The goal of science is a consensus of rational opinion over the widest possible field.

Ziman (1978)

Few, if any, insect species occur everywhere. At global scales, geographic ranges are generally small, and only a handful of species is distributed across several continents or oceans. Among the many reasons for this preponderance of narrow geographic distributions are the major barriers presented by continental and oceanic margins and the limited dispersal powers of many species. However, even within continents, range size frequency distributions are right-skewed. Most species have ranges much smaller than the total size of the continent (Fig. 1.1), which indicates that geographic ranges are limited.

The factors responsible for range size limitation have been much debated. For example, MacArthur (1972) suggested that for Northern Hemisphere species, biotic factors, such as predation, disease, or competition, limit ranges to the south, whereas the northern limits of species are set by abiotic factors such as temperature extremes. The ecological reasons for species' borders and the hypotheses proposed to explain borders are diverse (Hoffmann and Blows 1994). Horizontal (e.g. competition) and vertical (e.g. parasitism) interactions between species are thought to be significant factors limiting distributional ranges, as is the influence of the abiotic environment (e.g. Quinn et al. 1997; Davis et al. 1998; Hochberg and Ives 1999; Gaston 2003). Moreover, in many instances species interactions and abiotic conditions probably combine to produce range margins (Case and Taper 2000). This complexity means that range margins are unlikely to be set for the same reasons in any two species. Nonetheless, abiotic conditions probably play a role in setting at least a part of many species geographic range limits (Gaston 2003). That is, species are unable to

Geographic range size

Figure 1.1 Species range size frequency distribution for keratin beetles (Trogidae) in Africa. The units are numbers of grid cells of 615 000 km2.

Geographic range size

Figure 1.1 Species range size frequency distribution for keratin beetles (Trogidae) in Africa. The units are numbers of grid cells of 615 000 km2.

Source: Gaston and Chown (1999b).

survive and reproduce under the full range of abiotic conditions that might be found on a continent. Quite why this view is held is largely a consequence of clear instances of species borders being correlated with a given climatic variable (e.g. Robinson et al. 1997), species range shifts being associated with changes in climate (usually temperature), both currently (Parmesan et al. 1999) and in the past (Coope 1979) (reviewed in Chown and Gaston 1999), and experimental work involving caging studies (« reciprocal transplants) showing that individuals often struggle to survive a short distance beyond their natural ranges (Jenkins and Hoffmann 1999).

These findings, in turn, raise the question of why species have been unable to alter their tolerances of abiotic conditions and in so doing expand their ranges. From an evolutionary perspective there are several reasons why this might be the case, including genetic trade-offs, low heritability, low levels of genetic variation, and the swamping of marginal by central populations (reviewed in Hoffmann and Blows 1994; Gaston 2003). Indeed, there appear to be many grounds for assuming that species might never alter their physiologies in response to environmental change. However, this has clearly not happened, or else large parts of the Earth would be uninhabited. Thus, although there are limits to the conditions that insects can tolerate, they are not helpless in the face of changing abiotic conditions. Apart from moving, which many species clearly do, insects can respond to changing conditions in two ways. Over the short term, they can do so by means of phenotypic plasticity or short-term changes in their phenotype. Over the longer term their responses include adaptation, either of basal responses to the environment or of the extent of plasticity, or both. The outcome of these responses and the limited dispersal abilities of most organisms are what we see today—spatial and temporal variation in diversity, including physiological diversity.

1.1 Physiological variation

There are many ways in which physiological diversity (or variation) is manifested (Spicer and

Gaston 1999). Individuals vary through time for several reasons. Some individuals appear to be intrinsically variable, such as those of the cockroach Perisphaeria sp. (Blaberidae), where single individuals can show as many as four gas exchange patterns at rest (Fig. 1.2) (Marais and Chown 2003). Individual-level variation also arises as a consequence of ontogeny. For example, cold hardiness varies substantially between larval instars and between adults and larvae in sub-Antarctic flies (Vernon and Vannier 1996; Klok and Chown 2001). A variety of physiological traits also show changes associated with responses to the immediate environment. Indeed, such individual-level variation has been demonstrated in many species over a variety of time-scales, ranging from a few hours to an entire year (Lee et al. 1987; Hoffmann and Watson 1993; van der Merwe et al. 1997; Klok and Chown 2003). This phenotypic flexibility can be reversible, in which case it is considered acclimation or acclimatization, or fixed, and is then referred to as a developmental switch or polyphenism (Huey and Berrigan 1996; Spicer and Gaston 1999).

Cross-generational effects are also a form of phenotypic flexibility. Although the extent and significance of cross-generational effects, or the influence of parental or grandparental environmental history on an individual, has been widely examined for insect life history traits, the same is not true for physiological characters. Nonetheless, there have been some investigations of this kind, mostly involving Drosophila flies (Huey et al. 1995; Crill et al. 1996; Watson and Hoffmann 1996). Parental and grandparental exposure to stress can have substantial influences not only on early offspring fitness, but also on development time, and these responses can often differ in size and direction between male and female parents and between grandparents and parents (Magiafoglou and Hoffmann 2003).

Physiological variation is also a consequence of genotypic diversity. While this variation can differ considerably depending on the trait of interest (Falconer and Mackay 1996), and whether the species is clonal (parthenogenetic) or not, it is nonetheless of considerable importance. Consistent among-individual variation (which can be assessed as repeatability of a trait) is a prerequisite for natural selection (Endler 1986), which in turn is one of the major reasons why there is such a range of physiological diversity today. Curiously, the relationship between within- and among-individual variation in insect physiological traits has not been widely assessed despite its importance in determining whether the conditions for natural selection are met (Bech et al. 1999; Dohm 2002). It might be argued that laboratory selection (see Gibbs 1999) has adequately demonstrated that the conditions for selection have been fulfilled, but this is true only of a restricted set of traits and a relatively small number of taxa which are at home in the laboratory. For many other taxa, adaptation is simply assumed. To date, investigations of repeatability in insect physiological traits have largely been restricted to gas exchange and metabolic characteristics (Buck and Keister 1955; Chappell and Rogowitz 2000; Marais and Chown 2003), which have shown that repeatability tends to be both significant and high.

Individuals also vary through space as a consequence either of their membership of different populations of a given species or as consequence of their species identity. Indeed, inter-population and interspecific differences in physiological traits have been widely examined in insects over many years, and this comparative approach forms a cornerstone of physiology (Somme 1995; Chapman 1998; Nation 2002). While the criteria by which this variation is judged to be adaptive have become more stringent (Kingsolver and Huey 1998; Davis et al. 2000), and the tools on which these decisions can be based more sophisticated (Garland et al. 1992; Freckleton et al. 2002), spatial variation in physiological characteristics, over large (Fig. 1.3), and small (Fig. 1.4) scales, continues to form much of the foundation for modern understanding of the evolution of physiological diversity (Spicer and Gaston 1999; Feder et al. 2000a).

1.2 How much variation?

Given that physiological traits can vary with time of day (Sinclair et al. 2003a), season (Davis et al. 2000), instar (Klok and Chown 1999), stage (Vernon and Vannier 1996), and through space, it might be tempting to conclude that physiological traits are highly subtle in their variation. In other words, that insect physiological traits show such a bewildering complexity of variation (Hodkinson 2003), that they defy anything other than a species by species, stage by stage, season by season investigation, if their evolution and variation are to be comprehended.

If this were the case, physiological ecology could claim to know very little about insects. The reasoning is simple. Despite the many studies that have been done on a broad array of physiological traits, comprehensive knowledge is available for just a few taxa. These are the vinegar fly (Drosophila melanogaster), honeybee (Apis mellifera), Colorado potato beetle (Leptinotarsa decemlineata), migratory locust (Locusta migratoria), American cockroach (Periplaneta americana), tobacco hornworm (Manduca sexta), South American assassin bug (Rhodnius prolixus), and the mealworm (Tenebrio molitor) (see Chown et al. 2002a). Much of modern physiological understanding, which forms the foundation for physiological ecology, is based on investigations of these and a few other species (e.g. Grueber and Bradley 1994). However, even conservative estimates place insect species richness at roughly two million (Gaston 1991). Do these estimates mean that little is known about insect physiological responses to their environments, and that generalizations are not possible? To paraphrase Lawton (1992), the answer plainly lies in whether there are 10 million kinds of insect physiological responses. In our view there are not (see also Feder 1987). Rather, just as there is only limited variation about several population dynamic themes (Lawton 1992), there is limited variation about several major physiological themes for the traits in question (Chown et al. 2003). Several lines of evidence support this view:

1. Recent work has demonstrated that a considerable proportion of the variation in several physiological traits is partitioned at higher taxonomic levels (Chown et al. 2002a) (Table 1.1). Indeed, it is clear that a majority of the variation is often partitioned above the species level, and often at the family and order levels. This would not be


Was this article helpful?

0 0

Post a comment