Version classiqueVersion mobile
OpenEdition Books

AGRUMED: Archaeology and history of citrus fruit in the Mediterranean

Véronique Zech-Matterne
Girolamo Fiorentino

The INRA-CIRAD citrus germplasm collection of San Giuliano, Corsica

François Luro, Emmanuel Bloquel, Bruno Tomu, Gilles Costantino, Isabelle Tur, Simone Riolacci, François Varamo, Patrick Ollitrault, Yann Froelicher, Franck Curk et Olivier Pailly

Texte intégral

1Citrus production represents a major fruit crop in the Mediterranean as well as in the rest of the world, and citrus genetic resources are the cornerstone of citrus research, cultivation and breeding programs. To develop these activities, ex-situ germplasm collections were constructed in several countries. For more than fifty years, the SRA (Station de recherche agronomique) INRA (Institut national de la recherche agronomique)-CIRAD (Centre international de recherche et d’aide au développement) research centre has established, in accordance with appropriate sanitation controls, a healthy open-field citrus collection. Today, more than 1100 accessions, introduced from many areas of citrus cultivation across the world and representing the majority of the wild and cultivated species of the three major genera – Citrus, Poncirus and Fortunella – are maintained. The accessions are characterized according to the IPGRI citrus descriptors, supplemented by biochemical and molecular markers, and assessments on their resistance or susceptibility to biotic and non-biotic stresses. Molecular markers are routinely used to optimize the management of citrus germplasm as tools of traceability suitable for genetic identification at an early stage of plant development. This germplasm is exploited for research and breeding programs such as the genetic studies of phylogeny and heredity of agronomical traits or the production of triploid populations. It is also exploited for plant propagation by national and international nurseries, cultivators, professionals of industry and transformation, or individual amateurs. To manage the citrus germplasm more efficiently, a computer system is used to register information on every accession concerning introduction, sanitary status, pest control results, availability and localization of material, passport data, characterization, genetic identity and distribution.

1. Establishment of a healthy citrus collection in San Giuliano

2Corsica is a Mediterranean island free of naturally transmitted European and exotic Citrus quarantine diseases. The INRA-CIRAD citrus germplasm was founded at the beginning of the 1960s with the aim of avoiding introducing pests and microorganisms responsible for diseases transmissible by grafting. The major objective of this gathering of citrus varieties was to promote the development of citriculture in Corsica by selecting the most adapted varieties and rootstocks for the local environment. Four different periods can be identified in the establishment of this germplasm, corresponding to different waves of introduction and technical evolution. Soon after the creation of the research unit of SRA at San Giuliano, between 1959 and 1964, introductions were focused on citrus from Morocco and Algeria germplasms. These were submitted to a selective indexing procedure on plant indicators to eliminate individuals affected by graft-transmissible diseases. During the second phase, up to 1981, introductions of cultivars were performed using the seeds from a number of worldwide citrus producing countries: America, Asia, Africa and Australia. However, only polyembryonic varieties producing nucellar embryos, in addition to the sexual embryo, were introduced; only by using seedlings can healthy clonal propagation can be obtained. The selection of nucellar progeny to preserve clonal multiplication was made by selecting plants with a uniform phenotype. When the shoot-tip grafting procedure was adapted, since 1981, further introductions were made in the form of budsticks submitted to quarantine procedure. This new technology of healthy plant regeneration enlarged the collection diversity to include monoembryonic species and cultivars such as citrons, clementines, pummelos and bergamote, etc. Since the shoot-tip grafting procedure also eliminates the contamination risks by non-Mediterranean pathogens such as Liberibacter bacteria (Huanglongbing disease), tatter leaf virus and citrus canker bacteria, introductions from Asia have been thus possible.

3If the disease controls are negative, each accession is registered by the attribution of a unique SRA number and is amplified by grafting onto a suitable rootstock. If the accession is related to an “acid citrus” (lemon, lime and citron), sour orange (C. aurantium) or Volkamer lemon (C. limonia) are used as rootstocks. In other cases, except compatibility restriction, Carrizo citrange (Citrus sinensis x Poncirus trifoliata) is systematically used. One tree is then maintained as a “rescue” in a greenhouse or an insect-proof house, while the rest, around three or four, are planted in the arboretum orchard. This number of trees is necessary to reduce the risk of loss, to promote bud production in high quantity and to allow for enough replicates for experimentation taking into account statistical variance. When the trees in the field are growing, the plant rescue is discarded, or maintained in an insect-proof house if the accession is of great scientific or agricultural interest. Regular tests are carried out to maintain the healthy status of the open-field repository: annual controls for CTV (Citrus Tristeza Virus) and every triennial for CEVd (Citrus Exocortis Viroids). In 1999, the Ministry Office for Plant Protection decided to remove the quarantine facilities from the various INRA centres and centralize this function for every perennial crop in the centre of France, at Clermont-Ferrand. From 1999 to 2010, no citrus has been introduced from outside the EU into the Corsican germplasm, aside from a few seed-polyembryonnic varieties. Since 2011, collection enrichment with varieties introduced from other EU countries using budsticks is again possible.

4When the sanitary conditions are displayed, the availability of the accession for diffusion depends on its varietal authentication by phenotype characterization after fructification. As with many perennial fruit trees, the citrus has a juvenile period that delays fruiting for between four and eight years after seed germination depending to varieties and environmental factors. If the accession corresponds to the expected phenotype when compared to the data provided by the sender, it can be distributed. This whole process is time consuming, taking between seven and ten years, from the request of introduction to the availability of the accession (fig. 1).

Fig. 1 - General scheme of citrus introduction into INRA-CIRAD germplasm and duration of each step.

Fig. 1 - General scheme of citrus introduction into INRA-CIRAD germplasm and duration of each step.

2. Richness of the SRA citrus repository

5The INRA-CIRAD citrus collection currently maintains 1161 accessions, representing approximately 800 different wild or cultivated genotypes introduced from almost 50 places worldwide. The number of accessions for each taxonomic group or species is shown below (fig. 2).

Fig. 2 - Number of accessions for each citrus taxonomic group in the INRA-CIRAD collection (San Giuliano).

Fig. 2 - Number of accessions for each citrus taxonomic group in the INRA-CIRAD collection (San Giuliano).
  • 1 Tanaka 1961.

6Rootstock resources are represented by Poncirus trifoliata (trifoliate orange) with 82 accessions, Citrus aurantium (sour oranges) with 41 accessions and Citrus x Poncirus intergeneric hybrids such as citranges, citrumelos, citrandarins or citremons, with 120 accessions. Mandarins are the richest group with 340 accessions including 63 accessions of clementines. Tangors which are mandarin x orange hybrids, and tangelos which are mandarin x grapefruit hybrids, represent 68 accessions, and can be included in the mandarin group, giving a great specific characteristic to this collection. In terms of origins, many mandarins were introduced directly from South East Asia, the origin area of citrus, or indirectly via Morocco and the Ivory Coast, and many cultivars from the United States of America. The sweet orange group ranks second of the most represented taxa with 146 accessions. Conversely, pummelo (C. maxima), citron (C. medica) and the papeda species, three of the major Citrus species, have low numbers of accessions with only 19, 19 and 17 genotypes, respectively. More than 80 species from the 156 species of Tanaka Citrus taxonomy1 are represented in this germplasm. Other citrus genera such as Microcitrus and Eremocitrus, and relatives of the Aurantioideae sub-family are represented by only 16 accessions.

7The trees of each accession are planted successively in a row, every four metres (the two rows are six metres spaced). At the present time, in 2017, the collection covers 13 hectares with more than 4000 trees. Plots are mainly organized by citrus type (e.g. limes and lemons/mandarins/sweet oranges) or by use (rootstock/cultivars) .

3. Flow of shipped material from the citrus repository

8Many different kinds of material are available from the repository. In the case of tree cultivation, there are seeds and budsticks. If budsticks have been ordered, they are picked during spring when the first vegetative flush is occurring, and diffused immediately without storage step, whereas seeds are extracted and stored when the fruit has ripened. Seed harvesting begins in October with the trifoliate oranges (P. trifoliata) and ends in March with the Volkamer lemon (C. limonia). The nursery staff needs high quantities of seeds for rootstock amplification; therefore, a specific orchard is dedicated to seed production of the most requested rootstocks. After fruit harvest, the seeds are mechanically extracted, washed, air dried, treated for antifungus and stored at 5°C, 60-70% relative humidity in 1 kg plastic bags. A germination test is made by seed sampling those at the storage room’s entrance to verify the correct status of the seeds. The amount of total seed production per year is around 450 kg from approximately 20 different genotypes (fig. 3). Among them, Carrizo citrange and Volkamer lemon are the most requested rootstocks (55% of the total production). The seeds are mainly distributed to domestic nurseries (50%). In the EU, 93% of seed production is distributed to Mediterranean countries, with Italy being the main procurer (28%). The majority of bud shipments concern clementine (60%) and sweet orange (24%). In terms of diversity, 220 varieties/accessions have been out worldwide diffused between 2006 and 2012 with approximately 25% of sweet oranges and 25% of mandarins (fig. 3).

Fig. 3 - Details of plant material shipping flows between 2006 and 2012.

Fig. 3 - Details of plant material shipping flows between 2006 and 2012.

9Fruit are mainly requested for transformation (cosmetic, alcohol or jam production, ice cream makers or restaurants) and sometimes for research, such as the composition analysis of essential oils. Leaves are requested by researchers as sources of DNA or essential oils, and occasionally by restaurants. The fruit and leaves are also used during public events such as thematic fairs on science, agriculture or citrus production, where INRA/CIRAD staff inform visitors on their research and citrus resource preservation

4. Characterization and evaluation of accessions

4.1. Phenotypic description

10The description of each citrus accession is made on the basis of the “citrus descriptors” published by the IPGRI (International Plant Genetic Resources Institute).2 In comparison with field crops, perennial genetic resources have certain specificities mainly due to the existence of individuals within an accession. In term of management, this induces that some data may concern the accessions such as name, origin and phenotype characterization, but others have to be registered at the individual tree level, such as plantation date, death date and sanitary control.

11The organization of the description process depends on the tree growth and fruit maturity. Even if some vegetative characters such as tree or leaf shapes and dimensions may be observed before the fruiting age, the phenotypic description is always made when the tree is sexually mature. The internal branches are avoided for observation to reduce any environment influences and the effects of vigor associated with intense sap flow close to the trunk. The description is made on the three replicates of each accession with an equivalent number of samples per tree (10 leaves or fruit per tree, harvested around the periphery of the canopy). As far as possible, all characters are measured and the values are converted into classes as indicated by the “citrus descriptors”. Some fruit characters evolve during fruit development and maturation (e.g. size, juiciness, acidity, sweetness, rind colour) and, for many accessions, the commercial maturity and physiological status are unknown because they are not edible or were never cultivated in Corsica. In this situation, only sugar content (estimated by refractometer) and titratable acidity are measured on three different periods during fruit maturation. The rind colour is described using a chromameter (L*, a* and b* indices) when the chlorophyll is fully degraded and the rind homogeneously coloured.

4.2. Integration of molecular markers in germplasm management and accession description and identification

  • 3 Froelicher et al. 2008; Luro et al. 2008; Garcia-Lor et al. 2012; Ollitrault et al. 2010; 2012a; 20 (...)
  • 4 Froelicher et al. 2010; Garcia-Lor et al. 2012.

12Molecular markers have been proven useful for the assessment of genetic variation in germplasm collections. A large range of molecular markers are available, but among them, SSR (Single Sequence Repeats), InDels (Insertion/Deletion) and SNP (Single Nucleotide Polymorphism) markers are widely used in citrus collection management because of their reproducibility, multiallelic nature, codominant inheritance, monolocus location, relative abundance and good genome coverage.3 In addition, a few markers of chloroplastic and mitochondrial genomes are used to study the maternal heredity of citrus accessions.4 Citrus genotyping is particularly useful for genotype identification and for ensuring traceability in collection management.

4.2.1. Distinguishing zygotic from nucellar genotypes

13Many citrus cultivars and wild taxa produce polyembryonic seeds with one embryo originating from fertilization (zygotic embryo), and additional ones coming from the organogenesis of somatic cells (nucellar embryos). Usually, some citrus trees, such as rootstocks, are regenerated by polyembryonic seedlings. Morphological characters are mainly used to distinguish the two types of embryonic origins but many errors result from this, particularly with non-hybrid cultivars and self-fertilization. Initially, isozymes markers were used to distinguish nucellar and zygotic seedlings. Nowadays, this is supplanted by SSR or SNP markers due to their higher level of polymorphism. Every seedling of amplified variety is tested with at least five genetically independent SSR markers to certify, with a high probability (close to 98%), the nucellar origin of each seedling.

4.2.2. Cultivar identification

  • 5 Ollitrault et al. 2003.
  • 6 Butelli et al. 2012.

14Not all the citrus genotypes can be distinguished by SSR markers. Many cultivated citrus such as grapefruits, sweet oranges, lemons, sour oranges and clementines are known to be differentiated by somatic mutation.5 For example the blood-coloured pulp of some oranges cultivars is related to a natural mobility within the genome of a transposable element which was interposed in the promotor region of the Ruby gene.6 Except for cases of blood oranges, none of the phenotypes relating to mutations have been molecularly elucidated. Consequently, no markers, or very few, can distinguish these cultivars. Conversely, pummelos, citrons, limes and mandarins have mainly evolved by sexual hybridizations and therefore many cultivars of these citrus taxa are distinguishable with molecular markers.

15Thus, few homonyms have been detected among the existing trees of the citrus collection by different genetic profiles using SSR markers (fig. 4). These homonyms can be related either to misidentification that has occurred in different countries during citrus characterization, or to the selection of zygotic embryo seedlings from polyembryonic seeds. Molecular markers are of great interest for standardizing the genotype identification between repositories and for promoting the comparative analysis of phenotype variation under different climates, soils and agronomical practices.

Fig. 4 - Detection of three Sunki mandarin homonyms by genotyping with the 07B05 SSR marker. The different amplified DNA fragments and their sizes are indicated (arrows + base number). Photographs F. Luro.

Fig. 4 - Detection of three Sunki mandarin homonyms by genotyping with the 07B05 SSR marker. The different amplified DNA fragments and their sizes are indicated (arrows + base number). Photographs F. Luro.

5. Major research undertaken courtesy of the INRA-CIRAD citrus collection

5.1. On systematics and organization of diversity

  • 7 Ollitrault et al. 2003; Luro et al. 2002; 2008, Ollitrault et al. 2012a; 2012b; Garcia-Lor et al. 2 (...)
  • 8 Luro et al. 2012.
  • 9 Garcia-Lor et al. 2015.
  • 10 Curk et al. 2015.
  • 11 Ben Yahmed et al. 2016.
  • 12 Luro et al. 2012.
  • 13 Curk et al. 2015.

16The content of this collection was used to provide information on systematics, evolution and diversity organization of Citrus genus and species. Concerning the diversity of the Citrus genus, all the studies made with different markers reveal the same organization: a strong structure around three major species, pummelos (C. maxima), mandarins (C. reticulata) and citrons (C. medica), and the additional group of Papeda.7 The intraspecific diversity of these four taxa is high, while that of the cultivated species is zero or almost zero. By selecting markers with alleles specific to ancestral species, the diversity of some groups have been studied: citrons,8 mandarins,9 lemons and limes.10 These have all confirmed the hypothesis of the ancestral species status of pummelos, citrons, mandarins and papedas and their implication as genitors of important cultivated citrus, called secondary species. For Poncirus trifoliata (trifoliate orange), two major genetic sub-groups were detected clearly differentiated by SSR analysis.11 Due to cloning propagation by apomixis, the natural hybrid genotypes have been fixed (excepted mutations) allowing us to detect the alleles of ancestral species and then identify their putative parents. For example, citron cv. Corsican putatively originates from two successive self-fertilizations from citron cv. Poncire, an Italian citron cultivar;12 the Volkamer lemon (C. limonia), used as a rootstock of lemon varieties, is a hybrid of an unknown mandarin maternal parent and an unknown citron male parent.13

5.2. On tolerance to abiotic stress

  • 14 Hussain et al. 2012; 2015.

17Citrus trees are sensitive to the salinity of Mediterranean coastal lands, where citrus are widely cultivated and severely exposed. The selection of tolerant rootstocks is the best way to resolve this abiotic constraint. Sources of salinity tolerance were sought and cellular mechanisms of detoxification were studied in different species.14

  • 15 Ben Yahmed et al. 2015.

18The selection of grafted cultivar tolerant to salinity can be also sought to improve the global tolerance of the tree. Thus, an in-vitro evaluation was performed on sixteen mandarins,15 revealing variable behavior patterns within this Citrus taxa.

  • 16 Ben Yahmed et al. 2016.

19A physiological study of different Poncirus trifoliata accessions exposed to water shortage was conducted in pots under a greenhouse.16 This experiment concluded the lower sensitivity of one trifoliate orange subgroup to drought and validated the increased tolerance of tetraploid genotypes.

  • 17 Santini et al. 2012; 2013.

20Citrus cultivars are also known to be sensitive to cold temperatures due to their tropical origins. Furthermore, the diversification area of the Citrus ancestral species and species of another genera is wide and climatically varied. The photosynthesis activity of these ancestral species and the Fortunella genus, under optimal growing conditions and under oxidative stress conditions (cold temperatures), was studied directly on the tree collection.17 Each taxa had a particular physiology according to the thermal conditions, and the enzymatic detoxification system activity varied according to their tolerance to the cold.

5.3. On tolerance to biotic stress

  • 18 Vogel, Bové, 1986.
  • 19 Bernet et al. 2008.

21The sanitary status of the citrus collection does not allow experiments on quarantine pathogens in the research unit. However, the behavior of large samples of citrus accessions from the INRA-CIRAD collection has been analyzed with non-quarantine pathogens, as in the study of 39 trifoliate oranges which had been infected by viroids.18 Less sensitivity was detected both in the tetraploid genotypes and in clones of the large-flower group of trifoliate oranges. Other studies, outside of France, were performed on quarantine pathogens where the CTV inoculated 20 accessions of sour oranges originating from the INRA-CIRAD citrus collection.19 CTV tolerance appears related to slow-growing genotypes.

5.4. On breeding and selection

  • 20 Aleza et al. 2011.
  • 21 Mouhaya et al. 2010; Ben Yahmed et al. 2016.

22Natural chromosome doubling occurs spontaneously at the level of the nucellar embryo in polyembryonic varieties, leading to the development of tetraploid trees. This event can be very frequent (10 to 20%), depending on the genotype and low temperatures during the blossoming period.20 These tetraploid genotypes present a reduced sensitivity or a higher tolerance to some constraints, such as salinity and drought, than those of homologous diploid genotypes.21 The enrichment of the tetraploid genotypes (self-double diploids) collection in each taxa is performed by analyzing the ploidy level of polyembryonic citrus seedlings with a flow cytometer.

  • 22 Ollitrault et al.1998; 2007; 2008.

23Production of triploid genotypes gives rise to sterility (seedless fruit). If the spontaneous citrus are mainly diploid, very few natural citrus are triploid genotypes such as Tahiti lime (Citrus latifolia), for instance. These originated from an unreduced gamete (2n) fertilized by a normal gamete (1n). The citrus collection supports the triploid breeding program by making trees from different accessions available for cross pollinations.22 Another way for triploid production was explored by crossing tetraploid genotypes with diploid ones. Several thousand triploid hybrids have been produced since the beginning of the 21th century and are in the evaluation phase to select those that will combine the best agronomic traits requested by producers and consumers (ease of peeling, maturity spread over time, homogeneity of production, juicy, good flavor and aroma...).

  • 23 Jacquemond, Rocca Serra 1992.

24A selection program was also conducted by testing rootstocks or cultivars maintained in the citrus collection. For instance, over the last forty years, more than 150 rootstocks have been experimented on, in combination with clementine cultivars.23 Some clones of trifoliate oranges and citranges have been classified as the best rootstocks for clementine production in Corsican citriculture conditions.

5.5. On metabolite compositions

  • 24 Luro et al. 2011.
  • 25 Albertini et al. 2006.

25Maintaining the citrus collection in orchards is useful for fruit yield and fruit composition analysis under conditions of normal tree production. A study on juice acidity and sweetness was made at different times of fruit maturity for several Citrus species.24 Four phenotypic profiles were observed in terms of the proportion of these two characters as well as in the evolution of these characters during ripening. The composition of sugar and organic acids in an acidless citrus genotype group was compared to that of acidic genotypes homologous during fruit development.25 Two models of primary metabolic pathway were suggested, one for the acidic citrus and one for the acidless citrus.

  • 26 Dugrand et al. 2013; Durand-Hulak et al. 2015.

26In another study, the composition of polyphenols in the different fruit layers was analyzed using UPLC (Ultra Performant Liquid Chromatography) among major Citrus species.26 Furocoumarins are toxic components for human health and are produced specifically by pummelos and its hybrids (grapefruits and sour oranges) but also by limes and bergamots.

  • 27 Fanciullino et al. 2006a.
  • 28 Luro et al. 2012.
  • 29 Paoli et al. 2015.
  • 30 Lota et al. 2002.

27Fruit or leaf essential oils are components present in great quantities, responsible for the citrus fragrance. Quantitative and qualitative analysis have been carried out on different Citrus species: mandarins,27 citrons,28 grapefruits29 and limes.30 Some essential oil components can be used as diagnostic profiles of the Citrus species, but they are less powerful for intraspecific diversity analysis when the genotypes had close genetic relationships.

  • 31 Fanciullino et al. 2006b.
  • 32 Fanciullino et al. 2007.

28The composition in carotenoids, the major components responsible for fruit peel and pulp colours, were investigated in the three ancestral species and in secondary cultivated species.31 The diversity in carotenoid composition led to a structure very close to the organization revealed by molecular markers of genes of the carotenoid biosynthetic pathway.32

6. Certification of the INRA-CIRAD citrus collection

29Since August 2014, part of the INRA-CIRAD citrus collection (1083 accessions out of 1161) was certified by the Biological Resources Center (BRC) according to the French standard NF S96-900 entitled “Quality of biological resource centers (BRCs) – Management system of a BRC and quality of biological resources”. Until now, the Citrus BRC of San Giuliano is the only INRA Perennial plants BRC to be certified. An ISO (International Organization for Standardization) standard is currently being discussed on the basis of this standard, meaning that every NF S96-900 certified BRC would probably also be ISO certified. With increasing controls and restrictions on worldwide vegetal movements to avoid pathogen and disease propagation, providing certified disease-free plant material will probably become imperative for the BRC.

30The Citrus BRC activities are divided into three operational processes (introduction, conservation, diffusion), four support processes (communication, local and material management, financial management, information system) and two management processes (human resources, quality management). Two other processes, characterization and sanitary control, are not included in the certification perimeter but are linked through written specifications. Every process is under the responsibility of a pilot to insure its efficiency.

31The Citrus BRC governance is provided by a steering committee made up of all the pilots of the various processes of BRC and representatives of interested parties to the BRC. The steering committee meets three to four times a year.

7. Computerization of the citrus genetic resource management

32A system called EGID was developed in San Giuliano during the 1990s to manage the repository (e.g. entry, localization, identity, sanitary test results and shipments), varietal characterization and passport data. This system is no longer used due to the inability of the programming language to fit with computer system upgrades.

33However, recently we have successively developed two different databases:

34 A Microsoft Access Database as a management tool for the Citrus BRC, interconnected with a national INRA multicrop data system project (Siregal:​Projects/​URGI-softwares/​Siregal) to facilitate curated data transfer (via CSV Files). The data were collected from our citrus collection according to the IPGRI citrus descriptor and adapted to this citrus BRC specificity, such as tree traceability. The interface linked to the Access database allows each person involved in the Citrus BRC to read and/or input information according to the processes required by the certification mentioned above.

35 A web database, developed in 2014, whose technical basis is one of several computer server Linux, Apache, MySQL, PHP (LAMP). As for the Access database above, data comes from the first developed data sytem (EGID) and tables have been reviewed to fit the Siregal database. The main reason for developing this new database was to harness the web as it is the most powerful way to improve visibility and possibilities to request data and/or genetic resources; which is why the website for the Citrus BRC was created at the same time.

36The final objective is to solely use the web database and website either for internet access, as it is now, or for intranet access as the management tool for our BRC. The architecture of the website has been designed with this in mind and will be adapted for mobile and tablet use. Thus, we will get a full web toolbox, accessible from every computer, mobile phone or tablet for the management of the Citrus BRC that can be presented, and possibly shared, with other BRC managers worldwide

8. Perspective in security and long-term storage/concept of core collection

37The preservation of citrus diversity in the form of trees planted in open fields offers numerous opportunities for researchers and breeders, is suitable for the diffusion of large amounts of material, and assists demonstrations to citrus growers, scholars, students or other visitors. Unfortunately, it is also exposed to the climate risks and pathogen attacks, reducing the opportunity of the perennial safe preservation of genetic resources. However, additional forms of citrus collection preservation are under development. A selection representing the 300 most diffused accessions was amplified by grafting and transferred in an insect-proof greenhouse. This preservation form protects against insect-transmitted pathogens but not against freezing temperatures.

38To promote more secure long-term preservation, liquid nitrogen cryobanking is under investigation. Unfortunately, citrus seeds are recalcitrant or semi-orthodox (sensitive or semi-sensitive to dehydration), and are therefore not adapted to tolerate an intense desiccation process. A study on dehydration conditions suitable to maintain the seed germinate capacity after thawing was performed on different citrus species and genera. After desiccation at 75-78% of relative humidity, seeds from about 70% of the tested genotypes could partially support the deep-freezing process at a very low temperature (-196°C) in liquid nitrogen. Some taxa such as P. trifoliata and Fortunella sp. are very sensitive to dehydration, so they are not adapted to this procedure. The cryo-banking project of the citrus collection is still in progress and alternative methods to seed cryopreservation will investigated for recalcitrant, sterile or monoembryonic genotypes (eg. apex).

  • 33 Garcia-Lor et al. 2017.

39In this perspective, the definition of core collection is complementary to the efforts to preserve genetic resources by defining a group in reduced number (approximately 20% of the total) citrus representative of the major diversity. Methodological studies are carried out in order to select the best methods and the best selection criteria (genetic markers, phenotypic characters, agronomic traits).33

9. The citrus germplasm as a communication and education support

40The INRA-CIRAD Citrus germplasm bank is also a pedagogical and communication support. Every year, five to ten school classes, students from the University of Corsica Pasquale Paoli in Corte and from agricultural schools are informed by scientists and germplasm curators on the process of germplasm management, citrus diversity, history and phylogeny and agriculture practices. Growers and nurseries from the Mediterranean often organize tours to our BRC, looking for diversification in citrus production, new varieties and genetic material supplying. Transformers, such as confectioners, ice cream makers, restaurants, pastry chefs or perfumers, looking for colours, textures, flavours or original fruit flavours, also visit the BRC to learn about Citrus diversity, phylogeny, characteristics and uses. Finally, we also organize group requests from amateurs and interested individuals.

41Moreover, a dozen TV shows, radio broadcasts, magazines reports or daily press articles from French speaking countries (e.g. France, Canada and Switzerland) are made each year directly inside the citrus germplasm plots of San Giuliano. A presentation of the citrus fruit diversity of this germplasm is also presented at public events related to citrus, with the most representative and diversified varieties (fig. 5). In addition, we try to organize an annual open-door day to provide an opportunity for the general public to visit the INRA-CIRAD Citrus germplasm.

Fig. 5 - Fruit diversity of citrus accessions harvested in the INRA-CIRAD citrus germplasm. Photograph F. Luro.

Fig. 5 - Fruit diversity of citrus accessions harvested in the INRA-CIRAD citrus germplasm. Photograph F. Luro.


42The INRA-CIRAD citrus germplasm bank is one of the greatest citrus repositories worldwide, and is well known for its large diversity of the Citrus genus, particularly within the mandarin group. With the long-term benefit of a pathogen-protected environment, together with rigorous phytosanitary processes, many accessions of this collection have been distributed around the world. The great diversity maintained in open fields is the basis of many research programs producing in-depth knowledge regarding citrus phylogeny, genetics, physiology and agronomy.

43Unfortunately, because of climate change and the spread of pathogens, this open-field collection is exposed to risks; therefore, its diversity must be preserved by developing new secure forms of conservation and by researching new methods of communication or germplasm discovery (Internet virtual visit).

44For about fifty years, scientists and technicians at San Giuliano have improved the knowledge, techniques, methodologies and strategies to introduce, preserve, share and diffuse healthy characterized citrus resources. Through this article, we wanted to give you an overview of what has been achieved and what is still being achieved thanks to their hard work.

More details of the citrus germplasm can be found at:​Outils-et-Ressources/​Conservatoire-Agrumes
If you require specific information, please contact us using the following e-mail addresses: (steering committee) (general information) (genetic resources request)


Albertini et al. 2006: M.V. Albertini, E. Carcouet, O. Pailly, C. Gambotti, F. Luro, L. Berti, Changes in organic acids and sugars during early stages of development of acidic and acidless citrus fruit, Journal of Agricultural and Food Chemistry, 54, 21, 2006, p. 8335-8339.

Aleza et al. 2011 : P. Aleza, Y. Froelicher, S. Schwarz, M. Agust, M. Hernandez, J. Juarez, F. Luro, R. Morillon, L. Navarro, P. Ollitrault, Tetraploidization events by chromosome doubling of nucellar cells are frequent in apomictic citrus and are dependent on genotype and environment, Annals of Botany, 108, 1, p. 37-50.

Barret, Rhodes 1976: H.C. Barrett, A.M. Rhodes, A numerical taxonomic study of affinity relationships in cultivated Citrus and its close relatives, Systematic Botany, 1, p. 105-136.

Ben Yahmed et al. 2015: J. Ben Yahmed, P. Novillo, A. Garcia-Lor, A. Salvador, M. Ben Mimoun, F. Luro, M. Talon, P. Ollitrault, R. Morillon, Salt tolerance traits revealed in mandarins (Citrus reticulata Blanco) are mainly related to root-to-shoot Cl-translocation limitation and leaf detoxification processes, Scientia Horticulturae, 191, p. 90-100.

Ben Yahmed et al. 2016: J. Ben Yahmed, G. Costantino, P. Amiel, M. Talon, P. Ollitrault, R. Morillon, F. Luro, Diversity in the trifoliate orange taxon reveals two main genetic groups marked by specific morphological traits and water deficit tolerance properties, The Journal of Agricultural Science, 154, 3, 2016, p. 495-514.

Bernet et al. 2008: G.P. Bernet, M.T. Gorris, E.A. Carbonell, M. Cambra, M.J. Asins, Citrus tristeza virus resistance in a core collection of sour orange based on a diversity study of three germplasm collections using QTL-linked markers, Plant Breeding, 127, 4, p. 398-406.

Butelli et al. 2012: E. Butelli, C. Licciardello, Y. Zhang, J. Liu, S. Mackay, P. Bailey, G. Reforgiato-Recupero, C. Martin, Retrotransposons Control Fruit-Specific, Cold-Dependent Accumulation of Anthocyanins in Blood Oranges, The Plant Cell, 24, 3, 2012, p. 1242-1255.

Citrus descriptor, IPGRI:

Curk et al. 2014: F. Curk, G. Ancillo, A. Garcia-Lor, F. Luro, X. Perrier, J.P. Jacquemoud-Collet, L. Navarro, P. Ollitrault, Next generation haplotyping to decipher nuclear genomic interspecific admixture in Citrus species: analysis of chromosome 2, BMC Genetics, 15, p. 152 (

Curk et al. 2015: F. Curk, A. Garcia-Lor, H. Snoussi, Y. Froelicher, G. Ancillo, L. Navaro, P. Ollitrault, New Insights on Limes and Lemons Origin from Nuclear and Cytoplasmic Markers Genotyping and Targeted Nuclear Gene Sequencing, in B. Sabater-Muñoz et al. (eds.), Proc. XIIth Intl. Citrus Congress, Acta Horticulturae, 1065, ISHS.1, p. 113-124.

Dugrand et al. 2013: A. Dugrand, A. Olry, T. Duval, A. Hehn, Y. Froelicher, F. Bourgaud, Coumarin and furanocoumarin quantitation in citrus peel via ultraperformance liquid chromatography coupled with mass spectrometry (UPLC-MS), Journal of Agricultural and Food Chemistry, 1, 2013, p. 10677-10684.

Durand-Hulak et al. 2015: M. Durand-Hulak, A. Dugrand, T. Duval, L.P.R. Bidel, C. Jay-Allemand, Y. Froelicher, F. Bourgaud, A.L. Fanciullino, Mapping the genetic and tissular diversity of 64 phenolic compounds in Citrus species using a UPLC-MS approach, Annals of Botany, 115, 5, 2015, p. 861-877.

Fanciullino et al. 2006a: A.L. Fanciullino, F. Tomi, F. Luro, J.M. Desjobert, J. Casanova, Chemical variability of peel and leaf oils of mandarins, Flavour and Fragance Journal, 21, 2006, p. 359-367.

Fanciullino et al. 2006b: A.L. Fanciullino, C. Dhuique-Mayer, F. Luro, J. Casanova, R. Morillon, P. Ollitrault, Relationships between juice carotenoid profiles and genetic diversity within cultivated citrus, Journal of Agricultural and Food Chemistry, 54, 12, 2006, p. 4397-4406.

Fanciullino et al. 2007: A.L. Fanciullino, C. Dhuique-Mayer, F. Luro, R. Morillon, p. Ollitrault, Carotenoid biosynthetic pathway in Citrus genus: number of copies and phylogenetic diversity of seven genes, Journal of Agricultural and Food Chemistry, 55, 18, 2007, p. 7405-7417.

Froelicher et al. 2008: Y. Froelicher, D. Dambier, G. Costantino, S. Lotfy, C. Didout, V. Beaumont, P. Brottier, A.M. Risterucci, F. Luro, P. Ollitrault, Characterization of microsatellite markers in Citrus reticulata Blanco, Molecular Ecology Notes, 8, 1, 2008, p. 119-122.

Froelicher et al. 2010: Y. Froelicher, W. Mouhaya, J.B. Bassene, G. Costantino, M. Kamiri, F. Luro, R. Morillon, P. Ollitrault, New universal mitochondrial PCR markers reveal new information on maternal citrus phylogeny, Theoretical and Applied Genetics, 7, 1, p. 49-61.

Garcia-Lor et al. 2012: A. Garcia-Lor, F. Luro, L. Navarro, P. Ollitrault, Comparative use of InDel and SSR markers in deciphering the interspecific structure of cultivated citrus genetic diversity; A perspective for genetic association studies, Molecular Genetics and Genomics, 287, 1, p. 77-94.

Garcia-Lor et al. 2013: A. Garcia-Lor, F. Curk, H. Snoussi-Trifa, R. Morillon, G. Ancillo, F. Luro, L. Navarro, P. Ollitrault, A nuclear phylogeny: SNPs, indels and SSRs deliver new insights into the relationships in the “true citrus fruit trees” group (Citrinae, Rutaceae) and the origin of cultivated species, Annals of Botany 111, 1, p. 1-19.

Garcia-Lor et al. 2015: A. Garcia-Lor, F. Luro, G. Ancillo, P. Ollitrault, L. Navarro, Genetic diversity analysis and population structure of the mandarin germplasm by nuclear SNP markers, in B. Sabater-Muñoz et al. (eds.), Proc. XIIth Intl. Citrus Congress, Acta Horticulturae, 1065, ISHS.1, p. 105-111.

Garcia-Lor et al. 2017 : A. Garcia-Lor, F. Luro, P. Ollitrault, L. Navarro, Comparative analysis of core collection sampling methods for mandarin germplasm based on molecular and phenotypic data, Annals of Applied Biology, 171, 3, p. 327-339 (

Herrerro et al. 1996: R. Herrero, M.J. Asins, E.A. Carbonell, L. Navarro, Genetic diversity in orange subfamily Aurantioideae I: Intraspecies and intragenus genetic variability, Theoretical and Applied Genetics, 92, 5, 1996, p. 599-609.

Hussain et al. 2012: S. Hussain, F. Luro, G. Costantino, P. Ollitrault, R. Morillon, Physiological analysis of salt stress behavior of citrus species and genera: low chloride accumulation is an indicator of salt tolerance in citrus, South African Journal of Botany, 81, p. 103-112.

Hussain et al. 2015: S. Hussain, R. Morillon, M.A. Anjum, P. Ollitrault, G. Costantino, F. Luro, Genetic diversity revealed by physiological behavior of citrus genotypes subjected to salt stress, Acta Physiologiae Plantarum, 37, p. 1740.

Jacquemond, Rocca Serra 1992: C. Jacquemond, D. de Rocca Serra, Citrus rootstocks selection in Corsica for 25 years, in Proceedings of the VIIth International Citrus Congress, Acireale, International Society of Citriculture, 21, p. 246-251.

Lota et al. 2002: M.L. Lota, D. de Rocca Serra, F. Tomi, C. Jacquemond, J. Casanova, Volatile components of peel and leaf oils of lemon and lime species, Journal of Agricultural and Food Chemistry, 3, 50-4, 2002, p. 796-805.

Luro et al. 1995: F. Luro, F. Laigret, J.-M. Bové, P. Ollitrault, DNA amplified fingerprinting, a useful tool for determination of genetic origin and diversity analysis in Citrus, HortScience, 30, 5, p. 1063-1067.

Luro et al. 2002: F. Luro, D. Rist, P. Ollitrault, Evaluation of genetic relationships in Citrus genus by means of sequence tagged microsatellites, in C. Doré, F. Dosba, C. Baril (eds.), Proceedings of the international symposium on molecular markers for characterizing genotypes and identifying cultivars in horticulture, Montpellier 2000, Acta Horticulturae, 546, p. 237-242.

Luro et al. 2008: F. Luro, G. Costantino, X. Argout, Y. Froelicher, J. Terol, M. Talon, P. Wincker, P. Ollitrault, R. Morillon, Transferability of the EST-SSRs developed on Nules clementine (Citrus clementina Hort ex Tan) to other Citrus species and their effectiveness for genetic mapping, BMC Genomics, 9, 2008, p. 287.

Luro et al. 2011: F. Luro, J. Gatto, G. Costantino, O. Pailly, Analysis of genetic diversity in Citrus, Plant and Genetic Resources, 9, 2, p. 218-221.

Luro et al. 2012: F. Luro, N. Venturini, G. Costantino, J. Paolini, P. Ollitrault, J. Costa, Genetic and chemical diversity of citron (Citrus medica L.) based on nuclear and cytoplasmic markers and leaf essential oil composition, Phytochemistry, 77, p. 186-196.

Mouhaya et al. 2010: W. Mouhaya, T. Allario, J. Brumos, F. Andrés, Y. Froelicher, F. Luro, M. Talon, P. Ollitrault, R. Morillon, Sensitivity to high salinity in tetraploid citrus seedlings increases with water availability and correlates with expression of candidate genes, Functional Plant Biology, 37, 7, p. 674-685.

Mohammadi, Prasanna 2004: S.A. Mohammadi, B.M. Prasanna, Analysis of genetic diversity in crop plants – salient statistical tools and considerations, Crop Sciences, 43, p. 1235-1248.

Ollitrault, Luro 2001: P. Ollitrault, F. Luro, Citrus, in A. Charrier, M. Jacquot, S. Hamon, D. Nicolas (eds.), Tropical plant breeding, Montpellier, p. 55-77.

Ollitrault et al. 1998: P. Ollitrault, D. Dambier-Sudahono, F. Vanel, F. Mademba Sy, F. Luro, B. Aubert, Biotechnology for triploid mandarin breeding, Fruit, 5, 53, p. 307-317.

Ollitrault et al. 2003: P. Ollitrault, C. Jacquemond, C. Dubois, F. Luro, Citrus, in P. Hamon, M. Seguin, X. Perrier, J.-C. Glaszmann (eds.), Genetic Diversity of Cultivated Tropical Plants, Boca Raton, p. 193-217.

Ollitrault et al. 2007: P. Ollitrault, Y. Froelicher, F. Luro, S. Yamamoto, I. Khan, Seedlessness and Ploidy Manipulations, in I.A. Khan (ed.), Citrus genetics, breeding and biotechnology, Wallingford, p. 197-218.

Ollitrault et al. 2008: P. Ollitrault, D. Dambier, Y. Froelicher, F. Luro, Ploidy manipulation for seedless triploid Citrus breeding, Plant Breeding Reviews, 30, p. 323-352.

Ollitrault et al. 2010: F. Ollitrault, J. Terol, J.A. Pina, L. Navarro, M. Talon, P. Ollitrault, Development of SSR markers from Citrus clementina (Rutaceae) BAC-end sequences and interspecific transferability in Citrus, American Journal of Botany, 97, 11, p. 124-129.

Ollitrault et al. 2012a: F. Ollitrault, J. Terol, A.A. Martin, J.A. Pina, L. Navarro, M. Talon, P. Ollitrault, Development of indel markers from Citrus clementina (Rutaceae) BAC-end sequences and interspecific transferability in Citrus, American Journal of Botany, 99, 7, p. 268-273.

Ollitrault et al. 2012b: P. Ollitrault, J. Terol, A. Garcia-Lor, A. Berard, A. Chauveau, Y. Froelicher, C. Belzile, R. Morillon, I. Navarro, D. Brunel, M. Talon, SNP mining in C. clementina BAC-end sequences; transferability in the Citrus genus (Rutaceae), phylogenetic inferences and perspectives for genetic mapping, BMC Genomics, 13, 2012, p. 13 (

Paoli et al. 2015: M. Paoli, F. Tomi, O. Pailly, F. Luro, Agreement between chemical composition of the leaf essential oil of grapefruit (Citrus paradisi Macf.) and their diversification process, Submitted to Journal of Essential Oil Research.

Santini et al. 2012: J. Santini, J. Giannettini, S. Herbette, O. Pailly, P. Ollitrault, F. Luro, L. Berti, Physiological and biochemical response to photo-oxidative stress of the fundamental citrus species, Scientia Horticulturae, 147, p. 126-135.

Santini et al. 2013: J. Santini, J. Giannettini, O. Pailly, S. Herbette, P. Ollitrault, L. Berti, F. Luro, Comparison of the photosynthesis and antioxidant performances in basic true species of Citrus and Fortunella growing under chilling conditions in a marginal zone, Trees, 27, 1, p. 71-83.

Tanaka 1961: T. Tanaka, Citologia: Semi-centennial commemoration papers on citrus studies, Osaka.

Vogel, Bové 1986: R. Vogel, J. Bové, Influence de l’Exocortis sur la croissance et la production du clémentinier greffé sur 39 lignées de Poncirus trifoliata, Fruits, 41, 11, p. 639-647.


1 Tanaka 1961.


3 Froelicher et al. 2008; Luro et al. 2008; Garcia-Lor et al. 2012; Ollitrault et al. 2010; 2012a; 2012b; Curk et al. 2014.

4 Froelicher et al. 2010; Garcia-Lor et al. 2012.

5 Ollitrault et al. 2003.

6 Butelli et al. 2012.

7 Ollitrault et al. 2003; Luro et al. 2002; 2008, Ollitrault et al. 2012a; 2012b; Garcia-Lor et al. 2012; 2013.

8 Luro et al. 2012.

9 Garcia-Lor et al. 2015.

10 Curk et al. 2015.

11 Ben Yahmed et al. 2016.

12 Luro et al. 2012.

13 Curk et al. 2015.

14 Hussain et al. 2012; 2015.

15 Ben Yahmed et al. 2015.

16 Ben Yahmed et al. 2016.

17 Santini et al. 2012; 2013.

18 Vogel, Bové, 1986.

19 Bernet et al. 2008.

20 Aleza et al. 2011.

21 Mouhaya et al. 2010; Ben Yahmed et al. 2016.

22 Ollitrault et al.1998; 2007; 2008.

23 Jacquemond, Rocca Serra 1992.

24 Luro et al. 2011.

25 Albertini et al. 2006.

26 Dugrand et al. 2013; Durand-Hulak et al. 2015.

27 Fanciullino et al. 2006a.

28 Luro et al. 2012.

29 Paoli et al. 2015.

30 Lota et al. 2002.

31 Fanciullino et al. 2006b.

32 Fanciullino et al. 2007.

33 Garcia-Lor et al. 2017.

Table des illustrations

Titre Fig. 1 - General scheme of citrus introduction into INRA-CIRAD germplasm and duration of each step.
Fichier image/jpeg, 144k
Titre Fig. 2 - Number of accessions for each citrus taxonomic group in the INRA-CIRAD collection (San Giuliano).
Fichier image/jpeg, 152k
Titre Fig. 3 - Details of plant material shipping flows between 2006 and 2012.
Fichier image/jpeg, 248k
Titre Fig. 4 - Detection of three Sunki mandarin homonyms by genotyping with the 07B05 SSR marker. The different amplified DNA fragments and their sizes are indicated (arrows + base number). Photographs F. Luro.
Fichier image/jpeg, 320k
Titre Fig. 5 - Fruit diversity of citrus accessions harvested in the INRA-CIRAD citrus germplasm. Photograph F. Luro.
Fichier image/jpeg, 2,0M


UMR AGAP INRA-CIRAD Corse, Équipe SEAPAG, station INRA 20230 San Giuliano, France;

UE Citrus INRA 20230 San Giuliano, France

UE Citrus INRA 20230 San Giuliano, France

UMR AGAP INRA-CIRAD Corse, Équipe APMV, station INRA 20230 San Giuliano, France

UMR AGAP INRA-CIRAD Corse, Équipe APMV, station INRA 20230 San Giuliano, France

SDAR Centre INRA de Corse 20230 San Giuliano, France

SDAR Centre INRA de Corse 20230 San Giuliano, France

UMR AGAP INRA-CIRAD, Station de Roujol, 97170, Petit-Bourg, Guadeloupe, France

UMR AGAP INRA-CIRAD Corse, Équipe APMV, station INRA 20230 San Giuliano, France

UMR AGAP INRA-CIRAD Corse, Équipe APMV, station INRA 20230 San Giuliano, France

UE Citrus INRA 20230 San Giuliano, France

© Publications du Centre Jean Bérard, 2017

Conditions d’utilisation :


Volume papier