Chapter 2: Legal issues in plant germplasm collecting

G. Moore
Bioversity International, Maccarese, Rome, Italy
E-mail: g.moore(at)cgiar.org
K. A. Williams
USDA-ARS National Germplasm Resources Laboratory, BARC-West Beltsville, MD 20705, USA
E-mail: Karen.Williams(at)ars.usda.gov

	2011 version (0.6 MB)	1995 version (0.3 MB)
	Open the full chapter in PDF format by clicking on the icon above.

This chapter is a synthesis of new knowledge, procedures, best practices and references for collecting plant diversity since the publication of the 1995 volume Collecting Plant Diversity: Technical Guidelines, edited by Luigi Guarino, V. Ramanatha Rao and Robert Reid, and published by CAB International on behalf of the International Plant Genetic Resources Institute (IPGRI) (now Bioversity International), the Food and Agriculture Organization of the United Nations (FAO), the World Conservation Union (IUCN) and the United Nations Environment Programme (UNEP). The original text for Chapter 2: Legal Issues in Plant Germplasm Collecting, authored by IPGRI (now Bioversity International), FAO, IUCN and UNEP, has been made available online courtesy of CABI. The 2011 update of the Technical Guidelines, edited by L. Guarino, V. Ramanatha Rao and E. Goldberg, has been made available courtesy of Bioversity International.

Please send any comments on this chapter using the Comments feature at the bottom of this page. If you wish to contribute new content or references on the subject please do so here.

Back to list of chapters on collecting

References for this chapter

Internet resources for this chapter

 

	A wild sunflower (Helianthus simulans) in Louisiana, US. Sunflower is one of the crops included in Annex 1 of the Treaty.(Photo: K.A. Williams)

Abstract

The legal environment governing the collection and conservation of germplasm has changed extensively since the Technical Guidelines were first published in 1995. This chapter looks at these changes and the implications for germplasm collection of the Convention on Biological Diversity, the Global Plan of Action for the Conservation and Sustainable Utilization of Plant Genetic Resources for Food and Agriculture, the International Treaty on Plant Genetic Resources for Food and Agriculture, and the Nagoya Protocol on Access to Genetic Resources and the Fair and Equitable Sharing of Benefits Arising from their Utilization.

 

Introduction

Many developments have occurred in the legal environment governing the collection of plant germplasm since the first edition of these Technical Guidelines, including the adoption of the Global Plan of Action for the Conservation and Sustainable Utilization of Plant Genetic Resources for Food and Agriculture (hereinafter “the Global Plan of Action”) in 1996, the entry into force of the International Treaty on Plant Genetic Resources for Food and Agriculture (hereinafter “the Treaty”) in June 2004, the establishment of the Global Crop Diversity Trust in 2004, and the adoption of the Nagoya Protocol on Access to Genetic Resources and the Fair and Equitable Sharing of Benefits Arising from their Utilization (hereinafter “the Nagoya Protocol”) in 2010. There have also been developments and new activities in the area of plant breeders’ rights.

The Global Plan of Action was adopted by 150 countries at an international technical conference in Leipzig and reflects consensus on the main priority areas covering in situ conservation and development, ex situ conservation, plant genetic resources utilization, and institutions and capacity building. It is now in the process of being updated.

The entry into force of the Convention on Biological Diversity (CBD) in 1992 (with its emphasis on the sovereign rights of countries of origin of genetic resources and on the premise that access should be on the basis of prior informed consent and mutually agreed terms) raised concerns about the potential impact of an essentially bilateral system of access and benefit-sharing on the flow of plant genetic resources for food and agriculture (PGRFA) so critical to food security and the development of sustainable agriculture. The response of the international agricultural community was to call for the revision of the International Undertaking on Plant Genetic Resources, with its emphasis on the free flow of PGRFA, to bring it into harmony with the CBD while still allowing for the exchange of the PGRFA most essential to food security without incurring prohibitive transaction costs. The product was the International Treaty on Plant Genetic Resources for Food and Agriculture (www.planttreaty.org/texts_en.htm), which was adopted by the FAO Conference in 2001 and now has some 127 Contracting Parties.

The Treaty establishes a multilateral system of access and benefit-sharing (MLS) for the plant genetic resources of a number of the crops and forages that are most important for food security and on which countries are most dependent (listed in Annex 1 of the Treaty, which can be updated from time to time by consensus of all Contracting Parties). All Annex 1 PGRFA that are under the management and control of Contracting Parties and in the public domain are automatically included in the MLS, and Contracting Parties are to encourage other holders of Annex 1 PGRFA to place them in the MLS. Access and benefit-sharing for the plant genetic resources of the genera of some 64 crops and forages listed in Annex 1 to the Treaty, as between Contracting Parties to the Treaty, is to be on the basis of multilaterally agreed terms and conditions as set out in a standard material transfer agreement (SMTA), thus reducing the transaction costs involved in negotiating access and benefit-sharing on a strictly bilateral basis. Facilitated access under the Treaty covers only the crops and forages listed in Annex 1, although a number of Contracting Parties, mostly developed countries, have chosen to extend facilitated access under the SMTA to the plant genetic resources of other crops.

The MLS was designed primarily with access to ex situ collections in mind. Access to the plant genetic resources of Annex 1 crops and forages from in situ conditions is to be provided in accordance with national legislation or (in its absence) in accordance with such standards as may be set by the Treaty’s Governing Body. Nevertheless, these resources are still covered by the MLS. The exact implications of the reference to national legislation in connection with access and benefit-sharing for PGRFA found in in situ conditions have not yet been considered by the Governing Body of the Treaty. In any case, much germplasm collection will normally be carried out in areas not covered automatically by the MLS (e.g., non-Annex 1 PGRFA or PGRFA in famers’ fields or in countries not yet party to the Treaty). The legal environment governing the collection of germplasm will thus inevitably be a patchwork of norms under both the Treaty and the CBD.

The CBD itself is a framework convention setting out the main principles governing the conservation and sustainable use of all genetic resources: access to genetic resources outside the scope of the Treaty is governed by the CBD and will be subject to the terms of the Nagoya Protocol once that protocol enters into force. Access to genetic resources and the fair and equitable sharing of the benefits arising from their use is to be on the basis of prior informed consent and on mutually agreed terms. Exactly what this means in practice is the subject of the Nagoya Protocol, which sets out in more detail the process for obtaining access to genetic resources and related traditional knowledge (as well as the fair and equitable sharing of benefits arising from their use) on the basis of prior informed consent and mutually agreed terms.

The following update concentrates on the impact of the Treaty and the Nagoya Protocol. Work has not yet started on the development of standards for access to PGRFA in in situ conditions under the MLS, which will inevitably throw more light on the meaning and implications of the provisions of the Treaty that deal with this issue. The following update also examines recent developments regarding ex situ conservation, including the adoption of the Global Plan of Action and the establishment of the Global Crop Diversity Trust. It also briefly describes new developments in the area of plant breeders’ rights and ex situ conservation techniques.

 

	Collecting germplasm of a wild bean in Florida (Photo: M. Welsh)
	Collecting germplasm of a wild chile in Paraguay (Photo: K.A. Williams)

Current Status

The International Treaty on Plant Genetic Resources for Food and Agriculture

The Treaty entered into force on 29 June 2004 and (as of 15 June 2011) has 127 Contracting Parties. The objectives of the Treaty are the conservation and sustainable use of PGRFA and the fair and equitable sharing of the benefits arising out of their use (in harmony with the CBD) for sustainable agriculture and food security. Articles 5 and 6 set down general provisions regarding conservation, exploration, collection, characterization, evaluation and documentation of PGRFA and their sustainable use. Article 5.1.(b) requires each Contracting Party, subject to national legislation, to promote the collection of PGRFA that are under threat or are of potential use, along with relevant associated information. Under paragraph 5.1.(e), Contracting Parties are required to cooperate to promote the development of an efficient and sustainable system of ex situ conservation, giving due attention to the need for adequate documentation, characterization, regeneration and evaluation, and to promote the development of appropriate technologies for this purpose. They are also to monitor the maintenance of the viability, degree of variation and genetic integrity of collections of PGRFA (Article 5.1.(f)). Nothing further is said in the Treaty as to what should constitute an “efficient and sustainable” system of ex situ conservation, but more indications are given in the Global Plan of Action adopted in 1996 (see below).

Part IV of the Treaty establishes the MLS for the plant genetic resources of crops and forages listed in Annex 1 to the Treaty: the genera of some 35 crops and 29 forages. The Contracting Parties agree to grant other Contracting Parties, or legal and natural persons under their jurisdiction, facilitated access to PGRFA included in the MLS in their countries for the purpose of conservation and utilization for research, breeding and training for food and agriculture. All PGRFA of Annex 1 crops and forages that are under the management and control of the Contracting Party and in the public domain are automatically included in the MLS. Other holders of Annex 1 PGRFA are invited to include their holdings in the MLS and Contracting Parties agree to take measures to encourage them to do so. Some light has been shed on the meaning of these terms by the Ad Hoc Advisory Technical Committee on the Standard Material Transfer Agreement and the Multilateral System of the Treaty (hereinafter “the Ad Hoc Advisory Technical Committee”) in the report on its first session in 2010 (ftp://ftp.fao.org/ag/agp/planttreaty/gb4/AC_SMTA_MLS1/ac_smta_mls1_repe.pdf).

All PGRFA in the MLS are to be made available in accordance with standard multilaterally agreed terms and conditions as set out in the SMTA adopted by the Governing Body of the Treaty at its first session in June 2006 (www.planttreaty.org/smta_en.htm). These include the requirement that the PGRFA shall be used only for research, breeding and training for food and agriculture, that access shall be accorded expeditiously and free of charge apart from administrative expenses, and that the recipients should not claim any intellectual property rights over the material received that would limit facilitated access by others to the PGRFA in the form received from the MLS. The MLS also provides for benefit-sharing on a multilateral basis. Facilitated access to PGRFA is expressly recognized as being a major benefit of the system. Other forms of benefit-sharing include exchange of information, access to and transfer of technology, capacity building and the sharing of monetary and other benefits of commercialization.

In this context, the MLS provides that a recipient who commercializes a product that is itself a PGRFA and that incorporates material accessed from the MLS should pay an equitable share of the benefits arising from the commercialization into an international fund set up for this purpose. The payment, which was set by the Governing Body at its first session at the rate of 1.1% of the gross sales of the product less 30% (i.e., 0.77%) is mandatory where the recipient takes action to restrict the availability of the product for further research and breeding (e.g., by taking out certain types of patents over the product). Where availability of the product is not so restricted, then the payment is to be voluntary, though encouraged. In addition to the monetary payments, recipients of material from the MLS are required to make available to the MLS all non-confidential information resulting from the research and development carried out on the material.

As noted by the Ad Hoc Advisory Technical Committee (FAO 2010a) in the report of its first session, the term “PGRFA under the management and control of the Contracting Parties”, encompasses both PGRFA in in situ conditions and that held ex situ”. The Treaty provides that access to PGRFA found in in situ conditions will be provided according to national legislation, or in the absence of any such legislation, in accordance with such standards as may be set by the Governing Body. This provision is without prejudice to the other provisions of the MLS. While the exact meaning of this is not entirely clear, it does imply that facilitated access should still be granted to Annex 1 PGRFA that meet the conditions for inclusion in the MLS, but that the way in which access is to be granted, including the actual process of any collection, would be subject to the requirements of national legislation. The Governing Body has not yet authorized the development of standards regarding the collection of PGRFA from in situ conditions, whether this would consist of the revision of the Code of Conduct on Plant Germplasm Collection or the development of entirely new standards. At the same time, at its second session, the Ad Hoc Advisory Technical Committee (FAO 2010b) introduced a further note of caution, suggesting that the whole question of the interrelationship of the provisions of Article 12.3.(h) regarding access to PGRFA in in situ conditions with the rest of the provisions of the MLS should be further worked on.

Whatever the outcome of the future discussions in the Governing Body on the subject of the scope and implications of Article 12.3.(h), it is clear that the international legal framework governing the collection of PGRFA from in situ conditions will be a patchwork of norms from both the Treaty and the CBD. In the end, all will depend on how these provisions are interpreted and implemented at the national level through national legislation. In any event, it is clear that the coverage of the MLS provisions will be only partial, given that non-Annex 1 material will not be covered, nor will Annex 1 PGRFA found in farmers’ fields, unless these are placed in the MLS through voluntary action or as a result of collection by national authorities and subsequent incorporation in governmental ex situ collections.

The Nagoya Protocol on Access to Genetic Resources and the Fair and Equitable Sharing of Benefits Arising from Their Utilization

Under the CBD, access to genetic resources and benefit-sharing is subject to national legislation, to prior informed consent and to mutually agreed terms with the country that is the country of origin of those resources or that has acquired them in accordance with the CBD. The rather general CBD provisions regarding access and benefit-sharing have been recently fleshed out in the Nagoya Protocol (www.cbd.int/abs/text), which was adopted in September 2010 and has now been signed by 65 states. The Protocol will enter into force 90 days after the deposit of the 50th instrument of ratification, acceptance, approval or accession.

The Nagoya Protocol lays particular stress on the fair and equitable sharing of benefits arising from the utilization of genetic resources. In this context, it draws on previous work undertaken within the framework of the CBD, including the Bonn Guidelines on Access to Genetic Resources and Fair and Equitable Sharing of the Benefits Arising out of Their Utilization (www.cbd.int/abs/bonn) (hereinafter “the Bonn Guidelines”) adopted by the Conference of Parties to the CBD in 2002. The list of possible monetary and non-monetary benefits set out in the Annex to the Nagoya Protocol is drawn directly from the Bonn Guidelines. It also deals in a parallel way with access to genetic resources and to traditional knowledge related to those resources: in both cases prior informed consent is required, and access and benefit-sharing is to be on mutually agreed terms. Where indigenous and local communities have an established right to grant access to particular genetic resources and associated traditional knowledge, then the prior informed consent of those communities must be obtained, and access and benefit-sharing must be negotiated with them on mutually agreed terms.

The Nagoya Protocol seeks to ensure legal certainty with respect to the terms and procedures for access and benefit-sharing, as well as to the genetic resources having been acquired with the prior informed consent of the country providing them and in accordance with mutually agreed terms. Thus, Parties to the Protocol are required to

• provide for legal certainty, clarity and transparency of their domestic access and benefit-sharing legislation or regulatory requirements
• provide for fair and non-arbitrary rules and procedures on accessing genetic resources
• provide information on how to apply for prior informed consent
• provide for a clear and transparent written decision by a competent national authority, in a cost-effective manner and within a reasonable period of time
• provide for the issuance at the time of access of a permit or its equivalent as evidence of the decision to grant prior informed consent and of the establishment of mutually agreed terms, and notify the Access and Benefit-sharing Clearing-House accordingly
• where applicable, and subject to domestic legislation, set out criteria and/or processes for obtaining prior informed consent or approval and for the involvement of indigenous and local communities in acquiring access to genetic resources
• establish clear rules and procedures for requiring and establishing mutually agreed terms (which are to be set out in writing)

Central to the whole process will be the Access and Benefit-sharing Clearing-House set up under Article 11 of the Nagoya Protocol to act as a central information point for legislative, administrative and policy measures on access and benefit-sharing, including procedures, requirements and information on national focal points and national authorities, and for notifications of the issuance of permits or certificates of compliance with requirements for those accessing genetic resources.

In developing and implementing legislation and regulatory schemes for access and benefit-sharing, Parties are required to create conditions to promote and encourage research that contributes to the conservation and sustainable use of biological diversity, particularly in developing countries. This includes developing simplified measures for access for non-commercial research purposes, taking into account the need to address a change of intent for such research. Parties are also required to pay due regard to cases of present or imminent emergencies that threaten or damage human, animal or plant health, as determined nationally or internationally. Parties may take into consideration the need for expeditious access to genetic resources and expeditious fair and equitable sharing of benefits arising out of the use of such genetic resources, including access to affordable treatments by those in need, especially in developing countries. Finally, they are required to consider the importance of genetic resources for food and agriculture and their special role for food security.

With respect to traditional knowledge associated with genetic resources, Parties are to take into consideration the laws, protocols and procedures of indigenous and local communities and to support, as appropriate, the development by those communities of community protocols in relation to access and benefit-sharing, of minimum standards for mutually agreed- terms for benefit-sharing and of model contractual clauses for benefit-sharing.

Parties to the Nagoya Protocol are required to take appropriate, effective and proportionate measures to provide that genetic resources and associated traditional knowledge utilized within their jurisdiction have been accessed in accordance with prior informed consent and that mutually agreed terms have been established, as required by the domestic access and benefit-sharing legislation of the country providing the genetic resources. They are also required to monitor the utilization of genetic resources, including establishing one or more checkpoints related to prior informed consent, the source of the genetic resources, the establishment of mutually agreed terms and the utilization of the genetic resources. Users will be required to provide pertinent information at these checkpoints, including internationally recognized certificates of compliance where they are available. Such certificates of compliance will serve as evidence that the genetic resources covered by the certificates have been accessed in accordance with prior informed consent and that mutually agreed- terms have been established as required. They are to include the following minimum information:

• issuing authority
• date of issuance
• provider
• unique identifier of the certificate
• the person or entity to whom prior informed consent was granted
• subject-matter or genetic resources covered by the certificate
• confirmation that mutually agreed terms were established
• confirmation that prior informed consent was obtained
• commercial and/or non-commercial use

Guidelines for following the access and benefit-sharing requirements of the Convention on Biological Diversity (CBD) for collecting germplasm in situ Prior informed consent (PIC) must be obtained before an exploration takes place. Access is under the control of the national government in countries that are Parties to the CBD. PIC is required even in countries that do not have specific national legislation on access to genetic resources. The need for PIC covers wild plants, as well as traditional crop varieties obtained from farmers or markets. The exact situation under the Treaty regarding access to PGRFA of crops listed in Annex 1 and found in in situ conditions is still not totally clear, as discussed above. In particular, no standards regarding access to PGRFA found in in situ conditions have yet been set by the Governing Body of the Treaty. However, it is clear that prior approval must always be sought before an exploration mission is launched. The following condensed steps for a potential collector to request PIC are based on the CBD, the Bonn Guidelines and the Nagoya Protocol: 1. In consultation with host-country collaborators, develop a proposal describing the collection activities. Include taxa and associated information to be collected, foreign and host-country participants, dates, locations to be visited, collecting protocol, possible benefits to the host country, intended use of the germplasm and budget. 2. Identify one of the following authorities or focal points in the host country on the CBD website (www.cbd.int/information/nfp.shtml): - Competent National Authority on Access and Benefit Sharing (ABS CNA) - National Focal Point to the Intergovernmental Committee for the Nagoya Protocol on Access and Benefit-sharing (ICNP ABS NFP) - Primary National Focal Point to the CBD (CBD NFP) 3. Send your proposal to the national authority or focal point and request information on the procedure for obtaining PIC and establishing mutually agreed terms. (Note: In some countries, no national authorities or focal points have been designated. In those cases, contact the ministry – usually the ministry of environment, agriculture or forestry – that issues scientific research or collecting permits, or which is most relevant to your work, and inquire about the procedure for requesting PIC.) 4. Follow the requirements of the host-country authority to request PIC. 5. The host-country authority may respond to your request with a letter, permit, license, material transfer agreement* or other documentation. Carefully review the documentation received to be sure that you and your institution can abide by all the terms, including the procedures for collecting the germplasm, limitations on the use of the germplasm, and the specified benefit-sharing. Additional communication may be necessary to reach a mutual agreement on terms. 6. PIC from other types of authorities may also be required. Lower administrative levels, such as provincial authorities, require separate permits in some countries. Collection of germplasm in protected areas requires permits in most countries. Collection of traditional crop varieties often requires PIC from indigenous and local communities. 7. Take measures to ensure that all the mutually agreed terms of the PIC are followed both during and after the exploration. Carry the documentation providing PIC with you on the exploration. Follow all requirements of the CBD, including those concerning the rights of indigenous and local communities and the protection of biological resources. 8. Share the germplasm, documentation and subsequent results of research with cooperators in the host country. Involve host-country stakeholders in the research as much as possible. The exact process to obtain PIC varies from country to country. In some countries, host-country collaborators rather than foreign participants may be required to request permission for access from the national authorities. Future developments on implementation of the Nagoya Protocol and the Treaty regarding the collection of genetic resources from in situ conditions should be monitored to determine whether modifications are needed in the process. - - - - - - - - - - - - *The authorities in a country may choose to use the SMTA to provide access to genetic resources listed in Annex 1 to the Treaty collected in situ (for example, by depositing the germplasm in the national genebank and then distributing it from there).

 

Presumably, the SMTA will constitute an internationally recognized certificate of compliance insofar as access to PGRFA in the MLS is concerned. In any case, the Nagoya Protocol specifically recognizes that where a specialized international access and benefit-sharing instrument, such as the Treaty and its SMTA, applies, the Nagoya Protocol will not apply for the Party or Parties to that specialized instrument: the Protocol and other specialized instruments are to be implemented in a mutually supportive manner.

The Global Plan of Action for the Conservation and Sustainable Utilization of Plant Genetic Resources for Food and Agriculture

As noted above the Global Plan of Action adopted at the Fourth International Technical Conference on Plant Genetic Resources in Leipzig in 1995 provided a sound framework for the development of an efficient and sustainable global system of ex situ conservation. The relevant provisions are contained in the policy/strategy sections for sustaining existing ex situ collections (Activity 5) and regenerating threatened ex situ accessions (Activity 6) (http://typo3.fao.org/fileadmin/templates/agphome/documents/PGR/GPA/gpaeng.pdf):

82. Policy/Strategy: The international community has interests in and responsibilities for the ex situ conservation of plant genetic resources for food and agriculture. It is this understanding which provides the basis for an effective, integrated and rational global plan to secure existing collections. Countries have national sovereignty over, and responsibility for, their own plant genetic resources for food and agriculture.

83. Full use should be made of appropriate existing facilities, including national, regional and international centres. Conserved materials should be, as appropriate, replicated and stored in long-term facilities meeting international standards, in accordance with applicable international agreements. Unintended and unnecessary duplications between collections within the networks should be reduced to promote cost efficiency and effectiveness in global conservation efforts. Countries could be assisted in identifying which genetic resources are already stored and duplicated in long-term facilities.

84. FAO in cooperation with countries and with relevant institutions should facilitate the formalizing of agreements to safeguard diversity in ex situ collections in conformity with applicable international agreements. This would allow those countries so desiring to place collections voluntarily in secure facilities outside their boundaries.

The Global Plan of Action has now been revised by the FAO Commission on Genetic Resources for Food and Agriculture for adoption by the FAO Council in November 2011.

The Global Crop Diversity Trust

Central to the implementation of the global system of ex situ conservation will be the Global Crop Diversity Trust (hereinafter “the Trust”), an international endowment fund set up by international agreement in 2004 and recognized by the Governing Body of the Treaty at its first session in 2006 as being an essential element of the funding strategy of the Treaty in relation to the ex situ conservation and availability of PGRFA. The objective of the Trust is to ensure the long-term conservation and availability of PGRFA with a view to achieving global food security and sustainable agriculture and, in doing so, to promote an efficient, goal-oriented, economically efficient and sustainable global system of ex situ conservation in accordance with the Treaty and the Global Plan of Action. For this purpose, it runs an endowment fund to secure the long-term financing of eligible collections of PGRFA. At present, the endowment fund stands at some US$117 million, and long-term grants are made at the rate of over US$2.3 million a year.

New developments in technical knowledge on ex situ conservation

Technical knowledge on ex situ conservation has advanced considerably since the publication of the first Technical Guidelines in 1995. Much progress has been made in the development of techniques for in vitro storage of germplasm, which has become standard practice for many species in genebanks. In vitro samples are useful for conservation of genetic resources that are (1) vegetatively propagated or have short-lived recalcitrant seeds, (2) rare or endangered or (3) the products of biotechnology, such as elite desirable genotypes maintained as clones. In addition to their utility for conservation in active and base collections, in vitro techniques are useful for a number of other purposes, including rapid multiplication, production of disease-free material, safety duplication and distribution.

New developments in the area of plant breeders’ rights

International treaties related to plant breeders’ rights were introduced in the first version of the Technical Guidelines. As of April 2011, 69 states were members of the International Convention for the Protection of New Varieties of Plants (UPOV), with different members having acceded to different acts. Over 150 states are now obligated to comply with the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPs) as a result of their membership in the World Trade Organization (WTO). The World Intellectual Property Rights Organization (WIPO) established a committee in 2000 to discuss the relationship between intellectual property, genetic resources, traditional knowledge and expressions of folklore. The WIPO Intergovernmental Committee on Intellectual Property and Genetic Resources, Traditional Knowledge and Folklore (IGC) is currently working on an international legal instrument to protect genetic resources, traditional knowledge and traditional cultural expressions.

Linkages and possible inconsistencies between the international treaties protecting intellectual property and the regimes established by the CBD have been discussed in several fora. It is uncertain what impact the stronger intellectual property protections mandated by TRIPs, UPOV and WIPO will have on the CBD objective of fair and equitable sharing of the benefits arising from the use of biological genetic resources. One measure that has been suggested as a way to support the principles of the CBD through intellectual property systems is a requirement for the origin of genetic materials used in inventions to be disclosed. The WIPO IGC has considered the relationship between access to genetic resources and the need for disclosing in patent applications the origin of genetic resources and traditional knowledge, but no legally binding obligations have been approved. Thus far, the UPOV Council has asserted that the CBD and UPOV are already mutually supportive and has resisted including the disclosure of origin as an additional condition for protection.

The most discussion on the issue of disclosure has taken place in the TRIPs Council, which has discussed many times the relationship between TRIPs and the CBD, and the need to amend TRIPs to support compliance with the CBD’s access and benefit-sharing obligations. Recent negotiations in the TRIPs Council have focused on whether TRIPs should include a requirement for disclosure (in patent or other types of applications protecting intellectual property rights) of the origin of genetic resources and traditional knowledge, possibly including proof of prior informed consent and benefit-sharing with the source. Members opposing this change take the position that contracts between countries and national access and benefit-sharing systems are already adequate to prevent misappropriation of genetic resources. An international certificate of origin has been proposed as one means of providing proof of compliance with national legislations on access and benefit-sharing that could be used to provide disclosure in applications for intellectual property rights. The International Regime on Access and Benefit Sharing outlined in the Nagoya Protocol lacks a mandatory requirement for disclosure in patent applications. Instead, it has a general description of checkpoints that are to be used to monitor compliance with national laws related to access and benefit-sharing, which has reinforced the push by some countries to add the disclosure requirement to TRIPs.

 

Future challenges/needs/gaps

What is now required is consolidation and practical implementation of the legal environment governing the collection of plant germplasm. It is especially important to clarify the way the regimes established by the CBD and the Treaty will operate with respect to the collection of PGRFA from in situ conditions. The interrelationship between Article 12.5.(h) of the Treaty and the rest of the provisions relating to the multilateral system and the development of standards for plant collecting and transfer will be of particular importance in this respect. The development of model contractual clauses (including clauses covering benefit-sharing arising from the utilization of genetic resources and traditional knowledge held by indigenous and local communities) will also be important.

 

Conclusions

Many significant developments have taken place in the legal environment governing plant germplasm collecting since the first edition of the Technical Guidelines in 1995:

• the entry into force of the Treaty, which established a multilateral system of access and benefit-sharing for the plant genetic resources of crops and forages listed in Annex 1 (chosen on the basis of their importance for food security and the degree to which countries are dependent on them)
• the adoption of the Nagoya Protocol on Access and Benefit-sharing

At present, the legal environment governing plant germplasm collecting is a patchwork of provisions from the CBD, the Nagoya Protocol and the Treaty. A great deal of work will be needed to ensure that these regimes are implemented in a harmonious and cooperative manner to ensure a proper basis for food security and sustainable agriculture.

Meanwhile, as a practical approach, it is suggested that collectors should be careful to ensure that prior informed consent is always sought from the country where collecting missions are to take place and that the conditions set by the country for the collecting mission are scrupulously adhered to. Collectors are referred to the box, which provides practical guidance for seeking prior informed consent for collecting missions.

Back to list of chapters on collecting

 

References and further reading

FAO. 2010a. First Meeting of the Ad Hoc Advisory Technical Committee on the Standard Material Transfer Agreement and the Multilateral System of the Treaty. IT/AC-SMTA-MLS 1/10/Report. International Treaty on Plant Genetic Resources for Food and Agriculture, Food and Agriculture Organization of the United Nations, Rome. Available online (accessed 30 September 2011): ftp://ftp.fao.org/ag/agp/planttreaty/gb4/AC_SMTA_MLS1/ac_smta_mls1_repe.pdf.

FAO. 2010b. Second Meeting of the Ad Hoc Advisory Technical Committee on the Standard Material Transfer Agreement and the Multilateral System of the Treaty. IT/AC-SMTA-MLS-2/10/Report. International Treaty on Plant Genetic Resources for Food and Agriculture. Food and Agriculture Organization of the United Nations, Rome. Available online (accessed 30 October 2011): www.planttreaty.org/sites/default/files/ac_smta_mls2_repe.pdf.

Moore G, Tymowski W. 2005. Explanatory Guide to the International Treaty on Plant Genetic Resources. IUCN Environmental Policy and Law Paper No. 57. IUCN–The World Conservation Union. Gland, Switzerland. Available online (accessed 3 October 2011): www.iucn.org/dbtw-wpd/edocs/EPLP-057.pdf.

 

Internet resources

Bonn Guidelines on Access to Genetic Resources and Fair and Equitable Sharing of the Benefits Arising out of Their Utilization: www.cbd.int/abs/bonn

Explanatory Guide to the International Treaty on Plant Genetic Resources: www.iucn.org/dbtw-wpd/edocs/EPLP-057.pdf

International Treaty on Plant Genetic Resources for Food and Agriculture, full text: www.planttreaty.org/texts_en.htm

International Treaty on Plant Genetic Resources for Food and Agriculture, SMTA: www.planttreaty.org/smta_en.htm

National Focal Points of the Convention on Biological Diversity: www.cbd.int/information/nfp.shtml

The Global Plan of Action for the Conservation and Sustainable Utilization of Plant Genetic Resources for Food and Agriculture: http://typo3.fao.org/fileadmin/templates/agphome/documents/PGR/GPA/gpaeng.pdf

The Nagoya Protocol on Access to Genetic Resources and the Fair and Equitable Sharing of Benefits Arising from their Utilization: www.cbd.int/abs/text

 

Back to top

Chapter 6: Strategies for the collecting of wild species

Mauricio Parra-Quijano
Departamento de Biología Vegetal Universidad Politécnica de Madrid, Ciudad Universitaria Madrid, Spain
E-mail: mauricio.parra(at)agrobiodiversidad.org
José María Iriondo
Área de Biodiversidad y Conservación, Departamento de Biología y Geología Universidad Rey Juan Carlos, Madrid, Spain
E-mail: jose.iriondo(at)urjc.es
Elena Torres Lamas
Departamento de Biología Vegetal Universidad Politécnica de Madrid, Ciudad Universitaria Madrid, Spain
E-mail: elena.torres(at)upm.es

	2011 version (0.5 MB)	1995 version (0.5 MB)
	Open the full chapter in PDF format by clicking on the icon above.

This chapter is a synthesis of new knowledge, procedures, best practices and references for collecting plant diversity since the publication of the 1995 volume Collecting Plant Diversity: Technical Guidelines, edited by Luigi Guarino, V. Ramanatha Rao and Robert Reid, and published by CAB International on behalf of the International Plant Genetic Resources Institute (IPGRI) (now Bioversity International), the Food and Agriculture Organization of the United Nations (FAO), the World Conservation Union (IUCN) and the United Nations Environment Programme (UNEP). The original text for Chapter 6: Strategies for the Collecting of Wild Species, authored by R. von Bothmer and O. Seberg, has been made available online courtesy of CABI. The 2011 update of the Technical Guidelines, edited by L. Guarino, V. Ramanatha Rao and E. Goldberg, has been made available courtesy of Bioversity International.

Please send any comments on this chapter using the Comments feature at the bottom of this page. If you wish to contribute new content or references on the subject please do so here.

Back to list of chapters on collecting

References for this chapter

Internet resources for this chapter

 

Abstract

	Collecting seeds from a population of Arabidopsis thaliana growing among a stand of Lupinus hispanicus subsp. bicolor, Barco de Avila, Spain, 2004. (Photo: J.B. Beck/Wichita State University)

Crop wild relatives (CWR) have received increasing interest from the plant genetic resources community since the first Global Plan of Action in 1996 called attention on its poor representation in germplasm collections around the world. Collecting has become a strategic activity to improve the ex situ representation of wild germplasm, but improvements are now more focused on qualitative issues (i.e., introducing novel genetic diversity, useful traits, environmental adaptations) than being simply quantitative (i.e., aimed only at the number of accessions). Since the publication of these Technical Guidelines in 1995, there have been some significant technical and methodological advances in novel CWR collecting and evaluation of past collecting efforts.

In the last decade, several tools and methodologies that have been successfully tested in related disciplines (such as plant biology, ecology and biodiversity conservation) have been gradually introduced and/or popularized in plant genetic resources, such as the use of geographic information systems, ecogeographical analysis or species distribution models. Thus, current scientific knowledge and technical capabilities have allowed collecting missions for CWR following research or breeding interests to be carried out with “surgical” precision. Collecting for genetic diversity and/or conservation is the area where the new advances have been used more intensively, allowing for the design of effective and efficient collecting missions, the detection of collection biases in existing germplasm collections, the assessment of their representativeness and, therefore, the need for additional germplasm collecting.

The original version of this chapter (authored by von Bothmer and Seberg) provides an adequate framework for the integration and discussion of these new advances, and the identification of future challenges. This update presents the most significant contributions on CWR collecting that have taken place in the last 15 years. However, instead of keeping the focus on the tribe Triticeae of the Poaceae family as in the original chapter, it provides examples of a broad taxonomic range.

 

Introduction

Wild plant species, in a broad sense, are a relatively modern target in the history of plant genetic resources for food and agriculture (PGRFA). However, for those interested in the conservation and use of agricultural biodiversity, only some of the wild species that are genetically related to domesticated or cultivated species deserve major attention. They are referred as crop wild relatives (CWR) and include progenitors and other species belonging to the primary, secondary and tertiary genepools of crop species (Maxted et al. 2008b).

CWR are a valuable source of traits for resistance to and tolerance of biotic and abiotic stresses – traits that contributed to increases in yield and quality of about USD$350 million per year in the USA alone in the mid-1980s (Prescott-Allen and Prescott-Allen 1986). It can be assumed that this value has further increased in the last decades as we headed towards the current scenarios of global change. CWR became a matter of concern at a global scale when, at the end of the 1990s, the scientific community circulated several warnings about the risk of extinction of many CWR species and the consequent loss of useful genetic material. FAO’s Global Plan of Action (1996) provided one of the most important calls for attention to the low representation of CWR in ex situ germplasm collections and the need to collect them. Consequently, many organizations interested in PGRFA conservation have focused their efforts on improving the representation of CWR species in genebanks.

As the Second Report on the State of the World’s Plant Genetic Resources for Food and Agriculture (SoWPGR-2) notes, 240,000 new accessions were collected over the period 1996–2007, and approximately 40% of them were CWR (FAO 2010). This means that CWR were the predominant type of germplasm collected over the last 15 years. At the same time, SoWPGR-2 underlined the fact that most CWR germplasm was collected by crop genebanks as a matter of local or national interest. Some botanical gardens also operate seed banks and, therefore, collect seeds of wild plants (not only CWR), but there is no global data (on number of species and accessions) available about their conservation efforts. In addition, botanical gardens focus their attention mainly on threatened, rare or endemic wild plants (Bacchetta et al. 2008) and tend to represent species richness more than intraspecific genetic variation (Maunder et al. 2004).

Despite the increasing importance of CWR, information on CWR collecting is scarce and poorly represented in the scientific media. Scientific reports about collecting missions of plant genetic resources (PGR) are often undervalued and published just as “short communications” in most cases. Furthermore, most local/national initiatives involved in the collecting of CWR germplasm do not produce easily accessible outputs, and the results remain as grey literature or internal reports. As a result, CWR collecting expeditions are difficult to follow and quantify.

Fortunately, unlike the reports of collecting missions, some relevant collecting methodologies for PGR have been published in recent years, many of them focused on CWR. New, available methodologies are the result of previous joint developments, including recent advances in geographical information system (GIS) tools and data accessibility and sharing (through the internet). Progress in this area comes from individual and collaborative efforts involved in projects on CWR conservation (Heywood et al. 2007).

Most new advances on CWR collecting have considered representativeness as a main goal. In genetic terms, representativeness in ex situ conservation refers to the attempt to capture the genetic variability of the species to the greatest extent possible. However, until a complete view of the existing genetic variability across species is available, it is hardly possible to determine gaps in genetic representativeness (GR). The assessment of the genetic variability of all genebank samples of a particular species might be affordable but it would be difficult to achieve this for all the populations of the species. For this reason, most advances have focused on assessing ecogeographical representativeness (ER) as a surrogate of GR. Because there is a link between genotype or phenotype and environmental conditions (because of the selective pressures and local adaptation processes that affect wild populations) the use of ecogeographical data as a surrogate of genetic data can be justified. These relationships have been addressed by several authors, such as Steiner (1999), and supported by a constant number of publications over time (see, for example, the review for Triticum wild relative species by Nevo [2011]). In addition to this, the advantages of using ER are linked to the low information requirements about the target species’ populations. Georeferenced collecting sites are the baseline of any ecogeographical study in PGR, and geographic coordinates are the most appropriate data to define unequivocally any collecting site (Hijmans et al. 1999). When collecting is carried out following ER parameters, the principle is simple: sampling CWR of distinct ecogeographical areas should ensure sampling of representative intraspecific genetic variation (Greene and Hart 1999; Jones et al. 1997; Parra-Quijano et al. 2008; Steiner and Greene 1996).

Collecting CWR following representativeness criteria is obviously carried out within the framework of “collecting for genetic diversity study and conservation”, as established in the original chapter in 1995. For other collecting alternatives (i.e., collecting for specific use in breeding programmes or for taxonomic/phylogenetic/biosystematics research), the most significant innovations also come from similar methodologies and tools (GIS, ecogeographical approach or species distribution models).

 

	Collecting wild relatives of a legume (Lupinus spp) at the edge of a cereal field in Cadiz, Spain. Collecting design was optimized and location data loaded into the GPS unit the collector is holding. (Photo: M.P. Quijano)

Current Status

In this section, the most remarkable advances in CWR collecting methods published in the last 15 years are briefly described, following the structure of the original 1995 chapter.

Collecting for taxonomic, phylogenetic and biosystematics research

There are few publications on expeditions for CWR collecting following these interests, although some remarkable examples can be cited, such as the continued efforts on collecting wild relatives of Solanum tuberosum by Spooner et al. (1992, 1994, 2000). These articles describe exploration criteria, routes, taxonomic and/or phylogenetic interests and collecting results in Ecuador, Bolivia and Mexico in detail.

On the other hand, van Treuren et al. (2011) nicely showed how to use passport data from PGR documentation systems and other sources to determine collection gaps in crop and wild leafy vegetable species based on the genepool structure of the target crop species.
Collecting for taxonomic research frequently demands drawing the distribution of known species and assessing species richness in the territory of study. In this sense, Hijmans and Spooner (2001) established the geographic distribution of about 199 wild potato species and determined the areas with the highest species richness. This work was supported by GIS tools. Similarly, Sawkins et al. (1999) analysed the case of some Stylosanthes species, where, in addition to using GIS tools, they introduced a predictive distribution model (based on Principal Component Analysis) to estimate the potential distributions of these taxa.

In other cases, the objective (besides mapping species distribution) has been to identify environmental adaptations and/or climatic ranges for sets of CWR species. Segura et al. (2003) linked the distributions of five Passiflora species to several temperature and rainfall variables through GIS tools and species distribution models. With this information about species adaptation, they planned ex situ and in situ conservation strategies. Similarly, Ferguson et al. (2005) used GIS tools, species distribution models and climatic variables to study the biogeography of 69 wild Arachis species. These kinds of studies are also helpful for meeting the objectives of the types of collecting interests that are presented next.

Collecting for genetic diversity study and conservation

“Capture of genetic diversity” commonly refers to between- and within-population sampling procedures within a species. Molecular methods for characterizing genetic diversity developed enormously since 1995, and techniques such as ISSRs, microsatellites or AFLPs, which were not mentioned in the 1995 version of this chapter, have been extensively applied (Spooner et al. 2005). Nevertheless, at this level, collecting strategies in wild species is still a matter of determining the size of the sample per site (population/subpopulation) and the number of sites that need to be sampled to ensure appropriate allelic retention (Brown and Briggs 1991). Sampling theory on collecting PGR was extensively detailed by Brown and Marshall (1995) in chapter 5 of "Collecting Plant Genetic Diversity: Technical Guidelines", as well as by von Bothmer and Seberg in the 1995 version of this chapter. Since then, there have been few new developments. For instance, optimum sample size has been documented for some species, e.g., wild populations of Phaseolus lunatus (Zoro Bi et al. 1998) or Zizania texana (Richards et al. 2007), while other studies have explored different models to determine minimum sample size under specific conditions (e.g., ploidy level, matting system or multiple loci) (Sapra et al. 1998; Lockwood et al. 2007).

In spite of the extraordinary advances of molecular methods for characterizing genetic diversity, when it comes to planning the collection of seeds of target CWR for PGR conservation, the norm is that there are no previous genetic diversity data published on the target species or that published data are inadequate. Therefore, in most cases, it is not possible to rely on such data to design the sampling strategy. Even if the genetic diversity of the species has been previously studied, it should be noted that neutral genetic variation, as characterized using current widespread molecular techniques, does not necessarily correlate with adaptive genetic variation, which is the type of genetic variation that is of greatest interest for breeding purposes. In this regard, the studies performed by Storfer (1996), Pfrender et al. (2000) and Bekessy et al. (2003), among others, show that currently used molecular markers are not associated with genes of adaptive value and that the correlation between the indicators of diversity obtained in the two types of genes is usually low.

Concerning the use of ecogeographical data as a proxy for data on genetic diversity, in chapter 15 of the 1995 edition of "Collecting Plant Genetic Diversity: Technical Guidelines", Guarino (1995b) mentioned the potential of GIS tools in PGR collecting activities for the first time, and Greene et al. (1999a, 1999b) described the first case of the use of GIS tools in sampling intraspecific variation in CWR species of Trifolium, Lotus and Medicago. Afonin and Greene (1999) published a detailed description of the use of GIS in the design of CWR collecting. They focused their attention on the possible uses of GIS to (1) determine where to search through the use of predictive models, (2) determine where to search to collect germplasm adapted to abiotic stresses, (3) define a sampling framework and (4) assess the anthropogenic influence on target sampling areas.

Another important advance, also based on GIS techniques, was presented by Hijmans et al. (2000), concerning the identification of collecting biases in wild potato samples from Bolivia conserved in the world’s six main wild potato genebanks. Four types of bias were described: species, species-area, hotspot and infrastructure bias. Knowledge about the type and magnitude of collecting biases is relevant information that should always be taken into account when improving the representativeness of any genebank and designing new collecting expeditions.

Gap analysis is a comparative method that can be used to improve biodiversity conservation (Maxted et al. 2008a) and, specifically, to improve the efficiency in CWR collecting activities. Recently, Ramírez-Villegas et al. (2010) described the use of gap analysis to prioritize geographic areas for collecting wild Phaseolus species. This methodology has been used to determine “potential” gaps in several CWR species (see http://gisweb.ciat.cgiar.org/GapAnalysis). Whale potential gaps come from comparisons between collecting sites of germplasm that is already conserved (genebank data) and species distribution models, a “classic” gap analysis compares genebank collecting sites with data on species occurrence from other sources, such as herbaria, botanical databases and other types of presence data (see Parra-Quijano et al. 2008). Another example of the use of species distribution models in collecting CWR can be found in the case of Vasconcellea in Venezuela (Rodríguez et al. 2005).

The basis of ER studies in CWR ex situ germplasm collections is described by Parra-Quijano et al. (2008). The ER studies use gap-analysis techniques and ecogeographical characterization data to determine which areas should be explored to find and collect populations that grow in underrepresented environments. Ecogeographical characterization of the collection sites (Steiner and Greene 1996) or the territory where the target taxon occurs (Parra-Quijano et al. 2011a) is essential to ER analysis in genebanks. Ecogeographical variables are the first element to be considered. There are currently hundreds of ecogeographical variables available (in GIS layer format) for most countries, but only some of them might have a significant influence on the abiotic adaptation of the target species. Thus, Bennett and Bullita (2003) determined the most influential ecogeographical variables in six species of Trifolium collected in Sardinia.

Based on these premises, an optimized PGR collecting method has been described by Parra-Quijano et al. (2011b). This method was developed and validated in six Lupinus species (five CWR and one cultivated) in Spain. Optimized collection attempts to make collecting activities and field exploration as efficient as possible, combining spatial and ecogeographical gap analysis, species distribution models and GIS tools. Thus, optimized collecting designs will guide explorers in finding and collecting populations in underrepresented environments (from spatial and ecogeographical points of view). At this point, two possible cases of optimized collections that follow representativeness criteria can be identified: (1) collecting a CWR species that is already represented in the genebank and (2) collecting a new species that has no previous accessions in the genebank. The Lupinus example mentioned above belongs to the first case, since the objective there was to improve the ER of the Spanish Lupinus collection held by the Plant Genetic Resources Centre of the National Institute of Agriculture and Food Research (CRF-INIA). Based on similar ER principles, Ghamkhar et al. (2007) determined ecogeographical regions and subregions within the Mediterranean and Irano-Turanian ecoregions related to Trifolium spumosum adaptations and then identified areas to explore to fill gaps in the collection of the Australian Trifolium Genetic Resources Centre (ATGRC).

Some CWR species may stand out because of their rarity or threat degree, and seed collecting may be called for as part of the effort to conserve the species. This is the case of Capsicum flexuosum, a rare and endangered species that has low representation in Capsicum genebanks. In this context, Jarvis et al. (2005) developed a successful collecting strategy for sampling C. flexuosum in some areas of Paraguay, based on GIS and species distribution models.

Collecting for immediate use in a breeding programme

Reports for this type of collecting are scarce and difficult to find. This is in part due to the fact that breeding programmes are often conducted by private companies, which are zealous about not disclosing their activities and sources of germplasm. However, it is important to note that collecting expeditions that try to capture genetic diversity are frequently biased towards breeders’ interests since the ultimate goal of PGR conservation is the utilization of its resources.

The most novel strategy for collecting CWR for immediate use in a breeding programme is emerging from an extension of the focused identification germplasm strategy (FIGS) that originated as a way of selecting subsets of germplasm from genetic resource collections in such a way as to maximize the likelihood of capturing a specific trait at a higher frequency than if the subset were selected at random. This approach was initially presented as a core collection sampling strategy (Mackay 1986, 1990), but later, it was further developed (Mackay and Street 2004) to meet the needs of a plant breeder approaching a genebank for sources of genetic diversity for crop breeding programmes aimed at a specific trait. If we look at CWR populations in their natural habitats, the FIGS approach can be used to select the populations where seeds should be collected in order to maximize the likelihood of capturing the desired trait. In essence, this method can guide field explorations to populations exposed to specific environmental stresses of interest to breeders. This methodology is currently being tested in the context of the PGR Secure project (www.pgrsecure.org).

 

Future challenges/needs/gaps

Future predictable advances and their consequences in regard to CWR collecting activities can be listed as follows:

1. Presence data on CWR species will become more abundant, detailed and easily available, particularly data from hotspot areas and centres of crop origins and diversity. They will be useful for performing more accurate CWR gap analyses and for implementingoptimized collecting activities more efficiently.

2. Ecogeographical data in GIS layer format will improve in quality and resolution, and will be more easily available. This will allow ecogeographical representativeness studies to be performed – even at the within-population level – in the coming years.

3. Optimized methods of spatial interpolation of genetic diversity data will allow this information to be used as conventional GIS layers and will help improve the collection of genetic diversity in the field.

4. The development and popularization of non-neutral genetic markers linked to adaptive traits will enable a direct measure of the type of genetic variation that is sought for conservation in genebanks and use in breeding programmes. This will allow a much better assessment of the genetic diversity already stored in genebanks and will facilitate the design of more-efficient seed collecting strategies. It will also help calibrate the link with ecogeographical variables.

 

Conclusions

A significant change since the Technical Guidelines were published in 1995 is that CWR species conservation is now one of the highest priorities in plant genetic resources. Fifteen years ago, CWR was barely a rising topic. Several reasons for this huge change can be argued: the need for novel genes, the need of crop adaptation to extreme conditions, the breakdown of barriers to introgression, the increasing pressure on wild species populations brought about by global changes, and a growing awareness of the false sense of getting the job done with cultivated species. Although the effective conservation of CWR is the goal, lessons from cultivated plants have led to a focus on how to collect the most representative and useful germplasm of wild species.

Wild plant species are a target shared with other disciplines, such as biological conservation or plant ecology, so some of their tested tools have been gradually transferred to make possible the development of efficient methodologies for collecting CWR. As a result, a new range of methodological approaches based on ecogeographical data and plant adaptation is now available. GIS tools have provided a solid vehicle for facilitating the implementation and popularization of these methodologies. Their potential role in collecting plant genetic resources was highlighted by Guarino (1995a) in chapter 16 of these Technical Guidelines. Although genotypic and phenotypic approaches have also undergone significant changes (i.e., development of genomics, proteomics, new molecular markers, etc.), their effects have not been highly influential in the development of new strategies for collecting wild species.

Back to list of chapters on collecting

 

References and further reading

Afonin A, Greene SL. 1999. Germplasm collecting using modern geographic information technologies: directions explored by the N.I. Vavilov Institute of Plant Industry. In: Greene SL, Guarino L, editors. Linking Genetic Resources and Geography: Emerging Strategies for Conserving and Using Crop Biodiversity. CSSA Special Publication 27. American Society of Agronomy, Madison, Wisconsin. pp.75–85.

Bacchetta G, Bueno A, Fenu G, Jiménez-Alfaro B, Mattana E, Piotto B, Virevaire M. 2008. Conservación Ex Situ De Plantas Silvestres. Gobierno del Principado de Asturias, Obra Social La Caixa, Oviedo, Spain.

Bekessy SA, Ennos RA, Burgman MA, Newton AC, Ades PK. 2003. Neutral DNA markers fail to detect genetic divergence in an ecologically important trait. Biological Conservation 110(2):267–275.

Bennett SJ, Bullitta S. 2003. Ecogeographical analysis of the distribution of six Trifolium species in Sardinia. Biodiversity and Conservation 12(7):1455–1466.

Brown AHD, Briggs JD. 1991. Sampling strategies for genetic variation in ex situ collections of endangered plant species. In: Falk DA, Holsinger KE, editors. Genetics and Conservation of Rare Plants. Oxford University Press, New York. pp.99–119.

Brown AHD, Marshall DR. 1995. A basic sampling strategy: theory and practice. In: Guarino L, Ramanatha Rao V, Reid R, editors. Collecting Plant Genetic Diversity: Technical Guidelines. CAB International, Wallingford, UK. pp.75–91.

FAO. 1996. Global Plan of Action for the Conservation and Sustainable Utilization of Plant Genetic Resources for Food and Agriculture and the Leipzig Declaration. Food and Agriculture Organization of the United Nations, Rome. Available online (accessed 28 October 2011): http://typo3.fao.org/fileadmin/templates/agphome/documents/PGR/GPA/gpaeng.pdf.

FAO. 2010. The Second Report on the State of the World's Plant Genetic Resources for Food and Agriculture. Food and Agriculture Organization of the United Nations, Rome. Available online (accessed 28 October 2011): www.fao.org/agriculture/crops/core-themes/theme/seeds-pgr/sow/sow2/en.

Ferguson ME, Jarvis A, Stalker HT, Williams DE, Guarino L, Valls JFM, Pittman RN, Simpson CE, Bramel PJ. 2005. Biogeography of wild Arachis (Leguminosae): distribution and environmental characterisation. Biodiversity and Conservation 14(7):1777–1798.

Ghamkhar K, Snowball R, Bennett SJ. 2007. Ecogeographical studies identify diversity and potential gaps in the largest germplasm collection of bladder clover (Trifolium spumosum L.). Australian Journal of Agricultural Research 58(7):728–738.

Greene SL, Hart TC. 1999. Implementing a geographic analysis in germplasm conservation. In: Greene SL, Guarino L, editors. Linking Genetic Resources and Geography: Emerging Strategies for Conserving and Using Crop Biodiversity. CSSA Special Publication 27. American Society of Agronomy, Madison, Wisconsin. pp.25–38.
Greene SL, Hart TC, Afonin A. 1999a. Using geographic information to acquire wild crop germplasm for ex situ collections: I. map development and field use. Crop Science 39(3):836–842.

Greene SL, Hart TC, Afonin A. 1999b. Using geographic information to acquire wild crop germplasm for ex situ collections: II. post-collection analysis. Crop Science 39(3):843–849.

Guarino L. 1995a. Geographic information systems and remote sensing for plant germplasm collectors. In: Guarino L, Ramanatha Rao V, Reid R, editors. Collecting Plant Genetic Diversity Technical Guidelines. CAB International, Wallingford, UK. pp.315–327.

Guarino L. 1995b. Mapping the ecogeographic distribution of biodiversity. In: Guarino L, Ramanatha Rao V, Reid R, editors. Collecting Plant Genetic Diversity Technical Guidelines. CAB International, Wallingford, UK. pp.287–327.

Heywood V, Casas A, Ford-Lloyd BV, Kell S, Maxted N. 2007. Conservation and sustainable use of crop wild relatives. Agriculture, Ecosystems & Environment 121(3):245–255.

Hijmans RJ, Spooner DM. 2001. Geographic distribution of wild potato species. American Journal of Botany 88(11):2101–2112.

Hijmans RJ, Schreuder M, De la Cruz J, Guarino L. 1999. Using GIS to check co-ordinates of genebank accessions. Genetics Resources and Crop Evolution 46(3):291–296.

Hijmans RJ, Garrett KA, Huamán Z, Zhang DP, Schreuder M, Bonierbale M. 2000. Assessing the geographic representativeness of genebank collections: the case of Bolivian wild potatoes. Conservation Biology 14(6):1755–1765.

Jarvis A, Williams K, Williams D, Guarino L, Caballero PJ, Mottram G. 2005. Use of GIS for optimizing a collecting mission for a rare wild pepper (Capsicum flexuosum Sendtn.) in Paraguay. Genetics Resources and Crop Evolution 52(6):671–682.

Jones PG, Beebe SE, Tohme J, Galwey NW. 1997. The use of geographical information systems in biodiversity exploration and conservation. Biodiversity and Conservation 6(7):947–958.

Lockwood DR, Richards CM, Volk GM. 2007. Probabilistic models for collecting genetic diversity: comparisons, caveats, and limitations. Crop Science 47(2):861–868.

Mackay MC. 1986. Utilizing wheat genetic resources in Australia. In: McLean R, editor. Proceedings of the 5th Assembly, Wheat Breeding Society of Australia. Western Australian Department of Agriculture, Perth, Australia. pp.56–61.

Mackay MC. 1990. Strategic planning for effective evaluation of plant germplasm. In: Srivastava JP, Damania AB, editors. Wheat Genetic Resources: Meeting Diverse Needs. John Wiley & Sons, Chichester, UK. pp.21–25.

Mackay MC, Street K. 2004. Focused identification of germplasm strategy – FIGS. In: Black CK, Panozzo JF, Rebetzke GJ, editors. Proceedings of the 54th Australian Cereal Chemistry Conference and the 11th Wheat Breeders' Assembly. Royal Australian Chemical Institute, Melbourne. pp.138–141.

Maunder M, Havens K, Guerrant EO Jr, Falk DA. 2004. Ex situ methods: a vital but underused set of conservation resources. In: Guerrant Jr. EO, Havens K, Maunder M, editors. Ex situ Plant Conservation. Supporting Species Survival in the Wild. Island Press, Washington, USA. pp. 3–20.

Maxted N, Dulloo E, Ford-Lloyd BV, Iriondo JM, Jarvis A. 2008a. Gap analysis: a tool for complementary genetic conservation assessment. Diversity and Distributions 14(6):1018–1030.

Maxted N, Kell SP, Ford-Lloyd BV. 2008b. Crop wild relative conservation and use: establishing the context. In: Maxted N, Ford-Lloyd BV, Kell SP, Iriondo JM, Dulloo E, Turok J, editors. Crop Wild Relative Conservation and Use. CAB International, Wallingford, UK. pp.3–30.

Nevo E. 2011. Triticum. In: Kole C, editor. Wild Crop Relatives: Genomic and Breeding Resources. Cereals. Springer-Verlag, Heidelberg, Germany. pp.407–456.

Parra-Quijano M, Draper D, Torres E, Iriondo JM. 2008. Ecogeographical representativeness in crop wild relative ex situ collections. In: Maxted N, Ford-Lloyd BV, Kell SP, Iriondo JM, Dulloo ME, Turok J, editors. Crop Wild Relative Conservation and Use. CAB International, Wallingford, UK. pp.249–273.

Parra-Quijano M, Iriondo JM, Torres E. 2011a. Ecogeographical land characterization maps as a tool for assessing plant adaptation and their implications in agrobiodiversity studies. Genetic Resources and Crop Evolution DOI 10.1007/s10722-011-9676-7.

Parra-Quijano M, Iriondo JM, Torres E. 2011b. Improving representativeness of genebank collections through species distribution models, gap analysis and ecogeographical maps. Biodiversity and Conservation DOI 10.1007/s10531-011-0167-0.

Pfrender ME, Spitze K, Hicks J, Morgan K, Latta L, Lynch M. 2000. Lack of concordance between genetic diversity estimates at the molecular and quantitative-trait levels. Conservation Genetics 1(3):263–269.

Prescott-Allen R, Prescott-Allen C. 1986. The First Resource: Wild Species in the North American Economy. Yale University Press, New Haven, Connecticut.

Ramírez-Villegas J, Khoury C, Jarvis A, Debouck DG, Guarino L. 2010. A gap analysis methodology for collecting crop genepools: a case study with Phaseolus beans. PLoS ONE 5(10):e13497. Available online (accessed 28 October 2011): http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0013497.

Richards CM, Antolin MF, Reilley A, Poole J, Walters C. 2007. Capturing genetic diversity of wild populations for ex situ conservation: Texas wild rice (Zizania texana) as a model. Genetic Resources and Crop Evolution 54(4):837–848.

Rodríguez D, Marín C, Quecan H, Ortiz R. 2005. Áreas potenciales para colectas del género Vasconcellea Badillo en Venezuela. Bioagro 17(1):3–10.

Sapra RL, Narain P, Chauhan SVS. 1998. A general model for sample size determination for collecting germplasm. Journal of Biosciences 23(5):647–652.

Sawkins MC, Maxted N, Jones PG, Smith R, Guarino L. 1999. Predicting species distributions using environmental data: case studies using Stylosanthes Sw. In: Greene SL, Guarino L, editors. Linking Genetic Resources and Geography: Emerging Strategies for Conserving and Using Crop Biodiversity. CSSA Special Publication 27. American Society of Agronomy, Madison, Wisconsin. pp.87–99.

Segura S, Coppens d'Eeckenbrugge G, López L, Grum M, Guarino L. 2003. Mapping the potential distribution of five species of Passiflora in Andean countries. Genetic Resources and Crop Evolution 50(6):555–566.

Spooner DM, Castillo R, López LJ. 1992. Ecuador, 1991 potato germplasm collecting expedition: taxonomy and new germplasm resources. Euphytica 60(3):159–169.

Spooner DM, Berg RG, García W, Ugarte ML. 1994. Bolivia potato germplasm collecting expeditions 1993, 1994: taxonomy and new germplasm resources. Euphytica 79(1-2):137–148.

Spooner DM, Rivera-Peña A, van den Berg RG, Schüler K. 2000. Potato germplasm collecting expedition to Mexico in 1997: taxonomy and new germplasm resources. American Journal of Potato Research 77(4):261–270.

Spooner D, van Treuren R, de Vicente MC. 2005. Molecular Markers for Genebank Management. International Plant Genetic Resources Institute (IPGRI), Rome.

Steiner JJ. 1999. Exploring the relationship of plant genotype and phenotype to ecogeography. In: Greene SL, Guarino L, editors. Linking Genetic Resources and Geography: Emerging Strategies for Conserving and Using Crop Biodiversity. CSSA Special Publication 27. American Society of Agronomy, Madison, Wisconsin. pp.39–50.

Steiner JJ, Greene SL. 1996. Proposed ecological descriptors and their utility for plant germplasm collections. Crop Science 36(2):439–451.

Storfer A. 1996. Quantitative genetics: a promising approach for the assessment of genetic variation in endangered species. Trends in Ecology & Evolution 11(8):343–348.

van Treuren R, Coquin P, Lohwasser U. 2011. Genetic resources collections of leafy vegetables (lettuce, spinach, chicory, artichoke, asparagus, lamb's lettuce, rhubarb and rocket salad): composition and gaps. Genetics Resources and Crop Evolution DOI 10.1007/s10722-011-9738-x.

Zoro Bi I, Maquet A, Degreef J, Wathelet B, Baudoin JP. 1998. Sample size for collecting seeds in germplasm conservation: the case of the Lima bean (Phaseolus lunatus L.). Theoretical and Applied Genetics 97(1–2):187–194.

 

Internet resources*

AEGRO website: http://aegro.jki.bund.de/aegro

Bioversity Crop Wild Relatives Global Portal: www.cropwildrelatives.org

Crop Wild Relative Information System (CWRIS): http://aegro.jki.bund.de/aegro/index.php?id=168

European Crop Wild Relative Diversity Assessment and Conservation Forum: http://pgrforum.org

FAO Facilitating Mechanism for the implementation of the Global Plan of Action, Portal for Plant Genetic Resources for Food and Agriculture: www.globalplanofaction.org

FAO Global Plan of Action (full text from1996): http://typo3.fao.org/fileadmin/templates/agphome/documents/PGR/GPA/gpaeng.pdf

FAO, The State of the World’s Plant Genetic Resources for Food and Agriculture: www.fao.org/agriculture/crops/core-themes/theme/seeds-pgr/sow/en

FAO, The Second Report on the State of the World’s Plant Genetic Resources (SoWPGR-2): www.fao.org/agriculture/crops/core-themes/theme/seeds-pgr/sow/sow2/en

GapAnalysis: http://gisweb.ciat.cgiar.org/GapAnalysis

Genetic Reserve Information System (GenResIS): www.agrobiodiversidad.org/aegro

Global Biodiversity Information Facility (GBIF): www.gbif.org

PGR Secure: www.pgrsecure.org

 *Note that additional internet resources related to this chapter can be found in the update of chapter 15/16.

 

Back to top

Chapter 5: Basic sampling strategies: theory and practice

Jose Crossa
Biometrics and Statistics Unit, International Maize and Wheat Improvement Center (CIMMYT), 06600 Mexico, DF, Mexico
E-mail: j.crossa(at)cgiar.org
Roland Vencovsky
Dept. of Genetics ESALQ/Universidade de Sao Paulo, Piracicaba, Sao Paulo, Brazil
E-mail: rvencovs(at)esalq.usp.br

	2011 version (1 MB)	1995 version (0.3 MB)
	Open the full chapter in PDF format by clicking on the icon above.

This chapter is a synthesis of new knowledge, procedures, best practices and references for collecting plant diversity since the publication of the 1995 volume Collecting Plant Diversity: Technical Guidelines, edited by Luigi Guarino, V. Ramanatha Rao and Robert Reid, and published by CAB International on behalf of the International Plant Genetic Resources Institute (IPGRI) (now Bioversity International), the Food and Agriculture Organization of the United Nations (FAO), the World Conservation Union (IUCN) and the United Nations Environment Programme (UNEP). The original text for Chapter 5: A Basic Sampling Strategy: Theory and Practice, authored by A. H. D. Brown and D. R. Marshall, has been made available online courtesy of CABI. The 2011 update of the Technical Guidelines, edited by L. Guarino, V. Ramanatha Rao and E. Goldberg, has been made available courtesy of Bioversity International.

Please send any comments on this chapter using the Comments feature at the bottom of this page. If you wish to contribute new content or references on the subject please do so here.

Back to list of chapters on collecting

References for this chapter

Internet resources for this chapter

 

Abstract

Programmes for conserving genetic resources have collected, received and stored hundreds of thousands of accessions of different cultivated species and their wild and weedy relatives. Collection and regeneration protocols must consider the species (i.e., allogamous, partially allogamous, autogamous and dioecious) to ensure that the sample is representative of the population. Previous studies have used allelic richness as the basic parameter for determining sample sizes for genetic resource conservation. The concept of variance effective population size is important to the measurement of genetic representativeness and has been successfully used in genetic conservation (regeneration and collection). The aim of this chapter is to show how to practically apply the theory developed earlier and to demonstrate its use for answering practical questions that a manager of genetic resource conservation might pose when collecting and regenerating plant genetic resources. This chapter explains strategies for determining efficient sample size in order to maintain the representativeness of the original diversity when collecting and regenerating genetic resources.

 

[Download the PDF file for full text of updated version of Chapter 5 here]

Back to list of chapters on collecting

 

References and further reading

Allard RW. 1970. Population structure and sampling methods. In: Frankel OH, Bennett E, editors. Genetic Resources in Plants: Their Exploration and Conservation. Blackwell Scientific Publications, Oxford, UK. pp.97–107.

Bandel G, Gurgel JTA. 1967. Proporção do sexo em Pinheiro Brasileiro Araucaria angustifolia (Bert.) O. Ktze. Rev. Tec. Serv. Florestal do Est. de S. Paulo. Secr. Agr. Est. S. Paulo 6: 209–220.

Barret SCH, Yakimowsky SB, Field DL, Pickup M. 2010. Ecological genetics of sex ratios in plant populations. Philosophical Transactions of the Royal Society B. 365:2549–2557.

Bawa KS. 1974. Breeding systems of tree species of a lowland tropical community. Evolution 28(1):85–92.

Bawa KS. 1980. Evolution of dioecy in flowering plants. Annual Review of Ecology, Evolution, and Systematics 11:15–39.

Bawa KS, Perry DR, Beach JH. 1985. Reproductive biology of tropical lowland rain forest. I. Sexual systems and incompatibility mechanisms. American Journal of Botany 72(3):331–345.

Cockerham CC. 1969. Variance of gene frequency. Evolution 23:72–84.

Cockerham CC, Weir BS. 1984. Covariances of relatives stemming from a population undergoing mixed self and random mating. Biometrics 40:157–164.

Crossa J, Vencovsky R. 1994. Implications of the variance effective population size on the genetic conservation of monoecious species. Theoretical and Applied Genetics 89:936–942.

Crossa J, Vencovsky R. 1997. Variance effective population size for two-stage sampling of monoecious species. Crop Science 37:14–26.

Crossa J, Vencovsky R. 1999. Sample size and variance effective population size for genetic conservation. Plant Genetic Resources Newsletter 119:15–25.

Crossa J, Hernandez CM, Bretting P, Eberhart SA, Taba S. 1993. Statistical genetic considerations for maintaining germplasm collections. Theoretical and Applied Genetics 86:673–678.

Crow JF, Denniston C. 1988. Inbreeding and variance effective numbers. Evolution 42(3):482–495.

Crow JF, Kimura M. 1970. An Introduction to Population Genetics Theory. Burgess Publishing, Minneapolis, Minnesota.

Hernandez CM, Crossa J. 1993. A program for estimating the optimum sample size for germplasm conservation. Journal of Heredity 84:1.

Marshall DR, Brown AHD. 1975. Optimum sampling strategies in genetic conservation. In: Frankel OH, Hawkes JG, editors. Crop Genetic Resources for Today and Tomorrow. Cambridge University Press, Cambridge, UK.

Matallana G, Wendt T, Araujo DSD, Scarano FR. 2005. High abundance of dioecious plants in a tropical coastal vegetation. American Journal of Botany 92(9):1513–1519.

Moran PAP. 1950. Notes on continuous stochastic phenomena. Biometrika 37:17–33.

Namkoong G. 1986. Sampling for germplasm collections. Horticultural Science 23(1):79–81.

Oliveira PE. 1996. Dioecy in the cerrado vegetation of Central Brazil. Flora 191:235–243.

Oliveira PE, Gibbs PE. 2000. Reproductive biology of woody plants in a cerrado community of Central Brazil. Flora 195:311–329.

Queenborough SA, Burslem DFRP, Garwood NC, Valencia R. 2007. Determinants of biased sex ratios and inter-sex cost of reproduction in dioecious tropical forest trees. American Journal of Botany 94(1):67–78.

Sebbenn AM. 2006. Sistema de reprodução em espécies tropicais e suas implicações para a seleção de árvores matrizes para reflorestamentos ambientais. In: Higa AR, Silva LD, editors. Pomares de Sementes de Espécies Florestais Nativas. Fundação de Pesquisas Florestais do Paraná (FUPEF), Curitiba, Brazil. pp.93–108.

Sokal RR, Oden NL. 1978a. Spatial autocorrelation in biology. 1. Methodology. Biological Journal of the Linnean Society 10:199–228.

Sokal RR, Oden NL. 1978b. Spatial autocorrelation in biology. 2. Some biological implications and four applications of evolutionary and ecological interest. Biological Journal of the Linnean Society 10:229–249.

Vencovsky R. 1978. Effective size of monoecious populations submitted to artificial selection. Brazilian Journal of Genetics 1(3):181–191.

Vencovsky R, Crossa J. 1999a. Variance effective population size under mixed self and random mating with applications to genetic conservation of species. Crop Science 39:1282–1294.

Vencovsky R, Crossa J. 1999b. Medidas de representatividad. Workshop O Melhoramento de Plantas na Virada do Milenio. Universidad Federal de Vicosa, MG, Brasil.

Vencovsky R, Crossa J. 2003. Measurements of representativeness used in genetic resource conservation and plant breeding. Crop Science 43:1912–1921. doi:10.2135/cropsci2003.1912.

Vencovsky R, Nass LL, Cordeiro CMT, Ferreira MAJF. 2007. Amostragem em recursos genéticos vegetais. In: Nass LL, editor. Recursos Genéticos Vegetais. Embrapa, Brasília.

Vencovsky R, Chavez L, Crossa J. 2011. Variance effective population size for dioecious species. Crop Science (in press).

Viegas MP, Silva CLSP, Moreira JP, Cardin LT, Azevedo VCR, Ciampi AY, Freitas MLM, Moraes MLT, Sebbenn AM. 2011. Diversidade genética e tamanho efetivo de duas populações de Myracrodruon urundeuva fr. all., sob conservação ex situ. Revista Árvore 35(4):769–779.

Weir BS. 1996. Genetic Data Analysis II - Methods for Discrete Population Genetic Data. Sinauer Associates, Sunderland, Massachusetts.

Wright S. 1931. Evolution in Mendelian populations. Genetics 16:97–159.

 

Internet resources

MLTR (Multilocos Mating System Program) by Kermit Ritlan (a computer programme for analysing marker data): http://genetics.forestry.ubc.ca/ritland/programs.html

Resources for obtaining measures of genetic divergence among collection sites or subpopulations of a species in a given ecogeographic region, and of the level of natural inbreeding:

FSTAT by Jérome Goudet: www2.unil.ch/popgen/softwares/fstat.htm

GDA (Genetic Data Analyses) by Paul O. Lewis and Dmitri Zaykin: http://hydrodictyon.eeb.uconn.edu/people/plewis/software.php

  

Back to top

Copy-Chapter 5: Basic sampling strategies: theory and practice

Jose Crossa
Biometrics and Statistics Unit International Maize and Wheat Improvement Center (CIMMYT) Apdo. Postal 6-641, 06600 Mexico, DF, Mexico
E-mail: j.crossa(at)cgiar.org
Roland Vencovsky
Dept. of Genetics ESALQ/Universidade de Sao Paulo Cx P. 83,13400-970, Piracicaba, Sao Paulo, Brazil
E-mail: rvencovs(at)esalq.usp.br

	2011 version (1.2 MB)	1995 version (0.3 MB)
	Open the full chapter in PDF format by clicking on the icon above.

This chapter is a synthesis of new knowledge, procedures, best practices and references for collecting plant diversity since the publication of the 1995 volume Collecting Plant Diversity: Technical Guidelines, edited by Luigi Guarino, V. Ramanatha Rao and Robert Reid, and published by CAB International on behalf of the International Plant Genetic Resources Institute (IPGRI) (now Bioversity International), the Food and Agriculture Organization of the United Nations (FAO), the World Conservation Union (IUCN) and the United Nations Environment Programme (UNEP). The original text for Chapter 5: A Basic Sampling Strategy: Theory and Practice, authored by A. H. D. Brown and D. R. Marshall, has been made available online courtesy of CABI. The 2011 update of the Technical Guidelines, edited by L. Guarino, V. Ramanatha Rao and E. Goldberg, has been made available courtesy of Bioversity International.

Please send any comments on this chapter using the Comments feature at the bottom of this page. If you wish to contribute new content or references on the subject please do so here.

Back to list of chapters on collecting

References for this chapter

Internet resources for this chapter

 

Abstract

Programmes for conserving genetic resources have collected, received and stored hundreds of thousands of accessions of different cultivated species and their wild and weedy relatives. Collection and regeneration protocols must consider the species (i.e., allogamous, partially allogamous, autogamous and dioecious) to ensure that the sample is representative of the population. Previous studies have used allelic richness as the basic parameter for determining sample sizes for genetic resource conservation. The concept of variance effective population size is important to the measurement of genetic representativeness and has been successfully used in genetic conservation (regeneration and collection). The aim of this chapter is to show how to practically apply the theory developed earlier and to demonstrate its use for answering practical questions that a manager of genetic resource conservation might pose when collecting and regenerating plant genetic resources. This chapter explains strategies for determining efficient sample size in order to maintain the representativeness of the original diversity when collecting and regenerating genetic resources.

 

Introduction

The great genetic complexity of most plant populations and the many possible ways that genetic resources may be used in the future makes it difficult to provide simple and efficient sampling schemes and optimal sample sizes for the maintenance of all species (Namkoong 1986). Programmes for the conservation of genetic resources and genebanks around the world have collected (in their centres of origin or elsewhere), received and stored hundreds of thousands of accessions of different cultivated species and their wild and weedy relatives. These accessions represent a wide spectrum of population diversity. Collection and regeneration protocols must consider the species (including allogamous, partially allogamous, autogamous and dioecious species, plus the type of materials being collected and regenerated) to ensure that the sample is representative of the population.

Previous studies have developed equations for collection and regeneration to be applied when using practical breeding schemes or methods for conserving genetic resources, mainly of monoecious species. The studies and research previously done on conservation of genetic resources concentrated on developing sample sizes based on maximizing allelic richness or the number of distinct alleles at a single locus (Marshall and Brown 1975).

Another important measurement of representativeness is related to how stochastic changes in allelic frequency – caused by sampling error in small populations (random genetic drift) – lead to continuous fixation and loss of alleles, and reduce the proportion of heterozygous individuals in the population (Crow and Kimura 1970; Wright 1931). These random changes in allelic frequency affect the genetic representativeness of the finite population. In a large mating population of N individuals, the reduced number of progenitors whose offspring will constitute the next generation causes random genetic drift, which is quantified and predicted using the parameter called "effective population size" (N_e). The effective population size is the size of an ideal population that has the same amount of drift in allelic frequency or the same rate of decrease in heterozygosity as the actual population. Several factors affect the effective size of a population, including the number of parents per generation, the number of corresponding offspring, the number of male and female gametes contributed per individual in the parental population, the mating system of the species, and so on. The effective population size, taken as a measure of the genetic representativeness of a seed sample, can be adapted to specific aspects of plant breeding and conservation of genetic resources, such as seed regeneration and collection.

During the process of collecting, storing, regenerating and restocking germplasm, genetic drift occurs and affects a population’s genetic integrity in a number of ways. First, when collecting germplasm in the field, an appropriate sampling strategy should be used to avoid, as much as possible, dramatically reducing the size of the population (bottleneck). Second, when accessions are stored, the different survival rates of genotypes and the accumulation of mutations affect the genetic integrity of the accessions. Third, when seed of an accession is drawn for regeneration, the sample needs to be optimum in order to avoid genetic drift due to sampling or to differential survival (germination) or fecundity, leading to changes in the genetic constitution of the accession. Fourth, when plants of accessions are being regenerated in the field, insects, diseases, and other environmental factors can affect the plant stand, thereby limiting the accessions’ gametic contribution (offspring) to the next generation. Since we do not know which alleles will prove useful in the future, it is essential that sampling be done efficiently and that populations be of sufficient size to maintain as much genetic diversity as is practical. Large samples are expensive and difficult to manage, but if samples are too small, genetic diversity may be affected by the loss of valuable alleles through random changes in allelic frequency. Maintaining allelic diversity during regeneration depends on factors such as sampling procedures, seed viability and mating systems, all of which cause random genetic drift.
Concerning the sample size for collecting germplasm, Allard (1970) pointed out that most plant species contain remarkable stores of genetic variation and consist of millions of different genotypes. Plant collectors can hope to sample only a fraction of the variation that occurs in nature. It is important that this fraction be as large as possible and contain the maximum amount of useful (now and in the future) variation. Allard (1970) also recognized that collectors as well as end-users of germplasm have limited time and resources at their disposal. Thus, the problem is to define a sampling procedure that yields the maximum amount of useful genetic variation, within a specified and limited number of samples (Marshall and Brown 1975).

The aim of this chapter is to show how to apply the theory and equations developed earlier and to demonstrate their use for answering the practical questions that a manager of genetic resource conservation might pose when collecting and regenerating plant genetic resources. It includes theoretical considerations for probability models that compute the required sample size for conserving alleles at loci and for developing strategies for efficient sample size to maintain the representativeness of the original diversity when collecting and regenerating genetic resources. It also provides theoretical concepts on the variance effective population size (N_e(v)), and describes the derivation of manageable equations for collection and regeneration of germplasm. We give practical examples of how to (1) compute the required sample size for maintaining, with a certain probability, at least one copy of each allele at independent loci in the accession to be regenerated, (2) assess practical procedures for the collection and regeneration of genetic resources that will increase the N_e(v) of random-mating populations and the N_e(v) of species with mixed self- and random-mating systems of reproduction and different natural rates of self-fertilization, and (3) examine practical collection and regeneration procedures that will increase the N_e(v) of dioecious species with different proportions of male and female plants.

 

Theory

Probability models for number of alleles in the sample

We are interested in finding a sample size n_g such that the probability of detecting at least  one allele of each allele class is greater than a quantity 1- α or P(n)>1-α (for small α). Except in the case of synthetics derived from inbred lines, in real situations, allele frequencies are unknown; therefore, some simplification is necessary. Assuming that k-1 alleles occur at an identical low frequency of p₀ and that the kth allele occurs at frequency  of 1-[(k-1)p₀], Crossa et al. (1993) showed that the sample size (n_g gametes) required to retain with a probability P=1-α that at least one copy of each of the k allelic classes in each of the m loci should be larger than

where it is assumed that loci are independent and that diploid individuals (not gametes) are sampled. The assumption of loci independence is unrealistic, since genetic drift may cause non-random associations between linked loci. If the population is under random mating (a cross-pollinated species), linkage equilibrium can be assumed for all possible pairs of loci, so it is expected that, for each pair of loci, a similar number of coupling and repulsion combinations occur. This obviously does not hold for selfpollinated species or species with mixed self- and random-mating systems of reproduction. Therefore equation 1 only offers approximate guidelines for the range of sample sizes, which are greatly dependent on the frequency of rare alleles. Furthermore, sampling ng= 2N gametes is equivalent to sampling (½)n_g= N individuals (diploid zygotes) only for panmictic populations (i.e., idealized random-mating populations of
infinite size with no association between any two genes within individuals). In this case, the number of diploid individuals that need to be sampled is exactly half the number of gametes (n_g).

It should be pointed out that the range of sample sizes given by equation 1 is based solely on probability models and does not consider the genetic structure of the population or specify how well a particular sample represents the reference population in terms of genetic parameters such as variance of allelic frequency, inbreeding, random genetic drift due to sampling error, genetic linkage, etc. Furthermore, probability models do not, by themselves, specify appropriate mating and reproduction systems such as panmixia, mixed self- and random mating, self-fertilization, and so forth, which could, under specific circumstances, maximize the genetic representativeness of the sample.

Therefore, it is clear that for an initial census of a population of size N, it is very likely that N_e < N. For studying factors that will make N_e > N, aspects of population genetics theory must be incorporated into the probability models in order to facilitate the estimation of parameters such as the mean and variance of male and female contributed gametes and the mating systems that would facilitate control of male and female gamete contributions and thus allowing to increase N_e .

Measure of representativeness: variance effective population size [N_e(v)]

The concept of effective population size N_e(v) is useful when studying the breeding structure and genetic representativeness of an actual population, as related to an ideal population in which individuals mate at random with no variation in fertility: the number of progeny per parent has a binomial distribution (which approaches Poisson for large sample sizes) and its size is kept constant through time (Wright 1931). The effective population size determines the amount of sampling error between generations that causes random fluctuations in allelic frequency. On the other hand, natural and breeding populations fluctuate in size from one generation to another, and some individuals might not produce gametes or might have different natural inbreeding levels and different natural self-fertilization rates than others. Thus, measures of effective size often deviate from the size of the actual population.

Crow and Kimura (1970) derived models for studying the genetics of finite populations and assessed the relationship between the actual size of a population, N, and the distribution of progeny among parents by estimating its mean and variance. They derived an expression for the variance effective population size (N_e(v)) when a random sample of individuals is taken from an original population of size N, assuming that all the individuals potentially contribute male and female gametes,

where α measures the departure from Hardy Weinberg equilibrium in generation t -1,

; n is the number of offspring in generation t (or total number of seeds collected from the parents); and k is the number of gametes contributed by a particular parent. This can be considered a binomial random variable in the case of a random sampling of gametes with no variation in fertility, where

are the mean and variance of the gametes contributed by the parents, respectively.

For large N,

Crow and Denniston (1988) provided expressions of the inbreeding and the variance effective number that are general for populations under random mating, and presented formulas as functions of means, variances and covariances of the number of gametes contributed by parents to the offspring.

Genetic models based on the number of male and female gametes contributed by individuals of monoecious plant species have been developed for variance effective population size applied in the context of artificial selection (Vencovsky 1978) and to specific aspects of genetic resource conservation (i.e., collection and regeneration) (Crossa and Vencovsky 1994; Vencovsky 1978). Later, Crossa and Vencovsky (1997, 1999) and Vencovsky and Crossa (1999a, 1999b, 2003) showed the theoretical developments and practical applications of variance effective population size when drift occurs at two stages: when sampling parents for reproduction and when sampling gametes (offspring) from those parents for monoecious populations and for populations under mixed self- and random mating.

Two-stage sampling model for random-mating species

Crow and Kimura (1970) assumed that all N individuals in the parental generation potentially contribute gametes to the next generation. However, in real-life situations of plant breeding, regeneration and collection of plant genetic resources, some plants may fail to produce gametes (offspring) due to external factors such as poor seed viability or germination, insects and diseases in crossing blocks or the systematic exclusion of a fraction of parents, etc. Therefore, a more realistic model would consider that the overall effect of sampling on allelic frequency drift is the summation of drift occurring at two stages: (1) when sampling parents from the original population (first stage) and (2) when sampling gametes from these parents (second stage).

Vencovsky (1978) first proposed models for computing V_(k) and N_e(v) for the two-stage sampling model for monoecious species that considers that some plants might not contribute gametes to the next generation. Crossa and Vencovsky (1994) provided approximate equations for computing V_(k) and N_e(v) for germplasm collection and regeneration with and without control of contributed male and female gametes. Crossa and Vencovsky (1997) provided the theoretical development of the two-stage model, gave an alternative derivation based solely on the theory of random sampling within a finite population, and demonstrated practical procedures for seed collection and regeneration.

Consider an initial set of N diploid monoecious plants (reference population), where P plants contributing male and female gametes are randomly selected (0 < P ≤ N). From the remaining N-P plants, R contributing only male gametes are also sampled at random [0≤ R ≤ (N-P)]. Therefore, M =P+R plants contribute male gametes, P plants contribute female gametes, and N-M plants do not contribute any gametes at all. Thus, the proportion of functional seed parents is u=P/N (or P =uN) (where 0< u ≤ 1) and the proportion of functional pollen parents is v= (P+R)/N =M/N (or M =vN) (0 < v ≤ 1). With these proportions, it is possible to simulate several important practical situations that arise when regenerating and collecting genetic resources.

For example, the condition given by Crow and Kimura (1970) when developing equation 2 (i.e., that all N individuals potentially produce male and female gametes) is achieved by considering P=M=N or u=v=1, that is, a perfect stand of plants in the field. For seed regeneration, this means that all plants in the field are used as male and female parents. On the other hand, if P<N, M<N such that u<1, v<1, some plants fail to produce gametes due to poor germination, poor seed viability or a poor stand of plants in the field because of insects, diseases or other environmental factors.

In seed collection activities, parameters u and v are fractions of seed and pollen parents, respectively, that effectively contribute gametes for generating the sample of seeds collected. They should be measured in relation to the size of the reference population under natural conditions. For monoecious species, Crossa and Vencovsky (1994) derived the following N_e(v) expressions for different alternative sampling schemes of female and male gametes. As shown in Annex A, we considered three different cases, depending on how male and female gametic control is performed.

• Case 1: Plant-to-plant hand pollination is practiced and equal numbers of seeds are taken from each plant (female plus male gametic control; FGC + MGC scheme) (equation 3).

• Case 2: Pollination is random and the same number of seeds is taken from each seed parent (female gametic control: FGC scheme) (equation 4).

• Case 3: Pollination is random and n seeds are randomly sampled from a bulk of seeds stemming from P seed parents (RS schemes) and M pollen parents (equation 5).

For seed collection, it can be assumed for random-mating species that the number of pollen parents (M) is very large; then potentially M=N and v=1. This over-simplification is used for calculating an upper limit for N_e(v). Also, the number of seed parents can be considered to be much smaller than N (P<<N) and u=0. Thus, assuming that M is very large, M=N, and v=1, the approximation to N_e(v) is given by equations 6 and 7. Also, annex A shows how equation 6 can be written when the average number of seeds per seed parent is given by n/P

Two-stage sampling model for mixed self- and random-mating species

Many self-compatible species have a high natural rate of self-fertilization (s) and thus may be considered as having a mixed self- and random-mating reproduction system. Logically, the genetic structure of a mixed self- and random-mating species is complex because not all plants have the same level of natural inbreeding. Furthermore, the progeny of a population of a mixed self- and random-mating species will have a mixture of selfed seed with proportion s and half-sibs or outcrossed seeds with proportion 1- s. This model excludes biparental crosses between parents. Proportion s can be artificially manipulated by hand-pollination so that it becomes s = 1 when selfing all plants or s = 0 when crossing all plants. However, s varies from 0 to 1 in natural populations, which are assumed to be in inbreeding equilibrium when they reproduce naturally. In this case, f = s/(2-s) [or s = 2f/(1+f)]. Vencovsky and Crossa (1999a) extended the two-stage sampling models for random-mating species to include mixed self- and random-mating species (0 ≤ s ≤ 1). They also developed estimates and direct expressions for computing V_(k) and N_e(v) with applications to specific aspects of germplasm regeneration and collection. The authors considered an initial set of N diploid plants where P plants (contributing male and female gametes) are randomly sampled from N and, subsequently, R plants (contributing only male gametes) are randomly sampled from N-P. Proportions u and v are defined as in two-stage sampling for random-mating monoecious species.

The model allows a correlation between the numbers of female and male gametes contributed per individual within the set of P plants (a correlation assumed to be zero for random-mating species). Therefore, the natural or artificial rate of self-fertilization (assumed to be constant over parents) is s, the proportion of crossing plants is 1-s, and n is the total number of seeds collected from the set of P parents. Then, 2ns is the expected total number of female and male gametes contributed from selfing and 2n(1-s) is the expected total number of male and female gametes contributed from crossing: the overall total of contributed gametes is 2ns+2n(1-s) = 2n.

Vencovsky and Crossa (1999a) considered random sampling of seed (RS) and female gametic control (FGC) under unrestrictive inbreeding and under inbreeding equilibrium. Here, we will consider only RS and FGC under inbreeding equilibrium. The general equations for a fraction u of potential parents contributing female gametes and v, the fraction of potential parents contributing male gametes, are given in Annex B. Equations 8 and 9 can be adapted for seed collection with the assumptions that the number of seed parents (P) is a very small fraction of the entire population (u = 0) and that the number of pollinating parents (for s < 1) is sufficiently large, allowing the assumption that M = N, such that v = 1. This last assumption tends to inflate the corresponding N_e(v) value. In practical situations, the number of pollen parents for each seed parent can be estimated using molecular markers, but it requires a sample of adult plants and the corresponding maternal offspring to be genotyped.

Based on these assumptions, equations 10 and 11 were obtained for collecting n seeds from P seed parents (Vencovsky and Crossa 1999a) of a population in inbreeding equilibrium with a natural rate of self-fertilization s. A fundamental N_e(v) expression is given by Cockerham (1969) for a group of n individuals or seeds differentiated from a reference population of infinite size solely by drift. See equation 12, which can be adapted for measuring the degree of representativeness contained in a single maternal descendant (equation 13), which is equivalent to equation 10 or 11 when P = 1.

Seed collection from several subpopulations

Under natural conditions, there may be situations in which a large population of a species, or meta-population, is made up of a group of subpopulations. When seeds are collected in such a situation, the number of subpopulations from which seeds should be sampled is an additional sampling unit to be considered and incorporated for estimating  N_e(v). Annex C shows some of the equations obtained by Vencovsky and Crossa (2003) when extending the theory developed by Cockerham (1969).

Vencovsky and Crossa (2003) derived  N_e(v) equation 14 (Annex C), where S* is the total number of subpopulations in the region and C² is the squared coefficient of variation of the number of seeds collected from the S subpopulations, which is n_i for subpopulations i, such that n= Σ n_i (i = 1, 2, …, S). With equal n_i, C² = 0. Parameters F_ST and F_IT are Wright’s measures of the divergence among subpopulations and the total inbreeding coefficient, respectively (Weir, 1996). These F statistics can be estimated for the parental generation using codominant genetic markers.

When seeds are not collected in bulk but on a progeny basis, an additional sampling unit is incorporated, namely the number of seed parents. Now the expression for estimating N_e(v) is equation 15 (Vencovsky et al. 2007), where C²_s is the square of the coefficient of variation of the number of seeds among subpopulations and C²_m is the square of the coefficient of variation of the number of seeds among seed parents. Also, is the average coancestry among offspring within progenies, which can also be estimated using codominant molecular markers. Equations 14 and 15 (annex C) are useful for planning the collection sampling strategies necessary to obtain a desired  N_e(v) value.

Two-stage sampling model for dioecious species

Originally Crow and Kimura (1970) derived N_e(v) expressions for species with separate sexes based on simplified assumptions. Later Crow and Denniston (1988) made the necessary correction of this earlier work. The aim of Vencovsky et al. (2011) was to consider the statistical properties of the number of contributed gametes in various practical situations. To this end, they transformed and adapted the formulas of Crow and Denniston (1988) in such a way that they are easy to apply in practical breeding situations and when collecting and regenerating genetic resources. Vencovsky et al. (2011) consider a finite reference population such that all individuals are potentially functional parents; they also assume that, of the reference population, only a fraction of individuals is taken as functional parents. The derivations and results given by Vencovsky et al. (2011) refer only to dioecious species, and only the variance effective population size is considered. The reference population is considered under random mating, a condition also assumed by Crow and Denniston (1988).

Some of the derivations of Vencovsky et al. (2011) are shown in Annex D and refer to populations of arbitrary size, excluding some parents, that is, a fraction of parents is systematically discarded or rejected, such that only a subset of the initial N_f female and N_m male parents effectively participate in the reproduction process (Annex D, equations 16–18). Consider that, of the initial N_f female parents, only P participate in reproduction such that the fraction is

For male parents (M), the fraction is
 .

For the rejected N_f - P females and N_m - M males, the number of gametes contributed will be zero. Consider that f (female) and m (male) offspring are sampled, with a total of t = f + m, and also that the sex ratio is r =f / (f + m).

In another case, all the parents are included (Annex C, equations 19–21). In both cases, when some parents are excluded or all the parents are included, three cases of gametic control are considered, no gametic control or random sampling (RS), female gametic control (FGC), and female gametic control and male gametic control (FMC+MGC)

 

Practical applications

Probability models for number of alleles in the sample

The statistical genetic models used by Crossa et al. (1993) and based on equation 1 indicate that sample sizes of 160 to 250 plants of a random-mating population are required for capturing alleles at frequencies of 0.05 or higher in each of 150 loci, with 90%–95% probability. These formulas consider the sample size for conserving at least one allele per locus, but they do not quantify the probability of conserving two, three or more alleles per locus. The genetic diversity of a population depends on the number and frequency of alleles at a locus and across loci, and determining a sample size depends on whether estimates of allele frequencies are available (Hernandez and Crossa 1993).

However, when the required probability for conserving alleles at different loci increases or the frequency of a rare allele (p₀) drops to 1%, larger sample sizes than those specified earlier are required. For example, for unknown associations between genes within individuals and p ₀ =0.001, sample sizes of between 533 and 1708 individuals will retain, with a probability of 0.9999, at least one allele from each class, considering a wide range of alleles, k = 2 to 20, and a number of loci ranging from m = 5 to 1000. Sample size is always more affected by low allele frequency than by the number of alleles or the number of loci. To maintain alleles at 3% frequency with a 0.9999 probability, assuming unknown associations between genes within individuals, the required sample sizes are between 177 and 564 individuals for 2 to 20 alleles per locus and for 5 to 1000 loci. For retaining alleles at a frequency of 5%, 105 to 335 individuals are required. For 6 alleles per locus with allele frequencies between 0.03 and 0.10 and for 1 to 20,000 loci, a sample size of between 84 and 750 individuals will preserve at least one copy of each allele class in each locus with a probability of 0.9999999.

Assuming two alleles at each of the 20,000 loci and one of them at a 0.05 frequency, 186 individuals will preserve this allele at each locus with a 95% probability. However, a sample size of 172 individuals will only retain an allele at 5% frequency, assuming no association between genes within individuals at any locus. However, this does not seem to be sufficient if associations between genes within individuals exist at some loci and/or if maintaining alleles at frequencies between 3% and 1% is required.

Assuming that k-1 alleles occur at an identical low frequency for all loci makes the required sample size estimated by equation 1 conservative, because it is likely that some alleles will have higher frequencies, at least at some loci. The assumption of loci independence is unrealistic, since genetic drift may cause non-random associations between linked loci. If the population is under random mating (a cross-pollinated species), linkage equilibrium can be assumed for all possible pairs of loci, so that a similar number of coupling and repulsion combinations can be expected to occur for each pair of loci. This obviously does not hold for self-pollinated species or species with mixed self- and random-mating systems of reproduction.

Recommendations for estimating general sample size using probability models

General recommendations and/or guidelines concerning sample sizes should be given

• within ranges (or intervals), depending on model assumptions and biological considerations such as number of alleles, loci, genetic linkage, etc.

Furthermore, it is important to point out that

• certain sample sizes (n) obtained from the probability models might or not provide a similar value for N_e(v).

Two-stage sampling model for random-mating species – regeneration

For germplasm regeneration of N (finite) plants in the field, equations 3, 4 and 5 (annex A) can be adapted for different situations and mating designs, for monoecious species. Assuming that accession size is kept constant (n = N), these expression are much simplified, as illustrated in the following cases.

Two general cases are considered.

Case 1

The integrity of the accession is affected (u = P/N≤1), and there is loss of seeds from the original accession, but pollen and seeds are contributed by all remaining plants. Then P = M = uN (P = M; u = v). This is similar to the case where seeds of an accession are lost due to environmental causes, during storage in the genebank, or while accession plants are growing in the field. The exception is when u = 1, as shown in Table 5.1.

Table 5.1: Variance Effective Population Size (N_e(v)) for Regenerating Accessions of a Monoecious Species Using Expressions Given for Cases 1a-c and Cases 2b-c for n=N=1000 and Different Proportions of Parents Contributing Male and Female Gametes to the Next Generation (u), Assuming Constant Accession Size

Case 2

The integrity of the accession is not affected; that is, pollen is produced by all plants (M = N) but seed is collected from only P plants. Then, M = N, v = 1, but P = uN. This case simulates a situation where only a portion of the total number of pollinated plants in the field is harvested.

For Case 1, three mating designs based on different types of gametic control are considered:

1. female and male gametic control (full-sib mating system), FGC + MGC
2. female gametic control only (half-sib mating system), FGC
3. no female and male gametic control, that is, random open pollination, and unequal numbers of seeds are randomly taken from the set of P parents, RS

For Case 2, only schemes FGC and RS (b and c) are considered here because a scheme with control of both types of gametes (FGC+MGC) is not possible because, with Nu female parents available, only Nu pollen parents will be taken for obtaining full-sib progenies and not N.

In Case 1a (seed and pollen from P = M = uN), where female and male gametic control involves full-sib crosses and taking equal numbers of seeds from P = M = uN plants, N_e(v) = N[2u/(2-u)]. In Case 1b (seed and pollen from P = M = uN), where random pollination involves half-sib crosses among P = M = uN plants, and female gametic control implies taking equal numbers of seeds from P = uN plants, N_e(v)= N[4u/(4-u)]. In Case 1c (seed and pollen from P = M = uN), with random pollination and unequal numbers of seeds randomly taken from P parents, N_e(v) = Nu.

In Case 2b (seed from P = uN and pollen from P = M), where random pollination includes half-sibs among all plants, but female gametic control implies taking seeds from only P = uN pollinated plants, N_e(v)= N[4u/(1+2u)]. In Case 2c, with random pollination and unequal numbers of seeds randomly taken from P = uN parents, N_e(v) = N[4u/(1+3u)].

The above expressions for Cases 1 and 2 and the three mating systems, considering a range of accession losses (from 90% [u=0.l] to 0% [u=1, full stand of plants in the field]), are given in Table 5.1. For Case 1, with a perfect stand of plants (u=1), taking equal numbers of seeds per pollinated plant (female gametic control) and using hand-pollination (full-sibs) (male gametic control) produces N_e(v) = 2N; with only female gametic control and random pollination (half-sibs), N_e(v) = (4/3)N; and without male and female gametic control, N_e(v) = N. Values of N_e(v) for Case 2 (last two columns in Table 5.1) tend to be larger than their corresponding columns for Case 1 (the first three columns in Table 5.1) because the pollen pool is larger when M=N than when M=uN (except when u=1).

How can a genebank curator regenerate an accession of a monoecious plant species that has lost a percentage of its seeds or plants, and still attain N_e(v) = N?

In Case 1, a loss of 30% (u = 0.7) of the plant accessions can be recovered by male and female gametic control (N_e(v)=1077 when n = 1000), whereas a loss of 20% (u = 0.8) can be recovered by simply taking equal numbers of seeds from the 800 randomly pollinated plants that remained in the field (N_e(v) =1000).
Furthermore, in Case 2, if all 1000 plants are left to pollinate randomly (half-sibs), but equal numbers of seeds are taken from only 500 of them (u = 0.5), N_e(v)=1000 = n. This indicates the importance of controlling female gametes as a mechanism for recovering the desired magnitude of N_e(v) because of the possible loss of seeds and/or plant accessions before and during the growing cycle (Case 1) or the potential problems encountered when harvesting only a portion of the total number of pollinated plants in the field (Case 2).

Table 5.1 shows that N_e(v) is larger for female gametic control with half-sibs (Case 2 [fifth column in Table 5.1]) than for female and male gametic control with full-sibs (Case 1 [second column in Table 5.1]). This is true up to an accession loss equal to or greater than 30% (u ≤ 0.7) (1167 versus 1077 for u = 0.7). When u > 0.7 under female and male gametic control with full-sib families from P = M = u N plants, Case 1 is superior to that of half-sib families from M = N plants, but female gametic control on only P = uN families (1333, 1636 and 2000 versus 1231, 1286 and 1333 for u = 0.8 , 0.9 and 1.0, respectively). The half-sib mating system of Case 2 is the best option for compensating for insufficient sampling of female parents (smaller values of u) through increasing N_e(v), in contrast to the full-sib alternative of Case 1, which becomes important only for smaller accession losses (u > 0.8). On the other hand, female and male gametic control with the full-sib system in Case 1 (second column in Table 5.1) is always superior to female gametic control with the half-sib system in Case 1 (third column in Table 5.1) for all values of accession loss, as well as for the situation of no gametic control (fourth column in Table 5.1).

As already pointed out, when comparing half-sibs with female gametic control (Case 2) versus full-sibs and total gametic control (Case 1), the former is preferred if deterioration of the accession is intermediate to severe (0 < u < 0.7). Accession losses in Case 1 increase drift that is attributable to sampling parents (first stage of sampling), but controlling female and male gametes should decrease drift due to sampling of gametes (second sampling stage) and, therefore, should help to control overall drift. The advantage of the half-sib system is that it does not require hand-pollination because pollen from all plants in the field can be used. It is a valuable alternative for decreasing the contribution to drift attributable to the sampling of parents (first stage).

Recommendations for regenerating random-mating monoecious species

Generally speaking, in the regeneration process discussed so far, two aspects are fundamental for maintaining adequate representativeness of accessions:

• avoiding excessive deterioration of the accession

• practicing female gametic control when sampling seeds after reproduction

Two-stage sampling model for random-mating species – collection

In collection activities, obtaining exact estimates of N_e(v) when collecting seeds is not possible in most cases. This is due to the fact that the real size of the reference population is not known, and fractions u and v, therefore, cannot be estimated. Equations 6 and 7 (Annex A) are only approximations, since the number of female parents sampled for seed collection (P) is considered to be a very small fraction of the entire population (u=P/N = 0) and the number of pollen parents is admittedly very large, such that v=M/N=1.
Effective numbers, consequently, are overestimated. In any case, equations 6 and 7 can be considered for showing that under these assumptions, N_e(v) is dominated by the number of seed parents (P). As an example, consider a sample of 1000 seeds taken from P = 100 (10 seeds per parent), P = 200 (5 seeds per parent) and P = 500 (2 seeds per parent). With female gametic control, the resulting N_e(v) values are 308, 500 and 800, respectively (equation 6). If seeds are bulked and then sampled, the resulting N_e(v) values for n = 1000 are 286, 447 and 667, respectively (equation 7). Once again, results demonstrate the importance of female gametic control in these activities.

The alternative form of equation 6 can be used for planning collection strategies. If, for instance, collecting 15 seeds per parent ( =15) is feasible, and a value of N_e(v)=1000 is required for representing the reference population, then N_e(v) = 4P ( /+3) = 4P (15/18) = 1000, and P = 300 is the number of seed parents necessary to achieve N_e(v)=1000.

The total number of seeds required in this case is n = 4500, and the average N_e(v) per maternal progeny is _e(v) = 3.33. An additional condition not mentioned earlier and required for the validity of computing P is that the seed parents should be genetically unrelated (negligible coancestry among all parents).

There should be a minimum distance among seed parents during seed collection in order to have negligible coancestry among adult individuals. Investigations on pollen distance and on the reproductive neighbourhood area are generally recommended for estimating this distance (Viegas et al. 2011). Molecular markers can be used for this propose.

Recommendations for collecting random-mating monoecious species

The main lessons for collectors of genetic resources are that N_e(v) can be increased by

• increasing the number of seed parents (P)

• taking an equal number of seeds per seed parent, that is, exercising female gametic control

Two-stage sampling model for mixed self- and random-mating species – regeneration

Equations 8 and 9 (Annex B) show that N_e(v) increases linearly with sample size n, for a given s, when regeneration occurs before any loss of plants in the field, with all parents contributing male and female gametes. Both equations assume inbreeding equilibrium and can be adapted to germplasm regeneration and collection, considering the effect of (1) an arbitrary rate of natural self-fertilization (s) and (2) random sampling of seeds or female gametic control (FGC). We will examine the same two cases we studied when considering the two-stage model for random-mating species, that is,

• Case 1: loss of accessions given by u < 1 and v = u, with seed parents P = uN, and pollen parents M = uN

• Case 2: no loss of accessions but seeds taken from uN pollinated plants; that is, seed parents P = uN and pollen parents M = N, when s < 1. An additional effect that can be considered is given by the fact that the number of collected seeds (n) that can be taken is equal to N (constant n=N) or arbitrary (n ≠N).

Results obtained by Vencovsky and Crossa (1999a) indicated that FGC gives higher values of N_e(v) than RS. Female gametic control combined with adequate levels of u (u > 0.7) becomes an effective combination for maintaining values of N_e(v) equal to or higher than n, when n=N; whereas, with RS, values of N_e(v) are always smaller than sample size (n). In Cases 1 and 2 (here and below), RS of cross-fertilizing species (s = 0.0) is less affected by accession deterioration than are more self-fertilizing species at any level of accession deterioration (u). However, in Case 1 and Case 2, FGC of cross-pollinating species gives slightly higher values of Ne(v) than that of more self-fertilizing species but only for severe values of accession deterioration; the reverse is true for mild accession loss. Vencovsky and Crossa (1999a) found that Ne(v) = n for high values of u and when the species approaches panmixia (s = 0.0), under RS.

Other findings from Vencovsky and Crossa (1999a) relative to germplasm regeneration can be pointed out, as given below.

Female gametic control (FGC) always gives higher N_e(v) values than random sampling (RS) in all circumstances. This practice produces a rapid increase in N_e(v) for autogamic species when u > 0.7.

In Cases 1 and 2, when seeds are taken randomly (RS), accession regeneration is more efficient for panmictic species than for autogamic ones, under any level of accession deterioration. This is also true with FGC and severe deterioration loss. On the other hand, autogamic species with FGC produce higher N_e(v) in regeneration than allogamic species when u is mild.

For germplasm collection, the results obtained by Vencovsky and Crossa (1999a) contradict those for regeneration; that is, attaining adequate Ne(v), even with FGC for collecting autogamic species, is more difficult than regenerating accessions of this category when u > 0.8. The reverse is true for panmictic species.

For constant accession size (n=N) and FGC, considered here as a reference point, the minimum value of u (u*) acceptable for having N_e(v)=n is given by u*=4/(s2+5). Evaluating this quantity gives u*= 0.8, 0.76 and 0.67 for s = 0.0, 0.5 and 1.0, respectively. This indicates that more autogamic species permit slightly higher levels of accession deterioration, compared to more panmictic ones. Values given in Table 5.2 confirm this.

Increasing sample size (n) during regeneration increases N_e(v), but only when s < 1. For perfectly autogamic species (s = 1), increasing n has no effect on N_e(v), as can be seen in Table 5.2.

Table 5.2: Variance Effective Population Size (N_e(v)) When Regenerating N= 500 Plants and Harvesting an Arbitrary Number of Seeds (n) for Different Values of Accession Loss or Proportion of Functional Parents (u) and Natural Rate of Self-Fertilization (s), Assuming Inbreeding Equilibrium

How can a genebank curator regenerate an accession of a mixed self- and random-mating species that has lost a percentage of seeds or plants and still attain N_e(v)=n?

We can study the advantages of FGC over RS systems under Cases 1 and 2. For this situation, we consider an arbitrary value of n (n≠N). Values of N_e(v) for regenerating a hypothetical accession of N=500 plants in the field for Cases 1 and 2, under RS and FGC for various values of accession loss (1-u), natural rate of self-fertilization (s) and sample size (n) are shown in Table 5.2, assuming that plants are in equilibrium with respect to the mating system.

In Table 5.2, effective sizes for Case 2 are always higher than the corresponding ones for Case 1, because of the larger pollen pool in Case 2, where M=N as compared to M=uN in Case 1. With complete autogamy (s = 1), the condition M=N is impossible because there is complete self-fertilization. This situation (M=N with s = 1) was wrongly included by Vencovsky and Crossa (1999a). The advantage of Case 2 over Case 1 can be visualized by comparing N_e(v) values within the RS and FGC systems.

As already mentioned, considering only Case 1, Table 5.2 shows that for small losses in accessions (large u) it is easier to regenerate more autogamic species, whereas for large accession losses (small u) it is easier to regenerate more panmictic species.

If we collect n=N=500 seeds from 250 plants (u = 0.5) by taking two seeds per plant (FGC) of a panmictic species (s = 0) but using pollen from all M=N=500 plants (Case 2), the required condition N_e(v)=N=500 is achieved (N_e(v)=500.3 in the last column of Table 5.2). On the other hand, if we collect n=500 seeds from 250 plants, by taking two seeds per plant (FGC) of an autogamous species (s = 1) (Case 1), the required condition is never achieved because N_e(v)=(1/2)N=250, irrespective of sample size n. As expected, this indicates that severe accession loss has a greater impact on autogamic species than on panmictic ones. If pollen from all 500
plants of a panmictic species (s = 0) is used and 12 seeds (FGC) are taken from only 250 plants (u = 0.5 and n = 3000) (Case 2), N_e(v) = 1333.6 = 2.67N and N_e(v) = 1200.7= 2.40N with RS (Table 5.2). In Case 1, where the pollen pool is smaller than in Case 2 (for s < 1), with FGC, the condition that N_e(v) ≥ N is only achieved for a smaller accession loss (u > 0.7) than that allowed for Case 2 and u = 0.5.

In Case 1, when there is an accession loss of 10% (u = 0.9) and 3000 seeds are collected under FGC from a panmictic species (s = 0), N_e(v)= 2,118.5 > 4N. In Case 2, a 30% accession reduction on the female side (u = 0.7) and collecting n = 3000 will produce N_e(v) = 2154.6 = 4.31N.

For the three types of species considered, the condition that N_e(v) ≥ N is reached for u ≥ 0.7, under Case 1.
In Table 5.2, it can be seen that the highest N_e(v) values are obtained by practicing FGC and that fractions of accession loss are very small (u = 0.9). In general terms, the combination of female gametic control with the possibility of having a larger pollen pool, as in Case 2, produces important increases in N_e(v) for u = 0.7, 0.8 and 0.9, especially for sample size n >N. It is worth mentioning that the ideal situation for regenerating an accession occurs with FGC in autogamic species (s = 1) and no accession loss (u = 1). In such a case, there is no drift between the original and the regenerated accession and N_e(v) = _∞ (equation 9). Practicing random sampling in such a situation is inadequate.

Recommendations for regenerating mixed self- and random-mating species

• Taking an equal number of seeds (FGC) gives higher values of N_e(v) than a random number of seeds (RS).

• Female gametic control combined with adequate percentage germination (u >0.7) becomes an effective combination for maintaining values of N_e(v) equal to or higher than n, when n=N; whereas, with RS, values of N_e(v) are always smaller than sample size n.

• Random sampling of cross-fertilizing species (s = 0.0) is less affected by accession deterioration than it is with more self-fertilizing species at any level of accession decrease in percent germination.

• The combination of female gametic control with the possibility of having a larger pollen pool produces important increases in N_e(v). The ideal situation for regenerating an accession occurs with FGC in autogamic species (s = 1) and no accession loss (u = 1). In such a case, there is no drift between the original and the regenerated accession and N_e(v)= _∞ . Practicing random sampling in such a situation is inadequate.

Two-stage sampling model for mixed self- and random-mating species – collection

For collection activities, equations 10 and 11 (Annex B) are used to perform some numerical evaluations. It is assumed that n = 1000 seeds are collected from P=100 and P=20 adults, assuming random sampling (RS) and female gametic control (FGC), for different values of the natural rate (s) of self-fertilization. In natural conditions, the population is assumed to be in equilibrium relative to the mating system. Corresponding N_e(v) values are given in Table 5.3.

Table 5.3: Effective Sizes N_e(v) of n=1000 Seeds Collected from P=100 and P=20 Seed Parents, for Random Sampling (RS) and Female Gametic Control (FGC), from a Population in Inbreeding Equilibrium and Rate of Self-Fertilization (s)

 

 As can be seen in Table 5.3, the dominating factors in the effective size values, for n =1000 seeds collected, are the number of seed parents and the population’s rate of self-fertilization. Practicing gametic control has a positive effect for increasing N_e(v); however, that diminishes as the number of seed parents decreases.

Values such as those given in Table 5.3 can be used as a guide for planning sampling procedures when a specific N_e(v) value is desired. With FGC, P=100 (10 seeds per parent) and under panmixia (s = 0), we have an average effective size value of N_e(v)=3.1 per seed parent. To attain a value of N_e(v)=1000, in this case, P = 322 seed parents would be necessary under the supposition that they are genetically unrelated (negligible or zero coancestry among them). For autogamy (s =1.00) and P =100, we have an average
_e(v) = 0.5 per seed parent, meaning that P =2000 would be required to reach a total N_e(v) of 1000. This specific example reinforces the fact that obtaining seed samples with adequate representativeness is more difficult in more autogamic species.

Effective size values for a single maternal progeny, under the assumption of mixed self- and random mating, can be obtained directly from equation 13 (Annex B). Table 5.4 gives an N_e(v) that is applicable to this case for several sample sizes (n) and self-fertilization rates (s). Considering the same example as in the preceding paragraph, with n =10 seeds per seed parent and s = 0.00, N_e(v) = 3.1, as found earlier. The following numbers of seed parents (P) would be necessary for having a joint value of N_e(v)= 1000, for s = 0.0, 0.25, 0.5, 0.75 and 1.0, respectively, with n =10 seeds collected per seed parent: P =322, 526, 833, 1250 and 2000. For s =1.00, N_e(v)= 0.5 per progeny, which corresponds to a single gamete sampled from an idealized population.

In a broad review involving 30 tree species conducted by Sebbenn (2006), the number of seed parents (P) was computed for attaining an effective size of N_e(v)= 150. Values were obtained based on estimates of the average coancestry ( ) and the inbreeding coefficient (F), using molecular markers. The number of seeds collected per tree was considered large, such that (n-1)/n » 1 and (1+ F )/2n is negligible. Values of the number of seed parents (P) varied considerably among species, between 44 and 144, with an average of = 67. Most species showed a predominantly allogamic mating system with s varying between 0.0 and 0.25 and an average value of = 0.08.

Table 5.4: Variance Effective Population Size (N_e(v)) Values of a Single Maternal Progeny for Several Sample Sizes (n) and Natural Rates of Self-Fertilization (s), Reference Population of Infinite Size in Equilibrium under Mixed Self- and Random Mating

Recommendations for collecting mixed self- and random-mating species

The dominating factors in the effective size values are

• the number of seed parents

• the population’s rate of self-fertilization (s)

While practicing gametic control has a positive effect for increasing N_e(v), that diminishes as the number of seed parents decreases.

Two-stage sampling model for dioecious species

Several dioecious plant species are economically important, such as some domesticated dioecious arboreal species (Bandel and Gurgel 1967). Accounting for approximately 6% of angiosperms, dioecious species evolved from hermaphrodites in multiple independent events (Barret et al. 2010). Bawa (1980) mentions that dioecy is not as rare as is generally assumed. Bawa et al. (1985) found that there is a high level of dioecy (> 20%) among 333 tropical tree species. A similar percentage (22%) was found by Bawa (1974) among 130 lowland tropical tree species, and in the tropical savannah, it was found that about 15% of woody species are dioecious (Oliveira 1996; Oliveira and Gibbs 2000). Studying tropical coastal vegetation, Matallana et al. (2005) found 14% dioecy among 566 species and a higher percentage (35%) among dominant woody plants. In general, the percentage of dioecy among tropical forest trees and shrubs varies between 16% and 28% (Queenborough et al. 2007).

Regenerating dioecious species

Table 5.5 gives the results obtained by studying the effect of reducing seed viability or the germination rate (u; v) of an accession initially containing N_f =N_m = 100 individuals or seeds. Functional female and male parents are P=uN_f and M=vN_m, respectively (Annex D, equations 16–18).

Suppose enough resources are available for a plant-to-plant crossing scheme between the N_f = 100 females and males, aimed at achieving a total number of offspring of at least t = 200. After planting the seed, it is verified that only 50 of the male and female plants germinated. If female and male gametic control is practiced (by hand-pollinating female plants using plant-to-plant crosses and by taking an equal number of seeds per female plant), the resulting effective size is N_e(v)= 133.3 (Table 5.5) for r = 0.5 in the offspring generation. Equation 18 gives this value by taking t = 200, M = P = 50, u = v = 0.5 and r = 0.5. To achieve N_e(v) = 200 under these conditions, the number of parents should be higher, namely P=M = 66.7, such that u=v=0.67. For P=M=70 and u= v = 0.7, N_e= 215 for r = 0.5 (Table 5.5). To reach N_e(v)= t = 200 and the initial sizes, the following number of functional parents of each sex (P=M) would be necessary: 66.7, 68.6, and 76.4, for sex ratios r = 0.5, r = 0.4 and r = 0.3, respectively. These results indicate that losses of 25% to 30% of plants in the field due to poor germination or to any environmental cause might be compensated for, at least in part, by female and male gametic control. If all female and male parents are used in the crossing block and no loss of plants has occurred, the effective population size is doubled for r = 0.5 and reaches N_e(v) = 369.2 and N_e(v) = 289.6 for r = 0.4 and r = 0.3, respectively, with gametic control on both sexes. All values given here refer to a sample size of t =200. As expected, when the sex ratio deviates from r = 0.5, there is a reduction in N_e(v).

Table 5.5: Variance Effective Population Size (N_e(v)) Values for Examples of Genebank Accession Regeneration with an Original Dioecious Reference Population with N_f = N_m = 100, under Random Mating, Assuming That t = 200 Offspring Are Sampled; Reducing Values of u and v (u = v); Three Sex-Ratio Levels (r = 0.3, r = 0.4 and r = 0.5) and Three Procedures of Gametic Control Random Sampling (RS), Female Gametic Control (FGC ) and Female Plus Male Gametic Control (FGC + MGC)

 

As an alternative to the previous examples, consider that of the N_f =100 only 33 are available. With a germination rate of only 50%, the remaining functional parents will be P =16 and M =50, such that u=0.16 and v=0.5. Since plant-to-plant crosses are not possible, male gametic control is not practicable and the option is to apply female gametic control. With these values and using equation 17, for t =200 and r =0.5, for the offspring generation, the resulting effective size is N_e(v)=51.7, which is relatively small due to the considerable reduction in the number of female parents and the impossibility of controlling the number of contributed male gametes. With the basic quantities remaining the same but a germination rate of 70%, P = (0.7)(33)23, M=(0.7)(100)=70, u=0.23 and v = 0.70, for a result of N_e(v)= 75.9.

When all individuals of the reference population are functional parents, the following values are obtained for t =f +m and r =0.5, N_e(v)=t for RS, N_e(v)=1.33t for FGC, and N_e(v)=2t for FMC+MGC (see Annex D, equations 19, 20 and 21).

Recommendations for regenerating dioecious species

Maintaining high levels of germination and viability is fundamental for preserving adequate Ne(v) values of accessions in genebanks. In the example,

• a 20% increase in the germination rate leads to an increment of 24.2 units, or 46.8%, in N_e(v);

• deviations of the sex ratio from r = 0.50 only have sizable effects on N_e(v) for larger values of u and v or when the viability of the accession is higher than 50% or 60%.

Collecting dioecious species

When collecting germplasm, the number of seed parents from which the seed samples are taken is, in general, a small fraction of the total population size, such that u tends to be small or negligible. In relation to the pollinator plants, the corresponding fraction v can also be smaller or larger than u, depending on the factors affecting pollen dispersal, the structure of the population and the size of the reference population.

Suppose that resources are available for collecting only t = 200 seeds, stemming from P = 100 female parents randomly crossed with M = 100 males within an ecogeographic area. Assume that the population of females and males is very large (N_f = N_m = 10,000) such that only 1% of the total population is sampled (u = v = 0.01), and that female gametic control is used. The expected population size will be N_e(v) = 115.1 (equation 17). For such small fractions u and v, the effective size is smaller than the sample size (t = 200) even with female gametic control (Table 5.6). Control of both types of gametes is rarely applicable in collection activities, but just for comparison, let us assume complete control of gametes. Effective size would then be N_e(v) = 134.2 (equation 18), with a gain of 16.6% relative to N_e(v) = 115.1. In this example, it was assumed that the sex ratio is r = 0.5 for both offspring and parents.

Table 5.6 shows some applications of equations 16 and 17, with a constant number of functional parents but a decreasing degree of representativeness (u = v) and increasing size of the reference population (N_f ; N_m). It is assumed that the sex ratio remains the same between generations. Here parameter r is included as a peculiarity of some species. As already seen in previous items, as r deviates from the value r = 0.5, N_e becomes smaller. It should also be noted that sampling 0.1% or 5.0% of individuals in the parental generation has little effect on N_e(v). Female gametic control is required for obtainingN_e(v) = t, but the effect is noticeable only when u and v are large.

Estimating N_e(v) in dioecious species requires distinguishing the sex of the offspring. Since this is generally difficult or impossible in plant seedlings or seeds, some strategies are necessary. An indirect procedure is estimating r at the adult stage in a natural population and transferring this value to the offspring.

Another procedure is achieving some protection through gametic control. Establishing that a sample of t seeds should have an effective size N_e(v) ≥ t and if the entire set of parents participate in reproduction, the given condition will be achieved through female gametic control for a range of r values between 0.28 and 0.72. With control of both types of gametes (FGC+MGC), this range is increased to 0.21 ≤ r ≤ 0.79 gametic control. Therefore, gametic control is also important when the sex ratio is not known.

Recommendations for collecting dioecious species

As r deviates from the value r = 0.5, N_e(v) becomes smaller.

• Female gametic control is required for obtaining increased values of N_e(v) ; however, the effect is noticeable only when the percentage of seed germination is large.

• Sampling 5% or 0.1% of adult plants of the population for seed collection has negligible effect on N_e(v) .

Table 5.6: Variance Effective Population Size (N_e(v)) Values for Examples of Germplasm Collection in Dioecious Species, with Increasing Size of the Reference Random Mating Population (N_f ; N_m), a Constant Total Number of Parents (P+M=200) and a Reducing Fraction of Functional Parents Sampled (u and v). Total number of seeds collected: t=200.

 

Sampling seeds from subpopulations

As already mentioned, there are situations in which seed samples are intended to represent a large population composed of subpopulations or fragments in natural conditions. In such a case, the number of subpopulations (S) is an additional sampling unit to be taken into account.

In this example, we consider that a sample of n =1000 seeds is taken from S subpopulations. We assume that an equal number of seeds (n/S) is sampled in bulk from each subpopulation and obtain N_e(v) values by applying equation 14 (Annex C). For evaluation, four degrees of genetic divergence due to drift among subpopulations (F_ST = 0.02, 0.05, 0.10 and 0.15) are incorporated. These subpopulations are considered to be in equilibrium with respect to the mating system, with a constant natural rate of self-fertilization (s), such that, within each one, the natural level of inbreeding is F_IS = s/(2-s). The relationship (1-F_IT)=(1-F_ST)(1-F_IS) (Weir 1996) was used to obtain the total level of inbreeding F_IT, which is also necessary for applying equation 14. To widen the example, we include two other potential situations, namely, (1) that the number of populations existing in the region is very large (S*=  ∞ ) or (2) that this number is relatively small (S*=50). Pertinent  N_e(v) values obtained for this example are given in Table 5.7. In equation 14, we considered that C² = 0.

A striking outcome in this example is that  N_e(v) is strongly dependent on the number S of subpopulations sampled. This can be seen even when the divergence among subpopulations is only a small fraction of the total genetic diversity, such as when it is only 2% (F_ST = 0.02). Effective size barely reaches N_e(v) =500 for s < 50 and S*= ∞ , under random mating within subpopulations (s = 0) and with a mixed mating system (s = 0.5). The degree of among-subpopulation divergence (F_ST) also shows a predominant influence upon N_e(v).

Equation 14 can be used for determining the number S necessary for obtaining a given N_e(v) value. If n is sufficiently large and S*= ∞ , this expression leads to S = 2F_ST N_e(v). If, for instance, F_ST = 0.05 and the desired N_e(v) is 1000, then the necessary number is S =100 for large n, such that (1+F_IT)/2n is negligible.

When the reference population contains a finite set of subpopulations (S*= 50 in the example), the general trend remains the same. Effective size values for subpopulations with a mixed mating system (s = 0.5) are always smaller than the corresponding values obtained for s = 0, as was observed for the infinite model (S*= ∞ ). Also, in general, N_e(v) diminishes as among-subpopulation divergence (F_ST) increases. An exception occurs when all subpopulations are included in the sample (S = S*= 50). As expected, it is much easier to represent a population in a sample when it is composed of a finite set of subpopulations than when a population includes a very large number of subpopulations. Such results demonstrate the importance of clearly defining the reference populations when computing an N_e(v) value.

Table 5.7: Effective Size N_e(v)) Values for Examples of n=1000 Seeds Collected from S Subpopulations, for Increasing Among-Subpopulation Divergence (F_ST), Two Levels of Self-Fertilization (s) within Subpopulations and Two Models of Natural Population Size (S*)

So far, we have stressed the importance of having a sufficient number S of subpopulations as sources for seed collection. In practice, we understand that such a sample of S subpopulations should not be taken at random. The fact is that, very often, adjacent subpopulations have a certain degree of genetic resemblance, a phenomenon detectable when genetic distances between all pairs of subpopulations are correlated with the corresponding geographic distances. This resemblance can be evaluated using morphological measurements or genetic data based on molecular markers. This technique involves estimating spatial autocorrelation, as shown by Sokal and Oden (1978a,b) and originally by Moran (1950). The ideal is to collect seeds from a subset S of subpopulations having the smallest possible degree of resemblance to each other. General results obtained by applying equation 14, as well as those given in Table 5.7, are applicable as strategies for in situ preservation.

• N_e(v) is strongly dependent on the number S of subpopulations sampled.

• The degree of among-subpopulation divergence (F_ST) also shows a strong influence upon N_e(v). N_e(v) diminishes as among-subpopulation divergence (F_ST) increases. An exception occurs when all subpopulations are included in the sample. As expected, it is much easier to represent a population in a sample when it is composed of a finite set of subpopulations than when a population includes a very large number of subpopulations.

• It is important to define the reference populations when computing N_e(v).

• When possible, collect seeds from a subset S of subpopulations having the smallest possible degree of resemblance to each other.

 

General recommendations

Regeneration and seed collection

• Practicing female gametic control (harvesting an equal or approximately equal number of seeds per seed parent) is always recommended for all species.

• When possible and especially for valuable materials, always obtain genetic information concerning the mating system of the species, based on molecular markers. This is especially important for species with mixed mating systems.

• A computer programme for analysing marker data is available for this purpose: MLTR win (Multilocos Mating System Program) by Kermit Ritland (http://genetics.forestry.ubc.ca/ritland/programs.html).

Regeneration

• Avoid a loss of viability or germination below 60%–70%during storage of germplasm accessions.

• In addition to female gametic control, also apply male gametic control when species is a cross-fertilizer. This requires plant-to-plant hand pollination and is important for accessions that have low germination (high degrees of deterioration).

Collection

• In the case of a single population, avoid harvesting large amounts of seeds from a small number of seed parents. This is valid for all species. To attain reasonable representativeness, harvest seeds from at least 25 seed parents for cross-fertilizing species and 50 seed parents for autogamous species.

• In the case of large and genetically structured populations in natural conditions, seeds should be collected from the largest possible number of sites or subpopulations. This is recommended when no additional information is available about the genetic structure of the population.

• When possible, always obtain measures of the genetic divergence among collection sites or subpopulations of the species in a given ecogeographic region, and of the level of natural inbreeding. This can be done as an activity parallel to collection. The following computer programmes are available for this purpose: FSTAT by Jérome Goudet (www2.unil.ch/popgen/softwares/fstat.htm) and GDA (Genetic Data Analyses) by Paul O. Lewis and Dmitri Zaykin: (http://hydrodictyon.eeb.uconn.edu/people/plewis/software.php).

Future challenges/needs/gaps

Making general considerations about future challenges in regeneration and collection activities is not simple. Every programme in the area of genetic preservation and conservation has its own peculiarities, objectives and difficulties. There are, however, some goals all curators have in mind when collecting seeds for a genebank. One is to apply strategies that permit obtaining the best possible samples under specific circumstances. We believe that this goal can be achieved when the strategies are developed on the basis of genetic information furnished by molecular markers. It is worth remembering how important this information can be for conservation activities. The following points can be clarified using molecular techniques:

• the mating system

• the degree of natural inbreeding

• the level of genetic divergence among collection sites or subpopulations; area of seed and pollen dispersal; the degree of genetic relationship or coancestry between adult plants in natural conditions; the neighbourhood within which adults are genetically related

There are three basic units to be considered in seed collection:

• the number of seeds per seed parent

• the number of seed parents per subpopulation

• the number subpopulations

Molecular genetic information allows one to find the ideal combination of these units for maximizing the representativeness of the seed sample. It also helps to define where the samples should be taken.

We understand that the collection of valuable materials or of populations endangered by human activity should be supported by estimates of the parameters of population genetics obtained from molecular markers. Along with the use of dense molecular markers, which are becoming cheaper with time, the use of valuable bioinformatics visualization tools and efficient biometrical-statistical methods are indispensable tools that can assist genebank curators in making appropriate decisions about optimum sampling strategies for regeneration, collection and subpopulation structures.

 

Acknowledgements

This work was partly funded by CNPq (the Brazilian Council for Scientific and Technological Development). The authors would like to thank Evandro V. Tambarussi and Fernando H. Toledo for their assistance.

Back to list of chapters on collecting

 

References and further reading

Allard RW. 1970. Population structure and sampling methods. In: Frankel OH, Bennett E, editors. Genetic Resources in Plants: Their Exploration and Conservation. Blackwell Scientific Publications, Oxford, UK. pp.97–107.

Bandel G, Gurgel JTA. 1967. Proporção do sexo em Pinheiro Brasileiro Araucaria angustifolia (Bert.) O. Ktze. Rev. Tec. Serv. Florestal do Est. de S. Paulo. Secr. Agr. Est. S. Paulo 6: 209–220.

Barret SCH, Yakimowsky SB, Field DL, Pickup M. 2010. Ecological genetics of sex ratios in plant populations. Philosophical Transactions of the Royal Society B. 365:2549–2557.

Bawa KS. 1974. Breeding systems of tree species of a lowland tropical community. Evolution 28(1):85–92.

Bawa KS. 1980. Evolution of dioecy in flowering plants. Annual Review of Ecology, Evolution, and Systematics 11:15–39.

Bawa KS, Perry DR, Beach JH. 1985. Reproductive biology of tropical lowland rain forest. I. Sexual systems and incompatibility mechanisms. American Journal of Botany 72(3):331–345.

Cockerham CC. 1969. Variance of gene frequency. Evolution 23:72–84.

Cockerham CC, Weir BS. 1984. Covariances of relatives stemming from a population undergoing mixed self and random mating. Biometrics 40:157–164.

Crossa J, Vencovsky R. 1994. Implications of the variance effective population size on the genetic conservation of monoecious species. Theoretical and Applied Genetics 89:936–942.

Crossa J, Vencovsky R. 1997. Variance effective population size for two-stage sampling of monoecious species. Crop Science 37:14–26.

Crossa J, Vencovsky R. 1999. Sample size and variance effective population size for genetic conservation. Plant Genetic Resources Newsletter 119:15–25.

Crossa J, Hernandez CM, Bretting P, Eberhart SA, Taba S. 1993. Statistical genetic considerations for maintaining germplasm collections. Theoretical and Applied Genetics 86:673–678.

Crow JF, Denniston C. 1988. Inbreeding and variance effective numbers. Evolution 42(3):482–495.

Crow JF, Kimura M. 1970. An Introduction to Population Genetics Theory. Burgess Publishing, Minneapolis, Minnesota.

Hernandez CM, Crossa J. 1993. A program for estimating the optimum sample size for germplasm conservation. Journal of Heredity 84:1.

Marshall DR, Brown AHD. 1975. Optimum sampling strategies in genetic conservation. In: Frankel OH, Hawkes JG, editors. Crop Genetic Resources for Today and Tomorrow. Cambridge University Press, Cambridge, UK.

Matallana G, Wendt T, Araujo DSD, Scarano FR. 2005. High abundance of dioecious plants in a tropical coastal vegetation. American Journal of Botany 92(9):1513–1519.

Moran PAP. 1950. Notes on continuous stochastic phenomena. Biometrika 37:17–33.

Namkoong G. 1986. Sampling for germplasm collections. Horticultural Science 23(1):79–81.

Oliveira PE. 1996. Dioecy in the cerrado vegetation of Central Brazil. Flora 191:235–243.

Oliveira PE, Gibbs PE. 2000. Reproductive biology of woody plants in a cerrado community of Central Brazil. Flora 195:311–329.

Queenborough SA, Burslem DFRP, Garwood NC, Valencia R. 2007. Determinants of biased sex ratios and inter-sex cost of reproduction in dioecious tropical forest trees. American Journal of Botany 94(1):67–78.

Sebbenn AM. 2006. Sistema de reprodução em espécies tropicais e suas implicações para a seleção de árvores matrizes para reflorestamentos ambientais. In: Higa AR, Silva LD, editors. Pomares de Sementes de Espécies Florestais Nativas. Fundação de

Pesquisas Florestais do Paraná (FUPEF), Curitiba, Brazil. pp.93–108.

Sokal RR, Oden NL. 1978a. Spatial autocorrelation in biology. 1. Methodology. Biological Journal of the Linnean Society 10:199–228.

Sokal RR, Oden NL. 1978b. Spatial autocorrelation in biology. 2. Some biological implications and four applications of evolutionary and ecological interest. Biological Journal of the Linnean Society 10:229–249.

Vencovsky R. 1978. Effective size of monoecious populations submitted to artificial selection. Brazilian Journal of Genetics 1(3):181–191.

Vencovsky R, Crossa J. 1999a. Variance effective population size under mixed self and random mating with applications to genetic conservation of species. Crop Science 39:1282–1294.

Vencovsky R, Crossa J. 1999b. Medidas de representatividad. Workshop O Melhoramento de Plantas na Virada do Milenio. Universidad Federal de Vicosa, MG, Brasil.

Vencovsky R, Crossa J. 2003. Measurements of representativeness used in genetic resource conservation and plant breeding. Crop Science 43:1912–1921. doi:10.2135/cropsci2003.1912.

Vencovsky R, Nass LL, Cordeiro CMT, Ferreira MAJF. 2007. Amostragem em recursos genéticos vegetais. In: Nass LL, editor. Recursos Genéticos Vegetais. Embrapa, Brasília.

Vencovsky R, Chavez L, Crossa J. 2011. Variance effective population size for dioecious species. Crop Science (in press).

Viegas MP, Silva CLSP, Moreira JP, Cardin LT, Azevedo VCR, Ciampi AY, Freitas MLM, Moraes MLT, Sebbenn AM. 2011. Diversidade genética e tamanho efetivo de duas populações de Myracrodruon urundeuva fr. all., sob conservação ex situ. Revista Árvore 35(4):769–779.

Weir BS. 1996. Genetic Data Analysis II - Methods for Discrete Population Genetic Data. Sinauer Associates, Sunderland, Massachusetts.

Wright S. 1931. Evolution in Mendelian populations. Genetics 16:97–159.

 

Internet resources

MLTR (Multilocos Mating System Program) by Kermit Ritlan (a computer programme for analysing marker data): http://genetics.forestry.ubc.ca/ritland/programs.html

Resources for obtaining measures of genetic divergence among collection sites or subpopulations of a species in a given ecogeographic region, and of the level of natural inbreeding:

FSTAT by Jérome Goudet: www2.unil.ch/popgen/softwares/fstat.htm

GDA (Genetic Data Analyses) by Paul O. Lewis and Dmitri Zaykin: http://hydrodictyon.eeb.uconn.edu/people/plewis/software.php

  

Back to top

 

Annex A: Two-stage sampling model for random-mating species

Annex B: Two-stage sampling model for mixed self- and
random-mating species

Annex C: Seed collection from several subpopulations

Annex D: Two-stage sampling model for dioecious species