International concern for tropical forests has risen sharply during the past decade. This period has seen the start of the Tropical Forestry Action Programme (TFAP) as a framework for action in tropical forestry, and the creation of the International Tropical Timber Organisation (ITTO) as a forum of consumers and producers of tropical timber moving in international trade, and discussion of the problems of deforestation at the highest international levels, in the World Commission on Environment and Development, in the UN General Assembly, in the Intergovernmental Panel on Climate Change, at summit meetings of the group of seven leading industrialised countries, and last but not least the United Nations Conference on Environment and Development in Rio de Janeiro, Brazil in June 1992. At the same time the FAO Forest Resources Assessment 1990 Project has revealed that the rate and scale of tropical deforestation in the past decade have been even greater than was previously predicted (FAO 1992c).
Central to any solution to this problem must be the reconciliation of the need for development and increased living standards in the tropical countries, based on the sustainable utilization of each nation's resources, including the natural forests, with the conservation of their capability for renewal and adaptation to changing conditions and needs: their genetic resources. The problem is complex, being bound up with broader aspects of land use and population increase, as well as the impacts of international debt, trade and aid relationships. Part of the difficulty in the search for quick and effective solutions is the very complexity of the tropical forests themselves. In terms not only of the many varieties of dry forests, moist forests, hill and coastal formations, but also the different social and economic conditions that surround them, there is not one but many thousand forest problems, each in need of separate analysis and management.
One aspect of this complexity of the tropical forests that has now attracted international interest is their biological diversity. This brings a new appreciation of this unique value of the natural forest ecosystems, in addition to their productive and environmental values, which have often proved insufficient to ensure their protection. Diversity considerations are generally poorly, or not at all, represented in formulae used to determine the economic value of an area, and have been discounted in competition with the immediate needs for land, food production and revenue from intensive timber production, for example through fast-growing plantations, in place of natural forest. As human populations continue to increase the opportunity cost of setting aside large areas purely to conserve biological diversity for ethical reasons and often un-specified “future needs”, becomes increasingly difficult to defend. At the same time pressures to maximise short-term timber volumes and revenue from production forests may increasingly threaten the continued existence of natural forests under selective harvesting systems, in which funds and revenue are re-invested in care and management of the areas harvested. However to the extent that the conservation of genetic diversity can be seen as one of the services of natural forest under multiple-use management, including appropriate levels of sustainable timber production, the reconciliation of developmental and conservation objectives may be easier to achieve. The exceptional wealth of the tropical forests in biological diversity makes this a strong possibility, given adequate provision for effective management.
In the wake of preferences for other land use options, the progressive reduction in size and increasing isolation of the remaining areas of natural forest in most countries imposes the need for increased efficiency in their management. This is partly to ensure the maximum levels of production, to defend the forest against increasing demands of increasing populations, but also to the increased stresses on the ecosystem resulting from the loss of the “buffering” effects of large areas of surrounding forest. This changed situation may have direct impacts on the environmental conditions around the forest, and also indirect influence on the breeding and seed dispersal systems of tree species dependent on animal vectors.
The component of the genetic diversity in the forest which is of actual or potential use, either for production or to maintain the forest as a functioning system, constitutes its genetic resources. The conservation of genetic resources is absolutely fundamental to the long-term success of all other forms of diversity conservation (Riggs 1990) and is also essential to the sustainable and productive management of the forest ecosystem in which they occur. To this extent therefore in situ conservation of forest genetic resources should reinforce the conservation and management of production forests, and vice versa. At the same time the systems of National Parks and other fully Protected Areas also have an essential role, particularly in the conservation of biological diversity of very uncertain use value or existence value, as well as the genetic resources of species of economic importance. However the coverage of the tropical forest biome within such systems is limited and they will never satisfy the full needs of genetic resource conservation. New approaches are therefore needed to integrate national conservation activities for maximum effect in both production forests and fully Protected Area systems.
FAO has the leading international role in linking forest management to the conservation of forest genetic resources, particularly through the work of the FAO Panel of Experts on Forest Gene Resources, established in 1968. Following publication in 1975 of the book, “Methodology of Conservation of Forest Genetic Resources” (FAO 1975), “Guide to in situ Conservation of Genetic Resources of Tropical Woody Species” was published by FAO in 1984 (Roche and Dourojeanni 1984). These documents reviewed the issues and activities involved in genetic resource conservation in forestry. Three case studies, from Cameroon, Malaysia and Peru, were published in the later document, which further reviewed the role of Protected Area systems, including those within production forests, such as the Virgin Jungle Reserves in Malaysia, which were discussed in detail. The principles and procedures set out in both documents still provide valuable guidance to the subject. Since then the compilation and publication of reviews of management practices in tropical moist forests in Africa (FAO 1989b), Asia (FAO 1989c) and Latin America (FAO 1992a) together with other broader reviews of the subject from both the managerial and the ecological viewpoints (e.g. Mergen and Vincent 1987; Wyatt-Smith 1987a; Poore 1989; Whitmore 1990; Gomez-Pompa et al 1991), have provided new opportunities to consider the possible role of production forests in relation to the conservation in situ of forest genetic resources. The FAO Panel, at its 7th Session in December 1989, endorsed proposals to update and expand the guidance given in the 1984 publication on in situ conservation of tropical woody species. In doing so it was decided to concentrate on the role of tropical forests under management for the production of timber or other products and services in the conservation of the genetic resources of the woody species. These, and particularly the largest and longest-lived trees, are dominant in determining the unique genetic potential of the system.
The diversity of tropical forests is a product not only of the large number of species present on a given area but of successional change over time, from the colonisation of gaps or cleared areas by pioneer species, through complex successional stages, to the mature “climax” forest. Economic forces and market demands have produced management systems aimed to simplify and to truncate the natural complexity and successional stages in the forest, to concentrate the growth potential into relatively few species, and short cutting cycles, with very limited reinvestment in management activities. This situation, in which short-term economic considerations have frequently overridden ecological concerns, has been governed by the apparently low value of the production per unit area of forest, as expressed in fees and log values in the forest. This has been generally incompatible with the level of investment in management needed to conserve the genetic resources even of the most important economic species, except perhaps and incidentally, those of the pioneer and fast-growing gap-opportunist guilds. Even on the basis of the very limited knowledge of the species composition and intraspecific variation in the forests it is clear that large parts of the useful or potentially useful genetic resources are in danger of being lost. Resolution of this problem is dependent on a wide range of integrated action, within which scientific research and professional forest management are only part of the solution.
The rapid pace of change in market demands and opportunities has often rendered invalid both the initial objectives of management and the programmes of data collection or silvicultural action undertaken to achieve them. The danger of adopting too narrow and short-term objectives is even greater in respect of genetic resource conservation, which has to consider possible changes in needs and opportunities over much longer periods than one or two rotations. These are likely to include rapid advances in the means to handle, manipulate and recombine genetic material. At the same time advances in information technology are already transforming the possibilities to handle and interpret large and complex arrays of data for the better understanding and use of functional relationships in the management of the forest. These developments greatly improve the opportunities for multiple-use management, including the conservation of genetic resources. A very important aspect of such multiple-use approaches, both for the long-term security of the forest ecosystem and for the conservation of genetic diversity, is the incorporation of non-timber forest products and the interests of local communities within the management system.
Whereas in the past, with more limited resources for managing complexity in the system, and narrower management objectives, the response to the forest's diversity was to simplify the system, by concentrating action on few species, often with unpredictable and unexpected results, much of the current and future value of the natural forest is seen to lie in its genetic diversity. A more appropriate future response to the management of diversity will be greater diversity of management. This may be achieved in various ways and at various levels, from multiple-use of the same area of forest, either simultaneously or in successional stages, through separate management by compartments or working circles, to the integrated and diversified management of the entire national forest estate, embracing production forests, protection forests, genetic conservation areas, Protected Area networks, and areas combining two or more of these functions.
Even for the management of genetic resources of species whose natural range transcends national boundaries the national borders define the ultimate unit of effective management, whose freedom and effectiveness to act may be further enhanced through international cooperation. It is ultimately the national government which holds the power to formulate the necessary policies in land and resource use which govern the possibilities both for sustainable forest management and for the conservation of the nation's biological civersity and genetic resources.
Each country, and to a large extent each forest area, is unique in terms of its genetic resources and of the appropriate strategies at both national and local level to manage the forests, both for production and for conservation objectives. Attempts at management solutions at the level of the individual forest, alone and in isolation from its setting in national development policies, not only for forest and land use but embracing forest industry, trade and the linkages with other sectors, can only have very limited chance of success.
For this reason a detailed case study of the problems facing the conservation of forest genetic resources in one country (Ghana) and of approaches to their solution, with shorter contrasting examples from two other countries (India, Brazil), form a large part of this overall study of policies and activities needed to link conservation with forest production through appropriate management. These case studies are presented in Part II of the present document.
In Part I, Capter II briefly summarises the essential elements of forest genetic resources. Chapter III considers the nature and extent of the impacts of management for timber production on the genetic resources of the forest, and some possibilities for multiple-use management. Chapter IV attempts to foresee the future for the natural forests, in terms of both international and domestic demand for timber and other products, and the probable influence of other factors such as population increase, and environmental concerns, including international interest in the conservation of biological diversity. Finally in Chapter V suggestions are made for the formulation of National Strategies for the Conservation of Forest Genetic Resources.
Appendix 1 provides an example of the careful research needed to provide information on the reproductive biology and genetics of individual target species within a national strategy.
It is hoped to complement the present study in the future with more detailed guidelines aimed at harmonising sustainable forest management for productive purposes with the conservation in situ of the species, provenances and populations under use, for the benefit of these two complementary objectives.
In the same way that the term “forest resources” refers to the usefulness of the forests for the production of timber or other products for human benefit, the term “genetic resources” implies that elements of the genetic variability of the trees and other plants and animals will be used to meet human needs and objectives. However as compared with the harvest of timber or the collection of other products from the forest, which can be used immediately, the benefits from the genetic resources can be used not only in current programmes but are essential for future development in the next and subsequent generations; and for the continued adaptation of the resources to changing environmental conditions and human needs.
The other important aspect of the genetic resources of natural forests, especially the tropical forests, is their great diversity, and this range of variation provides the basis for selection and improvement of the products and other benefits to meet future needs, so far as they can be foreseen. Therefore the greater the uncertainty over future demands, the greater the potential value of conserving diversity. Apart from the relatively small number of tree species of current economic importance, or domestic importance to local communities, which must be the main focus of conservation programmes, there are likely to be several hundred or even thousands of other species present with lesser or unknown values. Some of these might form an important part of future harvests of wood, timber and other forest products, in response to changing environmental conditions or market demands.
Genetic diversity occurs at various levels of organisation from the ecosystem, through its component species, their sub-specific populations (provenances), family groups and individual genotypes to the molecular level, of the gene. While ecosystem conservation may also achieve the conservation of some included species and genotypes others might be lost unless, for example in the case of important timber trees, the species and its genetically distinct component populations, are also targeted for specific conservation measures (FAO 1989a). Within such populations there is likely to be substantial variation between individual trees (genotypes). Depending on the pattern of distribution of such variation through the stand, as a result of the nature of the species' breeding and dispersal systems, highly valuable genetic resources at this level may in turn be lost, even if the population as a whole survives. In this respect, special care is needed to ensure that progeny of the best individuals of desirable species, when logged at maturity, is adequately represented in the existing regeneration, to avoid dysgenic effects and loss of diversity.
At the level of the gene allelic differences could be the basis of valuable traits, for example in resistance to insect pests or to severe environmental stress, of great potential value for adaptation to changing environmental conditions and for future use. The rapid advances in genetic engineering may permit new combinations of genetic characteristics which could transform the possibilities for highly productive or otherwise desirable plantations, making best use of deforested and degraded land. The chances of recognising such genetic variation at that level in the natural forest are clearly remote. However the strong probability of such potentially valuable genetic resources occurring in highly diverse populations is a further reason to guard against inadvertent loss of diversity.
It is therefore essential that all levels of genetic diversity be considered, and to the extent appropriate and practicable, included in the objectives and the activities of a conservation programme (Namkoong 1990). Moreover it is the organisation and structure of genetic diversity at the various levels that underlie both the functioning of the ecosystem and the approaches to the conservation of the genetic resources of individual species (Riggs 1990).
The genetic structure of a species is defined by the way in which genetic variation is distributed between and within populations. This structure is the result of mutation, migration, selection and gene flow between separate populations and is strongly influenced by the genetic system, embracing the mating system and dispersal systems for pollen and seed. Information on the diversity and distribution of genes within species and their local populations is essential to the management and conservation of their genetic resources, but such information is very limited or non-existent for most tropical tree species.
As stressed above, the conservation of the diversity of the natural forests is dependent on maintaining all essential functional components of the ecosystems in situ. These are likely to involve a range of ecological interactions, particularly symbiotic relationships and interdependent connections, for example between tree species and their animal pollinators, seed dispersers and so forth. Although the objective may in many cases be the conservation of particular target species and populations, in practice this is likely to involve conserving whole communities, at least until we have a more complete understanding of ecosystem dynamics. Such research as has been done commonly indicates hidden complexities and interactions, for example, among so-called “plant-web” or “food-web” systems (Gilbert 1980; Terborgh 1986; Whitmore 1990). A wide variety of animals, including many groups of invertebrates, birds, bats and other mammals, are involved in pollen and seed dispersal of trees of known economic importance and may in turn be dependent for their survival, in the areas of forest where they are needed, on sources of food or breeding niches provided by quite different and apparently insignificant trees or other plant species. The disappearance of such “keystone” or “pivotal” species might then lead to the loss of plant species, including timber trees, dependent on the animals for pollination or seed dispersal (Howe 1990).
Although the great majority of species in mixed tropical forests may play no essential part in ecosystem function, and may therefore be in that sense redundant, the present level of knowledge and understanding is inadequate to determine with certainty all the key components of the ecosystem. The precautionary principle therefore requires that management and harvesting practices should conserve as wide a spectrum of species of as yet uncertain linkages and values as is practicable. Such an approach is consistent with the growing international concern for the conservation of biological diversity in general. Tropical forests and woodlands, in addition to the genetic resources of woody species they contain, are the habitat for a wealth of other plant and animal species, all of which contribute to the total sum of genetic diversity and resources on earth.
We may therefore distinguish four arbitrary categories of forest genetic resources for conservation in situ: (i) the principal economic tree species (ii) other trees, plants and animal species of known but lesser value in the national economy or to local communities (iii) key functional species for sustainability of the ecosystem (iv) other elements in the total biological diversity, of as yet uncertain value.
Just as the overall objectives of forest management for productive purposes will give preference to the desirable trees of the principal economic species these species will also be a main target for in situ conservation of the forest's genetic resources. This will aim to maintain viable breeding populations, and to favour reproduction of the better individual trees, insofar as this may be judged from their phenotypic appearance in the forest. This objective should be a normal component of all sustainable forest management but in practice is likely to increase the cost of harvesting and management operations as they are currently conducted in most countries. To add further to the costs of management and harvesting, through greater care to avoid damage to seed trees and regeneration of additional “non-economic” species, would therefore require the imposition of special conditions, through incentives or constraints. Nevertheless the maintenance of viable breeding populations of “keystone” species and those important for domestic use by local communities may be essential for long-term sustainable management of the forest for all its values, including the production of timber.
Some overall loss of biological diversity must be inevitable as a result of harvesting and management practices, at least for a period and in the areas affected by these operations. A single management system, if consistently practised, will affect all stands similarly. The timing, frequency, intensity and extent of logging will be the principal factors which determine the severity of the impact on overall biological diversity. However the use of diverse management systems can increase the variation between management units. The impact on the principal target species will depend also on the level of understanding of the pattern of variation within the species, both within and between different populations in the forests, and of its interactions with other species in the ecosystem, and then the willingness and capability to adjust management and harvesting operations to minimise losses in diversity.
The practical implications of the need to understand the genetic structure include the likelihood that genetically distinct populations of a species may exist in different forest areas within the species' range (i.e. different provenances of the species), as a result of isolation and/or adaptation to different environmental conditions. These distinct populations may differ substantially in their socio-economic value and production potential, or offer possibilities for improvement by combination of their qualities through breeding programmes. The design of in situ conservation strategies is therefore particularly dependent on estimating the likelihood and location of such intra-specific variation.
Another important aspect of the genetic systems of individual species in determining patterns and levels of diversity within provenances and populations is the extent to which pollination takes place between different individual trees (outbreeding) or the frequency of self-pollination, or even the production of seed without fertilisation (apomixis). Although very few species have been adequately studied many scientists who have undertaken such investigations have concluded that there is a strong tendency for cross-pollination and outbreeding among tropical tree species (Bawa 1974; Bullock 1985; Bawa et al 1985; Bawa and Krugman 1990; Janzen and Vasquez-Yanes 1990). Some authorities believe that the high gene flow between individuals resulting from the predominance of cross-pollination contributes to the high levels of genetic diversity characteristic of the tropical forests. Others have attributed high levels of speciation in tropical rainforests to the isolation of the individual trees of a species, with consequent reliance on inbreeding or apomixis (for a discussion of these views see Whitmore 1990 and Bawa et al 1990).
The practical implications are that without specific safeguards, based on adequate knowledge of the genetic systems and variation patterns of important tree species, the effects of logging or extensive deforestation on the population, through alterations in the levels and patterns of outcrossing or inbreeding, are likely to increase with increased severity of disruption of the populations. As a result there could be a lack of adequate fertile seed production, or excessive inbreeding, which could endanger the viability of the species or the provenance in the longer term. The effects might not necessarily be negative, since in some circumstances the removal of closely related individuals could encourage wider outcrossing. However the essential aspect is that the effects will be unpredictable in the absence of adequate information on the species' population biology.
The absence of such information in respect of almost all tree species in most areas of tropical forest must severely limit the possibilities for effective management of their genetic resources. The safest practicable conservation strategy must be to maintain the broadest possible genetic base in the species as a whole by conserving as wide as possible a range of provenances over the natural geographical and ecological range. In doing so it is also desirable to protect the integrity of each such population against genetic pollution which can be caused by the introduction of outside provenances of the species, such as might occur for example in the course of gap planting or enrichment planting within the forest, or following the establishment of plantations using outside, hybridising species or provenances in an adjacent area within the range of natural pollen distribution.
The relative value of different individual trees or populations of a species is very difficult to judge in natural stands, where uncertainties over age and past history are compounded by accidental differences in site quality and competition with other vegetation. For the few tropical tree species which have been the subject of provenance trials and breeding programmes highly significant increases in productivity and related economic and social benefits have been achieved, based on selection from the very diverse provenances sampled. These gains have been achieved through careful comparative trials and the application of the results in fast-growing plantations. The genetic resources of other tropical tree species to be conserved in situ in the natural forests might provide the basis for similar increased productivity if subjected to trials ex situ to determine their heritable qualities and performance in given environments.
Economists recognise two main types of value: use values and non-use (existence) values. Use values may be further sub-divided into those available for known and immediate uses and those which might become available in the future (option values). While the direct use value of the genetic diversity in the forests can best be measured in respect of the few most marketable species the option value of species not currently in demand may be quite high. The latter could become important to adapt to possible climatic changes, for example, or to meet changed market demands for timber or non-timber forest products (NTFP). There have been many examples in recent years of species earlier regarded as unmarketable or as “weed” species, but which are now highly valued. The effect is to increase the options open to the forest manager, as a result of having conserved a wider range of genetic diversity in the natural forests than those species previously known to be of immediate use (see the Ghana case study in Part II).
Among the important indirect (non-consumptive) use values of the genetic diversity of tropical forests is its possible contribution to the stability of the ecosystem particularly in the face of climatic changes at global or regional levels. Virtually nothing is known of the functional role of overall diversity in ecosystem stability, or of the acceptable levels of species loss or the thresholds of irreversible change with decreasing levels of genetic diversity in natural forests. Similarly it is impossible to forecast with certainty the response of tree species to the likely climatic changes. These high levels of uncertainty, combined with the dangers of irreversibility in the reduction of genetic diversity place emphasis on following the precautionary principle of avoiding unnecessary loss of such diversity and of genetic resources.
The value of conserving the genetic resources of a range of populations of a species of established socio-economic importance may also be greater in the context of expected climatic changes. Particular interest attaches to the edge of a species' natural range, where the local populations may be adapted to more extreme environmental stresses, for example in the transition zones between moist and dry forest types, or between dry savanna woodland and more arid thorn scrub or desert formations. Moreover recent developments in molecular genetics and genetic engineering may give additional value to such populations which, although they may exhibit very slow growth and poor stem form or other production limitations, may contain valuable genetic material conferring characteristics such as resistance to drought, or to high salt levels in the soil. Similar considerations apply to populations of species in other transition zones, for example, in the inter-tidal zone of coastal forest formations, where the effects of sea-level rise associated with global warming could be severe.
The final category of value - existence value - is likely to be relatively high for the rarer and more precious tropical tree species most vulnerable to genetic impoverishment, or even to extinction, through intensive harvesting without adequate management, and without adequate attention to the conservation of their genetic resources. It is such existence values, coupled with concern for the existence of tropical forest ecosystems as a whole, that are uppermost in the minds of the public and the media in many developed countries, particularly in regard to tropical rain forests.
While there may be future benefits from the conservation of a wide spectrum of genetic diversity, including some of possible indirect value in ecosystem stability, and others of uncertain option value, in respect of as yet unidentified market opportunities, there are likely to be immediate financial costs. These may be both direct costs of protection and management, and also indirect costs of opportunities foregone in the immediate term through lower levels of production from the forest as a whole. In the short term the stability of the forest ecosystem is unlikely to be affected by a reduction in species diversity. Although this could lead to the progressive elimination of e.g. some slower-growing cabinet or joinery timber species, especially if accompanied by more intensive harvesting systems that tend to favour fast-growing pioneer species, the overall levels of total wood production may also be largely unaffected by loss of species diversity.
The greatest diversity of tree species is found in the lowland humid tropics, for example at Yanamomo, in the Peruvian Amazon, where 283 species of trees of 0.1 m in diameter or larger were recorded on a one hectare plot (Whitmore 1990). Historical and environmental influences have led to the concentration of species diversity in certain areas, as a result of evolutionary pressures, related for example to long periods of environmental stability or to periodic disturbance, isolation, the removal of barriers between populations, migration and other influences. Although the level of information on species numbers and distribution patterns is still very sparse biogeographic divisions between the world's major terrestrial ecosystems may be used as an initial stratification, and with other available data provide a basis for selecting areas of high levels of diversity, or of exceptional degrees of endemism, or both together, as high priorities for ecosystem conservation. These two criteria, together with the degree of depletion or of threat to the genetic resources of the area, have been used by various authors and organisations to single out countries and locations to be accorded high priority (e.g. Myers 1988; Reid and Miller 1989; McNeely 1990). Such areas are most appropriate for conservation within fully Protected Area systems, such as Nature Reserves or National Parks, rather than areas managed for the production of timber or other products. The criteria suggested above may be the most appropriate to select a single location of high priority for ecosystem conservation but if more than one area can be selected as part of an integrated system of reserves the criteria should be extended to include a diversity of sites to capture also intra-specific variation of target species. Moreover the association of adjacent areas of natural forest under productive management as “buffer zone” to the fully Protected Area may provide a valuable extension to the range and size of populations of many tree species, and to the efficient coverage of their intra-specific variation.
The role of managed production forests in genetic resource conservation is particularly important for the conservation of intra-specific variation at the population level, of tree species of known or probable value. Ideally the location of areas serving such conservation objectives should be determined from data on patterns of variation, or of genetic structure and gene flow, since the latter should reveal the geographic scale over which populations may diverge from each other. In the absence of such information, some reasonable assumptions can be made from geographical and ecological data.
Differentiation between populations may develop in response to selection pressures resulting from local environmental conditions. There is thus generally a correlation between geographic/ecological factors on the one hand and inherent morphological or physiological properties of the local population of a species on the other (FAO 1989a). While the fully Protected Area system of Nature Reserves, for example, may include some part of the range of a species the effective conservation of the gene pool (i.e. the total sum of genetic materials) of the species as a whole requires the inclusion of a much wider range of populations representative of possible genetic differences which can only be surmised from their geographical or ecological situations (Frankel 1970). This is likely to require a number of conservation areas distributed over the entire natural range of the species and to the extent that such a strategy might be practicable in any given case it is likely that most of these would have to serve multiple objectives, including production of timber as well as other forest products. The desirable number and location of such conservation areas must be determined for each species and in the absence of specific data it may be assumed that for widespread and strongly outcrossing species a few locations in each major ecological/geographical zone might be adequate. For strongly inbreeding species, and those exhibiting scattered and isolated occurrences, more conservation sites might be needed.
Much thought has been given, particularly in respect of populations of large animals, to the minimum size of a viable population that would allow scope for continued undirected evolution, and related to that the minimum conservation area needed. Other calculations have been made based on known or probable levels of inbreeding and the size of population needed to minimise consequent loss of genetic variability, over a given number of generations. A commonly quoted figure is a minimum of 50 breeding adult individuals for short-term maintenance of fitness in the population, and 500 to sustain long-term genetic adaptability to change (FAO 1989a). A figure of 1 000 individuals has been suggested as desirable to maintain “evolutionary potential”, while at the other extreme it has been estimated that a genetically effective size of less than 50 may be adequate for several generations and that nearly all the genetic variability of a population can be conserved temporarily in only a few breeding individuals (Wilcox 1990).
For long-lived species such as trees and those under some degree of management over much of their natural range, namely those of recognised economic or local value, individual population size is less important than the distribution of conservation sites to sample likely patterns of diversity across the species' range. In the absence of information on patterns of variation within a species the conservation of populations of a few hundred individuals at the extremes of the geographical and ecological ranges is the most practical option. However for highly diverse rainforest containing many hundreds of tree species each of which may exist normally at very low frequency, conservation areas of 5 000 ha have been suggested as a “rule of the thumb”, based on estimates that this would cover 95% of the species (Ashton 1984).
An essential feature of in situ conservation is that it provides for “continuing evolution” (Frankel 1981). In this sense both the genetic resources themselves and the practice of their conservation are essentially dynamic and should not be seen as an attempt to preserve a fixed and finite resource. At the same time conservation implies the avoidance of the rapid erosion of genetic variability, for example through the extinction of species or unique populations, or strong directional change in the genetic composition of a population as a result of severe reduction in numbers, causing increased inbreeding and/or genetic drift, as a consequence of isolation. As the possibilities for setting aside substantial, additional fully Protected Areas for conservation are progressively reduced, the necessity for the more deliberate management of socio-economically important species must continue to increase.
Changes in species composition by the ecological succession of different plant communities on the same area are a common feature in any type of vegetation, as the colonising pioneer species and communities give way to later stages. Local extinctions of pioneer species are a necessary part of this process but provided that the natural or artificial disturbances in the climax formations leave opportunities for cyclical recolonisation by pioneer species in other areas the overall composition of the forest over large areas is not affected. Such natural processes in forest dynamics are generally accepted to occur in the most biologically diverse formations, notably in the tropical rain forests, where the most species-rich areas are likely to be those which include patches of secondary forest in various stages of recovery following disturbance, as well as patches of mature-phase forest (Whitmore 1990).
The goal for genetic resource management is thus to maintain a dynamic system (Namkoong 1986), which may entail the deliberate removal of trees in the later stages of natural succession, as well as the deliberate conservation of some elements of the mature-phase forest. Depending on the exact systems of management, and the degree of understanding of the forest dynamics on which they are based, genetic diversity and specific genetic resources may be enhanced or reduced in specific areas of forest, over specific periods of time. The lack of active management, for example through the complete exclusion of human intervention, may tend to reduce genetic diversity within any given area, although in certain circumstances such (usually temporary) non-intervention may be a conscious management decision aimed at conserving specific genetic resources within the framework of an overall conservation strategy.
Selective logging in mixed tropical forests might, in theory, be managed to maintain an optimal balance between the various stages of ecological succession to allow for maximum genetic diversity and the conservation of the genetic resources of both pioneer and late succession species. This might be achieved either by clear-cutting, at very long intervals, to allow each felled area in turn to revert eventually to the mature condition, or by careful opening of small gaps by the removal of individual trees, or various possible intermediate patterns and levels of harvesting that may favour the advanced regeneration of different species. However this presupposes not only the willingness to subordinate short-term financial gain to long-term ecological objectives but also a degree of understanding of forest composition and dynamics.
In addition to the immediate effect of logging on the advanced regeneration and environmental conditions, such as light, temperature and humidity on the forest floor, all of which may affect the regeneration of different species in different ways, there will be further effects on the density and spacing of populations of the species felled. These changes may influence flowering and fruiting patterns and the mating relationships within the population. In addition to such direct effects there may be impacts on the populations of pollen vectors or seed dispersers of tree species through the removal of “keystone” plant species on which these populations depend.
One often inadvertent and usually severe effect of human intervention which may follow logging is increased susceptibility to fire. While some forest formations are adapted to survive periodic burning, and may be fire disclimax communities with particular qualities of genetic resources of potential value associated with their capabilities to colonise fire-prone areas, the adverse impact of indiscriminate fire in other more complex forest formations can severely reduce the genetic resources of the more valuable tree species. In extreme cases whole populations may be lost through fire following the felling of all adult trees of a species in the area concerned.
Nevertheless with adequate control and based on a sufficient understanding of the ecological processes involved logging and timber extraction can be used to assist the conservation of a wide spectrum of genetic resources of the principal tree species. Both the efficiency with which this can be achieved, and the security against accidental loss of substantial elements of the gene pool, will be dependent on the management of a network of conservation sites in both the production forests and the fully Protected Area systems, extending across the natural range of the principal species.