Willem de Winter
Johnson & Johnson
Issues of size, shape, form and function continue to intrigue and divide biologists. Ranging over seven orders of magnitude in mammals alone, body size represents the most conspicuous – and most easily quantified – source of variability between different organisms, as well as between different developmental stages within a single organism. In order to compare features of different organisms, therefore, size differences have to be taken into account through some scaling procedure.
Broadly speaking, there are two approaches to scaling for body size differences. One is to represent the size or dimension of the trait under study as a proportion of body size or some other, often functionally related trait in the same individual (e.g. femur length to tibia length, size of a brain part relative to the sizes of its major inputs). Such proportionally scaled measures of relative size are themselves unique, empirically measurable properties of a given individual organism, and as such their variability between organisms is readily amenable to functional interpretation.
The other approach, referred to as allometric scaling, is to measure the same trait in a broad sample of organisms and to regress the variability of its size or dimension against the variability in body size in the same sample (typically after a log-log transformation). The absolute dimension of this trait in any particular organism can then be evaluated relative to the expected dimension for that trait given its body size. In contrast to the proportional approach, however, such allometrically scaled dimensions are a property not only of the organism itself, but also of the sample of organisms against which it is scaled. This fact has important ramifications for both the empirical assessment of the allometric constants and for their functional interpretation. Some of these methodological issues, including choice of statistical regression procedures, (non-)independence of data points, selected sampling and dependence of allometric parameters on phylogenetic closeness, will be touched upon briefly.
A more fundamental issue with allometric scaling is its tendency to obscure the contribution of adaptation as a source of variability between species and developmental stages. In the allometric approach, the sheer magnitude of body size variation tends to swamp all other sources of variability, even if these carry much greater functional significance. Regardless of body size, however, every organism is the product of millions upon millions of years of adaptive evolution, tuning it in every aspect, every minute detail during every stage of its lifecycle, to its particular way of life in its particular environment. Moreover, in their natural environments, different organisms, but also the same organism at different developmental stages, face different sets of challenges and opportunities, are exposed to different predators, diseases or chemical substances, and can profit from different sources of shelter, food and protection. Therefore, body size alone can never be expected to fully explain, let alone predict, all functional differences between organisms. Previous work of the author on mammalian brain evolution will be shown to illustrate how decades of allometric research failed to uncover a wealth of adaptive variability in the functional organization of mammalian brains.
Reference: PAGE 18 () Abstr 1674 [www.page-meeting.org/?abstract=1674]
Poster: Oral Presentation: Pediatrics