So that is how species are defined; but how do new species form? It is abundantly clear that genetic differences among populations, along with the potential for reproductive isolation, can be enhanced by long periods of geographical separation—'allopatry'. An early advocate of the importance of significant geographical barriers in driving spe-ciation was German naturalist and explorer Moritz Wagner, who, working before the publication of Origin of Species, was struck by the distribution of different species of wingless darkling beetles on different sides of rivers.39-41 Likewise, Alfred Russel Wallace noted that the geographic boundaries of primate and bird species in Amazonia tended to coincide with major rivers.42,43 These ideas have become incorporated into a rule that has become known as Jordan's law, although its namesake David Starr Jordan was quick to recognize the priority of his predecessors44:
Given any species, in any region, the nearest related species is not to be found in the same region nor in a remote region, but in a neighboring district separated from the first by a barrier of some sort or at least by a belt of country, the breadth of which gives the effect of a barrier.45
Several different versions of allopatric speciation have been presented, and all, by definition, involve long periods with no gene flow between the populations of interest due to some form of spatial separation, or temporal separation of reproducing forms. For example, 'vicariant speciation' (derived from the Latin for 'interchange') is said to occur when a single reasonably large population becomes split into two or more populations due to the appearance of a significant barrier to dispersal such as an ice sheet, a desert, a river, or mountain chain. Many examples of vicariant allopatric spe-ciation have been proposed, including the different species of snapping shrimp in the Caribbean and Pacific that arose when the population was cleaved apart following the gradual formation of the Isthmus of Panama about 3-10 million years ago46,47 and the evolution of related seaweed species in the North Pacific and North Atlantic Oceans.48 These examples and others clearly show how new species can evolve following sufficient periods of isolation of separated populations.
In the 'peripatric' version of allopatric speciation, a few dispersers by chance overcome a major geographical barrier such that at least one of the separated populations is small. A fascinating recent example comes from the study of lupins (genus Lupinus) in the high elevation Andes in South America.49 Over 80 related species of Lupinus are found in this area and almost certainly arose as a consequence of the uplift of the northern Andes, effectively creating competitor-free islands that could be colonized by lupins, with populations gradually diverging from one another over time.49 Perhaps the best-known examples of peripatric speciation, however, come from oceanic islands. Textbook cases include the radiation of Darwin's finches in the Galápagos and the radiation of drosophilid (fruit) flies in the Hawaiian Islands, both of which are thought to have been facilitated by geographical isolation.9 On-going collaborative work in one of our own research groups has looked at the distribution of about 30 endemic species of damselfly within the Fijian archipelago, with each species, based on both morphology and genetic analyses, classed within the genus Nesobasis (Fig. 4.4). What is remarkable is that while individual species are relatively broadly distributed within the islands where they are found, no species is found on both the main islands Viti Levu and Vanua Levu. Moreover, several closely related (sister) species are found on different islands,
Figure 4.4 An array of Fijian damselfly species, all from the genus Nesobasis. N. malcolmi (left upper panel), N. unds2 (as yet undescribed species 2, left middle), N. erythrops (left lower panel), N. rufostigma (right upper panel), N. brachycerca (right middle), and N. heteroneura (right lower panel). Photos courtesy of Hans Van Gossum.
further supporting Jordan's law and indicating a role of geographical isolation in species formation.50
What is it about isolation that facilitates speciation? When populations become separated, then different mutations in these populations can arise by chance and spread by selection and/or by genetic drift. First, it is possible that the form of natural selection is similar in both populations, but by chance they evolve different solutions to the same challenges. Second, it is possible that the isolated populations experience subtle differences in their environment, so that factors such as climate, salinity, or competitors cause them to experience slightly different forms of natural selection and that this helps drive the two populations apart. Third, it is also possible that the two populations experience rather different forms of sexual selection—for example, females in the two populations might evolve a preference to mate with a rather different type of male, so that differences between populations can be driven by mate choice. Note, however, that invoking sexual selection only pushes the problem back a step to asking how such differences in preference arose in the first place.
The small size of populations involved in peripatric speciation might lead one to suspect that the chance genetic make-up of the colonists ('founder effects'), coupled with their subsequent chance drift, plays some role in facilitating speciation. This founder-effect mechanism of speciation (often called the 'peripheral isolate theory') enjoyed considerable popularity for several decades of the past century, advocated by Ernst Mayr and influenced by his comprehensive observations of New Guinean birds (including paradise kingfishers that show striking differences in appearance when they are found on small islands, compared to their relatives on the mainland51). Here is not the place for a detailed critique of the theory, observations, and experiments to evaluate the role of founder effects and drift in speciation. Suffice to say, many researchers are currently sceptical that founder effects per se play an important part in speciation,9 in part because the very traits that facilitate reproductive isolation are typically under strong selection. Theory indicates that small populations do not readily become repro-ductively isolated through chance drift, and several experiments, notably those with replicated populations of fruit flies started from small numbers of founders,52,53 fail to show any incipient preferences of members of isolated populations for their own kind after many generations of being kept apart. This is perhaps unsurprising as such experiments have been performed unconsciously many times when researchers establish a laboratory stock of their favourite organism from a small collection from the wild or a shipment from a stock centre. In no case has a new species been reported to arise.9
Was this article helpful?