Conservation of Citrus Germplasm 3 страница

Protected collections are costly to establish due to the investments in screen- houses, and more expensive to maintain than fi eld collections, since they need to be cared for by labour with higher specializa- tion. However, they are highly recom- mended for germplasm conservation. They are the only way to maintain healthy geno- types in areas with naturally spreading dis- eases. This is needed in order to release material for certifi cation programmes and for a safe exchange of germplasm. A pro- tected collection is the only way to guaran- tee germplasm conservation in areas with severe naturally spreading diseases that may kill fi eld trees. In addition, plants can be much more easily protected against abi- otic stresses. An additional advantage of these collections is that parthenocarpy can be easily evaluated.

Protected collections vary somewhat. Some general information on suitable facil- ities and phytosanitary and security prac- tices was presented above. Within the facilities, the manner in which the trees are grown may vary somewhat. Maintaining trees in pots has some obvious advantages with regard to disease prevention, fertiliza- tion, frost protection, the ability to manipu- late and move the trees, etc. However, use of pots limits the size of the tree to a certain extent. The largest practical size for pots for this use is about 25 l. Pot sizes larger than this are more diffi cult to move, and trees maintained therein would be more diffi cult to trim and train. In other words, at a cer- tain point, it is probably more effi cient to maintain large trees planted directly in the soil rather than in extremely large contain- ers. Spacing of trees under screen needs to balance the need for adequate space for growth (mainly, providing enough space

such that crowding and shading is not a problem) with the need to maintain acces- sions in a limited area. Spacing will depend on tree size and vigour.

Trees maintained in pots do not have to be propagated exactly as if they were to be planted in the fi eld. The increased phy- tosanitary standard that is possible in a pro- tected facility makes a suitable tree for a protected collection somewhat different from a tree destined for fi eld planting. Rootstock selection is not as important as with fi eld trees, but in many, if not most, cases, the rootstocks of choice for fi eld trees are also very well suited and perhaps the best choice for a protected collection. A rootstock that produces a compact, long- lived tree is preferred. Excessive distance between bud union and crown is undesir- able in a protected collection both due to height limitation from the structure and from the standpoint of ease of manipulation of the trees. Therefore, the crown of the tree should be formed closer to the bud union than for a tree for the fi eld. The use of a ster- ilized soil mix makes practical a lower bud- ding height. Regular pruning creates a more compact tree as well as stimulating produc- tion of budwood. Pruning has to be done so as to allow fruit production, which is nec- essary to ascertain trueness to type. Re-pot- ting to larger pots may be necessary as the trees increase in size. Even when pot size is no longer increased, periodic re-potting is useful in order to prune excessive root growth and observe abnormalities in the roots.

With training such as that described above, it is possible to maintain for a con- siderable period of time a citrus tree at a size suitable for maintenance under screen. Currently, the oldest and largest trees main- tained under screen at Riverside are approximately 15 years old and are main- tained in 20 l pots. In Spain, the oldest trees are 22 years old; they have been maintained in 25 l containers without re-potting and they grow and produce fruits normally. Of course, at some point, it will be necessary to re-propagate these trees.

Trees maintained under screen are gen-

erally not suitable as sources of fruit for accurate characterization and evaluation purposes. However, they help greatly in ascertaining the trueness to type of geno- types, detecting possible misidentified accessions and detecting possible chimeras. These are very important considerations when releasing budwood for certifi cation. Similarly, due to the relatively large amount of manipulation to which they are subjected, trees in a protected collection are not suitable for characterization of most vegetative characteristics. These types of observations should be made on field- grown trees. Trees in a protected collection should only be propagated or repropagated from pathogen-tested budwood from a reli- able source.

Long-term preservation

Genetic resources of most plant species are stored in seedbanks. Seeds are dehydrated and stored in sealed containers at low tem- peratures (usually about –20˚ C). This is a very safe and inexpensive procedure but requires that seeds be tolerant to desicca- tion and freezing temperatures. These are the so-called ‘orthodox’ seeds, as opposed to ‘recalcitrant’ seeds that lose viability when the moisture content is reduced. For many years, citrus seeds were considered recalcitrant because they could be stored only for short periods of time. As a conse- quence, citrus genetic resources are main- tained as living plants in fi eld or protected collections, with the associated high cost and the risk of losing genotypes due to cli- matic or biological hazards.

This situation is common to many veg- etatively propagated crops, particularly fruit trees. During the last two decades, there have been investigations into the use of tissue culture methods, including slow growing vegetative shoots and cryopreser- vation of several tissues, for dealing with the problem of long-term conservation of genetic resources of these crops (Withers, 1992; Engelmann, 1997). In citrus, most approaches have been investigated with variable results (Duran-Vila, 1995, 1997)

and they are briefl y reviewed in the follow- ing sections.

SLOW GROWTH OF in vitro VEGETATIVE TISSUES. With

some species, it is possible to establish pro- cedures for slow growth of shoots, usually using low temperatures and decreased light intensity in conjunction with some modifi - cations in the culture medium. Re-culture of the explants is done at intervals of about one year. This procedure allows a routine con- servation of germplasm of several species in vitro (Engelmann, 1997). Attempts to follow this approach in citrus were not successful, because the procedure developed for juve- nile material involved a complicated cycle of producing shoots from nodal segments, root- ing of the shoots and re-culturing of nodal segments (Marí n and Duran-Vila, 1991). In addition, there was a low effi ciency in both rooting and shoot production from nodal explants, and this procedure could not be used with adult material.

CRYOPRESERVATION OF APICAL SHOOT-TIPS. Shoot-

tips of adult material are the best tissue for cryopreservation of citrus genetic resources, because regenerated plants will not have juvenile characteristics and will be readily available for breeding. Recently the success- ful cryopreservation of shoot-tips from juve- nile plants of P. trifoliata using the encapsulation dehydration technique has been reported (Gonzá lez-Arnao et al., 1988). However, this work could not be applied at IVIA to other species or adult material (M.

T. Gonzá lez-Arnau et al., unpublished results). The main limitation is that follow- ing freezing and thawing, only a portion of the shoot-tip may survive. In citrus, small shoot-tips do not regenerate in vitro (with the exception of some species such as P. tri- foliata) and it is necessary to regenerate plants by shoot-tip grafting, which cannot be done with only a portion of the shoot-tip.

CRYOPRESERVATION OF SEEDS. Although citrus seeds were considered recalcitrant, some limited studies have demonstrated that there are Citrus species tolerant to desicca- tion, others partially tolerant and still others recalcitrant (i.e. they lose viability

very quickly with desiccation) (Mumford and Grout, 1979; King et al., 1981; Pé rez, 1995). Seeds of species tolerant to desicca- tion, such as lemon, Mexican lime and sour orange, can be cryopreserved with good survival rates by direct immersion in liquid nitrogen after desiccation. Seeds of species partially tolerant to desiccation (e.g. sweet orange and common mandarin) have vari- able survival rates depending on the degree of dehydration and are diffi cult to establish on a large scale. No survival was obtained with recalcitrant species (e.g. Cleopatra mandarin, P. trifoliata) (Pé rez, 1995). Cryopreservation of seeds has not been used in practice for conservation of citrus genetic resources.

CRYOCONSERVATION OF OVULES. Under-

developed ovules are possibly a good tissue for conservation of genetic resources of citrus polyembryonic genotypes. They are easily excised from immature fruits and efficient plant regeneration can be achieved. There was an early report claim- ing survival of ovules by direct immersion in liquid nitrogen (Bajaj, 1984). However, survival of ovules has not been achieved in several experiments done at IVIA using dif- ferent freezing protocols, including fast and slow cooling rates, and vitrifi cation (unpub- lished results). Following the encapsula- tion–dehydration technique, occasionally a few ovules have survived (1–16%) but not at a rate high enough for practical applica- tion (Gonzá lez-Arnao et al., 2003).

CRYOPRESERVATION OF SOMATIC EMBRYOS.

Cryopreservation of somatic embryos pro- duced by ovules of sweet orange cultured in vitro was achieved some time ago (Marí n and Duran-Vila, 1988; Marí n et al., 1993). However, survival rates were low and erratic, and a careful selection of embryos in early developmental stages was needed to achieve survival. The procedure was dis- carded as a practical method for germplasm preservation. Recently, cryopreservation of somatic embryos of several Citrus species produced by ovule culture in vitro has been accomplished by the encapsulation–dehy-

dration technique with high survival rates (75–100%) (Gonzalez-Arnao et al., 2003). This method does not require a careful selection of embryos. Recovery of plants was rapid, since the whole embryo sur- vived the treatment and germinated readily. In addition, somatic embryos of polyembry- onic varieties are easily produced by ovule culture in vitro. Although additional exten- sive work with a wider range of genotypes needs to be done, this procedure is very promising for the conservation of citrus germplasm. However, there is still the limi- tation that regenerated plants have juvenile characters.

CRYOPRESERVATION OF EMBRYOGENIC CALLUS.

Embryogenic callus produced by ovules cultured in vitro of several citrus species and some related genera have been success- fully cryopreserved, either directly or in cell suspension cultures, in a number of laboratories using several procedures, including slow cooling and vitrifi cation (Kobayashi et al., 1990; Sakai et al., 1990, 1991a, b; Aguilar et al., 1993; Engelman et al., 1994; Pé rez et al., 1997, 1999; Xiancai, 1997). Survival rates are very high, and nor- mally 100% of the calluses included in the cryotubes produce embryos and plants after thawing and reculture. The disadvantage of this procedure is that the production of embryogenic callus by ovules cultured in vitro is genotype dependent and time con- suming (Pé rez et al., 1998) and that recov- ered plants have juvenile characters. The advantage is that germplasm is maintained as embryogenic callus, a very valuable and diffi cult to obtain tissue that is the basic material for somatic hybridization (Grosser et al., 2000). Maintaining embryogenic callus under normal growing conditions requires monthly subcultures. This is a time-consuming procedure that may result in problems associated with somaclonal variation, decrease or even loss of their embryogenic potential, and risk of loss by contamination due to in vitro manipula- tions. After thawing, cryopreserved callus has been proved to be excellent material for protoplast isolation (Olivares-Fuster et al.,

2000) and has been used for exchange with other laboratories. Embryogenic callus also can be used for genetic transformation (Fleming et al., 2000). Taking into account these advantages and the fact that somatic hybridization and genetic transformation programmes are being carried out at IVIA, this method has been adopted to establish a collection of cryopreserved embryonic callus of more than 40 accessions.

The standard cryopreservation proto- col used at the IVIA germplasm bank was described by Pé rez et al. (1997). Briefl y, 150–200 mg of loose cells are cryoprotected in 1.8 ml of liquid basal medium supple- mented with 10% (v/v) dimethylsulphox- ide (DMSO) and maintained at 4˚ C for 30 min. Cryoprotected cultures are frozen by slow cooling in 2 ml cryotubes using a pro- grammable freezing unit. Samples are cooled from 4 to –40˚ C at –0.5˚ C/min in order to obtain a smooth cooling curve. The samples are fast cooled from –40 to –150˚ C at –20˚ C/min and then placed in liquid nitrogen and stored. Cultures are thawed by immersion of the cryotubes in a 37˚ C water bath for 5 min. The cryoprotectant solution is removed from the cryotubes and cells are washed three times with 1.8 ml of liquid basal medium and transferred to solid basal medium for production of embryos and plants.

Characterization and evaluation

The effi cient and effective utilization of citrus germplasm requires sound and accu- rate knowledge and documentation of its traits, i.e. it entails a description of what is in a collection. Descriptions of a germplasm resource are conveyed by descriptors based upon passport data, characterization, and evaluation of the germplasm. Passport data include basic information on the origin and type of the germplasm. Management data trace the history of an accession, the han- dling of its propagative units, its distribu- tion, regeneration, etc. This ensures that users of germplasm are handling the mate- rials that they believe they are.

A distinction between characterization and evaluation is sometimes made, although this is somewhat arbitrary and the bound- aries somewhat blurred. Characterization refers to documentation of characters that are highly heritable, are easily identifi ed (usually qualitative) and are expressed in all environments, while evaluation consists of documentation of additional characters (often quantitative) which are thought desir- able by a consensus of users of the crop. Traditional phenotypic and modern molec- ular characterization of citrus is discussed by Gmitter et al. (1999), and some of the extensive documentation of characteriza- tion and evaluation data of citrus and related genera that has been generated over the last century is shown in Table 4.3.

Responsibility for characterization and evaluation varies. The curator is usually involved with characterization (usually the more basic attributes), while advanced or complex evaluations may be beyond the curator’s capabilities and/or resources. Curators have the primary responsibility for documentation, which increasingly is via computerized databases, such as the Germplasm Resources Information Network (GRIN) system (Germplasm Resources Information Network, 1995; Mowder and Stoner, 1989), or the specifi c citrus data- bases GERMO developed at IVIA (< https://www.ivia.es/deps/biot/germop.ht m>) and EGID developed at CIRAD, France (< https://www.corse.inra.fr/sra/>).

The descriptors most widely used for citrus are those developed by the International Board for Plant Genetic Resources (IBPGR) (1988; recently revised as International Plant Genetic Resources Institute (IPGRI), 1999), which are a slightly modified and expanded version of the ‘Fruit Description Outline for Citrus’ devel- oped many years ago by H. J. Webber of the University of California Citrus Research Center (Webber, 1943; Hodgson, 1967). These descriptors are chiefl y concerned with documentation of passport data and basic morphological traits. Other systems of descriptors developed independently are often fairly similar to these standardized

descriptors, since the basic attributes to be described are fairly intuitive and obvious. These basic descriptors fall under ‘charac- terization’.

The descriptors are adequate for describing the basic morphology of citrus. However, they do not address some very basic characteristics (e.g. growth rate), and their treatment of important physiological, pathological, horticultural and genetic characteristics is limited to a few items tacked on to the bottom of the morphologi- cal descriptors. One obvious example is that, except for one item dealing with one specifi c scion/rootstock compatibility, root- stock characteristics or other traits that might infl uence an accession’s suitability as a rootstock are not dealt with. Yet, for many accessions, these would be the traits of most interest and importance. Another issue with the IBPGR/IPGRI descriptors is their utility for the citrus relatives.

Some of the shortcomings of the descriptors are due more to their applica- tion than to the descriptors themselves. There are two major considerations in assessing the value of descriptor data: geo- graphical area and time. Geographical area can affect citrus morphology and growth via environmental (climatic) effects, soil conditions, water quality, air quality, pest and/or disease endemism, etc. Climatic fac- tors (including temperature, photoperiod, rainfall, humidity and soil temperature) affect vegetative growth, fl owering, fruit set, fruit composition, fruit growth and fruit morphology (Reuther et al., 1969; Reuther, 1973; Germaná and Sardo, 1988). Soil con- ditions interact with water quality in affect- ing citrus morphology via fertility level (Embleton et al., 1973a; Reitz and Embleton, 1986); presence or absence of vesicular-arbuscular mycorrhizae (Menge et al., 1977); salinity (Bernstein, 1968; Maas, 1993); and water relations (Kriedemann and Barrs, 1981; Syvertsen and Lloyd, 1994). Air pollution can also affect the mor- phology of citrus (Thompson and Taylor, 1969; Olszyk et al., 1988; Yelenosky, 1991). Time can also affect morphological obser- vations. Due to patterns of seasonal growth,

the time of the season in which an observa- tion is made will affect such characteristics as fruit maturity and composition, fl ower- ing, percentage fruit set, size of various organs, etc. There are probably also less obvious effects on such parameters as pest and disease resistance, physical properties, etc. that are mediated more by environmen- tal conditions than the time of season per se. The age of the trees upon which the observations are made are another facet of the effects of time which should be taken into account.

Another weakness of much descriptive work is that it is performed in one (or a lim- ited number of) site(s) and for one (or a lim- ited number of) year(s). Repetition of observations over time is easily done given adequate resources. However, dealing with the effects of location (i.e. climate) is less easily accomplished due to the fact that germplasm collections usually exist in only one location, and due to cost and other con- straints it is often not feasible to replicate collections in a number of diverse climates. In any case, evaluation of perennial crop germplasm is a long-term endeavour. It would also be diffi cult to account for differ- ences in cultural practices, pests and dis- eases, etc.

Another constraint caused by resource limitation is the small number of trees that can be maintained and hence evaluated. Many germplasm collections maintain only a few trees of each accession. This is proba- bly suffi cient to make general morphologi- cal evaluations, but not for more complex types of evaluations that require replicated trials. These include such important traits as pest and disease resistance, rootstock characteristics, responses to environmental conditions or cultural practices, etc. These types of trials, while important in evaluat- ing germplasm and determining its value, are sometimes outside the scope of what can be evaluated within a germplasm col- lection. Most such collections consist of only a few specimens of each accession, and destructive (or potentially destructive) trials compromise the integrity of a collec- tion. Although these investigations are out-

side the scope of what can be investigated within a germplasm collection, they may not be outside the responsibility of the repository scientists, depending upon the areas of expertise. However, they would have to be performed on trees planted specifi cally for trials and not on trees in the collection. More complex evaluations (dis- ease resistance, physical properties, etc.) are very important but often have to be investigated as stand-alone research proj- ects outside of an established repository or germplasm conservation system. These types of investigations require more resources than the initial characterizations, since they are complex, intensive, multi- year projects in many different areas. These types of investigations are by nature open ended and often yield new questions to investigate, all requiring adequate resources. Obviously a complete evaluation of genotypes in germplasm banks is a huge task that usually requires replication trials over a period of time and that is far beyond the responsibilities and resources of germplasm banks. Evaluation is the respon- sibility of all the citrus scientifi c commu- nity. The main problem is how to coordinate and place all the scattered infor- mation together in order to make it easily available. Evaluation is the black hole of genetic conservation.

Despite the limitations of phenotypic descriptors, they are quite useful for man- agement of germplasm. They are currently the only means to differentiate accessions of certain groups of species that have been produced by natural budsport mutations (e.g. sweet oranges, satsumas or clemen- tines), and likewise to identify duplications in the collection. Elimination of duplicates is very important because this reduces the cost of maintenance. Molecular markers have not at this time been developed to the point that they can be utilized for these purposes.

The last several decades have seen the evolution of biochemical and molecular markers as tools with great potential appli- cation to germplasm characterization. In contrast to morphologically based pheno-

typic characterization, molecular markers are generally unaffected by the many fac- tors able to infl uence plant or organ charac- teristics. This allows comparisons between accessions within a collection or among collections at different locations at any time of year, while phenotypic characteristics can be masked by environmental or cultural affects.

Molecular characterization has a number of applications in the management of germplasm collections. These include elucidating systematic relationships between accessions; assessing gaps and redundancies in the collection; develop- ment of core subsets; characterizing newly acquired germplasm; maintaining trueness to type; monitoring shifts in population genetic structure in heterogeneous germplasm; monitoring genetic shifts caused by differential viability in storage or in vitro culture; exploiting associations among traits of interest and genetic mark- ers; and genetic enhancement (Bretting and Widrlechner, 1995; Ayad et al., 1997). One of the most important potential uses of molecular markers is their use in breeding programmes. Identifi cation of genes and markers associated with quantitative traits will greatly increase the effi ciency of a breeding programme.

The use of molecular markers for char- acterization and management of citrus germplasm is currently in its early stages. There are many reports on the development or use of molecular markers in citrus breed- ing, phylogenetic studies, etc., but many of these do not deal with citrus germplasm collections per se. Some of the work that has been done in this area includes the use of ISSR markers to analyse trifoliate acces- sions (Fang et al., 1997), certain mandarin accessions (Fang et al., 1998b) and miscel- laneous accessions (Fang and Roose, 1997), and the use of isozymes, ISSR and simple sequence repeat (SSR) markers to analyse lemon germplasm (Gulsen and Roose, 2001) maintained by the University of California and USDA in Riverside, California. Isozymes have also been used to analyse a large number of accessions of the IVIA

germplasm bank (Herrero et al., 1996a, b) and copia -like retrotransposon sequences to study clementine accessions of this bank (Bretó et al., 2001). Some of these studies suggested that there is less diversity among certain groups of accessions than previ- ously believed, as noted in the discussion of Poncirus above. However, at this point, the technology of molecular analysis of germplasm accessions is not suffi ciently developed to ‘fi ngerprint’ accessions. Many methodologies yield inconsistent results.

These reports deal with an analysis of genetic diversity among groups of acces- sions. Potentially, larger subsets and even entire collections can be analysed. In Riverside, Barkley et al. (2003) surveyed approximately 400 apparently sexually derived accessions using 25 SSR markers. This is an extremely robust data set that, when fully analysed, revealed a great deal about various aspects of the genetic diver- sity and relationships among these acces- sions. The information generated in some cases confi rmed existing ideas derived from other types of observations, and in other cases called this into question. The molecu- lar data have to be looked at in a broad con- text in combination with other types of observations and data.

Molecular markers potentially have a number of other uses in the management of citrus germplasm collections, but are per- haps less developed for these uses than for other crops. This is due to the small size of the worldwide citrus industry as compared with, for instance, wheat and maize. However, these types of uses are bound to increase for citrus. Krueger et al. (2003) and Krueger and Roose (2003) reported the use of various molecular markers to reduce redundancies in potential new accessions received as seed, resolve the identity of mislabelled accessions and detect pathogens. One can conceive of many other potential uses for molecular markers, and their use will undoubtedly increase in the future. They will never replace, but rather will complement, traditional morphologi- cal, horticultural and phytopathological evaluations.

The main limitation today of molecular markers for germplasm management is that they are not able to differentiate and fi nger- print close genotypes within species, par- ticularly in the case of cultivated sweet oranges, clementines, satsumas or lemons, where cultivars have been produced by spontaneous budsport mutations in the fi eld. Most markers used today in citrus are the so-called neutral or anonymous, which are not related to specifi cally known genes. As citrus genomic projects taking place in several countries advance, many sequence- based markers will be available and this probably will allow a more effi cient molec- ular genotyping of most accessions, facili- tating both germplasm management and utilization.

As molecular methodologies improve, identifi cation of redundancies and gaps in collections will improve. A more complete understanding of the relationships between accessions will make it possible to identify more accurately accessions that are com- pletely or basically the same. There is an interaction here between molecular and ‘traditional’ data. In some cases, traditional observations may indicate that accessions are very similar or identical, whereas molecular analysis may reveal differences that may or may not be significant. Conversely, molecular analysis may show differences that have no practical signifi - cance. This is an area of inquiry in which a great deal of progress remains to be made.

Documentation and databases

There is sometimes a tendency to think of documentation of germplasm resources being associated only with ex situ collec- tions, but this is too narrow a viewpoint. Documentation of germplasm resources in the broad sense includes documentation both of the overall status of a particular species and its local status, as well as formal collections. Various publications by entities such as IPGRI, IUCN, various gov- ernmental and non-governmental agencies and organizations, etc. address some of