Note: Tabulated values are percentage of the total variance accounted for at each successive level. The species level includes the error term in the data. *p < 0.05; **p < 0.01.

Note: Tabulated values are percentage of the total variance accounted for at each successive level. The species level includes the error term in the data. *p < 0.05; **p < 0.01.

-1.8 I I I I I I I I I I—I I I I I I I I I I I I I I I I I I I—r~i

Mean ambient temperature (°C)

Figure 1.5 Scatterplot showing the negative relationship in insects between mean annual temperature and metabolic rate, expressed here as the residuals from a generalized linear model including body mass, trial temperature, respirometry method, and wing status. The fitted line serves to illustrate the trend.

Source: Addo-Bediako et al. (2002). Functional Ecology 16, 332-338, Blackwell Publishing.

Nonetheless, the fact that there is variation about the major physiological themes has several important consequences. Most significant among these is that a comprehensive understanding of the ways in which insects respond physiologically to the world around them cannot be achieved solely by the study of model organisms (Feder and Mitchell-Olds 2003). The grounds for this view are not only rooted in the arguments provided above, but also go beyond these. In the first instance, it is clear that spatial variation among species at large geographic scales means that more than a handful of species must be investigated if this variation and its ecological implications are to be comprehended (Section 1.3). Indeed, most model taxa share a set of characteristics that are certainly not representative of all species. Model taxa tend to be widely distributed, fast-growing, relatively generalist in their food preferences, and not particularly fussy about the conditions under which they will survive, develop, and reproduce. If this were not the case, they would not be handy laboratory subjects, or model organisms. These very characteristics make them different from most other species.

Among the insects, most species are narrowly distributed, many feed on a limited range of host plants, and comparatively slow growth rates and long generation times are not unusual. Moreover, the large majority of species are less than happy to make anything but the most salubrious of laboratory environments their home. These characteristics are reflected in their physiologies too, and consequently they contribute substantially to physiological diversity. That there are likely to be important differences in the physiologies of common (and weedy) insects, and those that are rarer (and less ruderal) has long been appreciated. For example, both Lawton (1991) and Spicer and Gaston (1999) have argued that a comprehensive understanding of physiological diversity requires investigations of rare and common species. They have also suggested that such investigations could contribute substantially to our understanding of why some species are rare and others common, and why among the rare species some can survive rarity while others decline out of it and into extinction (Gaston 1994). These differences might also have potentially significant consequences for understanding the likely responses of species to environmental change (Hoffmann et al. 2003a).

Furthermore, while laboratory selection using model organisms is a useful approach for investigating the evolution of physiological attributes, and one that can circumvent many of the problems associated with comparative studies (Huey and Kingsolver 1993; Gibbs 1999), it is not without its problems, including those of repeatability and laboratory adaptation (Harshman and Hoffmann 2000; Hoffmann et al. 2001a). Among the insects, most laboratory selection experiments have also been based on a single insect order, the Diptera, and indeed often on a single family, the Drosophilidae. Because variation in many physiological characteristics is partitioned at the family and order levels (Chown et al. 2002a), results based on a single family or order might not be reliably generalized across the insects. Therefore, the generality of the findings based on model organisms must obviously be confirmed by investigations of physiological traits in selected higher taxa, and especially in non-model organisms that are often unwilling inhabitants of the laboratory (Klok and Chown 2003). If such investigations are undertaken within a modern comparative framework, they can complement and validate laboratory selection experiments (Huey and Kingsolver 1993), and supplement comparative work on other taxa. Thus, by broadening the scope of comparative work (and laboratory selection), a comprehensive understanding of physiological variation can be achieved (Kingsolver and Huey 1998).

Investigations at a variety of scales are required to develop a full picture of the patterns of physiological variation and the mechanisms underlying them (Chown et al. 2003). These range from mechanistic molecular studies (Flannagan et al. 1998; Jackson et al. 2002; Storey 2002), through intra-specific comparisons, to broad-scale comparative analyses. To some extent, this approach presupposes that the question being posed determines the scale of the study (as does the magnification used in a microscopic examination—10 000 x is just no good for understanding variation in the body form of dung beetles; Chown et al. 1998). Combining different-scale investigations into a single study can also be particularly powerful, as recent work on the likely responses of Drosophila birchii to environmental change has shown. This species shows geographic variation in desiccation resistance that tracks environmental variation in water availability. However, laboratory selection experiments have indicated that populations at the range margin are unlikely to be able to further alter their resistance owing to a lack of genetic variation in this trait (Hoffmann et al. 2003a).

1.3 Diversity at large scales: macrophysiology

While it is clear that studies at a variety of scales are required to understand physiological variation and its causes, studies at large geographic scales (in the ecologist's sense of large scale—that is, covering large geographic areas such as whole continents) are comparatively recent. Similar approaches were occasionally adopted in the past (Scholander et al. 1953), and qualitative comparisons of insect responses across large scales have also been undertaken since then (e.g. Zachariassen et al. 1987). However, extensive quantitative studies of broad-scale physiological variation are new.

The impetus for this work has come, first, from macroecology: the investigation of large-scale variation in species richness, and species abundances, distributions, and body sizes, through space and time (Brown and Maurer 1989; Gaston and Blackburn 2000). One of the many strengths of macro-ecology is its ability to reveal 'where the woods are, and why, before worrying about the individual trees' (Lawton 1999). In doing so, macroecological studies make several assumptions, some of which have to do with large-scale variation in physiological responses. It is these assumptions that have prompted large-scale investigations of variation in insect physiological traits, which likewise worry less about the trees than the woods.

For example, one of the explanations proposed for the latitudinal gradient in species richness relies on the Rapoport effect, or the increase in the latitudinal range sizes of species towards higher latitudes (Stevens 1989). The mechanism that is apparently largely responsible for producing this macroecological pattern is straightforward. To survive at higher latitudes individual organisms need to be able to withstand greater temporal variability in climatic conditions than at lower latitudes. This idea, known as the climatic variability or seasonal variability hypothesis, rests on two critical assumptions. First, that towards higher latitudes climates become more variable (Fig. 1.6). Second, that species at higher latitudes have wider climatic (physiological) tolerances than those at lower latitudes (Addo-Bediako et al. 2000). The second assumption plainly can only be examined by adopting a large-scale comparison of physiological tolerances. Few studies have made such broad-scale comparisons, but in those cases where this has been done the physiological variability assumption appears to have been met (Gaston and Chown 1999a; Addo-Bediako et al. 2000) (Fig. 1.7).

A further example of macroecological questions that rely on information concerning geographic variation in physiological traits concerns spatial variation in body size (Gaston and Blackburn 2000). The mechanisms underlying body size variation, at either the intraspecific or interspecific levels, are firmly rooted in assumptions concerning geographic variation in physiological traits. These assumptions generally have (incorrectly) to do with thermoregulation, or correctly in the case of insects, with interactions between the temperature dependence of growth and development, and resource availability (Chown and Gaston 1999).

Was this article helpful?

0 0

Post a comment