It has been proposed that marine centres of origin have been functioning as evolutionary engines (Briggs, 2003). Four such centres have been recognized: the Antarctic, the North Pacific, the East Indies and the Southern Caribbean. Each centre has produced dominant, successful species that have spread over large geographical areas. More recently, Goldberg et al. (2005) have used the alternative term ‘macroevolutionary source’ for regions that have high rates of origination, and suggested the complementary term ‘macroevolutionary sink’ in reference to regions that obtain taxa through immigration. Their palaeontological contribution provided additional evidence that age distributions of extant taxa can be used to estimate rates of origination, dispersal and extinction. However, their substitution of an alternative term for centres of origin, a descriptive name that has been used since Darwin's time, appears to be less useful. Also, one may question their claim that there has been a ‘traditional assumption’ that centres of origin demonstrate high levels of endemism. In fact, it has been pointed out that within the tropics almost all reef areas that harbour relatively large numbers of endemics are located on the fringes, not within the centres of origin (Briggs, 2002).
The centre of origin in the East Indies Triangle was completely established only c. 10 Ma, but its predecessor had existed in the Tethys Sea between Africa and Eurasia since the early Cretaceous. An Indo-Mediterranean Region within the early Cretaceous Tethys was recognized by Kauffman (1979) and, by c. 132 Ma, Eastern and Western Mediterranean subprovinces could be distinguished. By c. 124 Ma, a separate Caribbean Province was formed. The endemism that distinguished the latter was a reflection of the increasing distance between the New and Old World as the Atlantic Ocean became wider.
During the early Palaeogene (65–45 Ma), many extant families and genera evolved in the Indo-Mediterranean Region. They included the earliest coral reef fish assemblage consisting of eight families determined to be of Eocene age (Bellwood, 1996), with subsequent discoveries of parrotfishes (Streelman et al., 2002) and the family Triacanthidae (Santini & Tyler, 2003). The gastropod family Cypraeidae (Kay, 1990) and numerous bivalve and echinoderm higher taxa (Kafanov, 2001) were formed in that area. The climate grew colder beginning in the mid-Eocene, and Piccoli et al. (1987) were able to find fossil evidence of the displacement of molluscan taxa from the Indo-Mediterranean to the East Indies. The early Miocene collision between Africa and Eurasia eliminated the Tethys Sea and established the Mediterranean. The tropical biota trapped in the Mediterranean was gradually eliminated as the climate got colder.
In the meantime, the Tethyan fauna that had become isolated to form the Caribbean Province became divided into the West Central American and Antillean subprovinces (Kauffman, 1979). Some early contributions to the East Indies fauna may have been received from the Caribbean Province via trans-Pacific migration (Hallam, 1994). The late Cretaceous subdivision of the Caribbean Province may have been in response to the formation of a Central American archipelago. An early Central American isthmus may have also been formed at that time (Briggs, 1995), but its existence is still speculative. If there had been an isthmus, it disappeared in the Palaeocene, leaving a common American tropical fauna, until it was divided by the rise of the Panamanian Isthmus in the Pliocene.
It may be concluded that the modern, high-diversity biota of the Southern Caribbean was, in large part, derived from the Caribbean Province of the Tethys Sea, although some of it came from the Western Pacific across the East Pacific Barrier prior to the rise of the Panamanian Isthmus. The rich biota of the East Indies was inherited primarily from the Indo-Mediterranean Province via the Indian Ocean. In the northern part of the Western Atlantic, Pliocene–Pleistocene fluctuations in temperature resulted in many extinctions, but the Southern Caribbean was little affected and new species continued to originate (Allmon et al., 1996). From the time that the East Indies centre was well established in the late Miocene, and the Southern Caribbean centre in the Pliocene, dispersal of species from each centre began to augment the total diversity.
Considering the fact that the average life span of a marine species is c. 4 Myr (Sepkoski, 1998), and that the Pliocene began more than 5 Ma, this means that the origin and distribution of the great majority of our present species probably took place in the Pliocene and Pleistocene. Distributional patterns suggest that a large portion of the present species diversity originated in the two tropical centres. For example, with regard to the East Indies Triangle, the species diversity of reef fishes is likely to exceed 3000 (Briggs, 2005). Allen & Adrim (2003) estimated the number for the entire Indo-Pacific to be 3764. This means that more than three-quarters of the Indo-Pacific reef fishes, inhabiting a region that stretches more than two-thirds of the way around the world, are also present in the East Indies centre of origin. Also present are c. 450 species of hermatypic corals (Veron, 1995).
The concentration of species in the East Indies does not, in itself, prove that their formation took place in that area. However, several biogeographical patterns strongly point to the East Indies (Briggs, 1999): (1) a distance-correlated decrease in species diversity that extends outward from the East Indies, (2) an average outward increase in geological age, (3) historical dispersal tracks that radiate outward, (4) phylogenetic regression patterns indicating the peripheral locations of plesiomorphic species, (5) extinction patterns that originate in the East Indies and progress outward, and (6) indications from the mitochondrial DNA of some species that a decreasing gradient of genetic diversity extends outward.
Alternative explanations for the high diversity in the East Indies have been offered. These have usually involved theories dependent on: (1) an overlap of Indian Ocean and Western Pacific faunas, (2) a passive accumulation of species that had been formed elsewhere, and (3) historic movements of species via plate tectonics. The lack of evidence for these theories has been discussed (Briggs, 2004). More recently, the plate tectonic hypothesis involving island integration has been reiterated (Carpenter & Springer, 2005). But it is difficult to relate the present distribution of species, most of which are probably < 5 Myr old, to ancient tectonic movements. Bellwood et al. (2005) examined coral reef biodiversity in relation to energy, the mid-domain effect and reef area. Their best model incorporated area and the mid-domain effect, but did not include energy. However, it does not seem possible to apply the mid-domain effect when the longitudinal diversity peak on the equator lies far to the west of the midpoint of the Indo-Pacific (Mora et al., 2003). At a given time, one might expect to find a positive relationship between habitable area and species diversity, but this relationship can still exist and not affect the function of the East Indies as an evolutionary engine – the system is dynamic, so that origins are almost equalled by extinctions.
Mora et al. (2003) conducted a survey of coral reef fishes that extended all the way across the Indo-Pacific. They examined the longitudinal and latitudinal ranges of 1970 fish species, and found that on both axes the range midpoints clustered directly on the East Indies. Although the latitudinal peak was on the equator, the longitudinal peak was far to the west of the midpoint of the Indo-Pacific. This departure from a mid-domain effect indicated strong origination activity in the East Indies. Also, Mora et al. (2003) found 90 species that were endemic to the East Indies. These were apparently neo-endemics that had not begun to disperse outward. Their report concluded that the East Indies had played the major role in assembling communities throughout the Indian and Pacific oceans. Recent work on two invertebrate groups, cuttlefishes (Sepiidae) and cone shells (Conidae), also provides support for the East Indies centre of origin hypothesis (Neige, 2003; Vallejo, 2005).
Another interesting biogeographical pattern is morphological disparity. Within a given clade, it appears that high species diversity, as found in the East Indies, tends to be accompanied by a low level of morphological contrast among the species. Conversely, species located out on the horizontal periphery or in deep water often exhibit a greater variety in their structure, Why should this be so? Where speciation is most active, one might expect to find large numbers of closely related species. Over time, as species extend their range from their place of origin, extinction would take its toll leaving relatively few survivors. As a result, the remaining peripheral species often consist of plesiomorphic relicts that represent various evolutionary stages. Therefore a greater disparity might be expected. Examples of this pattern may be found in the strombid gastropods (Roy et al., 2001) and in the cuttlefishes (Neige, 2003).
The East Indies Triangle is the major centre of origin in the marine world because its influence, although predominant in the tropical Indo-West Pacific, extends to other tropical regions and to higher latitudes. Two principal barriers separate the Indo-West Pacific from the other tropical shelf regions. They are the deep-water East Pacific Barrier that lies between Polynesia and the New World, and the Old World Land Barrier comprising Africa and Eurasia. A variety of Indo-West Pacific species have crossed the former to become established in the Eastern Pacific. These include 80 fishes (Robertson et al., 2004) and 61 gastropod molluscs (Emerson, 1991). Conversely, there have been few successful migrations in the opposite direction. Twenty-two shore fishes have apparently migrated westward, but 12 of them do not extend past Hawaii. None of the molluscan lineages in the Eastern Pacific has been able to invade westward (Vermeij, 2004).
About 95% of the Eastern Pacific coral species are recent immigrants from the Indo-West Pacific (Robertson et al., 2004). This predominantly eastward invasion has occurred despite the fact that the north equatorial current has been shown to carry larval stages from the Eastern Pacific towards the Indo-West Pacific (Scheltema, 1988). Similarly, the other boundaries of the Indo-West Pacific appear to function essentially as one-way filters. Twenty-one gastropod species and a similar number of fishes from the Indo-West Pacific have been able to round the Cape of Good Hope to colonize the tropical Atlantic, but there seems to have been no successful migration in the opposite direction (Vermeij, 2004). Since the opening of the Suez Canal in 1869, more than 200 species of Indo-West Pacific organisms have established themselves in the Mediterranean, but only about a dozen have taken the reverse course (Galil, 1994). When the biota of the other three tropical regions (Eastern Pacific, Western Atlantic, Eastern Atlantic) is examined, the relationship to the Indo-West Pacific seems obvious. Most of the families and a large fraction of the genera are shared among all four.
Although the rise of the Isthmus of Panama did not form a complete marine barrier until the late Pliocene, there are indications that the reef fauna of the Caribbean began to separate from that of the Eastern Pacific in the late Miocene, c. 7 Ma (Muss et al., 2001). During the intervening time a new evolutionary relationship became established within the Atlantic. A diversity centre developed within the southern part of the Caribbean Sea, but its species richness is far less than that of the East Indies. For example, the latter probably supports more than 3000 fish species compared with c. 700 in the Caribbean (Rocha, 2003). Similarly, there are c. 450 hermatypic coral species compared with c. 50 in the Caribbean (Veron, 1995).
The shelf fauna of the Eastern Atlantic tropics is separated from that of the Caribbean by the deep-water Mid-Atlantic Barrier. The overall species diversity of the Eastern Atlantic is only about one-third that of the Caribbean (Briggs, 1985). Recent research (Joyeux et al., 2003) has revealed the presence of 92 trans-Atlantic species of reef fishes, the great majority belonging to well developed Caribbean genera. Circumstantial evidence indicates an eastward migration, with only four species apparently having travelled in the opposite direction. Within the Western Atlantic, Southern Caribbean species have apparently been penetrating northward into Florida and Bermuda and southward into Brazilian waters (Rocha, 2003). During the past 2 Myr, the Eastern Atlantic has received at least 41 molluscan species from the Western Atlantic, and c. 14 migrated in the opposite direction (Vermeij, 2004). These predominantly outward migrations are indications that the Southern Caribbean has been operating as a centre of origin.
In addition to the enormous diversity of species that exist in the shallow waters of the tropics, there is good evidence that, over time, tropical species have been successful in invading deeper waters and the temperate waters of higher latitudes. Jablonski et al. (1983) drew attention to onshore–offshore patterns in the evolution of Phanerozoic shelf communities. In more recent years, several palaeontologists provided additional evidence of onshore-to-offshore replacements, and discovered that, in general, new species, genera and families had evolved under high-diversity conditions. This led to the common observation that ‘diversity begets diversity’.
Over time, the onshore-to-offshore replacement sequence has had a cumulative effect in the deep sea. Although the cold-water centres in the North Pacific and Antarctic have made significant contributions to the slope and abyssal faunas, so have the tropics. There are many families of deep-sea animals that live only at low latitudes, so the faunistic centres in the Pacific and Atlantic may be contributing directly to the deep sea (Zezina, 1997). On the other hand, there is no evidence of any deep sea clade being able to establish itself in shallow, tropical waters (Vermeij, 2004).
With regard to the relationship of the shallow-water tropical and temperate biotas, it has been noted that the warm-temperate zone that borders the tropics to the north and south, possesses genera and families that are mostly of tropical affiliation. It has also been discovered that the historical movements of species between the tropics and temperate waters have apparently been entirely one way. In a survey of molluscs and barnacles that originated from 20 to 25 Ma, Vermeij (2004) found that 29 American and nine Asian clades in warm-temperate to tropical waters had given rise to cold-adapted species, but no temperate clades had spawned tropical species. Crame (2000) concluded that the steep latitudinal gradients of the youngest bivalve clades provided additional information on dispersal, and that the tropics could be seen as a species pool supplying bivalve taxa to high-latitude and polar regions.
As can be noted from the family history of marine diversity (Clarke & Crame, 2003), a significant increase began with the continental break-up in the early Jurassic, but the steepest rise started in the early Cenozoic. The Cenozoic increase was especially marked at the lower taxonomic levels, with species diversity rising perhaps an order of magnitude (Crame, 2004). There were significant radiations of neogastropods, heteroconch bivalves, cheilostome bryozoans, decapod crustaceans and teleost fishes. Considering that these comparatively young groups show strong latitudinal diversity gradients, it can be inferred that their major radiation events were centred in the tropics.
Within the Indo-West Pacific, it is apparent that the centre of diversity and of evolutionary innovation lies within the bounds of the East Indies Triangle. This concentration of power is exemplified by the recent study of the ranges of fish species across the Indo-Pacific (Mora et al., 2003) where, on average, 86% of the species found in the outlying areas were also present in the Triangle. If, as the fossil data indicate, diversity does beget diversity at species, family and generic levels, then the East Indies centre may be the place of origin of most of the young lineages that have successfully penetrated the cooler waters of higher latitudes and greater depths. The Southern Caribbean, the secondary tropical centre of origin, could also have been effective in this manner.
A considerable amount of tropical species diversity is due to the evolution of herbivory, a feeding habit that has become widespread in the tropics but remains relatively rare in temperate waters. It has been noted (Vermeij, 2004) that herbivore-containing clades exist within the molluscs, annelids, arthropods, echinoderms and vertebrates. Furthermore, the herbivore taxa occupy the more derived positions within their respective evolutionary trees, and most are unknown prior to the Cenozoic. Herbivory is found within 10 clades of bony fishes that occur mainly in the tropics. Most of the herbivorous clades in the Indo-West Pacific are represented by more species in the East Indies centre than anywhere else.
Investigations of fish diets have revealed that tropical reef fishes show an evolutionary trend towards taking advantage of relatively low-energy food resources such as algae, sea grasses, sponges, detritus and cnidarians (Harmelin-Vivien, 2002; Floeter et al., 2004). In some locations as many as 57–79% of species were dependent on such resources, especially algae and sea grasses. This suggests that a considerable portion of the tropical marine diversity may be attributed to the presence of species that have evolved, by means of ecological specialization, to utilize low-energy food resources. This shift to an alternative food supply under highly competitive conditions suggests that sympatric speciation may have been involved. It appears that, when such tropical lineages attempt to invade cooler waters, the increased energy demand constitutes a difficult physiological barrier. Some clades manage to exist at higher latitudes by becoming omnivorous or by a seasonal switching between herbivory and carnivory.
Temperate and polar centres
In the cold and cold-temperate zones, the other two centres of origin have also been effective in adding to Cenozoic diversity. Of the two, the North Pacific centre has been the most broadly influential. It was formed c. 40 Ma but did not achieve its present influence until the opening of the Bering Strait permitted the Great Trans-Arctic Biotic Interchange. The invasion into the Arctic-North Atlantic produced enormous ecosystem changes. Some of the migrations may have begun as early as 12 Ma, but apparently most took place c. 3.5 Ma, before the Arctic Ocean was ice-covered and when the temperature was still in the cold-temperate range. New information regarding molluscan species in the North Atlantic (Vermeij, 2004) indicates that at least 143 invading species colonized European shores, while 176 settled in eastern North America. At most, 24 species invaded the North Pacific from the Atlantic. On the eastern American rocky shores, the invaders now comprise the majority of common species.
The North Pacific has contributed a variety of organisms to the cold-temperate southern hemisphere and even to the Antarctic. The migrations took place primarily via isothermic submergence, whereby the species concerned could maintain a suitable temperature by moving beneath the tropics at great depth. For example, two fish families of North Pacific origin are represented by numerous species around the Antarctic continent. There are 90 species belonging to the families Zoarcidae and Liparididae, together comprising > 40% of the Antarctic fish fauna. Many other clades of North Pacific origin have reached the Southern Ocean, but most did not penetrate as far south as the Antarctic.
The external influence of the Antarctic Region, with respect to the surface waters, has generally been limited. Many species produced in the region have managed to migrate to the surrounding cold-temperate waters of the Sub-Antarctic. A few species, such as some penguins, fishes and molluscs, have extended their range much farther. The greatest external influence has been in the deep sea. A variety of deep benthic organisms were apparently introduced to that habitat via thermohaline currents originating around Antarctica.
In inquiring about the origin of contemporary marine species diversity, one needs to take into consideration the evident difference in modes of speciation between the centres of origin and other areas. Recent studies combining the phylogeny and ecology of natural populations have emphasized the role of ecological factors in speciation (Via, 2002). Studies of parallel speciation have provided a strong case for sympatric speciation and for natural selection in generating reproductive barriers (Johannesson, 2001). Briggs (2005) proposed that, within the East Indies, the high species diversity, the production of dominant species, and the presence of newly formed species are due to natural selection being involved in reproductive isolation, the first step in the sympatric speciation process.
The Darwinian view maintains that natural selection directly favours the multiplication of species during ecologically based sympatric speciation (Coyne & Orr, 2004). Under this view, allopatric species are accidental by-products of genetic divergence. In other words, when a physical barrier separates a formerly continuous population, natural selection is not involved in the process of reproductive isolation. If natural selection usually produces better adapted species, then those produced by allopatry are at a disadvantage. Furthermore, species formed by sympatry usually evolve much more quickly. In sympatric Drosophila populations, speciation is completed in c. 200,000 years, but allopatric populations require c. 2.7 Myr (Coyne & Orr, 2004).
The East Indies example does not mean that sympatric speciation is confined to centres of origin. This mode was suggested by the presence of sympatric sibling species in many locations (Knowlton, 1993), and has now been proven in several places (Hellberg & Vaquier, 1999; Hendry et al., 2000; Dawson et al., 2002; Jones et al., 2003) and strongly suggested in others (Munday et al., 2005; Rocha et al., 2005). Sympatric speciation usually takes place when there is opportunity for the occupation of additional habitat or when there is an alternative food supply available. The same ecological opportunities apply to parapatric speciation. Species flocks resulting from rapid bursts of cladogenesis are well known among freshwater fishes, but have been investigated only recently in various marine clades. By using molecular phylogenies to study variation in diversification rates among lineages, Ruber & Zardoya (2005) were able to find elevated rates of cladogenesis in six different groups of marine fishes. Explosive radiation leading to the formation of a species flock has been found in a marine gastropod genus (Duda & Rolan, 2005). Sexual selection under sympatric conditions has contributed to rapid speciation in freshwater fishes (Salzburger & Meyer, 2004), and should be anticipated in marine clades.
There is a difficulty with terminology that arises in descriptions of the speciation process. While allopatry, whether caused by dispersal or vicariance, is clearly understood, various terms have been employed when speciation occurs within a continuous population. The designations parapatric, sympatric, competitive and ecological are all being used currently to describe speciation events that occur without initial physical separation. But all four terms essentially define the same process: the achievement of reproductive isolation by means of selection for an alternative environment or food source. And speciation that is initiated by natural selection is fundamentally different from that initiated by allopatry. Although one result of such a sympatric type of speciation may be the formation of adjacent or parallel species, often called parapatric, their origination does not appear to differ from that of other non-allopatric species.
As most species now isolated in regions and provinces by physical barriers (temperature, distance, currents) were probably formed by allopatry, they would be relatively slow in producing endemics. Therefore the species-packing process would be comparatively slow. Its effectiveness on the general level of species diversity could theoretically be calculated by adding together the numbers of all the endemic species that exist in all the provinces – unfortunately an impossible task at the present state of our knowledge.