Ecology. Michael Begon

variation on a small scale

      Differentiation within a species can occur over a remarkably small geographic scale. In the case of sweet vernal grass, Anthoxanthum odoratum, growing along a 90 m transition zone between mine and pasture soils at the Trelogan zinc and lead mine in Wales, there was a striking increase in evolved tolerance to zinc, at otherwise toxic concentrations, over a distance of only 3 m within the zone. In this case, any counteracting mixing and hybridisation of the ecotypes was reduced because plants growing on the mine soil tended to flower later than their counterparts in the pasture (Antonovics, 2006).

      …. and a large scale

In a study with a much broader geographic range, common frogs (Rana temporaria) were monitored over a latitudinal gradient encompassing Sweden and Finland. Geographic variation within species is generally studied both in situ and using a ‘common garden’ approach, where individuals from different sites are transplanted and grown together, thus eliminating any influence of immediate environments. In this case, while there was considerable variation in tadpole development time (from complete gill absorption to emergence of the first foreleg), no consistent trend with latitude was evident (Figure 1.2a). However, when tadpoles from different sites were reared in a common environment, at a range of temperatures, those from higher latitudes developed significantly faster. There had clearly been local adaptation, and frogs experiencing colder temperatures (at higher latitudes) had evolved compensatory increases in development rate. The net result was that development times were similar at different latitudes.

Figure 1.2 At a given temperature, tadpoles from higher latitudes developed faster than those from lower latitudes. (a) Tadpoles from ponds in two areas of Sweden, in the south, and from Finland, in the north, showed variation in development times but no consistent trend with latitude. (b) When tadpoles from sites at various latitudes were reared in the laboratory at different temperatures, those from higher latitudes consistently developed fastest. Temperatures: 14°C (yellow circles), 18°C (blue circles), and 22°C (red circles).

      Source: From Laugen et al. (2003).

      APPLICATION 1.1 Selection of ecotypes for conservation

The sapphire rockcress, Arabis fecunda, is a rare perennial herb restricted to calcareous soil outcrops in western Montana (USA) – so rare, in fact, that there are just 19 existing populations separated into two groups (‘high elevation’ and ‘low elevation’) by a distance of around 100 km. Whether there is local adaptation is of practical importance for conservation: four of the low‐elevation populations are under threat from spreading urban areas and may require reintroduction from elsewhere if they are to be sustained. Reintroduction may fail if local adaptation is too marked. Observing plants in their own habitats and checking for differences between them would not tell us if there was local adaptation in the evolutionary sense. Differences may simply be the result of immediate responses to contrasting environments made by plants that are essentially the same. But once again, the ‘common garden’ approach circumvents this problem. The low‐elevation sites were more prone to drought – both the air and the soil were warmer and drier – and the low‐elevation plants in the common garden were indeed significantly more drought tolerant (Figure 1.3). More generally, we need to improve our understanding of local adaptation, and its genetic basis, because of their importance for the conservation and restoration of genetic resources, and for crop and animal production, and this is of particular significance in a changing climate (McKay et al., 2005; Savolainen et al., 2013).

Figure 1.3 Local adaptation of rare sapphire rockcress plants. When plants of the rare sapphire rockcress from low‐elevation (drought‐prone) and high‐elevation sites were grown together in a common garden, there was local adaptation: those from the low‐elevation site had significantly better water‐use efficiency as well as having both taller and broader rosettes.

      Source: From McKay et al. (2001).

      the balance between local adaptation and hybridisation

      On the other hand, local selection by no means always overrides hybridisation. In a study of Chamaecrista fasciculata, an annual legume from disturbed habitats in eastern North America, plants were grown in a common garden that had been derived from the ‘home’ site or were transplanted from distances of 0.1, 1, 10, 100, 1000 and 2000 km (Galloway & Fenster, 2000). The study was replicated three times: in Kansas, Maryland and northern Illinois. Five characteristics were measured: germination, survival, vegetative biomass, fruit production and the number of fruit produced per seed planted. But for all characters in all replicates there was little or no evidence for local adaptation except for transplant distances of 1000 km or more. There is ‘local adaptation’ – but in this case it was clearly not that local.

We can also test whether organisms have evolved to become specialised to life in their local environment in reciprocal transplant experiments: comparing their performance when they are grown ‘at home’ (i.e. in their original habitat) with their performance ‘away’ (i.e. in the habitat of others). In his meta‐analysis of 74 reciprocal transplant studies (50 concerning plants, 21 animals, two fungi and one protist), Hereford (2009) reported that local adaptation was common (71% of studies) but not ubiquitous. On average, local populations had 45% greater fitness than non‐local populations. And crucially, there was a small but significant positive association between fitness differences and the magnitude of environmental differences between parental sites (‘environmental distance’ measured using composite values for up to four environmental variables, such as soil moisture, annual rainfall, elevation and frequency of predation) (Figure 1.4). The magnitude of local adaptation does not seem to be correlated with geographic distance (Leimu & Fischer, 2008), so Hereford’s results emphasise the role of ecological factors, not separation itself, as drivers of adaptive differentiation.

Figure 1.4 Meta‐analyses reveal generalities about local adaptation. Regression of local adaptation on environmental distance between sites in a meta‐analysis of reciprocal transplant experiments (P = 0.003). Local adaptation is the difference in relative fitness between a native population and a non‐native population in the native’s environment. To standardise measures of environmental difference between sites, Euclidean distances from the means of environmental variables were calculated for all sites in each study.

      Source: