Intro: Structural attributes of forests

Further strong support for the evolution of gliding in rainforest is provided by the fact that most of the other gliding vertebrates are largely restricted to rainforest; however, their biogeographical distribution in tropical forests is not uniform. For example, the colugos, which are by far the most welldeveloped gliding mammals, live in evergreen rainforests of South-East Asia, while most of the flying squirrels in South-East Asia and the scaly-tailed squirrels of tropical Africa also occur in rainforest. In contrast to the marsupial Northern Glider, the other gliding marsupials live in relatively open eucalypt woodlands. Similarly the American Northern Flying Squirrel and Southern Flying Squirrel live in open coniferous and mixed woodlands and the flying squirrels of Japan and Europe typically occur in coniferous, broad-leaved or mixed forest with evergreen and deciduous trees.

The overlapping tree crowns in tropical forests often give the illusion of a continuously connected canopy; however, the mid and lower storeys consist largely of open space that promotes the selection of long-distance gliding, especially where lianas to link up adjacent trees are scarce. As a result, the distribution of gliding species in tropical rainforests appears to be correlated with differences in canopy structure and particularly liana density. Groundlevel liana densities have been shown to be highest in tropical African forests, somewhat lower in non-tropical rainforests, and lowest in Indo-Malayan forests. It has been suggested that prehensile-tailed vertebrates are favoured in African and especially South American forests where they can bridge canopy gaps using lianas, whereas in the relatively more open Asian forests gliding is favoured.

Another aspect of forest structure that appears to be important (at least in some cases) is the greater height of lowland Indo-Malayan rainforests. Although systematic comparative surveys of tropical forest heights are not available, several observers have noted the extreme heights of Indo-Malayan forest canopies due to the abundance of trees of the family Dipterocarpaceae. The top of the canopy of lowland dipterocarp forests is typically 45 metres in height, with some individual trees reaching 60–70 metres. Similarly, canopy heights of four rainforests in New Guinea ranged from 40–60 metres in height. In contrast, canopy heights of African and South American rainforests are significantly lower because there is no predominant family of tall trees analogous to the Dipterocarpaceae of Indo-Malaya. For example, on Barro Colorado Island in the Republic of Panama, canopy trees in the old forest are 30–40 metres tall. Similarly, the average height of rainforest in the Amazonian Terra Firme Forest basin is 30–45 metres, with individual trees of Dinizia excelsa (Leguminosae) and Bertholletia excelsa (Lecythidaceae) reaching 50 metres in height. Observations of African rainforests are similar to those of South America.

A consequence for gliding animals of an increased forest height, especially from emergent tree species, is that there is a higher point of take-off for larger gliders to travel above the thicker rainforest below. The high point of launch from these emergent trees allows longer glides to be made and results in substantial energetic advantages, especially for larger species. In vertical climbing, energetic costs are proportional to the climbing speed, and vertical climbing prior to a glide is the principal energetic cost of locomotion for gliding animals. Therefore the energetic cost of transport can be significantly reduced in gliding animals by increasing glide lengths and, indirectly, by increasing heights of take-off points.