Should ICRAF wish to acquire further material for these taxa, and assuming that material is available and comes with good passport data, criteria that could be employed in a gap analysis to identify priority wild genotypes and cultivars for acquisition include the following: material of specific geographical or environmental origin (e.g. genotypes adapted to particular regions, countries or localities; cultivars optimal for specific environments); material with useful traits of importance in breeding programmes (e.g. important existing cultivars; material with desirable characteristics for future breeding programmes and/or research); and material with genetic variability (i.e. genotypes/cultivars with alleles not already in the collection).
6. ICRAF GRU collections policy
In developing an acquisition, retention and discard collections policy for ICRAF’s GRU, we are making the following assumptions based on ICRAF’s Agroforestry Tree Genetic Resources Strategy (see Section 2.3):
That ICRAF has a global agroforestry tree mandate, which includes potential users in all sectors (public, private, civil) in developing countries throughout the world;
That, although Article 15 of the International Treaty confines itself to food, forage and fibre species, ICRAF is not confined to working solely under the Treaty and therefore no group of useful tree species (e.g. timber species) is excluded from the mandate;
That ICRAF’s GRU has both a conservation and a service provision role; and
That ICRAF wishes to work in partnership with other genebanks in the CGIAR and beyond to improve its conservation and service provision functions.
6.1. Acquisition policy
There are a number of ways in which ICRAF may acquire material. These include through:
Proactive acquisition by field collection, purchase or exchange;
GRU participation in ICRAF research and development projects
In all cases, ICRAF’s GRU will have to decide whether the material is of high, medium or low priority. Criteria are described below for each category and are summarised in Table 6.1.
Table 6.1. Criteria defining ICRAF GRU priority species and collections
High priority collections
Medium priority collections
Low priority collections
Widely used in global agroforestry but not currently represented in the GRU collections
User’s priority species and cultivars
Species/cultivar in high demand
Threatened agroforestry species/genotype/cultivar
Unique or rare genotype/cultivar with specific traits not currently represented in the GRU
A collection important for the delivery of a project but which does not fulfil any of the high priority criteria
Other collections verified as being useful but not meeting the high priority criteria
All collections that do not meet high or medium priority criteria
Material should be a high priority and proactively acquired by the ICRAF GRU if:
It is a ‘gap’ species, genotype or cultivar that is widely used in agroforestry but is not currently represented in the GRU collections (SOWFGR Top 500 – Section 5.1);
It is a species, genotype or cultivar in high demand (ICRAF’s records of requests – Section 5.2);
It is a useful agroforestry species or genotype known to be threatened in the wild or to be in decline (Section 3.4); and
It represents a unique or rare genotype or cultivar not currently represented in the collections or with specific desirable traits, e.g. adaptive potential (Section 5.3)
Medium priority material that might be acquired under certain circumstances, e.g. by donation or through a specific project, will include:
A species, genotype or cultivar important for the delivery of a project but which does not fulfil any of the criteria above;
Although it is a user’s priority species, genotype or cultivar, it is not frequently requested (Section 4.2 & Annex 2); and
A species that SEPASAL or other economic botany databases list as being useful (Section 3.2)
Collections from all other species, genotypes or cultivars should be regarded as low priority and should not be acquired or retained.
6.2. Managing the collections cost-effectively, including decisions about retaining or discarding collections
In order for high and medium priority collections to be managed efficiently and cost-effectively, management decisions need to be based on:
The most cost-effective conservation methodology (seed bank versus field genebank)
The most cost-effective multiplication methodology (re-collection or regeneration).
A management decision tree, taking these criteria into account, is presented in Figure 6.1. These criteria are discussed below.
6.2.1. Criteria for determining cost-effective storage and multiplication methodologies
Seed storage behaviour. Seed conservation is by far the most cost-effective methodology for storing germplasm long term. Desiccation tolerant (orthodox) seeds can be dried and stored at sub-zero temperatures for many years without significant loss of viability. However, desiccation sensitive (recalcitrant) seeds cannot be stored under conventional seed bank conditions of low temperature and humidity without losing viability, sometimes quickly. The most comprehensive source of information on seed behaviour is Kew’s Seed Information Database (SID, http://data.kew.org/sid/sidsearch.html). Other sources of information include ICRAF’s own genebank records and species information sheets such as those produced by Plant Resources of Tropical Africa (http://www.prota.org/). Various relevant databases are given in ICRAF’s Agroforestry Species Switchboard.
Expensive/difficult to re-collect. Regeneration of material may be a more cost effective option than re-collecting seed for species, genotypes or cultivars that are difficult or expensive to sample. These may include taxa with small, scattered populations (including rare and threatened species), that occur only in inaccessible or dangerous places; and that rarely set seed.
Taxa with long regeneration times. Even where taxa are expensive to collect seed from, if regeneration times are long, then it will usually be more cost effective to re-collect seed (that is if it still available in natural ecology). Taxa with long regeneration times will include any species that take longer than five years to set seed. This will be the case for many perennials and most trees and shrubs.
True to type. Open pollinated plants will result in random allele selection, meaning that traits based on genic combinations may not be passed from parent to seed. If material exhibiting specific, selected traits is required then field genebanks may be the most cost-effective option for long term storage and the supply of material to users. Under controlled, field genebank conditions, cuttings can be taken or controlled pollination can be carried out on plants of known provenance. Alternatively, cuttings or controlled pollination can be carried out with species in the wild but this is difficult for inaccessible or rare species.
Figure 6.1. ICRAF Germplasm management decision tree
6.2.2. Retaining or discarding material
As indicated in Figure 6.1, orthodox seeds from low priority collections (see Table 6.1) should be discarded or archived (i.e. not actively managed but not actually discarded). In the case of recalcitrant-seeded species, only high priority collections should be retained because of the additional expense associated with maintaining such collections in field genebanks. Given the expense associated with establishing field genebanks, an option for existing, low priority genebanks is to reduce their maintenance costs by keeping them secure but no longer actively managing them. Table 6.2 summarises criteria that should be applied to retaining, discarding or archiving material. Other factors that should be taken into account in deciding whether to retain, archive or discard material include:
Material quality. Seed collections of low viability should be discarded or archived. International standards vary between 75% and 85% as a ‘pass rate’ for viability. However, ICRAF may want to lower the bar for certain high priority collections. Care should be taken to ensure that viability is genuinely low before seeds are discarded. Sub-optimal germination protocols may lead to false negative results, and complementary Tetrazolium or cut tests are recommended before material is discarded. When in any doubt, archiving material by continuing to store it but ceasing to actively manage it (e.g. no longer carrying out viability testing) is an option. In the case of field genebanks, trees in poor condition will have limited utility.
Diseased or infested material. Material which is not clean (i.e. carries disease or is infested with a pest) should not only be discarded, but incinerated as a biohazard. ICRAF already has procedures in place to detect such material.
Invasive alien or other proscribed species. ICRAF already has measures in place to ensure that prohibited alien invasive taxa are never distributed. Such collections should be discarded from ICRAF’s collections unless they are supporting research into control of such taxa. Taxa that are not prohibited but that have the potential to be invasive are already flagged by ICRAF with cautionary warnings on the seed packet and with further information available through the Agroforestry Species Switchboard.
Data quality. Material with limited passport data is of limited utility. In some cases, it is possible to characterise material retrospectively. However, if vital information such as date, locality of collection or sampling methodology is lost, then material has reduced utility.
ABS compliant material. The status of ICRAF’s material under Article 15 of the International Treaty is discussed in greater detail in Section 6.3.3. However, material acquired by ICRAF in the future may be subject to further national laws or regulations under the Nagoya Protocol of the Convention on Biological Diversity. It is not uncommon, for example, for material to be supplied only for the duration of a research project after which it will need to be repatriated or destroyed.
Table 6.2. Criteria relevant to retaining/discarding material
Non-ABS compliant material (or as stipulated in a bilateral agreement)
6.3. Other considerations
6.3.1. Infrastructures and facilities
As already discussed in Section 2.2.2, 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 overseas (Svalbard, Kunming and the Millennium Seed Bank). Given the decline in global seed conservation capacity, ICRAF’s GRU could take the decision to invest in long term storage capacity thereby fulfilling an important global role in conserving agroforestry tree germplasm. If ICRAF’s leadership decides that the organisation should (a) significantly broaden and strengthen ICRAF’s collections, and (b) play a more active role in long term seed conservation for agroforestry tree species, genotypes and cultivars, then it will be cost effective to build larger storage facilities, including long term seed storage capacity at ICRAF’s HQ in Nairobi. Given the lack of long term conservation capacity in most NARS, this is an option that should be given serious consideration.
Regardless of whether the ICRAF GRU decides to strengthen its collections and/or play a greater role in the long term conservation of agroforestry tree germplasm, we recommend that ICRAF strengthens its partnerships with other genebanks to improve both the range of collections available to users and the long term conservation of agroforestry tree species.
Using the Gene Bank CRP and existing tools such as GENESYS and GRIN Global, it should be possible to rationalise and share the task of supplying material to users. For example, if ICRAF, ILRI, CIAT and large NARS such as USDA’s Fort Collins were to combine their seed lists so that the user’s request was automatically directed to the genebank with the relevant accession, this would help to avoid duplication of effort and a prolonged process trying to find the right seeds for the user.
Furthermore, some kind of rationalisation between the relevant CGIAR genebanks would make economic and logistic sense. For example, if CIAT were responsible for agroforestry collections from tropical America, ICRISAT from Asia, ICRAF from tropical Africa and ILRI from drylands Africa, non-essential duplication could be avoided and the costs of acquiring material reduced.
Finally, partnership makes sense where safety duplication is required. All high and medium priority collections should be held to international standards in at least two seed banks. The current arrangements for safety duplication that ICRAF has with Svalbard, the Kunming Institute of Botany and the Millennium Seed Bank may well be adequate. However, if they are not, then existing partnerships should be strengthened or new partnerships developed.
6.3.3. ABS compliance
ICRAF’s Tree Genetic Resources Policy (2014) states (Section 2.2) that “ICRAF considers its tree germplasm collections as global public goods and will manage them in ways that maximize their global accessibility and/or ensure that they lead to the broadest possible impact on target beneficiaries in furtherance of the ICRAF and CGIAR vision.’
It goes on to say (section 2.3) that “ICRAF reaffirms its continuing commitment to the principles of its 26th October 1994 agreement with FAO, which placed CGIAR germplasm ex situ collections under the auspices of FAO. ICRAF holds designated ex situ germplasm collections of agroforestry tree species “in trust” for the benefit of the international community, in particular developing countries, under Article 15 of the ITPGRFA, in accordance with ICRAF´s agreement with the Governing Body of the ITPGRFA, signed on 16th October 2006.” Under this agreement, all food, forage and fibre material and associated data acquired by ICRAF prior to October 2006 is subject to Article 15 of the ITPGRFA and can be supplied under the Treaty’s SMTA without the need to renegotiate access with the country of origin.
For material not covered by Article 15 and for material acquired after 2006, however, ICRAF is subject to the relevant national laws and regulations that govern access to material under the Nagoya Protocol of the Convention on Biological Diversity. ICRAF’s Tree Genetic Resources Policy (Section 2.5) states: ”ICRAF recognizes the sovereign rights of states over their natural resources and that the authority to determine access to genetic resources rests with the national governments and is subject to national legislation. It also recognizes that access to such resources is subject to the prior informed consent of the country of origin and to the fair and equitable sharing of benefits deriving from their use, in accordance with the CBD and, as applicable, its Nagoya Protocol, without prejudice to the application of the ITPGRFA. All agroforestry tree germplasm accessed after the coming into force of CBD (on 29th December 1993) and not covered by the use of the SMTA shall be provided according to national ABS laws and in accordance with CBD and, as applicable, its Nagoya Protocol.”
Although ICRAF is explicit about its adherence to the principles and practice of the Nagoya Protocol, it is not clear to what extent this approach has been tested, i.e. how much material during recent years has been acquired under bilateral agreements with sovereign governments and the conditions attached to the use of such material, particularly with regard to supply of that material to third parties. It is possible that the bureaucratic burden associated with negotiating bilateral agreements under Nagoya and the conditions under which such material may be supplied to third parties will severely restrict ICRAF’s ability to grow its collections and its impact.
ICRAF HQ Seed Bank
FAO (2014). State of the World’s Forest Genetic Resources. Commission on Genetic Resources for Food and Agriculture, Food and Agriculture Organisation of the United Nations. Rome, 2014.
Franzel, S. and Kindt, R. (2012). Species priority setting procedures. In: Dawson, I., Harwood, C., Jamnadass, R., Beniest, J. eds. Agroforestry tree domestication: a primer. Nairobi. World Agroforestry Centre ICRAF, Kenya p. 36-45.
ICRAF (2012). ICRAF Agroforestry Tree Genetic Resources Strategy 2013-2017. http://www.worldagroforestry.org/products/grunew/downloads/ICRAFGeneticResourcesStrategy2013.pdf
ICRAF (2014). Tree Genetic Resources Policy. http://www.worldagroforestry.org/sites/default/files/users/admin/Genetic%20Resources%20Policy.pdf
Smith, P. and Thomson, L. (2015). External Review of the ICRAF Genetic Resources Unit. Commissioned by the Global Crop Diversity Trust, Bonn.
Van Noordwijk, M. (2014). Agroforestry as plant production system in a multifunctional landscape. Inaugural lecture upon taking up the special professorship in Agroforestry at Wageningen University 16 October 2014. Wageningen, Netherlands. Wageningen University. 24 p.
Van Noordwijk, M., Coe, R., Sinclair, F. (2016). Central hypotheses for the third agroforestry
paradigm within a common definition. World Agroforestry Centre (ICRAF) Working paper 233, Bogor