Session 2.1. Methods for identification, germplasm collection and selection

Training Workshop on Allanblackia Domestication, 23 to 27 Oct 2006

Ian Dawsonc/o The World Agroforestry Centre, Nairobi, Kenya

Learning objectives

Participants will be able to:
  • Relate the importance of germplasm identification, collection and selection to the Allanblackia agri-business.
  • Explain the key issues that need to be considered during identification, collection and selection.
  • Relate the actual approaches used for Allanblackia collection and selection to date.
  • Explain ‘best practice’ for practical deployment and key constraints/gaps in understanding that require further attention.

Instructional methods

45 min classroom presentation (PowerPoint) with open discussion.
Instructional materials
‘Lecture’ note, copy of PowerPoint presentation and references.

Summary

The Allanblackia agri-business requires than several million trees be planted across countries participating in the initiative each year from 2008 onwards. Germplasm of the right species must be provided in a suitable form, in sufficient quantities, and in a timely manner for farmer planting. At the same time, supplied planting material should be capable of consistently producing good quantities of high quality food oil.
Identification of the right Allanblackia species for planting is important in order to properly address quality and performance issues, and may be important to prevent outbreeding depression and properly facilitate circa situ conservation. Identification of different Allanblackia species is relatively straightforward, with few problems anticipated, even in cases where species distributions are sympatric (overlap) in Cameroon and Tanzania.

Germplasm collection of Allanblackia has been undertaken for a number of reasons, including for raising planting material for distribution to farmers, for tree management research, for genetic improvement programmes and for ex situ conservation. Collections have used a number of approaches, with material sampled as seed and vegetatively, and using random and targeted methods. A brief description of the theory behind the different approaches that have been applied, each of which has potential advantages and disadvantages, is given here. Vegetative sampling carries a number of advantages compared to seed collection in relation to selection, gender determination, and the physiology of collected material (earlier fruiting?), but requires more resources, both during and after collection, to undertake properly. Whilst targeted collection appears an attractive approach, it also takes more resources than random sampling to do properly, and re-collection (on return to a site) can be more difficult. Following this analysis a description of key issues important to all collections, irrespective of the approach used, is given.

Selection of Allanblackia for seed yield is underway during some collections, based on observations on fruit size during surveys of natural stands. General issues related to selection are here considered, including a consideration of the value of applying selection during seed and vegetative sampling. Whilst selection from natural stands appears an attractive proposition, there are reasons why it may not be as effective as collectors may assume, especially if germplasm is sampled as seed.
Summaries of current best practices for collection and selection, and key gaps in understanding that in the future need to be addressed, are given. For collection, determination of the efficiency of different approaches to sampling is a key issue. For selection, wider consideration of the suitable characteristics by which ‘superior’ material is defined (e.g. stability of production) is very important.

Key references

  • Dawson I, Were J (1997) Collecting germplasm from trees. Agroforestry Today,9 (2), 6-9.

Lecture note

The Allanblackia agri-business requires than several million trees be planted across countries participating in the initiative each year from 2008 onwards, for a number of years. Germplasm of the correct species must be provided in a suitable form, in sufficient quantities, and in a timely manner for farmer planting. At the same time, supplied planting material should be capable of consistently producing good quantities of high quality food oil.

In order to undertake this process, proper identification, collection and selection of Allanblackia germplasm is required. In the below sections, we consider each of these in turn, addressing aspects of theory and actual practice on Allanblackia. Key lessons for practical deployment are considered, and key gaps where our current understanding is limited, and where further work is therefore required, are related.

1. Species identification

When establishing both harvesting and cultivation programmes for a new tree product, it is important to be able to properly identify the tree concerned.

Allanblackia has features (e.g. large fruit, a characteristic flower and a particular tree architecture) that allow it to be distinguished relatively easily from other genera in African forests. More important is the ability to be able to distinguish between different species (apparently nine) within the Allanblackia genus, for the following reasons:

  • Although little is currently known about the oil quality attributes of different Allanblackia species, they are likely to vary. In the future, production will likely focus on better quality species, and distinguishing these from other Allanblackia taxa when they co-occur will be necessary.
  • In order to protect the performance of cultivated stands, the species-identity of material entering cultivation should ‘match’ the identity of trees growing wild in a given area. This is to avoid problems of outbreeding depression (loss of performance because mating trees are too different) when cultivated trees are pollinated by wild stands. In a ‘local’ domestication strategy, in which material for cultivation comes from neighbouring natural stands, no problems should be encountered. However, if germplasm translocations take place over large distances, difficulties may arise if species are not properly discriminated between.
  • Proper circa and ex situ conservation of Allanblackia requires that species identity be accurately applied to natural stands, in order that priorities can be set for bringing material into local cultivation (for circa situ conservation) or into live gene banks (ex situ conservation). If identity is not accurately assigned, ‘rare’ populations may be ‘missed’.

In practice, the situation with species identification within Allanblackia is much more straightforward than for many other tree genera. This is because, of the countries currently involved in the Allanblackia initiative, in two (Ghana and Nigeria) only a single species is believed to occur in each, so no opportunity for ‘mixing’ of species arises (see Table 1; except in the event that germplasm is transferred between countries, which currently appears unlikely for Allanblackia).

Only in Cameroon and Tanzania do multiple species coexist, with three found in the former country and two in the latter (Table 1). In these countries, species are on occasions sympatric (close or overlapping distributions), a situation where proper identification is particularly important. In practice, however, species normally have rather different ecological requirements and overlapping distributions are limited. For example, in Tanzania, A. stuhlmannii appears to be primarily ‘submontane’ (800 to 1500 m), while A. ulugurensis primarily grows in ‘montane’ or ‘upper montane’ (> 1500 m) regions. Furthermore, there are generally other clear differences between species. For example, in Tanzania, A. stuhlmannii has larger fruit than A. ulugurensis. Cameroon appears to face the most difficult situation for discrimination, since A. floribunda and A. stanerana both widely occur in lowland forest, but fruit size and other features can discriminate between these species.

In summary, then, identification of, and discrimination between, different species should be a relatively straightforward exercise in all of the countries currently participating in the Allanblackia initiative.

Table 1. Allanblackia species and their apparent distributions in project countries

Country / Species present / Notes on distribution/identity
Cameroon / A. floribunda
A. gabonensis
A. stanerana / Allanblackia gabonensis is found in ‘submontane’ forest (> 500 m altitude), A. stanerana has identifiably smaller fruit than other species (?) and is found in coastal forest, mostly in west and southwest Cameroon, sympatric in part with lowland forest A. floribunda.
Ghana / A. parviflora / Single species means identification straightforward.
Nigeria / A. floribunda / Single species means identification straightforward.
Tanzania / A. stuhlmannii
A. ulugurensis / Sometimes find only one species present on a particular mountain in the Eastern Arc, sometimes find both (e.g. in the Uluguru’s). In the Uluguru’s, although sympatry is possible, A. stuhlmannii appears to be primarily ‘submontane’ (800 to 1500 m), A. ulugurensis primarily ‘montane’ or ‘upper montane’ (> 1500 m). Allanblackia stuhlmannii is distinguishable because it has larger fruit than A. ulugurensis.

2. Collection

Within the context of a domestication and sustainable use programme, germplasm collection of Allanblackia is important for a number of reasons, including for:

  • Raising seedling/other propagules for distribution to farmers (farmers may collect their own germplasm or it may be collected for them by an extension or other agency).
  • Tree management research (e.g. on how to germinate seed or vegetatively propagate trees).
  • Genetic improvement programmes (e.g. for the establishment of provenance field trials from which superior material can be selected through exploiting intraspecific variation in a species).
  • Ex situ conservation (e.g. establishment of field gene banks for long-term management of genetic resources).

Collections undertaken to date on Allanblackia have used a number of approaches. Material has been sampled as both seed and vegetative propagules, and both random and targeted methods to collection have been adopted. In the below, a brief description of the theory behind the different approaches that have been applied, each of which has potential advantages and disadvantages, is given.

2.1. Seed and vegetative collection

To date, most large-scale germplasm collections of Allanblackia that have had both practical deployment and research objectives (in Ghana and Tanzania) have been based on sampling seed (see Box 1 for some general guidelines for Allanblackia seed collection). In Ghana, for example, ITSC, in collaboration with local communities, has collected hundreds of thousands of seed from hundreds of trees, for establishment in nursery beds in their central facility at Offinso.

Similarly, in Tanzania, for example, ANR and ICRAF have collected and planted hundreds of thousands of seed in nurseries beds at ANR’s central facility and (with TFCG) in medium-scale community nurseries, in the East Usambara. In Tanzania, community nurseries have also become directly involved in seed collection, sometimes adopting innovative methods to sampling and handling that may be of more general relevance to the Allanblackia initiative (such as harvesting from rat caches/burrows, and fruit burial as a seed pre-treatment).

Box 1. Some general guidelines for Allanblackia seed collection (taken from current collection and propagation extension guidelines)
Time of collection
  • Allanblackia appears to shows one main fruiting season (~ December through February), sometimes with a second more minor fruiting period (~ May through July).
  • Collection is easiest during the main fruiting season (~ December to February).
  • Fruit should be collected from trees immediately after it falls to the ground, to prevent it going mouldy or seed being eaten by animals (repeated regular visits to populations are necessary to maximise collection opportunities).

Transportation

  • If transporting material large distances, it is better to transport seed within fruit rather than extracting it at the collection point.
  • During transportation, allow air to circulate between fruit to prevent overheating, and do not expose unnecessarily to sunlight.

Processing

  • On arrival at the seed extraction site, store fruit in an open shaded area protected from animals until it can be processed.
  • Extract seed from fruit by crushing between hands and then hand-separate seed from pulp.
  • Bury seed in moist sand until further processing or planting out can be undertaken.

It is anticipated that, as the Allanblackia initiative develops, a greater focus will be placed on vegetative- rather than seed-based propagation during deployment. In Tanzania, the collection of vegetative propagules for deployment purposes on a significant scale has already begun: leafy cuttings are currently being collected from 100 mature cut and coppicing female trees in farmland in the East Usambara region, and are being rooted in non-mist propagators (see Box 2 for some general guidelines for Allanblackia vegetative collection).

Box 2. Some general guidelines for collecting Allanblackia vegetatively (cuttings) (taken from current collection and propagation extension guidelines; for more details see other presentations from this course)
Cutting mature trees for source material
  • Cut mature female trees to a stump height of ~ 75 cm at the start of the rainy season (cutting at this time facilitates re-sprouting). Trees should be cut at a slight angle (about 15 degrees) to facilitate run-off (prevent rotting).
  • Leave trees to re-sprout/coppice, monitoring progress (it may take several months before any material can be collected).
  • Remove the buds of dominant shoots to encourage bushy development of regrowth (provides more material for harvesting).
Collecting and handling shoots cuttings
  • Take regrowth that is 2.5 to 12 cm long, with a stem diameter at the base of ~ 4 to 8 mm. Cut from the stock plant using sharp secateurs or a sharp knife.
  • Moisten cut shoots and place them in a plastic bag to prevent them from drying out. Keep the plastic bag in the shade to prevent shoots from overheating.
  • Material should be collected before it becomes too woody, since otherwise it will become more difficult to root.
Transportation
  • Transport cuttings to the nursery site as soon as is practicable. This may require a vehicle to move backwards and forwards between collection site and nursery area as harvesting proceeds.
Processing
  • On reaching the nursery area, ensure cuttings are placed into non-mist propagators as soon as possible (certainly on the same day as harvested). This requires good prior preparation at the nursery area – with propagators and necessary materials all ready for the receipt of cuttings.
  • Make a new clean cut to cuttings and follow the appropriate procedure for rooting leafy stem cuttings in propagators (see others sessions of this course on vegetative propagation).

Seed and vegetative approaches to tree germplasm collection both carry their own advantages and disadvantages (summarised in Box 3 below), some of which are particularly pertinent for Allanblackia (as highlighted). The key advantages associated with vegetative propagation for Allanblackia appear to be accelerated maturation (reduced time to fruiting), control of gender (Allanblackia having separate male and female trees) and the ability to circumvent the difficulties observed with seed germination. A key disadvantage for Allanblackia, when collecting large quantities of vegetative material, is the significant extra resources required during and after collection compared to seed sampling, especially in the establishment and maintenance of non-mist propagators.

Box 3. Advantages and disadvantages of a vegetative- compared to a seed-based approach to tree germplasm collection (with particular relevance for Allanblackia in bold)
Advantages associated with vegetative sampling may include:
  • Accelerated expression of important characteristics may be exhibited by vegetative material. For example, material may fruit earlier than if sampled as seed. This has advantages both for the evaluation of germplasm in tree improvement programmes and in the direct distribution of material to users. In the latter case, farmers may more quickly receive benefits from collected material. This issue is of particular importance for Allanblackia, with trees established from seed expected to fruit after ~ 12 years, and material propagated vegetatively (leafy cuttings taken from mature females) expected to fruit before the 5th year after establishment.
  • During collection, an exact genetic copy of the sampled tree is taken. This could be an advantage when carrying out selection for superior trees, providing increased efficiency during targeted sampling (see following sections; favourable genes and adaptive gene complexes in mother trees may be lost during the collection of seed because the pollinating parent is unknown). Collection of an exact copy also means that the sex of propagules taken from dioecious trees is known, which is an important consideration in efficient management of such species. The ability to control sex during the collection of Allanblackia planting material is significant, because it allows more female than male trees to be planted on farms. This should allow greater unit area productivity from planted stands.
  • Collection is possible when no seed is available or where seed handling is difficult. For some species, the appropriate time for fruit collection may be difficult to predict and may vary greatly between years. Some species may not fruit in some years, or, if outbreeders, be unable to produce fruit at all, due to genetic isolation as a result of population fragmentation. In these instances, the ability to collect vegetative material provides the only method of obtaining germplasm. Seed of fruit trees in particular can sometimes be difficult to handle, making vegetative propagation an attractive option. This issue is particularly relevant for Allanblackia, the seed of which can take considerable time to germinate, with low absolute levels of germination sometimes also observed. In addition, Allanblackia seed must be collected at just the right time, which requires careful (and sometimes difficult) synchronisation (when fruit has fallen from trees, but before it is eaten by animals).
Disadvantages associated with vegetative sampling may include:
  • Vegetative collection may suffer from practical difficulties. The techniques involved in collection may be difficult (possibly requiring considerable prior research for optimisation) and time consuming. Vegetative material is perishable – it must therefore be handled carefully in the field and can not normally be stored for long periods of time. Material may be bulky and difficult to process, requiring special equipment such as nursery propagators. Regulations on germplasm movement may be stricter when compared to seed, due to the increased potential for the transmission of viruses or other diseases. The need to build thousands of non-mist propagators for rooting of germplasm collected as leafy stem cuttings is a particular practical issue for Allanblackia.
  • Vegetative sampling is often combined with a targeted approach to germplasm collection, but the effectiveness of selection from natural stands is not well known (this issue is addressed further in the following sections).
  • Vegetative collection tends to focus on a small number of trees from any given provenance. Likely, this leads to a narrowing of the genetic base in collected material. Although reduced genetic variation may be an advantage in certain situations (e.g. when markets require a product of uniform size and character), it may also lead to a reduced capacity of germplasm to adapt to varying environmental conditions or user requirements. One approach to avoid a narrowing of the genetic base during vegetative sampling is to collect, evaluate and distribute a greater number of clones from provenances. However, this practice has rarely been followed. In the case of Allanblackia, genetic narrowing is unlikely to be a major issue simply because of the very large numbers of propagules required in project countries, which will require that material is cloned from many different trees.

Apart from collection via seed and vegetative means, some more minor collections based on wilding transplantation have also been successful. This is particularly the case in Tanzania, where community nursery groups around ANR have transplanted Allanblackia wildings into nurseries very local to forest sampling sites, using locally collected forest soil as the potting substrate. Wilding collection may provide a ‘stop-gap’ method while seed and vegetative propagation approaches are further developed, but it is unlikely to become a major collection approach in the Allanblackia initiative, and is not considered further in this session.