{short description of image}

The Application of Genetic Fingerprinting Techniques to Plant Conservation Problems

Fay, M.F., Richardson, J.E., Cowan, R.S. & Stranc, P.A.
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK
Smith College, Northampton, MS 01963, USA

Home | Contents | Abstract | Introduction | Case Histories | Medusagyne Oppositifolia | Endemic Rhamnaceae from St Helena | Tulipa Sprengeri | Conclusions | Acknowledgements | References


The advent of molecular techniques has had a major effect in the way in which questions concerning plant conservation can be approached. In particular, the development of PCR allows the use of minute quantities of plant tissue and makes DNA sequencing and genetic fingerprinting studies possible even when the amount of material of a rare species is extremely limited. At Kew, we utilise sequencing techniques at the phylogenetic level to clarify relationships of taxa and to identify isolated lineages of high conservation importance and fingerprinting techniques (predominantly AFLPs) to look at population-level questions. Information gained from these studies can then be incorporated into management strategies for conservation. This paper is illustrated with examples of work from Kew investigating Medusagyne oppositifolia (Medusagynaceae), Nesiota elliptica and Phylica polifolia (Rhamnaceae) and Tulipa sprengeri (Liliaceae).

{short description of image}


Botanic gardens have long played an important role in both in situ and ex situ conservation, and a glance through the titles of papers presented at the Fourth International Botanic Gardens Conservation Congress in 1995 (Touchell & Dixon, 1997) will show the wide ranging nature of this role.

One area is the study of genetic variation in rare species, and the number of techniques available is increasing rapidly. Until recently, allozymes were the main tool available for such studies, but in the last decade, DNA-based techniques for examining the relationships of plants to each other at various levels have become widely available and are being used to answer questions relating to taxonomy, hybridisation, conservation etc. The background to the development of these techniques has been reviewed in several recent publications and will not be discussed in detail here. Further details can be found in Schaal et al. (1991), Weising et al. (1995), Fay & Chase (1997) and Qamaruz-Zaman et al. (1998a). In this paper we will discuss the use of two techniques in conservation studies with case histories from the work carried out at Kew.

DNA sequencing has been widely used in phylogenetic studies and is an increasingly important tool in plant classification, being the basis of the new system put forward by the Angiosperm Phylogeny Group (1998). In a conservation context it is useful in the identification of distinct lineages, and we believe that in a world of limited resources isolated lineages should be accorded higher priority for conservation. Thus an endangered monotypic family such as Medusagynaceae (see below) has greater significance in terms of biodiversity than a rare subspecies of an otherwise common species, and this significance should be reflected in the establishment of conservation priorities (Vane-Wright et al., 1991). DNA sequencing allows us to investigate these issues by placing narrow endemics within a larger context (Fay et al., 1996; Fay & Chase, 1997).

Methods of sequencing DNA have evolved rapidly over the last decade, and most studies now involve the use of automated sequencers, removing the requirement for radioactive labeling of the DNA. The different approaches are described in a recent book and will not be detailed here (Karp et al., 1998).

Different DNA loci evolve at different rates and so it is important to choose a locus that evolves at a rate appropriate to the question being asked. Thus different loci can be used to investigate relationships at different taxonomic levels. Slowly evolving genes such as rbcL (coding for the large subunit of the photosynthetic enzyme rubisco) have been used extensively to investigate relationships between genera, families and orders (e.g. Chase, Soltis, Olmstead et al., 1993), whereas more rapidly evolving regions (often non-coding) such as the internal transcribed spacer (ITS) of nuclear ribosomal DNA have been used to study relationships at the interspecific and intergeneric levels (e.g. Baldwin, 1992). In the studies described here, we have used rbcL and ITS, in additional to two non-transcribed regions of plastid DNA (the trnL-F region and the rpl16 intron). further details about these loci can be found in papers by Taberlet et al. (1991) and Jordan et al., 1996.

The second technique to be discussed here is a method of DNA fingerprinting called amplified fragment length polymorphisms (AFLPs). This is a recently developed technique (Vos et al., 1995) and, since the Fourth International Botanic Gardens Conservation Congress in 1995, it has become one of the more widely used methods of DNA fingerprinting of crops and wild species. This and other methods for genetic fingerprinting are discussed in various other publications, to which we refer the reader (Weising et al., 1995; Fay & Chase, 1997; Karp et al., 1998; Qamaruz-Zaman et al., 1998a).

For AFLPs, genomic DNA is restricted with two different restriction endonucleases and then a subset of these are amplified using a modified polymerase chain reaction (PCR) and visualised using radioactivity, silver staining or fluorescent dyes for use with an automated sequencer. AFLPs have a number of advantages, detailed in Qamaruz-Zaman et al. (1998a), over pre-existing techniques. The main advantages for conservation studies are:

  1. only small quantities of DNA are required because the technique is based on the PCR, meaning that tiny quantities of plant material are required,
  2. the fingerprint traces are highly reproducible and consist of many markers, allowing greater discernment between closely related plants than other techniques including RAPDs and microsatellites.
As a result of these advantages, AFLPs have now been used widely in conservation-related projects on a range of different species. Examples include Astragalus cremnophylax (Travis et al., 1996), Orchis simia (Qamaruz-Zaman et al., 1998b), Populus nigra var. betulifolia (Winfield et al., 1998), and Populus euphratica (Fay et al., in press).

{short description of image}

Case Histories

Medusagyne Oppositifolia (Medusagynaceae)

Medusagyne oppositifolia is the sole species in the family Medusagynaceae, the affinities of which have been obscure. An endemic from the Seychelles, it is only found on the island of Mahé, where it is known as the jellyfish tree or bois méduse. The botanical and common names refer to the gynoecium which resembles a gorgon’s head or the umbrellashaped larval stage of a jellyfish due to the presence of up to 25 stalked capitate stigmas. The unusual mode of dehiscence of the valves of the septicidal capsule from the base results in a structure at maturity that resembles the ribs of an umbrella. Medusagyne oppositifolia was assumed extinct from 1903 until its rediscovery in 1970 (Robertson et al., 1989).

Originally described by Baker (1877) as a member of Ternstroemiaceae, M. oppositifolia was judged to be the sole member of a separate family by Engler & Gilg (1924). A fuller account of Medusagynaceae was published by Engler & Melchior (1925), and the family has been recognised in all subsequent systems of classification (Dickison, 1990a). Affinities have been suggested to, among others, Caryocaraceae, Clusiaceae, Eucryphiaceae, Ochnaceae, Paracryphiaceae and Quiinaceae (Robertson et al.; 1989, Dickison, 1990a, b). Molecular studies carried out at Kew (Fay & Chase, 1996; Fay et al., 1997) have shown that Medusagynaceae is most closely related to Ochnaceae and Quiinaceae, but that it is distinct from both. These molecular results indicate that M. oppositifolia is a relict palaeoendemic, and it should be given high priority for conservation.

The present habitats occupied by this species indicate that it originally flourished in exposed areas prone to drought. There has been no recent natural regeneration from seed, all plants apparently being a similar age (Wise, 1998). Propagation of the species is difficult, and, although it has been achieved from seed at Nancy and in tissue cultures derived from seed at Kew, it has thus far not proved possible to generate large numbers of plant ex situ. Given these problems, M. oppositifolia is in severe danger of extinction.

Medusagyne oppositifolia is only known in cultivation from three collections, one at Kew, donated by Whitehead in 1981, and two at Nancy, France, donated by Friedman in 1983 and 1984. However, it was unclear whether these collections are distinct from the remaining trees on Mahé, and consequently how significant they might be in the conservation of M. oppositifolia. To address the questions relating to the origin and distinctness of the cultivated plants and the level of genetic variation existing in this species we carried out genetic fingerprinting studies on cultivated material from Kew and Nancy and samples from the wild populations (Fay et al., 1999). The samples from wild plants, despite the fact that they came from three separate locations, formed a cluster of closely related genotypes. The material from Kew was closely related to this cluster, whereas those from Nancy represented distinct genotypes. While this could be an artefact of the level of sampling of wild material, the material in cultivation at Nancy and Kew should be treated as an important ex situ resource in the conservation of this important species. In particular, seeds should be collected from the plants at Nancy and stored in a seedbank as a resource for potential reintroduction programmes.

{short description of image}

Endemic Rhamnaceae from St Helena

The island of St Helena in the South Atlantic is famous for its flora, with more than 40 endemic taxa still in existence, and one species (Mellissia begonifolia, Solanaceae; see http://www.rbge.org.uk/news/boxwood.htm), thought to be extinct since the late 19th century, was rediscovered recently. However, largely due to human influence on the vegetation of the island, many of the endemic taxa are on the verge of extinction and all are rare or threatened.

Two species of Rhamnaceae, Nesiota elliptica and Phylica polifolia, are endemic to the island. There is only one species of Nesiota and therefore the genus is also endemic. There are questions at phylogenetic and population genetic levels and we have addressed both levels using molecular techniques (Richardson, 1999; Richardson et al., in press; Richardson et al., in prep.). Taxa thought to be close relatives of both species occur in South Africa and on other oceanic islands including Tristan da Cunha, New Amsterdam, Mauritius and Réunion. The first level of questions relates to the relationships of the two St Helenan species to each other and to the possible close relatives.

Are the two St Helenan taxa each other’s closest relatives? Is Nesiota a distinct genus? How distinct are these taxa from their closest relatives? Nesiota elliptica is known as the St Helena Olive. It is a small tree, once known from localised populations on the highest parts of the eastern central ridge, and its range and population size have probably always been restricted (Cronk, pers. comm.). Melliss (1875) found no more than 12-15 plants in existence, and the species was presumed extinct until 1977 when George Benjamin discovered a single tree near Diana’s Peak (Cronk, 1987). This wild tree died in 1994, and the species is therefore given the status EW (extinct in the wild). Before it died, attempts were made to propagate it, but few viable seeds were set despite many hand pollinations and only one cutting was successfully rooted. Attempts at micropropagation proved unsuccessful due to systemic fungal contamination with at least 14 species of fungi being isolated from the wild plant (Fay, 1989). A study of the genetic diversity of this species was considered desirable to assess its conservation genetic status. Samples of DNA from the cutting and the three existing seedlings were included in this analysis. We expected that the seedlings would show some variation from the cutting due to segregation at heterozygous loci. A genetic fingerprinting study using AFLPs was carried out to test these ideas.

The second species, P. polifolia, is known as wild rosemary. Melliss (1875) recorded that only 100 plants were remaining. It is now critically endangered with a CR C2a status (Oldfield et al., 1998). In the wild there are about 50 recorded plants remaining, and plants held ex situ include two plants on the island and material at Kew. There are phenotypic differences between the remaining populations on St Helena (Cairns-Wicks, pers. comm.) with individuals at Lot having a more upright growth form than the prostrate High Hill individuals. We wanted to investigate whether the phenotypic differences were reflected in the genetics of these plants, because they might be adapted to the areas where they occur. If this were the case, reintroduction of individuals from other areas could lead to loss of fitness in the populations. In this situation, seed orchards from the two populations would be best kept separate.

In the phylogenetic studies using DNA sequence data for two loci (rbcL and trnL-F), P. polifolia was shown to be closely related to P. arborea from the Tristan da Cunha archipelago and New Amsterdam and P. paniculata, a widespread species in South Africa. It is most likely to be a neoendemic, representing a relatively recent colonisation event of St Helena from South Africa. In contrast, N. elliptica is not closely related to the island species of Phylica, but is an isolated taxon, whose closest relative appears to be Trichocephalus stipularis, another South African member of Rhamnaceae sometimes included in Phylica. Thus N. elliptica appears to be a relict palaeoendemic, representing a separate, much earlier colonisation event. It has no extant close relatives and is best treated as a separate genus (Richardson, 1999; Richardson et al., in press; Richardson et al., in prep.). On the basis of these results N. elliptica should be treated as a high priority for conservation as it is the sole representative of an isolated lineage.

The results from the fingerprinting studies for the two species were also strongly contrasting. Despite the fact that three of the samples of N. elliptica were seedlings and therefore expected to show some degree of genetic variation from the cutting from the last wild tree, the fingerprint traces for the four samples were indistinguishable. Where similar situations have been shown in other species, this has been used to demonstrate clonal origin of populations e.g. in Salix spp. (Beissmann et al., 1998) and Populus euphratica (Fay et al., in press). In the case of N. elliptica, the plants are known to be seedlings and these results are therefore surprising. These results suggest either that this species has been inbreeding for several generations, with a resultant loss in heterozygosity, or that the seeds were produced apomictically. The first possibility is unlikely given the difficulty in obtaining seeds from self pollination (Jackson and Cairns-Wicks, pers. comm.). These results do not detract from the importance of N. elliptica as an isolated genetic resource, but may well indicate that it is too late to save this species from extinction.

With P. polifolia, the fingerprint traces were variable both within and between populations. The Lot population formed a distinct cluster of genotypes in the analyses, whereas the High Hill population appeared to consist of a number of closely related individuals and one or more outliers. Thus while the population at High Hill is more genetically variable than that at Lot, the latter adds to the overall genetic variability of the species and should be considered an important resource. Any ex situ conservation or reintroduction strategy should take account of the distinctness of the two populations, and care should be taken not to disrupt any genetic adaptation that the remaining plants have to their environment.

Genetic diversity measures may indicate which taxa will have a better chance of long term survival. In these studies, the phylogenetic analysis identified N. elliptica as an endangered palaeoendemic taxon and P. polifolia as a member of a larger, more recently derived group that also contains some other endangered taxa. Vane-Wright et al. (1991) suggested that taxa that are palaeoendemic or phylogenetically isolated should be given priority for conservation. On this basis, N. elliptica should be given a higher conservation priority than P. polifolia. However, further factors regarding the long-term survival chances of N. elliptica have to be taken into consideration before embarking on conservation programmes. The considerable efforts to increase the numbers of individuals of N. elliptica have so far proved relatively unsuccessful for reasons mentioned above. Even if propagation were successful, the long-term chances of survival of the species would be in doubt due to the lack of genetic variation detected. Similar efforts expended on P. polifolia are more likely to lead to successful restoration given the greater levels of genetic variation that this taxon exhibits. These studies illustrate the way in which the molecular techniques allow the parties involved to make important policy decisions on a sound scientific footing. however, it is a clear case of ‘the agony of choice’ referred to by Vane-Wright et al. (1991).

Tulipa Sprengeri

The genus Tulipa is generally split into two main sections, although some species do not fit clearly into either (Hall, 1940). Section Eriostemones is characterised by the presence of a boss clothed with short hairs at the base of the stamen filaments. The section includes many of the familiar ‘species tulips’ such as T. saxatilis, T. batalinii, and T. tarda in addition to the ‘wild tulip’ (T. sylvestris). Section Leiostemones consists of species which lack the hairy boss on the stamen filaments and include such species as T. gesneriana, the source of many commercial cultivars. Species in both these sections, lack a true style, although the ovary can be somewhat tapered in some species. In addition to these two sections, several small-flowered species which possess a true style are placed in section Orithyia. Finally, there are some species which do not fit neatly into any of these sections, leading to different treatments by different authors. One of these is T. sprengeri, described by Baker in 1894.

In the original description of T. sprengeri Baker stated that the stamen filaments were “without any basal tuft of hair”, i.e. it was a member of Leiostemones. However, in another article (Baker, 1896/7), he included T. sprengeri in a list of tulips with hairy stamens i.e. Eriostemones. Hall (1940) reported that the filaments were swollen into a boss at the base, but did not comment on the presence or absence of hairs. He stated that T. sprengeri was one of two species (the other being T. schmidtii) “which cannot be associated with any of the sections previously described, which again have no affinities to one another but appear to stand alone..... T. sprengeri from Asia Minor is unique in the character of its filaments and stamens, as well as in its general aspect”. Our observations on material at Kew show it to have a boss at the base of the filaments which lacks the tuft of hairs. Thus it combines the characteristics of the two sections.

Tulipa sprengeri was brought into cultivation in the late 19th century by Damann & Co. of Naples from Amasya in Turkey. It is recorded as having been collected until the outbreak of the 1st World War (Hoog, quoted in Hall, 1940), but no new stocks of wild origin appear to have come into cultivation after that, and it is thought to be extinct in the wild. In cultivation, it is considered highly desirable and commands high prices when it is commercially available (Brickell & Sharman, 1986). Much of the material in cultivation is believed to have come from the stocks held at Kew, where it has been grown since early this century. It naturalises there, propagating itself by seed and by bulb. It is also held by other botanic gardens The aims of this project were to establish the relationships of T. sprengeri to other species of Tulipa and to investigate the distribution of genetic variation in the cultivated material, so that maximum benefit can be gained from the remaining genetic diversity in future repatriation and reintroduction studies.

The sequencing study showed that the level of sequence divergence in the genus Tulipa is low, indicating that all the species studied are relatively closely related. Several species of section Eriostemones were indistinguishable using the locus used here (rpl16). Within Tulipa, section Eriostemones formed a distinct group. Rather than being a distinct species, T. sprengeri fell within this group, and it should be treated as a member of Eriostemones. The presence of a boss at the base of the stamen filaments thus appears to be more significant in the recognition of this section than the presence of hairs. In contrast, section Leiostemones did not form a distinct group, but the level of variation between these species was too low for a definite conclusion to be drawn. Further data will be necessary to clarify this point.

In the fingerprinting study, approximately two thirds of the bands were invariant. The maximum genetic distance between samples of T. sprengeri was approximately 12%. Individuals within several of the groups were genetically indistinguishable from each other, indicating that they are probably vegetative propagules from the same original bulbs. All the samples from other botanic gardens fell within the range of variation shown by plants at Kew.

From these studies, it appears that T. sprengeri is closely related to other members of the section Eriostemones, rather than being an isolated species within the genus. The material held at Kew appears to be as variable as that in other gardens tested so far. This is in contrast to the situation with some other plants, including Medusagyne oppositifolia (see above), where the level of genetic variability shown by plants at Kew is increased by the inclusion of plants from elsewhere. Much of the material of T. sprengeri in cultivation elsewhere may indeed come from Kew as has been suggested.

We intend to include other accessions of T. sprengeri as these become available for inclusion in the fingerprinting study to ascertain whether they represent different genetic lines. In the meantime, if material of T. sprengeri is to be repatriated to Turkey, our results indicate that a range of material taken from the Rock Garden at Kew will cover the range of available genetic variation.

{short description of image}


We hope that these examples show how research projects using molecular techniques can have a real contribution to the conservation of rare and endangered species. Techniques that explore relationships at different levels allow us to add to the information base about these species and this can significantly affect conservation policies and management of the remaining genepool. As new techniques continue to be developed, the role of molecular studies and our understanding of the processes involved in management of endangered species will only increase.

{short description of image}


We acknowledge the financial support of the Friends of Kew, through the Threatened Plants Appeal, without which some of these projects would not have been possible, and the Muriel Kohn Pokross Travelling Fellowship and the Howard Hughes III Internship Fund of Smith College, Massachusetts for providing financial support to PS. We also wish to thank the numerous individuals and institutions who have provided plant material for comparative purposes, Gloria Beltran, Jeff Joseph and Martyn Powell for technical support, the horticulturists who grow these rare plants at Kew and Mark Chase for useful comments on this paper and his ongoing support of the Conservation Genetics Programme

{short description of image}



{short description of image}

Copyright 1999 NBI