Over the past 30 years, the principal research problem in ecology has been the explanation of biodiversity. More specifically, the aim has been to explain the gradient of diversity from its high values near the equator to the very low values in subpolar regions. Most hypotheses about this assume that the present situation is in equilibrium with the environment, and thus try to explain matters in terms of currently operating processes. Although the complete explanation may well be multivariate— that is, involving the interaction of many separate factors—there has still been a tendency to search for an underlying mechanism.
One such mechanism that has been popular in the last 20 years is the refuge theory (Haffer, 1997; Haffer and Prance, 2001), which was one of the first ideas involving paleoecology. If the tropical rainforest was divided into many separate blocks by dry Pleistocene climates, this could have isolated populations for long enough to allow allopatric speciation. In fact, repetition of this by multiple Milankovitch cycles could have produced a "species pump" leading to many closely related species, which is what we find in the tropical rainforest. Despite the fact that the refuge theory had no explanation as to why it operated only in the tropics, and not also in temperate regions (Flenley, 1993), it remained popular even after it was found that there was little palynological support for the existence of the supposed dry phases (e.g., Bush, 1994; Bush et al., 1992). The theory only foundered completely when it was demonstrated that the Quaternary had not been a time of great speciation in the tropics at all. In fact, using palynological richness as a proxy for diversity (Flenley, 2005), it could be suggested that diversity had actually declined during the Quaternary (Morley, 2000; Bennett, 2004; Flenley, 2005). It seems that mutation rates were too low to produce the required genetic isolation in the time available (Willis and Niklas, 2004).
The opportunity therefore exists for promulgation of a new hypothesis, which might retain some of the attractive features of the refuge hypothesis—such as the species pump concept, related to cyclical climatic change—while avoiding the pitfalls of the earlier hypothesis. This is the aim of the present section.
To be at all plausible, the new hypothesis must account for the origin and survival of large numbers of closely related species, now living in close proximity with each other. Assuming normal evolutionary processes, the most likely procedure for achieving this is allopatric speciation—that is, the original population of a species must be split up and the sub-populations isolated. There must then be mutation at a sufficient rate within each sub-population such that when the sub-populations are recombined, they remain reproductively isolated.
Referring again to the geological record, and using palynological richness as a proxy for diversity (Morley, 2000; Flenley, 2005) we find that the times of rapidly increasing diversity were the Eocene and the Early Miocene, especially the former (Wilf et al., 2003). There is good evidence that both these periods were exceptionally warm. In the Eocene, megathermal forests spread well beyond their present limits (Morley, 2000). For instance, the tropical palm Nypa occurred in Britain. Even allowing for continental movement, world climates must have been much warmer than now, and this could have been the result of either greater insolation, or higher concentration of greenhouse gases, or both (Morley, 2000; Willis and Niklas, 2004). Since the altitudinal temperature lapse rate is dependent on the amount of atmosphere above the surface, it is difficult to see how the lapse rate could have differed very much from the present rate found in the wet tropics, of c. 6°C per 1,000m. At the peak of Eocene warmth, the sea surface temperature in the tropics rose as high as 32° C (Pearson et al., 2001). This is about 5°C higher than now. The Late Paleocene thermal maximum may have been even warmer (Zachos et al., 2001; Willis and Niklas, 2004).
Thus, using the above lapse rate, we might expect a MAT of 27°C to 21°C (the present range of lowland forest species) to occur at c. 800-m to 1,800-m altitude, and thus what are now lowland species could have grown at that altitude, assuming that their climatic tolerances have not changed. Since the usual MAT tolerance of a species has a range of about 6°C MAT (van Steenis, 1934-36), other species would have occurred at lower altitudes, although they might well have been congeners of the present day lowland species.
At 800 m to 1,800 m we might expect UV-B insolation to be slightly above that at sea level. This altitude effect is usually quite small, with values around 15% per 1,000 m (Blumthaler et al., 1997). Since, however, there were major volcanic eruptions in Greenland during the Eocene (Aubrey et al., 1998) there might have been a considerably greater temporary rise in UV radiation (see previous section).
UV-B is a well-known mutagen. Its effect in producing human skin cancer has been well-researched (Lodish et al., 2000) and it is clear that skin cancer is more
Eocene Warm phase of Milankovitch cycle
Eocene Warm phase of Milankovitch cycle
Eocene Cool phase of Milankovitch cycle
| ^ Distribution of present-day lowland species
Figure 8.5. Diagram to show how a combination of appropriate topography, Eocene warmth, and enhanced UV-B after vulcanicity could lead to isolation, mutation, allopatric speciation, a species pump, and increased biodiversity.
common in the tropics (Smith and Warr, 1991). The latitudinal gradient of UV-B is well-established (Caldwell et al., 1980). UV-B is widely used by geneticists and plant breeders to promote mutation (Jansen et al., 1998; Atwell et al., 1999). Since plant reproductive apparatus is necessarily exposed to the atmosphere (for pollination, whether by wind, insects, or other means) it seems possible that UV-B induced mutations in reproductive DNA would occur at an enhanced rate. The effect of elevated UV-B on plant reproductive apparatus has already been reported (van der Staaij et al., 1997). Extreme elevation of UV-B in the geological past is thought to have caused worldwide dieback of woody plants (Visscher et al., 2004).
There is evidence that Milankovitch cycles have affected world climate throughout geological time (Bennett, 1990; Pietras et al., 2003; Willis and Niklas, 2004). The effect of such cycles on Eocene megathermal forest taxa would have been to drive them up and down hills. In warm phases, what are now lowland taxa (already adapted to MAT of 27-21 °C) would have migrated to the hills. The possibility of isolation on individual peaks would occur and this would be repeated with each cycle (Figure 8.5). Thus, we have all the requirements for allopatric speciation and a species pump: mutation, geographic isolation, and cyclic change. Furthermore, this process would work preferentially in the area then covered by megathermal vegetation, which included mid-latitudes as well (Morley, Chapter 1 in this book). Outside that area, strong seasonality would have restricted tree growth to low altitudes.
It would support this hypothesis if present day tropical montane taxa were experiencing similar high speciation rates, since they experience high UV-B and the possibility of isolation on mountain peaks during interglacials. One does not have to look far for examples. The case of Espeletia in the Colombian Andes is well-known: while some species are widespread, others are restricted to single peaks (van der Hammen, 1974). It is likely that Rhododendron may provide another example, in New Guinea and the Himalayas (Sleumer, 1966; Leach, 1962). Interestingly, many Rhododendron species are epiphytes, and thus exposed to strong insolation in the canopy. Possibly, many epiphytic orchid genera would be another set of examples
(e.g., Cymbidium) (Du Puy and Cribb, 1988). The possible origin of varieties of Leptospermum flavescens on Mt. Kinabalu (Borneo) by the action of UV-B has already been proposed (Lee and Lowry, 1980a). The abundance of endemic species in several genera on individual peaks in the Andes has already been noted by Gentry (1989).
In summary, the hypothesis presented here is that major speciation in the tropics may have occurred especially in warm periods in the past (Eocene and Early Miocene), because in those times lowland taxa were able to live at higher altitudes where they experienced enhanced UV-B induced mutation rates (increased further by intermittent vulcanicity), and isolation on individual mountain peaks. Cyclical climate changes could have led to a species pump. Occasional major volcanic eruptive phases could have provided the "punctuated equilibrium" which is the pattern of evolution accepted by many (Eldredge and Gould, 1972). It is not suggested that this is a complete explanation for biodiversity. While it could apply to some insects and some other animals, it is unlikely to apply to marine life or to nocturnal or soil animals. Probably, the full explanation will turn out to be multivariate, but it is hoped that this hypothesis may make a contribution.
Was this article helpful?