A Global Conservation Strategy for Cassava and Wild Manihot ...

A Global Conservation Strategy for Cassava(Manihot esculenta) and Wild Manihot Species 1December 20101 Prepared by Clair H Hershey (chh23@cornell.edu), consultant to CIAT. A summary of stakeholder deliberationsand recommendations.

ContentsDisclaimer............................................................................................................................................... 4Executive Summary................................................................................................................................ 41 Introduction and background to the strategy ....................................................................................... 81.1 Goals and outputs ................................................................................................................................81.2 Building on progress, meeting new challenges ...................................................................................81.3 Precedents and resources ...................................................................................................................111.4 Sources of information ......................................................................................................................132 Cassava in the global economy and agro-ecosystem......................................................................... 142.1 Manihot species overview and the origins of cassava.......................................................................142.2 Production overview..........................................................................................................................162.3 Evolution and farmer selection..........................................................................................................252.4 Modern genetic improvement............................................................................................................252.5 Landraces and wild species at risk ....................................................................................................263 Cassava-related networks .................................................................................................................. 274 Cassava in the regional crop strategies.............................................................................................. 284.1 Americas............................................................................................................................................284.2 Eastern Africa ....................................................................................................................................284.3 SADC Region....................................................................................................................................294.4 South and Southeast Asia ..................................................................................................................294.5 Pacific Islands....................................................................................................................................304.6 Summary from the regional reports...................................................................................................305 Overview of cassava collection and conservation............................................................................. 315.1 Collecting strategies, techniques and priorities .................................................................................315.2 Conservation alternatives for cassava................................................................................................335.2.1 Field..............................................................................................................................................................345.2.2 In vitro..........................................................................................................................................................355.2.3 Cryopreservation ..........................................................................................................................................355.2.4 Seed ..............................................................................................................................................................365.2.5 Pollen............................................................................................................................................................375.3 Conservation costs.............................................................................................................................376 Overview of wild Manihot species collection and conservation....................................................... 387 Overview of current cassava genebanks............................................................................................ 397.1 Genebank holdings of landrace varieties and collection needs .........................................................397.2 Collaboration arrangements in conservation .....................................................................................417.3 The most important collections .........................................................................................................428 Overview of current wild Manihot species genebanks...................................................................... 439 Improving the efficiency of conservation.......................................................................................... 529.1 Core collections .................................................................................................................................529.2 Duplicate identification .....................................................................................................................539.3 Improved slow-growth conditions.....................................................................................................5310 Characterization and preliminary evaluation................................................................................... 5411 Distribution...................................................................................................................................... 5511.1 Benefits and risks.............................................................................................................................5511.2 Forms of exchange...........................................................................................................................5511.2.1 Vegetative ..................................................................................................................................................5511.2.2 Seeds ..........................................................................................................................................................5611.3 Quarantine considerations ...............................................................................................................5611.4 Procedures for distribution ..............................................................................................................5811.4.1 Sources .......................................................................................................................................................582

DisclaimerThis document, developed with the input of a large number of experts, aims to provide a framework for the efficientand effective ex situ conservation of globally important collections of cassava (Manihot esculenta) and Wild Manihotspecies.The Global Crop Diversity Trust (the Trust) provided support for this initiative and considers this document to be animportant framework for guiding the allocation of its resources. However, the Trust does not take responsibility for therelevance, accuracy or completeness of the information in this document and does not commit to funding any of thepriorities identified.This strategy document (dated December 2010) is expected to continue to evolve and be updated as and whencircumstances change or new information becomes available.Executive SummaryAs for any crop, the future potential of cassava to contribute to the sustainable benefit of humankind will relyfundamentally on the availability and use of broad-based genetic resources. These resources are basically thelandrace varieties that evolved for centuries under farmer and natural selection, and some 100 wild species of thegenus Manihot. The genus is native to the Americas, and most of the genetic diversification has occurred here.Traders introduced cassava into Africa in the 1500s and into Asia in the 1800s. Both have become importantsecondary centers of genetic diversity, especially Africa.Cassava is a vegetatively propagated crop, while all the wild species are seed-propagated in their naturalenvironments. In order to preserve the genetic integrity of a landrace, cassava must be conserved in vegetative form.The most common forms of conservation are as field-grown plants or as plantlets started from meristem tips, culturedon sterile artificial media, under light, temperature and media conditions that induce slow growth. For either field or invitro conservation, expensive periodic regeneration is required, at a much higher frequency (typically every 12-24months) than is typical for seed conservation.This report summarizes stakeholder input into a study whose objectives were defined jointly by the Global CropDiversity Trust (the Trust) and the International Center for Tropical Agriculture (CIAT): “To develop a strategy for theefficient and effective conservation of Manihot and cassava genetic resources and identify priority collections forlong-term support. The strategy will promote the rationalization of conservation at regional and global levels.”The proposed strategy is based on the literature, personal interviews, and on two recent activities sponsored by theTrust: a detailed survey sent to some 50 genebanks around the world, and a meeting of a small group of cassavaand Manihot genetic resources experts at CIAT from 30 April to 2 May 2008.Thirty-four surveys were returned for cultivated cassava, and the summarized highlights follow.• Most cassava-growing countries have established a genebank of local landraces, owned and maintained bygovernment organizations.• Most collections were established since the 1970s, but some as recently as a few years ago.• Most countries note collection gaps (less so for Asia), due to lack of funding, losses from natural disastersand social conflict, difficult access to areas for collecting, and inadequate collecting techniques of the past.• Information is generally managed manually, and even when managed electronically, is generally notavailable on the Internet.• Nearly all programs rely primarily on field-grown plants, but may have part of their collection in vitro as well.Globally, only about one-quarter of accessions held by national programs appear to be conserved in vitro.4

Cassava Conservation Strategy• Two international centers (CIAT and IITA) maintain regional collections for the Americas and Asia (CIAT)and for Africa (IITA).• There are very few national genebanks that have the capacity to carry out safe international exchange insituations where viruses of quarantine significance are present. Mostly this is done via the internationalcenters.• Human resources development is often identified as a critical area for the future success of germplasmconservation.• Most respondents see the value of a global network for cassava genetic resources, if it is adequatelyfunded.Out of the information from these surveys, and combined with information from other sources to fill in missing data,we developed a matrix of estimates, by country and region, of various parameters for cassava genetic diversity.About two-thirds of cassava is currently grown in Africa, but probably well over half the landraces occur in theAmericas. This is to be expected in view of origin of the species in the Americas. This study estimates some 27,000distinct landraces of cassava in situ, and about 10,000 maintained in genebanks. It is proposed that a total of about15,000 landrace varieties should be conserved ex situ in order to represent the complete genetic diversity of thespecies.A conservation strategy should consider security, cost and efficiency in its design. Security is a function of both thenumber of replications of a genebank (in different sites, or in different forms), and the management level of each.Field genebanks are the least secure, followed by in vitro slow growth, and finally, cryoconservation. Currently, onlythe IARCs have significant cryo genebanks for cassava.The baseline for our process of thinking about conservation was a meeting held at CIAT in 1992, “the first meeting ofthe International Network for Cassava Genetic Resources.” This group saw the need to decentralize conservationand strengthen the ability of national programs to conserve cassava germplasm, especially through training andinfrastructure development for in vitro technologies. Unfortunately, this was near the beginning of an era of seriouslydeclining public funds for agricultural research and development. The network did not make much progress towardsits goals, and really did not become a successful forum for cassava genetic resources.There are compelling reasons to rethink a decentralized strategy where each national program has the ability toconserve its germplasm in a highly secure system, which normally involves a field collection backed up by an in vitrocollection. There have been some significant changes in the world of cassava genetic resources, which impact thestructure of an optimum conservation strategy. First, the status of the collections maintained by the CGIAR has beenclarified. These collections are now part of the Multilateral System of the International treaty under its Article 15.Secondly, international exchange has become much safer and more acceptable with advances in virus indexing.This changed environment allows us to think in new ways about the optimum conservation system for cassava.Conservation in vitro (slow growth or cryopreserved) is highly non-site-specific and therefore large efficiencies can begained by centralization. This centralization in the international centers now becomes politically viable, becauseownership has been clarified, and international exchange is also clearer and more secure from a quarantineperspective. We now have an opportunity to develop a strategy that is biologically and economically rational, createsa structure of interdependence and collaboration among genebanks, and at the same time conforms to the newpolicy environment. We can think of this strategy as one of collaborative centralization.In a nutshell, the strategy that comes out of this reality could be the following:• Collecting in priority areas is carried out to fill gaps, with the aid of genetic diversity studies and GIS.• National program genebanks and international center genebanks are systematically compared for matchingand non-matching accessions, based on passport, morphological and molecular information. This wouldevolve into a common cassava registry at a global level.5

Cassava Conservation Strategy• CIAT and IITA duplicate all the landraces of national program collections, in their respective regions ofresponsibility (CIAT: Americas and Asia; IITA: Africa). Currently they appear to maintain about 50-60% ofthese accessions.• National programs commit to at least one working genebank that serves the purposes both of conservationat a moderate level of security, and evaluation.• CIAT and IITA maintain at least two forms of each accession. Currently this may be an in vitro activegenebank plus a black box duplicate kept in another center. In the future, cryopreserved accessions will beeither the main or the backup genebank.• CIAT and IITA commit to making the material they maintain available to national program genebanks, whenrequested.• CIAT and IITA commit to meeting the demands and phytosanitary requirements for international exchangeof cassava landrace varieties under terms of the International Treaty. Along with this, it is urgent to developprotocols for the safe movement of vegetative germplasm between the Americas and Africa.• There is a mechanism developed for periodic interaction among stakeholders. Most notably this will bebetween the international centers and the national programs. Each will have a formal responsibility toperiodically inform the other of the status of collections.This collaborative centralization strategy will lead to greater overall efficiency, but at the same time, new initiativesneed to be supported to get to this point of lower costs and higher security.Information sharing is the starting point, especially to develop the common cassava registry. This will involve somestandardization of information in order to succeed. The international centers should take the lead in this exercise.Germplasm indexing and transfer between the international centers and national programs will be required, formovement in both directions. Expansion and upgrading of some facilities will be needed to allow this.In the ongoing conservation process, there is research with a high payback in terms of greater security andefficiency. Duplicate identification, further improvements for in vitro slow growth techniques, improvingcryopreservation, and flower induction for seed conservation are all research areas outside the funding stream forroutine conservation, but which will contribute to greater conservation and use efficiencies in the long term. Having acoordinated centralized strategy will multiply these efficiencies and savings, especially in areas of safety duplication,duplicate identification, information management and international distribution. For example, the international centersand national genebanks can work jointly, on complementary tasks, for duplicate identification if there is a commoncassava registry.The international centers need to continue to evolve plans on how they are going to conserve the base collection,along with one or more systems of secure backup. A sensible starting point seems to be for the centers to have an invitro base collection and an in vitro backup collection in another location (the current situation). The planting of fieldcollections can be based on user demand for planting material for evaluations, and may be contracted out tobreeders, for example. The question of how the international centers jointly decide to manage the collections hasbeen based on historical regions of mandate, and the fact that there are viruses exclusive to either the Americas orAfrica. This division is reasonable into the medium-term future. As the protocols for virus detection and cleaningprogress, it is certainly feasible to see CIAT and IITA acting as two centers for safe conservation of the entire globalcollection, each acting as the safety backup for the other, and thereby eliminating the current black box system ofsafety duplication.Cryopreservation is clearly an option for effective, inexpensive, secure long-term conservation, but work remains tobe done on achieving an adequate recovery level for about one-third of accessions (based on results from CIAT’score collection). Research should continue on improving recovery of these recalcitrant types before committing tolarge-scale cryopreservation of any genebank.6

Cassava Conservation StrategyAs a future alternative to vegetative cassava genebanks, the seed from selfed accessions could be a less expensiveand efficient conservation method, and would be equally or more effective in breeding programs. Since manycassava accessions do not readily flower, there is a need for research on the induction of flowering. There could alsobe collaboration with national programs that maintain collections in sites where flowering tends to be more profuse.This is another example of the efficiencies that could be introduced with a common cassava registry. Long-term, wemight envision a conservation strategy that consists of a combination of cryopreserved meristem shoots, and seedmaintained in conventional cold storage. This would combine the advantages of both seed and vegetativeconservation in a low-cost, secure system.The wild species present both a simpler and a more complex situation compared to Manihot esculenta. It is simpler inthat only a handful of institutions are involved in conservation – mainly EMBRAPA and the University of Brasilia inBrazil, and CIAT. It is more complex in that:• The taxonomy of species is still poorly defined. An ongoing taxonomic revision of the genus is stalledbecause of the retirement of the main scientist, Antonio Allem of EMBRAPA.• The highest concentration of species is native to threatened habitats. This is especially true in south centralBrazil, in the campo cerrado, where the expansion of agriculture and urbanization are rapidly encroachingon the wild species habitats.• A secondary center of diversity, with a distinct set of species, exists in Mesoamerica. Here, and especially inMexico, cassava is a relatively unimportant crop, and it is difficult for these governments to justify investmentin Manihot conservation.• Fewer than half the species are conserved in vitro, and very few are protected in national or regionalreserves, in their native habitat.• Wild species conservation presents many challenges, especially with regard to regeneration. Progress isbeing made both in seed and in vitro propagation, but much remains to be done.• The value of the wild species is continually becoming more evident as new characters are identified withpotential for transfer to cassava, and the techniques for efficient transfer and selection of specific genes aredeveloped.7

Cassava Conservation Strategyto manage genetic resources for future needs. The foundation of the international centers (CIAT and IITA) that tookon cassava germplasm conservation on a global basis, gave considerable impetus to the interest and demand fromnational programs to invest in cassava and its improvement. Nevertheless, what scientists believe to be the bestmanagement of cassava and Manihot genetic resources has continued to evolve.A wide genetic base is not necessarily a prerequisite to short-term or even medium-term success of a crop. Both Asiaand Africa, by every measure, have a far narrower genetic base for cassava than the Americas, but the crop hassucceeded extremely well on both continents. But the need for broader diversity becomes ever more evident as timegoes on. For example, the challenges of evolving markets in Asia (especially for starch and ethanol), and changes inpest pressures in Africa (new races of cassava mosaic disease and recent outbreaks of brown streak disease) havebeen met with new genetic diversity from genebanks. We have seen many examples where weakness in a cropvariety is compensated by inputs such as chemicals to protect against pests, or nutritional supplements for a suboptimalnutritional content. This may lead to the false conclusion that broad genetic diversity is not essential forsustainable crop improvement. But genetic solutions to constraints in crop production or utilization are oftenoverwhelmingly superior to management that requires other types of purchased or labor-intensive managementinputs. The fact that there often can be a partial or complete genetic solution to problems is a strong incentive toassure access to appropriate genetic diversity.If it were possible to define today the genetic diversity that will be needed in the distant future, we could narrowlytarget the type of diversity to conserve, and let the rest disappear as farmers decide to switch to new hybrids. But it isclear from history that there are no completely reliable predictors of the specific genes that will be useful in the future,and the best sources of those genes are absolutely unknown today. Pests and diseases will not stop evolving, andwill always present new challenges to production practices. New markets, tastes and production practices likewiserequire new traits. There are many examples of needs for new genes that could not have been foreseen when thecassava collection was first established at CIAT in 1969, such as whitefly resistance or starch traits specifically suitedto specific products in the market place.This is why we preserve landrace varieties and related wild species of our crops, and why we expect to do so inperpetuity. We can have a reasonable level of confidence that greater success can be achieved when a wide geneticdiversity is securely conserved, understood, accessible, and wisely utilized.Vegetatively propagated crops are inherently more difficult to manage in genebanks than seed-propagated species,when the goal is for conservation of specific genotypes (clones). This requires continued vegetative conservation.Since all known cassava landrace varieties are highly heterozygous, the seed that results either from crossingbetween any two varieties, or from first generation selfing (S 1), will segregate widely. While some progeny mayresemble either of the parents in some or in many traits, none will be genetically identical to them. Vegetativeconservation may take several forms, including field or greenhouse-grown plants, slow-growth in vitro plantlets,protoplasts in culture media, undifferentiated tissue in culture media, or cryopreserved tissue. One of the keyrequirements is that the methodology must allow regeneration of whole plants that maintain genetic integrity, i.e. thatare genetically identical to the parent clone. To date, field-grown and in vitro plantlets best fit this criterion.The scientific bases for effective and efficient cassava genebank management continue to advance. We now havethe ability to conserve either vegetative or seed accessions at a very high level of security. However, as will bepointed out later, achieving this high security has been more difficult than anticipated; many institutions have found itdifficult to achieve conservation goals due to financial or human resources constraints. The tools and the theoreticalbackground are also advancing rapidly for understanding the genetic structure of germplasm, and applying thisinformation to conservation strategies. In vitro slow growth technologies are routine, when the infrastructure andmanagement expertise are available. Vegetative materials can be tested for most pests and pathogens, and they canbe eliminated when deemed necessary. Cryopreservation is being successfully applied to a wide range of genotypes,although more research is needed to further improve the recovery rate for broad and secure application as a9

Cassava Conservation StrategyTable 1. Summary of recommendations of the working group on cassava genetic diversity, of the first meeting ofthe International Network for Cassava Genetic Resources (IPGRI, 1994). aHuman resources development; train in:• Quarantine and pathology aspects• Conservation and exchange of germplasm• Pest and abiotic screening• Use of microcomputers in documentation• Analysis of genetic diversity by molecular and biochemical methodsLinkage with other network groups• Make formal contact with other groups of researchers or users whose interests and/or activities haverelevance to the Cassava Genetic Resources Network• Maintain continuing communication among the networksa Summarized by the author of the current report.Clearly, the International Network for Cassava Genetic Resources was an important precedent for the current effortsof the Trust and its partners. The recommendations, follow-up, successes and remaining tasks are highly pertinent tounderstanding the appropriate next steps to successfully conserving Manihot genetic resources in perpetuity. At thesame time, it is important to understand why this network had limited success, and to overcome those obstacles inthe present efforts.Both national programs and the international centers remain committed to the goals for Manihot genetic resourcesmanagement that have largely lain dormant for 15 years. At the same time, institutions are justifiably cautious aboutcommitting energy and resources without a high level of assurance that action will follow words. Ultimately, successcan only come to each program that makes a philosophical and financial commitment to this success. The currentglobal environment has a number of features that indicate new potential for progress. With high prices for agriculturalcommodities in general, and the inputs to produce them, there is a broad interest in renewing investment inagriculture as the best route to sustainable development, leading to adequate food for all and improved livelihoodswith minimal environmental impact. There is growing recognition that climate change is already affecting agricultureand our ecosystems, and that appropriate management of genetic resources – crops and non-crops alike – is vital tothe well-being of humankind. These motivating forces, along with the commitment of the Trust to support a rationalglobal system for genetic resources conservation, provide a renewed level of confidence in the ability of local,regional and global partnerships to succeed in these new goals for secure cassava and Manihot speciesconservation.1.4 Sources of informationThe principal sources of information for this report include a detailed survey sent by e-mail to the curators of most ofthe known cassava and wild Manihot genebanks around the world (Appendix III, IV and V; a consultation of expertsheld at CIAT headquarters in Cali, Colombia from 30 April to 2 May 2008 (Appendix I, II and VI), the literature, andpersonal contacts and personal experience of the author of this report.As of this writing, 34 surveys were returned for cassava, and three for the wild species (see more detail on sources inAppendix IV). Those for cassava are:AMERICAS: Bolivia, Brazil (three institutions), Costa Rica, Ecuador (two institutions), Guyana, Panama, Peru.AFRICA: Chad, Côte d’Ivoire, D.R. Rep. of Congo, Ghana, Guinea Conakry, Malawi, Mozambique, Niger, Nigeria,Sierra Leone, South Africa, Sudan, Swaziland, Togo, Zambia.13

Cassava Conservation StrategyASIA/OCEANIA: China, Indonesia, Malaysia, Papua New Guinea, Thailand, Vanuatu, Vietnam.INTERNATIONAL CENTERS: CIAT and IITA.For the wild Manihot species, surveys were returned from Brazil (EMBRAPA and the Universidade de Brasilia) andfrom CIAT.A survey is an inexpensive, but not entirely adequate, way of getting information on genebanks and theirmanagement. First, there are not many people who enjoy receiving or filling out surveys, especially extensive onesthat require some research or some thought. Many people simply will decline to respond, unless there is someassociated motivation, such as the possibility of getting future funding. Secondly, the person assigned to fill out thesurvey is not necessarily the person best qualified to answer the questions. Some people may want to provideanswers that give the best public view of their program. Alternatively, some may want to exaggerate problems, withthe thought that this may attract funding to improve genebank management. Thirdly, questions may be interpreted inslightly, or significantly, different ways, often because the language of the survey is not the respondent’s firstlanguage. This can also introduce errors, and make it difficult to make legitimate comparisons across all the returnedsurveys.2 Cassava in the global economy and agro-ecosystem2.1 Manihot species overview and the origins of cassavaManihot is a Neotropical genus, distributed in its natural habitat from the southern United States, throughMesoamerica, northern South America, and south through Brazil, Peru, Bolivia, Paraguay and Argentina. The latestmonograph by Rogers and Appan (1973) identified 98 species. The groups of species from the northern andsouthern hemispheres are markedly distinct. With the exception of M. esculenta, none of the northern Mesoamericanspecies are found naturally in South America, and only M. brachyloba occurs in both South and Central America. In amonograph still in progress, Allem reduces the size of the genus to 70 species, with 55 in South America and 15 inCentral and North America (Howeler et al., 2001).The species of Manihot are perennial and vary in form from acaulescent shrubs to trees with trunks 25 cm indiameter and a height of 10 to 12 m. They are generally sporadic in their distribution and never become dominantmembers of the vegetation. Most are native to dry regions, with a few in rainforest ecosystems. In general, thespecies appear to be shade-intolerant – capable of survival only with plenty of sunlight. They are not goodcompetitors with vigorous intercrops or with weeds. All the species are sensitive to frost, thus limiting their distributionto elevations below about 2200m. Since many of the species are found where long dry periods are common, theyhave evolved mechanisms of drought avoidance or drought tolerance. One of the most notable of these mechanismsis the production of storage roots where large amounts of starch are accumulated. In all species studied, thesestorage roots also contain the glucoside linamarin, which breaks down after cell injury to release prussic acid (HCN)(Rogers and Appan, 1973).The genus is clearly of New World origin, but further details of its evolution and distribution within the New Worldhave been poorly understood. Only since the 1990s, with the discovery of wild cassava, and the aid of molecularanalyses to examine relationships between the crop and the wild species, has there been better progress in definingan evolutionary history. Allem (2002) describes three important questions that need to be addressed concerning theobscure origins of cassava: botanical origin (i.e. the wild species which gave rise to cassava); the geographical origin(i.e. the area where the progenitor evolved); and the agricultural origin (i.e. the area of initial cultivation of the wildancestor by Amerindians).14

Cassava Conservation StrategyArchaeological evidence of cassava in northern South America indicates that its cultivation is of great antiquity there.Radiocarbon dates are much earlier than those from the Brazil/Paraguay region. But archaeological remains are rarein humid environments, so this does not give incontrovertible support to a northern South American origin. Studies ofmicro-fossils, such as starch granules in the case of cassava, allow us to partly overcome this problem. Nonetheless,since cassava was broadly cultivated in the New World since several thousand years ago, it is difficult to use thesparse archaeological remains to pinpoint the origin of the crop.There are few reliable phenotypic characters in the genus Manihot to indicate evolutionary relationships. Most of thespecies (including M. esculenta) show high intraspecific morphological variability. Because it was not possible toconfidently narrow origins with the use of morphology or archaeological evidence, a theory of multiple origins arose,but this was based less on positive evidence than on lack of evidence for alternative hypotheses.In what was to become the first insight into an entirely new perspective on cassava’s origins, Dr. Antonio Costa Allemof CENARGEN in Brazil, discovered a putative wild population of cassava in Goias state in 1982, described asManihot esculenta ssp. flabellifolia. Continued explorations showed that this subspecies was distributed in a zone oftransitional forest between the Amazon basin and the drier savanna to the south and east, including areas of thestates of Acre, Rondônia, Mato Grosso, Goiás and Tocantíns (Allem 1987; 1992; 1994).M. e. ssp. flabellifolia is similar to cassava morphologically, but cassava has greater root thickening, swollen leafscars and a stem morphology that is adapted to vegetative propagation (shortened internodes and thicker stems formore carbohydrate reserves). As with most Manihot species, M. e. ssp. flabellifolia is sporadic in its distribution; mostpopulations typically comprise fewer than 15 individuals.Early work with molecular markers to explore evolutionary patterns of Manihot indicated that South American andCentral American species form two distinct lineages, and cassava is more closely related to the South Americangroup. This work included RFLPs (Bertram, 1993; Fregene et al., 1994), AFLPs (Roa et al., 1997) and DNAsequences (B. Schaal, cited in Olsen, 2004).At the next level of molecular evolutionary studies, variations in SNPs (single nucleotide polymorphisms) and SSRs(simple sequence repeats) were used to explore cassava’s relationship to M. e. ssp. flabellifolia (Olsen, 2004). Thesestudies compared a presumed wide genetic diversity of cassava clones selected from CIAT’s core collection, andsamples from a range of M. e. ssp. flabellifolia genetic populations. The results appear to definitively place cassavawithin the range of genetic variation of the subspecies. Across the eight loci examined, the cassava clones contain anaverage of 18.8% of the total variation of the wild species. M. e. ssp. flabellifolia genetic variation is sufficient toaccount for cassava’s genetic diversity, without any need to involve a hybrid origin (Olsen, 2004). The composite ofevidence from molecular studies gives strong support to M. e. ssp. flabellifolia as the progenitor of cassava.Allem (2002) also provides interesting anecdotal evidence on the possibility that domestication of cassava from wildspecies is not that difficult and is in fact still taking place today in parts of Brazil. He proposes a transitional linkbetween cultivated cassava and its wild ancestor, in the form of a landrace called manipeba in northeast Brazil. Thislandrace (it is unknown how many distinct genotypes are involved) appears to be botanically and agronomicallyintermediate between wild and cultivated cassava, and as such gives a possible snapshot of the route to cassava’sdomestication.In nature, all the wild species appear to be principally seed-propagated. As a strategy for genetic resourcesconservation, there is probably little need for conserving individual genotypes through vegetative propagation. Inpractical terms, a strategy that combines seed, field and in vitro conservation will increase probability of success forconservation of many difficult-to-propagate species, as well as allowing field evaluations for traits of interest, andcrossing studies. CIAT has embarked on a detailed characterization of natural habitats of the wild species to betterunderstand their adaptation, and ultimately to tailor a conservation strategy to groups of species with similar15

Cassava Conservation Strategyrequirements (Jarvis and Guarino, 2008). This could also be an important part of a longer-term effort to produce aninventory of all existing Manihot populations.In summary, the accumulated evidence on cassava genetic diversity indicates the following 3 :• Manihot apparently originated in Mexico and Central America, and rapidly radiated in South America,perhaps after formation of the Isthmus of Panama (Duputié et al. 2008).• The genus Manihot appears to be recently evolved, without sharp genetic barriers among the species.• Cassava was domesticated in the Americas, most likely along the southwestern edge of the Amazonrainforest, from the wild species progenitor M. esculenta ssp. flabellifolia.• M. esculenta ssp. flabellifolia is distributed widely in the Americas, between the tropics of Capricorn andCancer.• The highest genetic diversity of cassava is in Brazil, with high diversity also noted for Central America.• A considerable amount of recombination continues to occur in farmers’ fields in some areas of Africa andLatin America.• Studies show high genetic diversity and low differentiation in all country studied, with the exception of agroup from Guatemala.• There is little substructure in American accessions (except for the Guatemalan group), but there is asubstructure in the African germplasm, explained by selection.• Neotropical accessions can be separated from African ones using molecular markers.• Historically, Asia received introductions from both Brazil (via Africa) and Mexico (via the Philippines).• Asia has far less genetic diversity and less differentiation compared to the Americas or Africa.• In modern times, massive introduction of diversity from the Americas to Asia by breeding programs hasgreatly enhanced diversity available for genetic improvement.• While much of the world’s ex situ germplasm has been evaluated for traditional major traits, there isprobably a wide array of unexplored variation, such as in traits for specialty markets (e.g., sugary, amylosefree,high protein, slow post-harvest deterioration).2.2 Production overviewCassava is the fourth most important supplier of food calories in the tropics. The principal economic product isstarchy roots, which are utilized in a wide range of end uses, most notably including human food, animal feed, andindustrial products. World production in 2006 of about 225 million tons is the energy equivalent of 80 to 85 milliontons of cereal grains (FAOSTAT, accessed 6/2008).The crop is important throughout the lowland humid and seasonally dry tropics. It extends to the limits of itsadaptation in the highlands of the Andean zone and East Africa (up to about 2200 masl), in the semi-arid tropicswhere rainfall may be as low as 400 mm annually, and into the subtropics, where its adaptation is limited by coolwinters and accompanying frost. The species has not succeeded much beyond the Tropics of Cancer or Capricorn,both because of the need for a long growing season, and also the difficulty of storing planting material for extendedperiods (during a cold winter, for example).In the last two decades, cassava’s importance has grown much more quickly in Africa than in either Asia or LatinAmerica. Nigeria alone plants an area 36% greater than the entire area planted in Latin America, and slightly morethan all of Asia. The three top producers in Africa – Nigeria, the Democratic Republic of Congo, and Mozambique –together plant about the same area of cassava as all of Latin America and Asia combined. In the past two decades,production has approximately kept pace with population increases in producing countries. Nonetheless, there arelarge imbalances among regions. In most of Africa, where the crop is utilized mainly for human consumption, yieldsare still well below those of Asia or South America (Table 2).3 Information taken mainly from presentations in the CIAT workshop, 30 April – 2 May 2008, except whereotherwise noted.16

Cassava Conservation StrategyTable 2. Production statistics for cassava, by region.Area Yield ProductionRegion/country (ha) (tons/ha) (tons)Latin America and CaribbeanArgentina 17,571 10.0 175,706Bolivia 36,858 10.1 373,612Brazil 1,901,561 14.0 26,713,038Colombia 180,000 11.1 2,000,000Costa Rica 20,000 15.0 300,000Cuba 79,648 5.6 450,000Dominican Rep. 15,435 6.1 93,609Ecuador 22,677 4.4 100,229El Salvador 1,288 12.5 16,102French Guiana 1,608 3.5 5,582Guatemala 6,279 2.8 17,578Guyana 2,995 11.1 33,294Haiti 71,270 4.6 326,821Honduras 5,075 4.0 20,300Jamaica 751 21.8 16,405Mexico 1,501 13.8 20,661Nicaragua 11,000 9.5 105,000Panama 2,341 11.8 27,693Paraguay 300,000 16.0 4,800,000Peru 86,000 11.0 945,000Puerto Rico 49 9.3 456Suriname 225 20.0 4,495Venezuela 41,641 11.7 489,047Sub-total: 2,806,835 13.2 37,041,521AfricaAngola 757,000 11.6 8,810,000Benin 173,450 14.6 2,524,234Burkina Faso 1,000 2.0 2,000Burundi 82,000 8.7 710,000Cameroon 350,000 6.0 2,100,000Cape Verde 11.7 3,500Central African Republic 190,000 3.0 565,000Chad 27,000 12.0 325,000Congo, Republic of 110,000 9.1 1,000,000Côte d’Ivoire 280,000 7.9 2,200,000D.R.Congo 1,845,510 8.1 14,974,470Gabon 45,541 5.1 231,816Gambia 2,500 3.0 7,500Ghana 790,000 12.2 9,638,000Guinea-Bissau 14.8 40,000Guinea Conakry 136,252 7.8 1,068,518Kenya 77,502 10.9 841,196Liberia 100,000 6.3 634,874Madagascar 388,779 6.1 2,358,77517

Cassava Conservation StrategyConsolidated estimatesEstimates of ex situ accessions based onof unique local landracedifferent sourcesEstimatedMainvarieties (excluding Consolidatedin situ Proposed Priority forduplicates and estimated landraceAreanationalaccessions minimum accessionInbreeding/experimental density (ha perplantedNg & GCDTprogramsmissing ex situ no. duplication Priority forIBPGR Ng survey b CGIARholdingmaterial) dlandrace) efrom CGIAR of accessionsg centers h collection i statusin CGIAR additional ITPGRFARegion/country (ha) acenters(1994) (2002) (2008)caccessions Ex situ In situ Ex situ In situ centers f1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Nigeria 3,810,000 417 435 40 547 NRCRI 500 800 7,620 4,763 253 200 ** SignatureRwanda 118,860 280 2 125 150 951 792 148 75 *** *Senegal 19,464 11 57 ISRA/CDH 10 50 1,946 389 50 25 * * RatificationSierra Leone 70,000 43 134 118 110 IAR 100 200 700 350 90 100 * AccessionSouth Africa 100 0 ARC 5 25 0 0 25 10 *Sudan 6,545 10 ARC/10 10 655 655 10 10 * * RatificationSSARTOSwaziland 10 10 20 20 10 * * SignatureTogo 135,820 734 209 176 ITRA 100 200 1,358 679 24 100 * RatificationUganda 379,000 200 413 14 RTCP 250 1,000 1,516 379 986 500 *** * AccessionU.R.Tanzania 670,000 215 254 3 RTCP 250 1,000 2,680 670 997 500 *** ** AccessionZambia 180,000 96 103 ZARI 75 200 2,400 900 200 150 ** * RatificationZimbabwe 46,839 6 6 25 7,807 1,874 25 10 ** Ratification0Sub-total: 12,110,694 2,498 6,398 1,272 2,112 3,743 7,480 3,236 1,619 5,368 3,675Asia – OceaniaCambodia 96,324 25 3,853 25 10 ** AcceptanceChina 265,800 28 86 4 2 SCATC/UCRI/GAAS10 15 26,580 17,720 13 20 * *Fiji Islands 2,223 6 6 5 25 445 89 19 20 *India 242,400 701 1,507 CTCRI 600 750 404 323 750 200 *** * RatificationIndonesia 1,222,814 157 251 130 136 CRIFC/MARIF150 1,000 8,152 1,223 864 500 ** ** AcceptanceMalaysia 37,719 55 92 52 61 MARDI 50 100 754 377 39 50 * AcceptanceMicronesia 1,100 25 44 25 10 *Myanmar 16,500 7 21 ARI 7 50 2,357 330 50 20 * Acceptance23

Cassava Conservation StrategyConsolidated estimatesEstimates of ex situ accessions based onof unique local landracedifferent sourcesEstimatedMainvarieties (excluding Consolidatedin situ Proposed Priority forduplicates and estimated landraceAreanationalaccessions minimum accessionInbreeding/experimental density (ha perplantedNg & GCDTprogramsmissing ex situ no. duplication Priority forIBPGR Ng survey b CGIARholdingmaterial) dlandrace) efrom CGIAR of accessionsg centers h collection i statusin CGIAR additional ITPGRFARegion/country (ha) acenters(1994) (2002) (2008)caccessions Ex situ In situ Ex situ In situ centers f1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16Papua NewGuinea12,500 95 PNG NARI 75 100 167 125 100 50 ** *Philippines 204,578 134 384 6 PRCRTC/IPB130 500 1,574 409 494 250 *** ** AcceptancePolynesia 1,327 0 25 53 25 25 10 * *Sri Lanka 23,560 56 128 CARI/PGRC 50 100 471 235 100 50 ** *Thailand 1,070,805 8 250 11 5 RFCRC 10 11 107,081 97,346 6 11 * * SignatureVanuatu 281 j 150 CTRAV-VARTC120 50 k 2 6 150 25Viet Nam 474,800 20 36 31 9 Hung LocAgr. Center20 50 23,740 9,496 41 25 * *Sub-total: 3,690,795 1,172 2,755 473 257 1,132 2,965 3260 1245 2,708 1,170WORLD 18,608,324 9,195 14,948 7,685 7,205 10,068 26,986 1848 690 19,954 14,791a Source: FAOSTAT (except where otherwise indicated); data for 2006, accessed 20 May 2008.bInformation provided by survey respondents indicated in Appendix III.cBased on reports from CIAT and IITA, in the 30 April – 2 May 2008 workshop at CIAT.d These are approximations based on region (primary or secondary center of diversity), area planted, ex situ accessions reported, and personal knowledge of the author about diversity in individual countries.eTotal area planted to cassava in the country (column 2), divided by estimates of landrace varieties (columns 8 and 9).fEstimates of in situ unique landrace varieties (column 9) minus number of accessions in CGIAR centers (column 6).g Estimated number of accessions that would be required to fully represent a country’s cassava genetic diversity. More accurate estimates will be possible as more molecular information becomes available ongenetic variation.h *=low; **=intermediate; ***=high; based on current availability in CGIAR centers, risk of loss in national program centers, and importance of diversity for cassava improvement programs.I *=low; **=intermediate; ***=high; based on expected importance of a country’s genetic variability available in situ, the amount of that variability already collected, and the risk of losses oflandraces in situ that are not already collected.j Source: CTRAV-VARTC.k There appear to be a large number of landrace varieties lost from farmers’ fields and home gardens in the past two decades due to declining cassava production.24

2.3 Evolution and farmer selectionThe genetic diversity that exists today in cassava germplasm is the consequence of natural selection over millennia,added to more recent farmer selection, and progress from breeding programs. One can only speculate as to the totalnumber of cassava clones cultivated worldwide. All current germplasm collections are sub-samples of the totaldiversity, albeit some more complete than others. Probably in only a few cases, with the greatest likelihood in someAsian countries, is it likely that all the landraces grown in a country have been collected and conserved in genebanks.In Latin America and the Caribbean, where the crop originated, there are probably on the order of 16,000 clones,based on numbers in existing ex situ collections, on reports from collectors and ethnobotanists from fieldobservations, and on more recent molecular information (see Table 3). In Africa, diversification seems to haveoccurred rather quickly, probably in response to the broad array of growing environments and market uses, but wasalso enabled by continuing small-scale introductions since the original introductions in the 16 th century. Because ofthe extent of cassava cultivation in Africa, the range of environments in which it is cultivated, and the knowndiversification through natural intercrossing of landraces in farmers’ fields, there has been an explosion of geneticallydistinct genotypes cultivated in farmers’ fields. It is still unclear, however, whether there are any genes in Africangermplasm which do not exist in Latin American landraces.The Collaborative Study of Cassava in Africa (COSCA) identified some 1,200 local varieties in 281 villages incountries representing 70 percent of the continent’s cassava. There are probably more than 7,000 distinct clonescultivated by farmers in Africa, excluding the minor variations that seem to have appeared through seedlingpropagation, which were later incorporated into variety complexes (see Table 3). In Asia and Oceania, the number oflandraces appears to be much more limited than in Latin America or Africa, perhaps on the order of 3,000 distinctgenotypes (see Table 3). Indonesia and India seem to be the main repositories of distinct landraces in Asia. AlthoughIndia has a larger ex situ germplasm collection, many accessions are the product of breeding programs.2.4 Modern genetic improvementInvolvement by trained plant breeders in cassava improvement began early in the twentieth century. It appears thatthe earliest programs were established in Brazil, India, Madagascar, Nigeria and Tanzania. Most were directedtoward starch industry interest in higher productivity. Some of these continue to modern times, but others werediscontinued after World War II. A renewed interest in cassava breeding followed establishment of root and tubercrop programs in two newly developed international centers in the late 1960s and early 1970s – CIAT in Cali,Colombia and IITA in Ibadan, Nigeria. Many national breeding programs were established or strengthened as a resultof support from these centers.Breeding goals across countries, as documented in regional workshops or symposia, appear to be remarkablysimilar. Nearly all programs include among their goals: high yield, high dry matter or starch, early maturity, toleranceto local pests and diseases, and adaptation to local environmental conditions. The widespread adoption of goals forstress tolerance and pest resistance reflects a recognition that most farmers apply few inputs to alleviate factorscausing yield and quality variations. In Africa, yield and resistance to the African cassava mosaic disease have beenthe primary breeding goals for many years, but after widespread success in developing resistance, goals havebroadened to include other traits, such as root quality. In Asia, there are few pest or disease constraints and attentionfocused on yield and starch content for industrial markets. The Americas present a broad range of major and minorconstraints, and goals have been more regionalized and more diverse.The establishment of the Cassava Biotechnology Network in 1988 was the first step in a long-term strategy to bringthe benefits of biotechnology to the most relevant research areas. This has now evolved into the Global CassavaPartnership, which held its first scientific meeting in Ghent, Belgium in July 2008. There are now a number ofgenetically transformed varieties in preliminary field trials, and it is expected that farm-level benefits from these newtechnologies will begin to become evident within a few years.25

Cassava Conservation StrategyCassava breeding has been a very successful enterprise, especially considering the relatively low support it hasreceived, in comparison to several other major world crops. While there are no detailed studies on the global impactof the benefits from breeding, it is known to be in the billions of US dollars (Kawano, 2003). Thailand is perhaps themost remarkable success, where nearly all the cassava area is planted to high-yielding improved varieties. Likewise,in the Americas and Africa, there are many examples of successful adoption and impact. But there are continual newchallenges and new needs as farming conditions, pest and disease pressures, and the market demands evolve.Breeders continue to rely heavily on landraces as sources of genes in improvement programs. As goals haveevolved more toward industrial uses of cassava (especially in Asia and Latin America), there is new interest inexploring the germplasm base for novel traits such as starch quality, micronutrient content, and rate of postharvestdeterioration. Cassava breeders universally appreciate and support comprehensive, well-managed genebanks as anessential part of their work, long into the future.2.5 Landraces and wild species at riskCassava may represent a unique situation compared to the major cereal crop species with regard to the relationshipbetween modern and traditional varieties (landraces). This uniqueness relates to the relatively recent, and low levelof, emphasis on cassava genetic improvement, as compared to other major crops. Most cassava production is stillbased on the planting of landraces, although this is changing quickly, especially in the past decade, and in selectedcountries like Thailand, Brazil, Colombia and Nigeria. There are no comprehensive global statistics on the adoption ofimproved varieties of cassava, but this is probably currently on the order of 20-25% of area where local landraceshave been replaced by selections out of breeding programs. Twenty-five years ago it was near zero, although therewere some notable early successes in extending improved varieties to farmers, such as from the InstitutoAgronómico de Campinas in São Paulo, Brazil, one of the world’s oldest continuing operating breeding programs(since the 1930s).The spread of new varieties (and displacement of local landraces) is fastest outside the center of diversity for thespecies – the Americas. Asia and Africa are seeing the highest rates of varietal replacement. Reasons for this arevaried. In the case of Asia, expansive industrial demand, beginning in the 1970s, along with the limited potential oflocal landraces, drove development of successful breeding programs and replacement of landraces by new hybrids.In Thailand, Asia’s largest producer, nearly all the current cultivated area is planted to the products of breedingprograms (Kawano, 2003). In Africa, the change from landraces to improved varieties was, and continues to be,driven by the devastations brought by diseases (especially cassava mosaic disease, and more recently cassavabrown streak disease), and the better resistance and yield potential offered by new varieties. Replacement oflandraces still appears to be low in Africa, but can become serious with progressive success of breeding andextension programs. One antidote to such genetic erosion, of course, is to have secure ex situ collectionsestablished. There is time, but perhaps not much, to assure the conservation of the world’s remaining cassavalandraces that are not yet in ex situ collections. This is a matter that merits close observation, documentation andaction. It is among the more important criteria for prioritizing further collecting.Landraces continue to be used extensively in breeding programs (i.e., used directly as parents in crossing nurseries),in contrast to several others of the major crop species, in which most breeding is done among advanced breedingmaterial, and landraces play little direct role. This means that cassava breeders in general have a very strong anddirect interest in the conservation and availability of landraces on an ongoing basis.As with the related species of many other crops, the main risks of genetic erosion for wild Manihot are theconsequence of the expansion of human activity into native habitats, mainly in the form of the expansion ofagriculture and urbanization. It appears that for Manihot, the former is by far the more pronounced. This relates to thenature of the distribution of the wild species – the largest center of diversity is in central Brazil, in the cerrado(savanna) region, where agriculture has expanded rapidly since the 1970s, as technology became available toprofitably produce crops, especially soybeans, sugarcane and citrus. For example, of 41 habitats identified and26

Cassava Conservation Strategysurveyed in the late 1970s, only one locality remained for wild Manihot twenty-five years later (Nassar, 2006).Habitat change is also leading to loss of wild Manihot in Mexico.3 Cassava-related networksThe formation of collaborative mechanisms that bring together a number of institutions working towards a commongoal, and with similar or complementary objectives, has proven to be a cost-effective means of enhancing theefficiency of research and development. Best and Henry (1994) noted four levels of cassava and cassava-relatednetworks, summarized and updated in Table 4.Table 4. Cassava and related networks.GlobalRegionalSubregionalNationalCassava R&D workersCassava Biotechnology NetworkCassava Genetic Resources NetworkInternational Society for Tropical Root Crops (ISTRC)Global Cassava PartnershipMOLCASAsian Cassava Research NetworkPanamerican Cassava Breeders’ NetworkLatin American Integrated Projects NetworkCollaborators in Root and Tuber Improvement and Systems (CORTIS)African Francophone Cassava Network (CORAF)African Branch of the ISTRCCLAYUCAEastern and Southern African Root Crop Research NetworkSouthern Cone Cassava Development NetworkCountry Root and Tuber Crop SocietiesSource: Modified from Best and Henry, 1994.International networks for cassava generally appear to be much appreciated by the participants, in view ofexperiences in participation and feedback. One of the key elements of success is clearly adequate andstable funding. Although, ideally, networks are funded by the participants who benefit, most cassavarelatednetworks have required continuing external funding due to the generally low level of resourcesavailable to cassava R&D institutions. In fact, few of the crop networks have realized stable funding, andtherefore their effectiveness has been variable over time. Some of the longer-term networks have beenable to get support from a series of different donors in order to remain viable. Others have evolved into newnetworks as priorities and funding sources changed. For example, the Cassava Biotechnology Networkevolved into the broader-based Global Cassava Partnership in 2004, in response to participant and donorinterest in better integrating biotechnology with a broader range of production, processing and utilizationactivities.Some of these networks help facilitate the conservation of cassava genetic resources; nearly all of them benefit fromit. The only network specific to genetic resources is the Cassava Genetic Resources Network, described earlier(Section 1.3), and no longer active as an organized entity. Nonetheless, many of the networks rely on and link togenetic resources activities. One of the key activities of CLAYUCA, for example, is to promote and facilitate the27

Cassava Conservation Strategyinterchange of germplasm among participating countries. The Asian Research Network has, since its beginning asthe Asian Cassava Breeders’ Network, had a strong emphasis on germplasm, through facilitating exchange ofselected materials among participating countries, and through duplication of national genebank accessions at CIAT.Perhaps the future of networking for cassava genetic resources conservation could usefully be left to these moreuser-focused groups.4 Cassava in the regional crop strategiesThe Trust commissioned the regional PGRFA networks around the world to develop regional strategies forconservation and use of PGRFA. Each of the following regional reports 4 listed cassava as a high-priority crop forconservation, though based on a somewhat different approach to defining priorities: the Americas; Eastern Africa;West and Central Africa; SADC region of Africa; the Pacific; and South, Southeast and East Asia (reports available athttp://www.croptrust.org/main/regional.php?itemid=83).4.1 AmericasDuring several working sessions, the network coordinators and other stakeholders defined several levels of criteria toprioritize crops and collections: criteria based on the principles of theTrust; criteria used to define important crops;criteria for assessment of the importance of the collections; and criteria for assessment of the quality of collections.From these criteria they developed a weighted scoring system, described in the report as “far from a solid scientificprocess, but nonetheless provides a useful entry point for prioritization and a reference milestone for futurediscussions.” In this assessment, the ten top priority genera/species, in order of weighted scores, are: Phaseolus,Capsicum, Zea, Manihot esculenta, Arachis, Glycine, Lycopersicon, Theobroma, Musa and Citrus.Table 5. Important collections in the Americas.Country Network InstitutionBolivia TROPIGEN/REDARFIT El VallecitoSeveral CAPGERNet SeveralBrazil TROPIGEN EMBRAPABrazil TROPIGEN IACBrazil TROPIGEN IPAGROColombia REDARFIT/TROPIGEN CORPOICA-CIATPeru REDARFIT/TROPIGEN INIEAParaguay REGENSUR DIAVenezuela REDARFIT/TROPIGEN INIAVenezuela TROPIGEN UCV-FAGRO4.2 Eastern AfricaThe EAPGREN Regional Steering Committee ranked crops based on: threats of genetic erosion; center of diversity;uniqueness; food security; status of PGR collections – health; status of characterization; conservation facilities – longterm; safety duplication; and regeneration needs. Cassava ranked 7 th , just behind banana (#5) and sweet potato (#6),and just ahead of rice (#8) and yam (#9). It is listed as important in Burundi, Kenya, Madagascar, Rwanda, Sudanand Uganda. It is noteworthy that several of the vegetatively propagated crops are grouped together among the top4 At this writing, the West Africa Report is brief and does not list priority crops. However, the overwhelmingimportance of cassava in this region makes it evident that the crop should be near the top in terms of priority forconservation of genetic resources.28

Cassava Conservation Strategypriority crops. In this region, the collection of most importance was noted as that of NARO, Uganda (300 accessions;50% of accessions with passport data).4.3 SADC RegionThe Southern African Development Community (SADC) comprises 14 member countries. Most are importantcassava producers. The SADC Plant Genetic Resources network has a broad mandate to maintain geneticresources of the region. The mandate list has over 3,000 species, out of which 27 are included in the Annex 1 of theITPGRFA. Cassava is among the twelve Annex 1 crops in the SADC region considered important at the regionallevel. The report lists several countries with cassava collections: Malawi, South Africa, Swaziland, and Zambia.The SADC report lists a number of priority areas for upgrading and capacity building. Most important among these inthe case of cassava are:1. Support to field genebanks for vegetatively propagated crops such as sweet potato, cassava, banana, potato,yams, and aroids, as well as coconut.2. Support to in vitro base collection of the priority field crops mentioned above.4.4 South and Southeast AsiaThe National PGR Coordinators defined criteria for crops of greatest importance to the agriculture of the SSEEAregion, or to a few countries in the region, as follows:• Center of diversity (primary or secondary)• Level of sub-regional, regional and global importance as food and nutritional crop (including feeds andfodder for animals)• Presence of regional and /or international collections• Usefulness as crops for marginal areas and subsistence agriculture• Livelihood security for smallholders• Threat to the genetic diversity in situ/on-farm• Crop with unique advantage to the sub-region or regionTable 6. Examples of priority crops and ranking assigned by individual sub-regional network.SANPGR RECSEA-PGR EA-PGRRice 1 1 2Wheat 3 1Maize 2 3Potato 15 10Sweet potato 19 3Cassava 18 6Source: SSEEA Regional Report, Table 5.The SSEEA report provides an integrated ranking among the sub-regions, where cassava is ranked as 13 th in termsof priority crops for support. Countries noted as requiring this support are Indonesia, Malaysia, the Philippines. Theinformation was further discussed to define collections of greatest importance and priority for support based on thefollowing criteria (Section 13 of SSEEA report):• Collections in the public domain• Distinct collections (landraces, wild relatives)• Collections with no safety duplication• Collections under threat• Collections with specific traits and from specific ecologies• Collections that meet all the eligibility criteria29

Cassava Conservation Strategy• Collections with sufficient eco-geographical representation/Size of the collections• Collections from institutions where regional/international collaborations are on-goingBased on these criteria, collections from the following countries were recognized as important: India, Indonesia,Malaysia, the Philippines and Sri Lanka.4.5 Pacific IslandsCassava is widely distributed throughout the Pacific Islands, but overall is not a major crop. The entire region plantsonly some 15,000 hectares (2006 data, FAOSTAT). However, it is listed as a priority 2 crop in the regional report,coming in behind coconut, sweet potato, banana/plantain, aroids, yam and breadfruit, which are ranked as firstpriority. Cassava is most important in the Cook Islands, Vanuatu and Papua New Guinea. Its importance is related todiversity (some evidence of unique material) and to income generation (an important cash crop in some places, anda developing export market in a few). Germplasm conservation and use activities in the Pacific are coordinated byPAPGREN – the Pacific Agricultural PGR Network. Field cassava collections are noted for: Cook Islands, FederatedStates of Micronesia, Fiji, New Caledonia, Palau, Papua New Guinea and Vanuatu. In addition, there is a regionalcollection of 28 accessions maintained in vitro at SPC-RGS, Suva, Fiji.4.6 Summary from the regional reportsThe working group for the Americas used a useful approach to regional analysis – a SWOT analysis to help describethe region in terms of PGRFA. The following is adapted from that analysis, applied strictly to cassava, and globally.STRENGTHS• Large number of genebanks with ample diversity of landraces.• Major IARCs to partner with: CIAT (headquarters in Latin America and regional office for Asia in Bangkok)and Bioversity International (various regional offices), IITA in Nigeria.• Some large well-established PGRFA programs in conservation and sustainable utilization, e.g. Brazil andIndia.• Regional networks on each continent that support PGRFA conservation.WEAKENESSES• Communication difficulties – issues related to language, infrastructure, documentation systems,geographical access, and logistics.• Vast geographical areas from the tropic of Cancer to the tropic of Capricorn, and including the range fromvery wet to very dry, lowland to highland regions, and wide soil variations, among others.• Limited knowledge and public awareness of genetic diversity issues.• Differences in policy approaches and the involvement in, or ratification of, international agreements andinstruments (ITPGRA, CBD, etc.).• Political stability and related issues.• Not all countries are parties to the ITPGRFA and thus may not be eligible for funding without agreeing to a“Solemn Undertaking”.• Limited human resources.• Inadequate germplasm enhancement efforts.OPPORTUNITIES• Large number of partnerships possible – bilateral and multilateral.• Well established sub-regional networks, but with varying degree of operation and coordination.• Creative skills of network members in solving problems.30

Cassava Conservation Strategy• Considerable potential for new uses – for direct and indirect use in markets, communities, farmers’ groups,etc.• Commitment to common goals, particularly as they relate to the goals of the GPA and the Global CropDiversity Trust.THREATS• Shortage of financial resources – unsustainability.• Range in development of technical expertise – variable levels between countries and sub-regions.• Range of infrastructure development – many of which require substantial input and/or investment.• Environmental variables, e.g. hurricanes, flooding, extreme heat, low temperatures, extreme wet and dryconditions.• Rapid genetic erosion, especially in areas of agricultural development and urbanization.• Safety duplication is not complete.The regional reports clearly indicate the high importance of cassava among priority crops for support for conservationin the Americas, Asia and Africa. The remainder of this report highlights general and specific goals and strategies toaccomplish secure global conservation.5 Overview of cassava collection and conservation5.1 Collecting strategies, techniques and prioritiesWe now know that new variation is continually arising through incorporation (intentional or not) of seed-derivedvolunteer plants in cassava plantations. Each of these is a new genotype. A variety may actually in some cases be amixture of similar genotypes rather than a clone. It would be very difficult and expensive to collect and conserve asample of every existing genetic variant of cassava or of the wild Manihot species. What to conserve and how toidentify and collect these materials are two of the most fundamental questions in developing a conservation strategy.The techniques and the process of cassava collecting, like those of most crops, have advanced considerably in thepast 25 years. Most early collecting was done simply by visiting villages in known cassava-growing regions andtaking a few stakes from each of what appeared to be distinct varieties, based either on morphological differences inappearance or information provided by the grower. Usually, the collector recorded the date, the name of the village,the name of the grower, and the local name of the variety. This is the extent of passport information for the largemajority of genebank accessions around the world. A brief history is given here of the early establishment of CIAT’scassava genebank, as a baseline for understanding how many collections began and evolved.CIAT’s first annual report (1969) stated among the goals of its newly established root and tubers program, “to exploreand collect cultivars and related wild species of Manihot in the countries where variability is present, with emphasis inthe primary centers of origin (Northern South America and Middle America), in order to establish a germplasm bankrepresentative of the world’s variability.” (p. 43). In fact, as noted earlier, it is now widely believed that cassava’sgreatest diversity is in Brazil. In any case, it was certainly understood at the time that there was major variability inBrazil, but importation of cuttings into Colombia was prohibited due to concern about introducing coffee rust (Nesteland Cock, 1976). Nonetheless, based on this goal, CIAT began the systematic collecting of cassava landraces inColombia, in May of 1969, in collaboration with the Secretaria de Desarrollo y Fomento del Valle. This organizationappointed Mr. Victor Manuel Patiño, director of the Cauca Valley Botanical Garden, to work with CIAT in thecollecting. In this initial phase of collecting, a total of 611 accessions from 20 departments were assembled andestablished on the CIAT farm near Palmira, Valle. The following year, the collection was extended to other countries,and by the end of 1970, the genebank consisted of the following accessions:Colombia 1,88431

Cassava Conservation StrategyEcuador 123Puerto Rico 60Panama 118Peru 8Venezuela 330Mexico 70It was already recognized at the outset that phytosanitary care would be critical for germplasm management, andCIAT initiated a cooperative project with the plant pathologists of ICA (CIAT, 1969). 1971 brought an outbreak ofcassava bacterial blight (not yet identified), and in order to prevent the spread of the disease to commercial andexperimental plantings, there was an intense eradication of material from the genebank, with attempts to recoverclean planting material through cutting plants back to just the woody stems, where there appeared to be lessprobability of bacterial inoculant (CIAT, 1971). By 1972, it was possible to eradicate the bacterial blight pathogenfrom the field collection, but there had already been significant losses from the collection. Most of these lossesoccurred in the earliest years as the techniques for management of pathogens were being developed. Nonetheless,there were continuing losses over the years, especially among accessions that originated from environments thatwere very different from those of the CIAT station where the genebank had been established in the field.CIAT’s breeding program relied very heavily on this initial set of introductions, through its first decade. Thereafter,when in vitro transfer techniques became available, the CIAT genebank steadily increased its accessions andbroadened the genetic base of materials used in breeding. Most notably among the newer introductions was asizable introduction from Brazil. By the mid-1980s to early 1990s, the collection had broadened to include modestnumbers of accessions from Asia and Africa, though the majority of these were bred materials rather than landraces.CIAT initiated an in vitro genebank in 1978, soon after the technology for slow growth in vitro culturing of cassavawas developed at the University of Saskatoon, in Saskatchewan, Canada by Kartha and Gamborg (1975). Since1998, CIAT’s Genetic Resources Unit has further extended the subculturing to 12 to 20 or more months. Clonessubcultured under this system have been found to be genetically stable (D. Debouck, pers. comm.). However, it wasnot until the early 2000s that the decision was taken to eliminate the field collection, due to a combination of factors.These included cutbacks in research budgets at CIAT, increasing difficulties of keeping pest and disease pressuresunder control (especially whiteflies and frogskin disease), the increasing security of the in vitro techniques and thecompletion of the in vitro collection, and finally, the establishment of a black box duplication arrangement with CIP inPeru. In addition, by the late 1990s, the cassava breeding project had finished with the routine morphoagronomiccharacterization of the collection. It appeared at that time that there might be less urgency to evaluate the collectionfor new traits. However, this philosophy changed somewhat as it became apparent that the cassava world wasmoving toward the development of new markets, and there were new needs to evaluate the germplasm, especiallyfor variations in starch quality.In 1982 an IBPGR-sponsored working group proposed a collecting format that greatly expanded the information to berecorded in the field, to include a broad array of information from the grower about the traits and performance ofindividual varieties (Gulick et al., 1983). These forms, or something similar, have been used in many subsequentcollecting expeditions. Nonetheless, probably fewer than 1000 accessions worldwide have this broader information.While the expanded passport information would be useful, it does not seem to be justified to re-collect extensively inareas already collected over the past decades, in order to obtain this information. The complexity and cost ofintegrating recollected materials into existing genebanks would be very high.Collecting additional cassava landraces is a critical part of a long-term conservation strategy. Priorities should bebased primarily on: (1) genetic diversity; (2) gaps, and (3) areas in danger of genetic erosion. Globally, it is estimatedthat some 5,000 more landrace varieties should be collected in order to fully represent the species genetic diversity(Table 3). Priority countries, based on the above criteria, are: Bolivia, Brazil, Colombia, Haiti, Nicaragua, Peru,Venezuela, Democratic Republic of Congo, Uganda, Tanzania and Mozambique. Collecting in these countries should32

Cassava Conservation Strategymake full use of the lessons from past collecting in terms of information to collect, as well as the available databasesand technologies such as GIS, on-site in vitro preparations, and on-site virus indexing.There is now good evidence that the incorporation of new genetic variability through volunteer seedlings may beeither a conscious or an unconscious practice, depending upon the level of knowledge of farmers and their traditions(McKey, 2008; Reichel-Dolmatoff, 1996). It is thus clear that cassava will continue to evolve in farmers’ fields, whichwould indicate the need to plan for continuing future collections. However, the rate of evolution slows with advancingagricultural production technology. The opportunities for mutant types to be selected and propagated by farmers isminimal in situations where known, modern varieties are cultivated in monoclonal situations, and the selection andpropagation of seedlings (seed-derived plants) does not consciously occur. Nonetheless, evolutionary forces willcontinue to produce new genetic variability in the species, in some locations, as they have done for thousands ofyears. Therefore, the secure conservation of currently existing landrace varieties would reduce the need for futurecollections to very sporadic and narrowly targeted expeditions. Completing collecting goals in the next decade wouldreduce to very low levels the need for future investments in this area. At the same time, it would not eliminate theneed to continue to monitor on-farm change in varieties, and the positive or negative effects of changes in geneticdiversity.Elite breeding lines may or may not become successful varieties. Those that do not reach farmers’ fields may be lostunless specific steps are taken to preserve them. If they are not successful, there may be little reason to save them ingenebanks, unless they are known to have specific important genetic value, and/or are difficult to obtain (e.g.interspecific crosses). For those that are successful, there is a need for systematic preservation so that a permanent,pure representation of the variety exists – the equivalent of breeder’s seed in a seed-propagated crop. Thisconservation may be the responsibility solely of a breeding program, or may be managed jointly by breeding andgenetic resources efforts. These stocks can be the basis of a seed program, for clarifying any possible futureproblems related to varietal contamination and as a source for distribution to other gene banks or breeding programs.The definition of “elite” is the key to a sensible conservation strategy. Only a very limited number of materials can beassigned elite status, or the costs involved in conservation quickly get out of hand. At CIAT, for example, a clonebecomes elite only after passing through all preliminary stages of selection, and multi-site selection in advanced yieldtrials for at least two years. On average, 10 to 15 clones a year enter the elite category. Even this relatively smallnumber can eventually become burdensome for conservation, and this group may be given a lower managementintensity (e.g., fewer replicates). Many countries include elite breeding material or experimental lines in theirgenebanks, but the criteria for inclusion are not well-documented.Genetic stocks can be temporary or permanent parts of a collection, depending upon objectives. CIAT (1994)reported on incorporating a mapping population into the germplasm collection as a way to ensure that it would bewidely available to participants in the Cassava Biotechnology Network. Stocks for a specific, limited study may notneed to be preserved at all.For small collections, all accessions can normally be treated with an equally high priority for conservation. In largercollections, one may gain efficiencies by assigning levels of importance to different groups and managing theirconservation distinctly. Local landraces are nearly always the top priority. Their conservation must be secured. Thismay be by two separate field locations, duplicate in vitro collections, or a field and an in vitro collection, for example.If a core collection (see later section) has been defined, this may get the highest of all priorities. Breeding lines andintroductions, especially if retained in a collection in the country of origin, may be given a lower status forconservation.5.2 Conservation alternatives for cassavaThe gene combinations found in cassava landraces are normally the result of many decades or even centuries ofselection by farmers. Since cassava is highly heterozygous, the only means of conserving the specific genecombinations of landrace varieties is through vegetative propagation. For some purposes it is necessary to maintain33

Cassava Conservation Strategythese specific gene combinations. If the goal is to select germplasm accessions to recommend directly to farmers,vegetative conservation is a necessity. For species or clones that rarely or never flower, there may be littlealternative to vegetative conservation, until such time as flower induction is possible. Currently, every cassavagenebank in the world conserves accessions in vegetative form. The international collections maintained as clonalmaterial by CIAT and IITA are registered into the Multilateral System of the International Treaty.Alternatively, if the interest is conservation of genes rather than genotypes, germplasm may be maintained as trueseed. Germplasm maintained in seed form would ultimately be useful principally as a source of genes in a breedingprogramme and not directly as a source of varieties. The exception could be if there were means of inducingapomixis, which would duplicate the exact genetic structure of the parent clone, and produce the same result asvegetative propagation.5.2.1 FieldCassava collections have traditionally been maintained in field plots. Stem pieces are used as the propagules just asin commercial production. Theoretically, such a collection could be maintained for many years without regeneration.In practice, maintenance problems often increase after a year or two, making replanting at more frequent intervalsnecessary. Common problems include lodging from excessive growth and build-up of pests and diseases. Adaptationproblems typically occur when the edapho-climatic characteristics of the genebank location are very different from thecollection site, where the variety is presumably adapted well enough to be selected and propagated year after yearby the growers.In general, field-managed material is not available for international shipment, which is a major limitation. Majoradvantages of field maintenance of collections are their technical simplicity and the availability of planting material forevaluations, breeding nurseries, or other uses.The following general recommendations apply to field conservation:(1) The area where materials are maintained should be as free as possible of diseases and insect pests thatcould cause losses of clonal material or create difficulties in the transfer of clean planting material to other sites.(2) A minimum of three to five plants is necessary for practical maintenance. If a plantation is also to be used asa source of production of stakes for planting of other trials, more plants may be required.(3) Cassava can be maintained in field plantings as a perennial plant, but periodic renewal every one or twoyears is desirable to avoid problems of excessive vegetative growth, cumulative disease and insect problemsand to facilitate maintenance generally.(4) The distance between plots of different clones should be adequate to prevent undue competition among theplots.(5) Accurate plot and accession labeling are crucial for long-term reliability of information on cassavagenebanks. Because plants may remain in the field two or more years, durable labels are important.In order to combine the benefits of lower space requirements with continual availability of planting material forexperimental use, CIAT devised a slow-growth system based on restricting root development in small planting pots(bonsai effect). Plants occupy only a small fraction of the space they would occupy if allowed unlimited growth in thefield.Maintaining a cassava germplasm collection in containers has the potential advantages of space savings; betterprotection against pests, diseases and weather-related damage; and labor savings. Disadvantages can includedifficulty in using plants as a source of planting material for field trials (generally small and weak stems), cost ofinfrastructure, and cost of materials.34

Cassava Conservation Strategy5.2.2 In vitroIn the mid-1970s, the University of Saskatoon (Canada) (Kartha and Gamborg, 1975) and CIAT developedtechniques for routine in vitro conservation of rooted plantlets of cassava. These plantlets can be derived in a numberof ways, but for phytosanitary reasons the recommended source is small meristem tips. These can easily be surfacesterilizedagainst superficial organisms, and many systemic pathogens do not advance into the new tissue of arapidly growing meristem. Extra precautions of chemo- or thermotherapy can also lower chances of contamination.Meristem tips are cultured in nutrient media in glass or plastic jars or test tubes, and maintained under controlled lightand temperature conditions. Under minimum growth conditions, cultures can be maintained 12–18 months beforerenewal. Renewal can be by planting stem pieces or meristem tips from the in vitro plantlet into new sterile media,without the need for a field propagation phase. Recent experiments with silver nitrate in the media show promise offurther extending the regeneration period for in vitro plantlets (Mafla, 2008). CIAT's facilities have the capacity to holdmore than 6,000 accessions in vitro at 20°C (day)/15°C (night) temperatures, 12-hour photoperiod and 500 to 1,000lux illumination.CIAT monitored genetic stability of in vitro cultures using a combination of morphological and biochemical traits, andDNA markers. All results have so far been negative, indicating a high level of genetic stability after as many as 15years of in vitro conservation and regeneration (CIAT, 1994).5.2.3 CryopreservationLiquid nitrogen storage of vegetative tissue tips should be the most secure and trouble-free system for conservationof clonal cassava germplasm. The major advantage is the virtual freedom from maintenance problems duringstorage, with the possible exception of low rates of mutation caused by background ionizing radiation. Conservationcould theoretically be carried out indefinitely with no need for renewal. Development of successful cryopreservationtechniques has been somewhat slow and sporadic, due to limited funding, as well as what appears to beconsiderable variation among genotypes in the recovery response. Various laboratories developed basiccryopreservation techniques for meristem tips in the 1980s, using chemical dehydration and programmed freezing inliquid nitrogen. With later developments in encapsulation and quick freezing, more than 80% of accessions tested atCIAT (mainly from the core collection), have recovery rates of greater than 30%. The minimum acceptable level for along-term conservation strategy is still a matter of some debate. CIAT’s Genetic Resources Unit believes this levelshould be at least 50% (D. Debouck, pers. comm.). Preliminary observations have shown no noticeable changes inplant characters after cryopreservation. However, a cryopreservation strategy would need to include periodicmonitoring of stability, using precise molecular measures.Somatic embryos already represent an efficient regeneration system for rapid propagation, and are a target fortransformation. They have the potential to serve as the basis for germplasm conservation as well, especially if theycan be adapted to and recovered from cryopreservation. Mycock et al. (1995) and Stewart et al. (2001) successfullycryopreserved somatic embryos, with a 40–60% post-thaw viability. Danso and Ford-Lloyd (2004) introduced newcryoprotection and dehydration techniques and obtained 95% post-thaw viability (albeit, with a limited range ofgenotypes). Rate of plant recovery from the cryopreserved embryos was comparable to that of non-preserved ones.The optimal protocol involved induction of embryogenic calli on an induction medium (Murashige and Skoog mediumsupplemented with 2,4-D and sucrose), cryoprotection on 0.3 M sucrose for 21 days, followed by 16 h of dehydrationand immersion in liquid nitrogen. Although plants recovered from somatic embryos appeared to be genetically stable,this needs to be further tested and monitored. Current evidence suggests that sucrose cryoprotection followed by airdesiccation provides a viable solution for long-term conservation of cassava genetic resources via cryopreservation.Cryopreservation research for cassava should receive continuing and increased support, given its potential tocontribute to long-term conservation for additional security.Cryopreservation is certainly a technology that is amenable to continued improvement, but the rate of progress in thepast 20 years has been disappointing. This is in part because of sporadic funding and in part because there justseems to be a set of germplasm (cassava) that is stubbornly recalcitrant. The good news is that there seems to be a35

Cassava Conservation Strategynew optimism about overcoming some of the hurdles in dealing with the recalcitrant types. The other silver lining isthat this is one more indication of the fantastic genetic variation that exists in the cassava landraces.There is little doubt that a protocol will eventually be developed whereby it will be possible to recover all, or nearly all,accessions. So there are two components that are relevant to the cassava conservation strategy: first, the continuingresearch to fine-tune the technology, and secondly, determining how it fits into an overall conservation strategy. Withabout 75% of accessions estimated to have an acceptable recovery rate of over 30% (Gonzalez-Arnao et al., 2007),it seems we should be getting close to the point of beginning to incorporate cryopreservation into a conservationscheme. However, taking this next step will be rather expensive. It would involve screening the entire collection oflandrace varieties for their recoverability under the established protocol. Given that there seems to be optimism aboutthe near-future potential for good progress in improving recoverability of the recalcitrant accessions (Gonzalez-Arnaoet al., 2007 and Escobar, April/May CIAT workshop presentation), it is probably wise to still wait some time before themass screening of the collection. Perhaps once the recovery rate of the core collection reaches about 95%, then itwould be an appropriate time to begin the mass screening to determine the recovery rate of each clone. The set ofclones with acceptable recovery rates could then become a secure backup to the collection. At this point, this groupmight replace the current black box in vitro collections at CIP (CIAT collection) and at Cotonou (IITA collection). Orthey could also easily be moved to another institution as well. It is important that the in vitro base collections and thecryo collections not be held in the same location, if they are to play their full role in secure conservation. Those +/-5% of clones that are still classified as recalcitrant would have to continue in some scheme of secure backup such asan in vitro collection, while research on the freeze/recovery protocol continues. A time-frame for all of this is ofcourse difficult to predict, however, it seems that we should be looking at this type of system to be in place withinseven to eight years.5.2.4 SeedMost seed storage is done by breeders in their work of crossing and genetic improvement. Seed storage as a methodfor germplasm conservation in cassava has received limited attention. Varieties have been selected and propagatedvegetatively to preserve specific gene combinations. After self- or cross-pollination, these genes are re-assorted intonew combinations. Seed conservation can be a means of preserving genes, but not the specific combinations of theparent clone(s).Cassava seeds are apparently orthodox in behavior and therefore can be stored under conventional conditions of lowhumidity and low temperatures (Ellis et al., 1981). IITA (1979) reported storing seeds at 5°C and 60% relativehumidity for up to seven years with no loss in germination ability. Seed can also be preserved in liquid nitrogen andrecovered with high viability (Mumford and Grout, 1978; Marin et al., 1990). Manihot wild species have not beensystematically tested for seed storage behavior, a critical need as collections proceed, along with regeneration ofseed from field-grown plants.Although the mechanics of seed storage appear to be straightforward, further studies are needed to defineappropriate methodologies from the standpoint of germplasm conservation theory. Various approaches are possible,including uncontrolled open pollination, selfing, or pollination among selected accessions. In each case, the numbersof seeds needed for a defined level of probability of conserving an adequate gene pool (preventing genetic drift) needto be defined.The simplest method would be open pollination. This would perhaps be an appropriate means of conserving a broadpool of genes in the case of some catastrophe that resulted in loss of clonal accessions. From the point of view ofpreserving the genetic integrity of an accession, the best approach would appear to be selfing, which retains most ofthe genes of the accession of interest without intermixing the genes of other accessions. In fact, from the plantbreeder’s point of view, seed from selfed genebank accessions may be considerably more valuable than theaccessions themselves as parents. Selfing should eliminate some level of deleterious genes, although the detailedstudies have not yet been done to quantify these effects. Because of the partial homozygocity of S 1 plants (50%, onaverage), there is a higher breeding value, i.e. greater likelihood that expressed traits (the phenotype) can be36

Cassava Conservation Strategytransferred to progeny, which of course is to great advantage of the efficiency of a breeding program. Pollinationamong selected accessions could be very complicated to design appropriately, such that flowering among selectedmaterials is synchronized, and that the particular crosses chosen are those that are most appropriate in terms ofgenetic resources conservation.For any pollination system designed for germplasm conservation, there would be a very large advantage to havingthe ability to induce flowering -- either to produce flowers in clones that normally do not flower, or to regulate thetiming of flowering. The limited research in the area of flower induction has had only moderate success.The long-term advantages of seed conservation warrant further work in all these areas. Initially seed production andstorage would be an additional cost, in combination with some form of vegetative conservation such as in vitro slowgrowth or cryopreservation. However, eventually it should be possible to partially replace some of the vegetativelymaintainedaccessions with seed collections, for a net cost savings.Conservation of the wild Manihot species as seed requires some different approaches. Since all the wild speciesseem to be seed-propagated in the wild, seed regeneration ex situ can be based on natural systems of crosspollination within heterogeneous populations. However, studies on pollination agents and isolation distances arepractically non-existent for Manihot. Systematic seed propagation and conservation of the Manihot wild species willrequire a number of studies on pollination behavior and seed response to different storage regimes.5.2.5 PollenCryopreserved pollen should be a good system of preserving populations of genes. Like selfing, this system wouldallow sampling the genes within each accession, without introducing genes from others. From a breeder’sperspective, pollen has the advantage that it can be used almost immediately, as opposed to seed or in vitroconservation, both of which require a cycle of regeneration, to the stage of producing flowers.One of the limitations to research on cassava pollen conservation remains the difficulty of viability testing. Neitherstaining nor in vitro germination is adequately reliable as an indicator. Protocols for efficient, large-scale and rapidviability testing will be a necessary prerequisite to effective pollen conservation. Leyton (1993) resorted to in vivopollination as a means of testing viability in a series of experiments on pollen cryopreservation. He obtained no seedsfrom any sub-zero pollen treatment (-4°, -12° or -70°C). Orrego and Hershey (1984) were unsuccessful in storingviable pollen after desiccation over silica gel.5.3 Conservation costsDesign of a conservation strategy needs to consider, first, the best technical approach to safely preserve the genesand genotypes of the collected materials, and secondly, the cost-effectiveness of that approach. Advantages of invitro conservation are the low space requirements and minimal possibility of loss of materials through diseases,pests, weather or soil factors. Disadvantages are the need for relatively sophisticated facilities for culturing sterileplantlets, and for maintaining reliable conservation conditions. Costs of field versus in vitro conservation are highlylocation-specific, depending upon local costs of labor, energy, supplies and infrastructure. Economies of scale arealso a factor. For most small national collections, in vitro conservation may not be justified, unless the laboratoryforms part of a conservation facility serving other crops as well. Cryoconservation, at this stage of its development, isprobably only appropriate for more advanced programs that are able to carry out the continuing research necessaryfor fine-tuning the technology. As procedures become more routine, cryoconservation should be more broadlyapplicable, especially to medium and large national programs.Conservation of vegetatively propagated species has always been laborious and costly relative to seed conservation.Koo et al. (2004) carried out a comprehensive study comparing costs of maintaining field, in vitro and cryoconservedcollections at CIAT (Table 7). Total costs per accession for conservation alone (excluding distribution) werecomparable for in vitro and field (US$10.34 and US$7.18, respectively), while cryoconservation with regeneration is37

Cassava Conservation Strategymuch more expensive, at US$40.31. The advantages of cryoconservation enter the picture as regeneration time isextended, since maintenance costs alone are very low. Costs, security and convenience will dictate differentstrategies in different situations.Table 7. Comparison of key costs (US$ per accession, 2000 basis) for conserving and distributing acassava accession for one year in CIAT’s genebank.ExistingaccessionsIn vitroNewaccessionsWithoutregenerationCryoconservationWithregenerationFieldgenebankCost categoryConservationStorage a 1.28 1.28 0.86 0.86 7.18Subculturing b 9.06 9.06Viability testing 7.96Regeneration (cryo.) 31.49Disease indexation 57.27Conservation total10.34 67.61 0.86 40.31 7.18costDistributionStorage 1.28 1.28Subculturing 9.06 9.06Dissemination 12.54 12.54Distribution total cost 22.88 22.88Source: Koo et al., 2004.aStorage costs for the field genebank are the same as the cost of field maintenance.bStorage and subculturing costs for in vitro are allocated equally between conservation and distribution.6 Overview of wild Manihot species collection andconservationMany wild Manihot species are notoriously difficult to maintain either as field collections outside their natural habitator as in vitro plantlets. Seed conservation remains the preferred system. Since all the wild species appear to be seedpropagated in nature, populations are assumed to be highly heterogeneous (as opposed to cassava, which normallyconsists of single clones or clonal mixtures in production fields). There has, however, been inadequate attentiongiven to population biology theory in order to collect, and later conserve, wild species populations by methods thatoptimally conserve an adequate sample of the available genes. It is certainly to be expected that the genetic variationwithin and among populations of plants will be very high, and sampling for ex situ conservation should have a goodtheoretical and practical knowledge base at hand to select plants and seeds for collection in the wild, andregeneration ex situ.Low seed production of some species often limits the ability of the genebank curator to optimize a sampling systemduring regeneration. For example, EMBRAPA collected seeds from wild species across the four locations where fieldgenebanks are established in Brazil. Number of seeds collected varied from 11 for M. irwinii to 16,503 for M.peruviana (Alves, 2008). Genetic drift is likely to be a major issue in wild Manihot species genebanks, although nospecific studies have been carried out in this area. With species that produce few seeds, continual field maintenancemay be the most efficient means of combining conservation and regeneration.Response to different treatments to improve seed germination varies among species (CIAT, 1993). M. quinquipartitaresponded to heat treatment and pre-germination at alternating temperatures. Several species benefit from embryorescue, but others do not. Microwave treatment and mechanical scarification were detrimental to most species. CIATrecommended using a sample for germination by direct seeding, and holding some seeds for reserve in casealternative methods are needed.38

Cassava Conservation StrategyWork on in vitro culture shows that species vary widely in their media requirements for optimum conservation andregeneration (CIAT, 1984). Research at CIAT (CIAT, 1993) on methods to improve vegetative establishment of wildspecies compared leaf buds, shoots from rooted stakes, air layering, shoots from source plant, kinetin treatment andHormonagro ® treatment. Air layering was the most broadly successful method across species, but still resulted in alow rate of multiplication.Several species have been recovered successfully from cryopreservation, including M. esculenta ssp. flabellifolia, M.esculenta ssp. peruviana and M. carthaginensis. If most or all of the wild Manihot species have orthodox seeds(which is the indication so far from preliminary experience), there may be little justification to develop alternatives toseed preservation in low temperature/low humidity conditions. Field growth is useful as a means of having materialfor study, and for regeneration, but in the longer term, just as for cassava, will probably not be a recommendablesystem for genetic resources conservation.7 Overview of current cassava genebanksThis overview of ex situ collections is based mainly on a survey supported by the Trust during the first half of 2008.The survey instrument is included in Appendix III. Additional sources were used to fill gaps, especially for countriesthat had not yet responded to the survey as of this writing. The main sources were two relatively recent publishedreviews, with global coverage, of the materials existing in ex situ collections. The report of the first meeting of theInternational Network for Cassava Genetic Resources held at CIAT in Colombia, in August 1992, covered genebanksglobally, by region. Hillocks et al. (2002) provided a more recent summary. However, both these sources provide onlyrudimentary genebank information -- mainly total number of accessions by country.The surveys were sent to the curators of 50 genebanks throughout the cassava-producing world. As of August 2008,34 surveys were returned, and these data are summarized in various sections of this report, a good rate. Severalother curators committed to returning the surveys at a later date. A consolidated summary of key information fromthese various sources is given in Table 3, and Appendix V summarizes additional information from the LatinAmerican, African and Asian collections based on survey returns.7.1 Genebank holdings of landrace varieties and collection needsGenebank curators often choose to conserve some combination of local landraces, introduced landraces, andbreeding or other experimental materials. The approaches vary widely in terms of emphasis on these differentcategories. Local landraces are the nucleus of most collections, especially in the Americas. Breeder or experimentalmaterial generally involves rearranging the genes found in landraces, but does not introduce new genes. The onenotable, but relatively rare, exception would be mutation-based breeding, which produces heritable new traits notnecessarily found in any landrace variety. Therefore, from the perspective of cataloguing genetic diversity, this studyconsiders a local landrace (collected within the country) to be the unit of interest.Most genebanks are centralized within a country, with possible duplications at other sites. Brazil has instituted aregional system as a rational way to deal with very broad genetic diversity in a large country. For field-growncollections, this has the major advantage that most accessions are maintained in an environment similar to that oftheir site of collection (Fukuda, April/May CIAT workshop).Generally genebanks aim to maintain all collected material (except possibly material positively identified as internalcopies or duplicates), but there is rarely, if ever, a precise count of the number of distinct landraces in a country. Thegoal of collecting and conservation should be to sample all the locally frequent genes available in landraces, but thesampling strategy to achieve that is normally not based on good genetic information. The molecular characterizationof landraces is only beginning, so it is not yet a rational guide for collecting, but may be in the future. The concern is39

Cassava Conservation Strategythat collecting in areas not yet covered should probably not await the level of scientific information that would beideal; there is too much risk of continuing losses of material while waiting for research to provide this information.It must be understood that much of the data in Table 3 are highly speculative - current “best guesses” that need to beresearched further and continually updated with greater precision. Information in the Table is referred to by its columnnumber in the following discussion.In order to place cassava conservation in the perspective of importance of the crop in each country, Column 2 showsarea planted for each country, according to 2006 FAO data. Columns 3 and 4 show the number of accessions basedon previous reports. Column 5 reports only those materials indicated as local landraces in the current surveyinformation. Column 6 is an estimate of the number of landraces from each country that are held by either CIAT orIITA.A global conservation strategy must rely on reasonable estimates of the number of materials that should bepreserved from each country. Columns 8 and 9 of Table 3 estimate the number of unique landraces (excludingduplicates) held in ex situ genebanks, and the total number of distinct genotypes that exist in situ in each country.Although in theory it would be possible to arrive at nearly exact numbers by meticulous collecting and extensivemolecular analysis, the costs of doing so could probably not be justified, relative to other research priorities. Theseestimates place total ex situ landrace accessions at 10,068, and total in situ landraces at 26,986.There are two principal issues that confound the interpretation of most published, summarized registers of cassavagenebanks, and probably for the genebanks of other species as well: in general, it is not possible to distinguish whatproportion of the accessions are landraces, as opposed to bred varieties, or some other form or origin of material.And often it is also not possible to distinguish local landraces from landraces introduced from another country. Thesefactors tend to inflate the level of genetic diversity held in genebanks. For the most part, it is the local landraces thatrepresent the breadth of the genetic diversity available within a species (although, of course, there may be muchbroader variability available in the wild species). The ideal information, in order to best assess genetic variability,would be a comprehensive list of the number of accessions of local landraces held in ex situ collections, compared tothe total number of landraces that exist in a country (collected or uncollected). The germplasm survey reported hereattempted to correct this deficiency. Table 3 attempts to estimate this information by combining results from thecurrent survey data, from previous genebank analysis, and from personal contacts and experience of the author. Thisinformation is combined to make estimates about the density of landrace accessions in each country (number ofhectares per landrace accession); total number of landraces in each country; proportion of these that are maintainedex situ, and conversely, proportion that remain to be collected; and number of accessions from each country that areconserved in one of the CGIAR centers (CIAT or IITA). From these data, we make estimates of the total number ofaccessions that are not in CGIAR centers, and what would be involved in obtaining secure, complete duplication ofall landrace accessions in the IARCs.In some cases, there is good agreement in numbers across different reports, but in many cases there are widevariations in terms of number of accessions being reported for a given genebank.One way to compare across countries and regions is to look at landrace density, e.g. the number of hectares, onaverage, occupied by a landrace in a given country or region (total hectares planted divided by number of landracevarieties). By these estimates, as one would expect, landrace density is much higher in the Americas (center oforigin) than either Africa or Asia. In situ densities (hectares per landrace) are estimated as follows: Americas – 176;Africa – 1,619; and Asia/Oceania – 1,245. On a global basis, the estimate is 690 hectares per landrace variety.These numbers are very rough estimates. Also, these averages are made up of some extremely variable withinregionestimates. For example, Thailand is estimated at 97,346 hectares per landrace. This is in fact a situationwhere it is known that there are few landraces and a very large area of production. Countries like Mexico, Puerto40

Cassava Conservation StrategyRico, Surinam and Vanuatu appear to have rather broad genetic diversity, but very few hectares planted, so that, forthese four countries, the estimated landrace density varies from 2 to 8 hectares per variety.The estimate of 26,986 total landraces in the world (Column 9) may at first seem daunting in terms of the resourcesrequired for conservation. But there is little justification for attempting to conserve every genetic variant that exists –our interest is instead primarily in conserving the totality of genetic variation, which will most likely be contained in asubset of the total number of genotypes. Column 13 suggests numbers that might represent the minimum number oflandrace varieties that would need to be collected in order to conserve nearly all the genes in each country’s total oflandrace varieties. On a global basis, this is just over half the total estimated number of landrace accessions –14,791 out of a total of 26,986. This in fact is not an unreasonably large number of accessions to consider conservingex situ, given that CIAT alone has nearly half that number in its in vitro gene bank.Column 14 indicates the importance of introducing national program accessions that are not yet represented in theinternational centers, for safety duplication (see also later sections). These low, medium and high priorities are basedon number of in situ or ex situ materials not yet at CIAT or IITA, and the relative importance of that country’s cassavagenetic diversity. Critical countries for further representation in the international centers are: Brazil, Peru, Republic ofCongo, Côte d’Ivoire, D.R. Congo, Malawi, Mozambique, Rwanda, Uganda and Tanzania, along with several othercountries of medium urgency.Priority for further collecting (Column 15) is based on likely genetic diversity that is not yet represented in any ex situcollections. Highest urgency is for Bolivia, Brazil, Haiti, Angola, D.R. Congo, Ghana, Madagascar, and Mozambique.Collecting of cassava has already advanced very well on a global basis. This report estimates that about 14,791distinct landraces should be conserved in genebanks in order to adequately represent global genetic diversity ofcassava. Currently there are probably about 10,000, or two-thirds of the goal (see Table 3). However, it is probablyfair to conclude that most of the “easy” collecting has been done. The remaining priority areas tend to be somewhatdifficult to access because of lack of infrastructure, or are in countries with very low levels of available funding andpersonnel for collecting and conservation activities.One of the lessons from the genebank surveys is that truly accurate information is difficult to obtain in a surveyformat. In order to obtain accurate information, each genebank should be visited, the curator extensively interviewed,and historical records studied. Updating and correcting the information in Table 3 is a daunting task, but with acombination of local and international expertise committed to obtaining these data, it should be possible to arrive atreasonably accurate estimates in the next two to three years. Having these data will contribute considerably to arational conservation strategy.7.2 Collaboration arrangements in conservationIt is common for cassava genebanks to be managed by scientists who may not have a background in geneticresources management. These managers are often breeders or agronomists, who were the main founders of earlygenebanks as a practical way of providing for their own needs for broad genetic diversity. While the largest banks,especially those of the international centers and a few national programs, are now managed by germplasmspecialists, most are not. Plant breeders have a good sense of the economically significant genetic diversity in acollection and its long-term importance, but may be less interested in providing the necessary management input toconservation of materials with less immediate use in genetic improvement. In the 1980s, as the technology for invitro conservation was developed for broad use, there was a trend for separation of management of cassavagenebanks -- field genebanks continued to be managed by plant breeders or agronomists, and in vitro collections byphysiologists or botanists, but not necessarily genetic resources specialists. Not only are the in vitro managers oftennot genetic resources specialists, but also they frequently are not at all familiar with the field-grown plants or thegeneral needs of germplasm users, such as entomologists, pathologists or breeders.41

Cassava Conservation StrategyThe lesson in this overview is that the effective conservation of cassava genetic resources is necessarily acollaborative venture. The good news is that there is considerable expertise around the world in this area, andcommunications technology is making it ever more readily available to whoever needs it. The less-than-good news isthat consistent, adequate, long-term funding to support cassava genetic resources conservation is often notavailable, and the integrity of genebanks suffers.Collaboration of several types can enter the equation to improve conservation – training, infrastructure support,holding of duplicate genebanks, phytosanitary status testing/monitoring, GIS analysis in support of collecting, andothers. The mechanisms to foster collaboration reside mainly in networks – formal or informal, and range from intrainstitutionalto global. The formal networks were described in Section 3, and generally genebank managers will beaware of the networks relevant to their own situation. What is often missing in genebank management is creating andtaking advantage of the less formal networks. The good management of Manihot genebanks is in the interest of awide range of people, including future generations of farmers, agronomists and breeders, and all these people shouldbe part of formal or informal networks to support genebank management. Many times this support can come frompeople within the same institution where the genebank resides, or in related institutions.Below are some hypothetical examples at different levels of collaboration that may fit the needs of specificgenebanks. The way this collaboration is arranged will depend on each institution’s structure and modus operandi.For example, in some situations, a contractual arrangement may work, such as a genebank curator contracting theagronomy department in the same institution to maintain a field collection, or a pathologist to develop an indexingprotocol for a virus. Other situations may call for a joint grant application. A soil scientist could request permission toevaluate the collection at the expense of his or her own project, for tolerance to a soil stress condition, underguidelines agreed with the curator, and later load the data into the genebank database. In general the genebankcurator will be responsible for seeking out and creating these collaborative relationships, but should also welcomeproposals from others.Intra-institutional or local inter-institutional• Field genebank establishment and agronomic management by agronomist• Pest and disease evaluations by specialists• Virus indexing protocols by a pathologist• In vitro conservation by a private firm with appropriate facilities and expertiseNational• Molecular characterization in a university laboratory• Starch quality analyses by a private company labInternational• Duplicate accessions held by an international center• Virus indexing in a third-party lab prior to import or export of materials• Molecular characterization in a university laboratory• Taxonomic studies in an university7.3 The most important collectionsCriteria to determine the most important collections need to be identified for defining programs that merit support forcassava conservation. The April/May CIAT workshop participants discussed this topic and arrived at the followingrecommendations:Criteria to define the most important cassava collections• Recognized as a collection from a country or region with highly significant genetic variability for cassava• Inclusion of a major part of the country’s total genetic diversity in the genebank42

Cassava Conservation Strategy• Greater than X number of accessions• Duplication of collection at an IARC, or intent to do so• Signatory to the International Treaty• Demonstrated long-term institutional support• Ability to maintain entire collection “safely” in vitro for long-term conservation• Facilities for phytosanitary testing of in vitro collection• Ability to distribute to requesting entities within country• Ability to distribute to requesting entities outside countryThis set of criteria does not name specific genebanks, but establishes criteria that can be used to evaluate thecharacteristics of the world’s cassava genebanks by donors. The Working Group was reluctant to indicate a specificnumber of accessions required to be considered “important.” Rather, it was felt that this number would need to beconsidered in the context of the balance of other criteria. For example, a small collection from a country with broad orunique genetic diversity should be classified as “important” for long-term conservation. In such case, however, it maybe more practical to emphasize support to a centralized institution that maintains a duplicate of such materials.It is also recognized that the list of “important” collections will be dynamic. Countries whose current collections maynot be considered important may have a large genetic variability of uncollected landrace varieties. There needs to besupport for collecting these varieties and establishing them in ex situ collections. At that point, they may beconsidered important, and worthy of international support for long-term secure conservation. For example, eventhough Guatemala is a small producer of cassava, and has only a modest genebank (see Table 3), recent diversitystudies have shown that the Guatemalan germplasm is apparently genetically distinct from other groups in eitherAfrica or Latin America (Hurtado and Fregene, 2008). By this definition, this collection would be considered asimportant for the global conservation effort.It should be pointed out that currently very few national collections, and perhaps none of them, meet all the criteria inthis list. This became apparent after the workshop, as more of the genebank surveys were returned and analyzed.For example, among the genebanks that returned the surveys, only a few of the smaller ones maintain their fullcollection in vitro. Many genebanks have some capacity for in vitro conservation, but it would need to be upgraded inorder to securely conserve the entire genebank. Alternatively, donors could consider support to genebanks whoseaccessions are conserved in vitro in a duplicate collection, for example in one of the international centers. There arealso few countries that have the ability, at a high confidence level, to index accessions for virus infection, aprerequisite for international exchange. Few are equipped for, or interested in, international exchange, other than tosend to or receive from the international centers.The implications of these conclusions concerning important collections are included in the final recommendations forcassava conservation (Section 13).8 Overview of current wild Manihot species genebanksThe wild Manihot species have long been a frustrating challenge to collectors, genebank curators and breeders.Funding for their collecting has been sporadic and inadequate. Their conservation in the field, in vitro or as seeds allpresent difficulties and a number of species-specific research approaches. Breeders have long viewed some of theirpotential traits with great interest, but have been generally reluctant to commit to the time and difficulty of recoveringthose traits from the poor agronomic genetic background that result from crossing with M. esculenta. In addition, thethreats to their existence in natural populations continue, and in many instances is increasing. Despite proposalssince more than 20 years ago to create in situ reserves for wild Manihot species, this has not been realized.On the positive side, technological progress in all these areas is creating renewed interest in the Manihot species.The habitats of many of these species are known. This potentially allows the use of GIS to target specific sites for43

Cassava Conservation Strategymore efficient collecting. Progress has been made in conservation in all forms, including field, seed and in vitro.Developments in marker assisted selection are enabling much greater rates of progress in recovering target genesfrom wide crosses. Brazil’s EMBRAPA has continued making periodic wild Manihot collections, and is giving renewedefforts to their safe conservation (Alves, 2008).During the founding meeting of the Manihot Genetic Resources Network in 1992, participants updated and revisedcollecting priorities. The group recommended that collecting be prioritized to solve bottlenecks affecting existingbreeding programs. For wild Manihot there are still too many unknowns to define a detailed strategy, so Allem (1994)proposed using crossability with M. esculenta as an initial guideline. He described Gene Pool 1 (GP1) as the speciesknown to cross readily with cassava and yield fertile offspring. In this GP1 he included only M. flabellifolia and M.peruviana, now believed to be the direct ancestors of cassava. Taxa crossing with difficulty with cassava but givingsome positive results make up GP2. This pool includes M. glaziovii, M. dichotoma, M. pringlei, M. aesculifolia and M.pilosa. In practical terms there should probably be a combined weight given to environmental threats to populationsin the wild, along with expected potential use in cassava genetic improvement.Table 8 is a complete list of the Manihot species, as per Rogers and Appan (with notes on more recent taxonomicupdates). The list notes where species are maintained in the principal ex situ collections, and species of concern forpotential loss in their natural habitats. There are several genebanks around the world that maintain a few species forexperimental purposes, but only EMBRAPA, Universidade de Brasilia (Nagib Nassar) and CIAT have a seriousprogram for long-term conservation of wild Manihot. IITA maintains eight species. The shaded bars in the Tableindicate species of concern, which have no apparent representation in ex situ collections. 5 This includes ten of theSouth American species. However, many of those species that are conserved ex situ are seriously underrepresentedin terms of genetic diversity of the wild populations.In view of this small number of genebanks, determining the wild Manihot species collections of importance is ratherstraightforward: they are all critical to the safe long-term conservation of the genus. Due to its mandata, the Trust iscontemplating support only of ex situ collections, but in situ conservation should be supported by other means.5 As of this writing, there is no information available on the ex situ genebank of Mexico. Therefore none of theMesoamerican species are highlighted as species of concern, but it is nearly certain that many of them will later beshown to be at risk.44

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1Species1 M. angustiloba (Torrey) Muell.-Arg.emend Rogers & AppanApproximate geographical range asouthwest USA, Mexico2 M. davisae Croizat southwest USA, Mexico x3 M. walkerae Croizat southwest USA, Mexico x4 M. aesculifolia (HBK) Pohl Mexico x x x5 M. auriculata McVaugh Mexico x6 M. caudata Greenman Mexico x7 M. chlorosticta Standley & Goldman Mexico x x8 M. crassisepala Pax & K. Hoffmann Mexico x9 M. foetida (HBK) Pohl Mexico x10 M. michaelis McVaugh Mexico x11 M. oaxacana Rogers & Appan Mexico x12 M. pringlei Watson Mexico x13 M. rhomboidea Muell.-Arg. Mexico x14 M. rubricaulis I.M. Johnston Mexico x x15 M. subspicata Rogers & Appan Mexico x16 M. tomatophylla Standley Mexico x17 M. websterae Rogers & Appan Mexico x18 Manihotoides pauciflora (T.S.Brandegee) Rogers & AppanMexico19 M. carthaginensis (Jacquin) Muell.- Colombia, Venezuela, West Indies xArg.20 M. tristis Muell.-Arg. Venezuela, northern Brazil x21 M. surinamensis Rogers & Appan Venezuela, Guayana, Suriname22 M. filamentosa Pittier Venezuela xCIATEMBRAPAU. Brasilia*IITAMexican species 2xBrazilian species 3Brazil maniçobas 4Primary gene pool 5Secondary gene pool 6High genetic erosion 745

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1SpeciesApproximate geographical range a23 M. maguireiana Rogers & Appan Venezuela24 M. brachyloba Muell.-Arg. Central America, West Indies,northern & central South America25 M. marajoara Chermonte de Mirandaapud Hubernorthern Brazil26 M. caerulescens Pohl northern, northeastern, and centralBrazil27 M. glaziovii Muell.-Arg. northeastern Brazil; introducedthroughout tropical America, Africa,India, Pacific IslandsCIATEMBRAPAU. Brasilia*IITAMexican species 2Brazilian species 3Brazil maniçobas 4Primary gene pool 5x x xx x x x xx x x x x28 M. brachyandra Pax & K. Hoffmann northeastern Brazil x29 M. catingae Ule northeastern Brazil x30 M. dichotoma Ule northeastern Brazil x x x x31 M. epruinosa Pax & K. Hoffmann northeastern Brazil x x x32 M. heptaphylla Ule northeastern Brazil x33 M. maracasensis Ule northeastern Brazil x34 M. pseudoglaziovii Pax & K.northeastern Brazil x xHoffmann35 M. quinquefolia Pohl northeastern Brazil36 M. reniformis Pohl northeastern Brazil37 M. zehntneri Ule northeastern Brazil38 M. acuminatissima Muell.-Arg. eastern Brazil39 M. handroana N.D. Cruz eastern Brazil40 M. janiphoides Muell.-Arg. eastern Brazil x x41 M. pilosa Pohl eastern Brazil x x x xSecondary gene pool 6High genetic erosion 746

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1SpeciesApproximate geographical range a42 M. pohlii Wawra eastern Brazil x43 M. sagittato-partita Pohl eastern Brazil x44 M. warmingii Muell.-Arg. eastern Brazil45 M. tripartita (Sprengel) Muell.-Arg. central and eastern Brazil x x46 M. quinquepartita Huber ex Rogers & northern and central Brazil x xAppan47 M. alutacea Rogers & Appan central Brazil x48 M. attenuata Muell.-Arg. central Brazil x49 M. cecropiaefolia Pohl central Brazil x x50 M. crotalariaeformis Pohl central Brazil51 M. divergens Pohl central Brazil x52 M. falcata Rogers & Appan central Brazil53 M. flemingiana Rogers & Appan central Brazil x54 M. fruticulosa (Pax) Rogers & Appan central Brazil x55 M. irwinii Rogers & Appan central Brazil x x56 M. jacobinensis Muell.-Arg. central Brazil x x x57 M. longepetiolata Pohl central Brazil x58 M. mossamedensis Taubert central Brazil x x59 M. nana Muell.-Arg. central Brazil60 M. oligantha Pax central Brazil x61 M. orbicularis Pohl central Brazil x62 M. paviaefolia Pohl central Brazil63 M. peltata Pohl central Brazil x64 M. pruinosa Pohl central Brazil x xCIATEMBRAPAU. Brasilia*IITAMexican species 2Brazilian species 3Brazil maniçobas 4Primary gene pool 5Secondary gene pool 6High genetic erosion 747

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1SpeciesApproximate geographical range a65 M. purpureo-costata Pohl central Brazil x66 M. pusilla Pohl central Brazil67 M. quinqueloba Pohl central Brazil68 M. reptans Pax central Brazil x69 M. salicifolia Pohl central Brazil70 M. sparsifolia Pohl central Brazil x x71 M. stipularis Pax central Brazil72 M. tomentosa Pohl central Brazil x73 M. triphylla Pohl central Brazil x x x74 M. weddelliana Baillon central Brazil75 M. violacea Pohl central Brazil x x x x76 M. xavatinensis Rogers & Appan central Brazil77 M. esculenta subsp. flabellifolia western and central Brazil x x x x x(Pohl) Ciferri78 M. stricta Baillon Peru, western and central Brazil79 M. leptophylla Pax Ecuador, Peru, western and centralBrazil80 M. grahami Hooker southeastern Brazil, northernArgentina, Paraguay, Uruguay81 M. inflata Muell.-Arg. southern Brazil82 M. corymbiflora Pax southeastern Brazil x83 M. leptopoda (Muell.-Arg.) Rogers &Appansoutheastern Brazil84 M. jolyana N.D. Cruz southeastern BrazilCIATEMBRAPAU. Brasilia*IITAxMexican species 2Brazilian species 3Brazil maniçobas 4Primary gene pool 5Secondary gene pool 6High genetic erosion 748

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1SpeciesApproximate geographical range a85 M. condensata Rogers & Appan Bolivia86 M. guaranitica Chodat & Hassler Bolivia x87 M. anomala Pohl central Brazil, Paraguay x x x x88 M. gracilis Pohl central Brazil, Paraguay x89 M. pentaphylla Pohl central Brazil, Paraguay x90 M. hassleriana Chodat Paraguay x91 M. mirabilis Pax Paraguay92 M. variifolia Pax Paraguay93 M. populifolia Pax Paraguay94 M. procumbens Muell.-Arg. southern Brazil, Paraguay95 M. affinis Pax southern Brazil96 M. tenella Muell.-Arg. southern Brazil97 M. hunzikeriana Martinez-Corvetto southern Brazil, Argentina98 M. anisophylla (Grisebach) Muell.- ArgentinaArg.CIATEMBRAPAU. Brasilia*IITAMexican species 2Brazilian species 3Brazil maniçobas 4Primary gene pool 5Secondary gene pool 6High genetic erosion 7Species that exist in at least one location: 42a Source: Halsey, M.E., K.M. Olsen, N.J. Taylor and P. Chavarriaga-Aguirre.2008. Reproductive biology of cassava (Manihot esculenta Crantz) andisolation of experimental field trials. Crop Sci. 48:49-58.Species added after Rogers and Appan99 M. baccata x49

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1SpeciesApproximate geographical range aM. compositifolia x100 M. peruvianua x x x x xM. diamantinensis x101 M. hastatiloba x102 M.neusana Nassar (CJPS,1985) x103 M.swaminii Nassar (CJPS, 2008) xCIATEMBRAPAU. Brasilia*IITAMexican species 2Brazilian species 3Brazil maniçobas 4Primary gene pool 5Secondary gene pool 6High genetic erosion 7Cultivated varieties with "wild" traits"maniçoba""pornúncia""sete anos"xxxSynthetic speciesM. rogersii Nassar xM. vieirii Nassar xInterspecific hybridM. oligantha X cassava xM. pilosa X cassava xM. pohlii X cassava xM. glaziovii X cassava xM. pseudoglaziovii X cassava xM. neusana X cassava xM. dichotoma X cassava x50

Cassava Conservation StrategyTable 8. Wild Manihot species in genebanks and of concern for genetic erosion. (Shaded rows indicate “Species of concern” thatare not represented in any ex situ genebank).Genebanks Species of concern 1SpeciesApproximate geographical range aM. caerulescens X cassava xM. aesculifolia X cassava xM. anomala X cassava xM. reptans X cassava xCIATEMBRAPAU. Brasilia*IITAMexican species 2Brazilian species 3Brazil maniçobas 4Primary gene pool 5Secondary gene pool 6High genetic erosion 7Polyploidized interspecific hybridM. oligantha X cassava xM. aesculifolia X cassava xM. cearulescens X cassava xM. anomala X cassava xM. glaziovii X cassava xApomictic clone 307 and its polyploid typex* Each species is represented in the field by about 30 plants (Nagib Nassar, pers. comm.)1 Source: Howeler et al., 20012Mexican species threatened because of development3 Brazilian species threatened because of development and cassava cultivation4 Species of maniçobas economically valuable to dwellers of Brazil's NE semi-arid region5 Species involved in the ancestry of cassava and constituting th wild primary gene pool of the crop6 The putative closest relatives of cassava and assumed to participate n the secondary gene pool of the crop7 According to Nassar, 1979 (cited in IPGRI, 1994)51

Cassava Conservation Strategy9 Improving the efficiency of conservationVarious management strategies are available to make germplasm conservation more efficient. Two of these emergedwith some frequency in the genebank surveys – the definition of core collections, and the identification of duplicates.9.1 Core collectionsThe core collection concept grew out of the need to streamline and prioritize conservation and evaluation, particularlyin large genebanks. Originally conceived by Frankel (1984), a core collection would represent "with a minimum ofrepetitiveness, the genetic diversity of a crop species and its wild relatives." These collections are normally 5-10% ofthe total. CIAT defined a core collection of 630 accessions based on geographic origin, morphological diversity,diversity of esterase isozyme banding patterns, common landraces, and elite breeding lines (Hershey et al., 1994).CENARGEN, in coordination with CNPMF in Brazil, defined a core collection of Brazilian accessions.Defining a core collection has several implications for management of the whole collection. Conservation strategiesmay be tailored to give a higher priority for the core. The core may be duplicated in several institutions, or it may beheld in various forms (e.g. field and in vitro), while the remainder is kept only in vitro or in seed form. Use of the coreas a conservation strategy should be temporary, at best. For example, based on recommendations of the ManihotGenetic Resources Network (IPGRI, 1994), CIAT sent the core collection to Brazil and Thailand for duplication.Longer-term plans call for Africa to also receive this subcollection, when introduction of vegetative material isaccepted and managed more routinely.A core collection permits a better understanding of genetic diversity in the whole collection, through more efficientuse of resources for evaluation. Evaluation of the core for a given trait should indicate total diversity, and may directthe scientist to specific geographical areas or groups of germplasm with special promise for further study. It is acommon experience in large collections that evaluation for a single trait can take years and considerable resources.Evaluation of the core as a first step can be far more efficient. CIAT began extensively evaluating the core collectionsoon after its formation, especially for traits that had previously been considered too costly to evaluate in the entirecollection.As with any sampling procedure, a core collection definition is subject to sampling errors. The probability of a geneoccurring in the core collection is significantly different from its frequency in the entire collection only if that gene isrelated in some way to the criteria for defining the core. One of the main risks is the difficulty of identifying rare genesthat may escape inclusion in the core.Defining a core collection is strategically useful only in large collections. While a number of factors may influence thedecision, as a rule, defining a core collection may not be very worthwhile except in cassava collections of about 1,000or more accessions, and where most of the accessions are landraces.The surveys identified considerable interest in the development of core collections by the national programgenebanks, although only Brazil and India appear to have already established core collections. Some of the interestcould be the result of an incomplete understanding of the functions of a core collection. In fact, for practical purposes,there is probably limited value in defining a core collection for all but a very few of the largest collections. This wouldappear to limit the utility of core collections to CIAT and IITA, plus Brazil, and Peru (Table 3).There is some question about the need for, and the efficacy of, core collections for purposes of safe conservation,given the development in recent years by both IITA and CIAT of black box duplicates in other institutions. This doesnot mean that core collections may not have other purposes, especially with regard to determining likely areas tosearch for particular traits, or for other types of genetic diversity studies. Of course, the ideal would be a global corecollection, including material from multiple collections around the world. But centralized and standardized datamanagement in a global register would be needed for that. There is more on this in later sections.52

Cassava Conservation Strategy9.2 Duplicate identificationIt is common during collection expeditions to sample inadvertently the same genotype more than once. Situationsthat increase the probability of collecting duplicates are: (1) different local name for the same clone; (2) a clonewidely grown across a region; (3) collecting expeditions to the same region at different times; (4) clones sensitive toenvironmental variations and displaying variable phenotypes across microenvironments; and (5) inexperiencedcollectors.Hershey (1994) estimated that CIAT's global collection could be reduced by 20-25% by identifying duplicates. Thishas to be done with great care, however, and by relying on methods that will identify genetic duplicates with a highdegree of confidence. CIAT established a four-step procedure (Hershey et al., 1991; CIAT, 1993): (1) identification ofcandidate duplicate genotypes by comparison of eight key morphological characteristics; (2) side-by-side fieldcomparison of putative duplicates grown together in the same year; (3) re-characterization for morphological traits;and (4) characterization of putative duplicates with molecular markers. The efficacy of the molecular probe M-13 wasdemonstrated in that 20% of genotypes identified by other criteria as probable duplicates showed distinct fingerprints.Ocampo et al. (1993) analysed 4,304 accessions from CIAT’s germplasm collection, with the -esterase isozymesystem. From a total of 22 distinct bands, accessions grouped into 2,146 different banding patterns. A furtheranalysis of the Colombian accessions within the collection, using molecular markers, indicated a likelihood of about10% duplication (Debouck, 2008). If this figure is extrapolated across the CIAT collection, some 500 accessions(10% of about 5,000 landrace accessions in the genebank) could be combined or merged, with an annual savings ofsome $5,000 per year in conservation costs.In similar work, Sumarani et al. (2004) analyzed 70 sets of tentative duplicates (total of 139 accessions from 786indigenous accessions in India’s national collection). The esterase isozyme system produced a maximum of fivebands per accession, and among the multiple sets, a total of 35 bands, proving to be a highly polymorphic system.Altogether, 62 out of 218 accessions (28%) were found to be duplicates. The authors suggest that duplicateidentification should proceed in a logical manner from creating tentative groupings among a large number ofgenotypes, with rapid and inexpensive methods, to isozyme analysis with a reduced number of clones, and finally,confirmation by molecular probes. If facilities for molecular probes are readily available in a potential collaboratinginstitution, isozyme analysis might be eliminated altogether.Duplicate accessions may either be eliminated, assigned to a lower level of conservation priority, or combined into asingle accession. As the number of molecular markers is now in the thousands, the level of confidence in identifyingduplicate clones is becoming quite high. It is a question of whether the cost of identifying these duplicates is lowerthan the cost of maintaining them in the field or in the laboratory. As the cost of processing molecular markers fallsdramatically, their application to identifying duplicate accessions becomes very practical.9.3 Improved slow-growth conditionsThe annual cost of sub-culturing in vitro plantlets is about seven times that of storage (Koo et al., 2004). The mainway to reduce this cost will probably be to further slow the growth rate. This has been one of the main goals ofcassava in vitro research since its initial development 30 years ago, and constant progress is being made. This isresearch that can have a high return, and needs to continue in the main institutions that employ in vitro storage. Theincorporation of silver nitrate into the media is one recent example of new techniques with promise for extending thetime until regeneration (Mafla, 2008).53

Cassava Conservation Strategy10 Characterization and preliminary evaluationA germplasm collection is useful as a resource to its users only when accessions are well-described in terms ofcharacteristics of interest. Bioversity International (and its predecessors IBPGR and IPGRI) has developedstandardized descriptor lists for many crops, but has not yet done so for cassava. As noted in Appendix I, theApril/May CIAT workshop on cassava germplasm conservation was followed by a mini-workshop on cassavadescriptors, facilitated by Bioversity. This initiative should lead to a minimum descriptor list for cassava.There are two generally recognized basic categories of documentation for germplasm collections, apart frompassport data: (1) characterization – those characters that are highly heritable, clearly visible and are expressed inall environments; and (2) preliminary evaluation – a limited number of additional traits of lower heritability considereddesirable by a consensus of users of the crop. Characterization is important basically as a tool for varietaldescription, identification of duplicates in a collection, monitoring genotypic stability of clones stored in vitro or inother non-conventional forms, and varietal fingerprinting. Preliminary evaluation is often the starting point forbreeders to identify an accession’s potential value in a breeding program. A breeder's general objective is typically toidentify clones that can be used directly as recommended varieties, or as parents in a breeding program. Many othercrucial decisions hinge on this general objective, related to target production areas and their physical and biologicalcharacteristics, management practices to be employed, and processing and marketing characteristics.Most germplasm curators will see characterization, as a means of describing and cataloguing an accession, asclearly part of their mandate, but often draw a line at evaluation, considering that the preserve of breeders. There isroom for attitudinal change here, on both sides, and greater collaboration.Preliminary evaluation consists of six broad categories: (1) general adaptation, (2) resistance, (3) plant architecture,(4) yield, (5) root quality, and (6) other locally important traits. 6 The procedures for evaluation of germplasmaccessions may be very similar, or identical to evaluation of breeding lines. Much of the detail on evaluation andselection given in later chapters can be applied also to germplasm collections. There is, however, an importantprocedural difference: all germplasm accessions should be equally and fully evaluated. On the other hand, breedinglines may be pre-selected on the basis of a few key criteria, and only those passing this first step receive furtherevaluation. If large numbers of germplasm accessions need to be evaluated, some compromises may be made withregard to level of precision. With up to a few hundred accessions, multi-location evaluation in replicated trials may bepossible. If accessions number in the thousands, the breeder or germplasm curator may only be able to manageunreplicated single row trials.Many characters may appropriately be evaluated within a field-planted genebank itself. Stresses that impose risks tothe collection, and may result in accession losses if uncontrolled, should be evaluated in separate, specially designedtrials. Serious pests and diseases or major soil problems are examples. The field collection often is not anappropriate place to evaluate yield or quality because of inappropriate plot design or the need to leave plants in theground well beyond the normal harvest period.Since the mid 1990s, the accession information and some evaluations from CIAT’s germplasm collection have beenavailable on-line at www.singer.cgiar.org. This website is managed by the System-wide Information Network forGenetic Resources, the germplasm information exchange network of the System-wide Genetic ResourcesProgramme of the CGIAR. While this is a reasonable first step to search for traits of interest, it is best done withadditional consultation with breeders and germplasm curators who are familiar with the details of the evaluations andthe germplasm itself. Clearly there is great value in the germplasm information database. At the same time it will bemost useful to a breeding program if the evaluations are understood in the context of a complete picture that includesagro-climatic conditions, and the complete range of traits that are of importance to the breeder. To that end, a6 For example, consumer preferences for eating quality, etc., or color traits such as yellow roots.54

Cassava Conservation Strategycassava common registry has been recently established, with passport and preliminary characterization dataprovided by CIAT and IITA at this stage (see http://www.cassavaregistry.com).There is an urgent need to digitize data currently held only as hand-written files. Many genebanks are still managedwithout benefit of computerized information. Few of those that do maintain digitized files make this informationbroadly available online.11 DistributionMany genebanks and breeding programs obtain new genetic diversity through introduction from outside sources. Theprinciples and methods associated with germplasm exchange are fundamental to the functioning of most genebanks.The large majority of cassava genebanks distribute materials only within the county, and only a few engage in regularinternational exchange. This discussion focuses on international cassava germplasm movement.11.1 Benefits and risksThe potential benefits of germplasm introduction are essentially a function of the genetic variability available in localgermplasm, and more specifically, of the strengths and weaknesses of that germplasm. Typically, the range ofvariability in local germplasm depends on the region, with the highest variability usually in the crop's centre of origin,the Americas. Even in areas of high variability, there can be large advantages to germplasm introductions or tocapitalize upon advances made in breeding programs elsewhere to introduce specific characters.Two types of risks accompany germplasm introductions: phytosanitary and genetic. The phytosanitary risks –introducing new biotypes or species of pests or pathogens – are of paramount importance. Minimizing these risksmust take very high priority in any germplasm exchange. The genetic risks are the risks of knowingly or unknowinglyintroducing undesirable alleles along with the known desirable ones. Undesirable alleles may be those that confersusceptibility or non-tolerance to a particular environmental factor; alleles for poor quality; or in general, any that areconsidered less desirable than those controlling the same traits in local germplasm. These genetic risks areminimized by an appropriately designed evaluation and selection program.11.2 Forms of exchange11.2.1 VegetativeGenerally, the international exchange of genebank accessions between institutions is through in vitro culture. Theprincipal advantage of in vitro introduction is phytosanitary. Insects, mites, bacteria and fungi are easily eliminated,and cultures can be indexed for several viruses to provide a high level of assurance of pest and pathogen-freematerial. From a standpoint of international quarantine, in vitro introductions are widely accepted within and amongAsian and Latin American countries. Regulations on exchange within Africa, and between Africa and othercontinents, are more variable and generally more restrictive. No method is free of risk, but the technology fordetecting pathogens is well-advanced.In vitro introductions first need to be propagated and grown in specialized conditions, resulting in some delay untilagronomically useful evaluations can be made. Under ideal conditions, and using rapid propagation techniques,agronomic trials can be established within one and a half years after in vitro introductions. Normally, however, threeor more years are required to obtain sufficient planting material (including one cycle of field propagation to obtainlignified stakes). Many scientists receiving in vitro cultures have initially been too optimistic about the time requiredfor regeneration and evaluation.55

Cassava Conservation StrategyIn the case that international exchange is for the purpose of introducing new variability to genebanks that serve theneeds of breeders, the amount of genetic variability that can be managed is a major limitation for vegetativeexchange. For large numbers of clones, expense of preparation and difficulty of management by the recipient, maybe prohibitive. This generally means that only a limited number of clones are sent in any given shipment, usually onthe order of ten or less, but up to a few hundred in special cases.11.2.2 SeedsThe two outstanding advantages of seed introductions are ease of handling broad genetic variability and therelatively high tolerance of seeds to storage and shipping. The fundamental property of seed introductions is that anyplant derived from a botanical seed of cassava is a new, distinct genotype – necessarily different from the clone fromwhich it was derived. Seed introductions into a genebank will not duplicate the accessions of the donor bank. Oneconstraint for some programs to utilize seed introductions is the need for specialized training for management of seedand seed-derived plants. Some programs combine both seed and vegetative introductions, taking advantage of thepositive features of each.11.3 Quarantine considerationsThe exchange of cassava stem cuttings through unofficial means (farmers, tourists, entrepreneurs) is probably themajor means of disseminating pathogens and pests across international boundaries. The bacterial blight pathogencan survive in the xylem vessels of infested stems for months. Cassava viruses and mycoplasmas are efficientlyharbored in stem cuttings from infected plants and readily transferred to new plants via infested cuttings. Thecassava green mites, mealybugs and scale insects can survive for months, feeding on the lateral buds of stemcuttings. Introductions of green mites, mealybugs and the bacterial blight pathogen into Africa are importantexamples of the risks of inappropriate and unmonitored germplasm movement.Some pathogens can be disseminated through botanical seeds. These fit into two broad groups: (1) those that infestthe seed; and (2) those that infect it. Infestation may follow fruit dehiscence. If the seeds fall to the ground, theprobability of infestation is higher than when the seeds are collected prior to dehiscence and stored under controlledconditions. Pathogens of cassava that can most effectively infest the seeds and survive on them are those producingabundant mucilaginous propagules, such as Colletotrichum, Phoma and Diplodia spp., and Xanthomonas campestrispv. manihotis. Infestation of storage containers is also a risk. Disinfested seed should be repacked in cleancontainers.Pathogens that infect seeds include X. campestris pv. manihotis, Diplodia manihotis, Fusarium spp. andCladosporium spp. However, the limited research in this area does not preclude the possibility of other fungal andbacterial pathogens.Determination of the potential for seed transmission of all cassava viruses is essential for the safe interchange ofbotanical seeds, but information is far from adequate. The main virus concerns, namely, cassava mosaic virus,cassava common mosaic virus and frogskin virus, are apparently not transmitted via cassava seeds. Two morerecently discovered nepoviruses, the cassava green mottle virus (apparently a minor virus limited to some SouthPacific islands), and the cassava American latent virus, found in Brazil and Guyana, raise some concern about seedtransmission in view of the type of virus.A few mycoplasma-like organisms (MLOs) affect cassava, causing antholysis (leaf distortion) and witches' broomdiseases. These MLOs are not seed-transmitted.There are few insects that attack cassava seeds so the risk of disseminating arthropod pests is relatively low. Seedsmay however be superficially infested, especially with mites. A seed insecticide/miticide treatment is recommended56

Cassava Conservation Strategyas a precaution, especially for any seed to be shipped internationally. However, some quarantine agencies, includingthat of Brazil, have expressed concern about exposing quarantine personnel to pesticides as they examine seeds.FAO and IPGRI jointly published technical guidelines that include general and technical recommendations forcassava exchange (Frison and Feliu, 1991). 7 These phytosanitary measures, independent of others legallyestablished by quarantine regulations of importing countries, would reduce the risk of disseminating pathogens andpests through propagative material of cassava. Their effectiveness depends on the strict application of suchmeasures by both the sender and the recipient. Technical recommendations are provided for: (1) seeds; (2)pathogen-tested in vitro cultures; (3) cuttings from pathogen-tested in vitro cultures; and (4) untested vegetativematerial. By the end of 2008, CIAT intends to have the entire in-trust collection of cassava landraces disease testedand ready for international shipment (Cuervo, 2008)The importing country should be especially cautious in the introduction of cassava propagating material fromcountries or areas where exotic diseases exist. For example, because of cassava mosaic disease, vegetativematerial should not be imported from Africa or India, except after very thorough virus indexing both at the source andin a third party institution in a non-cassava growing country. Indexing can be done in the mother clone (a field-grownplant for example), in in vitro tissue in the country of origin, and/or in in vitro material by the third party. A third partyshould be an independent entity that has a high level of trust by both the country of origin and the recipient country.Detection methods can be based on the observation of symptoms in the mother plants, symptoms in grafts orindicator plants, or on the detection of virus particles and viral products. The reliability of detection methods based onplant symptoms can be increased by growing plants under optimal conditions for symptom expression. For example,the symptoms of cassava mosaic disease are poorly expressed at temperatures above 28°C. In this case plants maybe grown in a cooler environment to enhance symptom development.The bioassay of mechanically transmissible cassava viruses to indicator hosts is a sensitive indexing method if a verysusceptible host is available, virus concentration in the test plant is high and environmental conditions are optimal forsymptom expression. The Nigerian isolate of cassava mosaic disease produces a severe, systemic infection ininoculated Nicotiana benthamiana plants. The Kenyan isolate of cassava brown streak virus can be bioassayed on N.debneyi.Grafting is a method for indexing viruses and virus-like agents that are not mechanically transmissible. Graft indexingis very sensitive if a highly susceptible indicator clone is used in the graft. The native Colombian clone Secundina ishighly susceptible to frogskin disease (the same causal agent as for the Caribbean mosaic). When a Secundinascion is grafted onto an infected rootstock, leaves express moderate to severe mosaic symptoms. Although a graftindexingprogramme requires minimal facilities and training, the procedure is labor intensive and indexing results arenot available for several weeks. Another major constraint can be the difficulty of maintaining virus-free stocks of theindicator clone.Sensitive serological tests are available for viruses that have been isolated, purified and an antiserum produced.ELISA is a highly sensitive, efficient and rapid method for detecting CMD and cassava common mosaic virus(CCMV). The immunoabsorbent electron microscopy (ISEM) test can also be used for detecting CMD and CCMV.ELISA is suited to a large-scale virus-indexing programme, where hundreds of plants can be tested in a day withresults available within 36 hours. The preparation of test material and examination of grids is simple and rapid.Although ISEM is not as sensitive as ELISA, it has the advantage of providing results within several hours.Nucleic acid or spot hybridization and isolation of viral-specific double-stranded RNAs (dsRNAs) can detect somecassava viruses. Spot hybridization has been adapted for detecting CMD. The procedure is based on the use of a7 There are plans to update the guidelines in the near future (D. Debouck, pers. comm.).57

Cassava Conservation Strategyradioactively labelled DNA molecule that is complementary to the viral genome, to probe spots of leaf sap for thepresence of viruses. The test is highly sensitive and suited for processing large numbers of samples.Isolation of dsRNAs is especially suited to detecting uncharacterized viruses for which an antiserum or nucleic acidprobe is not available. The extraction and analysis of dsRNAs are somewhat laborious, making the test moreappropriate for indexing a limited number of mother plants rather than as a general screening method.One of the principal concerns of shipments from Latin America to either Asia or Africa has been frogskin disease.Until recently, detection was done mainly by grafting onto a sensitive scion. A new test for the virus (rt-PCR) hasbeen standardized and the results are in the process of being published, as the first step in gaining internationalacceptance for certification of frogskin-free material. The method cuts the diagnostic time from more than 20 weeks(for a grafting-based test) to just a few days (Cuervo; Debouck, workshop presentations).11.4 Procedures for distribution11.4.1 SourcesThere are few genebanks with the capacity to act as sources of cassava germplasm on a regular basis and providethe essential phytosanitary safeguards. These functions have been assumed mainly by the international centers –CIAT for Asia and Latin America, and IITA for Africa. Both centers use the latest indexing and preparation techniquesto give the highest possible assurance that material being distributed is pathogen-free.The Field Crops Research Institute of the Thailand Department of Agriculture, in collaboration with the CIAT AsiaCassava Program, has distributed germplasm from various sources (mainly Thailand and CIAT breeding program)throughout Asia. Several quality in vitro laboratories are situated in the region, and the generally low level ofproblems of quarantine significance simplifies distribution. Within the context of a major international project forcassava breeding in the 1990s, CNPMF and CENARGEN of Brazil developed a protocol for distribution of cassavaseed to Africa. A few other countries may respond to germplasm requests, but normally are not prepared to do so ona regular basis. The international centers can often act as intermediaries to facilitate germplasm exchange betweentwo countries that may not have complete capacity for pathogen indexing.11.4.2 Legal aspectsThe international agricultural research system (including formal or informal collaboration among national programs,universities, the private sector and international centres) has depended on the free exchange of materials andinformation for continued success. The results of plant breeding research, both from private and public sectors, areincreasingly protected with various forms of intellectual property protection, including patents, material transferagreements, plant breeders’ rights and trade secrets. Since implementation of the International Treaty on PlantGenetic Resources for Food and Agriculture (www.planttreaty.org), the principal means of formalizing exchange ofcassava germplasm has been the standard material transfer agreement (SMTA), required for material in theMultilateral System for access and benefit-sharing (see ftp://ftp.fao.org/ag/cgrfa/gb1/SMTAe.pdf). Patents and tradesecrets associated with genetically modified plants or tissues are coming more into play, but to a far lesser degreethan for crops important in temperate agriculture.Africa, in particular, has less capacity to replicate research results patented elsewhere, for the benefit of poor farmers(Devries and Toenniessen, 2001). While there are many publicly funded partners who would be willing in theory toshare their most important discoveries freely, they are often unable to do so because of agreements made withprivate donors who want to protect their market advantages.58

Cassava Conservation StrategyTable 3 (Column 16) indicates the ITPGRFA status of the world’s major cassava-producing countries. Only twentyfive(about one-third) of these countries have ratified the Treaty, so there is clearly a constraint to the free exchangeof cassava germplasm under the Treaty’s terms.It is important to note that wild cassava is explicitly excluded from the Annex 1 list of crops included in the Treaty’sMultilateral System. Such material in the international collections maintained by CIAT and IITA is included in the MLSvia Article 15, but the same cannot be said of wild Manihot in national programme genebanks, unless the countryexplicitly includes it.12 Documentation of germplasm managementGenerally, an institution assumes responsibility for germplasm management as a permanent, ongoing activity. Aneffective information management system becomes a critical part of the process. Survey returns indicated that mostnational programs have only rudimentary information management systems in place for genebank management.Most, however, also indicated an intention to computerize information in the next three years. This is an area wheregenebanks could greatly benefit by availability of standardized procedures and protocols, while at the same timehaving the flexibility to utilize locally familiar systems.Accuracy in information management is critical. For example, the cumulative effects of even a low error rate in theidentification of accessions will have devastating effects on the validity of information in the long term. Historically,there is often considerable instability in the personnel responsible for conservation of collections, and this cancontribute to some lack of consistency in information management. Information for many small collections is not wellorganized, and the evaluation data are of dubious quality.As electronic information management becomes more widely available, including positive ID systems like bar codesor RFID, genebank information management should improve. A database integrating information across allcomponents of germplasm management (Perry, 1994) would provide a means to:• assess the current status of conservation and characterization of the genetic resources in all participatingcollections;• provide an indication of gaps that may exist in geographical representation or phenotypic/genotypicvariability inherent in the collection;• provide an indication of duplication (including intentional security duplication) of material betweencollections;• assess the regeneration requirements at international level.Some of the above objectives are currently filled by the Cassava Common Registry (accessible freely athttp://www.cassavaregistry.com ), with data inputs provided so far by CIAT and IITA. There one can consult thecassava databases of both institutions and introduce cassava germplasm requests to any of them.13 Rationalizing a conservation strategy – Manihot esculenta13.1 Elements of a conservation strategyA conservation strategy for cassava should be comprehensive, secure, efficient and cost-effective. It needs to beagreed upon by all partner institutions that maintain or utilize genebanks. The current collections held by nationalprograms, state or regional research and extension programs, universities, international centers, or other entities,vary considerably in the types of materials that are maintained and in the protocols for conservation and distribution.To some degree, each is free to determine independently what is in their best interest in this regard. However, thosethat have signed the International Treaty are committed to its standards and protocols. The international centers –CIAT and IITA – have broad international obligations to conserve cassava in a manner which meets long-term needs59

Cassava Conservation Strategyof the global community. The range of holdings reflects these needs and interests. They include local landraces,introduced landraces, bred varieties (released materials), experimental breeding materials, mapping populations, andothers.A global conservation strategy should focus on landraces. These represent the range of diversity that has evolvedover time in farmers’ fields, under a combination of natural and human selections. All current genotypes (exceptingthose few that are the result of mutation or transformation events) are the result of recombinations of existing orextinct landrace varieties. The large majority of future genotypes will, as well, consist of genes that derive fromexisting or extinct landrace varieties. This is the set of genotypes that is of highest priority for secure conservation inperpetuity. Table 3 of this report gives some preliminary estimates of number of landrace varieties in most cassavaproducingcountries, but ultimately this number will be determined by extensive field collecting. Perhaps a moreimportant question for conservation, however, is, how many genotypes are needed to represent the diversity of alllandrace varieties of a country or region? This is a question that can only be answered with an extensive base ofinformation from evaluations and molecular studies on diversity. For the present, we can only provide an educatedguess. For the long-term, this will be a critical basis for determining the resources required for conservation. Thenumbers presented in Table 3 need to be broadly reviewed and evaluated by experts familiar with the diversity ofeach country. Although a draft of this report was widely distributed, feedback on this particular element ofconservation was minimal. It appears to be an aspect of cassava conservation that has not received much thought oranalysis up to now. The Table 3 estimate of almost 15,000 accessions needing to be conserved on a global basis, inorder to fully represent cassava genetic diversity, should be considered a tentative number, but at the same timeprovides a starting point in planning for resource requirements in a global conservation strategy.In the long term, a critical question to be answered is the desirability of conservation of genes versus conservation ofgenotypes. The question is fundamental to a conservation strategy both in terms of genetics and of financialresources. The conservation of genes can be done in either vegetative or seed form, whereas the conservation ofspecific genotypes (where perpetual regeneration is practiced) is only possible in vegetative form. The implications ofthis choice need to be a basic part of the discussions among cassava germplasm curators and users in the comingyears (and probably of other vegetatively-propagated crop species as well). Insofar as breeders make directrecommendations of landraces to farmers, it is essential that these landraces be conserved vegetatively. But this is astrategy likely to be of diminishing importance, as continual breeding improvements are made and landraces play acontinually lesser role for direct use as new varieties. However, their role as sources of genes will continue as long ascassava breeding continues. The genes of value can be derived from either seed-derived material, or from theoriginal landraces maintained continually in vegetative form. Based on current information, seed conservation shouldbe possible at a far lower cost than vegetative conservation. While it appears that a cassava conservation strategyshould evolve in the long term toward seed conservation, both for reasons of genetic utility and of efficient resourcesmanagement, in the short and medium term, we should continue a strategy based on vegetative conservation.Understanding that the longer term aim is seed conservation is an important factor in the design of a vegetativeconservation strategy. This report has focused on a vegetative strategy, since it is the one of immediate concern, andneeds to be established as a precursor to a seed conservation strategy.13.2 Conservation scenariosThe elements of a strategy can be brought together in multiple ways. Some of these options are presented here asscenarios, that present choices based on activities, partnerships and funding. There is no single best choice. In part,the pathways chosen will depend on level of external funding available, and the level of commitment by participatinginstitutions. Optimizing the organization of shared services can create substantial cost benefits and efficiencies, butat the same time requires a large investment in institutional planning and interaction. Networks abound that may playa role in conservation planning and activities, but there seems to be a balance between investments in assuring thatnetworks function, and in investments that are directly targeted to operational matters. There is little point in fundingplanning meetings while the lack of funding for basic conservation functions like field genebank maintenance results60

Cassava Conservation Strategyin germplasm losses. One of the more ambitious networks of the past 25 years, the International Network forCassava Genetic Resources, was conceived well in every aspect except for the unexpected sharp decline of nationaland international funding that would be available. The entry of private funding into cassava research anddevelopment has been very slow, but that now appears to be changing. The starch and animal feed industriesespecially are dynamic and enthusiastic about the potential of cassava in these markets. Both have shownwillingness to invest in genetic resources in a narrow way – where the benefits have a clear possibility of helping theirindustry – but these initial experiences can and should lead to productive partnerships between private and publicsector institutions for cassava conservation.Networks will be critical in establishing standards and agreements on conservation strategies and methods. Theseneed not be formal networks, but rather groups that have the ability to meet, or communicate electronically, todevelop agreements on conservation standards, on information management standards, on germplasm exchangeprocedures and standards, and on the means to achieve long-term support for conservation. South-southpartnerships should be a significant part of the future of Manihot genebanks. The regions of less variation (e.g. Asia)should see a role in supporting conservation in the Americas for their own future benefit.Table 9 summarizes some of the short- and medium-term research needs in cassava and Manihot conservation, whocan do the research, and key elements of funding to achieve success. The most notable point is that there is a widerange of needs, and these will need to be tackled by an equally wide range of institutions doing diverse research.Table 9 indicates a series of alternative scenarios in the broad scheme of a global conservation strategy for cassava,involving national programs and international centers, and a range of conservation methods. The following highlightssome of the main points to consider:Table 9. Alternative scenarios for conservation, based on an individual landrace variety.Alternative conservation sites and methodsNationalprogram International center Black box Analysis of the global systemEfficiency aField In vitro Field In vitro Cryo Seed In vitro Cryo Security CostCurrent situation: b Low na na7,000 2,000 2,000 7,500 670 0 6,000 0 Med Med MedAlternative future scenariosx x Low Low Lowx x x Med Med Medx x x x High Med Medx x x x x High High Medx x High Med Medx x x High Low Highx x x High Low Highx x x x High Med HighaCost-effectiveness of obtaining the designated level of ex situ security.b Uncollected landrace varieties (existing in situ only) are estimated at about 17,000.Security of conservation is a major consideration in a global strategy. This is basically a product of the level ofsecurity that any particular method provides, and the number of sites where each accession is replicated. Securityinvolves inherent characteristics, management factors, and the unknown or unpredictable externalities. For example,recovery from cryopreservation is known to be successful only for a certain percentage of genotypes. This is aninherent trait of each clone (for a given protocol). Management has a critical influence on all forms of conservation.Externalities could include serious threats like war or long-term electrical outages. Therefore, any single site forconservation is at some level of risk, and the most cost-effective means of obtaining a high level of security is to61

Cassava Conservation Strategyconserve accessions in different sites and in different forms. There is no magic formula to determine the right balancebetween level of security and cost. It seems clear that more than one site is essential, but it is also evident thatresources are not unlimited, and there can be little justification for more than two or three sites for secureconservation of a given clone. Based on the Table 3 estimates, nearly two-thirds of landrace varieties exist only insitu. We did not have enough information to estimate number of varieties that are held in only one ex situ genebank.In the long-term, a highly secure system for vegetative conservation could include in vitro conservation in one siteand cryopreservation at two sites.• There is a large number of uncollected landrace varieties in farmers’ fields, which currently fall outside therealm of management by ex situ genebanks.• Under the current situation, clones that are duplicated between national programs and IARCs are relativelysecure, but there are probably a few thousand accessions in national programs that do not exist in theIARCs. These are in various states of security.• There are multiple scenarios by which a high level of security of conservation can be achieved. Betweennational programs and IARCs, there should be at least three replications of each landrace accession, inorder to achieve optimum conservation security.• Greater than three replications ex situ is probably not cost efficient in terms of security, but it may be costefficient for other reasons, such as specific research objectives of a genebank.• The ideal combination of high security, low cost and high efficiency can be achieved with carefulcoordination between national genebanks and international centers.• The international centers should provide a secure duplicate for national program genebanks, such thatnational programs can reduce their investment in duplication, while at the same time having access to fieldcollections for efficient evaluation and breeding.• Any national program is of course free to duplicate their collection as often as they wish, for whateverreason, in their own country.• While cryoconservation can provide a high level of security at a low cost, there will probably always be anadvantage in having in vitro slow growth or field collections at the national program level, as a means ofeasier facility of field regeneration for evaluation purposes.13.3 A consensus strategy: collaborative centralizationIn terms of a global strategy, there appears to be a relatively logical way forward that meets the cassava geneticresources conservation needs expressed by the scientists providing input to this report. The components of thisstrategy are:• Collecting in priority areas is carried out to fill gaps, with the aid of genetic diversity studies and geographicalinformation systems (GIS).• National program genebanks and the CG centers (CIAT and IITA) develop a common cassava registry at aglobal level, based on passport, morphological and molecular information.• The CG centers continue their long-term commitment to the secure conservation of landraces. In the shorttomedium-term, they should continue with their continental mandates – Asia and the Americas for CIAT andAfrica for IITA – due to the virus problems exclusive to each continent, and the risks associated with movingvegetative material.• In the long-term, the CG centers should consider a unified conservation strategy, where each holds a globalcollection, and each serves as a backup to the other.• The CG centers will duplicate landraces that are currently held only by national programs. Currently thecenters appear to hold about 50-60% of the total ex situ accessions.• The CG centers formalize a long-term commitment to replacing any materials lost from national programcollections, at the request of the national program.62

Cassava Conservation Strategy• The CG centers maintain at least two forms of each accession. Currently this may be an in vitro activegenebank plus a black box duplicate kept in another center. In the future, cryopreserved accessions will beeither the main or the backup genebank.• The CG centers commit to meeting the demands and phytosanitary requirements for international exchangeof cassava landrace varieties under terms of the International Treaty. Along with this, it is urgent to developprotocols for the safe movement of vegetative germplasm between the Americas and Africa.• The national program genebanks commit to long-term conservation of their national genetic resources, at alevel of security that assures low risk to the collection, but without the need to invest in expensiveinfrastructure, or in a duplicate collection within the country.• Many national programs will find that field collections are the most cost effective and practical means ofconservation. This system normally allows a combination of moderately secure year-to-year conservation ifproper precautions are followed, and has the additional advantage that planting material is readily availablefor evaluations.• A structure is developed for periodic interaction among stakeholders. Most notably this will be between theCG centers and the national programs. Each will have a formal responsibility to periodically inform the otherof the status of each collection.This strategy could be described as one of collaborative centralization. There are compelling reasons to rethink adecentralized strategy where each national program has the ability to conserve its germplasm in a highly securesystem, which normally involves a field collection backed up by an in vitro collection. There have been somesignificant changes in the world of cassava genetic resources that impact the structure of an optimum conservationstrategy. First, the status of the collections maintained by the CGIAR has been clarified. These collections are nowpart of the Multilateral System of the International treaty under its Article 15. Secondly, international exchange hasbecome much safer and more acceptable with advances in virus indexing.This environment allows us to think in new ways about the optimum conservation system for cassava. Conservationin vitro (slow growth or cryopreserved) is highly non site-specific and therefore large efficiencies can be gained bycentralization. This centralization in the international centers now becomes politically viable, because ownership isclarified, and international exchange is clarified and more secure from a quarantine perspective. We now have anopportunity to develop a strategy that is biologically and economically rational, creates a structure of interdependenceand collaboration among genebanks, and at the same time conforms to the new policy environment.13.4 Implementation and fundingThe first step in implementation of the collaborative centralization strategy will be to formalize agreement at theinternational level, among the CG centers and at least the major national cassava genebanks. This formal discussionand agreement is important, since this recommendation represents a major shift in strategy, not just strengthening ofa current strategy. The consensus among the participants in the cassava conservation workshop (CIAT, 30 April – 2May 2008), and among the respondents to the genebank survey serves as an important step toward formalagreement. It is this formal international agreement that will provide the basis and the impetus to initiate a broadrange of conservation-related activities. This one step is the most fundamental action required, since it is aprerequisite to setting in motion most of the subsequent activities in the global strategy.The greatest needs in cassava genetic resources conservation are often not those of greatest technical difficulty.Rather, they tend to be issues of inadequate funding for fairly simple and routine tasks. Substantial progress hasbeen made in some of the more complex issues of cassava germplasm management, such as virus identification andindexing, fine-tuning of in vitro media for slower growth, cryopreservation pre- and post-treatments, and molecularfingerprinting. It is true, of course, that there are still many challenging scientific questions that will require significantinvestment.63

Cassava Conservation StrategyAt the same time, many programs struggle with the basic ability to maintain a field collection, manage the informationsystematically and accurately, and evaluate it in appropriate environments for the benefit of growers and consumers.The most urgent challenges in terms of securing the long-term preservation of cassava germplasm appear to be inproviding support to the programs most at risk of permanently losing accessions. Providing support to programs thatmeet the criteria for standard conservation protocols and for international exchange attacks the other side of thesame problem. This dual thrust of reducing the risks of loss of genetic diversity within at-risk genebanks, along withsupporting the ability to duplicate these materials in an international center for secure back-up, will go a long waytoward preventing serious erosion of cassava genetic resources.It must be emphasized that the conservation strategy proposed here by the stakeholders of cassava germplasmrepresents a major shift in the operational structure of the global conservation system. The emphasis placed oncentralized secure conservation, and the role of the international centers, collaboratively with the national programs,means that there will need to be high level discussions to put into place the agreements that will allow these newefficiencies to take place. This strategy does not simply involve a series of projects to strengthen existingcomponents of a conservation strategy, but changes the strategy itself. Table 10 lists a set of key short- and mediumtermresearch areas for cassava conservation and related strategies – activities that can only proceed afteragreements are reached at the research director level (national and international institutions) to implement the newstrategy.Table 10. Key short and medium-term (up to 10 years) research needs in cassava conservation andrelated activitiesResearch area PriorityLeadinstitutions Collaborators Funding needsComprehensive register ofglobal genebank holdings High IARCs Cassava genebanksConsultant(s) to visit orcorrespond with majorgenebanksDuplication of natl. progr.Preparation and shipping;High IARCs Cassava genebankscollections in IARCsexpanded in vitro storageIdentify priority collectionNational PGRMolecular studies ofareas (see also Section HighCassava genebanks; IARCsprogramslandraces; GIS studies5.1)Collection in priorityNational PGRExpeditions; post-collectionHighCassava genebanksregionsprogramsmanagementEstablish mechanisms forPeriodic meetings andIARCs and National PGRgenebank communication High PGR Networksconsultations; publishedprogramsand coordinationproceedingsEvaluation for traits ofNationalIARCs and National PGR Support for field and labimportance, especially for High breedingand programsstudiesfuture novel usesprogramsIn-country safety backupNational PGRField or in vitro facilities andMedCassava genebanksof natl. progr. collectionsprogramspersonnelCoordinatedIARCs; CassavaMedNational PGR programs Workshops; consultationsdocumentationgenebanksCryopreservation Med IARCs National PGR programs Pre/post freezing researchIn vitro very slow growthCassava genebanks; Media and environmentMed IARCsuniversitiesstudiesFlower induction for seedUniversities; cassava Physiology and hormoneMed IARCsgenebankgenebanksstudies64

Cassava Conservation Strategy14 Rationalizing a conservation strategy – Manihot wildspeciesApart from the fact that cassava and the wild species are part of the same genus, there is not much resemblancebetween the two groups in terms of conservation strategies. First, cassava is included in Annex I of the InternationalTreaty, while the wild Manihot species are not (though the wild accessions maintained by the CG Centres comeunder the Multilateral System through Article 15). This has implications for the support that may be available for thewild species management, and for international exchange. It is important that this be a topic of continuingconversation in the Treaty deliberations in the future.The wild species are seed-propagated in nature and cassava is universally vegetatively propagated in productionsystems (excepting the occasional volunteer seedlings that are found and tended by farmers). While cassavacollections are held in some 75 genebanks around the world, there are only a handful of wild Manihot collections.Brazil (EMBRAPA and Universidade de Brasilia) and CIAT hold the only major collections. While in vitro culture isroutine for cassava, most of the wild species have not been tested for their suitability to this conservation system.Where species have been tested, there are frequently species-specific requirements for optimum in vitro growth.Many of the conservation techniques and procedures are still experimental in the case of the wild species, sostandardizing recommendations for a coordinated global system is somewhat more difficult.Given the difficulties and costs of Manihot wild species conservation, it is expected that only a few institutions willtake on these challenges in a comprehensive manner, for the long-term future. Table 11 lists the critical researchareas for these species. It should be noted that even though there are only a few wild species genebanks, there is abroad array of research needs, in which many institutions can participate. Given the importance of the wild species tothe long-term goals of cassava improvement, all the tools need to be made available for their secure conservation,taxonomic classification, phylogeny, evaluation and use in breeding.Table 11. Key short and medium-term (up to 10 years) research needs in wild Manihot conservation and relatedactivities.LeadResearch area Priority institutions Collaborators Funding needsTaxonomyNational PGR Universities; botanical Post-graduate research andHighprograms gardens; Manihot genebanks national program scientistsThreats to habitatSupport to ongoing studies,Universities; nationalNational PGResp. Brazil and Mexico, toHighenvironmental agencies;programsinclude Manihot; policyManihot genebanksadvocacyCollection in priorityNational PGRSupport to national PGRHighManihot genebanksregionsprogramsprograms for direct costsCrossability studiesNational PGRSupport for field and labHighIARCsprogramsstudiesEvaluation for traits ofNational PGRSupport for field and labHighIARCsimportancePhylogeny studiesPopulation genetics forconservationSeed physiologyIn vitro techniquesMedMedMedMedprogramsNational PGRprogramsNational PGRprogramsManihotgenebanksManihotgenebanksUniversities; botanicalgardens; Manihot genebanksUniversities; ManihotgenebanksUniversitiesUniversities; IARCsstudiesPost-graduate research andnational program scientistsTheoretical studies; field testBasic standard lab studiesMedia and light/temperaturestudies65

Cassava Conservation StrategyTable 11. Key short and medium-term (up to 10 years) research needs in wild Manihot conservation and relatedactivities.LeadResearch area Priority institutions Collaborators Funding needsCryopreservationManihotLowUniversities; IARCsPre/post freezing researchgenebanksField propagationManihotLab/field studies under semicontrolledMedUniversities; IARCstechniquesgenebanksconditions66

Cassava Conservation StrategyAppendix I. Workshop programManihot Genetic Resources: Strategies for Long-Term Conservation30 April - 2 May 2008Calima RoomCIAT Headquarters, Cali, ColombiaWednesday 30 April 2008Presenters and ModeratorsWorkshop Opening8:00 - 8:15 Registration8:15 - 8:30 Welcome and Workshop Opening Joe Tohme8:30 - 9:00 Workshop goals and organization;program overview; introduction ofparticipants9:00 - 9:20 Goals and strategies of GCDT for CGRconservation; specific goals of studyClair HersheyLuigi Guarino9:20 - 9:50 Keynote Anthony BellottiSESSION I: Understanding Manihot Genetic DiversitySession Chair: Alfredo Alves9:50 - 10:20 Studies on Manihot evolution; mutants andnovel traits recently discovered10:20 - 10:40 Coffee10:40 - 11:10 Quantifying genetic diversity of landracevarieties: experiences and conclusionsfrom Latin America and Africa, andimplications for conservation strategies11:10 - 11:40 Conclusions from molecular fingerprintingof CIAT core collection and localgermplasm11:40 - 12:10 Functional classification of cassava genesand implications for the germplasm baseof cassava12:10 - 13:10 Lunch - CIAT restaurantLuiz CarvalhoPaula Ximena Hurtado andMartin FregenePeaingpen SarawatGerman Plata and JoeTohme67

Cassava Conservation Strategy13:10 - 13:40 Geographical Information Systems supportto cassava and wild Manihot collection andin situ conservation13:40 - 14:10 Cassava pre-breeding at CIAT andimplications for germplasm conservation,evaluation and exchange14:10 - 14:30 CoffeeLuigi Guarino and AndrewJarvisHernan CeballosTOUR14:30 - 16:30 Tour of CIAT Genetic Resources Unit Daniel Debouck, GracielaMafla, Maritza Cuervo,Roosevelt Escobar, CesarOcampo, EricsonAranzales, AngelaHernandez16:30 - 17:30 Wrap of GRU tour: Criteria and studies fordecision-making in cassava conservationDaniel DebouckThursday 1 May 2008SESSION II: Status and Needs of Existing GenebanksSession Chair:Daniel Debouck8:00 - 8:30 Status and needs of cassava germplasmconservation in Brazil8:30 - 9:00 Wild Manihot collection and conservationin Brazil9:00 - 9:30 Status and needs of cassava germplasmconservation in Meso-America and theCaribbean9:30 - 10:00 Status and needs of cassava germplasmconservation in Venezuela, Colombia,Ecuador and the GuyanasWania FukudaAlfredo AlvesSergio RodriguezAntonio Lopez10:00 - 10:20 Coffee10:20 - 10:40 Status and needs of cassava germplasmconservation in Peru, Bolivia, Paraguayand Argentina10:40 - 11:10 Status and needs of cassava germplasmconservation in AfricaLlerme Rios LobosPaul Ilona11:10 - 11:40 Overview of IITA cassava genebank Dominique Dumet68

Cassava Conservation Strategy11:40 - 12:10 Status and needs of cassava germplasmconservation in AsiaPeaigpen Sarawat12:10 - 13:10 Lunch – CIAT restaurant1:10 - 2:30 Group discussion: Genebank surveys --status and follow-upModerator: Clair HersheySESSION III: Elements of a Long-Term Conservation Strategy (Part I) (Group discussions)Session Chair:Llerme Rios14:30 - 15:00 Characterization, evaluation andinformation managementModerator: Antonio Lopez15:00 - 15:20 Coffee15:20 - 16:10 Regeneration, conservation and safetyduplication needs16:10 - 16:40 Networks and international cooperativeprogramsModerator: SergioRodriguezModerator: DominiqueDumet16:40 - 17:00 Capacity-building needs Moderator: Alfredo Alves19:00 Depart to Cali20:00 Dinner in CaliSESSION III: Elements of a Long-Term Conservation Strategy (Part II) (Group discussions)Session Chair:Wania FukudaFriday 2 May 20088:00 - 8:20 Bioversity International: goals andstrategies for crop genetic resources in theAmericas8:20 - 8:50 International exchange: genebank needsand obligations8:50 - 9:20 Conservation and germplasm users:developing a collaborative relationship9:20 - 9:50 Genebank services and service providers:a look at future scenarios9:50 - 10:10 Coffee10:10 - 10:40 Criteria for the most important collectionsfor GCDT support: grouprecommendationsXavier ScheldemanModerator: Paul IlonaModerator: SergioRodriguezModerator: Antonio LopezModerator: Luiz Carvalho69

Cassava Conservation Strategy10:40 - 11:10 Conservation strategies for smallercollections11:10 - 11:40 Conservation in perpetuity: overview ofwhat is requiredModerator: PeaingpenSarawatModerator: Daniel Debouck11:40 - 12:30 Wrap-up and conclusions Clair Hershey and DanielDebouck12:30 - 13:30 LunchCassava Descriptors - a Mini-Workshop of Bioversity International13:30 - 14:30 A strategy for developing cassavadescriptorsXavier Scheldeman andClara Ines QuinteroGonzales70

Cassava Conservation StrategyAppendix II. Workshop list of presenters and participantsBRAZIL* Alfredo Augusto Cunha AlvesResearch ScientistCurator – Manihot wild speciesEmbrapa - Cassava and Tropical Fruits (CNPMF)Caixa Postal 00744.380-000 Cruz das Almas, BahiaBRAZILaalves@cnpmf.embrapa.br* Luiz Joaquim Castelo Branco CarvalhoResearch ScientistEmbrapa – Genetic Resources and Biotechnology (CENARGEN)Biotechnology Building, Laboratory of Biochemistry and BiophysicsCaixa Postal 0237270770-900 Brasilia, DFBRAZILcarvalho@cenargen.embrapa.br* Wania Maria Gonçalves FukudaResearch ScientistCurator – National Cassava GenebankEmbrapa - Cassava and Tropical Fruits (CNPMF)Caixa Postal 00744.380-000 Cruz das Almas, BahiaBrazilwfukuda@cnpmf.embrapa.brCOLOMBIA* Antonio Lopez MontesAgroecosystems Research ScientistCorpoica - TuripanáMonteria, BolivarCOLOMBIAajlopez@corpoica.org.co71

Cassava Conservation StrategyCUBA* Sergio J. Rodríguez MoralesDirector – INIVITSanto Domingo, Villa ClaraCUBAsergio@inivit.co.cuPERU* Llermé Ríos LobosSpecialist – Genetic Resources and Biotechnology (SUDIRGEB)INIA, La MolinaLimaPERUrioslobo@hotmail.comTHAILAND* Peaingpen SarawatSenior Agricultural ScientistKhon Kaen Field Crop Research CenterOffice of Agricultural Research and Development, Region 3Khon Kaen, 40000THAILANDpeaingpen@yahoo.co.ukUSA* Clair HersheyWorkshop coordinator2019 Locust Grove RdManheim, PA 17545USAchh23@cornell.eduGlobal Crop Diversity Trust* Luigi GuarinoGlobal Crop Diversity Trustc/o FAOViale delle Terme di Caracalla00153 RomeItalyluigi.guarino@croptrust.org72

Cassava Conservation StrategyBioversity InternationalRegional Office for the AmericasCIATAA 67-13Cali, ValleColombia* Xavier Scheldeman x.scheldeman@cgiar.org* Clara Ines Quintero Gonzalez claraines88@hotmail.comIITAIITAPMB 5320Ibadan, Oyo StateNIGERIA* Dominique DumetHead of the Genebankd.dumet@cgiar.org* Paul IlonaHead – Cassava Regional Trials Networkp.ilona@cgiar.orgCIATAA 67-13Cali, ValleColombiaGenetic Resources Unit* Daniel Debouck d.debouck@cgiar.org* Angela Hernández a.hernandez@cgiar.org* César Ocampo c.ocampo@cgiar.org* Ericson Aranzales e.aranzales@cgiar.org* Graciela Mafla g.mafla@cgiar.org* Maritza Cuervo m.cuervo@cgiar.org* Roosevelt Escobar r.escobar@cgiar.orgJosefina Martínez Realpe (sec.)j.m.realpe@cgiar.orgAgrobiodiversity and Biotechnology Project73

Cassava Conservation Strategy* Joe Tohme j.tohme@cgiar.org* German Plata g.a.plata@cgiar.orgLee Calvertl.calvert@cgiar.orgCassava Project* Hernan Ceballos h.ceballos@cgiar.orgMartin Fregenem.fregene@cgiar.org* Anthony Bellotti a.bellotti@cgiar.orgElizabeth AlvarezDominique DufourSarah AdeyemoOlalekan AkinboJuan PérezFernando CalleGustavo JaramilloNelson MoranteTeresa Sáncheze.alvarez@cgiar.orgd.dufour@cgiar.orgs.adeyemo@cgiar.orgo.akinbo@cgiar.orgj.c.perez@cgiar.orgcallecallef@hotmail.comgjo97@hotmail.comnelmorante@hotmail.comtesa045@hotmail.com* Paula Ximena Hurtado p.x.hurtado@cgiar.orgJuliana Chacónj.chacon@cgiar.orgCLAYUCABernardo Ospinab.ospina@cgiar.orgLand Use ProjectAndrew Jarvisa.jarvis@cgiar.orgICAIsabel Natalia Salas T.CoordinadoraOficina del ICA en CIATConvenio ICA-CIAT___________________isabelnatalia@hotmail.com* Presenters (25)74

Cassava Conservation StrategyAppendix III. Cassava and wild Manihot survey formFeb. 2008A Survey to Build a Global Conservation Strategy for Cassava and Wild Manihot SpeciesBackgroundThe Global Crop Diversity Trust is supporting efforts to develop strategies for the efficient and effectiveconservation of crop diversity on both a regional and global crop basis. This questionnaire has beendeveloped in order to seek the advice and input of representatives of the world’s major cassava and wildManihot collections in the development of the conservation strategy. In particular the questionnaire seeksto assess the status of cassava conservation throughout the world and to identify major needs. It isintended that the Global Crop Diversity Trust (Trust) will base its support for the conservation of cassavagenetic resources on this strategy, once developed and adopted. We kindly request you to reviewquestionnaire in advance, improve it and use it as a reference for your presentation or the discussion. Weare keen to ensure your active participation in the development of the global cassava conservationstrategy.This survey is divided into two sections.A. Cultivated cassava collectionsB. Wild Manihot species collectionsPlease fill out all sections that are relevant to your situation. If you manage only cultivatedspecies, there is no need to complete Section B, or if you manage only wild species, there is noneed to complete Section A.SECTION A. CULTIVATED CASSAVA COLLECTIONS1. Institutional information1.1. Name and address of organization holding/maintaining the collectionName:Address:City:Postal Code:Country:Web site:Curator in charge of the collection:Name:Address:City:Telephone:Fax:Email:Name of respondent to this questionnaire if different than aboveContact details:Date ofresponse:75

Cassava Conservation Strategy1.2. Is the organization holding the cassava collection:1 An independent organization1 Part of a larger organization1 A government organization1 Other (specify): _____________________________________In the case of (B) please provide the name and address of the larger organization:_______________________________________________________________________________________________________________1.3. Who is financing of the conservation of the cassava collection?1 Government ________%1 Private sector ________%1 International or regional funding ________%1 Other (specify): ___________________________ __________%1.4. Who is the legal owner of the collection?1 Institution in charge1 Other (specify): _____________________________________1.5. How much time is devoted to the management of the cassava collection?_____ Full time equivalent (fte) per year (1 fte means that a person is working for 100% on thecassava collection)2. Details on the cassava collection2.1. Year of formal establishment of the collection:____________________2.2. What is the main objective of the conservation of the collection (in terms of use and ofconservation): ___________________________________________________________2.3. Present size of the cassava collection:Type of cassava germplasmFarmers’ varietiesBreeders’ varietiesExperimental materialsOthersTotalTotal numberof accessions% available fordistribution2.4. Origin of the collection. Please state the percentage of accessions included in thecollection of:o Local origin previously collected in own country: _____%o Introduced from abroad from the centre of diversity: _____%o Introduced from abroad, outside the centre of diversity: ____ %o Other origin _____%3. Management of the cassava collection3.1 Acquisition76

Cassava Conservation Strategy3.1.1. Was the collection increased during the last 10 years with new accessions (after theinternational agreement on genetic resources movement)?1 yes 1 no- If yes, how many new accessions were included of the following:o Landrace varieties: _______o Modern varieties: ________o Breeding material: _______o Other: ______3.1.2. How was the acquisition of the newly obtained germplasm conducted?1 Collecting in own country1 Collecting in other countries1 Introduction from other collections, institutes or private organizations1 Other sources – please specify: _________________________________3.1.3. Are there important gaps in the collection?1 yes 1 noo If yes, what are the main gaps: ____________________________________________________________________________________________3.1.4. Do you plan to fill in these gaps in the next 5 years? 1 yes 1 partly 1noo If yes, how1 Collection1 Introduction1 Othero If no, what are the main reasons:_____________________________________3.2 Storage and maintenance (seed, in vitro, field)3.2.1. Please indicate how germplasm is maintained for long- and medium-term storage (check allboxes that apply and indicate percent of total)Type of germplasmBotanicalseedsField In vitro Greenhouse/screenhouseCryo-conservationFarmers’ varieties 1 %____ 1 %____ 1 %____ 1 %____ 1 %____Breeders’ varieties 1 %____ 1 %____ 1 %____ 1 %____ 1 %____Experimental1 %____ 1 %____ 1 %____ 1 %____ 1 %____materialsOthers 1 %____ 1 %____ 1 %____ 1 %____ 1 %____3.2.2. What are the storage facilities and conditions of the cassava genebank?Type offacilityTemperature(ºC)RH %PackingmaterialBotanical seedsFieldIn vitroGreenhouse/screenhouseCryo-conservation3.2.3. What is the field (F) or greenhouse/screenhouse (G) maintenance protocol for the cassavagenebank?77

Cassava Conservation StrategyNumber ofplants peraccession*Distancebetween rowsDistancebetween plantsFarmers’ varieties F:____ G:____ F:____ G:____ F:____ G:____Breeders’ varieties F:____ G:____ F:____ G:____ F:____ G:____ExperimentalF:____ G:____ F:____ G:____ F:____ G:____materialsOthers F:____ G:____ F:____ G:____ F:____ G:____* In case of 1, it would be considered as 1 plant/pot within a greenhouse/screenhouse.3.2.4. Do you apply tests to control the quality of stored germplasm?1 yes 1 partly 1 noIf yes or partly, which tests are conducted?1 Germination test of sexual seed1 Control of the vitality and health of stem cuttings1 Control of true-to-type-ness of in vitro plantlets1 Other ______________________________________________________________3.3 Regeneration3.3.1. Method of regeneration: Please indicate how the cassava germplasm is regenerated.o As population (sexual seed): 1 yes 1 noo Vegetative by means of stem cuttings or other: 1 yes 1 noo In vitro: 1 yes 1 noNote: More than one option for the same type of material is possible3.3.2. On how many plants (pl) is the regeneration (population) normally based?1 < 10 pl 1 10- 20 pl 121 – 30pl 1> 30 pl3.3.3. How many cuttings (cu) are planted for the next vegetative regeneration?1 < 15 cu 1 15 –30 cu 131 to 45 cu 1> 45 cu3.3.4. How many plantlets (pl) are maintained for in vitro regeneration?1 < 10 pl 1 11 –30 pl 1 >30 pl3.3.5. Annual capacity of regeneration/multiplication (number of accessions)Type of germplasmAs population(sexual seed)Farmers’ varietiesBreeders’ varietiesExperimental materialsOthersNote: More than one option for the same type of material is possibleVegetative bymeans of cuttingsIn vitro3.3.6. Percentage of the collection that needs to be urgently regenerated:ooPrimitive forms _______%Modern varieties _______%78

Cassava Conservation StrategyoOthers & research material, etc_______%3.4 Identification (classification) and characterization3.4.1. Which type of material of the cassava collection is characterized?Type of germplasmMorphologicalcharacterizationMolecularcharacterizationFarmers’ varieties 1 yes 1 no 1 yes 1 noBreeders’ varieties 1 yes 1 no 1 yes 1 noExperimental materials 1 yes 1 no 1 yes 1 noOthers 1 yes 1 no 1 yes 1 no3.4.2. Which type of descriptor list is used for characterization?1 Standard list of IPGRI1 Your own independently developed list1 List developed by another organization (specify): ……………………….3.5 Documentation and access to information about the collection3.5.1. Do you use a database information system for the management of the cassava collection?1 yes 1 partly 1 noIf yes, what software is used for the documentation? _______________________3.5.2. Which kind of data of the collection has been computerized? Please check the appropriateanswer.Type of germplasm Passport data Characterization/ Management data*evaluation dataFarmers’ varieties 1 yes 1 partly 1 no 1 yes 1 partly 1 no 1 yes 1 partly 1 noBreeders’ varieties 1 yes 1 partly 1 no 1 yes 1 partly 1 no 1 yes 1 partly 1 noExperimental1 yes 1 partly 1 no 1 yes 1 partly 1 no 1 yes 1 partly 1 nomaterialsOthers 1 yes 1 partly 1 no 1 yes 1 partly 1 no 1 yes 1 partly 1 no* data related to storage, germination, distribution, etc.3.5.3. In case the cassava collection is not computerized, are there plans to do so in the future?1 No plans1 Computerization planned within 3 years3.5.4. Is information of the cassava collection accessible through the Internet?1yes 1partly 1 no URL: ____________3.5.5. Are data of the cassava collection included in other databases? If yes or partly, specify thedatabaseo National 1 yes 1 partly 1 no _______________________________o Regional 1 yes 1 partly 1 no _______________________________o International 1 yes 1 partly 1 no _______________________________3.6 Health of germplasm3.6.1. Is the cassava collection affected by diseases that can restrict the distribution of the germplasm?79

Cassava Conservation Strategy1 yes 1 slightly, or only few accessions 1 noIf yes or slightly, which types of diseases are causing this restriction?1 Seed-borne diseases in sexual seed1 Infection in maintained plants1 Virus or viroid-infected in vitro plantlets3.6.2. If in vitro samples are distributed within the country are they virus indexed?1 yes 1 some 1 no3.6.3. If in vitro samples are distributed outside the country are they virus indexed?1 yes 1 some 1 no3.6.4. Is knowledge available at your institution and are there facilities for eradication of these diseases?1 yes 1 limited 1 no3.6.5. Do you need assistance to improve the health status of the cassava collection?1 yes 1 limited 1 noIf yes, what type of assistance will be required?1) ____________________________________________________________________2) ____________________________________________________________________3) ____________________________________________________________________3.7 Distribution3.7.1. Do you distribute material outside your institute, within the country? 1 yes 1 no(If no, please go to Section 3.7.3)3.7.2. How many accessions have you distributed within the country in the past 5 years to thefollowing users?Sent to: Wild species LandracevarietiesBreeders, otherresearchersFarmersGenebanksExtensionistsOthers (specify)ExperimentalmaterialsOther3.7.3. Do you distribute material outside the country? 1 yes 1 no(If no, please go to Section 3.8)3.7.4. How many accessions have you distributed outside the country in the past 5 years to thefollowing users?Sent to: Wild species LandracevarietiesBreeders, otherExperimentalmaterialsOther80

Cassava Conservation StrategyresearchersFarmersGenebanksExtensionistsOthers (specify)_________________3.7.5. Do you set specific conditions for distribution? 1 yes 1 noIf yes, please specify: ___________________________________________________________3.7.6. Is the germplasm sufficiently available for distribution?o Sexual seed: 1 yes 1 partly 1 noo Cuttings: 1 yes 1 partly 1 noo In vitro plantlets: 1 yes 1 partly 1 no3.7.7. Compared to 5 years ago, are you now distributing less, the same amount, or moregermplasm? 1 less 1 same 1 more3.7.8. Do you expect to distribute less, the same amount, or more germplasm 5 years from now?1 less 1 same 1 more3.7.9. Do you keep records of the distribution? 1 yes 1 No3.7.10 What information is kept in these records?____________________________________________________________________________________________________________________________________________________________________________________________________________3.7.11 Do you request and get any feedback from the recipients? 1 yes 1 No3.7.12. What use is made of the information? ________________________________________________________________________________________________________3.7.13. How are the services of the collection publicized to users and how effective are thesemethods in terms of increased use of the collection?Scientific publicationsInstitution reportsExtension leafletsOral presentationsVisits to collectionOther (specify)___________________High impact Medium impact Low impact Don’t know3.7.14. Have any requests for material been refused? If yes, specify: _____________________________________________________________________________________________3.7.15. How do the users of the germplasm influence the management of the collection? Indicateyes or no in table below.81

Cassava Conservation StrategyBreeders, otherresearchersFarmersGenebanksExtensionistsOthers (specify)______________Throughfeedback onmaterial?Through formalconsultationsThroughparticipating nthe governingbody of thegenebankOther(specify)________________________3.8 Safety duplication3.8.1. Are the accessions of the cassava collection safety-duplicated in another genebank?1 yes ________% duplicated 1 no 1 uncertainIf yes, please specify where the germplasm is safety-duplicated _______________________Storage conditions ___________________________________________________________3.8.2. Is there any germplasm of other cassava collections safety-duplicated at your facilities?1 yes 1 noIf yes, can you specify the name of the holder of the cassava collection safety-duplicated at yourgenebank including the number of accessions duplicated?Collection holder: ______________________________ Accessions duplicated (no.): _____3.9 General management3.9.1. How many staff are working on the collection (full-time staff equivalents)?Mark the appropriate boxes with an X.5In the field or greenhouse/screenhousescientiststechniciansfield workersstudentsIn the labscientiststechniciansstudents3.9.2. Have you established a quality management system or written procedures and protocolsfor: (check each that applies)1 Acquisition (including collecting, introduction and exchange)1 Regeneration82

Cassava Conservation Strategy1 Characterization1 Storage and maintenance1 Documentation1 Health of germplasm1 Distribution1 Safety duplication3.9.3. In case you have procedures and protocols, are you able to provide the Trust with thisinformation or include a copy of it? 1 yes 1 no3.9.4.Does the existing capacity in numbers and skills meet the needs of the collection in the longterm (e.g. greater than 10 years)? 1 yes 1 noIf no, please describe what is needed. ______________________________________________________________________________________________________________4. Utilization of the cassava collection4.1. For what purposes is the cassava collection used? Check all that apply.1 Research (e.g. taxonomic, biosystematic, inheritance, evolutionary studies)1 Characterization1 Evaluation for important productivity & quality traits1 Plant breeding1 Biotechnology, e.g. gene isolation, molecular studies, functional genomics, etc1 Distribution to farmers1 Return of germplasm to country of origin4.2. Do you have a systematic evaluation program to evaluate the collection for traits?1 yes 1 planned 1 noIf yes, can you list the most important traits the cassava collection is evaluated for?___________________________________________________________________________________________________________________________________________________________________________________________________5. Networks of cassava genetic resources5.1. Do you collaborate in (a) network(s) as a cassava collection holder?1 yes 1 noExchange ofgermplasmExchange ofinformationTrainingNational level Regional level Global level NoneOther (specify):______________5.2. What are the major objectives of the network(s) in which you participate?83

Cassava Conservation Strategy1 Joint conservation of cassava germplasm1 Evaluation or characterization of cassava germplasm1 Establishment of central cassava database1 Rationalization of the collections1 Safety duplication of cassava germplasmRemark: more than one option is possible5.3. Do you consider a worldwide network for cassava genetic resources important and would youconsider participating in such network?1 yes 1 no5.4. What will be your major interest for participation in a cassava genetic resources network?6. Policies with regard to access of the cassava collection6.1. What is your policy regarding distribution of cassava germplasm?1 Distribution to any bona fide users, without further conditions1 Distribution to any bona fide users after signing of a MTA (Material Transfer Agreement)1 Distribution only to users in own country1 Distribution only to users in certain countries after signing of a MTA1 Distribution only on a mutually agreed exchange basis1 Other flows of distribution, please specify: ____________________________6.2. Cost for distribution of cassava germplasm:1 No cost, free distribution1 No cost, but only on the basis of reciprocal exchange of material1 Request to contribute for processing and shipping, specify amount: ________1 Request to pay for each requested accession, specify amount: ____________1 Other conditions requested, please specify: ___________________________6.3. Please attach examples of your organization’s long-term commitment to long termconservation of cassava collection, for example:1 Legal status1 Institutional constitution1 Mandates1 Published strategic plans1 National conservation strategy1 Action plans1 Other: ___________________________7. Future developments regarding the cassava collection7.1. Will the cassava collection be extended with new material or rationalized in the next fiveyears?1 The collection will keep approximately the same size1 The collection will be expanded to a limited extent (5-10 %)1 The collection will be substantially increased (> 20%)1 The collection will be reduced due to duplication with other collections and internalrationalization1 The collection will be reduced as a result of lack of funding or facilities7.2. Are there any constraints for maintenance of the cassava collection? 1 yes 1 noIf yes, what type of constraints do you face?1 Insufficiently trained staff84

Cassava Conservation Strategy1 Regeneration capacity to maintain the collection limited1 Facilities for optimal maintenance of the collection not satisfactory1 Others (please specify): ______________________________________________7.3. Will some of the above constraints result in a loss of cassava germplasm?1 yes 1 only incidentally 1 noIf yes, what is the most important constraint, which may contribute to genetic erosion within thecollection? _____________________________________________________8. Core collection8.1. Have you identified a core collection within the total collection? 1 yes 1 no8.2. If no, is there any plan to do so in the future? 1 yes 1 no8.3. If yes, how many accessions are included in the core?Number ____ Percent of total ______8.4. What criteria were used to define the core collection?1 Geographic or agro-ecological origin1 Morphological characterization1 Molecular characterization1 Agronomic and market traits1 Others (please specify): ______________________________________________9. Further remarksIf you are responsible for a collection of wild Manihot species, please complete Section II.Otherwise, please see directions at the end of the document for returning the completed survey.85

Cassava Conservation StrategySECTION B. WILD MANIHOT COLLECTIONS1. Institutional information1.1. Name and address of organization holding/maintaining the collectionName:Address:City:Postal Code:Country:Web site:Curator in charge of the collection:Name:Address:City:Telephone:Fax:Email:Name of respondent to this questionnaire if different then aboveContact details:Date ofresponse:1.2. Is the organization holding the Manihot collection:1 An independent organization1 Part of a larger organization1 A government organization1 Other (specify): _____________________________________In the case of (B) please provide the name and address of the larger organization:_______________________________________________________________________________________________________________1.3. Who is financing of the conservation of the Manihot collection?1 Government ________%1 Private sector ________%1 International or regional funding ________%1 Other (specify): ___________________________ __________%1.4. Who is the legal owner of the collection?1 Institution in charge1 Other (specify): _____________________________________1.5. How much time is devoted to the management of the Manihot collection?_____ Full time equivalent (fte) per year (1 fte means that a person is working for 100% on thecassava collection)86

Cassava Conservation Strategy2. Details on the Manihot collection2.1. Year of formal establishment of the collection: _______________________________2.2. What is the main objective of the conservation of the collection (in terms of use and ofconservation): ______________________________________________________________Note: Please attach or insert a list of the number of accessions of each species, along with theirorigin.2.3. Present size of the Manihot collection:Number of species: _________Total number of accessions: __________% available for distribution: _________2.4. Origin of the collection. Please state the percentage of accessions included in the collectionof:Local origin previously collected in own country: ____ %Introduced from abroad from the centre of diversity: ____ %Introduced from abroad, outside the centre of diversity: ____ %Other origin: ____ %3. Management of the Manihot collection3.1 Acquisition3.1.1. Was the collection increased during the last 10 years with new accessions (after theinternational agreement on genetic resources movement)? 1 yes 1 noIf yes, which species were included, and how many new accessions of each (insert or attach separatelist if desired):3.1.2. How was the acquisition of the newly obtained germplasm conducted?1 Collecting in own country1 Collecting in other countries1 Introduction from other collections, institutes or private organizations1 Other sources – please specify: _________________________________3.1.3. Are there important gaps in the collection? 1 yes 1 noIf yes, what are the main gaps: ________________________________3.1.4. Do you plan to fill in these gaps in the next 5 years? 1 yes 1 partly 1noIf yes, how1 Collection1 Introduction1 OtherIf no, what are the main reasons:_____________________________________3.2 Storage and maintenance3.2.1. Please indicate how germplasm is maintained for long- and medium-term storage87

Cassava Conservation Strategy(check all boxes that apply and indicate percent of total)Botanical seeds:1 %____Field:1 %____In vitro:1 %____Greenhouse/screenhouse: 1 %____Cryo-conservation:1 %____3.2.2. What are the storage facilities and conditions of the Manihot species genebank?Type of facilityPart ofcollectionrepresented(%)Temp.(ºC)RH(%)PackingmaterialBotanical seedsFieldIn vitroGreenhouse/screenhouseCryo-conservation3.2.3. What is the field (F)or greenhouse/screenhouse (G) maintenance protocol for the Manihotgenebank?Number of plants per accession*:Distance between rows:Distance between plants:F:____ G:____F:____ G:____F:____ G:____* In case of 1, it would be considered as 1 plant/pot within a greenhouse/screenhouse.3.2.4. Do you apply tests to control the quality of stored germplasm? 1 yes 1 partly 1 noIf yes or partly, which tests are conducted?1 Germination test of sexual seed1 Control of the vitality and health of stem cuttings1 Control of true-to-type-ness of in vitro plantlets1 Other ______________________________________________________________3.3 Regeneration3.3.1. Method of regeneration: Please indicate how the Manihot germplasm is regenerated.As population (sexual seed)1 yes 1 noVegetative by means of cuttings or other 1 yes 1 noIn vitro1 yes 1 noNote: More than one option for the same type of material is possible88

Cassava Conservation Strategy3.3.2. On how many plants (pl) is the regeneration (population) normally based?1 < 10 pl 1 10- 20 pl 121 – 30pl 1> 30 pl3.3.3. How many cuttings (cu) are planted for the next vegetative regeneration?1 < 15 cu 1 15 –30 cu 131 to 45 cu 1> 45 cu3.3.4. How many plantlets (pl) are maintained for in vitro regeneration?1 < 10 pl 1 11 –30 pl 1 >30 pl3.3.5. Annual capacity of regeneration/multiplication (number of accessions)As population (sexual seed) ________ (no. of seeds)Vegetative by means of cuttings ________ (no. of cuttings)In vitro _________ (no. of in vitro plantlets)Note: More than one option for the same type of material is possible3.3.6. Percentage of the collection that needs to be urgently regenerated: _______ %3.4 Identification (classification) and characterization3.4.1. Is the collection of wild Manihot species taxonomically classified?1 yes 1 no 1 partial (percent classified _______%)3.4.2. Do you have assistance of a taxonomist for the classification of the Manihot germplasm?1 yes 1some 1 no3.4.3. Which type of characterization is done?Morphological characterization 1 yes 1 noMolecular characterization 1 yes 1 no3.4.4. If morphological characterization is done, which type of descriptor list is used forcharacterization?1 Standard list of IPGRI1 Your own independently developed list1 List developed by another organization (specify): ……………………….3.5 Documentation and access to information about the collection3.5.1. Do you use a database information system for the management of the Manihot speciescollection?1 yes 1 partly 1 noIf yes, what software is used for the documentation? _______________________3.5.2. Which kind of data of the collection has been computerized? Please check the appropriateanswer.Passport dataCharacterization/ evaluation data1 yes 1 partly 1 no1 yes 1 partly 1 no89

Cassava Conservation StrategyManagement data*1 yes 1 partly 1 no* data related to storage, germination, distribution, etc.3.5.3. In case the Manihot collection is not computerized, are there plans to do so in the future?1 No plans1 Computerization planned within 3 years3.5.4. Is information of the Manihot collection accessible through the Internet?1yes 1partly 1 no URL: ____________3.5.5. Are data of the cassava collection included in other databases? If yes or partly, specify thedatabaseNationalRegionalInternational1 yes 1 partly 1 no _______________________________1 yes 1 partly 1 no _______________________________1 yes 1 partly 1 no _______________________________3.6 Health of germplasm3.6.1. Is the cassava collection affected by diseases that can restrict the distribution of the germplasm?1 yes 1 slightly, or only few accessions 1 noIf yes or slightly, which types of diseases are causing this restriction?1 Seed-borne diseases in sexual seed1 Infection in maintained plants1 Virus or viroid-infected in vitro plantlets3.6.2. If in vitro samples are distributed within the country are they virus indexed?1 yes 1 some 1 no3.6.3. If in vitro samples are distributed outside the country are they virus indexed?1 yes 1 some 1 no3.6.4. Is knowledge available at your institution and are there facilities for eradication of these diseases?1 yes 1 limited 1 no3.6.5. Do you need assistance to improve the health status of the cassava collection?1 yes 1 limited 1 noIf yes, what type of assistance will be required?1) ____________________________________________________________________2) ____________________________________________________________________3) ____________________________________________________________________3.7 Distribution3.7.1. Do you distribute material outside your institute, within the country?1 yes 1 noIf no, go to Section 3.7.390

Cassava Conservation Strategy3.7.2. How many accessions have you distributed within the country in the past 3 years to thefollowing users?Sent to: Wild species LandracevarietiesBreeders, otherresearchersFarmersGenebanksExtensionistsOthers (specify)ExperimentalmaterialsOther3.7.3. Do you distribute material outside the country?1 yes 1 noIf no, go to Section 3.83.7.4. How many accessions have you distributed outside the country in the past 3 years to thefollowing users?Sent to: Wild species LandracevarietiesBreeders, otherresearchersFarmersGenebanksExtensionistsOthers (specify)_________________ExperimentalmaterialsOther3.7.5. Do you set specific conditions for distribution? 1 yes 1 noIf yes, please specify: ___________________________________________________________3.7.6. Is the germplasm sufficiently available for distribution?Sexual seed: 1 yes 1 partly 1 noCuttings: 1 yes 1 partly 1 noIn vitro plantlets: 1 yes 1 partly 1 no3.7.7. Compared to 5 years ago, are you now distributing less, the same amount, or moregermplasm? 1 less 1 same 1 more3.7.8. Do you expect to distribute less, the same amount, or more germplasm 5 years from now?1 less 1 same 1 more3.7.9. Do you keep records of the distribution? 1 yes 1 No3.7.10. What information is kept in these records?________________________________________________________________________________________________________________________________________91

Cassava Conservation Strategy3.7.11. Do you request and get any feedback from the recipients? 1 yes 1 No3.7.12. What use is made of the information? ________________________________________________________________________________________________________3.7.13. How are the services of the collection publicized to users and how effective are thesemethods in terms of increased use of the collection?Scientific publicationsInstitution reportsExtension leafletsOral presentationsVisits to collectionOther (specify)___________________High impact Medium impact Low impact Don’t know3.7.14. Have any requests for material been refused? If yes, specify: _____________________________________________________________________________________________________3.7.15. How do the users of the germplasm influence the management of the collection? Indicateyes or no in table below.Breeders, otherresearchersFarmersGenebanksExtensionistsOthers (specify)______________Throughfeedback onmaterial?Through formalconsultationsThroughparticipatingin thegovernance ofthe genebankOther(specify)________________________3.8 Safety duplication3.8.1. Are the accessions of the Manihot collection safety-duplicated in another genebank?1 yes ________% duplicated 1 no 1 uncertainIf yes, please specify where the germplasm is safety-duplicated _________________Storage conditions _____________________________________________________3.8.2. Is there any germplasm of other Manihot collections safety-duplicated at your facilities?1 yes 1 noIf yes, can you specify the name of the holder of the cassava collection safety-duplicated at yourgenebank including the number of accessions duplicated?Collection holder: ______________________________ Accessions duplicated (no.): _____92

Cassava Conservation Strategy3.9 General management3.9.1. How many staff are working on the collection (full-time staff equivalents)?Mark the appropriate boxes with an X.5In the field or greenhouse/screenhousescientiststechniciansfield workersstudentsIn the labscientiststechniciansstudents3.9.2. Have you established a quality management system or written procedures and protocols for(check each that applies):1 Acquisition (including collecting, introduction and exchange)1 Regeneration1 Characterization1 Storage and maintenance1 Documentation1 Health of germplasm1 Distribution1 Safety duplication3.9.3. In case you have procedures and protocols, are you able to provide the Trust with thisinformation or include a copy of it? 1 yes 1 no3.9.4. Does the existing capacity in numbers and skills meet the needs of the collection in the longterm (e.g. greater than 10 years)? 1 yes 1 noIf no, please describe what is needed. ______________________________________________________________________________________________________________4. Utilization of the Manihot collection germplasm4.1. For what purposes is the Manihot collection used? Check all that apply.1 Research (e.g. taxonomic, biosystematic, inheritance, evolutionary studies)1 Characterization1 Evaluation for important productivity & quality traits1 Plant breeding1 Biotechnology, e.g. gene isolation, molecular studies, functional genomics, etc1 Distribution to farmers1 Return of germplasm to country of origin4.2. Do you have a systematic evaluation program to evaluate the collection for traits?1 yes 1 planned 1 noIf yes, can you list the most important traits the cassava collection is evaluated for?93

Cassava Conservation Strategy___________________________________________________________________________________________________________________________________________________________________________________________________5. Networks of Manihot genetic resources5.1. Do you collaborate in (a) network(s) as a cassava collection holder?1 yes 1 noExchange ofgermplasmExchange ofinformationTrainingNational level Regional level Global level NoneOther (specify):______________5.2. What are the major objectives of the network(s) in which you participate?1 Joint conservation of cassava germplasm1 Evaluation or characterization of cassava germplasm1 Establishment of central cassava database1 Rationalization of the collections1 Safety duplication of cassava germplasmNote: more than one option is possible5.3. Do you consider a worldwide network for Manihot genetic resources important and would youconsider participating in such network?1 yes 1 no5.4. What will be your major interest for participation in a Manihot genetic resources network?____________________________________________________________________________6. Policies with regard to access of the Manihot collection6.1. What is your policy regarding distribution of Manihot germplasm?1 Distribution to any bona fide users, without further conditions1 Distribution to any bona fide users after signing of a MTA (Material Transfer Agreement)1 Distribution only to users in own country1 Distribution only to users in certain countries after signing of a MTA1 Distribution only on a mutually agreed exchange basis1 Other flows of distribution, please specify: ____________________________6.2. Cost for distribution of Manihot germplasm:1 No cost, free distribution1 No cost, but only on the basis of reciprocal exchange of material1 Request to contribute for processing and shipping, specify amount: ________1 Request to pay for each requested accession, specify amount: ____________1 Other conditions requested, please specify: ___________________________6.3. Please insert or attach examples of your organization’s long-term commitment to long termconservation of Manihot collection, for example:94

Cassava Conservation Strategy1 Legal status1 Institutional constitution1 Mandates1 Published strategic plans1 National conservation strategy1 Action plans1 Other: ___________________________7. Future developments regarding the Manihot collection7.1. Will the Manihot collection be extended with new material or rationalized in the next fiveyears?1 The collection will remain approximately the same size1 The collection will be expanded to a limited extent (5-10 %)1 The collection will be substantially increased (> 20%)1 The collection will be reduced due to duplication with other collections and internalrationalization1 The collection will be reduced as a result of lack of funding or facilities7.2. Are there any constraints for maintenance of the Manihot collection? 1 yes 1 noIf yes, what type of constraints do you face?1 Insufficiently trained staff1 Regeneration capacity to maintain the collection limited1 Facilities for optimal maintenance of the collection not satisfactory1 Others (please specify): ______________________________________________7.3. Will some of the above constraints result in a loss of germplasm?1 yes 1 only incidentally 1 noIf yes, what is the most important constraint, which may contribute to genetic erosion within thecollection? _____________________________________________________8. Further remarks----------------------Please send the completed questionnaire as an e-mail attachment to:Clair Hersheychh23@cornell.edu95

Cassava Conservation StrategyAppendix IV. Register of cassava and Manihot species survey respondentsCultivated cassava1 Bolivia Instituto de Investigacionesagrícolas “El Vallecito”Universidad Gabriel RenéMoreno1 Bolivia Instituto de Investigacionesagrícolas “El Vallecito”Universidad Gabriel RenéMorenoMa. Lizzie Cuellar Gutierrez(in vitro collection)Mateo Vargas Banco (fieldcollection)mlizzie@hotmail.commateovar@yahoo.com2 Brazil IAC Teresa Losada Valle teresalv@iac.sp.gov.br3 Brazil CNPMF (Embrapa Mandioca e Wania Maria Gonçalves Wfukuda@cnpmf.embrapa.brFruticultura Tropical)Fukuda4 Brazil Genetics Department – “Luiz deQueiroz College of Agriculture /University of São PauloElizabeth Ann Veaseyeaveasey@esalq.usp.br5 Chad Institut Tchadien de Recherche MBAILAO Kemdingao mbailaok@yahoo.frAgronomique pour leDeveloppement6 China Liu Guo-dao, Hainan Li Kai-mian likaimian@sohu.com7 CIAT CIAT, Cali, Colombia Daniel G. Debouck d.debouck@cgiar.org7 CIAT CIAT, Cali, Colombia Graciela Mafla g.mafla@cgiar.org8 Costa Rica Centro Agronómico Tropical deInvestigación y Enseñanza(CATIE)9 Côte d’Ivoire Centre National de RechercheAgronomique (CNRA)Carlos Alberto CorderoVargascordero@catie.ac.crN’ZUE Boni nboni1@yahoo.fr boni.nzue@cnra.ci10 D.R. Congo INERA Bidiaka Mpansu mbidiaka@hotmail.com mbidiaka@yahoo.com11 Ecuador INIAP. Estación Experimental Ing. Jacqueline Benitez jackyiniap@yahoo.comSanta Catalina. Departamento debiotecnología12 Ecuador INIAP, Estación Experimental Francisco Hinostroza iniapeeportoviejo@yahoo.comPortoviejoGarcía13 Ghana Crops Research Institute Joe Manu-Aduening jmaduening@yahoo.co.uk96

Cassava Conservation Strategy14 Guinee-ConakryInstitut de RechercheAgronomique de Guinée (IRAG)Bah El Sanoussyelsanoussy@yahoo.com15 Guyana National Agricultural Research Cleveland R. Paul crpaul6@hotmail.com crp6@cornell.eduInstitute16 IITA IITA, Ibadan, Nigeria Dominique Dumet d.dumet@cgiar.org17 Indonesia ILETRI (Indonesian Legume andTuber Crops Reseach Institute)Dr. Sholihin Sholhalim@yahoo.com Blitkabi@telkom.net18 Malawi Malawi Plant Genetic ResourcesCentre19 Malaysia Malaysian Agricultural Research& Development Institute (MARDI)20 Mozambique IIAM21 Niger Institut National de rechercheAgronomique du Niger (INRAN)22 Nigeria National Root Crops ResearchInstitute (NRCRI)23 Panama Instituto de InvestigaciónAgropecuaria de Panamá(IDIAP)24 Papua NewGuineaNational Agricultural ResearchInstitute (PNG NARI)Lawrent PungulaniTan Swee LianSeyni SirifiEKE-OKOROOKECHUKWUJosé Antonio Aguilar LópezRosa Kambuoulawrentp@.yahoo.co.uksltan@mardi.gov.mySsirifi2002@yahoo.frekeokorono@yahoo.comjaal0917@gmail.comrosa.kambuou@nari.org.pg25 Peru Henry Williams Vivanco Mackie Llermé Ríos Lobo llrios@inia.gob.pe rioslobo@hotmail.com26 Sierra Leone Institute of Agricultural Research Festus B. Massaquoi iarsl@sierratel.sl(IAR)27 South Africa ARC – Institute Industrial Crops T Vorster tomv@arc.agric.za28 Sudan Agric Res. Corp."ARC"/ South George Louis Tokporo georgetokp@yahoo.com tototadu@hotmail.comSudan AgricRes.tech. Org."SSARTO"Tadu29 Swaziland Malkerns Research Station Thembinkosi Gumedze Mrs@realnet.co.sz tgumedze@yahoo.co.uk29 Swaziland Malkerns Research Station Cinisani Tfwala cinisanitfwala@yahoo.co.uk30 Thailand Department of Agriculture (DOA) Prapit Wongtiem rayong1@doa.go.th ryfcrc@hotmail.com30 Thailand Department of Agriculture (DOA) Atchara Limsila rayong1@doa.go.th ryfcrc@hotmail.com31 Togo Institut Togolais de RechercheAgronomique (ITRA)Komi SOMANA somanaeric@yahoo.fr esomana@caramail.com32 Vanuatu CTRAV-VARTC Roger MALAPA malapa.roger@vanuatu.com.vu32 Vanuatu CTRAV-VARTC Vincent Lebot lebot@vanuatu.com.vu97

Cassava Conservation Strategy33 Viet Nam HungLoc Agricultural ResearchCenter34 Zambia Zambia Agriculture ResearchInstituteNguyen Phuongphuongdtg@yahoo.comMartin Chiona rtip@zamnet.zm mtas@zamnet.zmWild species1 Brazil Universidade de Brasilia Nagib Nassar nagnassa@rudah.com.br2 Brazil EMBRAPA/CNPMF Alfredo Alves aalves@cnpmf.embrapa.br3 CIAT CIAT, Cali, Colombia Daniel G. Debouck d.debouck@cgiar.org3 CIAT CIAT, Cali, Colombia Graciela Mafla g.mafla@cgiar.org98

A Global Conservation Strategy for Cassava and Wild Manihot ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?