A vital fuelwood gene pool is in danger

Christel Palmberg

CHRISTEL PALMBERG is in charge of seed procurement and genetic resources in the FAO Forestry Department.

Few arid and semi-arid tree species have been explored well enough genetically and botanically for efficient utilization as fuelwood. A pioneer effort is now underway.

More than 1500 million people in the developing countries depend on wood for cooking and heating (Arnold, 1978; Arnold and Jongma, 1978). In some arid and semi-arid tropical areas, a dangerous imbalance between requirements and available supply of fuelwood already exists. In others a deficit is gradually developing because of growing population pressure. The effects of the subsequent scarcity of wood are numerous and self-reinforcing, and will eventually be felt in all facets of rural life. Trees and shrubs in rural areas traditionally serve not only as fuel, but also as shade and shelter, as building materials, as food or fodder, and as a source of a range of important products such as gums, resins and medicines. Over the long term, depletion of the natural vegetation will increase ecological fragility and contribute to gradual degradation of the resource base as well as the natural resources themselves.

In addition to an array of intertwined institutional, educational and technological solutions presently being sought to the problem, it is clear that biological aspects will have to play a predominant role in alleviating and eventually solving "the poor man's energy crisis" (Eckholm, 1975). Vigorous action is needed to protect, conserve and efficiently utilize existing natural resources.

Since the establishment of the FAO Panel of Experts on Forest Gene Resources in 1968, considerable progress has been made in the exploration, collection, evaluation, conservation and utilization of forest genetic resources (FAO, 1975a, 1978; Palmberg, 1980a; UNEP, 1980).

At first, in accordance with prevailing needs, emphasis was placed on fast-growing pioneer species suitable for large-scale plantations (FAO, 1969, 1972). There was already some understanding of the variation patterns and biology of these species. In addition, conservation and utilization were facilitated by knowledge of the methodology of seed storage and of techniques for the establishment and management of plantations and/or natural stands.

While stressing the importance of continuing work with fast-growing plantation species, the FAO Panel, at its Third Session in 1974, recommended the inclusion of slower-growing species adapted to difficult sites, such as those found in arid and semiarid zones (FAO, 1974a).

With the acceleration of the fuelwood crisis, the FAO Panel, at its Fourth Session in 1977, reinforced its recommendations, urging early action for arboreal species grown for fuel, food, fodder, soil stabilization, shade, shelter and farm forestry (FAO, 1977). The Panel also drew up a short list of priority species falling into these categories (see Table 1), placing a deliberate bias on species in arid and semiarid zones where environmental measures were likely to have the greatest effects.

On the basis of the recommendations, and with financial assistance from the International Board for Plant Genetic Resources (IBPGR), FAO's Forestry Department initiated in 1979 a project for the conservation and better utilization of genetic resources of arboreal species for the improvement of rural living. The main emphasis was placed on fuelwood species. After an initial phase, consisting of an on-the-ground survey of needs and possibilities, a second, operational phase was started in January 1981.

The situation for genetic resources of fuelwood species is clearly a serious one. No one disagrees. Yet, it has been difficult to obtain financial support for this project. The projects that receive priority in international development today are those requested by governments of developing countries to meet pressing short-term needs. Ironically, more widespread needs, ones that are serious, long-term and essentially international may not be seen. It is a classic example of the forest not being seen for the trees, and by foresters at that. Furthermore, national projects involving impressive amounts of financial and physical resources tend to crowd out those that are not only long-term and international in scope but which do not involve large amounts of money. This project, for instance, is costing less than $400000 for two years, of which $250000 is from IBPGR and the balance from the FAO Regular Programme, mainly the Forestry Department's Seed Funds. Normally, the regular programme budget does not finance projects. The United Nations Development Programme pays for most of FAO's projects. And $450000 for a two-year project is small indeed.

Since dry regions face more difficult fuelwood problems than either the humid tropical lowlands or the tropical highlands, it was decided that the project should give priority to the 600 million hectares of arid and semiarid woodlands receiving less than 500 mm of annual precipitation (NAS, 1980) and to the additional areas of seasonally dry tropical regions which suffer six or more rainless months.

The main aims of the FAO/IBPGR project are to act as a catalyst for gathering genetic information on arid and semi-arid zone species, and to aid countries in the practical application of any results which may become available.

Based on the list drawn up by the FAO Panel, and in accordance with the wishes expressed by potential cooperators, priority has been given to the exploration, collection and conservation of a few, select species in the genera Acacia, Eucalyptus and Prosopis (FAO, 1980a). Some countries will also collect seed of some sympatric species such as Atriplex, Azadirachta and Capparis (see Table 2).

Countries cooperating in the exploration/collection, conservation and evaluation phases are: Australia (collection/conservation only), Chile, India, Israel. Mexico, Peru, People's Democratic Republic of Yemen, Senegal and Sudan. However, other countries in arid and semi-arid areas will also benefit from the project since information and results will be made freely available. Moreover any seed in excess of that needed to establish a network of trials in cooperating countries will be distributed to any country interested in evaluating it on a "first come, first served" basis.

Every one of the tropical and subtropical tree species explored and collected during the last decade by cooperating governments under FAO coordination has been found to be in danger of genetic depletion, extinction or contamination in at least part of its natural range. Where the gene pool is not thought to be in danger of extinction, the population has often been so heavily reduced that supplies of seed are very limited (FAO, 197380; FAO, 1975c; Keiding and Kemp, 1978; Palmberg, 1980a).

Foresters all over the world at present acknowledge the fact that intra-specific variation of tree species may be as great as, or in some cases greater than, variation between closely related species. Marginal or isolated populations have often proved to possess specific characteristics (e.g., drought resistance or tolerance to alkaline or saline soils) of value for planting programmes on similar sites in other countries or regions (Stern and Roche, 1974).

Matching species and provenance with site and end use, and ensuring their conservation at the provenance as well as at the species level, is only possible if some basic knowledge exists regarding the biology, distribution and natural variation of potentially useful species (FAO, 1975a; Burley and Styles, 1976; Sneep and Hendrikson, 1979; Lamprey, 1975). As genetic variation does not necessarily manifest itself in situ, taxonomic and genecological exploration, followed by systematic sampling throughout the area of natural distribution and evaluation in field trials over a range of potential planting sites, are of fundamental importance (Palmberg, 1980b; Willan, 1973).

The practical importance of systematic exploration/evaluation has been convincingly demonstrated through a number of internationally coordinated species and provenance trials (Palmberg, 1980a). The earliest of these trials were started in the mid 1960s using provenance seedlots of Eucalyptus camaldulensis, an Australian species which for decades has been widely planted as an exotic and for the provision of fuel, shelter, posts, poles and pulpwood. The results, established on 32 sites in 18 countries, showed that the potential gains in productivity which can be achieved simply by the selection of the best-adapted provenances for prevailing environmental conditions, could amount to several hundred percent 300 percent in Nigeria and 800 percent in Israel, for example (Lacaze, 1970, 1978; Turnbull et al., 1980).

These results, although even more impressive than expected, had been foreseen. In addition to their practical value in drawing up programmes for both conservation and utilization of E. camaldulensis in Australia and elsewhere, they have served to underline the urgent need for systematic action in other dry-area species, a need compounded by increasing fuelwood shortages and the growing investments in fuelwood plantations which at present are generally established with seed from the most readily available provenances.

The general strategy of the present FAO/IBPGR Project follows recognized steps considered essential for any successful genetic resources programme, viz. (i) exploration and collection; (ii) evaluation; (iii) conservation; and (iv) utilization. Other activities will include training, the dissemination of information, and overall coordination.

Table 1. 41 tree species of arid and semi-arid environments that are particularly important for fuelwood

Trees	Utilization **
* Acacia albida	Fd	Fu	FF
* A. aneura	Fd	Fu	Sh	SS
* A. saligna (syn. A. cyanophylla)	Fu	Sh	SS
A. ligulata	Fu	Sh	SS
* A. nilotica	Fd	Fu	FF
A. peuce	Fu	Sh	SS
A. salicina	Fu	Sh	SS
* A. senegal	Fu	FF
* A. tortilis	Fu	FF
Anacardium occidentale	Fo	FF
Argania sideroxylon	Fd	Fu	SS
* Atriplex spp.	Fd	SS
* Azadirachta indica	Fu	Sh	FF
Calligonum spp.	SS
Casuarina decaisneana	Fu	Sh	SS
Ceratonia siliqua	Fd	Fu
Conocarpus lancifolius	Fu	Sh	SS
Eucalyptus astringens	Fu	Sh	SS
E. brockwayi	Fu	Sh	SS
* E. camaldulensis	Fu	Sh
E. gomphocephala	Fu	Sh	SS
E. intertexta	Fu	Sh	SS
E. leucoxylon	Fu	Sh
E. loxophleba	Fu	Sh	SS
* E. microtheca	Fu	Sh	SS
E. occidentalis	Fu	Sh	SS
E. ochrophloia	Fu	Sh	SS
E. salmonophloia	Fu	Sh	SS
E. salubris	Sh	SS
E. sargentii	Fu	Sh	SS
E. sideroxylon	Fu	Sh
E. tereticornis	Fu	Sh
* Gleditsia triacanthos	Fd	Fu	Sh	SS
Haloxylon spp.	Fd	Fu	SS
Kochia spp.	Fd	SS
Morus alba	Fo	Fu	FF
* Leucaena leucocephala	Fd	Fu	FF	SS
	(for wetter areas)
* Prosopis spicigera (syn. P. cineraria)	Fd	Fu	Sh	SS
* Prosopis spp.	Fd	Fu	Sh	SS
Tamarix aphylla	Fu	Sh	SS
Zizyphus spp.	Fd	Fu	SS

* Selected as priority species for the improvement of agricultural environments and rural living.

** Fo = Food; Fd = Fodder; Fu = Fuelwood; Sh = Shelterbelt; SS = Soil stabilization; FF = Farm forestry.

Source: Report of the 4th Session of the FAO Panel of Experts on Forest Gene Resources, 1880, Rome.

The main threat to arid and semi-arid zone tree species is the destruction of natural ecosystems as a consequence of changing land-use patterns and increasing human pressure.

Exploration and collection

Although it is often said that establishment of fuelwood plantations does not present serious technical problems (Anon., 1980a), planting programmes of non-industrial species, including arid-zone fuelwood species, are in practice still generally conducted at a species rather than at a provenance level, mainly due to lack of basic biological information. With the exception of some eucalypts (Turnbull, 1978), few arid and semi-arid tropical tree species have been sufficiently well explored botanically and genealogically to provide a solid basis for their efficient utilization and conservation.

Great confusion exists in the nomenclature of Prosopis spp. and some tropical Acacia species. This has been compounded by early introductions of unknown origin from one country to another, as well as by introgression and the formation of "land races" (i.e., exotic plantations which to various degrees have adapted to local conditions as a response to natural and sometimes artificial selection). Since early clarification will be important, an element on exploration/taxonomy has been included in the present FAO/IBPGR project.

Collection of seed for the evaluation of genetic variation will initially comprise range-wide sampling of the species listed in Table 2 on a fairly coarse grid, i.e., the collection of relatively small samples of seed from a relatively large number of seed sources. Seed will be collected from both natural stands and from "land races." Forestry institutes in the countries concerned will do the seed collection. FAO is coordinating the work.

A second stage of collections, involving sampling limited parts of the species ranges on a finer grid, may be necessary after the results of first-stage provenance trials have become available. In the second-stage collections or, alternatively, at a later stage, seed may be kept separate by mother trees to enable the evaluation of genetic variation within, as well as between, provenances.

Environmental gradients will be used initially to determine the number and location of populations to be sampled. Sampling within each population will be done at random. Seed, whenever possible, will be collected from a minimum of 25 trees, spaced no fewer than 100 m apart to avoid genetic relatedness of the mother-trees. Special caution will have to be observed in sampling of species within populations which are able to reproduce by root suckers, such as Acacia albida.

Table 2. National assignments for provenance collections

Nine countries cooperating in the FAO project on genetic resources of arid and semi-arid tree species

Country	Species	Observations
Australia	Eucalyptus camaldulensis Dehnh	11 new provenances collected for project with special emphasis on arid zones.
	E. microtheca F. Muell	73 seedlots collected for first-stage evaluation.
	Acacia aneura F. Muell. ex Benth	5 provenances collected by early 1980.
Chile	Acacia caven Mol.
	Atriplex repanda Phil.
	Prosopis tamarugo F. Phil.
	Prosopis spp. ("Algarrobo")	May include several species, P. atacamensis, P. siliquastrum, P. chilensis, P. burkartii.
India	Acacia nilotica (L.) Willd. ex Del	ssp. indica/var. vediana; var. jaquemontii; var. cupressiformis
	A. senegal (L.) Willd.	"Land Race"
	A. tortilis Hayne	"Land Race"; according to some sources may in fact be A. raddiana Savi.
	Prosopis cineraria (L.) Druce
	(syn. P. spicigera L.)
Israel	Acacia albida Del.
	A. raddiana Savi
	(syn. A. tortilis (Forsk) Hayne ssp. raddiana (Savi) Brenan)
	A. tortilis Hayne
	(syn. A. tortilis (Forsk) Hayne ssp. tortilis Hayne) Brenan)
Mexico	Atriplex canescens
	Prosopis spp. ("Mezquite")	May include several species, P. juliflora, P. glandulosa, P. alba, P. torreyana.
Peru	Capparis angulata
	Prosopis spp. ("Algarrobo")	May include several species, P. chilensis, P. limensis, P. juliflora.
Senegal	Acacia albida Del.
	A. nilotica (L.) Willd. ex Del	incl. var. adansonii
	A. raddiana Savi
	A. senegal (L.) Willd.
	A. tortilis Hayne
Sudan	Acacia nilotica (L.) Willd. ex Del	ssp. nilotica; ssp. tomentosa; ssp. astringens
PDR Yemen	Prosopis cineraria (L.) Druce
	Acacia nilotica (L.) Willd. ex Del	"Land Race"
	A. senegal (L.) Willd.
	A. tortilis Hayne
	A. tortilis Hayne	"Land Race"

Conservation, understood in its proper sense, embraces both preservation and utilization. For instance, some species of trees are endangered through overgrazing and others from lack of grazing.

Evaluation

Collection of range-wide samples will be followed by the establishment of species/provenance trials aimed at revealing potentially useful genetic variation, the degree of adaptability to a range of environmental conditions, and the economic or social value of the species/provenances.

The exact number of sites and the statistical layout to be used will partly depend on the amount of seed available. However, cooperating countries have preliminarily specified more than 40 potential planting sites on which they would be willing to test the species and provenances (FAO, 1980b).

The statistical design most likely to be recommended is a "randomized complete block" design. The advantages of this robust design are its suit ability for a wide variety of experimental situations and the ease of analysis and interpretation of test results, which even allow for the failure of one or more populations (Burley and Wood, 1976).

The decision to regenerate with nonnative species should be taken with great care. Local species must always be included in the evaluation. Only if exotics prove to be notably superior for all specified end uses as compared to natives tested under the same environmental conditions (including the use of optimal nursery, establishment and management techniques), should they be given priority in planting programmes. Conservation of representative samples of the native species must always be given high priority.

Conservation

The main threat to arid and semiarid zone arboreal species is the destruction of natural ecosystems as a consequence of changing land-use patterns and increasing human pressure. As evidenced by IUCN's Red Data Book, which classifies, for example, some species as being endangered from over-grazing and others from a lack of grazing (IUCN, 1978a), the key to conservation is appropriate management aimed at sustained utilization rather than passive protection or preservation.

No matter which strategies are chosen, conservation must always be an integral part of development. It must consider the users of the resources as well as the resources themselves (Sastrapadja, 1978). Conservation, understood in its proper sense, therefore embraces both preservation and utilization (Palmberg, 1980a).

A compromise between biological, technological, socio-economic and administrative goals is unavoidable when choosing strategies for the conservation and utilization of genetic resources. This implies balancing existing competing values against one another, while at the same time providing flexibility and security for future changes in markets and environments (Namkoong, 1978; Lamprey, 1975; IUCN, 1978b).

Unlike agricultural crops, which are usually conserved by storing seed, the conservation of forest trees and shrubs is generally done in conservation stands in situ or ex situ. This difference stems mainly from practical difficulties. Whereas plant gene banks can easily regenerate their seed collections whenever the viability falls by, at most, fifteen percent (Wang, 1978; IBPGR, 1976), the long vegetative period most tree species require before they produce viable seed manes seed regeneration an excessively long and expensive procedure.

As noted by the FAO/UNEP Expert Consultation on in situ Conservation of Forest Genetic Resources held in Rome in December 1980 (FAO, 1981), in situ and ex situ conservation are complementary: whereas in situ may be the only possible or realistic method for climax species without immediate economic importance or for species for which neither seed storage nor plantation techniques are known, ex situ is often an invaluable method in the conservation of species under heavy pressure in their natural ranges, as well as for species which may not be of present interest to their native countries but which are of proven socio-economic importance elsewhere, e.g., Acacia albida in Israel (FAO, 1980b). Both strategies will be used for the genetic conservation of the species included in the present project.

Conservation in situ is the more desirable method, provided both that an area can be given adequate protection and that the genetic material conserved is made available for use both within and outside the country of origin (FAO, 1975a; Frankel, 1978; Lamprey, 1975; Anon, 1980b).

Where a population is endangered but conservation in situ is not feasible due, for instance, to strong population pressure, changing land use priorities, or the danger of the local gene pool being contaminated by outside pollen (see Libby et al., 1975), early collection of substantial quantities of seed or other propagating material for subsequent establishment of ex situ conservation stands will be necessary.

Little information exists on the variation patterns and on genotype/environment interactions of the species included in this project. In addition, they have generally been subjected to unrecorded utilization and varying degrees of negative selection, which leave the poorest individuals as parents for the next generation. Therefore, random sampling within populations (i.e., sampling for variation rather than selective sampling) will be recommended in seed collections for conservation purposes. This will mean that further narrowing of the genetic base is avoided and that provenance-specific gene frequencies are maintained (Franker, 1970; Namkoong et al., 1980; Burley and Namkoong, 1980).

Even if great care is taken to avoid biased selection in the collection stage, gene frequencies will change over time in the ex situ conservation stands in response to natural selection (Namkoong et al., 1980). This problem will be at least partly overcome by establishing the stands in a number of different countries and environments, in which the selection pressures will vary and favour different genetic combinations.

At least some conservation measures have already been initiated in most of the cooperating countries. Although there is a general awareness by cooperating governments and institutes of the importance of conservation, severe pressures on the land will warrant a more systematic approach to both in situ and ex situ conservation. As noted above, efficient planning and action will, however, be dependent on more detailed knowledge of the distribution, variation and degree of threat to the species and provenances in question.

Species which to date have been identified by cooperating countries as being in need of urgent action (FAO, 1980c) are:

a) Acacia caven, Atriplex repanda and Prosopis chilensis in the Coquimbo region in northern Chile.

b) Some provenances of Acacia albida in Israel.

c) Acacia tortilis and parts of the range of A. nilotica in the Sudan.

d) Prosopis cineraria in Democratic Yemen.

Conservation action taken to date on an international level to safeguard specific provenances of dry-area species of socio-economic value includes the collection of seed from two eucalypt species (Eucalyptus camaldulensis and E. tereticornis). This action, taken under the auspices of an FAO/UNEP project with the collaboration of the Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia, will lead to the establishment of 11 ex situ conservation stands of some 10 ha each (FAO, 1977; Palmberg, 1980a). Although these stands will be located in countries different from those cooperating in the FAO/IBPGR project, valuable practical experience will be gained in the methodology of establishing and managing such stands.

SEED COLLECTION IN AUSTRALIA FROM Casuarina decaisneana at the basis of good forestry

Firewood Crops The selection and management of tree species considered particularly useful for fuelwood is the subject of Firewood Crops, Shrub and Tree Species for Energy Production, a book published by the National Academy of Sciences, Washington, D.C., 1980. It was reviewed in the Books section of Unasylva Vol. 33, No. 131, Wood Energy No. 1. The review carried a list of fuelwood species recommended by ecological groupings, i.e., best suited for humid tropics, tropical highlands and arid and semi-arid regions.

Utilization

As information about the most suitable seed sources becomes available, emphasis is expected to switch gradually from experimental seedlots for evaluation to the utilization of bulk and semi-bulk quantities of seed of provenances found to be well-adapted to the various planting sites. Although the supply of bulk quantities of reproductive materials for large-scale planting will remain the responsibility of government forest services or commercial seed merchants, it is envisaged for later to help countries in the procurement of semi-bulk quantities of seed (for plantations up to 500 ha) for the establishment of local conservation of stands, selection/seed stands and pilot plantations.

Gains likely to be achieved through tree improvement can be substantial, and they are of a lasting nature.

The long-term objective of the tree breeder is the generation of optimum breeding populations to create cumulatively better and more uniform genotypes for plantation establishment (Namkoong, 1978). In practice, however, this means that the genetic base of the population is deliberately narrowed to meet present-day needs.

This creates a serious conflict between short-term and long-term gains. The conflict can be mitigated by maintaining a hierarchy of separate populations side by side, representing a series of increasing selection intensity by decreasing population size (Burdon et al., 1978). These populations will, in principle, include the following:

Gene pool. A population in which the full range of genetic variation is maintained.

Selection population (base population). A large population (often a plantation of known origin) within which phenotypically superior individuals are selected. About one million trees is suggested as minimum size (500-1000 ha).

Breeding population. A population derived from the selection population, containing a minimum of 200-300 unrelated individuals chosen for their genetic superiority and used for further breeding. It is used in whole or in part to generate subsequent selection populations.

Seed-production population. A population of 30-50 individuals derived from the selection population or the breeding population to produce seed for plantations. In tree improvement programmes, seed production populations are generally represented by seed orchards.

Depending on the biology of the species, it may be possible to eliminate one of the parallel populations. For example, in a species propagated vegetatively, the seed production population would be eliminated; also, the gene pool may, in some cases, be equal to the selection population. On the other hand, the presence of strong genotype/environment interactions will necessitate the creation of several parallel breeding and seed production populations.

The majority of tree species studied to date suffer depression of growth upon inbreeding. It will therefore be essential to maintain a broad genetic base in the selection populations to allow for mating between non-relatives. Small plots established, for instance, with the primary purpose of species/provenance testing should never be used as the sole basis for selection and further breeding, although some genetic material selected in such plots may be used to enrich subsequent breeding populations.

Selection and breeding for traits such as high survival, resistance to pests and diseases, fast growth, and acceptable stem form under given site conditions are of universal value for any end use. Additional desirable characteristics in fuelwood species will include (Burley, 1978; Willan, 1977):

High specific gravity. Combining specific gravity with volume yield gives dry-weight yield per unit area, this is probably the most important single criterion for fuelwood. When making estimates, it should be noted that specific gravity in fast-grown, short-rotation crops may be different from that in mature or over-mature natural stands.

High calorific value of the wood. This is inversely related to its moisture content and directly related to content of extractives. Where there is a fuelwood shortage, plantation fuelwood is likely to be burnt soon after harvesting, and calorific value of green wood will be more important than calorific value of dry wood; where there is an opportunity to dry the wood before its use, natural durability during the drying period will be of importance.

Strong coppicing ability. This will spread the costs of establishment over a number of rotations.

Local acceptability. For example, trees free of thorns are easier to handle and transport; wood that burns without excessive sparking and odour is more agreeable. Species which will provide additional products or services such as food, fodder, shelter and soil amelioration or protection will be given priority in the improvement programmes.

What to do

Potential. The total existing area of all non-industrial plantations (fuelwood, shelter, protection, etc.) was estimated in 1975 at 3.3 million hectares, based on a survey covering 97 developing countries in Latin America, Africa south of the Sahara, Asia and the Far East (Lanly and Clement, 1979). Assuming that the contribution of wood to energy supplies were to be maintained at present per caput level within identical conditions of use, some 50 million hectares of additional fuelwood plantations would be needed to meet expected demands by the turn of the century in the developing countries alone. The cost of establishing fuelwood plantations in these countries (at 1980 cost levels) ranges from 200 to US$2000 per hectare (FAO, 1980c; Anon, 1980a), or a total of between US$10-100 thousand million.

Costs of plantation establishment will be the same regardless of the seed source used. The selection and use of optimal species/provenances will, on the other hand, increase potential production manifold, and these gains can be compounded by intra-population selection and breeding.

When viewed in the light of the above projections for areas and costs, and considering potential returns as evidenced by results from selection and breeding of industrial tree species (Carlisle and Teich, 1978), the investments urgently needed in intensive species/provenance research, breeding and conservation of fuelwood species are, in relative terms, of minor significance. Present resources are, however, grossly inadequate.

The area at present causing the greatest concern lies in the arid and semiarid climatic zones. It is hoped that the FAO/IBPGR project described in this paper will help to focus increased attention on the urgent, global needs of action in systematic exploration, collection, evaluation and conservation of previously little-known dry-area arboreal species.

References

ARNOLD, J.E.M. 1978, Wood energy and rural communities. 8th World Forestry Congress, Djakarta, Indonesia. FRC/3-0.

ARNOLD, J.E.M. & JONGMA, J. 1978, Fuelwood and charcoal in developing countries: an economic survey. Unasylva, 29 (118): 2-9.

BURDON, R.D., SHELBOURNE, C.J.A. & WILCOX, M. 1978, Advance selection strategies. Proc. 3rd World Consultation on Forest Tree Breeding. FO-FTB-77-6/2. CSIRO, Canberra, Australia.

BURLEY, J. 1978, Selection of species for fuelwood plantations. 8th World Forestry Congress, Djakarta, Indonesia. FRC/3-6.

BURLEY, J. 1980, Obstacles to tree planting for wood fuel in the Arid Zone with particular reference to India and Kenya. UNU/CEGET/CRNS Workshop on Fuelwood in Rural Communities of the Third World, Bordeaux, France, 5-9 May 1980.

BURLEY, J. & NAMKOONG, G. 1980, Conservation of forest genetic resources. Paper presented at the 11th Commonwealth Forestry Conference Trinidad.

BURLEY, J. & STYLES, B.T. (eds). 1976, Tropical trees: variation, breeding and conservation. Linnean Society, Oxford, UK.

BURLEY, J. & WOOD, P.J. (eds). A manual of species and provenance research with particular reference to the tropics. CFI Tropical Forestry Paper No. 10. University of Oxford, UK.

CARLISLE, A. & TEICH, A.H. 1978, Analysing benefits and costs of tree-breeding programmes Unasylva, 30 (119-120): 34-38.

ECKHOLM, E.P. 1975, The other energy crisis: firewood. Worldwatch Institute, Washington, D.C.

FAO 1969, Report on the 1st Session of the FAO Panel of Experts on Forest Gene Resources. FO:FGR/1/Rep. Rome, FAO.

FAO 1972, Report on the 2nd Session of the FAO Panel of Experts on Forest Gene Resources. FO:FGR/1/Rep. Rome, FAO.

FAO 1974a, Report on the 3rd Session of the FAO Panel of Experts on Forest Gene Resources. FO:FGR/3/Rep. Rome, FAO.

FAO 1974b, Report on the FAO/DANIDA Training Course on Forest Tree Improvement, Limuru, Kenya, September-October 1973. Rome, FAO.

FAO 1975a, The methodology of conservation of forest genetic resources, edited by L. Roche. FO:MISC/75/8. Rome, FAO.

FAO 1975b, Report on the FAO/DANIDA Training Course on Forest Seed Collection and Handling, Vol. 2, Chiang Mai, Thailand, February March 1975. Rome, FAO.

FAO 1975c, Proposals for a global programme for improved use of forest genetic resources. In Forest Genetic Resources Information No. 4. Forestry Occasional Paper 1975/1. Rome, FAO.

FAO 1977, Report on the 4th Session of the FAO Panel of Experts on Forest Gene Resources. FO:FGR/4/Rep. Rome. FAO.

FAO 1978, Genetics: selected papers from the Third World Consultation on Forest Tree Breeding. Canberra, Australia, March 1977. Unasylva, 30 (119-120).

FAO 1980a, Wood for energy. Secretariat Note, 5th Session of the Committee on Forestry, COFO 80/2, March 1980. Rome, FAO.

FAO 1980b, Genetic resources of tree species in arid and semi-arid areas: a surrey for the improvement of rural living in Latin America, India and Southwest Asia. Rome, FAO.

FAO 1980c, Mejora Genética de Arboles Forestales. Estudio FAO: Montes No. 20. Rome, FAO.

FAO 1980d, United Nations Conference on New and Renewable Sources of Energy: Report on the Second Meeting of the Technical Panel on Fuelwood and Charcoal. FAO, Rome. (Unpublished)

FAO 1981, Report on the FAO/UNEP Expert Consultation on in situ Conservation of Forest Genetic Resources. Rome, FAO. (In press)

FAO 1973-80, Forest Genetic Resources Information, Nos 1-9. Rome, FAO.

FRANKEL, O.H. 1970a, Genetic Conservation in Perspective. In: Genetic resources in plants - their exploration and conservation (eds. O.H. Frankel and E. Bennett). IBP Handbook No. 11. Oxford and Edinburgh, Blackwell Scientific Publications.

FRANKEL, O.H. 1978, Philosophy and strategy of genetic conservation in plants. Proc. 3rd World Consultation on Forest Tree Breeding. FO-FTB-77-1/1. CSIRO. Canberra, Australia.

IBPGR 1976, Report of IBPGR Working Group on Engineering, Design and Cost Aspects of Long-Term Seed Storage Facilities. Rome, FAO.

IUCN 1978a, The IUCN Plant Red Data Book. Morges, Switzerland, IUCN.

IUCN 1978b, Sourcebook for a world conservation strategy: drylands. General Assembly Paper GA.78/10, Add. 2. Morges, Switzerland, IUCN.

IUCN 1980, World Conservation Strategy: Living resource conservation for sustainable development. Document prepared by IUCN with the collaboration of UNEP, WWF,. FAO and Unesco. Gland, Switzerland.

KEIDING, H. & KEMP, R.H. 1978, Exploration, collection and investigation of gene resources: tropical pines and teak. Proc. 3rd World Consultation on Forest Tree Breeding. FO-FTB-77-1/3. CSIRO, Canberra, Australia.

KEMP, R.H. 1980, Increasing the supply of forest products. Paper presented at the 11th Commonwealth Forestry Conference, Trinidad.

LACAZE, J.F. 1970, Projet N° 6: Etude de l'adaptation écologique des eucalyptus. FAO/Comité de la Recherche Forestière Méditerranéenne. FO:SCM/FR 70/2/10. Rome, FAO.

LACAZE, J.F. 1978, Etude de l'adaptation écologique des eucalyptus; étude de provenances d'Eucalyptus camaldulensis. Proc. 3rd World Consultation on Forest Tree Breeding. FO-FTB-77-2/29. CSIRO, Canberra, Australia.

LAMPREY, H.F. 1975, The distribution of protected areas in relation to the needs of biotic community c conservation in Eastern Africa Occasional Paper No. 16. Morges, Switzerland, IUCN.

LANKY, J.P. & CLEMENT, J. 1979, Present and future forest and plantation areas in the tropics. FO:MISC/79/1. Rome, FAO.

LIBBY, W.J., KAFTON, D. & FINS, L. 1975, California conifers. In The methodology of conservation of forest genetic resources. FO:MISC/75/8. Rome,
FAO.

NAMKOONG, G. 1978, Choosing strategies for the future. Unasylva, 30 (119-120): 3841.

NAMKOONG, G., BARNES, R.D. & BURLEY, J. 1980, A philosophy of breeding strategy for tropical forest trees. CFI Tropical Forestry Papers No. 16. Department of Forestry University of Oxford, UK.

NAS 1980, Firewood crops: shrubs and tree species for energy production. National Academy of Sciences. Washington, D.C.

PALMBERG, C. 1980a, Principios y estrategia para el mejor aprovechamiento de los recursos genéticos forestales. In: Mejora Genética de Arboles Forestales. Estudio FAO: Montes No. 20. Rome, FAO.

PALMBERG, C. 1980b, Progress in the global programme for improved use of forest genetic resources. Paper presented at IUFRO Symposium and Workshop on Genetic Improvement and Productivity of Fast-Growing Tree Species, Aguas de São Pedro, Brazil.

SASTRAPRADJA, S., ADISOEMARTO, S., KARTAWINATA, K. & TARUMINGKENG, R.C. 1978, The conservation of forest animal and plant genetic resources. 8th World Forestry Congress, Djakarta, Indonesia. FQL/26-0.

SNEEP, J. & HENDRIKSEN, A.J.T. (eds). 1979, Plant breeding prospectives. Centre for Agricultural Publishing and Documentation, Wageningen, Netherlands.

STERN, K. & ROCHE, L. 1974, Genetics of forest ecosystems. Ecological Studies 6. Berlin, Heidelberg, New York, Springer-Verlag.

TURNBULL, J.W. 1978, Exploration and conservation of eucalypt gene resources. Proc. 3rd World Consultation on Forest Tree Breeding. FO-FTE-77-1/4. CSIRO, Canberra, Australia.

TURNBULL, J.W., NIKLES. D.G. & BROWN, A.G. 1980, Better use of the gene resources of some Australian forest trees. Paper presented at the 11th Commonwealth Forestry Conference, Trinidad.

UNEP 1980, Genetic resources: an overview. UNEP Report No. 5. Nairobi, Kenya, UNEP.

WANG, B.S.P. 1978, Procurement, handling and storage of tree seed for genetic research. Proc. 3rd World Consultation on Forest Tree Breeding. FO-FTB-77-4/8. CSIRO, Canberra, Australia.

WRIGHT, J.W. 1976, Introduction to forest genetics. New York, San Francisco, London, Academic Press.

WILLAN, R.L. 1973, Forestry: improving the use of genetic resources. Span, 16 (3): 119-122.

WILLAN, R.L. 1977, Note on tropical fuel wood species. Unpublished note for the file, Forest Resources Development Branch. Rome, FAO.

WILLAN, R.L. & PALMBERG, C. 1974, Improved use of forest genetic resources. In Report on the FAO/DANIDA Training Course on Forest Tree Improvement, Limuru, Kenya, September-October 1973. Rome, FAO.