Arnold's new book on hybridization[1xSee all References][1], recently reviewed by Ritchie in TREE[2xRitchie, M. Trends Ecol. Evol. 1998; 13: 123–124Abstract | Full Text PDFSee all References][2], has brought attention to an old, controversial but revitalized topic in evolutionary biology. The two reviews[2xRitchie, M. Trends Ecol. Evol. 1998; 13: 123–124Abstract | Full Text PDFSee all References, 3xBarton, N.H. Genet. Res. 1997; 70: 178–180CrossRefSee all References]we have read share a sceptical attitude towards studies of hybridization lying outside the hybrid zone theory. They consider Arnold's book too `opinionated'[2xRitchie, M. Trends Ecol. Evol. 1998; 13: 123–124Abstract | Full Text PDFSee all References][2]and `an argument for a greater emphasis on the positive role of hybridization in evolution' rather than a `comprehensive review'[3xBarton, N.H. Genet. Res. 1997; 70: 178–180CrossRefSee all References][3]. However, one of the merits of the book is that it devotes a great deal of effort towards reconciling divergent approaches to the topic. Arnold dedicates equal importance to animal and plant studies, as acknowledged by Ritchie[2xRitchie, M. Trends Ecol. Evol. 1998; 13: 123–124Abstract | Full Text PDFSee all References][2], and also puts much emphasis on the analysis of hybrid zones. In fact, when proposing a model for the birth of new evolutionary hybrid lineages he places his `new conceptual framework (the evolutionary novelty model)' within the hybrid zone framework. Our criticism of this proposal is that it may be too rigid to fit scenarios departing from the specific model of hybrid speciation proposed by Grant[4xSee all References][4], that is, recombinational speciation.In our opinion, relying exclusively on the hybrid zone framework to assess the role of hybrids in evolution is misleading. Hybrid zones usually imply relatively recent events and species with strong reproductive barriers. The tension zone model assumes that hybrid zones are maintained by a balance between selection against hybrid individuals and dispersal of parental individuals into the hybrid zone[5xBarton, N.H. and Hewitt, G.M. Annu. Rev. Ecol. Syst. 1985; 16: 113–148CrossRefSee all References][5]. As we allow other factors (basically, interactions between the hybrids and a heterogeneous environment) to play a role in the model, we move into the other interpretation of hybrid zones. The latter, often called the bounded-hybrid-superiority model, seems more accurate—at least when applied to plants[6xWang, H. et al. Evolution. 1997; 51: 95–102CrossRefSee all References][6]. Yet, dedicating much effort to studying the genetic and ecological factors involved in the maintenance of hybrid zones does not justify neglecting the ample molecular and morphological evidence available.From a pattern-oriented approach, it is true that reticulation may preclude a ready interpretation of results and that alternative explanations, such as `lineage sorting following speciation in a polymorphic ancestor', are sometimes possible. Adopting a phylogenetic approach has risks and difficulties, some of which are inherent to the subject of hybridization: (1) phylogenetic inference methods are designed to detect hierarchical structure resulting from divergence; (2) no single predictable pattern can undeniably reveal the occurrence of reticulation. However, a combination of different sources of data may be conclusive to document reticulate evolution as the references gathered in Arnold's book convincingly show. Certainly, the difficulties involved in a phylogenetic approach do not justify Ritchie's untenable statement that `most of the evidence of the significance of hybridization relies on well studied hybrid zones'[2xRitchie, M. Trends Ecol. Evol. 1998; 13: 123–124Abstract | Full Text PDFSee all References][2]nor his conclusion that he `need[s] more persuading that hybridization is a creative force in speciation before [he]can recommend this [book] as essential primer reading'.We do not claim that the pattern-versus-process dichotomy is all that matters in the way the impact of hybridization is perceived. Process-oriented studies have produced insights both at the genomic and populational levels. Our point is that this kind of study, as exemplified by the excellent research on sunflowers[7xRieseberg, L.H. et al. Science. 1996; 272: 741–745CrossRef | PubMedSee all References][7]and Galapagos finches[8xGrant, P.R. and Grant, B.R. Science. 1992; 256: 193–197CrossRef | PubMedSee all References][8], need not be immersed in the hybrid zone paradigm.The hybrid zone approach is a critical avenue to study the initial stages of the formation of hybrid lineages under certain circumstances. However, relying exclusively on studies framed by the hybrid zone theory is insufficient for a comprehensive perception of a complex phenomenon like hybridization. Furthermore, it is certainly inadequate to test or document the occurrence of past events of reticulation between divergent lineages.
Read full abstract