Subproject E15

Dr. Christoph Oberprieler & Dipl. Biol. Jörg Meister, Institut für Geobotanik und Botanischer Garten, Martin-Luther-Universität Halle-Wittenberg, Neuwerk 21, 06108 Halle (Saale), from 1.9.2001: ZE Botanischer Garten und Botanisches Museum Berlin-Dahlem, Freie Universität Berlin, Königin-Luise-Str. 6-8, 14191 Berlin, Germany

Phylogeographical and population genetic aspects of phytodiversity in paleo-African refugia in the coastal mountains of southern Yemen and on Socotra (Republic of Yemen)

Subproject E15 undertakes population genetic analyses using molecular markers which are aimed at a reconstruction of the chronological fragmentation of the relict habitats under study. Their intention is to elaborate a hypothesis of the temporal and spatial development of these habitats and will contribute to a discussion of aspects of taxonomic differentiation, floristic evolution (together with subproject E13), functional morphological adaptations (together with subproject E14), and enables the comparison of phylogeographic patterns in plants and animals of S Arabia (together with subproject E16). Population genetic studies of the partial project E15 aim at an assessment of effects of areal fragmentation on the genetic diversity of plants in semi-arid habitats. The taxa under study have been chosen in collaboration with E13 to satisfy the following demands: the taxa should represent a variety of different families of flowering plants, different life forms (annuals, perennials), and reproductive types (autogamous, xenogamous). Under these experimental conditions, the phylogeographical and population genetic analyses will show whether all taxa support the same underlying biogeographical pattern or whether they represent different scenarios of temporal and spatial development (fragmentation of an originally connected area vs. invasion by long-distance dispersal after fragmentation).
Plant species under study:

•	Gossypium stocksii / G. incanum / G. aresianum (Malvaceae) These three species (together with the N Somalian G. somalense) form a closely knit group of species vicariantly distributed in the refugia under study. While G. stocksii is characterised by a disjunct areal in S Arabia and Pakistan, the other two species are endemics in the study area.
•	Maytenus senegalensis / M. dhofariensis (Celastraceae) These two shrub species also form a closely related group characterised by vicariance. While M. senegalensis is widespread throughout S and W Arabia and Africa, M. dhofariensis is endemic to the Dhofar and SE Yemen. The two species are the essential host plant for the development of the diurnal moth Reissita simonyi studied in subproject E16. These circumstances will allow joint phylogeographical analyses to find correlations between faunistic and floristic data.
•	Launaea crassifolia (Compositae) This annual or short-lived perennial plant species is endemic to S Arabia, Socotra, and N Somalia and shows a considerable variation in morphological respects. While plants in most populations are characterised by wind-dispersed fruits, some populations in S Arabia and Socotra are formed by plants with unairworthy achenes. Therefore, this species is especially interesting to study correlations between morpho-functional adaptation, genetic diversity, and areal fragmentation.
•	Euryops arabicus (Compositae) This afrotropical shrub species is distributed in Somalia, Djibouti, Socotra, and W and S Arabia. Since it is not restricted to the refugia under study, we will use this species as a tool in the phylogeographic studies to provide a geographically more resolved hypothesis of the temporal and spatial differentiation of the S Arabian flora.
•	Euclea schimperi (Ebenaceae) This afrotropical tree is characteristic for the montane sclerophyllous woodlands of Ethiopia, Eritrea, S Sudan, and the mountain ranges in S and W Arabia. On Socotra it is replaced by the two endemic species E. balfourii and E. laurina which will be also included in our sampling. This species represents the long-lived life form of a tree in the study.
•	Justicia areysiana (Acanthaceae) A mesic shrub endemic to the coastal mountains of S Arabia. It occurs in thickets of the escarpment woodlands and is very scatteredly distributed between Jebel Urays in Abyan, SW Yemen, and Jebel Samhan in Dhofar, Oman, providing a good example for a strongly fragmented distribution.

Methods (briefly):
For the phylogeographical questions of our subproject we will use the method of polymerase chain reaction - restriction fragment length polymorphism (PCR-RFLP) to avoid extensive sequencing of molecular markers. In a first step, for each species under study a set of intergenic spacers of the chloroplast genome will be amplified using PCR. Following amplification, the DNA fragments are then individually digested with 4-6 restriction enzymes with a four-base recognition site, the products are analysed electrophoretically and screened for any eventual polymorphism between samples. After screening for suitable, species-specific combinations of PCR-amplified markers and restriction enzymes and the following definition of species-specific chloroplast haplotypes detectable by these combinations, all collected plants from populations and refugia under study (and additional herbarium specimens from outside the study area) will be genotyped (here: haplotyped) in order to get a spatial representation of haplotypes. Cladistic analyses will lead to phylogenetic interpretations of haplotype relationships and will be used in conjunction with the geographical arrangement of these haplotypes to allow the reconstruction of the chronological and spatial differentiation of the areal for each species or species group under study. In a final step, we are planning to compare with each other the phylogeographical reconstructions of all species under study to generate more general hypotheses on the fragmentation of the relict habitats.

For the population genetic analyses which aim at the assessment of effects of fragmentation on the genetic diversity of the species under study, we plan to use the method of amplified fragment length polymorphism (AFLP). This method provides individual-specific genetic fingerprints which can be analysed in order to estimate the inter- and intrapopulational genetic variability (e.g. by an analysis of molecular variance, AMOVA) and its dependence from factors like population size, geographical arrangement of populations (e.g. occurrence of metapopulations), and life history traits (e.g. life form, reproductive types). Results of these analyses may help in the discussion of minimum sizes of populations for the persistent conservation of the flora in the refugias concerned.