Ian & Stuart's Australian Mac: Not for Sale/Another.not.for.sale (Australia).iso/Dr. Doyle/Cotton Response

Text File | 1994-06-22 | 97KB | 1,936 lines

NBIAP BBS Note - (tm) = Trademark




              Response to Calgene Petition P93-196-01
              for Determination of Nonregulated Status
                          of BXN(tm) Cotton





                            Prepared by
              United States Department of Agriculture
             Animal and Plant Health Inspection Service
       Biotechnology, Biologics, and Environmental Protection































  I.  Executive Summary


The Animal and Plant Health Inspection Service (APHIS) has
 determined, based on a review of scientific data, that
 trademarked cotton lines, designated BXN(tm) cotton, do not
 present a plant pest risk and are therefore no longer
 regulated articles under its regulations at 7 CFR Part 340.


APHIS' determination has been made in response to a petition
 from Calgene, Inc., of Davis, California, received on July
 15, 1993.  The petition seeks a determination from APHIS
 that BXN(tm) cotton does not present a plant pest risk and is
 therefore not a regulated article.  On September 8, 1993,
 APHIS announced receipt of the Calgene petition in the
 Federal Register (58 FR 47249-47250) and stated that the
 petition was available for public view.  APHIS also
 indicated its role, as well as those of the Environmental
 Protection Agency (EPA) and the Food and Drug Administration
 (FDA), in regulation of BXN(tm) cotton, food products derived
 from it, and the herbicide bromoxynil that may be used, if
 a new label for the herbicide is approved, in conjunction
 with it.  APHIS invited written comments on whether BXN(tm)
 cotton poses a plant pest risk, to be submitted on or before
  November 8, 1993.

BXN(tm) cotton, as defined by its developer (Calgene, Inc., of
 Davis, California), is any cotton cultivar or progeny of a
 cotton line containing the BXN gene (a gene, derived from
 the soil microbe Klebsiella pneumoniae subsp. ozaenae that
 encodes the enzyme nitrilase, which can degrade the
 herbicide bromoxynil) with its associated regulatory
 sequences, i.e., sequences that allow for expression of the
 gene's enzyme product.  By definition, BXN(tm) cotton may also
 contain:  the kanr marker gene (encoding the enzyme
 aminoglycoside 3'-phosphotransferase II, which confers
 resistance to the antibiotic kanamycin) with its associated
 regulatory sequences; a DNA fragment containing the origin
 of replication of the pRi plasmid from Agrobacterium
 rhizogenes; T-DNA left and right border sequences from an
 Agrobacterium tumefaciens Ti plasmid; a segment of DNA from
 transposon Tn5; a portion of a synthetic polylinker sequence
 from lacZ'; and a segment of DNA containing the origin of
 replication of plasmid pBR322.  Expression of the BXN(tm) gene
 and the kanr gene is directed by copies of the promoter from
 the 35S gene from cauliflower mosaic virus and terminated
 using sequences derived from the tml gene from the
 octopine-type Ti plasmid pTiA6 from A. tumefaciens.  Each of
 these sequences will be discussed in detail in section IV of
  this determination.

APHIS regulations at 7 CFR Part 340, which were promulgated
 pursuant to authority granted by the Federal Plant Pest Act
 (FPPA), (7 U.S.C. 150aa-150jj) as amended, and the Plant
 Quarantine Act (PQA), (7 U.S.C. 151-164a, 166-167) as
 amended, regulate the introduction (importation, interstate
 movement, or release into the environment) of certain
 genetically engineered organisms and products.  An organism
 is not subject to the regulatory requirements of 7 Part 340
 when it is demonstrated not to present a plant pest risk.
 Section 340.6 of the regulations, entitled  Petition Process
 for Determination of Nonregulated Status,  provides that a
 person may petition the Agency to evaluate submitted data
 and determine that a particular regulated article does not
 present a plant pest risk and should no longer be regulated.
 If the agency determines that the regulated article does not
 present a risk of introduction or dissemination of a plant
 pest, the petition would be granted, thereby allowing for
  unregulated introduction of the article in question.

BXN(tm) cotton lines have previously been field tested under
 15 APHIS permits.  Permitted field tests took place at a
 total of 57 sites in 13 states.  Field tests were also
 conducted in Argentina, Bolivia, and South Africa in
 accordance with national regulatory requirements.
 Additional demonstration trials using BXN(tm) cotton were also
 performed under notification during the growing season just
 ended.  All field trials were performed essentially under
  conditions of reproductive confinement.

BXN(tm) cotton contains components from organisms that are
 known plant pathogens, i.e., the bacterium Agrobacterium
 tumefaciens and cauliflower mosaic virus.  BXN(tm) cotton has
 therefore been a regulated article under APHIS jurisdiction
 and its field tests in 1989, 1990, 1991, 1992, and 1993 have
 been in accordance with APHIS regulations.  APHIS'
 determination that BXN(tm) cotton does not present a plant pest
 risk is based on an analysis of data provided to APHIS by
 Calgene and other relevant published scientific data
 obtained by APHIS concerning the components of BXN(tm) cotton
 and observable properties of the cotton lines themselves.
 From this review, we have determined that these BXN(tm) cotton
 lines:  (1) exhibit no plant pathogenic properties; (2) are
 no more likely to become a weed than their nonengineered
 parental varieties; (3) are unlikely to increase the
 weediness potential for any other cultivated plant or native
 wild species with which the organism can interbreed; (4)
 will not cause damage to processed agricultural commodities;
 and (5) are unlikely to harm other organisms, such as bees
 and earthworms, that are beneficial to agriculture.  In
 addition, we have determined that there is a reasonable
 certainty that new progeny BXN(tm) cotton lines bred from these
 lines will not exhibit new plant pest properties, i.e.,
 properties substantially different from any observed for the
 BXN(tm) cotton lines already field tested, or those observed
  for cotton in traditional breeding programs.

Calgene has provided general information and data from field
 testing of some of the cotton lines fitting their definition
 of BXN(tm) cotton and intended to be representative of all
 those lines.  Our determination, therefore, applies to
 cotton lines that fit Calgene's definition of BXN(tm) cotton
 and that have been field tested under permit.  The effect of
 this determination is that such cotton lines will no longer
 be considered regulated articles under APHIS regulations at
 7 CFR Part 340.  Permits under those regulations will no
 longer be required from APHIS for field testing,
 importation, or interstate movement of those cotton lines or
 their progeny.  Normal agronomic practices involving these
 BXN(tm) cotton lines, e.g., cultivation, propagation, movement,
 and cross-breeding with other cotton lines, can now be
 conducted without APHIS permit.  (Importation of BXN(tm) cotton
 [and nursery stock or seeds capable of propagation] is
 still, however, subject to the restrictions found in the
 Foreign Quarantine Notice regulations at 7 CFR Part 319.)
 Variety registration and/or seed certification for
 individual cotton lines carrying the BXN(tm) gene may involve
 future actions by the U.S. Plant Variety Protection Office
  and State Seed Certification officials.

The potential environmental impacts associated with this
 determination have been examined in accordance with
 regulations and guidelines implementing the National
 Environmental Policy Act of 1969 (42 U.S.C. 4321 et seq.; 40
 CFR 1500-1509; 7 CFR Part 1b; 44 FR 50381-50384; and 44 FR
 51272-51274).  An environmental assessment (EA) was prepared
 and a finding of no significant impact (FONSI) was reached
 by APHIS for the determination that BXN(tm) cotton is no longer
 a regulated article under its regulations at 7 CFR Part 340.
 The EA and FONSI are available from APHIS upon written
  request.

The body of this document consists of two parts:  (1)
 background information which provides the regulatory
 framework under which APHIS has regulated the field testing,
 interstate movement, and importation of BXN(tm) cotton, as well
 as a summary of comments provided to APHIS on its proposed
 action; and (2) analysis of the key factors relevant to
 APHIS' decision that BXN(tm) cotton does not present a plant
  pest risk.



 II.  Background

 USDA Regulatory Framework

APHIS regulations, which were promulgated pursuant to
 authority granted by the Federal Plant Pest Act (FPPA), (7
 U.S.C. 150aa-150jj) as amended, and the Plant Quarantine Act
 (PQA), (7 U.S.C. 151-164a, 166-167) as amended, regulate the
 introduction (importation, interstate movement, or release
 into the environment) of certain genetically engineered
 organisms and products.  A genetically engineered organism
 is deemed a regulated article either if the donor organism,
 recipient organism, vector or vector agent used in
 engineering the organism belongs to one of the taxa listed
 in  340.2 of the regulations and is also a plant pest; if
 it is unclassified; or if APHIS has reason to believe that
 the genetically engineered organism presents a plant pest
  risk.

Prior to the introduction of a regulated article, a person
 is required under  340.1 of the regulations to either (1)
 notify APHIS in accordance with  340.3 or (2) obtain a
 permit in accordance with  340.4.  Introduction under
 notification ( 340.3) requires that the introduction meets
 specified eligibility criteria and performance standards.
 The eligibility criteria impose limitations on the types of
 genetic modifications that qualify for notification, and the
 performance standards impose limitations on how the
 introduction may be conducted.  Under  340.4, a permit is
 granted for a field trial when APHIS has determined that the
 conduct of the field trial, under the conditions specified
 by the applicant or stipulated by APHIS, does not pose a
  plant pest risk.

The FPPA gives USDA authority to regulate plant pests and
 other articles to prevent direct or indirect injury,
 disease, or damage to plants, plant products, and crops.
 The PQA provides an additional level of protection by
 enabling USDA to regulate the importation and movement of
 nursery stock and other plants which may harbor injurious
 pests or diseases, and requires that they be grown under
 certain conditions after importation.  For certain
 genetically engineered organisms, field testing may be
 required to verify that they exhibit the expected biological
 properties, and to demonstrate that although derived using
 components from plant pests, they do not possess plant pest
  characteristics.

An organism is not subject to the regulatory requirements of
 7 CFR Part 340 when it is demonstrated not to present a
 plant pest risk.  Section 340.6 of the regulations, entitled
  Petition Process for Determination of Nonregulated Status,
 provides that a person may petition the Agency to evaluate
 submitted data and determine that a particular regulated
 article does not present a plant pest risk and should no
 longer be regulated.  If the agency determines that the
 regulated article does not present a risk of introduction or
 dissemination of a plant pest, the petition will be granted,
 thereby allowing for unregulated introduction of the article
 in question.  A petition may be granted in whole or in part.


BXN(tm) cotton lines have been considered  regulated articles
 for field testing under Part 340.0 of the regulations in
 part because of the vector system used to transfer the
 bacterial nitrilase gene into the recipient cotton.  The
 vector system was derived from A. tumefaciens, which is on
 the list of organisms in the regulation and is widely
 recognized as a plant pathogen.  In addition, certain
 noncoding regulatory sequences were derived from plant
 pathogens, i.e., from A. tumefaciens and from cauliflower
  mosaic virus.

APHIS believes it prudent to provide assurance prior to
 commercialization that organisms, such as the BXN(tm) cotton,
 that are derived at least in part from plant pests, do not
 pose any potential plant pest risk.  Such assurance may aid
 the entry of new plant varieties into commerce or into
 breeding and development programs. The decision by APHIS
 that the BXN(tm) cotton is no longer a regulated article is
 based in part on evidence provided by Calgene concerning the
 biological properties of the BXN(tm) cotton and its similarity
 to other varieties of cotton grown using standard
 agricultural practices for commercial sale or private use.
 The BXN(tm) cotton has been field tested under 15 APHIS permits
 (92-106-01, 92-105-01, 91-357-01, 91-333-02, 91-329-04,
 91-329-03, 91-329-02, 91-329-01, 91-107-06, 91-035-07,
 90-303-02, 90-297-01, 90-016-04, 89-192-01, and 89-047-07).
 Permitted field tests took place at a total of 57 sites in
 the following 13 states:  Alabama, Arizona, Arkansas,
 California, Georgia, Hawai'i, Louisiana, Mississippi,
 Missouri, North Carolina, South Carolina, Tennessee, and
 Texas.  Field tests were also conducted in Argentina,
 Bolivia, and South Africa in accordance with national
  regulatory
requirements.  Calgene, in appendix 4 of its petition
 request, has provided field data reports from all of the
 above listed field trials.  Additional demonstration trials
 using BXN(tm) cotton were also performed under notification
  during the 1993 growing season.

The fact that APHIS regulates genetically engineered
 organisms having plant pest components does not carry with
 it the presumption that the presence of part of a plant pest
 makes a whole plant pest or that plants or genes are
 pathogenic.  The regulations instead have the premise that
 when plants are developed using biological vectors from
 pathogenic sources, use material from pathogenic sources, or
 pathogens are used as vector agents, that they should be
 evaluated to assure that there is not a plant pest risk
 (McCammon and Medley, 1990).  For each APHIS performs a
 review that allows a verification of the biology and
 procedures used; assesses the degree of uncertainty and
 familiarity; and allows the identification of any hazards,
 should they be present and predictable.  The overall aims of
 APHIS regulations in the Code of Federal Regulations at 7
 CFR Part 340 are to allow for the safe testing of
 genetically engineered organisms under an appropriate level
 of oversight, and to enable any issues of potential or
 hypothetical risks to be addressed early enough in the
 development of the new organisms to allow for the safe
  utilization of the technology in agriculture.

A certification that an organism does not present a plant
 pest risk means that there is reasonable certainty that the
 organism cannot directly or indirectly cause disease,
 injury, or damage either when grown in the field, or when
 stored, sold, or processed.  APHIS' approach to plant pest
 risk is considerably broader than a narrow definition which
 encompasses only plant pathogens.  Other traits, such as
 increased weediness, and harmful effects on beneficial
 organisms, such as earthworms and bees, are clearly subsumed
 within what is meant by direct or indirect plant pest risk.
 In APHIS' regulations at 7 CFR Part 340, a  plant pest  is
 defined as:   Any living stage (including active and dormant
 forms) of insects, mites, nematodes, slugs, snails,
 protozoa, or other invertebrate animals, bacteria, fungi,
 other parasitic plants or reproductive parts thereof;
 viruses; or any organisms similar to or allied with any of
 the foregoing; or any infectious agents or substances, which
 can directly or indirectly injure or cause disease or damage
 in or to any plants or parts thereof, or any processed,
 manufactured, or other products of plants.   A determination
 that an organism does not present a plant pest risk can be
 made under this definition, especially when there is
 evidence that the plant under consideration:  (1) exhibits
 no plant pathogenic properties; (2) is no more likely to
 become a weed than its nonengineered parental varieties; (3)
 is unlikely to increase the weediness potential for any
 other cultivated plant or native wild species with which the
 organism can interbreed; (4) does not cause damage to
 processed agricultural commodities; and (5) is unlikely to
 harm other organisms, such as bees, that are beneficial to
 agriculture.  Evidence has been presented by Calgene that
 bears on all of these topics.  In addition, because the
 Calgene petition seeks a determination regarding new cotton
 varieties containing the BXN(tm) gene, it should be established
 that there is a reasonable certainty that any new cotton
 varieties bred from BXN(tm) cotton lines will not exhibit plant
 pest properties substantially different from any observed
 for cotton in traditional breeding programs or as seen in
 the development of the BXN(tm) cotton lines already field
  tested.

 Oversight by Other Federal Agencies

 Environmental Protection Agency (EPA).

The EPA regulates the use of pesticide chemicals, including
 herbicides, in the environment.  Under the Federal
 Insecticide, Fungicide, and Rodenticide Act (FIFRA) (7
 U.S.C. 136 et seq.), EPA has the authority to regulate the
 testing, sale, distribution, use, storage, and disposal of
 pesticides.  Before a pesticide may be sold, distributed, or
 used in the United States, it must be registered under FIFRA
 section 3.  To be registered, FIFRA requires that a
 pesticide will not, when used in accordance with widespread
 and commonly recognized practice, cause  unreasonable
 adverse effects.   For a pesticide that is already
 registered, the use of the pesticide on a new crop plant
 (i.e., use on a crop for which the pesticide is not already
 registered) requires EPA approval of an amendment to the
 registration.  In determining whether to approve the new use
 of the pesticide, EPA considers the possibility of adverse
 effects to human health and the environment from the new
 use.  Under the Federal Food, Drug and Cosmetic Act (FFDCA)
 (21 U.S.C. 201 et seq.), EPA also has responsibility for
 establishing tolerances for pesticide residues on food or
  feed.

Although Buctril (bromoxynil) is currently registered as an
 herbicide for use on a number of crop plants, including
 small grains (wheat, barley, oats, rye, triticale), seedling
 alfalfa, corn, sorghum, flax, garlic, onions, mint, and
 grasses grown for seed and sod production, it is not
 currently registered for use on cotton.  Any use of Buctril
 on cotton would require the approval by EPA of an amendment
 to the registration under FIFRA and a tolerance review under
 FFDCA.  (There are established tolerances for bromoxynil on
 alfalfa, barley, cattle, corn, flaxseed, garlic, goats,
 grass, hogs, horses, mint hay, oats, onions, rye, sheep,
 sorghum, and wheat.)  The Calgene petition states that
 Rhone-Poulenc Ag Company, the manufacturer of the herbicide
 Buctril, has submitted to the EPA an amended label and
  tolerance for Buctril use on transgenic cotton.

 Food and Drug Administration (FDA).

Food safety in the United States, for products other than
 meat and poultry, is assured by regulation under the FFDCA.
 FDA's policy statement concerning the regulation of foods
 derived from new plant varieties, including genetically
 engineered plants, was published in the Federal Register on
 May 29, 1992, and appears at 57 FR 22984-23005.  Regulatory
 oversight for the safety of any food or feed products
 derived from BXN(tm) cotton is under the jurisdiction of the
  FDA.

A food additive petition (originally submitted as a request
 for advisory opinion, Docket 90A-0416, November 26, 1990),
 prepared by Calgene with regard to use of the kanr gene in
  food, is pending at FDA.



 III.  Responses to Comments

APHIS received a total of 45 comments from State officials,
 universities, farmers associations and cooperative extension
 services, environmental and consumer organizations, and
 business and professional associations.  Among these
 commenters, 34 were in favor of granting the petition, 9
 were opposed, and 2 others addressed APHIS' decision on the
  petition itself only parenthetically.

At least two-thirds of the 34 comments in favor of granting
 the petition expressed the following views:  that BXN(tm)
 cotton behaves no differently than traditional cotton
 varieties; that it is no more persistent or invasive than
 traditional varieties; that it offers the potential for
 decreased overall herbicide usage in cotton agriculture; and
 that it is a useful tool in integrated weed management
 systems.  Each of the following assertions was made by two
 commenters:  that the BXN(tm) cotton poses no plant pest risk;
 that it exhibits equal or increased yield and lint quality
 with decreased herbicide use; and that there is no chance
 that the BXN(tm) cotton varieties will be weedy.  Additionally,
 one commenter noted that in consideration of the
 consequences of outcrossing of the transgenic cotton lines,
 it must be realized that such an eventuality is inevitable,
 but that management practices favor the maintenance of
 varietal purity, and in any event there is abundant
 nontransgenic cotton pollen present in cotton production
 areas.  Another commenter indicated the importance of
 affirming the petition in enabling easier on-farm testing
 for more thorough evaluation of weed management systems and
 extension service demonstration projects.  Another pointed
 out that the new cotton varieties could be important for
 conservation tillage in accordance with the Food and
 Security Act.  One other comment indicated that the
 cauliflower mosaic virus regulatory sequences used to direct
 expression of the nitrilase and kanamycin resistance genes
 are harmless in these plants.  APHIS is aware of no
 compelling contradictory data on any of these points, and
  concurs with these comments.

Of the nine comments opposing granting this petition, four
 indicated that the petition should be rejected because there
 is no strong Federal program, to which all transgenic crops
 should be subjected, to assess and minimize their risks.
 APHIS disagrees.  The Coordinated Framework for Regulation
 of Biotechnology (51 FR 23303-23350; June 26, 1986),
 developed by the Office of Science and Technology Policy
 (OSTP), establishes a clear system for product-based
 coordinated reviews of the products of agricultural
 biotechnology.  Roles are set out for the USDA, APHIS, EPA,
 and FDA, based on the existing legal authorities of the
 respective agencies for oversight over particular aspects of
 this economic sector.  These agencies are committed to
 working cooperatively to ensure adequate oversight for the
 products of agricultural biotechnology.  The system is
 entirely adequate to identify and address any significant
 potential risks that may be posed by any of the new products
 of agricultural biotechnology.  One comment suggested that
 APHIS' authority to regulate these products under the FPPA
 is questionable.  APHIS also disagrees with this comment.
  Our
responsibilities under the Act to protect against the
 introduction or dissemination of plant pests provide broad
 authority over any products that may have potentially
 significant impacts on the environment, based on the broad
  definition of  plant pest  in the statute.

Two comments indicated that transgenic BXN(tm) cotton poses a
 threat to ecological diversity akin to threats posed by the
 introduction of exotic plants into new environments.  Three
 other comments asserted that in coming to a determination
 that BXN(tm) cotton has no potential to pose a plant pest risk,
 APHIS needs to consider potential effects of introduction of
 the transgenic cotton on centers of cotton diversity in
 developing nations.  APHIS believes that it can be clearly
 established that BXN(tm) cotton poses no threat to ecological
 diversity in the United States.  As will be discussed below,
 there is no reason to think that the diversity or prevalence
 of wild cotton relatives will be affected by introduction of
 BXN(tm) cotton, and wild cotton (Gossypium hirsutum), which
 would also likely be unaffected in any case, only grows at
 sites distant from areas of cotton production.  The general
 matter of the putative analogy between the introduction of
 an exotic species into a new environment and the
 introduction of a genetically modified crop plant is not
 relevant to this determination because the attributes of
 BXN(tm) cotton fall within the ranges established for those
 traits in traditional cotton lines, except for tolerance to
 bromoxynil.  Any impacts of that trait have been carefully
  considered.

APHIS has considered the concern about the potential effects
 of the cultivation of BXN(tm) cotton outside the United States.
 A general consideration of this topic indicates to us that
 there are no necessary impacts on cotton diversity
 occasioned by this determination to allow the cultivation,
 without permit, of BXN(tm) cotton in the United States.  Even
 if BXN(tm) cotton were to be cultivated in agricultural regions
 around centers of cotton diversity, it seems extremely
 unlikely that BXN(tm) cotton would have any effect on wild
 progenitors of cultivated cotton.  The herbicide bromoxynil
 is generally used only in agricultural contexts rather than
 on uncultivated land, and there is no reason to believe that
 the bromoxynil resistance trait would impart any selective
 advantage to a recipient plant in the absence of bromoxynil
 application.  In any event, there is already considerable
 cultivation of nontransgenic cotton around most centers of
 cotton diversity.  The major threat to many relatives of
 cotton appears to be habitat destruction (Fryxell, 1979).
 Therefore, this topic will not be discussed in any greater
 detail.  We would, however, note these additional facts:
 (1) crop plants and seeds exported from the United States,
 whether transgenic or nontransgenic varieties, are still
 subject to the phytosanitary restrictions of the importing
 nation; (2) APHIS has no jurisdiction over agricultural
 practices in foreign nations and our action does not
 constitute approval for field testing or commercialization
 of this cotton in any other nation; (3) foreign laws
 restricting or regulating field testing and/or commerce with
 transgenic cotton are unaffected by our action; and (4)
 APHIS has no jurisdiction over approval for the use of
 bromoxynil on cotton plants in foreign nations.  Scenarios
 in which an impact of BXN(tm) cotton on wild cotton varieties
 is envisioned depend, at a minimum, on a biologically
 unlikely scenario coupled with a failure of regulatory
  oversight in a foreign nation.

One comment argued that if seeds of BXN(tm) cotton are to be
 exported, the United States government  should require that
 they be labeled to state that approval of the crop under
 U.S. law carries no implication of safe use in other
 countries.   APHIS does not believe that such labelling is
 desirable.  The commenter is correct, however, in indicating
 that this determination does not carry with it any foreign
 safety presumption, inasmuch as our authority only extends
 to the borders of the United States and its territories and
 possessions.  We do not believe that the absence of such a
 labeling requirement will lead to any lack of clarity on
 this matter.  With respect to the normal transit of
 agricultural commodities, USDA regulations in place require
 that certifications for export of those commodities meet the
 phytosanitary requirements of the importing nation.  We as
 a signatory nation under the International Plant Protection
 Convention certify that movement of plants or plant
 materials out of the United States will not present the risk
 of injurious plant pests.  The Agreement on Sanitary and
 Phytosanitary Measures in the General Agreement on Tariffs
 and Trade (GATT) also requires that the parties employ
 science based risk assessment methods for commodities.  More
 specifically, it should be noted that APHIS is in frequent
 contact with agricultural officials from many foreign
 nations, including those with interest in the cultivation of
 genetically engineered cottons.  We are actively involved
 with many countries as they develop national scientific and
 regulatory frameworks that will enable them to make their
 own scientifically credible decisions about the safety of
  new crop varieties.

Several comments indicated that Calgene did not adequately
 address the toxic or deleterious effects of the herbicide
 bromoxynil and the potential for its increased use on BXN(tm)
 cotton.  Some comments also questioned Calgene's assertion
 that commercialization of BXN(tm) cotton would lead to
 decreased herbicide use.  Since this determination does not
 authorize the use of bromoxynil on cotton, APHIS does not
 believe that this information is relevant to its decision.
 EPA has authority over the use in the environment of all
 pesticidal substances, including herbicides, under FIFRA; in
 particular, EPA has jurisdiction over registration of
 bromoxynil for use on cotton.  Approval by EPA of a
 particular label condition for a pesticide is granted when,
 under the specified conditions of use, it will not generally
 cause unreasonable adverse effects on the environment.  EPA
 considers both human health and safety as well as nontarget
 effects of both the herbicide and its breakdown products in
 making a decision on registration of an herbicide.  (If EPA
 has reason to believe that metabolism in a new crop would be
 significantly different, additional studies may be required.
 Residues of the metabolites and their effect on humans would
 be evaluated whether the source of the metabolites was from
 the plant or from the soil.)  Questions regarding potential
 increased use of the herbicide bromoxynil and aggregate
 herbicide use on cotton are more properly addressed to that
 agency.  Based on our knowledge of current cultivation
 practices and herbicide use on cotton, we expect that
 overall herbicide use on cotton would change modestly, if at
 all, if even a large proportion of cotton cultivated in the
 United States were to be tolerant to bromoxynil.  A
 reduction in frequency of herbicide use might be seen in
 cases when bromoxynil can be applied postemergence and
 replace preplant treatments with other herbicides at the
 site.  In any event, copies of all comments relating to
 potential adverse consequences arising from approval of a
 label for bromoxynil used on BXN(tm) cotton are being forwarded
 to EPA.  Additionally, EPA intends to request public comment
 on the use of bromoxynil on cotton prior to its decision on
  granting a registration for this use.

One comment asserted that, in response to this petition, an
 environmental impact statement needs to be prepared in which
 APHIS must consider  every relevant environmental effect  of
 the commercialization of BXN(tm) cotton.  The comment argued
 that there will be a significant impact to the environment
 based on increased use of bromoxynil with concomitantly
 increased potential for human health effects and
 environmental exposure of fish to the persistent aqueous
 breakdown product bromoxynil octanoate.  This comment also
 noted that the burden is on APHIS to demonstrate that any
 impacts are insignificant.  APHIS agrees that it has the
 responsibility to analyze the potential impacts of its
 determination, but notes that the basis for the
 determination is a question of fact, specifically, the plant
 pest status of an organism.  We disagree that speculative
 impacts relating to the commercialization of BXN(tm) cotton are
 relevant.  This determination does not authorize the use of
 bromoxynil on cotton, nor does it constitute a license to
 commercialize BXN(tm) cotton.  Moreover, lint from BXN(tm) cotton
 plants grown under permit may currently be sold as fiber,
 because devitalized plant material is not considered a
 regulated article under our regulations.  With respect to
 consideration of the impacts of increased bromoxynil use on
 the environment, we reiterate that EPA has the
 responsibility for ensuring that any uses of herbicide will
 not cause unreasonable adverse effects on the environment
 within the context of FIFRA.  The potential issuance by EPA
 of a new label for use of the herbicide bromoxynil on BXN(tm)
 cotton and this determination by APHIS regarding the cotton
 itself are separate decisions, under consideration by
 different agencies based on distinct regulations under
 unrelated legal authorities in response to requests from two
 separate corporate entities.  One comment asserted that
 approval of the petition would violate the White House
 commitment to reduce use of pesticides.  APHIS disagrees.
 Again, we note that this determination is not an
  authorization for the use of bromoxynil on BXN(tm) cotton.

One comment expressed the opinion that the asserted benefits
 from the use of BXN(tm) cotton are speculative and
 disingenuous, while the real benefits are likely to be
 minimal or nonexistent.  APHIS believes that any analyses of
 future benefits from the use of BXN(tm) cotton would be
 speculative, but in any case are again irrelevant to the
  questions of fact addressed in this determination.

One comment asserted that APHIS has failed to establish
  data requirements sufficient to permit it to conduct a
 scientifically credible risks [sic] assessment.   APHIS
 disagrees.  The data requirements spelled out in 7 CFR
 340.6 (b) of our final rule of March 31, 1993, were
 established through notice and comment rulemaking with input
 from the public.  Those data requirements put applicants on
 notice that they are required to address the widest variety
 of topics relating to effects both within and outside the
 agricultural milieu if there is any reason to consider that
 the subject organism would pose a greater plant pest risk
 than the unmodified organism from which it was derived.  We
 do not believe that additional protection of human health or
 the environment would result from imposing broad testing
 requirements for all petitions involving transgenic plant
 varieties.  Rather, we believe that it is more efficient and
 appropriate to require applicants to address whatever issues
 are raised based on the biology of the recipient organism
 and the donor genes.  This may be accomplished through field
 tests, laboratory experimentation, agronomic observations,
 review of the scientific literature, consultations with
 experts, or other methods, depending on the nature of the
  concern.

Seven comments indicated that Calgene's data on the effects
 of gene flow to wild but nonweedy relatives such as G.
 tomentosum are inadequate or that the spread of the
 introduced genes to other plants could have serious (but
 unspecified) consequences or unknown ecological risks.
 APHIS disagrees that there is inadequate information on the
 issue of gene flow to weedy or nonweedy relatives to reach
 a decision on Calgene's petition.  APHIS believes that there
 is no reason to require more testing than has been performed
 to date by Calgene.  In order to come to a conclusion
 regarding the probability and consequences of gene transfer
 to wild relatives, APHIS has considered a great deal of
 available information.  Some of the information considered
 relates to cotton biology, e.g., the documented lack of
 ability of cotton to overwinter in most of its growing areas
 (and the existence of regional requirements for cultivation
 practices to prevent its persistence in those areas, such as
 Arizona, where it can persist) and the known differences in
 the pollination behavior and habitats of G. hirsutum and the
 native Hawai'ian species, G. tomentosum, and of the most
 recent genetic and biochemical data relating to the lack of
 introgression of genes from G. hirsutum into the native
  species.

It is important to note that, while raising concerns about
 pollination of wild cottons by BXN(tm) cotton, none of the
 commenters identified a specific risk that would
 differentiate pollination by those cottons from pollination
 by other commercial varieties.  Cotton lines bred by
 traditional means, which should be no more or less likely to
 interbreed with G. tomentosum than BXN(tm) cotton, are not
 considered to pose a threat to the wild cotton and are not
 subject to particular State or Federal regulation on this
 basis.  In addition, there is no reason to believe that a
 selective advantage will be conferred on hybrid progeny
 carrying the nitrilase or kanr genes in the absence of
 bromoxynil or kanamycin selection in the environment.  The
 data provided by Calgene in its invasiveness studies are
 confirmatory of the predicted behavior of cotton.  APHIS
 does not mean to imply by this discussion that it believes
 that gene flow to wild relatives cannot occur.  Rather, gene
 flow to wild cottons could occur in some instances,
 particularly to wild G. hirsutum, but APHIS has concluded
 that these occasional events will not be of consequence.
 The introduced genes should provide no selective advantage
 to plants growing on noncultivated land that is not treated
  with bromoxynil.

Several of the comments also specifically noted that
 Calgene's invasiveness studies were conducted at only one
 site but should have been performed in a variety of
 environments.  APHIS agrees that, as with any agronomic
 studies, data obtained at a variety of sites is useful in
 order to confirm that local environmental variations will
 not materially alter the performance of the test crop.  The
 question of how much testing should be required to confirm
 negative results is a matter of balancing hypothetical risks
 against real financial burdens.  While additional tests of
 the same type might be interesting, APHIS does not believe
 that additional studies at other sites are warranted in the
 absence of any indication that data obtained at other sites
  should be different.

One comment noted Calgene's description of wild populations
 of G. hirsutum as strand vegetation in tropical Americas
 rimming the Gulf of Mexico and extending into Florida  in
 virtual isolation from areas of agriculture,  and indicated
 that this asserted isolation must be verified
 geographically.  APHIS has investigated the distribution of
 wild G. hirsutum.  According to Dr. Paul Fryxell of Texas
 A&M University (personal communication), who is a leading
 authority on the systematics and distribution of Gossypieae,
 wild cottons are found only in southern Florida (virtually
 exclusively in the Florida Keys), whereas cultivated cottons
 are found in northernmost portions of the State.  Other wild
 G. hirsutum found around the Gulf of Mexico is to be found
 along the Mexican coast, largely along the Yucatan, and
 populations do not extend as far north as the Texas border.
 However, even if the nonagricultural land containing these
 wild cotton populations were near sites of commercial cotton
 production, APHIS believes that its determination would not
 be altered, because:  (1) any potential effects of the trait
 would not alter the weediness of the wild cotton; and (2)
 the wild cotton populations in Florida are not being
 actively protected, but have in fact been subject in the
 past to Federal eradication campaigns, because they can
 serve as potential hosts for the boll weevil, Anthonomus
  grandis Boh.

One comment contended that tests on seed germination have
 inappropriate controls, i.e., transgenic plants were
 compared with standard varieties rather than nontransgenic
 parental ones.  APHIS believes this comment is in error.
 Because cotton varieties such as Coker are not genetically
 homogeneous, it is not possible to make exact comparisons
 with parent strains.  There is noticeable variation among
 progeny produced even with propagation of a single Coker
  strain  in a single cotton field.  Therefore, it is most
 appropriate to compare transgenic lines with the range of
 parentals to see whether the phenotype in question falls
  within the range expected for normal cotton varieties.

It was also suggested in this comment that:  germination
 data obtained in greenhouses may not accurately reflect
 field germination of seeds; that the acid treatment used for
 delinting the seeds in some of the studies could possibly
 obscure some differences in dormancy; and that better
 studies would have involved the use of buried seeds.  APHIS
 agrees that each of the first two assertions is possibly
 correct, and that buried seed studies might be preferable to
 the types of studies performed.  However, APHIS does not
 believe that the cited imperfections in Calgene's studies
 are sufficient to cast into doubt the other data on BXN(tm)
 cotton.  In particular, knowledge about the lack of
 persistence and weediness of cultivated cotton and about
 cultivation practices with the crop, and Calgene's field
 study on the lack of invasiveness of BXN(tm) cotton, indicate
 that additional dormancy, however unexpected, would not pose
  a plant pest risk.

Three comments noted that, while Calgene claimed that the
 properties of BXN(tm) cotton are unchanged from those of
 parental varieties except for the intended modifications,
 the company indicated in its discussions of its yield trials
 on page 135 of the petition that the transgenic strains may
 have smaller seeds.  In these comments, the opinions were
 expressed that the observation itself merits examination and
 that the initial claim that BXN(tm) cotton does not differ from
 other varieties except for bromoxynil resistance is
 unsubstantiated.  In addition, one of the comments noted
 that Calgene's data also indicated differences between some
 BXN(tm) lines and standard controls in weight of linters and
 hulls; and that in some of the BXN(tm) lines the levels of
 gossypol are not within the reported range for the toxin.
 In response, APHIS notes first that a considerable amount of
 variation is routinely observed during crop breeding.  This
 is a normal and expected occurrence, particularly for crops
 such as cotton, in which the breeding stocks are not
 homogeneous, in the breeding process, varieties having
 desirable characteristics are selected for further
 development, while those exhibiting less desirable ones are
 excluded from it.  Slight and unremarkable changes in
 agronomic properties such as may have been observed in some
 transformants are not unexpected and raise no grounds for
 concern.  These changes are not, in APHIS' view, any
 indication that there might be as yet undiscovered
 pleiotropic effects of the introduced genes in BXN(tm) cotton
 plants.  APHIS believes that the fact that BXN(tm) cotton lines
 may have very slightly smaller seeds, as inferred by
 Calgene, poses no plant pest risk.  Calgene has, APHIS
 believes, correctly argued that while a relationship between
 seed size and increase in weediness potential may apply in
 small-seeded crops, in which seed dispersal is affected by
 factors like wind, it does not in large-seeded crops like
  cotton.

One comment asserted that, for completeness, the applicant
 should perform floristic studies and provide experimental
 data that gene transfer to wild relatives will not occur
 under the unrestricted conditions of commercial use.  APHIS
 disagrees.  APHIS believes it is, in fact, possible that
 gene transfer could occur to wild relatives; it is likely
 that gene transfer to feral G. hirsutum plants will occur.
 However, APHIS does not believe that such gene transfer
 events will cause any plant pest risk or significant impact
 to the human environment.  There should be no appreciable
 selection for maintenance of the herbicide tolerance trait
 outside the agricultural environment on which bromoxynil is
 applied.  Standard field management practices can control
  any herbicide tolerant cotton plants on agricultural land.

Of the remaining two comments, one noted that the
 commenter's State retains the right to regulate BXN cotton
 regardless of actions by APHIS but added that no adverse
 effects had been noted during 3 years of field testing of
 BXN cotton at multiple sites in his State.  APHIS concurs,
 noting, for example, that the State has routine jurisdiction
 over intrastate quarantine matters and seed certification.
 The other comment requested that APHIS amend the Federal
 Plant Pest Act to include any organisms that can alter the
 environment, in order that there be no loopholes in
 oversight over scientific research.  APHIS disagrees that
 there are any material deficits in its oversight authority.
 The definition of plant pest in our regulations is very
 broad and our regulations already allow that we may regulate
 genetically engineered plants that we have reason to believe
  cause plant pest risk.



 IV.  Analysis of the Properties of BXN(tm) Cotton

Brief discussions of the biology of cotton and of cotton
 cultivation practices follow in the next paragraph to help
 inform the subsequent analysis.  This information is
 expanded in subsequent sections when it is relevant in
  addressing particular issues with respect to BXN(tm) cotton.

 Biology and Cultivation of Cotton

Four species of the genus Gossypium are known as cotton,
 which is grown primarily for the seed hairs that are made
 into textiles.  Cotton is predominant as a textile fiber
 because the mature dry hairs twist in such a way that fine,
 strong threads can be spun from them.  Other products, such
 as cottonseed oil, cake, and cotton linters are byproducts
 of fiber production.  Cotton, a perennial plant cultivated
 as an annual, is grown in the United States mostly in areas
 from Virginia southward and westward to California, in a
 region often referred to as the Cotton Belt (McGregor,
  1976).

Cotton belongs to the genus Gossypium, which includes 39
 species, four of which are generally cultivated (Fryxell,
 1984).  The most commonly cultivated species, G. hirsutum
  L., is the subject of this petition.

Other cultivated species are G. arboreum L., G. barbadense
  L., and G. herbaceum L.

Four species of Gossypium occur in the United States
 (Fryxell, 1979; Kartesz and Kartesz, 1980).  G. hirsutum is
 the primary cultivated cotton.  G. barbadense is also
 cultivated.  The other two species, G. thurberi Todaro and
 G. tomentosum Nuttall ex Seemann, are wild plants of Arizona
 and Hawai'i, respectively.  G. tomentosum is known from a
  few strand locations very close to the ocean.

At least seven genomes (chromosome sets with distinctive
 gene groupings), designated A, B, C, D, E, F, and G, are
 found in the genus (Endrizzi, 1984).  Diploid species
 (2n=26) are found on all continents, and a few are of some
 agricultural importance.  The A genome is restricted in
 diploids to two species (G. arboreum, and G. herbaceum) of
 the Old World.  The D genome is restricted in diploids to
  some species of the New World, such as G. thurberi.

By far, the most important agricultural cottons are G.
 hirsutum and G. barbadense.  These are both allotetraploids
 (plants with four sets of chromosomes derived by doubling of
 chromosomes from a hybrid plant) of New World origin, and
 presumably of ancient cross between Old World A genomes and
 New World D genomes.  The simplest forms of these plants
 have 52 chromosomes, and are frequently designated as AADD.
 Four additional New World allotetraploids occur in the
 genus, including G. tomentosum, the native of Hawai'i.  G.
 tomentosum, G. hirsutum, and G. barbadense have compatible
 genome types, and can be crossed to produce viable offspring
 (although crosses with G. tomentosum are only known with
 certainty from artificial crosses in breeding programs).
 Gossypium thurberi does not successfully cross with the
  allotetraploids.
G. hirsutum is generally self-pollinating, but in the
 presence of suitable insect pollinators can exhibit cross
 pollination.  Bumblebees (Bombus spp.), Melissodes bees, and
 honey bees (Apis mellifera) are the primary pollinators
 (McGregor, 1976).  Concentration of suitable pollinators
 varies from location to location and by season, and is
 considerably suppressed by insecticide use.  If suitable bee
 pollinators are present, distribution of pollen decreases
 considerably with increasing distance.  The isolation
 distances for Foundation, Registered, and Certified seed in
 7 CFR Part 201 are 1320 feet, 1320 feet, and 660 feet,
  respectively.

The growing period for cotton, from planting until removal
 of the last harvestable cotton boll, ranges from 140 to 200
 days, depending on the planting site in the Cotton Belt
 (El-Zik et al., 1989).  Cotton as a crop is highly
 susceptible to attack by insects and plant pathogens.
 Programs requiring particular management practices to combat
 particular cotton pests are in place in some States; for
 example, State programs for pink bollworm management in the
 Southwest require that the mature crop be defoliated or
 desiccated and that stalks be shredded and plowed into the
 soil to prevent harboring of the insect. Weed management is
 a major concern in the cultivation of cotton.  Weed
 management practices in cotton cultivation have changed over
 the years (Frans and Chandler, 1989; Ridgway et al., 1984).
 While hand hoeing was the primary means of weed control up
 through the 1950's, it has become a much more minor
 component with the increasing cost of labor.  Similarly,
 flame cultivation using butane or propane burners has also
 declined as fuel costs have risen (Ridgway et al., 1984).
 Current methods for weed control in cotton production are
 cultural practices (e.g., cultivar selection, seedbed
 preparation), mechanical tillage, and chemical control.
 Most cotton today is grown using herbicides:  88 percent of
 upland cotton acreage in the United States in 1992 received
 herbicide treatments (USDA, 1993).  In 1990, cotton farmers
 applied, on average, 2.1 herbicide treatments per acre per
 growing season (USDA, 1991).  Herbicides used in cotton
 cultivation may be applied in a variety of preplant,
 preemergence, or postemergence treatments.  Some of the
 herbicides currently used in cotton cultivation are
 trifluralin, fluometuron, prometryn, and mono- and disodium
 methylarsonate.  Continuous repeated use of the same
 herbicide over many growing seasons has been implicated in
 declining cotton yields (Frans et al., 1982; Rogers et al.,
 1983; Talbert et al., 1983).  There is increasing commercial
 interest over the past few years in the  organic
  cultivation of cotton.

It has been projected that without herbicide use, cotton
 production would be reduced by approximately 32 percent
 (Abernathy, 1981).  It was estimated that weed interference
 accounted for cotton production losses of 8.4 percent in
 1983, even with herbicide use (Whitwell and Everest, 1984).


To reach its determination that BXN(tm) cotton does not present
 a plant pest risk, APHIS has analyzed not only public
 comments and basic information on the biology of cotton, but
 also data presented by Calgene and scientific data on other
 topics relevant to each of the considerations previously
 listed as relevant to a discussion of plant pest risk.
 Based on the data described, APHIS has arrived at a series
  of conclusions regarding the properties of BXN(tm) cotton.

(1) Neither the introduced genes, their products, nor the
 added regulatory sequences controlling their expression
  presents a plant pest risk in these BXN(tm) cotton plants.

The disarmed Agrobacterium tumefaciens transformation vector
 does not present a plant pest risk in BXN(tm) cotton.  The
 vector system used to transfer the BXN(tm) gene into the cotton
 nuclear genome is based on the natural tumor-inducing (Ti)
 plasmid system used by the plant pathogenic bacterium A.
 tumefaciens for plant infection and gene transfer
 (Zambryski, 1988).  (A. tumefaciens is the causal agent of
 a plant disease called crown gall.)  Calgene has presented
 evidence that the Ti-plasmids that have been used in the
 construction of all BXN(tm) cotton lines that have been field
 tested (pBrx74 and pBrx75) have been disarmed, i.e., the
 natural pathogenicity genes which result in the
 characteristic symptoms of crown gall (e.g., overproduction
 of phytohormones in the plant resulting in unusual cell and
 organ overgrowth and the formation of galls, and synthesis
 of unusual, tumor-specific amino acids) in an infected plant
 have been removed from the transferred or T-DNA.  The
 natural gene sequences between the T-DNA border sequences
 can be deleted and replaced by DNA from other sources
 without affecting the ability of A. tumefaciens to transfer
 the T-DNA to plants (Caplan et al., 1983).  Only the border
 sequences of the T-DNA are required for transfer into the
 plant nuclear genome and generally only DNA located between
 the border sequences is efficiently transferred and
 integrated (Wang et al., 1984); genes inserted into the
 T-DNA region by conventional cloning techniques will be
 transferred and integrated into the plant nuclear genome
 using this vector system (Hernalsteens et al., 1980).  The
 vector system used by Calgene is said to be binary, i.e.,
 the genes to be transferred are found on one plasmid and the
 genes encoding functions necessary for transfer are found on
  a second plasmid.

The scientific literature, reviewed by Calgene for its field
 trials and previously evaluated by APHIS in environmental
 assessments relative to those field trials for BXN(tm) cotton
 under permit, indicates that only the T-DNA region is
 transferred into the plant genome and only the sequences
 contained between the border DNA sequences are integrated
 (Fraley et al., 1986).  It has been established in the
 scientific literature that the border sequences do not
 remain intact during the process of insertion of T-DNA into
 the plant cell genome, and therefore the inserted DNA is no
 longer a functional T-DNA.  In other words, the transferred
 T-DNA segment cannot be transferred a second time to a new
 recipient using the same mechanism that originally inserted
 it into the recipient plant genome (Zambryski et al., 1982).
 The plasmid vector by itself is not viable and can only
 replicate inside bacterial cells.  Calgene has found,
 however, in examining the physical structures of integrated
 DNA in cotton transformants carrying the BXN(tm) gene, that in
 some transformation events the integration event has not
 utilized the DNA sequences of the right T-DNA border and
 additional sequences beyond the right T-DNA border may be
 integrated as well.  The only additional sequences derived
 from outside the T-DNA borders that may be present in BXN(tm)
 cotton according to the definition include:  an origin of
 replication fragment from T-DNA from the related bacterium
 Agrobacterium rhizogenes; an origin of replication derived
 from the bacterial plasmid pBR322; a DNA segment derived
 from transposon Tn5 from the bacterium Escherichia coli; and
 a synthetic polylinker sequence modified from the gene for
 B-galactosidase (lacZ'), also from E. coli.  These segments
  will be considered in detail later in this section.

Calgene has presented evidence in table 4 of their petition
 that the transferred genetic material in BXN(tm) cotton is
 genetically stable and segregates in a Mendelian fashion,
 i.e., in a fashion consistent with integration of the added
 genetic material into nuclear chromosomal DNA.  Calgene has
 also analyzed the physical structure of integrated BXN(tm)
 genetic material in several transformant lines (See figure
 1, petition; and appendix 14).  In addition to these direct
 analyses, there is a wealth of data in the scientific
 literature, some of which is presented by Calgene, showing
 that A. tumefaciens T-DNA with or without genes for
 tumorigenicity becomes integrated into nuclear chromosomal
 DNA as part of the gene transfer process.  A single
 unconfirmed report has shown that T-DNA can insert into
 chloroplast DNA (de Block et al., 1985).  As integrated
 pieces of plant chromosomes, T-DNAs are subject to the same
 rules governing chromosomal rearrangements and gene
 stability as other plant genes.  Once integrated into plant
 chromosomes (as no other type of T-DNA maintenance in
 transformed cell lines has been demonstrated), T-DNA becomes
 no different than naturally occurring plant genes in terms
 of stability, or potential ability to persist in the
 environment outside of direct progeny of transformed plants.
 The T-DNA containing the BXN(tm) gene is transmitted through
 mitosis and meiosis as a new Mendelian locus that is an
  integral part of the transformed plant's genome.

Following the use of the disarmed Agrobacterium vector
 system for cotton transformation, the bacterium has been
 killed with the antibiotic carbenicillin so that subsequent
 infection or transformation by it will not be possible
 (Fillatti et al., 1987).  Calgene has further indicated in
 its field reports that none of the transgenic cotton plants
 show disease symptoms indicative of infection by A.
  tumefaciens.

The introduced coding regions do not confer a plant pest
 risk.  The cotton plants have been transformed with the BXN(tm)
 gene, a gene encoding the enzyme nitrilase isolated from a
 strain of the bacterium Klebsiella pneumoniae subsp.
 ozaenae.  This species is a soil microorganism which is not
 known to cause disease in animals or plants.  The enzyme
 nitrilase catalyzes a specific chemical reaction, namely the
 breakdown of the herbicide bromoxynil
 (3,5-dibromo-4-hydroxybenzonitrile) to
 3,5-dibromo-4-hydroxybenzoic acid.  There is no reason to
 believe that this gene or its protein product could impart
 any capability to a BXN(tm) cotton plant to cause disease or
 damage to any other plant.  The BXN(tm) cotton plants have also
 been transformed with a kanamycin resistance (kanr) gene.
 The kanr gene encodes the enzyme aminoglycoside
 3'-phosphotransferase II, which confers resistance to the
 antibiotic kanamycin.  (The kanr gene is also frequently
 referred to in the literature as neomycin
 phosphotransferase.)  This gene was introduced as a marker,
 i.e., as a tag enabling identification of cotton cells that
 had concomitantly taken up the BXN(tm) gene.  The kanr gene was
 isolated from a transposon contained in a strain of
 Escherichia coli K12 (Beck et al., 1982; Jorgensen et al.,
 1979).  E. coli, a common enteric bacterium found in the
 human gut, is not a regulated article.  The kanr gene has no
 involvement in plant disease or damage.  Also, its use does
 not result in the presence of the antibiotic kanamycin in
 BXN(tm) cotton and does not imply that kanamycin will be used
  in the cultivation of cotton.

It has been reported (Kobayashi et al., 1993; Mahadevan,
 1963) that a nitrilase enzyme is involved in the synthesis
 of the important plant auxin indole-3-acetic acid (IAA) in
 cruciferous plants.  (The predominant pathway for IAA
 synthesis in plants does not involve a nitrile intermediate
 (Schneider and Wightman, 1978; Cohen and Bialek 1984).)  A
 nitrilase enzyme is also found in plants in the mustard and
 banana families (Thimann and Mahadevan, 1964; Mahadevan and
 Thimann, 1964) and a gene encoding nitrilase has been cloned
 from the crucifer Arabidopsis thaliana (Bartling et al.,
 1992).  There is no published evidence about the existence
 of a comparable nitrilase-mediated pathway for IAA synthesis
 among plants of the mallow family (such as cotton).  Even if
 such a pathway exists in cotton, it is quite likely that
 introduction of the nitrilase gene from K. pneumoniae subsp.
 ozaenae has had no effect on auxin-mediated growth
 regulation in the recipient plant.  This conclusion is
 reached based on two lines of evidence.  First, cotton
 transformants that carried the K. pneumoniae nitrilase gene
 were subjected to careful agronomic investigations in
 several of the field trials, which involved extensive
 monitoring of physiological and morphological
 characteristics such as leaf size, internode distance, plant
 stature, flower morphology, fertility of flowers, relative
 flower and boll abortion rates, boll size, seed per boll,
 and total seed per plant.  No abnormal characteristics were
 observed in any of the advanced transformant lines selected
 for the studies.  Second, it is unlikely that the nitrilase
 enzyme from K. pneumoniae subsp. ozaenae can efficiently
 convert the IAA precursor indole-3-acetonitrile to IAA even
 if the pathway exists in cotton because of the enzyme's high
 substrate specificity for bromoxynil:  metabolism of
 compounds related to bromoxynil but missing one or two
 bromine atoms (3-bromo-4-hydroxybenzonitrile and
 4-hydroxybenzonitrile) takes place with 6-fold and 75-fold
 reduced efficiency, respectively, in comparison to
 bromoxynil.  In addition, it should be noted that even
 though many plants (particularly crucifers) possess
 toxicants that are, or are derived from, nitrile compounds,
 the most important toxicant in cotton, gossypol, is a
 phenolic compound containing no nitrogen that is synthesized
  via an unrelated isoprenoid pathway.

The introduced regulatory sequences do not confer a plant
 pest risk.  Some of the regulatory sequences fused to the
 BXN(tm) and kanr genes were derived from organisms that are on
 the list of regulated articles.  Specifically, 3'
 transcription termination and polyadenylation sequences from
 the tml gene from the octopine-type Ti plasmid pTiA6 (Barker
 et al., 1983) are derived from A. tumefaciens and the 35S
 promoter region is derived from the cauliflower mosaic virus
 (CaMV) (Odell et al., 1985).  In addition, as a consequence
 of the transformation process, portions of the T-DNA border
 sequences were transferred to the cotton genome.  Two other
 DNA sequences present, the transposon Tn5 (Beck et al.,
 1982) and the lacZ' gene fragment containing polylinker
 sequences (Yanisch-Perron et al., 1985) are  derived from
 the bacterium E. coli, which is not considered a plant pest.
 Neither of these sequences causes any plant or animal
 disease.  Polylinker, LacZ' DNA, and Tn5 DNA are present
 only to facilitate the processes of cloning of other gene
 sequences and identification of isolates that have taken up
 the desired sequences (McBride and Summerfelt, 1990).  (LacZ
 encodes the enzyme beta-galactosidase, but only a fragment
  of the gene is present, and no protein is produced.)

Some transformants may also contain DNA derived from outside
 the T-DNA borders on the transformation vector and encoding
 the origin of replication of the pRi plasmid from the
 bacterium Agrobacterium rhizogenes.  A. rhizogenes is a very
 common microorganism that has long been known (Riker et al.,
 1930) to be responsible for the formation of hairy roots
 when the microorganism infects any of a large number of
 dicotyledonous plants.  (Though a common infection, A.
 rhizogenes accounts for little or no crop loss in nearly all
 infected species, and production of extra roots has been
 suggested as a possible means of increasing drought
 resistance in plants (Jaynes and Strobel, 1981).)  Virulence
 genes on the A. rhizogenes pRi plasmid are responsible for
 the hairy root phenotype of bacterial infection.  None of
 the virulence genes from pRi is contained on the 7.5 kb DNA
 segment that may be present in some BXN(tm) cotton lines.  The
 genetic material derived from A. rhizogenes that may be
 present in some lines was introduced into the A. tumefaciens
 plasmid in order to allow the donor plasmids to be stably
 replicated in the latter bacterium.  This segment of DNA
 encodes several polypeptides (from open reading frames
 called repA, repB, and repC) that are highly homologous to
 known proteins, involved in replication and stability of
 large plasmids in bacteria, that are found on an A.
 tumefaciens plasmid and other plasmids from enteric bacteria
 (Nishiguchi et al., 1987; Tabata et al., 1989; Mori et al.,
 1986; Theophilus and Thomas, 1987).  Calgene has provided
 evidence that these genes are physically separate from the
 regions of pRi DNA known to be involved in plant disease
 (Jouanin et al., 1985).  In addition, there are no
 plant-specific control sequences present to allow expression
 of these genes.  Therefore, these genes have no potential to
 cause any disease symptoms in the recipient cotton plants.


DNA sequences from the plasmid pBR322 (Sutcliffe, 1979) may
 also be present in some BXN(tm) cotton lines.  The pBR322
 origin of replication sequence is present to facilitate
  replication of the vector agent plasmid in E. coli.

Despite the presence of certain pathogen-derived sequences
 in the BXN(tm) genome, no crown gall, hairy root, or CaMV
 disease symptoms were observed by Calgene in any BXN(tm) cotton
 plants during greenhouse or field studies.  Calgene further
 provides evidence that expression of any of the introduced
 genes does not result in disease symptoms or the synthesis
 of products toxic to other organisms.  Levels of toxins
 normally found in cotton, such as gossypol and
 cyclopropenoids, appear to fall within normal levels.  None
 of the regulatory sequences encodes any polypeptide product.


Calgene has also monitored its BXN(tm) cotton field trials to
 verify that the disease susceptibility of its transgenic
 plants did not differ from that of parental varieties.  No
 difference in disease susceptibility was observed for the
 following diseases:  damping-off diseases caused by
 Phytophthora, Pythium, and Rhizoctonia species; Fusarium and
 Verticillium wilts; and bacterial blight caused by
  Xanthomonas campestris.

There is no published evidence for the existence of any
 mechanism, other than sexual crossing of compatible
 Gossypium species, by which these genetic sequences can be
 transferred to other organisms. Comparative analyses of
 numerous gene sequences from microorganisms and plants to
 our knowledge have never yielded any published evidence of
 strong inter-kingdom gene homologies that would be
 indicative of recent or frequent gene exchanges between
 plants and microorganisms, except for Agrobacterium-mediated
 gene transfers.  A certain amount of information can be
 found in the scientific literature (e.g., Carlson and Chelm,
 1986; Wakabayashi et al., 1986; Doolittle et al., 1990) that
 provides a suggestion that transfer of genes from plants to
 microorganisms may have occurred over evolutionary time,
 i.e., in the eons since the various times of divergence
 between the kingdoms.  A single report (Bryngelsson et al.,
 1988) has suggested that plant DNA can be taken up by a
 parasitic fungus, but no further evidence has ever been
 forthcoming that such DNA uptake has resulted in the
 transfer of a functional DNA sequence.  Additionally, it has
 been recently observed (Stierle et al., 1993) that both the
 Pacific yew (Taxus brevifolia) and an endophytic fungus
 (Taxomyces andreanae) found in its inner bark both produce
 the unusual anti-cancer substance taxol, but there is no
 published information available about the homologies between
  the taxol-synthesizing enzymes from the two organisms.

Even if a rare plant-to-microbe gene transfer were to take
 place, there is no reason to believe that such a transfer of
 any of the sequences, including the kanr gene or BXN(tm) gene,
 would pose any plant pest risk.  Also, in its petition to
 APHIS, Calgene has presented a calculation of the potential
 contribution of kanamycin-resistant bacteria derived by
 horizontal gene movement from the genome of the genetically
 engineered cotton based on a worst case scenario which
 starts with the premise that gene transfer will undoubtedly
 occur.  Based on these calculations, they conclude that
 kanamycin resistant soil bacteria arising from
 transformation from plant debris would represent no more
  than
1.4 x 10-11% of the kanamycin resistant microbes already
 present.  Based on Calgene's calculations, as well as data
 in the scientific literature, we conclude that concerns
 regarding DNA transfer from BXN(tm) cotton to microorganisms
  are at best entirely speculative.

(2)  BXN(tm) cotton has no significant potential to become a
  successful weed.

Almost all definitions of weediness stress as core
 attributes the undesirable nature of weeds from the point of
 view of humans; from this core, individual definitions
 differ in approach and emphasis (Baker, 1965; de Wet and
 Harlan, 1975; Muenscher, 1980).  In further analysis of
 weediness, Baker (1965) listed 12 common weed attributes,
 almost all pertaining to sexual and asexual reproduction,
 which can be used as an imperfect guide to the likelihood
 that a plant will behave as a weed.  Keeler (1989) and
 Tiedje et al. (1989) have adapted and analyzed Baker's list
 to develop admittedly imperfect guides to the weediness
 potential of transgenic plants; both authors emphasize the
 importance of looking at the parent plant and the nature of
  the specific genetic changes.

The parent plant in this petition, G. hirsutum, does not
 show any appreciable weedy characteristics.  The genus also
 seems to be devoid of any such characteristics; although
 some New World allotetraploid cottons show tendencies to
  weediness  (Fryxell, 1979; Haselwood et al., 1983), the
 genus shows no particular weedy aggressive tendencies.  The
 standard texts and list of weeds give no indication that
 cotton is clearly regarded as a weed anywhere (Holm et al.,
 1979; Muenscher, 1980; Reed, 1970; Weed Science Society of
 America, 1989).  Any reports that cottons behave as a weed
 are rare and anecdotal, and vague as to the nature of the
  problem.

The trait of interest, bromoxynil tolerance, is unlikely to
 increase weediness of this cotton.  Bromoxynil would not be
 applied on BXN(tm) cotton for the purpose of controlling the
 cotton itself, but rather for controlling unrelated weeds in
 the field.  To increase weediness of the cotton plant there
 would have to be selection pressure on BXN(tm) cotton (Tiedje
 et al., 1989; Office of Technology Assessment, 1988)
 associated with bromoxynil use on it.  Because bromoxynil
 will not affect the survival of BXN(tm) cotton and because G.
 hirsutum is not itself weedy, this type of selection
 pressure does not now and is unlikely ever to exist.  Even
 if bromoxynil-resistant weedy plants were ever observed,
 bromoxynil treatment would not be the control method of
 choice; many other methods of control would be readily
  available.

Examination of the new genetic sequences besides the gene
 encoding bromoxynil resistance that may be present in the
 transgenic plant shows no likelihood to increase weediness
 potential.  None of these confers any traits in any way
  associated with weediness.

Calgene's data from greenhouse studies show variability in
 germination rates among transgenic seed lines but no
 evidence of specific changes in the rate from parent to
 transgenic plant.  Calgene's burial study shows no obvious
 increase in volunteer plants from buried seeds.  In
 addition, Calgene's field reports show no obvious increase
 in volunteers from seed, regrowth from stubble, or increase
 in seed dormancy.  Calgene does report on lint
 characteristics which may suggest a decrease in seed size.
 If such a decrease were real, APHIS believes that no
 competitive advantage affecting weediness would be conferred
 on the transgenic plants by this change.  Calgene has
 correctly argued, APHIS believes, that a relationship
 between seed size and increase in weediness potential should
 only apply in small-seeded crops, in which seed dispersal is
 affected by factors like wind, and not in large-seeded crops
  like cotton.

(3)  BXN(tm) cotton will not increase the weediness potential
  of any other plant with which it can interbreed.

As discussed under the section on lack of weediness of G.
 hirsutum, neither G. barbadense, G. thurberi, nor G.
  tomentosum shows any definite weedy tendencies.

Movement of genetic material by pollen is possible only to
 those plants of a compatible chromosomal type, in this
 instance only to those allotetraploid cottons with AADD
 genomes.  In the United States, this would only include G.
  hirsutum, G. barbadense, and G. tomentosum.

BXN(tm) cotton is chromosomally compatible with wild G.
 hirsutum.  However, according to Dr. Paul Fryxell of Texas
 A&M University (personal communication), a leading authority
 on the systematics and distribution of Gossypieae, wild
 cottons are found only in southern Florida (virtually
 exclusively in the Florida Keys), whereas cultivated cottons
 are found in northernmost portions of the State.  Other wild
 G. hirsutum found around the Gulf of Mexico is to be found
 along the Mexican coast, largely along the Yucatan, and
 populations do not extend as far north as the Texas border.
 Even if the nonagricultural land containing these wild
 cotton populations were near sites of commercial cotton
 production, this determination would not be altered, APHIS
 believes, because:  (1) any potential effects of the trait
 would not alter the weediness of the wild cotton; (2) no
 authorization exists, nor is any sought, for the use of
 bromoxynil on nonagricultural land; and (3) the wild cotton
 populations in Florida are not being actively protected, but
 have in fact been subject in the past to Federal eradication
 campaigns, because they can serve as potential hosts for the
  boll weevil, Anthonomus grandis Boh.

Gossypium thurberi, the native diploid from Arizona with a
 DD genome, is not compatible with G. hirsutum pollen, so
 that BXN(tm) cotton can have no effect on this species.
 Movement to G. hirsutum and G. barbadense is possible if
 suitable insect pollinators are present, and if there is a
 short distance from transgenic plants to recipient plants.
 Any physical barriers, intermediate pollinator-attractive
 plants, and other temporal or biological impediments would
  reduce the potential for pollen movement.

Movement of genetic material to G. tomentosum is more
 speculative.  The wild species is chromosomally compatible
 with G. hirsutum, but there is uncertainty about the
 possibility for pollination.  The flowers of G. tomentosum
 seem to be pollinated by moths, not bees, and they are
 reportedly receptive at night, not in the day.  Both these
 factors greatly lessen the probability of cross-pollination.
 There was a report (Fryxell, 1979) that G. tomentosum may be
 losing its genetic identity from introgressive hybridization
 of cultivated cottons by unknown means.  Additionally,
 Stephens (1964) reported probable hybrid populations of G.
 barbadense X G. tomentosum, in a study of morphological
 attributes.  However, the most recent data, from DeJoode and
 Wendel (1992), indicate that despite the morphological
 suggestion of such hybrid populations, biochemical
 (allozyme) studies show no evidence of any such
 introgression, even with the presence of clear
 species-specific allozyme alleles.  Major factors
 influencing the survival of G. tomentosum are construction
 and urbanization, i.e., habitat destruction (Fryxell, 1979).
 APHIS believes that it is these factors, rather than gene
 introgression from cultivated cottons, that are of real
 significance to this species.  Cotton lines bred by
 traditional means, which should be no more or less likely to
 interbreed with G. tomentosum than BXN(tm) cotton, are not
 considered to pose a threat to the wild cotton and are not
 subject to particular State or Federal regulation on this
 basis.  Neither the weediness nor the survival of G.
 tomentosum, therefore, will be affected by the cultivation
 of BXN(tm) cotton, based on the facts that: the transgenic
 variety poses no increased weediness itself; the two species
 are unlikely to successfully cross in nature; and the added
 traits will confer no selective advantage in the wild
  species habitat.

In contrast to the situation with G. tomentosum, gene
 movement from G. hirsutum to G. barbadense is widespread in
 advanced cultivated stocks.  However, it is conspicuously
 low or absent in material derived from natural crosses such
 as that from Central America or the Caribbean where G.
 hirsutum and G. barbadense grow together.  The absence of
 natural introgression may be caused by any one of several
 isolating mechanisms of pollination, fertilization, ecology,
 gene incompatibility, or chromosome incompatibility (Percy
 and Wendel, 1990).  Movement of gene material from BXN(tm)
 cotton to cultivated or occasional noncultivated G.
 barbadense would therefore not likely occur at a high level.
 Any movement of genetic material from BXN(tm) cottons into G.
 barbadense is likely to be the result of intentional
 breeding practice rather than accidental crossing.  Even if
 such movement did occur, it would not offer the progeny any
 clear selective advantage over the parents in the absence of
  sustained bromoxynil use.

Should a movement of genetic material take place to these
 receptive plants and bromoxynil resistance be transferred,
 no competitive advantage would be conferred, because
 bromoxynil is not used with these plants when they are found
 in nonagricultural areas.  In agricultural areas, such
  plants would be controlled by normal agronomic practices.

(4)  BXN(tm) cotton will not cause damage to processed
  agricultural commodities.

Information provided by Calgene regarding the components and
 processing characteristics of BXN(tm) cotton revealed no
 differences in any component that could have an indirect
 plant pest effect on any processed plant commodity.  Calgene
 evaluated the effects of the genetic modifications on BXN(tm)
 cotton by measuring fiber characteristics, seed processing
 characteristics, and the biochemical composition of oil and
 meal.  Fiber characteristics were measured and the results
 were reported in the literature (Baldwin et al., 1992; Kiser
 and Mitchell, 1991).  Fiber characteristics were measured
 and compared with Coker 315 control plants for nine lines of
 BXN(tm) cotton in 1991 and two lines in 1992.  Fiber
 characteristics measured included micronaire (a measure of
 fiber fineness), length, uniformity ratio, strength,
 elongation, leaf index, and color factors.  These measured
 fiber characteristics varied between the lines but were
 within the range of the controls.  Cottonseed is processed
 into four major products:  oil, meal, hulls and linters
 (Cherry and Leffler, 1984).  For the evaluation of seed
 processing characteristics, Calgene presented data on the
 processing of delinted seeds from three lines of BXN(tm) oil,
 meal, hulls and linters.  Although these data show
 considerable variability among the lines tested, the results
 were comparable to those from control plants.  Calgene
 presented data on the composition of oil and meal derived
 from BXN(tm) cotton seed.  Oil derived from nine BXN(tm) cotton
 lines was compared to oil derived from three Coker 315
 controls and to a refined food grade cotton seed oil.  The
 composition of the oil derived from the nine BXN(tm) cotton
 lines was comparable to the Coker 315 oil and the refined
 food grade oil.  The measured values for all of the oils
 were within the expected ranges for standard, edible
 cottonseed oil, according to Codex Stan 22-1981.  Total
 protein, total nitrogen and residual oil content of
 cottonseed meal were compared for three BXN(tm) cotton lines
 and two Coker 315 controls.  These measured components
  varied between the lines but were comparable to controls.

(5)  BXN(tm) cotton will not be harmful to beneficial
  organisms, including bees.

There is no reason to believe that deleterious effects on
 beneficial organisms could result specifically from the
 cultivation of BXN(tm) cotton.  The novel proteins that will
 be expressed in the BXN(tm) cotton, nitrilase and
 aminoglycoside 3'-phosphotransferase II, are not known to
 have any toxic properties.  Calgene has provided data to
 show that the expression levels of the two proteins in
 cotton leaves are less than 0.002 percent for nitrilase and
 less than 0.008 percent for aminoglycoside
 3'-phosphotransferase II.  The lack of known toxicity for
 these proteins and the low levels of expression in plant
 tissue suggest no potential for deleterious effects on
 beneficial organisms such as bees and earthworms.
 Additionally, Calgene provides data to show that the
 introduced nitrilase has a high specificity for bromoxynil
 (3,5,-dibromo-4-hydroxybenzonitrile).  The high specificity
 of this enzyme makes it unlikely that the nitrilase would
 metabolize endogenous substrates to produce compounds toxic
 to beneficial organisms.  APHIS has not identified any other
 potential mechanisms for deleterious effects on beneficial
  organisms.



 IV.  Conclusion

APHIS has determined that cotton plants fitting the
 definition of BXN(tm) cotton that have previously been field
 tested under permit will no longer be considered regulated
 articles under APHIS regulations at 7 CFR Part 340.  Permits
 under those regulations will no longer be required from
 APHIS for field testing, importation, or interstate movement
 of those cotton lines or their progeny.  (Importation of
 BXN(tm) cotton [and nursery stock or seeds capable of
 propagation] is still, however, subject to the restrictions
 found in the Foreign Quarantine Notice regulations at 7 CFR
 Part 319.)  This determination has been made based on an
 analysis which revealed that those cotton lines:  (1)
 exhibit no plant pathogenic properties; (2) are no more
 likely to become a weed than their nonengineered parental
 varieties; (3) are unlikely to increase the weediness
 potential for any other cultivated plant or native wild
 species with which the organisms can interbreed; (4) will
 not cause damage to processed agricultural commodities; and
 (5) are unlikely to harm other organisms, such as bees, that
 are beneficial to agriculture.  APHIS has also concluded
 that there is a reasonable certainty that new progeny BXN(tm)
 cotton varieties bred from these lines will not exhibit new
 plant pest properties, i.e., properties substantially
 different from any observed for the BXN(tm) cotton lines
 already field tested, or those observed for cotton in
  traditional breeding programs.





 John H. Payne, Ph.D., Acting Director
 Biotechnology, Biologics, and Environmental Protection

 Date:

  V.  References

Abernathy, J. R.  1981.  Estimated crop losses due to weeds
 with nonchemical management.  In:  CRC Handbook of Pest
 Management in Agriculture (Pimental, D., ed.) Volume I.  CRC
  Press, Inc., Boca Raton, Florida, pp. 159-167.

Baker, H. G.  1965.  Characteristics and Modes of Origin of
 Weeds.  In:  Baker, H. G., Stebbins, G. L., eds.  The
 Genetics of Colonizing Species.  pp. 147-172.  Academic
  Press, New York and London.

Baldwin, G., Kiser, J., Mitchell, J., Germany, A.  1992.
 Impact of Buctril herbicide spray on Stoneville cotton
 containing the BROMOTOL gene.  1992 Proceedings Beltwide
  Cotton Conference 2:555-556.

Barker, R., Idler, K., Thompson, D., Kemp, J.  1983.
 Nucleotide sequence of the T-DNA region from the
 Agrobacterium tumefaciens octopine Ti plasmid pTi15955.
  Plant Molecular Biology 2:335-350.

Bartling, D., Seedorf, M., Mithîfer, A., Weiler, E. W.,
 1992.  Cloning and expression of an Arabidopsis nitrilase
 which can convert indole-3-acetonitrile to the plant
 hormone, indole-3-acetic acid.  European Journal of
  Biochemistry 205:417-424.

Beck, E., Ludwig, G., Auerswald, E. A., Reiss, B., Schaller,
 H.  1982.  Nucleotide sequence and exact localization of the
 neomycin phosphotransferase gene from transposon Tn5.  Gene
  19:327-336.

Bryngelsson, T., Gustafsson, B., Green, B., Lind, M.  1988.
 Uptake of host DNA by the parasitic fungus Plasmodiophora
 brassicae.  Physiological and Molecular Plant Pathology
  33:163-171.

Caplan, A., Herrera-Estrella, L., Inze, D., Van Haute, E.,
 Van Montagu, M., Schell, J., Zambryski, P.  1983.
 Introduction of genetic material into plant cells.  Science
  222:815-821.

Carlson, T. A., Chelm, B. K.  1986.  Apparent eukaryotic
 origin of glutamine synthetase II from the bacterium
  Bradyrhizobium japonicum.  Nature 322:568-570.

Cherry, J. P., Leffler, H. R.  1984.  Seed.  In:  Cotton
 (Kohel, R. J., Lewis, C. F., eds.). pp. 511-569.  American
 Society of Agronomy, Crop Science Society of America, and
 Soil Science Society of America.  Madison, Wisconsin.  605
  pp.

Cohen, J. D., Bialek, K.  1984.  The biosynthesis of
 indole-3-acetic acid in higher plants.  In:  The
 Biosynthesis and Metabolism of Plant Hormones (Crozier, A.,
 Hillman, J. R., eds).  Volume 23, pp. 165-182.  Cambridge
  University Press, Cambridge, United Kingdom.

De Block, M., Schell, J., Van Montagu, M.  1985.
 Chloroplast transformation by Agrobacterium tumefaciens.
  EMBO Journal 4:1367-1372.

DeJoode, D. R., Wendel, F. F.  1992.  Genetic Diversity and
 Origin of the Hawaiian Island Cotton, Gossypium tomentosum.
  American Journal of Botany.  79:1311-1319.

de Wet, J. M. J., Harlan, J. R.  1975.  Weeds and
 Domesticates:  Evolution in the Man-Made Habitat.  Economic
  Botany.  29:99-107.

Doolittle, R., Feng, D. F., Anderson, K. L., Alberro, M. R.
 1990.  A naturally occurring horizontal gene transfer from
 a eukaryote to a prokaryote.  Journal of Molecular Evolution
  31:383-388.

El-Zik, K. M., Grimes, D. W., Thaxton, P. M.  1989.
 Cultural management and pest suppression.  In:  Integrated
 Pest Management Systems and Cotton Production, (Frisbie, R.
 E., El-Zik, K. M., Wilson, L. T., eds.).  John Wiley and
  Sons, New York.  pp. 11-36.

Endrizzi, J. E., Turcotte, E. L., Kohel, R. J.  1984.
 Qualitative Genetics, Cytology, and Cytogenetics.  pp.
 82-129.  In:  Kohel, R. J. and Lewis, C. F., (eds.).
 Cotton.  American Society of Agronomy, Crop Science Society
 of America, and Soil Science Society of America.  Madison,
  Wisconsin.  605 pp.

Fillatti, J., Kiser, J., Rose, R., Comai, L.  1987.
 Efficient transfer of a glyphosate tolerance gene into
 cotton using a binary Agrobacterium tumefaciens vector.
  Bio/Technology 5:726-730.

Fraley, R. T., Roger, S. G., Horsch, R. B.  1986.  Genetic
 transformation in higher plants.  CRC Critical Reviews in
  Plant Science 4:1-46.

Frans, R. E., Chandler, J. M.  1989.  Strategies and tactics
 for weed management.  In:  Integrated Pest Management
 Systems and Cotton Production, (Frisbie, R. E., El-Zik, K.
 M., Wilson, L. T., eds.).  John Wiley and Sons, New York.
  pp. 327-359.

Frans, R. E., Talbert, R., Rogers, B.  1982.  Influence of
 long term herbicide programs on continuous cotton.
 Proceedings of the Beltwide Production Research Conference,
 National Cotton Council of America, Memphis, Tennessee, pp.
  228-229.

Fryxell, P. A.  1979.  The Natural History of the Cotton
 Tribe (Malvaceae, Tribe Gossypieae).  Texas A&M University
  Press.  College Station and London.  245 pp.

Fryxell, P. A.  1984.  Taxonomy and Germplasm Resources.
 pp. 27-57.  In:  Kohel, R. J. and Lewis, C. F., Editors.
 Cotton.  American Society of Agronomy, Crop Science Society
 of America, and Soil Science Society of America.  Madison,
  Wisconsin.  605 pp.

Haselwood, E., Motten, B., Hirano, R.  1983.  Handbook of
 Hawaiian Weeds.  University of Hawaii Press, Honolulu. 491
  pp.

Hernalsteens, J. P., Van Vliet, F., DeBeuckeleer, M.,
 Depicker, A., Engler, E., Lemmers, M., Holsters, M., Van
 Montagu, M., Schell, J.  1980.  The Agrobacterium
 tumefaciens Ti plasmid as a host vector system for
  introducing foreign DNA in plant cells.  Nature 287:654-656.

Holm, L., Pancho, J. V., Herbarger, J. P., Plucknett, D. L.
 1979.  A Geographical Atlas of World Weeds.  John Wiley and
  Sons, New York.  391 pp.

Jaynes, J. M., Strobel, G. A.  1981.  The position of
 Agrobacterium rhizogenes.  In:  International Review of
 Cytology, Supplement 13:  Biology of the Rhizobiaceae
 (Giles, K. L., Atherly, A. G., eds.).  pp. 105-125.
  Academic Press, New York.

Jorgensen, R. A., Rothstein, S. J., Reznikoff, W. S.  1979.
 A restriction enzyme cleavage map of Tn5 and location of a
 region encoding neomycin resistance.  Molecular and General
  Genetics 177:65-72.

Jouanin, L., Vilaine, F., d'Enfert, C., Casse-Delbart, F.
 1985.  Localization and restriction maps of the replication
 origin regions of the plasmids of Agrobacterium rhizogenes
  strain A4.  Molecular and General Genetics 201:370-374.

Kartesz, J. T., Kartesz, R.  1980.  A Synonymized Checklist
 of theVascular Flora of the United States, Canada, and
 Greenland.  The University of North Carolina Press.  Chapel
  Hill.

Keeler, K.  1989.  Can genetically engineered crops become
  weeds?  Bio/Technology 7:1134-1139.

Kiser, J., Mitchell, J.  1991.  Bromoxynil safened cotton:
 Trialing history, yield and quality.  Beltwide Cotton
  Production Research Conferences.  pp. 566-568.

Kobayashi, M., Izui, H., Nagasawa, T., Yamada, H.  1993.
 Nitrilase in biosynthesis of the plant hormone
 indole-3-acetic acid from indole-3-acetonitrile:  Cloning of
 the Alcaligenes gene and site-directed mutagenesis of
 cysteine residues.  Proceedings of the National Academy of
  Sciences (USA) 90:247-251.

Mahadevan, S.  1963.  Conversion of 3-indoleacetaldoxime to
 3-indoleacetonitrile by plants.  Archives of Biochemistry
  and Biophysics 100:557-558.

Mahadevan, S., Thimann, K. V.  1964.  Nitrilase:  II.
 Substrate specificity and possible mode of action.  Archives
  of Biochemistry and Biophysics 107:62-68.

McBride, K., and Summerfelt, K.  1990.  Improved binary
 vectors for Agrobacterium-mediated plant transformation.
  Plant Molecular Biology 14:269-276.

McCammon, S. L., Medley, T. L.  1990.  Certification for the
 planned introduction of transgenic plants in the
 environment.  In:  The Molecular and Cellular Biology of the
 Potato (Vayda, M. E., Park, W. D., eds.).  CAB
  International, Wallingford, United Kingdom, pp.233-250.

McGregor, S. E.  1976.  Insect Pollination of Cultivated
 Crop Plants.  Agriculture Handbook No. 496.  U.S. Government
  Printing Office. Washington, D.C.  411 pp.

Mori, H., Kondo, A., Ohshima, A., Ogura, T., Hiraga, S.
 1986.  Structure and functioning of the F plasmid genes
 essential for partitioning.  Journal of Molecular Biology
  192:1-15.

Muenscher, W. C.  1980.  Weeds.  Second Edition.  Cornell
  University Press, Ithaca and London.  586 pp.

Nishiguchi, R., Takanami, M., Oka, A.  1987.
 Characterization and sequence determination of the
 replicator region in the hairy-root-inducing plasmid pRiA4b.
  Molecular and General Genetics 206:1-8.

Odell, J. T., Nagy, F., Chua, N-H.  1985.  Identification of
 DNA sequences required for activity of the cauliflower
  mosaic virus 35S promoter.  Nature 313:810-812.

Office of Technology Assessment, United States Congress.
 1988.  New Developments in Biotechnology  3.  Field-Testing
 Engineered Organisms:  Genetic and Ecological Issues.  U.S.
  Government Printing Office, Washington, DC.  150 pp.

Percy, R. G., Wendel, J. F.  1990.  Allozyme evidence for
 the origin and diversification of Gossypium barbadense L.
  Theoretical and Applied Genetics 79:529-542.

Reed, C. F.  1970.  Selected Weeds of the United States.
 Agriculture Handbook No. 366.  Agricultural Research
 Service, United States Department of Agriculture,
  Washington, D.C.  463 pp.

Ridgway, R. L., Bell, A. A., Veech, J. A., Chandler, J. M.
 1984.  Cotton protection practices in the USA and world.
  In:  Cotton, (Kohel, R. J., Lewis, C. F., eds.).  Agronomy
 Series No. 24.  American Society of Agronomy, Crop Science
 Society of America, and Soil Science Society of America,
  Madison, Wisconsin.  pp. 266-265.

Riker, A. J., Banfield, W. M., Wright, W. H., Keitt, G. W.,
 Sagen, H. E.  1930.  Studies on infectious hairy root of
 nursery apple trees.  Journal of Agricultural Research
  41:507-540.

Rogers, B., Talbert, R., Frans, R. E.  1983.  Long term
 effects of two herbicide programs in continuous cotton.
  Proceedings of the Southern Weed Science Society 36:18.

Schneider, E. A., Wightman, F.  1978.  Auxins.  In:
 Phytohormones and Related Compounds  a Comprehensive
 Treatise (Letham, D. S., Goodwin, P. B., Higgins, T. J. V.,
 eds.).  Volume 1, pp. 29-105.  Elsevier, Amsterdam, the
  Netherlands.

Stephens, S. G. 1964.  Native Hawaiian Cotton (Gossypium
  tomentosum Nutt.).  Pacific Science 18:385-398.

Stierle, A., Strobel, G., Stierle, D.  1993.  Taxol and
 taxane production by Taxomyces andreanae, an endophytic
  fungus of Pacific yew.  Science 260:214-216.

Sutcliffe, G.  1979.  Complete nucleotide sequence of the
 Escherichia coli plasmid pBR322.  Cold Spring Harbor
  Symposium on Quantitative Biology 43-77-90.

Tabata, S., Hooykaas, P., Ota, A.  1989.  Sequence
 determination and characterization of the replicator region
 in the tumor-inducing plasmid pTiB6S3.  Journal of
  Bacteriology 171:1665-1672.

Talbert, R., Frans, R., Rogers, B., Waddle, B., Oakley, S.
 1983.  Long term effects of herbicides and cover crops on
 cotton yields.  Proceedings of the Beltwide Cotton
 Production  Mechanical Conference, National Cotton Council
  of America, Memphis, Tennessee, pp. 37-39.

Theophilus, B., Thomas, C.  1987.  Nucleotide sequence of
 the transcriptional repressor gene korB which plays a key
 role in regulation of the copy number of broad host range
  plasmid RK2.  Nucleic Acids Research 15:7443-7450.

Thimann, K. V., Mahadevan, S.  1964.  Nitrilase:  I.
 Occurrence, preparation, and general properties of the
 enzyme.  Archives of Biochemistry and Biophysics
  105:133-141.

Tiedje, J. M., Colwell, R. K., Grossman, Y. L., Hodson, R.
 E., Lenski, R. E., Mack, R. N., Regal, P. J.  1989.  The
 Planned Introduction of Genetically Engineered Organisms:
 Ecological Considerations and Recommendations.  Ecology
  70:298-315.

United States Department of Agriculture, Economic Research
 Service.  1991.  Agricultural Resources:  Inputs Situation
  and Outlook.  Publication AR-21.  Washington, DC.  54pp.

United States Department of Agriculture, Working Group on
 Water Quality.  1993.  Waterfax:  017.  1992 Field Crops
  Summary, Cotton.  Washington, DC.  2pp.

Velten, J., Velten, L., Hain, R., Schell, J.  1984.
 Isolation of a dual plant promoter fragment from the Ti
 plasmid of Agrobacterium tumefaciens.  EMBO Journal
  3:2723-2730.


Wakabayashi, S., Matsubara, H., Webster, D. A.  1986.
 Primary sequence of a dimeric bacterial hemoglobin from
  vitreoscilla.  Nature 322:481-483.

Wang, K., Herrera-Estrella, L., Van Montagu, M., Zambryski,
 P. C.  1984.  Right 25 bp terminus sequence of the nopaline
 T-DNA is essential for and determines direction of DNA
 transfer from Agrobacterium to the plant genome.  Cell
  38:455-462.

Weed Science Society of America.  1989.  Composite List of
  Weeds.  WSSA.  Champaign, Illinois.

Whitwell, T., Everest, J.  1984.  Report of 1983 cotton loss
 committee.  Proceedings of the Beltwide Production Research
 Conference, National Cotton Council of America, Memphis,
  Tennessee, pp. 257-262.

Yanisch-Perron, C., Vieira, J., Messing, J.  1985.  Improved
 M13 phage cloning vectors and host strains:  nucleotide
 sequences of the M13mp18 and pUC19 vectors.  Gene
  33:103-119.

Zambryski, P., Depicker, A., Kruger, K., Goodman, H.  1982.
 Tumor induction by Agrobacterium tumefaciens:  Analysis of
 the boundaries of T-DNA.  Journal of Molecular and Applied
  Genetics 1:361-370.

Zambryski, P.  1988.  Basic processes underlying
 Agrobacterium-mediated DNA transfer to plant cells.  Annual
  Review of Genetics 22:1-30.