Phylogenetic and allometric constraints

The life histories that natural selection favors (and we observe) are not selected from an unlimited supply but are constrained by the phylogenetic or taxonomic position that organisms occupy. For example, in the entire order Procellariformes (albatrosses, petrels, fulmars) the clutch size is one, and the birds are 'prepared' for this morphologically by having only a single brood patch with which they can incubate this one egg (Ashmole, 1971). A bird might produce a larger clutch, but this is bound to be a waste unless it exhibits concurrent changes in all the processes in the development of the brood patch. Albatrosses are therefore prisoners of their evolutionary past, as are all organisms. Their life histories can evolve to only a limited number of options, and the organisms are therefore confined to a limited range of habitats.

organisms are prisoners of their evolutionary past

Figure 4.31 Phenotypic plasticity in the combination of clutch size and laying date in the kestrel, Falco tinnunculus, in the Netherlands. (a) Within particular territories (in this case, one of high quality) the expected, optimal combination is that with the highest total reproductive value (for calculation, see text). (b) Predicted (rectangles) and observed (points with standard deviations) combinations for territories varying in quality from high (left) to low (right). (After Daan et al., 1990; Lessells, 1991.)

Figure 4.31 Phenotypic plasticity in the combination of clutch size and laying date in the kestrel, Falco tinnunculus, in the Netherlands. (a) Within particular territories (in this case, one of high quality) the expected, optimal combination is that with the highest total reproductive value (for calculation, see text). (b) Predicted (rectangles) and observed (points with standard deviations) combinations for territories varying in quality from high (left) to low (right). (After Daan et al., 1990; Lessells, 1991.)

It follows from these 'phylogenetic' constraints that caution must be exercised when life histories are compared. The albatrosses, as a group, may be compared with other types ofbirds in an attempt to discern a link between the typical albatross life history and the typical albatross habitat. The life histories and habitats of two albatross species might reasonably be compared. But if an albatross species is compared with a distantly related bird species, then care must be taken to distinguish between differences attributable to habitat (if any) and those attributable to phylogenetic constraints.

4.14.1 Effects of size and allometry

One element of phylogenetic constraint is that of size. Figure 4.32a shows the relationship between time to maturity and size (weight)

in a wide range of organisms from viruses to whales. Note first that particular groups of organisms are confined to particular size ranges. For instance, unicellular organisms cannot exceed a certain size because of their reliance on simple diffusion for the transfer of oxygen from their cell surface to their internal organelles. Insects cannot exceed a certain size because of their reliance on unventilated tracheae for the transfer of gases to and from their interiors. Mammals, being endothermic, must exceed a certain size, because at smaller sizes the relatively large body surface would dissipate heat faster than the animal could produce it, and so on.

The second point to note is that time to maturity and size are strongly correlated. In fact, as Figure 4.32a-c illustrates, size is strongly correlated with many life history components. Since the sizes of organisms are constrained by phylogenetic position, these other life history components will be constrained too.






Mammals, invertebrates, reptiles, amphibians







8 10 Br



102 103 104 Body weight (g)

Diameter (m)

Diameter (m)

Body weight

Figure 4.32 Allometric relationships, all plotted on log scales. (a) Maturation time as a function of body weight for a broad range of animals. (b) Brood time as a function of maternal body weight in birds. (c) Maximum lifespan as a function of adult body weight for a broad range of animals. (After Blueweiss et al., 1978.) (d) The allometric relationship between tree height and trunk diameter 1.525 m from the ground, for 576 individual 'record' trees representing nearly every American species. (After McMahon, 1973.)

An allometric relationship (see Gould, 1966) is one in which a physical or physiological property of an organism alters relative to the size of the organism. For example, in Figure 4.33a an increase in size (actually volume) amongst salamander species leads to a decrease in the proportion of that volume which is allocated to a clutch of young. Likewise, in Figure 4.32b an increase in weight amongst bird species is associated with a decrease in the time spent brooding eggs per unit body weight. Such allometric relationships can be ontogenetic (changes occurring as an organism develops) or phylogenetic (changes that are apparent when related taxa of different sizes are compared), and it is the latter that are particularly important in the study of life histories (Figures 4.32 and 4.33).

Why are there allometric relationships? Briefly, if similar organisms that differed in size retained a geometric similarity (i.e. if they were isometric), then all surface areas would increase as the square of linear size, whilst all volumes and weights increased as the cube. An increase in size would then lead to decreases in the length : area ratios, decreases in length : volume ratios and, most important, decreases in area: volume ratios. Almost every bodily function depends for its efficiency on one of these ratios (or a ratio related to them). A change in size amongst isometric organisms would therefore lead to a change in efficiency.

For example, the transfer of heat, or water, or nutrients, either within an organism or between an organism and its environment, takes place across a surface, which has an area. The amount of heat produced or water required, however, depends on the volume of the organ or organism concerned. Hence, changes in area : volume ratios resulting from changes in size are bound to lead to changes in the efficiency of transfer per unit volume. Thus, if efficiency is to be maintained, this must be done by allometric alterations. Exact allometric slopes vary from system to system and from taxon to taxon (for further discussion see Gould, 1966; Schmidt-Nielsen, 1984; and, in a more ecological context, Peters, 1983). What, though, is the significance of allometry in the study of life histories?

The usual approach to the ecological study of life histories has been to compare the life histories of two or more populations (or species or groups), and to seek to understand the differences between them by reference to their environments. It must be clear by now, however, that taxa can also differ because they lie at different points on the same allometric relationship, or because they are subject to different phylogenetic constraints generally. It is therefore important to disentangle 'ecological' differences from allometric and phylogenetic differences (see Harvey & Pagel, 1991; Harvey, 1996; and also the summary in Stearns, 1992). This is not because the former are 'adaptive' whereas the latter are not. Indeed, we have seen, for example, that the basis for allometric relationships is a matching of organisms of different sizes to their respective environments. Rather, it is a question of the evolutionary responses of a species to its habitat being limited by constraints that have themselves evolved.

Log body volume (mm3)

Clutch volume Body volume

,0 Allometric relationship from (a)

Log body volume (mm3)

Figure 4.33 Allometric relationships between total clutch volume and body volume in female salamanders. (a) The overall relationship for 74 salamander species, using one mean value per species (P < 0.01). (b) The relationships within a population of Ambystoma tigrinum (X) (P < 0.01), and within a population of A. opacum (o) (P < 0.05). The allometric relationship from (a) is shown as a short dashed line: A. opacum conforms closely to it; A. tigrinum does not. However, they both lie on an isometric line along which clutch volume is 13.6% of body volume (---). (After Kaplan & Salthe, 1979.)

allometry defined why are there allometric relationships?

These ideas are illustrated in Figure 4.33a, which shows the allometric relationship between clutch volume and body volume for salamanders generally. Figure 4.33b then shows the same relationships in outline; but superimposed upon it are the allometric relationships within populations for two salamander species, Ambystoma tigrinum and A. opacum (Kaplan & Salthe, 1979). If the species' means are simply compared, without reference to the general salamander allometry, then the species are seen to have the same ratio of clutch volume : body volume (0.136). This seems to suggest that the species' life histories 'do not differ', and that there is therefore 'nothing to explain' - but any such suggestion would be wrong. A. opacum conforms closely to the general salamander relationship. A. tigrinum, on the other hand, has a clutch volume which is almost twice as large as would be expected from that relationship. Within the allometric constraints of being a salamander, A. tigrinum is making a much greater reproductive allocation than A. opacum; and it would be reasonable for an ecologist to look at their respective habitats and seek to understand why this might be so.

In other words, it is reasonable to compare taxa from an 'ecological' point of view as long as the allometric relationship linking them at a higher taxonomic level is known (Clutton-Brock & Harvey, 1979). It will then be their respective deviations from the relationship that form the basis for the comparison. Problems arise, though, when allometric relationships are unknown (or ignored). Without the general salamander allometry in Figure 4.33a, the two species would have seemed similar when in fact they are different. Conversely, two other species might have seemed different when in fact they were simply conforming to the same allometric relationship. Comparisons oblivious to allometries are clearly perilous, but regrettably, ecologists are frequently oblivious to allometries. Typically, life histories have been compared, and attempts have been made to explain the differences between them, in terms of habitat differences. As previous sections have shown, these attempts have often been successful. But they have also often been unsuccessful, and unrecognized allometries undoubtedly go some way towards explaining this.

4.14.2 Effects of phylogeny

The approach used with the salamanders, of comparing species or other groups in terms of their deviations from an allometric relationship that links them, has been applied successfully to a number of larger assemblages. By removing the effects of size, the approach searches for phylogenetic relationships beyond those associated with size. For example, Figure 4.34 shows, for a number of mammal species, that 'relative' age at first reproduction increases as 'relative' life expectancy increases (i.e. relative to a value expected on the basis of an underlying allometry). This shows a powerful relationship between these two life history characters once the confounding effects of size have been removed. It also reveals underlying similarities between species of very different sizes: elephants and otters, and mice and warthogs.

A further impression of the strength of the influence of phylogeny can be gained from analyses like those in Table 4.8 (Read & Harvey, 1989). A nested analysis of variance has been applied to the variation in seven life history traits amongst a large number of mammal species. This has led to the determination of the percentage of the total variance attributable to: (i) differences between species within genera; (ii) differences between genera within families; and so on. Species vary very little within genera; genera vary little within families. Far and away the largest part of the variance, for all the traits, is accounted for by differences between orders within the mammalian class as a whole. This emphasizes that in simply comparing two species from different orders, we are in essence comparing those orders (which probably diverged many millions of years ago) rather than the species themselves. It does not mean, however, that comparing species comparing salamanders: dangerous if allometries are ignored removing the confounding effects of size

Table 4.8 When nested analyses of variance are performed on data sets for a number of life history traits from a large number of mammal species, the percentage of the variance is greatest at the highest taxonomic level (orders within the class) and least at the lowest level (species within genera). (After Read & Harvey, 1989.)

Table 4.8 When nested analyses of variance are performed on data sets for a number of life history traits from a large number of mammal species, the percentage of the variance is greatest at the highest taxonomic level (orders within the class) and least at the lowest level (species within genera). (After Read & Harvey, 1989.)


Species within genera

Genera within families

Families within orders

Orders within the class

Gestation length





Age at weaning





Age at maturity





Interlitter interval





Maximum lifespan





Neonatal weight





Adult weight




^ African elephant

Hippo ^ Red squirrel






Rrobiscoidea ♦

Rerissodactyla Artiodactyla Carnivora

♦ Insectivora






Relative annual fecundity

Figure 4.34 After the effects of size have been removed, age at first reproduction increases with life expectancy at birth for 24 species of mammals. 'Relative' refers to the deviation from the underlying allometric relationship linking the character in question to organism size. (After Harvey & Zammuto, 1985.)

in the same genus, say, is only 'scratching the surface'. Even when two species are very similar in their life histories and habitats, if one makes a greater reproductive allocation and also lives in a habitat that is lower CR, then this allows us to build a pattern linking the two.

Moreover, the strength of these relationships at the higher taxonomic levels does not mean that attempts to relate life histories to lifestyles and habitats should be abandoned even there, since lifestyles and habitats are also constrained by an organism's size and phylogenetic position. There may still, therefore, at these higher levels, be patterns linking habitats and life histories rooted in natural selection. For example, insects (small size, many offspring, high reproductive allocation, frequent semelparity) have been described as relatively r-selected, compared to mammals (large size, few offspring, etc. - relatively K-selected) (Pianka, 1970). Such differences could be dismissed as being 'no more' than the product of an ancient evolutionary divergence (Stearns, 1992). But, as we have stressed, an organism's habitat reflects its own responses to its environment. Hence a mammal and an insect living side by side are also almost certain to experience very different habitats. The larger, homeothermic, behaviorally sophisticated, longer lived mammal is likely to maintain a relatively constant population size, subject to frequent competition, and be relatively immune from environmental catastrophes and uncertainties. The smaller, poikilothermic, behaviorally unsophisticated, shorter lived insect, by contrast, is likely to live a relatively opportunistic life, with a high probability of unavoidable death. Insects and mammals are prisoners of their evolutionary past in their range of habitats just as they are in their range of life histories - and the r/K scheme provides a reasonable (although certainly not perfect) summary of the patterns linking the two.

The same point is illustrated in a more quantitative way by an application of the 'phylogenetic-subtraction' method (Harvey & Pagel, 1991) to patterns of covariation in 10 life history traits of mammals (Stearns, 1983). In the unmanipulated data, the pattern to be expected under the influence of r/K selection was pronounced: it accounted for 68% of the covariation. This r/K influence was reduced to about 42% when the effects of weight were removed, and was reduced still further when the trait values were replaced in the analysis by their deviations from the mean value for the family to which the species belonged (33%), or their order (32%). In the first place, this reaffirms the importance of both size and phylogeny, since each clearly accounted for much of the interspecific variation. It is also significant that the r/K pattern remained clearly visible even after these other effects had been removed. But the strength of the pattern in the unmanipulated data set cannot simply be dismissed as an artefact arising out of phylogenetic naivety. It may well be that important differences in habitat, too, are associated with an organism's size, or its order or family.

It is undoubtedly true that life ... yes! history ecology cannot proceed oblivious to phylogenetic and allometric constraints. Yet it would be unhelpful to see phylogeny as an alternative explanation to a role, still, for habitat?...

habitat in seeking to understand life histories. Phylogeny sets limits to an organism's life history and to its habitat. But the essentially ecological task of relating life histories to habitats remains the most fundamental challenge.

Was this article helpful?

+1 0
Lawn Care

Lawn Care

The Secret of A Great Lawn Without Needing a Professional You Can Do It And I Can Show You How! A Great Looking Lawn Doesnt Have To Cost Hundreds Of Dollars Or Require The Use Of A Professional Lawn Care Service. All You Need Is This Incredible Book!

Get My Free Ebook

Post a comment