Species diversity versus phylogenetic diversity: A practical study in the taxonomically difficult genus Dactylorhiza (Orchidaceae)

Abstract

Setting conservation priorities in taxonomically complex groups such as the orchid genus Dactylorhiza is a difficult task. As an alternative to taxonomic diversity, we used here a molecular phylogenetic analysis and the results of a genetic investigation using plastid microsatellites with an extensive geographic sampling to assess in a more objective way the patterns of diversity within this genus. Although western Europe is thought to be the main diversity centre for the genus due to the large number of species found there, we found higher phylogenetic and genetic diversity as well as higher endemicity in the Caucasus and the Mediterranean Basin, two biodiversity hotspots. Species number seems to be correlated with taxonomic effort, tentatively estimated here by the number of herbaria, and is thus biased and not an appropriate measure of diversity. Our results show that phylogenetic analyses and genetic data obtained with molecular tools can offer an alternative measure of biodiversity that is not sensitive to taxonomic inflation. Conservation of allotetraploid taxa is also discussed, and it is recommended that sites in which polyploids are formed should be conserved rather than any specific allotetraploid taxon.

Introduction

Members of the genus Dactylorhiza Necker ex Nevski (Orchidaceae), the spotted and marsh orchids, are terrestrial orchids from the Northern Hemisphere. They occupy a wide range of open habitats from dune slacks to alpine meadows, including swamps and peat bogs. The subtribe to which they belong, Orchidinae, is most diverse in Eurasia, encompassing the majority of European orchids. According to Averyanov (1990), there are 75 species of Dactylorhiza worldwide and 58 in Europe, North Africa and the Near East (hereafter termed Europe and adjacent areas; Delforge, 2001). Dactylorhiza has been shown to be monophyletic if the former monotypic genus Coeloglossum Hartman is included in synonymy (Pridgeon et al., 1997, Cribb and Chase, 2001).

The distribution of Dactylorhiza, including D. viridis Bateman, Pridgeon and Chase formerly Coeloglossum viride Hartman, covers most of Europe, most of temperate Asia, North Africa, Japan, the Aleutian Islands and northern parts of North America (Fig. 1). Averyanov (1990) distinguished three centres of diversity: western Europe (including the British Isles, Germany and southern Scandinavia), the Carpathian Balkan area and Asia Minor (Fig. 2). This is in broad agreement with data given in Delforge (2001), according to whom the greatest species richness is found in northwestern Europe. For instance, nine species are endemic to the British Isles according to Delforge’s classification (2001). Dactylorhiza viridis, the species with the largest range, is the only one to become widespread in the New World (Luer, 1975). Dactylorhiza is thus unusual among European orchid genera, most of which show greatest diversity around the Mediterranean Basin.

Dactylorhiza is universally recognized as a taxonomically challenging genus (Bournérias et al., 1998, Pedersen, 1998, Delforge, 2001, Hedrén, 2001), as demonstrated by the differences in the number of species recognized by different authors (reviewed by Pedersen, 1998), from 12 to 75 worldwide and from 6 to 58 in Europe. There can even be important differences between treatments by the same author. Delforge (1995) for example, added nine species between his monographs of 1995 and 2001. This taxonomic complexity can largely be explained by the frequency of hybridization, and nearly all hybrid combinations are possible (Averyanov, 1990). Most Dactylorhiza species belong to the D. incarnata/maculata polyploid complex, which is composed of three broad groups: D. incarnata s.l., D. maculata s.l. and allotetraploids that are hybrids between the first two groups (Hedrén, 2001). The D. maculata group is itself composed of diploid and tetraploid species, delimitation of which is often difficult.

As with many other terrestrial orchids, populations of Dactylorhiza have decreased due to habitat loss. Many wetlands in Europe have been drained, and changing agricultural practices have led to the degradation of their habitats through use of fertilizers, early haymaking, etc. More recently, the decrease in agricultural pressure has had a counterintuitive effect: abandonment of grassland leads to forest expansion and fewer suitable habitats. However, a few species such as D. fuchsii and D. praetermissa have shown some ability to colonize human disturbed environments, but generally transiently. Another threat to Dactylorhiza is the collection of their tubers to make salep, used as food and medicine. This is a particularly important threat in the Himalayas (Srivastava and Mainera, 1994), where D. hatagirea or “panch aunle” is judged critically endangered (Biodiversity Conservation Prioritisation Project, 1998) due to over-collection. Thus, several species of Dactylorhiza are declining, and some are already protected at a national scale, e.g., in Belgium, Luxembourg, Nepal, and the UK.

Setting conservation priorities in taxonomically complex groups is an essential but especially difficult task because these species tend to be over-represented in red lists (Pilgrim et al., 2004). Hybridization has often made decision-making difficult in conservation (Rieseberg and Gerber, 1995, Wayne and Gittleman, 1995), and neglecting taxonomy can have disastrous effects on the conservation of a particular group, e.g., the tuatara (Daugherty et al., 1990). In the case of Dactylorhiza such problems have already been encountered; D. lapponica, formerly classed as a threatened species in Britain, proved to be indistinguishable from D. traunsteineri (a more frequent species) after morphological and molecular investigations (Bateman, 2001, Pillon et al., in press; Bateman, submitted). Thus, caution should be applied before setting taxon priorities, and molecular systematics can aid in this task (Soltis and Gitzendanner, 1999).

The aim of conservation biology is to preserve biodiversity: “the variety of life in all its manifestation” (Gaston and Spicer, 1998). Species richness is by far the most commonly used measure of biodiversity, but many others also exist (Purvis and Hector, 2000). Some approaches have proposed giving different weight to species because some species are more distinctive and genetically isolated than others, e.g., one species of apomictic Taraxacum may not deserve the same attention as the single species of Welwitschia (Vane-Wright et al., 1991). Because Dactylorhiza species are unequally distant from each other, some being barely distinguishable genetically and others relatively isolated, we thought that evolutionary history or phylogenetic diversity could be a better measure of the diversity of a region (Faith, 1992, Mace et al., 2003) than purely taxonomic diversity. A hierarchical taxic weighting approach (Vane-Wright et al., 1991) cannot be applied to Dactylorhiza because of reticulate evolution. Thus, we propose here to use neutral molecular markers to assess global diversity distribution within Dactylorhiza. Rate of evolution in DNA sequences is known to vary among lineages (e.g., Soltis et al., 2002), but genetic distances between taxa are nevertheless correlated with the time of their divergence and thus can act as a surrogate for genetic, morphological and biochemical distinctiveness. Here, the taxonomic status of the taxa in questions is not clear, so we do not know if we should be considering the variation we detect to be inter-or intraspecific. Previous studies have shown than molecular markers such as DNA sequences, plastid microsatellites and AFLPs are often linked with morphological (Barraclough and Savolainen, 2001, Rodríguez et al., 2003, Shipunov et al., 2004) or ecological (Kelly et al., 2003) diversity of species, but see e.g., Bonnin et al., 1996, Hamrick et al., 1991 for situations in which historical change is a better predictor of genetic diversity within populations.