Why we need gene conservation in tree breeding programs

Genetic diversity is widely recognised as the key component for long-term survival of most tree species. It is a foundation of sustainability because it provides raw material for adaptation, evolution, and survival of species and individuals, especially under changed environmental and disease conditions (e.g. Frankham et al. 2002; Reed and Frankham 2003). Genetic uniformity may leave a species vulnerable to new environmental and biotic challenges (FAO 2006). Long-term conservation of forest genetic resources (FGR) is a cornerstone of sustainable forest management. Through conservation and proper forest management, it is possible to maintain the evolutionary processes of species and the diversity of their gene pools for use now and in the future (Geburek and Konrad 2008). Traditionally, forest conservation has often been understood as a steady-state process, whose goal is to protect natural forests in their original state. Forests, however, are dynamic systems that are now being altered by human intervention as well as natural climatic changes. Human activities have modified forests and their biodiversity for millennia, even in seemingly pristine tropical forests (Kjaer et al. 2004).

In conserving forest genetic resources, we are concerned primarily with conserving genetic diversity within and among populations of a species. Our priorities are normally based on the economic and environmental importance of the species, its ecological functions, the level of risk, or other special features that contribute to the importance of a species. For example, quantitative traits (e.g. survival, resistance to biotic or abiotic stress and growth) are of economic significance and associated with population adaptation to different environments. Since different pictures of population genetic structure are often provided by molecular markers versus adaptive quantitative traits (e.g., Karhu et al. 1996; Merilä and Crnokrak 2001), integration of both kinds of information should be used to develop strategies for the genetic conservation of tree species (e.g., Frankham 2010). However, most studies aimed at formulating conservation strategies have used DNA markers to determine the level of genetic variation and population structure, although variation for such neutral markers may correlate poorly with traits under selection and does not necessarily reflect the evolutionary potential of the species (Reed and Frankham 2003; Leimu et al. 2006). Methods that assess important genetic variation are being developed (e.g., Avise 2010). These may include the application of single nucleotide polymorphisms (SNPs) (e.g., Neale and Savolainen 2004; Namroud et al. 2008). The end result of this is to ensure that functionally useful alleles will be identified and available in the future and can be captured in breeding programs within useful genotypic sets. For example, genes of intermediate frequency are relatively easy to conserve as they are commonly expressed. However, low-frequency genes (rare alleles) are more difficult to conserve even though they may well be needed for the future to meet new selection criteria such as resistance to new pests and diseases (e.g., Eldridge 1978; Burdon 1997; Yanchuk 2001; Gapare et al. 2005).

Gene conservation involves several sequential stages, ranging from the initial selection of target taxa and identification of conservation objectives, through field exploration and germplasm collection, to the actual storage and management of that germplasm over extended time periods. Existing forest resources and the genetic diversity they contain provide a basis for conservation measures. The selection of candidate species is one of the most important phases of the conservation process because future activities will be based on priorities set in the planning phase. As soon as priority setting has been completed, genetic diversity must be assessed and located for subsequent conservation activities. The selection of conservation measures depends on how the objectives of a conservation programme are defined and practical options available, especially in connection with in situ measures. These measures then provide a framework to implement practical conservation work.

Conservation of genetic resources can be accomplished through a variety of in situ and ex situ approaches (e.g., St. Clair and Howe 2011). In situ methods protect plants in their native habitats where they are subject to natural evolutionary processes. Ex situ methods involve storing genetic material in locations such as seed banks, genetic resource plantings (such as provenance and progeny trials and broadly-based gene-resource plantations), and seed- and breeding orchards. There is a general consensus among scientists and practitioners that no single conservation method is adequate, and that different methods should be applied in a complementary manner (e.g., Boffa et al. 2000). In situ gene conservation, however, has a number of benefits and so often forms the basis of conservation programs. It allows evolutionary processes to be maintained, including the adaptation of tree populations to changing environmental conditions. This is particularly important for breeding programs, since future human needs and environmental conditions are difficult to predict. As mentioned earlier, in situ and ex situ gene conservation should be used in a complementary fashion to conserve forest genetic resources. Both are an integral part of the conservation process, and both can be effective only after genetic diversity has been located and conservation priorities have been set. The main purpose of ex situ gene conservation is to capture and maintain a representative sample of the existing genetic diversity of a species. However, there will inevitably be selective pressure that does not match those conserved in situ. Moreover, the downside of any selection for commercial traits will tend to be offset by reducing the opportunity costs of maintaining a given area of gene-resource planting, which will hopefully enable greater genetic diversity to be maintained. And if one can maintain bigger areas of gene-resource plantings, then it is feasible to reduce the pressures of pollen contamination. For highly endangered tree species, ex situ conservation may be the only approach in the short to medium term. The main pitfalls of collecting germplasm samples for ex situ gene conservation are: (1) limited geographic coverage of genetic variation; (2) other biases in the collected plant material; and (3) samples that are too large to deal with (Brown and Hardner 2000). Because ex situ gene conservation plantings are typically (but not always) more costly than in situ gene conservation, it is particularly important that sampling of populations and germplasm within populations is given special attention to maximise the use of limited financial and human resources.

An important requirement in conserving quantitative variation is to maintain its dynamics (Yanchuk 2001). Maintenance of continuing evolution should be primarily emphasised in developing conservation strategy. Conservation ex situ in the forms of seed bank, and tissue bank, called static conservation, decouples from dynamic environments and implies a risk of maladaptation in the future and can easily result in large genetic differentiation between populations maintained ex situ and in situ, especially for loci with large or moderate effects (Hu and Li 2002). Adaptation of material in static conservation would be affected by the conditions in storage and the time elapsed. Pinus radiata seed had viability levels as low as 3 % after 25 years in cold storage (Gapare et al. 2012a). A robust gene conservation strategy combines elements of both approaches and is based on knowledge of the genetic structure of a species and the perceived threat to a species whether from natural disturbance processes, introduced insect and pathogens, or sensitivity to changing climate. Addressing these issues effectively depends on a well-integrated tree breeding strategy to serve the management of genetic resources.

As breeding programs have become more sophisticated, there have been moves towards expanding the genetic bases. For example, the original genetic base of the Australian breeding populations of P. radiata may have been of the order of 200–300 seed trees from a limited part of the natural range of P. radiata (Wu et al. 2007). The genetic base for the land race(s) and traditional breeding populations has been a very incomplete and unbalanced sample of the species’ natural range and hence of its total genetic diversity. In most cases, the genetic base of populations derived from second- and third- generation selections are probably too narrow and there is a need to infuse new germplasm to expand the genetic base.

Because of anticipated adverse climatic change and resulting increases in risk of disease and insect attacks in forest plantations in the future, there is need to maintain broad genetic bases for adaptability and pest resistance. Ironically, the recent influx of tree diseases and insects in plantations around the world and the threat they pose to major commercial tree species has been a catalyst in renewing interest in applied tree conservation programs as sources of new, better-adapted, genetic material (Burdon 2010; Dvorak 2012). For example, in Chile, a new disease (Daño Foliar del Pino’ (DFP)), caused by Phytophthora pinifolia (Pine needle and shoot blight) has severely affected more than 60,000 ha of prime P. radiata plantations in the country (Durán et al. 2008). Phytophthora pluvialis (Red Needle Cast) is an emerging disease threat of P. radiata in New Zealand (Reeser et al. 2013). In Australia and New Zealand, Dothistroma septosporum (Dothistroma needle blight) remains a challenge. The infection has a strong negative impact on growth. The economic impact of Dothistroma septosporum in New Zealand has been estimated at about $24 million per year, including chemical control costs of about $2.5 million per year in addition to direct growth loss costs (Bulman et al. 2004). Essigella californica (Monterey Pine Aphid) on P. radiata forests in New South Wales, Australia has been associated with severe upper crown yellowing and premature needle cast, resulting in some tree death. In South Africa, pitch canker fungus is now established on P. patula and P. greggii in nurseries and rarely (so far) is found on trees in the field (Coutinho et al. 2007; Roux 2007; Mitchell et al. 2011). It is also major problem on P. radiata in the nursery and sometimes on older trees in the field.

Climate change, and the associated shifts and fluctuations in temperature and precipitation patterns, are likely to affect substantial areas of forest plantation production in the tropics and subtropics (e.g., Leibing et al. 2013). Progressive climate change may make the choice of which species and provenance to plant increasingly difficult in native environments. However, in exotic conditions, use of results from provenances trials planted at different locations may guide decisions on which genotypes to plant. This highlights the importance of continued investment in provenance trials establishment. Increases in the frequency, duration, and/or severity of drought and heat stress associated with climate change could fundamentally alter the composition, structure, and biogeography of forests in many regions. There are increases in tree mortality associated with climate-induced physiological stress and interactions with other climate-mediated processes such as insect outbreaks and wildfire. Climate change will require trees to cope with new biotic and abiotic environments and stresses, including: habitat shifting and alteration, fragmentation, drought, temperature extremes, wildfire, and novel insect and disease pressures. The specific effects of climate change will vary greatly over time and space. According to climate change predictions by the intergovernmental panel on climate change, seed material that is well matched under the current climate will have to grow in sub-optimal conditions within the production cycles of the upcoming decades (McKenney et al. 2009). This suggests that forest plantations will need to take anticipated climate conditions into account in order to maintain productivity in future harvest cycles. Therefore, seed material, adapted to the novel climate conditions, needs to be immediately identified and selected (Spittlehouse and Stewart 2003).

Current in situ gene conservation efforts: challenges and opportunities

Geist and Lambin (2002) highlighted that tropical deforestation is still continuing and it is clear that many in situ conservation efforts in the tropics have not been effective. This lack of success can be attributed partly to inadequate participation by various stakeholders in natural resources management and conservation, and tree breeders (Koshy et al. 2002). For the last three decades, researchers and scientists have urged that the forest community be more pro-active in the development of sound gene conservation programs to promote sustainable forestry, which includes maintaining and protecting genetic base populations (e.g. Eldridge 1978, 1997; Dvorak 2012). However, there are differences on issues such as strategy and approaches (ex situ vs. in situ), focus on common versus low-frequency alleles, the cost/benefits of long-term programs to maintain gene pools, and the justification for protecting commercially important tree species represented by large areas of exotic plantations (Eldridge 1978; Dvorak 2012). Some have argued that rare alleles arise largely as unfavourable mutations, are merely evolutionary relics and contribute little to the overall fitness value (e.g., Brown 1989), whereas others (e.g., Gapare et al. 2005) suggested that rare alleles can be very important for long-term evolution. For example, recessive alleles for resistance to fungal pathogens are a very interesting case in point. Where they work, they tend to be much less prone to being ‘defeated’ by the pathogen mutation. If rare alleles become a target for future breeding objectives or for survival of populations, the possibility of relatedness among selected genotypes may not allow their effective use due to an excessive build-up of inbreeding. Burdon (1997) postulated that, with a mixed mating system for pines, the rare alleles can be expressed reasonably readily and may have played a role in some conifers flipping over into becoming selfing species. It is therefore desirable to raise the probability of maintaining rare alleles in sufficient unrelated genotypes for future use in breeding populations. Dvorak (2012) also argued that as forest diseases become more prominent in plantation forestry and our understanding improves about the gene action that controls resistance to tree diseases, we might need to rethink the value of capturing rare alleles for use in breeding programs. As an example, resistance to fusiform rust (Cronartium quercuum f. sp. fusiforme), the most commercially damaging disease that affects P. taeda in the southern USA, costs the industry million of dollars in annual losses (Schmidt 1998) and is probably controlled by rare alleles (Dvorak 2012). Eight genes/alleles have been identified at the present time, but probably more exist. It is unclear at this time what kind of alleles control resistance mechanisms in trees to pitch canker (Dvorak 2012).

Some challenges

Threats to the genetic diversity of forest trees are manifold, and for the main part directly or indirectly anthropogenic (Geburek and Konrad 2008). Changing land use (e.g., conversion of forests for agriculture) and logging increase fragmentation and isolation of tree populations in the tropics. Inevitably, fragmentation of populations results in the loss of genetic diversity through genetic bottlenecks, and increased inbreeding eventually leads to negative long-term effects on the tree population (Ezard and Travis 2006). For example, in hardwood species, associated losses of the populations of bat species that are efficient pollinators between widely scattered individuals can be a big factor in loss of genetic diversity. New pests and pathogens, which may have been introduced or are newly evolving, inevitably seem to be on the increase as a result of the global movement of goods, further challenging the adaptive potential of an increasing number of tree species (Brasier 2001; Burley 2001; Scherm and Coakley 2003; Goss et al. 2011). For example, there are several threats to the Mexican and southern pine species in their native habitats that make in situ conservation a challenge. Suffering diseases, bark beetle attacks and excessive fires describe the situation for most pine species in Mesoamerica. For example, some P. tecunumanii and P. maximinoi stands in their native habitats are under threat primarily from agricultural expansion but are also threatened by fires and insects, notably Dendroctonus sp. As such, conservation status for P. tecunumanii, P. maximinoi is now critically endangered and endangered, respectively (Camcore 2008). Both these species have proved to be alternative species for commercial forestry in the tropics, in particular in Southern Africa. P. tecunumanii is being used as a hybrid parent with P. patula in South Africa, highlighting the need for ex situ conservation of the species (Camcore 2010).

Monterey and Cambria populations of P. radiata are severely being encroached upon by urban development and exotic invasive organisms in California. P. radiata stands on Guadalupe Island have been severely impacted by grazing from feral goats, with fewer than 220 mature trees left based on a 2001 census (Rogers 2004), although there may have been some recovery since then (A.C. Matheson, pers. comm.). However, there are efforts to purchase land to protect the remaining radiata pine forests from degradation or conversion to other uses.

Because of the diseases and pests now present in native habitats, importation of seed and other material from any pine species is now subject to stringent quarantine controls in some countries. For example, there is currently a moratorium on importing P. radiata germplasm from native stands into Australia and New Zealand. Consequently, the genetic material of pine species currently in Australasia cannot readily be augmented by further importations from overseas. Therefore, conservation of genetic material currently in Australasia is likely to be important to support the long-term viability of plantation forestry using pine species in Australasia. Cook and Matheson (2008) estimate that delaying the entry and spread of pitch canker in Australia by as little as two or three years is worth nearly $13 million to the Australian forest industries. As a consequence of the extreme risk posed by pitch canker to the Australian forest industries, importation of seed is prohibited. Resistance to fungal diseases has been an important breeding objective in some tree breeding programs (Burdon and Wilcox 2007). Because fungal diseases can cause biotic crises, they can impose new and crucial breeding objectives that can place extreme demands on breeding populations (Burdon 2010; Ivković et al. 2010).

Maintaining unimproved gene-resource stands is expensive, and likely to become more so as progressive genetic gain increases the opportunity costs of such material. For example, the real significance of the boost in genetic diversity from broadening the provenance base from unimproved gene-resource stands may lie in recruiting genetic resistance to various biotic factors and abiotic factors (e.g., Gapare et al. 2011). In some cases, the unimproved gene resources, while promising low costs and the benefits of natural selection, are, if ex situ, highly vulnerable to pollen contamination which may badly affect their value as sources of new unrelated germplasm. Some tree-improvement programs have started with too narrow a genetic base, thus decreasing the value of their conservation efforts. There is now a realisation that for ex situ conservation plantings to succeed, they must hold some economic or aesthetic importance to the receptor institution (Dvorak 2012). However, there appear to be little information on the valuation of forest genetic resources. Laarman and Dvorak (1988) investigated the financial feasibility of participation in an international seed cooperative for tropical pines. They reported that it is financially viable at annual planting levels of 100–500 ha if the opportunity cost of capital is 8 %. A literature search revealed that valuation of genetic resources up to now has only been done in one woody species, namely coffee (Coffea arabica) (Hein and Gatzweiler 2006; Shumeta et al. 2012). The monetary value of coffee genetic resources was determined by calculation of the productivity gain through hypothetical breeding programs for increased yield and pathogen resistance. Geburek and Konrad (2008) suggested that a similar approach could be tried for forest tree species, such as in the simplified form of modelling the additional monetary profit gained by the use of superior clones or offspring originating from phenotypically superior (plus) trees. In such an approach, the profit could be transferred to the whole forest area from which valuable forest genetic resources (FGR) have been selected and used to indicate the value of the genetic information located there. Such valuation would provide decision makers with valid arguments that would allow them to defend conservation programs against budget cuts.

Evidently, the questions of requisite population sizes and appropriate management systems have become increasingly pressing. Setting the sizes of breeding populations and gene resources may be based on the size of the breeding program and may depend on: (1) acceptable probabilities of retaining alleles present at certain frequencies; (2) threshold frequencies at which alleles are worth trying to conserve; (3) expected ongoing and long-term utilisation of breeding populations; (4) acceptable costs (Burdon and Wilcox 2007). While breeding populations may contain a few hundred individuals, and gene resources a few thousand (e.g. White 2004), such numbers may be interdependent, since the appropriate size of a breeding population, for instance, may depend on what is available or can be maintained as gene resources.

Resources allocated for domestication of new species and gene conservation continue to decline. For example, there are several species with commercial potential that have been assembled by Camcore (an International Tree Breeding and Conservation Program) at North Carolina State University, USA, but they have not been domesticated (Table 1 in Dvorak 2012). Commercial forest land is expensive and use of any piece of land is expected to generate an income. A key challenge for managing ex situ populations is meeting the opportunity cost of using land for conservation. Management of ex situ gene resource stands, while crucial to long-term genetic gain, does not directly earn revenue for commercial breeding programs. It is therefore often more resource-limited than activities that generate revenue directly (e.g. seed-orchard production, mass clonal propagation) (Burdon and Wilcox 2007). Meanwhile, tree breeders must often deal with shareholders who seek guaranteed returns from the direct costs and opportunity costs involved in gene conservation activities. Better knowledge will therefore be extremely welcome, especially as the opportunity costs are being increased by genetic gain already achieved (Burdon and Wilcox 2007). A further challenge is of a more organisational nature: forestry is a long-term business, at least for the pines, with rotation ages around 25–30 years for sawlogs. Other deeper structural challenges exist within the industry. For example, on the one hand, it is indeed long-term, yet it tends to be very cyclical, with alternating shortages and over-capacity of very capital-intensive processing plants that has long lead times to come on stream. Forestry companies are changing ownership increasingly frequently, there is lack of continuity in management, and thus conservation plantings often do not get protected.

Some opportunities

Even though provenance/progeny trials were not necessarily set up with conservation in mind, they can be used for the purpose, provided that future propagation from them is done with careful selection of unrelated individuals. More recently, international multi-site provenance trials, ones that measure growth performances of different seed sources in a range of different climates, are being used to address climate change issues (for example: re-analyse spatial patterns of genetic variation in relation to predicted future climates; validate species-suitability models; and monitor germination, growth, phenology, and resistance to insects and diseases) (e.g., Leibing et al. 2009; van Zonneveld et al. 2009; Wang et al. 2010). These trials provide a wealth of information on the adaptive capacity of species and provenances that may facilitate selection of seed material for changing climatic conditions (Leibing et al. 2013). It is imperative, therefore, for tree-breeding programs to take pre-emptive action to protect genetic diversity for current and future generations, for a wide range of commercial species.

The provenance/progeny trials also offer some opportunities to examine trends of population genetic structures in time and space through new DNA marker technology (single nucleotide polymorphisms (SNP). The use of SNPs has much promise for characterising base populations and undomesticated pine species, helping to interpret the results of evaluation trials, and monitoring the maintenance of the genetic bases in future breeding. It may also help in making policy decisions on the size of the genetic resources to be used for breeding and in the actual selection (e.g., Burdon 1995). For example, more than 2000 SNPs responsible for corewood formation was discovered using a 25-year-old P. radiata provenance trial established in Batlow, New South Wales, Australia in 1980 (Dillon et al. 2010). Other studies have also utilised provenance-trial material to identify significant associations between SNP markers in P. taeda and wood properties in relation to carbon-isotope discrimination, height growth, or resistance to a fungal pathogen, Fusarium circinatum (González-Martínez et al. 2007, 2008; Eckert et al. 2010; Cumbie et al. 2011). Some members of Camcore have received more than 20 pine species and established provenance/progeny trials. Such a resource offers a great opportunity to identify candidate genes for growth and disease traits.

Most Mexican and Central American pine provenance/progeny trial studies have revealed significant amounts of genetic variation. For example, provenance differences for DBH growth between the worst and best provenances of species like P. tecunumanii and P. maximinoi are as much as 30 % (Hodge and Dvorak 2012). Similarly, great differences are also found in adaptability. For example, sources of P. leiophylla and P. herrerae from northern Mexico are more cold-hardy than southern sources when planted in South Africa (Dvorak 2012). These materials offer opportunities to broaden the genetic bases of most tree breeding programs. Broadening the genetic bases of breeding populations favours the retention of new alleles and pre-existing but rare ones (Burdon 1997; Otto and Whitlock 1997). In some cases, historical introductions have often been haphazard, and may not represent the best-adapted provenances even though the existing material may perform well. For example, using isozymes and cortical oleoresin monoterpenes, the present Australian and New Zealand plantations of P. radiata have been shown to be from Año Nuevo and Monterey with none from Cambria (Moran and Bell 1987; Burdon 1992). Infusion of Cambria material is bound to increase the overall genetic base of the radiata pine breeding programs with potential long-term benefits (Gapare et al. 2011), even though the adaptive advantages of pure Cambria provenance are often outweighed by adaptive problems. Similarly, P. patula introduced in Zimbabwe and South Africa in the 1920s to 1950s was reported to have come from a narrow base (Barrett and Mullin 1968) and later collections made by Camcore in the mid-1990s could broaden the base of the species (Nyoka et al. 1996).

Through genetic infusion, genes are introduced from external sources into a breeding population at advanced-generation level. Such infusion makes possible not only the maintenance of genetic diversity but also potential increase of genetic gain (Gapare et al. 2011). However, the greatest challenges are to develop practical approaches and overcome institutional inertia for infusing genes from base populations of unimproved material into more advanced-generation breeding programs without greatly affecting productivity gains (Gapare et al. 2012a). While genetic diversity is desirable in itself, maximising diversity is liable to include material that is markedly suboptimal, in respect of adaptation and/or quality traits, so there is no simple formula for an overall optimal genetic base. The real significance of the boost in genetic diversity from broadening the genetic bases may lie in recruiting genetic resistance to various biotic and abiotic factors and/or in crucial provenance differences in wood properties coming to light (Gapare et al. 2012b). For example, hybrids between P. patula and P. tecunumanii are being used to reduce the susceptibility of young P. patula seedlings to Fusarium circinatum (Nel et al. 2006; Kanzler et al. 2012).

The germplasm assembled by Camcore provides a basket of options—various species/populations to choose from in order to meet the challenges and demands of the future. This is a key avenue for ensuring sustainability of the forestry industry in a changing world. Only rarely are ex situ and in situ conservation used in a complementary manner as is done by Camcore, as part of a holistic approach to conserving important species. Moreover, the few complementary efforts that do exist are seldom based on a well-documented knowledge of the genetic diversity of a given species within a country, much less an entire region. More studies are needed on the distribution of genetic diversity in material assembled by Camcore, and how effective current in situ conservation efforts are at safeguarding this diversity.

Lastly, forest geneticists have to find a common language to convey their ideas and to become more effective in the decision-making process. Successful conservation programs for forest genetic diversity require the development of clear objectives. As a first step, a strict distinction between the protection of genetic resources for utilitarian and ethical conservation purposes needs to be made; a point that often fails to emerge clearly in discussions and that has probably also had a major influence on the effectiveness of gene resource conservation activities (e.g., Geburek and Konrad 2008). Because forest genetics often has its roots in economically-oriented provenance research, its context tends to focus on tree breeding rather than conservation. Therefore, it needs to be clarified whether or to what extent (1) a genetic resource (forest area) to be conserved has been chosen purely for a commercial purpose (as is the traditional way of selecting superior-quality stands or trees and thus serving immediate breeding purposes rather than conservation per se) or (2) whether a resource has been protected based on its (assumed) high genetic variability or genetic uniqueness (Garnier-Géré and Ades 2001). In the future all parties (e.g., foresters, geneticists, conservationists) must try harder to promote forest genetic resource conservation through collaboration. Instead, concerted gene conservation efforts between forest geneticists and conservation agencies are desperately needed.

Overall, goals such as conserving adaptive variation and adaptability are appreciated by everyone, but it is nearly impossible to map actions to this goal. Forest geneticists tend to avoid giving clear precedence to specific action, perhaps because they believe that all activities are equally important and will somehow help to reach the overall goal. Nevertheless, the consequence of this failure to assign weights to actions is that the decisions are often deferred and probably made later through a different rationale. Decision makers need smaller steps that they can “sell” or “push,” and to do so they rely on expert advice (e.g., McNeely and Vorhies 2000). Therefore assigning different weights to actions must be done wisely and responsibly, but it must be done. Forest geneticists should not be afraid to make mistakes. In fact, failures are almost inevitable if decisions are by default made later by those lacking expertise (e.g., Failing and Gregory 2003).

Radiata pine gene conservation and utilisation: a case study

The introduction, genetic base and state of P. radiata breeding populations in Australia and New Zealand

Pinus radiata was introduced into Australia and New Zealand, via England, in the mid 1800s (Wu et al. 2007; Johnson et al. 2008). Initially the species was used for ornamental purposes in botanic gardens. Further seed was imported directly from California to New Zealand in the 1870s with the largest consignment of 11 kg in 1876 probably originating from a large number of trees (Wu et al. 2007). From 1859 into the mid-1860s importations were made basically on a specimen-tree scale, from England (Wu et al. 2007; Burdon et al. 2008). The species thrived so well that by the late 1860s larger-scale importations had begun from California, and are known to have continued until 1882, by when New Zealand plantings could evidently meet the considerable demand for seed (Raymond and Henson 2009). However, records of the exact origin and breadth of the original genetic base have never been found. From the early 1900s seed from these early plantings was used to establish increasing areas of radiata plantation in southern Australia and New Zealand (Moran and Bell 1987; Johnson et al. 1997, 2008).

Small scale provenance research was embarked on in Australia and New Zealand from the late 1930s through into the 1950s, and there were small experimental plantings of island material in Australia. This, however, was done with no real commitment to provenance research, because of a prevailing belief that provenance would not be a real issue with radiata pine. A key change in thinking emerged after work published by Bannister (1959) that suggested that the entire geographic range of the species needed to be embraced, and should include both the Guadalupe and Cedros populations. During the 1960s two major collections were carried out—one by Forde was around 1962 and the second one by Dr Reid Moran on Guadalupe Island, supplemented by an expedition by him and Libby to Guadalupe and Cedros Islands in early 1964 (Forde 1964; Hood and Libby 1980). A Genetic Survey provenance/progeny experiment was subsequently established using seed arising from Forde’s cone sampling, a seed collection made in Guadalupe by Dr Reid Moran, incorporated two New Zealand landrace population samples as controls, and Cedros Island from the 1964 Libby collection.

Results from the Genetic Survey showed that New Zealand land-race material was very predominantly, if not entirely from Año Nuevo and Monterey. However, the isozyme data gave a pretty weak signal, in respect of the Año Nuevo/Monterey ratio. Isozyme studies also suggested that Australian radiata pine originated from Monterey and Año Nuevo, the two best adapted of the five natural populations (Moran and Bell 1987; Burdon 1992). The isozyme work also showed that the first generation of existing breeding programs had captured a substantial proportion of the genetic diversity (in terms of allelic representation) in Año Nuevo and Monterey provenances (Moran and Bell 1987). However, the three southern provenances (Cambria, Guadalupe and Cedros) did not appear to be represented in breeding programs. A combination of these early results from the Genetic Survey, monoterpene studies (Burdon et al. 1997a), some step-out provenance trials using spare stock and evidence that edaphic adaptation was an important issue, led to the joint New Zealand/Australia ‘Eldridge’ expedition. As breeding programs became more sophisticated in terms of pedigree, thought was also given to expanding the genetic base by re-collecting in the original stands in California, and this was also a contributing factor to the joint New Zealand/Australia expedition.

A comprehensive collection was carried out in 1978 (referred in this paper as the 1978 ‘Eldridge collection’). The collection was done as a collaboration between New Zealand and Australia with joint funding from United States of America/Australia Cooperative Science Program, CSIRO Division of Forest Research, New Zealand Forest Research Institute, and Forest Resources Division of Food and Agricultural Organisation. Seed was collected from a total of 621 mother trees, at least 22 trees from each native population and the trees were selected for having at least 40 collectable cones and being well grown and of good form (Eldridge 1978) (see Fig. 1; Table 1). Camcore and the Forest Genetic Center, Mexico made a further collection from 176 trees in all populations except Cedros in 1991 (Rogers 2002). A collection targeting the island populations (Guadalupe and Cedros) was carried out in 2001 by a multinational group of 14 scientists and conservationists, organised by Dr. Rodgers of the University of California, Davis Genetic Resources Conservation Program (Rogers 2002). The collection was in collaboration with scientists from the Colegio de Postgraduados in Montecillo, Mexico and CSIRO in Canberra, Australia.

Fig. 1
figure 1

Location of Pinus radiata populations: Año Nuevo, Monterey and Cambria native to California and Guadalupe and Cedros Islands in Mexico. Adapted from Ledig et al. (1998)

Full size image
Table 1 Seed-origin information for Pinus radiata populations sampled in California in 1978 (from Eldridge 1978)
Full size table

The 1978 ‘Eldridge collection’ comprised open-pollinated (OP) seed from individual trees from the Año Nuevo, Monterey, Cambria, Guadalupe and Cedros populations (Fig. 1). The 1978 ‘Eldridge collection’ was used to establish provenance trials and large-block gene-resource plantings between 1979 and 1982 at many locations in Australia, New Zealand, South Africa and Chile. The Australian plantings included provenance progeny trials, whereas New Zealand only did provenance trials (by ecological subdivisions) of mainland material (except for a single-site planting of Guadalupe families). The shorter-term objective was to evaluate the adaptive and commercial characteristics of the provenance material. The longer-term objective was to provide material for infusion into breeding populations (while maintaining largely unimproved gene resources), meeting that objective being informed by the results of meeting the shorter-term one.

Findings from 23 different sites in New Zealand involving just the Californian mainland provenances showed different profiles of site adaptation among provenances (Burdon et al. 1997b, 1998). For example, Año Nuevo and Monterey are the best adapted overall to New Zealand conditions, but Año Nuevo is much less adapted to P-deficient clays, yet better adapted to cold, snow-hazard sites. Cambria is susceptible to needle-casts, shoot dieback, and frost and snow damage, but has tolerated poor soils. In Australia, Monterey was usually the best of the natural populations, but Año Nuevo grew better on the highest productivity sites and Cambria grew well on some low productivity sites in New South Wales (Johnson et al. 1997). In five South African trials (Falkenhagen 1991), Monterey was overall the best of the three mainland populations at age 8 years, being significantly superior to Año Nuevo at three sites and Cambria at one site out of five. At age 6 years in three trials in Chile, Monterey was superior for volume, although not significantly different from the other mainland populations (Jayawichrama and Balocchi 1993). Monterey also ranked highest for growth traits in one trial in Turkey at age 6 years (Toplu et al. 1987). In Turkey, Año Nuevo population was slightly inferior overall to Monterey in height, and inferior to Monterey and Cambria in mean diameter and stand volume (Toplu et al. 1987). It was also slightly poorer than the other mainland populations for mean tree volume in Chile (Jayawichrama and Balocchi 1993). The fastest growing populations in Greece were Año Nuevo and Monterey while Cedros was completely unadapted to conditions in Greece (Matziris 1995).

Interpopulation hybrids including Cambria have grown well in comparison with pure Monterey population trees in a trial planted in 1973 on a fertile site in New South Wales (K.G.Eldridge unpubl. data). Such hybrids may combine better adaptation to certain environments with higher general growth rates, and could prove very useful in a generally adapted breed of P. radiata. Guadalupe population had promising performance at a low altitude site that represents low-rainfall area that is close to the hot and dry limit and soil that is sub-optimal for radiata pine (Boardman and McGuire 1997). A comparison study of native-population and New Zealand land-race samples of P. radiata using cortical oleoresin monoterpenes confirmed strong differences between Monterey and Cambria and also showed that local land race samples were evidently derived from Año Nuevo and Monterey (Burdon 1997).

The details of the base-population field trials and ex situ gene conservation trials where pedigree information is available are provided in Table 2 (Australia-based trials), Table 3 (New Zealand-based trials). The near-rotation ages of most provenance trials and conservation-block plantings, and a gradual loss of plantings at more than 20 sites highlighted the need to implement a strategy for re-conservation of radiata pine. The rest of the trials have reached rotation age and are due for clear-felling by 2015. The range-wide P. radiata 1978 ‘Eldridge collection’ also provided an opportunity for boosting the genetic base of the existing Australian and New Zealand breeding populations (Eldridge 1997). To ensure genetic sustainability and further increase genetic gain for P. radiata, infusion of new genetic material from the range-wide P. radiata 1978 ‘Eldridge collection’ was recommended (Burdon 1992; Eldridge 1997; Wu et al. 2007).

Table 2 Details of current base provenance trials and ex situ conservation trials (family identities records maintained) in Australia
Full size table
Table 3 Details of current base provenance trials (no family identity records) planted in New Zealand, mainland provenances only
Full size table

In order to explore what potential benefits infusions of native-population material into breeding programmes are likely to confer, initial characterisation of quantitative genetic variation of the native provenances would be critical in assessing the size and usefulness of genetic variation and for directing efficient selection of the genetic material. In addition, the ‘Eldridge collection’ germplasm had not yet been utilised in breeding programs. Infusion of conserved ‘wild’ genes from natural populations of radiata pine was seen as providing a new opportunity for improvement of Australian and New Zealand P. radiata. However, such expansion of genetic diversity will have a cost as germplasm in the breeding programs began at an estimated 2–5 generations are between three (Australia) and five (New Zealand) generations removed from the wild, and would be up to 2–3 additional generations thereafter and have been genetically improved for adaptation to their exotic environments and industrial objectives.

The case for allocating resources and effort for ex situ gene conservation of this important forest genetic resource is strong from both an international and national perspective, including documented threats to the native populations Año Nuevo, Monterey and Cambria, on the Californian mainland coast and Guadalupe and Cedros Islands, off the coast of Baja California. From a commercial forest industry perspective, this importance is due to the fact that: (1) the P. radiata resource underpins a significant industry; (2) the breeding programs in Australia and New Zealand appear to have a relatively restricted genetic base; and (3) opportunities to expand the genetic base through direct importation of native germplasm is now limited due to the risk of introducing pitch canker disease which is impacting native stands in California. The need for expanding the genetic base of the breeding populations is mainly argued on the grounds of needs to: (1) maintain genetic diversity for coping with climate change (including growth on drier sites) and conferring resistance pests or diseases; (2) make good any reduction in genetic variance in the current population; and (3) enhance opportunities of adaptation to changing market demands.

The project was carried out under a joint funding scheme by Forest and Wood Products Australia (FWPA), Southern Tree Breeding Association (STBA), Radiata Pine Breeding Company (RPBC), Scion (New Zealand Forest Research Institute Ltd) and Forests New South Wales (FNSW)). The objectives of the project were to:

  1. 1.

    Provide for P. radiata gene conservation for the future

  2. 2.

    Infuse new genotypes into breeding populations for increased genetic gain (especially increased growth and wood stiffness) and increase the effective population size.

Conservation

The project has made significant progress in the management, conservation and utilisation of native radiata pine germplasm. A catalogue of existing provenance and genetic conservation trials of P. radiata in Australia provides forest managers with a one-stop-shop that lists details of trial establishment, location, owners and collaborators of the trials, and proposed schedule of measures recommended for each trial, including possible prolonging of rotations and collecting seed at time of harvesting the trees. Large amounts of seed from all five provenances were collected from ex situ provenance large block plantings and are being kept in cold storage for long-term future use, while some genetic material is being used to establish second-generation gene-conservation plantings at different sites in Australia and New Zealand to ensure continued evolution and long-term genetic security. However, these populations are likely to diverge from the native stands. The conservation plantings at different locations in Australia and New Zealand offer the most secure gene conservation of all five populations. Pollen collected from multiple trees across all provenances will be used for controlled crosses in order to increase genetic diversity. The strategy should conserve not only existing genetic diversity, but enhance the genetic base needed to adapt to changing environments and market demands.

Utilisation

A comprehensive and large-scale genetic analysis of growth-, tree-form and wood-quality traits derived from the 1978 ‘Eldridge collection’ trials showed that was abundant genetic variation in these populations and provide an opportunity for breeding programs to select superior trees for solid-wood products that will combine superior growth with desirable wood traits (Gapare et al. 2011, 2012a, b). For example, progeny of 10 unrelated parents ranked above control seed lots from the older open-pollinated seed-orchard stock for DBH growth in New Zealand, and would be potential candidates for infusion. This result fitted very well with the RPBC breeding strategy which recommended infusion of 10 unrelated parents from the native population selections from the 1978 ‘Eldridge’ collection (Dungey et al. 2009). The selected parents are now being progeny-tested as part of 500 selections in the breeding-population trials. The 10 parents comprised six unrelated families from Monterey, one from Cambria and one from Año Nuevo, which all ranked above a control seedlot rated at GF14 (represents Growth and Form of older open-pollinated seed orchard stock) (see Fig 2 in Gapare et al. 2011). Two additional families from Monterey ranked just below GF14. The best individual trees from those 10 unrelated families would be potential candidates for infusion in line with the current breeding strategy.

The analysis of data on growth- and wood-quality traits for the Australia trials showed significant differences in wood properties among provenances, families and/or individual trees. This provided an opportunity for breeding programs to select superior trees for solid-wood products that will combine superior growth with desirable wood traits. The results from these analyses, including trial-specific estimates of additive genetic, and error variances, and between-trait correlations, were entered into DATAPLAN® (Powell et al. 2004) for TREEPLAN® runs by STBA to estimate individual-tree breeding values (McRae et al. 2003). The estimated breeding values were used to select elite trees for grafting into seed orchards. A total of 210 elite trees representing all five provenances of P. radiata were selected for possible infusion into existing breeding populations in Australia. The scions collected from the elite trees were grafted into the national genetic resource centre arboreta in South Australia for conservation and use in a controlled crossing program in order to increase genetic diversity. The infusion of the selected germplasm would extend the future responses to selection by contributing favourable genes not present in the existing breeding populations. For example, unique genes in the Guadalupe, Cedros and Cambria populations which are currently not in the breeding populations will now be included in the breeding populations and enable genetic adaptation to projected future environments (e.g. adaptation to climate change, pests and disease resistance).