MSB Millennium Seed Bank
NARS National Agricultural Research Systems
NGO Non-Governmental Organisation
NTSC National Tree Seed Centre
PGRFA Plant Genetic Resources in Food and Agriculture
SEPASAL Survey of Economic Plants in Arid and Semi-Arid Lands
SID Seed Information Database
SMTA Standard Material Transfer Agreement
SOWFGR State of the World’s Forest Genetic Resources
USDA United States Department of Agriculture
UTP Useful Tropical Plants
1.1. The World Agroforestry Centre’s Genetic Resources Unit
Established in 1993, the World Agroforestry Centre’s (ICRAF’s) Genetic Resources Unit (GRU) aims to collect, conserve, document, characterise and distribute a diverse collection of agroforestry trees, with a strong focus on indigenous species. The GRU comprises a medium-term cold storage facility located at ICRAF’s headquarters in Nairobi (ex situ genebank), short-term seed storage facilities in Bamako (Mali) and Lilongwe (Malawi), and 37 agroforestry field genebanks in 15 countries, mostly in tropical Africa, but with a few in addition in South America (Peru) and Asia (Vietnam and Bangladesh). In 2006, ICRAF signed an agreement (Article 15) with the International Treaty on Plant Genetic Resources for Food and Agriculture (ITPGRFA) Secretariat as one of a group of International Agricultural Research Centres (IARC). ICRAF’s collections falls under Article 15.1b for non-Annex 1 crops held by the IARC, which includes germplasm of multi-purpose uses such as for fruits, forages and fibre. As part of this agreement, ICRAF has agreed to manage and administer the germplasm held by its GRU in accordance with internationally accepted standards, in particular the Genebank Standards endorsed by the Commission on Genetic Resources for Food and Agriculture. ICRAF also agreed to avail germplasm in accordance with the provisions of the Material Transfer Agreement (later the Standard Material Transfer Agreement, the SMTA). Since 2012, the ICRAF GRU has received funding under the CGIAR Research Programme (CRP) for Managing and Sustaining Crop Collections (Genebank) to support the long-term conservation and sustainable utilisation of agroforestry tree germplasm. The fund is managed by the Global Crop Diversity Trust.
1.2. Why a GRU Collections Policy is needed
In June 2015 the Global Crop Diversity Trust commissioned an External Review of ICRAF’s GRU (Smith & Thomson, 2015) to:
Assess the operations and activities of the ICRAF genebanks;
Assess the roles, services and use of the genebanks, and the linkages with users and partners both within and outside the CGIAR;
Consider the status of the genebanks and/or individual collections, in the context of a global system for the long-term conservation and use of the species in question;
Review the status of the genebanks with respect to performance targets and the feasibility of proposed work plans to reach targets; and
Provide actionable recommendations related to all of the above.
The Review noted that:
“The GRU’s existing collections are globally significant, and their value will be increased once they are fully characterized, and their passport data is complete. However, in order for ICRAF’s GRU to fulfil its true potential, it needs to develop a collections acquisition and retention policy based on ICRAF’s global mandate, identified user needs and cost-effectiveness. ICRAF’s collections represent only a fraction of the several thousand tree species that deliver a whole host of products and services in agroforestry systems to smallholder farmers in developing countries. The recent State of the World’s Forest Genetic Resources Review (2014) concludes that around 8,000 tree species are used in forestry around the world. Only a proportion of these will be used in developing countries, and still fewer in agroforestry. However, the Review recommends that ICRAF considers adopting a more pro-active approach to acquire, as a minimum, genetically representative collections of the most important species globally in agroforestry systems. This would also include tree species which can ensure resilience in agroforestry systems in the face of predicted climate change and extremes.”
This document is written in response to this recommendation. It sets out a collections acquisition and retention strategy based on ICRAF’s global agroforestry mandate, user needs, gaps in collections and cost effectiveness.
2. ICRAF’s global agroforestry tree mandate
2.1. Defining agroforestry tree species
Noordwijk, Coe & Sinclair (2016) define agroforestry as “…the practice and science of the interface and interactions between agriculture and forestry, involving farmers, trees, livestock and forests at multiple scales.” Their definition incorporates three complementary agroforestry paradigms:
A set of specific practices that combine trees, crops and/or livestock with the aim of supporting positive interactions among these components;
The landscape level interface of trees and farms, farmers and forests, tree domestication, forest use rights, ecosystem services and markets for tree products; and
Agroforestry within the policy context of sustainable development goals, globalising markets and global climate change.
While this definition of agroforestry (outlined further in Figure 2.1) is helpful for understanding the systems that constitute agroforestry, it is less useful for deciding which tree species constitute ‘agroforestry species’ and which don’t, because it embraces multiple systems, formal and informal, for a multitude of purposes from revenue generation to ecosystem services, and at a wide range of scales.
Common denominators of agroforestry systems are that they constitute a specific response to the special needs of tropical developing countries, in particular, to low capital availability and to farming in degraded or marginal lands. From this context, we can start to set criteria that characterise ‘agroforestry tree species’. Normally they will be low input and well adapted to poor or marginal lands. They also will be genetically diverse and resilient. Ideally, they should be able to be integrated into rotation/mixed production systems. In addition, agroforestry species should have ecological as well as economic benefits. From this we might conclude that a typical agroforestry tree or shrub requires low inputs and delivers economic benefits (e.g. food, forage or fuel) and/or ecological benefits (e.g. shade, habitat or soil fertilisation). By the same reasoning, trees that do not qualify as agroforestry species are likely to be high input, highly bred commodity crop or forestry species that are planted at large scale in monocultures.
Figure 2.1. Three agroforestry concepts mapped on to the agriculture-forestry interface (from van Noordwijk, 2014)
2.2. The role of ICRAF’s GRU
2.2.1. ICRAF GRU’s current role
A recent survey of ICRAF genebank users (see Section 3.3 and Annex 1) indicated that:
ICRAF primarily supplies seed to African countries, particularly to Kenya;
50% of the users who responded to the survey come personally to ICRAF’s GRU to obtain seed, as many are recipients of seed for ICRAF projects;
Users are more likely to obtain seed from ICRAF than from government seed centres or commercial seed suppliers;
Seeds are primarily for on farm use, research and restoration/afforestation;
A wide range of species are used for a wide range of purposes;
Preferred species that are difficult to obtain are primarily multipurpose (37%), food (27%) and medicinal (16%) trees; and
The most problematic issues identified by users concerning germplasm acquisition are limited information on suitable species for a given region, lack of visibility of the collections and lack of characterisation and evaluation data.
2.2.2. ICRAF GRU’s potential role
In response to the question “What role do you think the ICRAF genebank should play in the supply of agroforestry seeds and information?”, the respondents in ICRAF’s user survey (also see Section 4.1) indicated that the primary roles of the GRU should be the supply of agroforestry germplasm, the dissemination of agroforestry knowledge, and user training and capacity building. Perhaps because of the way the question was phrased, only 2% of respondents saw the GRU’s primary role as research, only 2% cited conservation of germplasm as an activity that ICRAF should engage in, and only 6% identified monitoring and evaluation as ICRAF’s primary role (Total number of respondents =57).
The views of users on the importance of seed supply and capacity building are indicative of the lack of investment in recent years in National Tree Seed Centres (NTSCs) and National Agricultural Research Systems (NARS), reflected in crumbling infrastructures, fewer staff and the general lack of availability of high quality tree seed and associated information on appropriate use and management. There is also some evidence to suggest that the exchange of plant genetic resources between countries has become more difficult since the advent of the Convention on Biological Diversity and, specifically, since national laws and regulations related to the Nagoya Protocol have come in to force. The trend of disinvestment in NARS and NTSCs has seen a drastic decline in extension services advising and training farmers in the use of agroforestry trees. There has been reduced financial and technical support to NTSCs by donors, and the gap in funding has frequently not been filled by host governments.
How ICRAF can help address gaps in the dissemination of knowledge and in the development of capacity around tree germplasm is part of its research agenda (especially of the Tree Diversity, Domestication and Delivery science domain). Further discussion on this point is beyond the scope of this report, but that ICRAF’s GRU has an important role to play in the supply of agroforestry tree germplasm is beyond doubt. This is discussed in more detail below.
Important steps and/or decisions that ICRAF needs to make with regard to its future role in conserving and supplying agroforestry tree seed include:
Broadening the geographical scope of ICRAF genebanks. Currently, the ICRAF GRU supplies the vast majority (>90%) of its seed to Africa. As the World Agroforestry Centre, it has a much broader mandate than this.
Better defining which agroforestry species, genotypes and cultivars fall within the ICRAF GRU’s purview, e.g. tropical versus temperate taxa relevant to ICRAF’s geographical focus above. Similarly, which uses are appropriate? Article 15 only encompasses food, forage and fibre species, but ICRAF currently works with a much wider range of utilitarian species including timber and medicinal trees.
Considering a greater focus on long term seed conservation. ICRAF’s GRU is currently set up primarily for short to medium term seed storage, with limited back up of its collections in long term storage facilities elsewhere (Svalbard, Kunming and the Millennium Seed Bank). Given the decline in global seed conservation capacity, ICRAF’s GRU could take the decision to invest more in long term storage capacity to better fill an important global niche in conserving agroforestry tree germplasm.
Considering a wider range of storage approaches. For recalcitrant species, medium to long term storage or maintenance of vegetative material in cryopreservation in addition to field genebanks may be more important in the future.
Investing further in strategic partnerships. As part of the Global System for PGRFA, ICRAF has excellent opportunities to create strategic alliances with NARS and other CGIAR centres engaged in the conservation and supply of agroforestry tree germplasm. There is already some collection overlap with ILRI and CIAT, and here rationalisation of collections and roles would be invaluable. Similarly, ICRAF already partners with some NARS in co-managing field genebanks, but this could be expanded to include seed banks (e.g. Fort Collins, CSIRO Australia, EMBRAPA Brazil) that hold significant agroforestry tree collections, taking into account the regional importance of taxa and their ecological limitations.
Improving access to resources. ICRAF is in a position to facilitate access to a wide range of International Treaty species under the SMTA. For species not covered under the Treaty, the CGIAR centres, including ICRAF, have a potential role to play in facilitating access to third parties through bespoke bilateral or multilateral agreements.
These and other issues related to the acquisition, retention and distribution of agroforestry tree germplasm are discussed in more detail in Section 6 below.
2.3. ICRAF’s current Genetic Resources Strategy
ICRAF’s Agroforestry Tree Genetic Resources Strategy 2013-2017 sets out its goal for germplasm collection, acquisition and rationalisation. We recount sections of this strategy of relevance to the current document below, before considering issues further in following sections:
“The ICRAF collection of priority agroforestry tree species will be expanded to fill important gaps in the coverage of the genepool and rationalised as needed after review of diversity based on the results of characterisation and eco-geographic mapping.”
The strategy goes on to state:
“Although species priority setting has been an important activity for deciding which trees to promote and domesticate within ICRAF regional programmes (Franzel & Kindt, 2012), no specific priority setting exercise has been carried out to determine targets for conservation. Priority setting for conservation overlaps, but in important elements is different from that for domestication. As well as market considerations, it takes into account the level of diversity and degree of threat to the resource and existing conservation initiatives, as well as the potential for current and future use of the tree. A clear process of priority setting for conservation will be carried out by region over the next five years to determine which species to focus on. Priority setting is an ongoing dynamic process to meet current and future needs, and requires a broad perspective to respond to global challenges such as climate change. Ongoing consultations with partners and farmers will be needed to ensure that the collections remain relevant to their needs. Criteria used to determine key species for conservation during this process will include:
Market demand and importance to livelihoods (health, nutrition, income) of farmers
Domestication suitability for cultivation, production and processing
Level of coverage by other conservation efforts, including the long-term viability of in situ methods, and genebanks
Availability of potential partners for conservation
Gaps in collections
Level of threat, endangered nature of the species and its role in the environment for supporting the sustainable use of other flora and fauna
Presence of useful adaptive traits such as drought tolerance, salt tolerance and resistance to pests and diseases.
The amount of genetic diversity within the key agroforestry species and access to this diversity within the access and benefit sharing conditions of international treaties, [in addition to the] distribution [of] and knowledge about the species and funding will determine the size of species-specific collections. Representative samples will be collected across the distribution range to capture a wide range of diversity.”
The strategy also makes a commitment to address information gaps for agroforestry tree species and to facilitate access to materials from genebank collections. Regarding access to material, the strategy states:
“In line with the Consultative Group on International Agricultural Research (CGIAR) Intellectual Assets (IA) principles, ICRAF considers its germplasm collections as global public goods and is committed to the widespread diffusion and use of material and information to achieve the maximum possible access, scale, scope of impact and sharing of benefits to advantage the poor, especially farmers in developing countries.”
Regarding germplasm delivery systems, it says:
“Germplasm delivery systems for agroforestry trees generally do not follow a formal breeding and delivery approach. Lack of access to high quality planting material of a wide range of tree species to small-scale farmers constrains their adoption of agroforestry practices. Improving the ways by which growers gain access to planting material is therefore crucial in bringing trees into cultivation on farms, and thus is an essential consideration in tree domestication strategies. Considering the variation in species of interest between different countries and even different regions of the same country, local field genebanks and seed orchards will be used in addition to the conventional seed genebank to meet needs. In order to make the germplasm available to a diverse set of users, germplasm will be made available on a non-profit basis to cover the costs of production.”
Finally, the strategy makes a commitment to strategic partnerships, saying:
“Partnerships involving the sharing of responsibilities for conservation and management are central to this strategy. ICRAF will continue to work closely with existing national and international partners in the regions and will seek and initiate new partnerships for conservation as needed, based on the ICRAF partnerships strategy and guidelines. The modalities for involvement of different partners will vary depending on local conditions and needs. However, all partnerships will be governed by contracts that identify responsibilities and cost sharing arrangements. Service agreements will be concluded with service providers or partners taking on specific responsibility for conservation or testing.”
3. Prioritising collections based on ICRAF’s global mandate
As indicated in Section 2.3, ICRAF’s Agroforestry Tree Genetic Resources Strategy considers global importance, market demand/importance to livelihoods, level of threat and role in the environment as criteria that should be used to determine key tree species for conservation. In this regard, there are a number of sources of information on globally important agroforestry tree species, genotypes and cultivars that ICRAF may wish to use to prioritise which collections it should acquire and/or retain. These include the first State of the World’s Forest Genetic Resources Review and various useful and threatened species databases. We consider some of these below (further information can also be found at ICRAF’s online Agroforestry Species Switchboard, http://www.worldagroforestry.org/output/agroforestry-species-switchboard-13).
3.1. State of the World’s Forest Genetic Resources (SOWFGR) Review, 2014
The Food and Agriculture Organization of the United Nations (FAO) gathered country reports from 86 nations on the current status of forest genetic resource conservation and use to synthesise the recent first State of the World’s Forest Genetic Resources (SOWFGR) Review (FAO, 2014). Country reports mentioned 8,000 species of trees, palms and bamboos, and economic and conservation value were given as the two main reasons for nominating priorities for conservation and management. Of the 8,000 species, around 2,400 were described as being actively managed, i.e. managed specifically for their products and services (although not necessarily cultivated). The remainder were indicated to be harvested from the wild. The main products and functions targeted through management activities were reported by countries as timber (42%), non-wood forest products (41%) and energy (mainly fuelwood, 19%).
While the species listed in the country reports of high income nations are generally of limited interest, those cultivated by the more than 50 low and middle income countries who participated in the SOWFGR give a good indication of species that are of importance locally and globally in a development context. These lists also give some indication of user demand and importance to livelihoods of species, at least in qualitative terms.
We have obtained a database from FAO listing the species cited by the SOWFGR country reports from 58 low and middle income countries, and use these data in the gap analysis in Section 5 below.
3.2. Survey of Economic Plants in Arid and Semi-Arid Lands
The Survey of Economic Plants in Arid and Semi-Arid Lands (SEPASAL) database is the world's most comprehensive online source of information on useful 'wild' and semi-domesticated tropical and subtropical dryland plants, with a focus on Africa. It includes plants which humans eat, use as medicine, feed to animals, make things from, use as fuel, and many other uses. The SEPASAL database focuses on drier parts of the world because these are home to one sixth of the world's population, in some of the poorest countries. The database contains information on approximately 7,000 species. Registration is required to access the database, but entry is free. SEPASAL can be searched by scientific name, distribution and use. Relevant lists of useful species that can be generated include: food plants, food additives, animal food, bee plants, materials, fuelwoods, social uses, medicines and environmental uses (e.g. erosion control, shade, soil improvers). Lists can be filtered with useful criteria such as habit, conservation status, domestication status, and adaptive traits such as disease and pest resistance.
3.3. Useful Tropical Plants database
A useful counterpart to SEPASAL is the Useful Tropical Plants (UTP) database (http://tropical.theferns.info) that, unlike SEPASAL, includes species from the wet tropics. The UTP database currently includes records of over 11,000 useful plant species from the tropics. The database can be searched by use (e.g. medicine, food, green manure), and lists generated accordingly. Advanced search options enable the user to search on adaptive traits such as drought tolerance, nitrogen fixation and tolerance of poor soils, among others.
3.4. The IUCN Red List of Threatened Species
An important criterion in the ICRAF GRU strategy is level of threat. The IUCN Global Red List (http://www.iucnredlist.org) can be searched by taxonomy, location, ecosystem, habitat, threat, assessment and plant growth forms (e.g. shrub or tree). The Red List is not currently comprehensive for plants as it does not include national and regional assessments. However, it is the best source of information available for globally threatened plant species and is therefore relevant to ICRAF as a data source.
Botanic Garden Conservation International (www.bgci.org) will launch ‘ThreatSearch’, the most comprehensive list of threatened plant species, incorporating global, regional and national threat assessments, later this year. This list includes non-IUCN threat assessments such as the Australian and US assessments. This database should therefore also be a useful resource.