Charles Darwin was the first to recognize that orchids offer superb cases for illuminating evolutionary processes, devoting an entire volume to the Various Contrivances by Which Orchids Are Fertilized by Insects (Darwin 1877). Given this head start, and their extraordinary diversity, it might be expected that orchids would be a major target of evolutionary biology investigations. Interestingly, more than two decades ago, Dressler (1981) lamented that orchids had not received the research attention they deserved in any field. A database search suggests this remains true today. Figure 1 indicates that orchids remain under-represented in studies of speciation, evolution and reproductive isolation as well as in the development of molecular markers relative to other species rich plant families. A similar pattern of under-representation is apparent in the number of papers published on orchids in Molecular Ecology, and the number of primer notes for orchids published in Molecular Ecology Notes (Fig. 2).
Despite the under-representation of orchid research, several recent comprehensive and insightful reviews have summarized the current state of orchid research and explored hypotheses for why there are so many species of orchids. Cozzolino & Widmer (2005) propose that deceptive pollination is one of the keys to orchid floral and species diversity, but they also call for a re-assessment of the importance of pollinator specificity for prezygotic isolation in orchids. Tremblay et al. (2005) propose that sequential and rapid interplay between drift and natural selection, combined with the low fruiting success typical of orchids (especially deceptive species), may account for the diversity. Otero & Flanagan (2006) have noted that the role of obligate orchid–mycorrhizal interactions should not be neglected in consideration of the factors promoting speciation. Gravendeel et al. (2004) provide evidence that suggests epiphytism, rather than pollinator specialization (based on crude pollinator data), has stimulated the development of orchid species richness. This diversity of views indicates that a great deal remains to be learnt about the mechanisms of speciation in the orchids.
Moccia et al. (2007) have explored hybridization between two closely related and morphologically similar sister species of Italian food-deceptive orchids, Anacamptis morio and Anacamptis papilionacea (Fig. 3). The study exploited a ‘natural laboratory’ consisting of two strategically chosen sites of different geological age near Naples in Italy. The authors applied a combination of morphological and genetic analyses to assess the frequency of hybridization among the two study species. Species-specific nuclear ITS (internal transcribed spacer) polymorphism enabled identification of hybrids, while a polymorphic chloroplast DNA sequence identified the maternal lineage. Additionally, 100 polymorphic amplified fragment length polymorphism (AFLP) loci were screened and a molecular hybrid index developed. The molecular identification of hybridization was complemented by morphological analysis that revealed congruence between the two methods.
Numerous hybrid individuals were detected at the study sites with both parental species contributing maternal lineages. Surprisingly, the AFLP analysis indicated that the existing hybrids were exclusively first generation (F1). No evidence of introgression into the parental populations was found. An innovative bulk seed DNA extraction that circumvented the difficulty of seed germination revealed that hybridization is an ongoing process with approximately 7% of the 340 fruits assayed of hybrid origin. Consistent with the molecular findings, both artificial and natural F1 crosses produced viable seed, similar to the parents, but F2 and backcrosses effectively failed to set seed. The finding of only F1 hybrids cannot be explained by a limited geological time since co-occurrence of the parents for the establishment of later generation hybrids. The more likely explanation is indicated by a chromosome number difference between the species (2n = 32 vs. 2n = 36) that points to meiotic irregularity as the basis for strong postzygotic isolation in the hybrids (Moccia et al. 2007).
Both the frequency and ongoing nature of the hybridization suggests that pollinator-mediated reproductive isolation plays little or no part in this food-deceptive system. However, Moccia et al. (2007) found lower than average pollinator visitation among the hybrids, despite the morphological similarity of the parents and the known sharing of pollinators. Consequently, they suggest that pollinator-mediated reproductive isolation may exist at the hybrid stage after all, and in combination with hybrid sterility results in strong reproductive isolation among the parents. If pollinators do discriminate against hybrids, it will be of interest to determine the basis of this discrimination in the future.
The key findings of this study of food-deceptive orchids would not have been revealed without the combination of well-established techniques of field observation, hand pollination and morphological analysis in combination with nuclear and chloroplast molecular markers. The strategic choice of study sites based on knowledge of the underlying geological history and a keen eye for recognizing an ideal ‘natural laboratory’ experiment also underpin the success of this study. Finally, the best explanation for the severe depression of fitness stems from the cytological investigation.
The crucial multidisciplinary combination of old and new tools in the study of Moccia et al. (2007) offers a timely reminder that molecular analysis should complement rather than replace the seemingly ‘old fashioned’ fields of natural history, pollination ecology and cytology in studies of evolutionary ecology. Noor & Feder (2006) make a similar point, concluding in their formative review on the evolving approaches in speciation genetics that while modern molecular and genomics approaches are accelerating the pace and scale of data acquisition these fields have not yet (on their own) led to major conceptual advances. They further emphasize that future progress in understanding speciation is dependent on merging old and new tools in combination with creative imagination of researchers.
As recognized so long ago by Darwin, a better understanding of speciation mechanisms within the orchids will be illuminating for evolutionary biology more generally. It is therefore imperative that we take full advantage of the unparalleled diversity and extraordinary plant–animal interactions within the family (Schiestl 2005). Significant challenges will continue to confront orchid researchers. Not the least of these is the task of obtaining funding in a group for which a case of applied scientific outcomes is often difficult to make. A related challenge is that there is no ‘model species’ to exploit nor an orchid genome project or other major genomic resources in place [although the recently published full cpDNA sequence of Phalaenopsis aphrodite (Chang et al. 2006) is a positive step forward]. Peculiarities of the orchids themselves, such as deceptive pollination, low fruiting success, hyperdispersion and dependence on mycorrhiza also pose major challenges. Nonetheless, as aptly illustrated by Moccia et al. (2007), some of these challenges can be overcome by the creative integration of molecular and ecological tools combined with the strategic choice of target species and study sites.