No species occurs everywhere. Indeed, most are absent from the vast majority of sites across the globe. Those areas in which a species does occur constitute its geographic range. As such, the geographic range is one of the fundamental units in ecology. The sizes and distribution of geographic ranges give rise to patterns of species richness and change in species composition from site to site, and combined with their abundance and trait structure give rise to other spatial patterns in assemblages. Likewise, temporal changes in assemblages on both short and long time scales follow from changes in the size, position, and structure of geographic ranges.
area of occupancy. The area within the outermost geographic limits to the occurrence of a species over which it is actually found
extent of occurrence. The area within the outermost geographic limits to the occurrence of a species
intraspecific species-abundance distribution. The frequency of areas within a species’ geographic range in which it attains different levels of abundance
range edge or limit. The outermost geographic occurrences of a species, usually excluding vagrant individuals
species–range size distribution. The frequency of species with geographic ranges of different sizes
The sizes of the geographic ranges of species vary dramatically and can be characterized in two fundamentally different ways. Extent of occurrence is the area within the outermost limits to the occurrence of a species, and area of occupancy is the area over which the species is actually found. The latter will tend to be consistently smaller because no species is distributed continuously across space even within the broad geographic limits to its occurrence. The finer the spatial resolution and the shorter the time period over which area of occupancy is measured, the smaller will be the area over which the species is documented to occur, and the greater this disparity will be. At one extreme lie those, predominantly freshwater or terrestrial, species that are currently found occurring in a single small habitat patch (often with only a very small number of individuals), which are thus narrowly distributed in terms both of extent of occurrence and area of occupancy. At the other extreme lie some marine organisms. Species of microorganisms may be widespread across the oceans both in terms of extent of occurrence and area of occupancy, whereas some large-bodied species of vertebrate may have large oceanic distributions in terms of extent of occurrence but, because of the relatively low numbers of individuals, not area of occupancy.
Both within and across major taxonomic groups, the geographic ranges of the majority of species are relatively small, and only a very few are widespread. Indeed, within such groups species–range size distributions, the frequency of species with ranges of different sizes, are almost invariably strongly right-skewed. One important consequence is that the vast majority of occurrence records result from a small number of species. For example, by one estimation, at a spatial resolution of approximately 100 × 100 km, the 10% most globally widespread extant species of birds account for 50% of occurrence records. Given that the ratio of extents of occurrence to areas of occupancy may often be proportionately larger for rare species than for widespread ones, that is, they occupy their ranges less densely, the dominance of occurrence records by widespread species may increase when documented at finer spatial resolutions. This dominance may explain why it is the more widespread rather than, as often assumed, the restricted species that contribute disproportionately to spatial variation in species richness and related macro-ecological patterns.
The average sizes of geographic ranges can vary markedly between species in different major taxonomic groups. Thus, among nonmarine vertebrates, species of fish and amphibians tend naturally to have smaller ranges than do mammals, and mammals smaller ranges than do birds. However, within taxonomic groups, the extent to which the geographic range sizes of species exhibit phylogenetic constraint is contentious. Certainly range size is not as strongly conserved as are body size and many life history traits. Even where significantly conserved, it typically remains impossible to predict with any accuracy the range size of a species from that of its sister species or other close relatives, suggesting that such heritability has limited practical value (e.g., in estimating the range sizes of species whose distributions have not been well documented). This would tend to follow if the range sizes of different species are determined by the variable outcomes that result from the combinations of individual traits and environmental conditions occurring at particular times and places.
The mean size of the geographic ranges of the species within a higher taxon tends to vary spatially. Most obviously, ranges are typically smaller in situations in which dispersal and environmental conditions are geographically highly constrained, such as on islands and at high elevations, and in specialized habitats (e.g., desert springs, deep sea vents). However, more systematic spatial patterns have also been argued to occur, in particular, increases in the latitudinal extent of ranges from low to high latitudes, in their altitudinal extent from low to high elevations, and in their depth extent from shallow to deep waters. The first of these is a phenomenon termed Rapoport’s rule. The pattern appears to be most evident in the terrestrial northern hemisphere but may actually reflect a general trend for terrestrial ranges to increase from high southern to high northern latitudes. Although other factors may also have an influence, this trend is at least in part a result of changes in land area.
More difficult to establish than patterns of spatial variation in geographic range sizes are the long-term temporal trends. How the mean range sizes of the species in a higher taxon have changed over geologic time remains virtually unknown (although it is likely to have been marked, given changes in the distributions of land masses, water bodies, and climatic conditions). Little more is understood about how the range of an individual species changes in size between its origination and its extinction. However, best evidence suggests that geographic ranges typically undergo a rapid increase in size following speciation and then a slower subsequent, and perhaps prolonged, decline to extinction. This is supported by studies of species introduced into areas in which they previously did not occur, which have revealed that following an initial lag phase, during which a species tends to remain rather restricted to the locale of its introduction and densities there tend to build up, spread can occur across large areas very rapidly (both phases are extremely short in terms of evolutionary time).
When we focus on the events at the outset and conclusion of a species’ lifespan, geographic range size influences both the likelihood of speciation and that of extinction. At least when allopatric, the likelihood of speciation appears to be related to range size by a hump-shaped function. As ranges increase from small to moderate sizes, the likelihood of speciation increases because the chance of the range being bisected by a barrier to dispersal increases. However, at some point ranges will become sufficiently large that they will tend to engulf all but the largest potential barriers, such that they do not engender speciation, and the probability of division will decline. In addition, widespread species may have well-developed dispersal abilities and greater numbers of individuals that both help to maintain range contiguity and reduce speciation rates.
By contrast, the likelihood of extinction is strongly negatively correlated with geographic range size. Indeed, abundance and range size are in general the two best predictors of the probability that a species will go extinct in the near future (although there are examples of previously very widespread species that, usually as a consequence of anthropogenic pressures, have rapidly declined to extinction). Larger ranges typically comprise greater numbers of individuals and thus have a smaller probability of a random walk to extinction, and because of their greater areal coverage, they have a reduced risk that adverse conditions in one region will affect all individuals at the same time. This raises the possibility of species selection acting on range sizes.
Within taxonomic groups, interspecific variation in geographic range sizes is often correlated with such variation in other traits, including dispersal ability, breadth of resource use, or environmental tolerance, local abundance, and body size. In at least some cases, these relationships seem likely to be mechanistic, although the paths of causality may be variable. Thus, although it seems intuitive that a greater dispersal ability will tend to lead to a species becoming more widespread, other barriers may prevent this from occurring (see below), and if the structures associated with good dispersal abilities are costly to build or maintain, they may be reduced or lost. Indeed, there is evidence that when other limits are removed, species can show rapid acquisition of improved dispersal abilities.
Species that are able to exploit a wider variety of resources or persist under a wider range of environmental conditions should, all else being equal, be able to attain larger geographic ranges. Of course, all else may not be equal (e.g., extent of different resource types, dispersal abilities), which will tend to weaken any correlations between the level of such generalism and range size. The variety of resources and the breadth of environmental conditions that species can use are influenced both by the variety and breadth that can be exploited by individual organisms and by the differences in resource usage and tolerance of environmental conditions among individual organisms. The latter is probably a much more important influence on the relative geographic range sizes of species in many taxonomic groups, given evidence for marked spatial variation in the realized and fundamental niches of individuals, particularly of those species that are more widespread.
The geographic range sizes of species within taxonomic groups tend commonly to be positively correlated with their local density, such that widespread species not only have more individuals but disproportionately more so than restricted species. A number of plausible mechanisms have been proposed to explain such a pattern, including variation in niche breadth (the range of resources or conditions a species can exploit), niche position (how typical are the resources or conditions that a species can exploit), habitat selection (the tendency for species to use more habitats as they become more abundant), and metapopulation dynamics (in which dynamics in local populations depend on those in other such populations). It seems likely that a variety of potentially mutually reinforcing processes may be at work and that the pattern is an almost inevitable consequence of the aggregated spatial distributions of the individuals of most species.
Also within taxonomic groups, the body sizes and geographic range sizes of species tend to exhibit an approximately triangular form, such that although species of all body sizes may have large geographic range sizes (the upper limit normally being imposed by the size of the land mass or ocean mass), the minimum range size observed tends to increase with body size. A positive relationship likely occurs because, on average, larger-bodied species have larger-sized home ranges than smaller-bodied ones. They may thus also require larger total geographic range sizes if range-wide populations are to exceed some minimum viable size, which would tend to result in a positive interspecific range size–body size relationship, and indeed a triangular one because there is no necessary upper constraint on the range size of small-bodied species. One consequence of this mechanism is that the body sizes of the largest species tend to increase with the areas of the land masses on which they occur.
Regardless of their extents of occurrence or areas of occupancy, the geographic ranges of few species are entirely congruent. Rather, the bounds fall in different places and shift position on both ecological and evolutionary time scales. Why at any given time, or averaged over a particular period, the range edges of a particular species fall quite where they do has been much debated, and for surprisingly few species is it well understood. Part of the difficulty is that the question has a variety of answers, depending on the terms in which it is couched. First, one can determine whether there are abiotic and/or biotic factors that prevent further spread, and if so what these are. Second, one can consider how, in response to these factors, the population dynamics of a species change such that it is unable to persist beyond this point. Third, one can establish the genetic mechanisms that prevent a species from evolving capacities that would enable it to overcome any limiting abiotic and biotic factors that prevent it from expanding the limits to its geographic range and becoming more widespread.
Two principal groups of abiotic factors have been argued to limit geographic ranges, physical barriers and climate. In both cases, it is actually the interplay between these factors and the traits of the particular species that is of concern. Thus, given its dispersal abilities and behavioral tendencies, the spread of a species may be delayed or entirely prevented by expanses of inhospitable habitat, such as water for terrestrial organisms and land for marine ones, high elevations for lowland organisms and shallow water for deep-sea ones, or grasslands for forest species and forests for grassland species. In some cases, the role of behavior may be more significant than that of dispersal ability per se, with even quite small disjunctions in the distribution of suitable habitat greatly restricting spread. In many cases, rare long-distance dispersal events effectively define what does and does not constitute a barrier.
Climatic constraints on geographic ranges attract by far the majority of attention from ecologists, particularly because these may be modified by anthropogenic climate change, with implications for, among others, agriculture, forestry, fisheries, and human health. Climate doubtless limits the potential occurrence of all species because their physiological tolerances are constrained as a consequence of the costs of maintaining wide tolerances and the trade-offs associated with being adapted to particular conditions. However, demonstrating that climatic factors actually determine the position of the limits to the range of a species is more difficult. A variety of forms of evidence are strongly suggestive, including observations of systematic latitudinal variation in elevational and depth limits to species occurrences (suggesting elevational and depth responses to changes in climate with latitude), of spatial coincidence between the occurrence of range edges and particular climatic conditions, and of temporal covariation in the position of range edges and particular climatic conditions. However, such patterns could reflect covariation with some other factors, such as the distribution of a key resource, predator, or parasite, which itself is climatically limited. Rather more convincing are demonstrations of the more direct influence of climatic conditions at range edges on reproduction and mortality, perhaps most commonly reflected in the failure of species to be able to complete their life cycles beyond the range boundary.
A wide variety of biotic factors have been argued to limit the geographic ranges of species, including the absence of essential resources (e.g., nutrients, prey) and the presence of competitors, predators, or parasites. The role of resources in limiting ranges is perhaps most clearly demonstrated by specialist consumers, whose distributions must be contained within that of their host. Almost invariably, such species do not occur throughout the distribution of this resource, which in the absence of other factors presumably reflects spatial variations in the abundance and quality of the host. The roles of competitors, predators, and parasites in limiting ranges frequently involve interaction among three or more species, such that predators and parasites have alternative resources to exploit and competitors have their influence through predators and parasites (apparent competition).
Although it is easiest to consider them separately, in practice the limitation of the geographic ranges of individual species by abiotic and biotic factors may often be complex. Combinations of these factors may act synergistically, and different factors may be limiting on different parts of the range boundary and at different times, on both ecological and evolutionary time scales.
The edges of geographic ranges are formed at the point at which births and immigration in local populations are exceeded by deaths and emigration. If we assume that dispersal is sufficient for the establishment of a species in peripheral sites but does not otherwise influence numbers in those local populations, then the key issues are the factors that drive local extinction, which is commonly observed to be higher among populations at range edges. These factors can include demographic stochasticity (in the extreme, in small populations by chance during the same time interval all individuals may die, all may fail to breed, or sex ratios may become highly skewed), the mean environmental conditions (including resource availability, interspecific competition, predation), and the temporal variance in environmental conditions (even when conditions on average are suitable, high variance will increase the likelihood of local extinction).
Although such simple scenarios at range edges may occur, immigration may often play an important role in the dynamics of local populations, enabling them to persist even when the death rates of individuals exceed the birth rates. This leads to the notion of source–sink dynamics, in which the geographic range of a species can embrace unfavorable niche conditions, such that it occurs in environments at the range edges under which in isolation any local population would rapidly become extinct. This emphasizes the potential importance of thinking about range limits in terms of the interactions among multiple local populations. Indeed, under such a scenario, limits can be formed simply because the proportion of the landscape occupied by local populations becomes sufficiently low that the influence of dispersal on local populations becomes insufficient, and the ratio of population extinctions to colonizations too high.
A number of constraints have been suggested that prevent populations at range edges from evolving the capacity to spread further. Most attention is, however, focused on the possibility that if gene flow is sufficient, the occurrence at range edges of alleles that would otherwise enable range expansion may be swamped by alleles from other populations. This is particularly likely to be the case if gene flow occurs predominantly to edge populations from the typically larger numbers of local populations and individuals that do not occur at range edges, pushing the latter away from adaptation to local optima. However, a diverse array of other mechanisms have been suggested that include low levels of genetic variation in peripheral populations, that traits show low heritability as a consequence of directional selection in marginal environments, that traits show low heritability because of environmental variability in marginal environments, that changes in several independent characters are required for range expansion and so favored genotypes occur too rarely, that genetic trade-offs between fitness in favorable and stressful environments prevent the increase of genotypes adapted to stressful conditions, that genetic trade-offs among fitness traits in marginal conditions prevent traits from evolving, and that the accumulation of mutations that are deleterious under stressful conditions prevents adaptation.
Aside from their size and boundaries, geographic ranges are structured in complex ways. This structure is determined by the distribution of the individuals of a species within the range boundaries and by variation in the traits exhibited by those individuals.
Across its geographic range, a species is almost invariably rare in most of the places in which it occurs and relatively abundant in only a few. That is, intraspecific species-abundance distributions are strongly right-skewed. The notion has long prevailed that areas in which a species attains higher densities tend to lie toward the center of its geographic range, with the range edge being an area of lower density. This would seem likely to follow if conditions were most favorable in the range center and declined in all directions away from that core. However, whereas this might be a useful model in the abstract, the empirical evidence to support such a pattern of abundance is limited, and there are ample examples in which it does not occur, including cases in which high abundances are found close to range edges and of marked latitudinal trends in abundance across ranges. Overall, it seems that just as they take a wide diversity of shapes, the spatial abundance structures of ranges are also very varied. This is important from an applied perspective, as without local abundance data, it is difficult to target conservation or other activities at abundance hotspots.
Although local abundances may not tend consistently to decline toward range edges, there may be a greater likelihood that levels of occupancy do so. That is, although the densities of individuals within local populations, and therefore of intraspecific interactions, may not do so, the densities of local populations may change systematically. This would be consistent with a scenario in which range boundaries were more often determined by reductions in the availability of suitable habitat patches rather than in the quality of those patches where they do exist.
Just as there are interspecific abundance–range size relationships, there are intraspecific ones such that as the local density of a species increases through time, so does its occupancy. This has significant implications for understanding how the geographic ranges of species spread and decline and also for a variety of applied issues such as strategies for harvesting species. As geographic ranges change in size, they seem essentially to move back and forth along trajectories in abundance–range size space.
A number of systematic patterns of variation in the traits of individual organisms have been documented across the geographic ranges of species (not simply between the edges and the rest of ranges). These include spatial trends in morphology (principally body size), physiology, and life history. Indeed, some of these have been regarded as sufficiently general as to constitute ecogeographical rules. They include the neo-Bergmannian rule or James’s rule (an increase in the size of a species toward higher latitudes or lower temperatures), Foster’s or the island rule (smaller species become larger and larger species smaller on islands compared with mainland areas), Gloger’s rule (a tendency for endothermic animal populations in warm and humid areas to be more heavily pigmented than in cool dry areas), Jordan’s rule (fish species develop more vertebrae in cold environments than in warm ones), and one of Rensch’s rules (an increase in litter sizes of mammals and clutch sizes of birds in colder climates). In the main, such patterns in geographic range structure appear to be driven by broad spatial trends in environmental conditions or by spatial trends in the temporal variation (between seasons or years) in those conditions. They tend also to be observed predominantly among the more widespread species, whose geographic ranges extend over a greater range of environmental conditions. The trait structures of the ranges of the majority of species may thus be a good deal more complex.
The study of geographic ranges has been revolutionized by dramatic increases in available data on the occurrences of species and on the environments in which they occur (particularly from remote sensing), and in the technology available to handle those data. The broad-scale perspective that these have enabled has served to highlight the significance of geographic ranges as fundamental units in ecology, and much of that understanding of population and community ecology can usefully be cast in terms of the size, distribution, and structure of geographic ranges.
Brown, James H. 1995. Macroecology. Chicago: University of Chicago Press.
Gaston, Kevin J. 1994. Rarity. London: Chapman & Hall.
Gaston, Kevin J. 2003. The Structure and Dynamics of Geographic Ranges. Oxford: Oxford University Press.
Gaston, Kevin J., and Tim M. Blackburn. 2000. Pattern and Process in Macroecology. Oxford: Blackwell Science.
Lomolino, Mark V., Brett R. Riddle, and James H. Brown. 2006. Biogeography, 3rd ed. Sunderland, MA: Sinauer Associates.
Williamson, Mark. 1996. Biological Invasions. London: Chapman & Hall.