Josefa Fernández-López¹ and Ricardo Alía²

¹Centro de Investigaciones Forestales de Lourizán, Pontevedra, Spain
²Dpto. de Genética y Biotecnología, CIFOR-INIA, Madrid, Spain

Introduction

The overall objective of the conservation of wild chestnut forest genetic resources is the creation of good conditions for the future evolution of populations and the preservation of the present genetic constitution for conservation and genetic improvement purposes.

Local grafted varieties for nut production also represent an important chestnut genetic resource; their importance has been recently highlighted by the fact that traditional varieties are being abandoned and are in danger of disappearing. Thus, while conservation for fruit production is not the focus of the work carried out within the framework of the activities of the EUFORGEN Network, the inclusion of fruit-producing varieties in an overall conservation strategy will be important. Links need to be established between the sectors of forestry and horticulture to further genetic conservation of this multipurpose species for mutual benefit.

Castanea sativa is a Mediterranean species. The range of distribution extends from the Caspian to the Atlantic, including Madeira, Azores and Canary Islands, from 51º latitude in southwest Germany and south England, to 37º latitude in Tunisia (Mounts Tlecem). The species is found in north-facing slopes where the rainfall is greater than 600 mm, on moderately acid soils (pH 4.5-6.5) with a light texture. Chestnut needs high air humidity and avoids late frosts, sprouting very late. Chestnut roots are very susceptible to ink disease, caused by several species of Phytophthora, a disease which is very important in humid areas. It is also susceptible to Chryphonectria parasitica, which causes severe damages in some populations. is a Mediterranean species. The range of distribution extends from the Caspian to the Atlantic, including Madeira, Azores and Canary Islands, from 51º latitude in southwest Germany and south England, to 37º latitude in Tunisia (Mounts Tlecem). The species is found in north-facing slopes where the rainfall is greater than 600 mm, on moderately acid soils (pH 4.5-6.5) with a light texture. Chestnut needs high air humidity and avoids late frosts, sprouting very late. Chestnut roots are very susceptible to ink disease, caused by several species of Phytophthora, a disease which is very important in humid areas. It is also susceptible to Chryphonectria parasitica, which causes severe damages in some populations.

Chestnut is an important species, currently occupying more than 1 700 000 ha in southern Europe. It covers important areas in France, Italy, Spain, Portugal, Turkey, the United Kingdom and Greece. It is found mainly as cultivated varieties in grafted orchards to produce nuts, and in coppices to produce small pieces of wood. High forest, to produce high-quality timber, is very scarce but the area is increasing. The species has a discontinuous range, occupying hundreds of hectares on acid soils. In mixed broadleaved forests stands (i.e. with Quercus robur as the main species), chestnut could not be considered as a ‘social’ species, and in some cases it is rare, comparable to Acer pseudoplatanus or Prunus avium.

The natural or autochthonous origin of C. sativa in Europe has been discussed in numerous papers. Some authors assume that chestnut disappeared from southern Europe during the Würm glaciation, surviving only in northeast Turkey and in the Caucasus. From these two areas chestnut was introduced as a cultivated species by the Romans. According to this theory, populations existing in western Europe have a very restricted and unknown origin (either from Turkey or Caucasus or both), and its presence dates back less than 2000 years. Isoenzyme studies have shown that total genetic variability is much higher in East Turkey than in West Turkey, Italy and France (Villani et al. 1991a, 1991b; Manchon et al. 1996). These authors consider it as a proof of the migration of chestnut into western Europe.

Other authors, however, assume that chestnut remained in several refugia during the glaciations in southern Europe. Several fossils and remains of pollen have been found in the present area of distribution of the species. The species was present at least 2500 to 3000 years BP (Pitte 1986; Aira-Rodriguez and Ramil-Rego 1995), i.e. before Roman times. Thus, present populations in western Europe are the result of a mixture of native populations with the descendants of cultivated, grafted varieties. For example, in Spain, the continuity of palynological records in the Quaternary sediments precludes the introduction of the material by the Romans as a likely hypothesis (Gómez-Manzaneque 1997). Moreover, isoenzyme studies of material of Spanish origin (varieties and populations) have shown a reduction in allele richness compared with Turkish populations. In general, the number of alleles per locus is higher in Spanish origins than in the Italian and French (Fernández-López 1996; Pereira-Lorenzo et al. 1996b). In this sense Camus (1929) pointed out that chestnut is not an autochthonous species in France but it could be native in the northwestern area of Spain.

Chestnut is characterized by the diverse products that can be obtained. One of the most important is the fruit, of great importance in pastries and in human alimentation in the past. For this reason the distribution area of the species was increased. Many varieties have been described and propagated by grafting, most of them for nut production, some for wood or both (Breviglieri 1955; Bergonoux et al. 1978; Borghetti et al. 1983; Gomes et al. 1993; Pereira et al. 1996a; Frank and Radocz 1998; Solar et al. 1998). Although nut prices are very good, orchards are being abandoned because of rural depopulation.

Chestnut timber is straight-grained, closely resembles oak in colour and texture, and is highly valued, with high prices for big trees free of defects. The traditional use of the species for small pieces of wood, obtained from coppices, was in general associated with vineyards, and presently rotation periods are increasing, as is regeneration from seeds.

An important use of the species is landscaping, mainly as mixed forest, with some other broadleaves such as Quercus spp., Fraxinus spp., Betula spp., etc. In these types of forests, chestnut provides food for many game animals.

The extension of the ink disease resulted in the introduction of the resistant species Castanea crenata and C. mollissima and further in the use of interspecific hybrids to be used as rootstocks or clonal varieties for wood or nut production.

The overall goal of gene conservation in chestnut within the framework of the Noble Hardwoods Network can be divided into four clearly connected objectives.

1. Conservation of the present genetic constitution in clonal archives of plus trees or grafted fruit varieties, seed orchards and progeny tests. The plus trees are selected in seed populations and some grafted varieties of interest for wood production can be included. This conservation is the base for future breeding programmes of the species.

2. Conservation of the present genetic constitution of populations endangered by Chryphonectria parasitica or Phytophthora spp.

3. Conservation of the species creating good conditions for the future evolution of populations. This third objective would incorporate several populations of the previously described steps.

4. Conservation of the present genetic constitution in clonal archives of grafted fruit varieties.

To reach these objectives, different studies need to be carried out, mainly on the genetic structure of populations, but these are not further described in this paper.

The conservation of chestnut fruit varieties has already been considered by IPGRI and FAO. European chestnut collections are included in the Directory of European Institutions Holding Crop Genetic Resources Collections (Frison and Serwinski 1995).

Genetic knowledge

The multiple use of the species has determined the main genetic characteristics and variability of natural populations. The type and origin of the forest can be divided into the following types, ordered by importance of the resource:

1. Coppices, regenerated by stump shoots after each felling and with different levels of seed regeneration, depending on the density of the stands. Coppicing has been a traditional silvicultural system in many areas for hundreds of years. Depending on the origin of the coppice, cultivated varieties or natural stands, the genetic structure of the populations can be different.

2. Grafted orchards, obtained from a few high-quality wood- and nut-producing genotypes. Grafting has been a method of propagation for several centuries. In this case the level of genetic diversity has clearly changed. The preservation of the most important grafted varieties is the main concern. The type of grafted chestnut for wood production is actually very scarce. The chestnut orchards grafted with fruit varieties are a very important resource to be considered.

3. New natural populations regenerated from seeds: mostly from a few clones, where grafted orchards have been abandoned, a very common event in the chestnut areas; regeneration from seeds transported by birds under plantations of other species such as Pinus pinaster or Eucalyptus globulus. Genetic drift could be an important evolutionary force in this type of forest.

4. Mixed broadleaved forests, with chestnut as a rare species. These forests are perhaps those in which the genetic diversity has been maintained without human influence. Geneflow among individual trees and inbreeding are not known in this type of forest, but the situation is similar to the hardwoods concept used in the Network. This type of forest is represented by important marginal stands of the species.

5. Plantations, frequently from seeds coming from orchards or unknown provenances. Seed stands selection was made in France (Bilger 1998).

6. Pollen contamination from only one or a few C. sativa grafted genotypes, and of C. crenata or C. crenata × C. sativa hybrids, used in North Spain and France, may have affected the genetic structure of the stands.

7. Incidence of ink and canker diseases could be a cause of genetic drift in certain areas.

In conclusion, in the types of forests listed above, the mixed broadleaved forest, seed stands and grafted fruit varieties could be considered as the most important for the genetic conservation of chestnut.

Mating system
Chestnut is mainly an anemophilous species but it is also pollinated by insects. Some biological features force cross-pollination, as male sterility, dichogamy and an important degree of self-incompatibility. Therefore, high levels of heterozygosity and polymorphism have been found in isoenzyme studies.

Geneflow
The weight of chestnut fruits reduces seed dispersion, but dispersion by birds is highly effective, estimated to several hundred metres (Kollman and Schill 1996). The size of chestnut pollen (14-18 microns) permits the transportation of appreciable amounts at distances of about 100 km. Levels of among-population differentiation estimated from isoenzyme data are 0.22, 0.081 and 0.16 for Turkish, Italian and French populations, respectively. These values are similar to those reported in other species of the same family (Villani et al. 1991a and 1991b; Manchon et al. 1996).

Adaptive variation
Chestnut covers a broad range of site conditions, which may have resulted in adaptation to diverse conditions. Information from provenance tests showing selection responses of the most important traits is not yet published. The first provenance tests, established in south Germany, are still very young (Maurer and Tabel 1997). Important traits are tolerance to drought and to temperature regime, and resistance to Phytophthora spp. and Chryphonectria parasitica. Selection pressures act in different ways, resulting in different Selective Environmental Neighbourhoods (SENs). Differences in climatic characteristics of the natural range of the species suggest an important variation in these traits; however, the existence of additive genetic variance in the resistance to Phytophthora and Chryphonectria is not yet known.

Mutations
Mutation rate is supposed to be high, of the same level as in other tree species. Production of seeds starts early, at the age of 5 years, and fruit set is annual.

Genetic drift
Genetic drift is an important parameter in mixed forest, where chestnut is a rare species. Population decline due to diseases (Phytophthora spp. and C. parasitica) also must have caused genetic drift.

Phenotypic plasticity
Chestnut does not perform as a plastic species with respect to the soil pH and flooding.

Structuration of genetic variability
Genetic variability at isoenzymes along the range of C. sativa decreases from East Turkey, through West Turkey, to the western range of the species.

Some adaptive variation among populations located in discontinuous, isolated areas, subject to different ecological pressures, can be expected, but human interventions might have reduced part of the genetic variation. In continuous areas, more homogeneity can be expected owing to the influence of grafted chestnut orchards, the importance of geneflow and the impact on evolution caused by coppicing.

The number of grafted varieties is very high. They differ by size, shape, taste, conservation, peeling, etc. Although they are propagated by grafting, frequently they are polyclonal varieties.

To define the strategy of gene conservation from a European perspective, the following aspects need to be considered:

1. Objectives (breeding, conservation) and the different uses of the populations (timber or fruit production and landscaping) suggest using different populations to reach different goals.

2. Number of populations and genetic entries to be included.

3. Sampling of populations.

4. Activities to develop.

First of all, conservation concerns the species as a whole, and from a European perspective, reduction of the genetic entries needs to be considered.

As demonstrated by Williams et al. (1995) the Multiple Population Breeding System (MPBS) is an efficient method to combine both breeding and conservation of wild populations on a long-term scale. The method (Eriksson et al. 1993; Varela and Eriksson 1995) is not described in detail here.

In general, several populations will be selected, seed populations or mixed broadleaves forest, each with at least 50 genetic entries, and managed in different ways and with different main objectives. We can define at least 20 populations covering all the main situations to increase the genetic variability to be included. Designation of populations will be guided by discontinuous occurrence of chestnut areas and by ecogeographic gradients. Sampling inside each population will be guided by ecogeographic gradients and by existing knowledge of human influence. The choice of 50 genetic entries in each population will be carried out with the objective to obtain maximum genetic variability within the population. In this sense, a gradient from north to south within the range of distribution, and from east to west is to be distinguished.

Conservation of grafted local varieties for nut, wood production or both will be made in clonal archives. The main questions to design the conservation of fruit varieties are: how many varieties to conserve; what characteristics are the most important; how much 'intracultivar' variability to conserve?

Clonal archives of plus trees and local fruit varieties. This could be considered as a subpopulation, with the main objective of breeding and preserving the present composition of the chestnut forests and orchards. The main reason is to prevent their disappearance due to diseases or dysgenic selection. Each population will be conserved by the authority responsible for the collection and several copies of the selected plus trees will be established in the field collection. In vitro conservation can be considered, especially where C. parasitica has an important incidence.

Provenance tests will be established in contrasting environments not affected by any disease for ex situ conservation. The first objective is to study the variability of adaptive traits. A second objective is the preservation of the material. Castanea sativa shows a high variability in the Caucasus and Turkey, and therefore, some additional provenances should be collected from these regions. Design of the provenance trials will include material from all the populations selected (20). For this purpose, pooled material from 50 trees in each population will be collected. There must be no risk of pollen contamination with local provenances. 

Progeny tests of selected plus trees from several populations that will be tested in the sites where a breeding programme will be implemented. One of the populations will serve as a control, to be tested in different environments, and the others will be included depending on the interest of every country.

Managed stands. Conservation of different managed stands, focusing on the conservation of the genetic variation of the populations, will be established on the basis of the selected stands. In this case, a stand including at least 100 individuals will be included in such subpopulation. The subpopulation is the one sampled for the provenance test. 

• Distribution map of the species, including ecogeographic gradients and incidence of diseases.

• Studies of the mating patterns in different populations.

• Introgression with introduced species in Castanea sativa.

• Genetic relationship between wild populations and grafted (fruit) varieties.

• Protection of the species in seed populations (managed stands, provenance and progeny tests defining methods, number of populations and sampling of populations) and in clonal archives.

• Establishment of range-wide provenance tests to study the genetic variability of adaptive traits. It may be necessary to standardize the descriptors, methods of seeds conservation and design of field tests.

• Selection of plus trees: spatial distribution, selection methods, descriptors.

• Progeny tests: design, heritability of the main traits.

• Variability among and within fruit cultivars.

• Design of clonal archives: field collections, in vitro conservation methods, pollen conservation.

• Descriptors for clonal archives: morphology, phenology, biochemical and genetic markers.

• Promotion of regulations on the transfer of forest reproductive material. The trade with chestnut reproductive material is not regulated at the European level. The use of seeds from grafted trees instead of local provenances and the movement of plants among countries poses certain risks.

• Promotion of the use in reforestation of seeds from seed stands or seed orchards.

• Promotion of in situ conservation of fruit varieties.

Aira-Rodriguez, M.J. and P. Ramil-Rego. 1995. Paleobotanical data from northern Portugal (Baixo Minho) from pollen analysis and fossil seeds. Lagascalia 18(1):25-38.

Bergonoux, F., A. Verlhac, H. Breish and J. Chapa. 1978. Le châtaignier. INVUFLEC, Paris. 192p.

Bilger, I. 1998. Selection of chestnut stands: a first step toward the use of improved material for afforestation. COST G4, Multidisciplinary chestnut research. 2nd Meeting Genetic Resources. Santiago de Compostela, Spain, 14-16 May.

Borghetti, M., R. Giannini and C. Nocentini. 1983. Indagini preliminari sulla variazione di alcuni caratteri del frutto in popolazioni di Marrone Fiorentino. Monti e Boschi 1:49-52.

Breviglieri, N. 1955. Indagini ed osservazioni sulle migliori varietá italiane di Castagno. Pubblicazione del Centro di Studio sul Castagno, Firenze (Supplemento a la Ricerca Scientifica) 2:27-166.

Camus, A. 1929. Les Châtaigniers. In Encyclopédie économique de sylviculture. Lechevalier, Paris.

Eriksson, G., G. Namkoong and J.H. Roberds. 1993. Dynamic gene conservation for uncertain futures. For. Ecol. Manage. 62(1-4):15-37.

Fernández-López, J. 1996. Variability in isozyme, morphology and clonal selection in Castanea sativa Miller, C. crenata Sieb. et Zucc., C. mollissima and inter-specific hybrids. PhD thesis, Univ. Polit., Madrid.

Frank, N. and L. Radocz. 1998. European chestnut in Hungary and its local cultivars. COST G4, Multidisciplinary chestnut research. 2nd Meeting Genetic Resources. Santiago de Compostela, Spain, 14-16 May.

Frison, E. and J. Serwinski (editors). 1995. Directory of European Institutions Holding Crop Genetic Resources Collections, fourth edition. International Plant Genetic Resources Institute/FAO, Rome, Italy.

Gomes, J., C. Pinto and T. Valdiviesso. 1993. Chestnut selection in Portugal. Evaluation of some characteristics. International Congress on chestnut, Spoleto, Italy.

Gómez-Manzaneque, F. (coord.). 1997. Robledales y bosques mixtos de frondosas. Pp. 103-132 in Los bosques ibéricos. Planeta, Barcelona.

Kollmann, J. and H.P. Schill. 1996. Spatial patterns of dispersal, seed predation and germination during colonization of abandoned grassland by Quercus petraea and Corylus avellana. Vegetatio 125(2):139-205.

Manchon, N., L. Burel, M. Lefranc and N. Frascaria-Lacoste. 1996. Evidence of genetic drift in chestnut populations. Can. J. For. Res. 26:905-908.

Maurer, W.D. and U. Tabel. 1997. Genetic investigation on German sweet chestnut (Castanea sativa Mill.) provenances. Proceedings of the Workshop on tree physiology and genetic resources of chestnut. COST G4 Multidisciplinary chestnut research, Torino, Italy, June 1997. (in press).

Pereira-Lorenzo, S., J. Fernández-López and J. Moreno-González. 1996a. Variability and grouping of Northwestern Spanish chestnut cultivars. I. Morphological traits. J. Am. Soc. Hort. Sci. 121(2):183-189.

Pereira-Lorenzo, S., J. Fernández-López and J. Moreno-González. 1996b. Variability and grouping of Northwestern Spanish chestnut cultivars. II. Isoenzyme traits. J. Am. Soc. Hort. Sci. 121(2):190-197.

Pitte, J.R. 1986. Terres de castanide. Fayard. 364p.

Solar, A., F. Stampar and A. Podjavorsek. 1998. Pomological variability of the chestnut (Castanea sativa) in Slovenia. COST G4, Multidisciplinary chestnut research. 2nd Meeting Genetic Resources. Santiago de Compostela, Spain, 14-16 May.

Varela, M.C. and G. Eriksson. 1995. Multipurpose gene conservation in Quercus suber - a Portuguese example. Silvae Genet. 44:28-37.

Villani, F., S. Benedettelli, M. Paciucci, M. Cherubini and M. Pigliucci. 1991a. Genetic variation and differentiation between natural populations of chestnut (Castanea sativa Mill.) from Italy. Pp. 91-103 in Biochemical Markers in the Population Genetics of Forest Trees (S. Fineschi, M.E. Malvolti, F. Cannata and H.H. Hattemer, eds.). SPB Acad. Publishing, The Hague, Netherlands.

Villani, F., M. Pigliucci, S. Benedettelli, and M. Cherubini. 1991b. Genetic differentiation among Turkish chestnut (Castanea sativa Mill.) populations. Heredity 66:131-136.

Williams, C.G., J.L. Hamrick and P.O. Lewis. 1995. Multiple population versus hierarchical conifer breeding programs: a comparison of genetic diversity levels. Theor. Appl. Genet. 90:584-594.