elements in ‘what lives where’ are plainly apparent for many of the species groups in the figure. The earliest split within the picture‐wing clade occurred prior to the emergence of the mature Kauai, with separation of the basal adiastola group (Figure 1.16). The figure also shows that the basal species of the picticornis, planitibia and grimshawi groups are found on the ancient island of Kauai, with these groups separating 3.8–4.7 million years ago (mya). A second stage of diversification can be seen when the grimshawi subgroups split at 2.1–3.1 mya on Oahu. The planitibia group provides a particularly clear example of progression‐rule dispersal, with a split into two lineages on Kauai, followed by a split into three lineages on Oahu and subsequent dispersal to the younger islands. But such clear patterns are not always discernible, particularly in the grimshawi group. As new islands have been formed, rare dispersers have reached them and eventually evolved into new species, usually by becoming specialised on particular host plants. The arrival on Oahu around 3 mya of new plants upon which grimshawi species specialise (including Charpentiera and Pisona spp.) may have triggered a burst of speciation in the group. At least some of the picture‐winged species appear to match the same environment as others on different islands. Of two closely related species, for example, D. adiastola is only found on Maui and D. setosimentum only on Hawaii, but the environments that they live in are apparently indistinguishable (Heed, 1968). What is most noteworthy, of course, is the power and importance of isolation (coupled with natural selection) in generating new species. Thus, this island biota illustrates two important, related points: (i) that there is a historical element in the match between organisms and environments; and (ii) that there is not just one perfect organism for each type of environment.
1.4.3 Climatic history
Climatic variations have occurred on shorter timescales than the movements of landmasses. Changes in climate during the Pleistocene ice ages, in particular, bear a lot of the responsibility for the present patterns of distribution of plants and animals. Techniques for analysing and dating biological remains (particularly buried pollen) increasingly allow us to detect just how much of the present distribution of organisms is a precise locally evolved match to present environments, and how much is a fingerprint of the hand of history. As climates have changed, species populations have advanced and retreated, been fragmented into isolated patches, and may then have rejoined. Much of what we see in the present distribution of species represents a phase in the recovery from past climatic change (Figure 1.17).
the Pleistocene glacial cycles …
Techniques for the measurement of oxygen isotopes in ocean cores indicate that there may have been as many as 16 glacial cycles in the Pleistocene, each lasting for about 125 000 years (Figure 1.17a). Each cold (glacial) phase may have lasted for as long as 50 000–100 000 years, with brief intervals of only 10 000–20 000 years when the temperatures rose to, or above, those of today. From this perspective, present floras and faunas are unusual, having developed at the warm end of one of a series of unusual catastrophic warm periods.
Figure 1.17 Contrasting changes in the distribution of spruce and oak species in relation to the waning of an ice age. (a) Estimates of temperature during glacial cycles over the past 400 000 years, obtained by comparing oxygen isotope ratios in fossils taken from ocean cores in the Caribbean. Periods as warm as the present have been rare events, and the climate during most of the past 400 000 years has been glacial. The dotted line represents the temperature 10 000 years ago at the beginning of the present period of warming (b) Ranges in eastern North America, as indicated by pollen percentages in sediments, of spruce species (above) and oak species (below) from 21 500 years ago to the present. Note how the ice sheet contracted during this period.
Source: (a) After Emiliani (1966) and Davis (1976). (b) After Davis & Shaw (2001).
During the 20 000 years since the peak of the last glaciation, global temperatures have risen by about 8°C. The analysis of buried pollen – particularly of woody species, which produce most of the pollen – can show how vegetation has changed (Figure 1.17b). As the ice retreated, different forest species advanced in different ways and at different speeds. For some, like the spruce of eastern North America, there was displacement to new latitudes; for others, like the oaks, the picture was more one of expansion.
We do not have such good records for the postglacial spread of animals associated with the changing forests, but it is certain that many species could not have spread faster than the trees on which they feed. Some of the animals may still be catching up with their plants, and tree species are still returning to areas they occupied before the last ice age. It is quite wrong to imagine that our present vegetation is in some sort of equilibrium with (adapted to) the present climate.
Even in regions that were never glaciated, pollen deposits record complex changes in distributions. In the mountains of the Sheep Range, Nevada, for example, different woody species of plant show different patterns of change in the ranges of elevations that they have occupied as climate has changed (Figure 1.18). The species composition of vegetation has continually been changing and is almost certainly still doing so.
Figure 1.18 Contrasting changes between fossil and current distributions of 10 species of woody plant from the mountains of the Sheep Range, Nevada. The red dots represent fossil records, while the blue lines show current elevational ranges.
Source: After Davis & Shaw (2001).
The records of climatic change in the tropics are far less complete than those for temperate regions. It has been suggested, though, that during cooler, drier glacial periods, the tropical forests retreated to smaller patches, surrounded by a sea of savanna, within which speciation was intense, giving rise to present‐day ‘hot spots’ of endemism. Evidence for this in, for example, the Amazonian rainforest now seems less certain than it once did, but there is support for the idea in other regions. In the Australian wet tropics of Queensland, north‐eastern Australia, it has been possible to use present‐day distributions of forest to predict distributions in the cool‐dry climate of the last glacial maximum when forest contraction was greatest (about 18 000 years ago), the cool‐wet period around 7000 years ago when a massive expansion was likely, and the warm‐wet period around 4000 years ago when there was likely to have been another contraction (Figure 1.19a) (Graham et al., 2006). Putting the distributions together, then, allows each subregion of
