIS has been seeking previously, will reduce rather than increase a developer's data submission burden, and is intended to be sufficiently flexible to accommodate the nature of the plant being evaluated. Under the petition process, developers have had to submit data and information regarding a broad range of possible harms for evaluation by APHIS, regardless of whether the plant could plausibly pose a plant pest risk. The RSR process differs from the petition process in that APHIS is requesting much less information for the initial review, with no requirement for laboratory or field-test data. If APHIS is unable to identify a plausible pathway by which the GE plant could pose an increased plant pest risk in the initial review, developers will not be required to submit any additional information to APHIS. When there is a plausible pathway to plant pest risk identified, developers will receive feedback about the type(s) of information that APHIS would need to assess the identified plausible pathway and complete a plant pest risk assessment. This information could include field-test data, gene expression data, or other data relevant to assessing whether the GE plant could have increased importance as a host for plant pests. The preamble to the proposed rule discussed some of the types of information that might be required in this situation, but incorrectly made it appear as if this information would be required for all initial reviews. We now clarify that such information could be submitted during the initial review stage, but that any such submission would be optional. To clarify that additional data would be requested on the basis of identified plausible pathways to plant pest risk, APHIS has added the following language to the existing text in § 340.4(b)(3)(i): “APHIS may request additional information as needed to evaluate the factor(s) of concern.” We are revising the detailed information that will be published on the APHIS website to make this distinction clear.

One commenter found it difficult to understand how plant-trait-MOA could be adequately evaluated without field trials.

Data from field trials do not provide information about the plant-trait-MOA. As we noted in the preamble to the proposed rule, APHIS' experience in preparing risk assessments in accordance with the petition process indicates that field trial data are generally not necessary unless they address an identifiable plausible pathway to plant pest risk. The introduced trait and MOA provide the most reliable indicators of the organism's potential for plant pest risk. As we also noted in the June 2019 preamble, our conclusions are consistent with findings of reports of NAS.

7 8

7

National Research Council (NRC) 1989. Field Testing Genetically Modified Organisms: Framework for Decisions. Washington, DC. National Academy Press. 185 pp. Retrieved from

http://www.nap.edu/catalog/1431.html.

8

National Academies of Sciences, Engineering, and Medicine (NAS) 2016. Genetically Engineered Crops: Experiences and Prospects. Washington, DC: National Academy Press. 420 pp. doi: 10.17226/23395. Retrieved from

http://www.nap.edu/23395.

By having an understanding of the biology and any existing impacts of the plant, the genetic trait to be inserted into the plant, and the MOA, APHIS is able to conduct a review based upon a large body of scientific publications, as well as APHIS' knowledge and experience. Information from field tests would be unnecessary, in most cases, for a determination of regulatory status under these regulations. Accordingly, field test information would not be a generally applicable requirement for the initial RSR and would be requested only as needed when further analysis is required. This approach would not preclude developers from providing information from field tests that they consider pertinent to our analysis. For example, if a developer requested a reevaluation of a GE plant that APHIS had previously considered to be subject to regulation, field test information demonstrating a lack of plant pest risk could be provided in support of that request. Nor would the provisions preclude APHIS from asking for field test information if APHIS considers it necessary in order to conclude review of a particular request.

The revised detailed information requirements that will appear on the APHIS website are listed below.

1. A description of the comparator plant(s), to include common name(s), genus, species, and any relevant subspecies information that would distinguish the plant.

2. The genotype of the modified plant, including a detailed description of the differences in genotype between the modified and unmodified plant, specifically:

a. If genetic material is inserted into the genome, provide information on all inserted genetic material, including:

i. For genetic sequences, the name of the sequence, the donor organism(s) or source, the function of the sequence, the nucleotide sequence, and if applicable, the publicly available sequence identification, protein accession

number, and enzyme commission number. If inserted genetic sequences have been modified (

e.g.,

codon usage efficiency, gene shuffling), a statement regarding the nature and purpose of the modification, and identification of the modifications by submitting an alignment of the modified sequence with the unmodified sequence.

ii. For regulatory sequences, the function of each regulatory sequence as it relates to the gene sequence and the donor organism(s) or source of each regulatory sequence. Identify promoters as constitutive, inducible, developmental, or tissue specific. If developmental/tissue specific, describe the stage(s)/tissue(s) at/in which the promotor is intended to be active.

b. If genetic material is not inserted into, or was inserted and is no longer present in, the genome, and the genome is modified in a way that does not fall under the exemptions in § 340.1(b), provide:

i. The nature of the modification(s) and the gene(s) and function(s) being modified;

ii. For substituted based pairs, the number of substitutions;

iii. The original unmodified sequence aligned to the targeted modified sequence.

3. A detailed description of the new trait(s) of the modified plant, including:

a. The purpose and intended phenotype of the new trait and available information on the MOA by which the intended trait is conferred;

b. Any expected changes in metabolism, physiology, and development due to the trait/genetic modification, to the extent known;

c. Optional: Any additional experimental data, publications, and other science-based assessments that may be helpful for APHIS' evaluation of the potential of the plant to pose plant pest risks. Such information could include, to the extent that it is known, information about any new enzymes or other gene products produced; where, when, and at what level the introduced or modified genetic material is expressed in the plant; the biochemical action of the genetic material or its product; and how the genetic material or its product participates in or interacts with metabolic, physiological, or developmental processes in the engineered plant or in other organisms. (APHIS does not intend to require submitters to generate experimental data specifically for an RSR. However, if a submitter is aware of information or experimental data in the public domain that may support our assessment, the submitter may include the data.)

The June 2019 proposed rule specified, in § 340.4(b)(1)(i) through (iii), the factors that APHIS would consider when conducting an initial review of the plant pest risk posed by the GE plant and any sexually compatible relatives that could acquire the engineered trait, relative to that posed by their respective non-GE or other appropriate comparator(s). To provide context for the discussion that follows, we are listing those factors below, as they appeared in the proposed rule.

1. The biology of the comparator plant(s) and its sexually compatible relatives;

2. The trait and mechanism-of-action of the modification(s); and

3. The effect of the trait and mechanism-of-action on:

a. The distribution, density, or development of the plant and its sexually compatible relatives;

b. The production, creation, or enhancement of a plant pest or a reservoir for a plant pest;

c. Harm to non-target organisms beneficial to agriculture; and

d. The weedy impacts of the plant and its sexually compatible relatives.

Commenters had concerns and questions about some of the factors. One commenter stated that APHIS should clarify that a comparator could be a GE plant, even though Codex Food Safety Guidelines do not allow a GE crop to be a comparator, because the majority of certain crops, such as corn and soybean, are already GE.

APHIS agrees that in some circumstances a GE plant could be an appropriate comparator for the purpose of evaluating plant pest risk, and notes that the Codex Guidelines address food safety and do not address plant pest risk. Typically, a comparator plant is the non-GE plant from which the GE plant is derived. In some cases it may be appropriate to use another GE variety of the plant as a comparator. This could occur if, for example, a developer is using genetic engineering to add a new trait to an existing GE plant. To date, APHIS has not generally seen the use of a GE plant as a comparator, but this could change in the future as products of genetic engineering become more complex.

One commenter requested that APHIS define how it intends to determine “distribution, density, or development of the plant and its sexually compatible relatives and weediness across plant types.” Another suggested that we add a definition of

weediness

because it is mentioned in the context of the RSR.

APHIS is making no changes to the rule in response to these comments. The plant pest risk assessment framework document that accompanied the proposed rule described how the distribution (including density) of the GE plant and its sexually compatible relatives can be predicted by the biological properties of the plant compared with the known distribution and properties of the comparator(s), in the context of the receiving environment. The development of the GE plant and its sexually compatible relatives can similarly be predicted. Assessment of these factors is important for determining whether the GE trait(s) could increase the prevalence or alter the distribution of the plant or its sexually compatible relative(s) in such a way that they could have increased importance as hosts for plant pests. It is also important to point out that consideration of weediness in this manner has long been a part of the plant pest risk assessments conducted in response to petitions for nonregulated status since the 1990s, under the regulations that we are replacing in this final rule. This final rule does not change this analysis, and does not expand the scope of APHIS' consideration of weediness in evaluating plant pest risks as compared with the scope of consideration that was present in APHIS' exercise of its authority under the regulations that we are replacing.

Some commenters had concerns about the factor “harm to non-target organisms beneficial to agriculture,” and asked us to shift our focus to adverse effect on trophic functional groups beneficial to agriculture and to articulate a scientific rationale as to how a plant, whether GE or not, could pose a plant pest risk on the basis of its potentially harming an insect predator or pollinator.

Beneficial organisms such as predators and pollinators fall squarely under APHIS' authority because predators and pollinators are essential to plant health, and harm to these organisms may result in greater injury or damage to plants. APHIS analyses are based on whether a GE trait introduced into a plant will had adverse impacts on non-target organisms beneficial to agriculture. Non-target organisms beneficial to agriculture encompass a broad range of organisms that provide ecosystem services. Focusing on certain trophic guilds is not adequate to address all aspects of plant pest risk to non-target organisms beneficial to agriculture. For example, some GE traits may have greater effects on closely related groups of insects, regardless of the trophic guild of members of that group. Focusing on trophic levels may also expand the scope to impacts

outside of agriculture. When there is a scientifically plausible link to harm to non-target organisms beneficial to agriculture, the information needed for a plant pest risk analysis would be determined on a case-by-case basis, accounting for the particular biology of the GE plant, the MOA of the GE trait, and the environment.

In addition to listing the factors discussed above, proposed § 340.4(b) set out the components of the RSR process, including making determinations and providing public notice of such determinations. Proposed paragraph (b)(1) stated that when APHIS receives a request for an RSR, APHIS will conduct an initial review of the potential plant pest risk posed by the GE plant and any sexually compatible relatives that could acquire the engineered trait, relative to the plant pest risk posed by their respective non-GE or other appropriate comparator(s), based on the factors discussed above. Proposed paragraph (b)(2) stated that if APHIS is unable to identify potential plant pest risks in the initial review, the GE plant will not be subject to the regulations. Proposed paragraph (b)(3)(i) stated that if APHIS does identify potential plant pest risks in the initial review, APHIS will conduct an evaluation of the factor(s) of concern to determine the likelihood and consequence of the potential plant pest risk posed by the GE plant. Proposed paragraph (b)(3)(iii) stated that if the GE plant is found unlikely to pose a plant pest risk and, therefore, not to require regulation under part 340, then APHIS will post the finding on its website. Proposed paragraph (b)(3)(iv) stated that if APHIS is unable to find the GE plant unlikely to pose a pest risk, then the plant will require regulation, and its movement will be allowed only under permit in accordance with § 340.5.

Commenters expressed numerous concerns about this process as we described it in the proposed rule. Some thought that we provided insufficient detail, especially concerning the distinction between the initial review and the additional evaluation that some GE plants would need to undergo. Others took issue with some of the terminology that we used, stating that it lacked clarity and could lead to confusion about our regulatory focus and decision making process. Numerous commenters proposed alternative language, in some cases arguing that their proposed alternatives were more consistent with standard risk assessment terminology and the PPA than what we had proposed. Commenters also stated that in order for regulation to be appropriately calibrated with actual risk, our decision-making criteria should incorporate the concept that the plant pest risk posed by the GE plant should be greater than that posed by the plant from which it was derived.

APHIS agrees with many of these comments. In this final rule, we have amended § 340.4(b) to provide additional detail and clarity and to incorporate the concept that in order for regulation to be appropriate, the plant pest risk posed by the GE plant or its sexually compatible relatives must pose an increased plant pest risk relative to the comparator(s).

Regarding terminology, we have revised § 340.4(b) to indicate that in the initial reviews, we will make determinations concerning whether further review is necessary based on a finding of “plausible,” rather than “potential,” plant pest risks. We view the former term as more precise and more in keeping with standard risk assessment terminology. Further, since the RSR process will require that a scientifically plausible risk be identified in order to proceed with further risk assessment, the revision will ensure that the initial review will not be an open-ended evaluation of any conceivably possible scenario that could be imagined.

As noted earlier in this document, in connection with the discussion on confirmation letters, some commenters saw a need for timeframes for APHIS regulatory processes for purposes of predictability and business planning. Commenters raised the issue in connection with the RSR as well. We agree with the commenters on the need for timeframes and are adding them to paragraphs (b)(2) and (3), as discussed below.

Revised § 340.4(b)(1) contains provisions related to the initial review. The introductory text states that when APHIS receives a request for an RSR of a GE plant, APHIS will conduct an initial review to determine whether there is any plausible pathway by which the GE plant, or any sexually compatible relatives that can acquire the engineered trait from the GE plant, would pose an increased plant pest risk relative to the plant pest risk posed by the respective non-GE or other appropriate comparator(s), based on the factors listed in paragraphs (b)(1)(i) through (iii) (also listed above), which remain the same as those in the proposed rule.

Revised § 340.4(b)(2) provides that except in unforeseen circumstances, APHIS will complete the initial review within 180 days of receiving a request that meets the requirements specified in this section. If APHIS does not identify a plausible pathway by which the GE plant or its sexually compatible relatives would pose an increased plant pest risk relative to the comparator(s) in the initial review, the GE plant will not be subject to the regulations. APHIS will post information on the plant and trait and a general description of the MOA on its website.

Regarding the timeframe, while the RSR process is new to APHIS, we anticipate that in many cases the initial review may be completed rapidly (that is, within 60 to 90 days). However, for plants that APHIS has infrequently authorized in the past, we anticipate that additional time may be required to compile information on the appropriate comparator(s) needed to conduct the initial review. In addition, we anticipate that additional time may be required to compile the information on less familiar or more complex MOAs needed to conduct initial reviews. Based on our experience, we anticipate that we will generally be able to complete reviews of less familiar plants and MOAs within 180 days, barring unforeseen circumstances.

Revised § 340.4(b)(3)(i) states that if APHIS does identify a plausible pathway by which the GE plant or its sexually compatible relatives would pose an increased plant pest risk relative to the comparator(s) in the initial review, the requestor may apply for a permit and/or request that APHIS conduct an evaluation of the factor(s) of concern to determine the likelihood and consequence of the increased plant pest risk.

Revised paragraph (b)(3)(ii) states that for those GE plants for which such an evaluation is conducted, APHIS will publish the results of the evaluation in the

Federal Register

and will solicit and review comments from the public. Soliciting public comments will allow APHIS to collect information we might have missed and receive additional comment. Except in circumstances that could not reasonably have been anticipated, APHIS will complete these steps within 15 months of receiving a request for an RSR that meets our requirements. This evaluation will be similar to the current petition process, and will include, in addition to public notice and comment, preparation of any applicable National Environmental Policy Act (NEPA) analysis; hence, the longer timeline.

Revised paragraph (b)(3)(iii) states that if APHIS finds that the GE plant and its sexually compatible relatives are unlikely to pose an increased plant pest risk relative to their comparator(s), the GE plant is not subject to part 340 and APHIS will announce the final determination in a subsequent

Federal

Register

notice and post the finding on its website. If APHIS does not make such a finding, the GE plant will remain regulated, and its movement will be allowed only under permit in accordance with § 340.5.

Due to the changes made in § 340.4(b)(2) and (b)(3)(iii), we are not finalizing proposed paragraph (c), as it is no longer necessary. (There is a paragraph (c) in § 340.4 of this final rule, but it discusses when the section becomes applicable, and is discussed later in this document.) APHIS does not agree with other changes to the regulatory text suggested by some commenters. Specifically, the commenters recommended that we predicate our decisionmaking on whether the GE plant poses an “unacceptable plant pest risk” or an “unacceptable” or “unreasonable” “increase in plant pest risk.”

APHIS appreciates these comments and has given them full consideration. APHIS does not find these terms to be necessary for purposes of our decisionmaking, nor have we concluded that such terms would provide the necessary precision to become the foundation for regulatory analysis and decisionmaking. For example, these terms could be interpreted to take into account considerations unrelated to plant pest risk and, if used as a regulatory benchmark, could be used to attempt to place APHIS risk assessors in the position of deemphasizing scientific considerations. As such, APHIS does not make changes to the regulatory text under in part 340 as suggested by the commenters.

A commenter stated that just as the MOA for achieving a phenotypic trait in a GE organism should be taken into account, the MOA for achieving the genotype changes used to achieve those phenotypic traits should be taken into account as well. According to the commenter, the reason why APHIS regulations have historically been “event-specific”

9

is that genetic material is inserted into recipient plants in an essentially random manner during the genetic engineering process which can create mutations in recipients at rates of ~30-60 percent, and that uncharacterized genetic materia

This text is long and has been trimmed here. Open the source document for the complete record.