There is now a consensus that the dominance of Homo sapiens within the Earth's ecosystems during the last 50,000 years has precipitated an extinction event that rivals the ‘big five’ extinctions of the geological past (Mace et al., 1998). Hence the growth of the discipline of conservation biology, the aim of which is to identify the sorts of natural and anthropogenic processes that we need to be concerned about, and simultaneously to identify the strategies that should be used to minimize human impacts on the various components of biodiversity.
Surprisingly, most biogeographers and macroecologists have tried to avoid this problem by using data-gathering strategies to understand ‘natural’ processes alone, factoring out the potential problems caused by the addition of current anthropogenic effects. It is a common practice, for example, to use historical ranges to evaluate processes underlying range size, shape and placement, and to combine historical ranges to generate patterns of species richness. The reasoning that underlies this approach is that human impacts are ‘recent’, and that patterns of range shift from historical to current times are useful to distinguish natural and anthropogenic components of range collapses (Channell & Lomolino, 2000).
A related issue is the belief that human impacts are spatially restricted, or focused, and that, although their influence can obviously diffuse in space, in most cases the broad-scale consequence of this diffusion is a collapse by range fragmentation of native species [Gaston's (2003) Model I, see p. 169]. Indeed, extinction always starts as a consequence of local population dynamics and continues as long as processes related to extinction vortices (increased demographic stochasticity, inbreeding depression, reduced ability for adaptation) continue to reinforce the probability of extinction. However, the way local extinction processes spread throughout a species range is subject to many complex factors at the landscape level, both intrinsic and extrinsic to populations and species. Thus, under this scenario, the methodological strategy of using very large grain sizes (cell sizes) to obtain range data, based on crude measures of extents of occurrence, almost ensures that macroecological patterns will be minimally affected by ‘non-natural’ anthropogenic effects. Indeed, this was a very common practice in the first macroecological studies (see Gaston & Blackburn, 2000). As a consequence, it is expected that both climatic and historical factors usually provide powerful explanations for current diversity patterns at these large grain sizes, including for North American avifauna (see Hawkins & Porter, 2003; Hawkins, 2004).
Detailed sampling programs that include data collected at smaller grain sizes are starting all over the world, and access to digital data bases from collections and museums is increasing rapidly. Moreover, broad-scale remotely sensed environmental data are fast becoming available. These improved data are also being analysed with new powerful modelling approaches (Segurado & Araújo, 2004), allowing the development of robust and detailed descriptions of geographic ranges and other derived macroecological and biogeographical patterns, such as species richness. This framework has been used mainly in a conservation context (e.g. Araújo & Williams, 2000), but recent adoption of these practices by macroecologists and biogeographers highlights an old question: are the community-level and biogeographical patterns observed today truly independent of recent human influences and associated modern ecological processes, being driven only by deep-time ecological and evolutionary processes?
In a paper published in this issue of Journal of Biogeography, LaSorte (2006) addressed questions relevant to understanding the degree of human influence on current macroecological patterns. He used a detailed data set provided by the North American Bird Breeding Survey (BBS) (see http://www.pwrc.usgs.gov/bbs) to evaluate changes in the geographic ranges of 453 bird species over four time periods between 1968 and 2003. Two characteristics of species’ ranges, the extent of occurrence (the outermost limits of a species’ range) and the area of occupancy (the number of BBS routes in which a species appears), were investigated. These characteristics of species’ ranges were standardized to vary between 0 and 1, by making them relative to total area of occupancy and total number of BBS routes.
First, LaSorte (2006) showed that range expansion was more frequent (51% of the species) than range contraction (28% of the species). This pattern was a consequence of increases in both range extent and occurrence, which were accentuated and leveraged by a few species with extreme range shifts. He also studied the composition of species at the assemblage level, using the median of extents and occurrences of species across the 1673 BBS routes sampled, which also revealed the same consistent pattern in range expansion. More interestingly, average species richness for BBS routes characterized by range expansion did not change significantly between 1968 and 2003, whereas routes with contracting ranges tended to gain species. In short, high numbers of species tended to expand their ranges and become more common within the US, with an increased prevalence within BBS routes.
How can we interpret these results? First, a warning: it does necessarily mean that we do not need to worry about conserving US birds just because some of these species are now expanding their ranges and becoming more common. It is important to realize that the first time-step of the study was 1968, a time at which most of the natural areas in the US were already lost. We have no reason to suspect that comparing current and deep historical ranges for these species will not reveal the same pattern as described by Channell & Lomolino (2000) (indeed, they included a few North American bird species in their analyses). Many bird species were probably lost in the last ten thousand years or so, after the arrival of more dense human populations in North America. Furthermore, these patterns suggest a degradation of biodiversity in some aspects, especially β-diversity, since some species are becoming more widespread and can be found in many places, and these species replace rare and local species in some local assemblages.
On the other hand, of course, it might be possible to see an optimistic message in the high frequency of range increase and commonness. The gain in species for BBS routes is the result of a successful colonization of common species and the local presence of species that are becoming more common over time. There was indeed previous evidence that many species that are expanding their ranges and becoming more common are those benefiting directly or indirectly from anthropogenic effects related to changes in landscapes, in most cases related to interactions between preserved and disturbed areas at local scales. Understanding these interactions at the landscape level can provide further insights for improving conservation strategies for many species.
More importantly, from a theoretical and methodological perspective, LaSorte (2006) paper clearly shows that modern processes, most probably associated with human activities, have played an important role in shaping the current macroecological patterns of birds in North America. These processes have affected the structure of assemblages in their overall richness and their relative composition, in terms of distributions of rare and common species. Further studies are necessary to evaluate the real role of anthropogenic activities in the patterns found, because tests were indirect and conclusions regarding these effects were based on the assumptions of the study rather than on empirical analyses. LaSorte (2006) analysed only one level of anthropogenic activity, because the BBS routes chosen for the analyses were located in regions of ‘moderate’ human influence. Data across a gradient of human influence can clarify the mechanisms underlying changes in assemblage attributes at the landscape level. There is little doubt, however, that a major component of modern processes affecting landscapes is human activity. In addition, it remains a priority for us to continue to study how geographic range shifts are related to other attributes of ecological communities at more local scales (i.e. abundance, niche breadth, and so on), and how recent climatic changes (now also attributed in great extent to human activities), and even recent conservation actions, may be involved with these patterns too. Nevertheless, the study by LaSorte (2006) lends weight to Gaston's (2004) call for a better understanding of the role that humans and modern processes play in shaping macroecological patterns.