Cracking the problem of tree seeds

The ICRAF genebank hosts the world’s largest holdings of agroforestry tree genetic resources but encounters several unique challenges in getting some of the seeds in its collection of 190 species to germinate.

Michael Major, Crop Trust

BANG! BANG! BANG! People walking past the genebank of the World Agroforestry Centre (ICRAF) may need to block their ears when the technicians are opening a baobab fruit. Agnes Were, a laboratory assistant, wields a hammer to open the hard, woody exterior and reveal the fibrous seeds inside. BANG! BANG! BANG! A hammer works for the baobab, but that is just one of many tools and techniques that ICRAF genebank staff have developed to bring tree species into ex situ conservation.

“As a genebank for trees, we encounter several unique challenges,” said Alice Muchugi, the Genebank Manager of ICRAF’s Genetic Resources Unit. “Some of the normally routine genebank operations like regeneration are a major undertaking since trees take years to mature. And maintaining the integrity of an accession is difficult, as tropical trees are predominantly outcrossing. Seed germination is also an issue. The seed of each species has its own characteristics and there is limited information on most of them. Therefore, we need to carry out research on how best to unlock the seeds of each species and get them to germinate.”

The ICRAF genebank was established in 1997 and now hosts the world’s largest holdings of agroforestry tree genetic resources. ICRAF has a medium-term storage facility in Nairobi which safeguards more than 5,300 seed accessions of 190 species. Another 12,000 accessions of 67 species are in 36 field genebanks located in 16 countries in Africa, Latin America and Asia. The accessions are primarily indigenous tree species useful for many different purposes as sources of medicinal extract, fruit, fodder, vegetables, specialist woods, and timber. The diverse tree species also help in improving soil nutrient content, retaining water and providing shade. By conserving and making available the best specimens of these tree species, ICRAF helps farmers and other landowners to plant more trees.


“Trees are totally different from annual food crops, so the general model presented by crop genebanks for seed management and ex situ conservation doesn’t always apply to us,” revealed Alice. The large number of genera, long regeneration intervals, large growth forms, generally outbreeding reproductive systems and species-specific seed storage and regeneration requirements of agroforestry trees require a more complex, and more flexible, approach.”

Many tree species, for example, have recalcitrant seeds. Their seeds lose viability when stored as dried seeds. These species need to be maintained as living plants in field genebanks.

“We need to promote decentralized and complementary conservation methods,” she noted. “This could include seed genebanks, live ex situ field genebanks, conservation circa situ, which means in farmland around natural forest, in situ, which is in the wild and on farm.”

Ex situ seed storage is the most common conservation method for orthodox seeds. However, if the species is popular and the seed is demanded by users, then eventually the supply has to be regenerated. For trees, this is expensive and difficult, demanding not only customized germination protocols for each species but a large amount of space and time to keep the trees until they mature and bear fruit.

“We have to conduct a fair bit of research to establish dormancy-breaking techniques for many species before we can get them to germinate,” Alice reiterates. “We also have to keep monitoring their viability in storage because no one has really kept tropical tree seed on this scale or for this long before, and we don’t know how many years all the different species will keep their viability in these storage conditions.”

Nature and local knowledge help solve the Allanblackia challenge

Allanblackia is a good example of how ICRAF had to learn about breaking dormancy. This species is a native African tree that has been valued for centuries for its oil. It’s often called the tallow tree. In 2000, Unilever discovered what local villagers had known for generations: the oil is a “food technologist’s dream” and is ideally suited for spreads like margarine. Unilever began working with ICRAF to domesticate Allanblackia, but soon realized that getting the seed to germinate was no easy task.

“We tried the conventional approaches, like nipping, soaking in water and treating with chemicals to break seed dormancy,” said Alice. “Those approaches were unsuccessful, so we started looking both to nature and local knowledge and applied a lot of trial and error in our search for a solution.”

Some communities in Tanzania shared their knowledge and experiences on seed germination with the ICRAF researchers. “They noted that seeds that had been buried by giant rats and forgotten would germinate. So, we decided to mimic the process.”

Following these ideas and taking the example of another lucky experience led to an effective method. “A farmer in Ghana told us that he had stored some seeds in a black polythene bag. When he opened the bag a few weeks later, he discovered that the seeds had germinated.”

“We replicated the black polythene bag method in the genebank and finally succeeded in breaking dormancy and starting germination after about four weeks,” Alice explained.

Establishing SOPs

Understanding the biology of the lesser known tree species has been a challenge for ICRAF’s genebank staff, but once they determine the best techniques to break dormancy the procedures are documented. “We have worked with the Genebank Platform to document our protocols as standard operating procedures, which we refer to as SOPs,” said Zakayo Kinyanjui, the GRU’s Seed Physiologist. “SOPs ensure that all protocols are conducted consistently now and in the future. With so many types of seed to process, we’d be lost without our SOP documents.”

The ICRAF team has developed SOPs for seed acquisition, conservation, characterization and distribution with assistance from Janny van Beem, the quality management specialist working with the Crop Trust. “An audit system is currently in place to ensure that all SOPs adhere to relevant standards, policies and best practices,” said Janny. “ICRAF’s genebank is quite distinct from other genebanks so it is critical to get these procedures documented and audited.”

Kick-starting germination

Written procedures are critical when you have such a diversity of material. Eucalyptus species have seeds of less than one millimeter while the Allanblackia fruit can weigh an average of 5 kg. With most species, genebank staff can scarify the seed by puncturing the seed coat with a razor blade, scalpel or emery paper without damaging the embryo. However, some tree species have very hard seed cases that are impossible to scarify using normal abrasion techniques. So, over the years, staff at ICRAF have developed numerous distinct protocols for the varied species in their collection.

The ICRAF genebank has more tools than a medieval torture chamber, with hammers, saws, vices, garden secateurs, fingernail clippers and more (see photo gallery). For some species, only a few seeds are collected, so any seed used in germination tests are grown out and taken to the field nursery for research or distribution.

Multiplying for distribution

All this expertise in conservation techniques is fantastic for keeping the seeds healthy in storage for the long term. However, the demand for tree seed from ICRAF is increasing. In 2017 alone, ICRAF sent germplasm samples to 2,959 users around the world not including ICRAF’s own users. About 87% of those samples went directly to farmers. “Other genebank materials may first pass through a breeding pipeline before they turn up as improved varieties in farmers’ fields, but most trees are not yet domesticated so our users tend to be the farmers themselves,” said Alice.

ICRAF’s germplasm recipients can be divided into two main groups. Farmers may request the seeds for their own use. They raise the seeds in their nurseries and plant the trees on their farms. Alternatively, community-based organizations request material and then distribute to farmers. “These are mainly smallholder farmers who in the end may plant 5-10 trees of a species on their farms,” said Alice.

“If funds are available, we may monitor the performance of the distributed material in the farmers’ fields,” said Alice. Tree germplasm research projects tend to be very participatory. Trees require a lot of space and take years to mature. Most research centers do not have that much land available so most of the research is conducted on farmers’ fields.

Safeguarding tree diversity forever

Trees play a vital role in many agricultural systems and are key to climate change mitigation. By safeguarding their diversity, we improve our food security and our ability to adapt to our changing climate. ICRAF’s genebank has been cracking the problem of tree seed conservation for more than 25 years to ensure that famers continue to have access to tree diversity. By distributing thousands of tree seeds and cuttings every year, the ICRAF genebank is paving the way to a future filled with trees.