Do humans homogenize or differentiate biotas? It depends
2008; Wiley; Volume: 35; Issue: 11 Linguagem: Inglês
10.1111/j.1365-2699.2008.02011.x
ISSN1365-2699
Autores Tópico(s)Wildlife Ecology and Conservation
Resumo'Biotic homogenization represents a unique challenge to ecologists because it is a multifaceted process encompassing many dimensions of the modern biodiversity crisis, including species invasions and extirpations, and it requires the explicit consideration of how the identities of species (not species richness) change over both space and time.'Olden & Rooney (2006) p. 113. A common concern of globalization is that humans are homogenizing the planet. This is captured in such colourful descriptions of our future planet as: the geography of nowhere, planet of weeds, the new Pangea and the Homogecene (reviewed in Olden et al., 2005; Olden, 2006). This notion of a homogenizing biosphere is, however, seductively simplistic. As one begins to examine them in more detail, the effects of human activities on the biosphere are seen to be much more complicated than these metaphorical concepts imply. For example, although biological homogenization (BH) is commonly associated with the introduction of widespread (invasive) species, it is also increased by the extinction of local (endemic) native species (McKinney & Lockwood, 1999; Olden, 2006). Furthermore, the degree of biological homogenization depends on how we measure it. Genetic homogenization is the mixing into native gene pools of exotic genes (Olden & Rooney, 2006). Functional (or ecological) homogenization is the increase in similarity of ecological roles or traits between communities (McKinney & Lockwood, 1999; Olden & Rooney, 2006), such as the increase in generalist species (and the decline in specialist species) that occurs with increasing anthropogenic disturbance (Devictor et al., 2008). BH is most commonly measured in terms of 'taxonomic homogenization' (Olden & Rooney, 2006); that is, by the relative gain in the number of shared species between two areas as measured by pairwise similarity indices (e.g. Jaccard's). This metric can be extended to construct cumulative species–area curves and other methods to examine the loss of beta diversity (La Sorte et al., 2008). Yet another complication is that most studies of taxonomic homogenization have access only to species presence–absence data, whereas a fuller assessment of homogenization would include the abundance of shared and unshared species. Fortunately, the omission of abundance data may not change the basic results in most cases (Cassey et al., 2008). Aside from types of measurement, several other complexities arise in assessing human impacts on reshuffling biological communities. Two of the most important relate to scale, both temporal and spatial (Olden, 2006). Regarding temporal scale, anthropogenic homogenization almost certainly increases with time, as long as human disturbances (e.g. farming, urbanization) continue to intensify (McKinney, 2006). This has been a major obstacle for many homogenization studies, which typically compare existing communities, with little or no data on past (and future) changes in community composition. For example, the omission of data on the number of native species that have gone extinct from recent human activities would probably lead to an underestimation of previous homogenization if extinction is biased towards rare endemic species (probably not shared between the communities compared). Future homogenization will be underestimated where data on currently threatened species, likely to go extinct, are not included. Furthermore, many introduced species are just beginning to spread and will ultimately have a more homogenizing effect than their current distribution indicates. Regarding spatial scale, comparisons of communities separated by large distances show that species introductions consistently have a homogenizing impact (Qian & Ricklefs, 2006). This is because communities separated by large distances share few, if any, native species, so that even a few shared introduced species will increase community similarity. In contrast, communities separated by short distances often show reduced similarity (differentiation) from species introductions, because introduced species are shared to a lesser extent than native species owing to a high spatial autocorrelation of native communities (La Sorte et al., 2008). Fortunately, as interest in homogenization grows, researchers are employing increasingly sophisticated methods to study it. A good example is a recent study by Spear & Chown (2008), who examined the effects of ungulate introductions on biotic similarity across four spatial scales, at three resolutions within South Africa and among 41 nations. In addition to comparing various spatial scales, the study included the effects of temporal factors such as historical native species extirpations and future native species extirpations (based on hypothetical extirpations of currently threatened and endangered species). Furthermore, by examining ungulates, the study benefited by focusing on a group (mammals) that has generally been omitted from the homogenization literature. The large majority of studies thus far have focused on plants, fishes and birds (Olden, 2006). The Spear and Chown study thus provides a unique and exceptionally reliable insight into the effects of our propensity simultaneously to eradicate local species and import species from many different areas. It is therefore very significant that their findings generally reinforce previous findings based on other data sets involving different taxa. Notably, ungulate species introduced (translocated) from less distant localities (other areas of South Africa) tend to play a much larger role in homogenizing local faunas than do species introduced from more distant areas. Indeed, the latter (extra-regional) species often have the opposite effect of reducing similarity among South African ungulate assemblages. This pattern of differentiation by extra-regional introductions and homogenization by intra-regional introductions has also been found in plants and freshwater fishes (McKinney, 2005). The role of spatial scale is further demonstrated, with perhaps the best analysis of which I am aware, in showing that ungulate species assemblages separated by short distances tend to be differentiated by introduced species whereas distant assemblages are nearly always homogenized by introduced species. Although other studies have shown this effect in other taxa (e.g. in plants by Qian & Ricklefs, 2006), the Spear and Chown study has an enlightening analysis illustrating the role of spatial autocorrelation among native assemblages in causing this effect. Because neighbouring assemblages share many native species, there is a high probability that introductions will involve fewer shared species than the number of shared native species. This is especially true where human introductions are 'haphazard' (Qian & Ricklefs, 2006) and introduced species have not had time to disperse into adjacent areas. By contrast, native species assemblages that are widely separated (e.g. between distant nations) often have no species in common and therefore the introduction of even a few shared non-native species will have a homogenizing effect. Spear & Chown (2008) identify a 'hard boundary' of about 4000 km for ungulates such that comparisons of assemblages in excess of this distance apart generally reveal a homogenizing effect of introduced species because no native species are shared. Assemblage comparisons below this boundary often show a differentiating effect of introductions. These factors also influence the spatial resolution of assemblage comparisons. Coarse-grained scales of comparison tend to show homogenizing effects of introductions, whereas comparisons of communities at fine spatial grains often show differentiating effects (Spear & Chown, 2008). It may seem obvious that anthropogenic homogenization tends to increase with time but few studies have had the data to test this hypothesis. Spear & Chown (2008) showed that homogenization does increase through time and that, thus far, this has been driven primarily by introductions rather than by native extirpations. This latter finding is of considerable interest because homogenization can, theoretically, be driven by either (or both) introductions or extirpations (Olden, 2006). Teasing these variables apart empirically as done by their study is therefore a substantial contribution. Interestingly, this dynamic may change, as the future homogenization of ungulates could be driven more by native extirpations if species now at risk become extinct (Spear & Chown, 2008). Homogenization studies are clearly in their infancy but it seems obvious that this area of research can only grow exponentially as human influence, especially urbanization, spreads across the planet (McKinney, 2006). The complex dynamics of homogenization, driven by the complex interplay of species additions (introductions) and deletions (extirpations) across several scales of time and space, represents a significant challenge to biologists. Understanding these dynamics will require many studies of many taxa. It will also require greatly expanding the rather narrow ways that homogenization has been analysed. Studies of taxonomic homogenization should include species abundance data wherever possible. Genetic and functional (ecological) homogenization should also be examined to a much greater extent than is now typically the case. There are many practical benefits for conservation from homogenization studies (Olden, 2006). Conceptual benefits also abound. Perhaps most notably they provide a more complete way of visualizing human impacts on the biosphere than can be obtained by focusing only on species extinctions or species introductions as isolated and separate processes. Editor: John Lambshead
Referência(s)