A biogeographical perspective on ecological systems: some personal reflections


Robert E. Ricklefs, Department of Biology, University of Missouri–St. Louis, One University Boulevard, St. Louis, MO 63121–4499, USA.
E-mail: ricklefs@umsl.edu


During my graduate studies, I characterized patterns of geographical distribution and taxonomic differentiation in birds of the West Indies, which suggested that species undergo phases of expansion and contraction similar to the taxon cycles that E. O. Wilson had described for Melanesian ants. Fieldwork in the early 1970s with George Cox confirmed that these phases were associated with variation in habitat distribution and abundance on individual islands, tying together local ecology and biogeography. Because taxon-cycle stage was independent of taxonomic or ecological relationships among birds of the West Indies, George and I postulated that whether a species was in a phase of expansion or contraction reflected the outcome of coevolved relationships with antagonists, including pathogens. The taxon cycle concept had a cool reception initially, but subsequent phylogeographical analyses, beginning in the early 1990s with Eldredge Bermingham, provided a time scale that confirmed the relationship between taxon cycle stage and the relative age of the most recent population expansion. The discrete nature of islands allows one to visualize taxon cycles in island systems, but the principle should apply in a continental biota as well. The absence of strong phylogenetic effects in distribution and abundance is consistent with evolutionary lability caused by coevolutionary outcomes with specialized antagonists. Related species appear to compete for resources on a more-or-less equal footing across a broad range of environments, and their distribution at any particular time is likely to be determined primarily by their relationships with pathogens, among other antagonists. This model of distribution and abundance within a regional community is consistent with much of what we know about the interactions between pathogens and their host populations, but testing the model will require the development of a new research programme focused on endemic pathogen effects in natural communities.

My professional career stretches back almost 50 years, over which biogeographers and ecologists have learned an amazing amount about the diversity and distribution of organisms over the Earth. The increase in available information is remarkable in its own right, but this period has also seen the introduction of new, geographically explicit concepts and analytical methods, including phylogeography, metapopulations, source–sink relationships, metacommunities, species distribution models, and phylogenetic community structure. Many of these developments supported the emergence of macroecology (Brown, 1995; Gaston & Blackburn, 2000). Yet, looking back over the decades, I am also impressed by the staying power of some of ecology’s persistent questions: What is an ecological community? How do we define a species’ niche? Why are there so many species in the tropics? What controls the relative abundance of species? The fact that intelligent people can seriously consider a theory of biodiversity that contains no ecology – I am referring to Stephen Hubbell’s ‘neutral theory’ of zero sum ecological drift (Hubbell, 2001) – should cause us to reconsider what we know.

My purpose in this essay is to describe some of the directions my research has taken, beginning with my graduate studies, and to reflect on the ways my earliest professional experiences have shaped my perspectives on fundamental issues in biogeography, in its broadest sense. Most recently, I have been concerned with the nature of the ecological community, arguing, in the tradition of Gleason (1926), Whittaker (1953), and others, that the population, rather than the local assemblage of species, is the fundamental unit of community ecology (Ricklefs, 2008). In this view, populations belong to a regional community within which the distributions of species determine the composition of local assemblages. This perspective shifts questions of local diversity and relative abundance of species to factors that determine diversity and distributions within larger regions (Ricklefs, 2011), and it emphasizes the unity of evolution, ecology and biogeography. This perspective also brings to mind the first research that I pursued in graduate school at the University of Pennsylvania, under the mentorship of Robert H. MacArthur.

The year I arrived in Philadelphia, MacArthur and E. O. Wilson published their influential paper ‘An equilibrium theory of insular zoogeography’ (MacArthur & Wilson, 1963), later expanded as the first Princeton Monograph in Population Biology, entitled The Equilibrium Theory of Island Biogeography (MacArthur & Wilson, 1967). To my way of thinking at the time, certain aspects of the theory conflicted with ideas about community ecology, as it was developing out of simple competition models (Gause, 1934; Hutchinson, 1957; MacArthur & Levins, 1967), microcosm and field experiments (Vandermeer, 1969; Schoener, 1983), and observations of niche partitioning among species (Lack, 1971; Pianka, 1973; Cody, 1974). The competitive exclusion principle (Hardin, 1960) embodied the idea that local species interactions determined the ability of any particular species to occur at a particular locality, as well as the number of species that could coexist. From that perspective, the persistence of populations in the face of competition depended on resource specialization and niche partitioning among species (MacArthur & Levins, 1967; Vandermeer, 1972). At the same time, other authors argued that predation and disturbance influenced competitive relationships and were important components of the local community mix (Connell, 1961, 1973, 1978; Janzen, 1970; Paine, 1974).

According to the equilibrium theory of island biogeography, however, the number of species that inhabited an island reflected the rate of colonization, which was external to the local system, and the rate of extinction, which depended on, among other things, the size of the island and thus population sizes of individual species. Consequently, the number of species present on an island, and also in local habitats within an island, was not fixed by local interactions and niche partitioning but rather by an extrinsic driver (colonization) and a regional property of populations (number of individuals). One could argue that remoteness and limited island area reduced species richness on islands below some ecological saturation point (MacArthur & Wilson, 1963; Terborgh & Faaborg, 1980), or that the saturation point is lower on such islands (Lack, 1976), but empirical analyses have provided little evidence for this (Srivastava, 1999; Ricklefs, 2000; Hillebrand & Blenckner, 2002).

MacArthur and Wilson’s formulation of the equilibrium theory encouraged ecologists and many biogeographers to emphasize place (i.e. the island) at the expense of the ‘biogeography of the species’ in Wilson’s words (Wilson, 1959, p. 122); thus, the local assemblage (the island biota) became the focus of research into processes influencing diversity and relative abundance, and the distribution of individual species assumed a secondary position in community ecology. This preoccupation with place also obscured the importance of connections between global and local processes. While recognizing historical effects in patterns of diversity (e.g. the difference in tree species between Europe and North America), MacArthur explicitly ignored history because he was interested in general patterns (e.g. MacArthur, 1972). In addition, local processes were believed to approach equilibrium states very rapidly in the time frame of historical and evolutionary processes.

This partitioning of much of ecology from history and geography was unfortunate because ecologists lost track of their roots in the early traditions of biogeography (Kingsland, 1985). Certainly one did not forget Darwinian principles of adaptation and diversification, including character displacement, as processes that promoted species coexistence (Lack, 1947; Brown & Wilson, 1956; Grant, 1972, 1986). However, little was said among ecologists of: Alfred Russel Wallace’s emphasis on dispersal barriers and the global differentiation of biota (Wallace, 1876); William D. Matthew’s thesis on the role of climate cycles in dispersal from areas of origin (Matthew, 1915); John C. Willis’s much derided ideas about species formation and species range dynamics (Willis, 1915, 1922); George Gaylord Simpson’s insights from fossil taxa concerning the waxing and waning of clades over geological time (Simpson, 1953); Philip J. Darlington’s conclusions about overwater dispersal based on field studies in the West Indies (Darlington, 1957); and ideas about cycles of population expansion and contraction of species (Brown, 1957; Wilson, 1961; Dillon, 1966).

Upon entering this intellectual atmosphere, my own inclinations were initially towards evolution and biogeography, rather than community ecology. As a child, I was fascinated with maps and would study them for hours, sometimes while reading a relevant travel adventure; I was an enthusiastic birder at the time; stamp collecting rounded out my interest in distant places and the diversity of life and human culture. It was only natural when I arrived in Philadelphia that I should be drawn to the biogeography of West Indian birds: MacArthur was keen on island biogeography; the world’s expert on West Indian birds, James Bond (the namesake of Ian Fleming’s 007), was at the Academy of Natural Sciences of Philadelphia and had published a field guide and a detailed distributional checklist of the birds of the West Indies (Bond, 1956); the islands themselves were relatively accessible to the eastern United States. I began by mapping all the species. Believing at the time that taxonomic distinction (that is, ‘evolutionary divergence’) provided a reasonable indication of the relative ages of island populations, I kept track of the subspecies names as well.

After a few days of filling basemaps with species distributions, certain patterns began to suggest a temporal sequence of (I) dispersal through the islands followed by (II) reduced gene flow and differentiation (subspeciation) of island populations, (III) extinction of some island populations creating gaps in distributions, and finally (IV) contraction to a single-island endemic (Fig. 1). The existence of widespread West Indian endemic species also suggested that this sequence could restart, perhaps repeatedly. E. O. Wilson (1994, pp. 214–215) describes a similar ‘epiphany’ after pouring over maps of Melanesian ant distributions a few years earlier, acknowledging the influence of authors who wrote on global dynamics of species distributions, Matthew and Darlington among them (see also Lomolino & Brown, 2009). I thought the logic of this ‘taxon cycle,’ to use E. O. Wilson’s term, was unassailable, but MacArthur discouraged me from pursuing this basically historical, non-equilibrium phenomenon. Consequently, I shifted for my thesis to working on avian growth and development as life-history traits (e.g. Ricklefs, 1968), combined with fieldwork near Tucson, Arizona, on the physiological ecology of desert birds with fellow graduate student Reed Hainsworth (e.g. Ricklefs & Hainsworth, 1968).

Figure 1.

 Distributions of five species of bird in the Lesser Antilles in different ‘stages’ of the taxon cycle. Taxonomic differentiation based on Bond (1956) is indicated by shades of grey. Populations of the house wren on Guadeloupe and Martinique are thought to have become extinct in the 20th century. Based on Ricklefs & Cox (1972). Gray kingbird: Tyrannus dominicensis; house wren: Troglodytes aedon; Adelaide’s warbler: Dendroica adelaidae [subsequently split into endemic island species]; Martinique oriole: Icterus bonana; Lesser Antillean bullfinch: Loxigilla noctis. The two most southern islands on the maps are Trinidad (continental land bridge) and Tobago.

In 1965, I had the good fortune to spend a month with another grad student, Henry Hespenheide, collecting plants throughout the island of Jamaica. I should say that while Henry pursued plant collecting, I entertained myself mostly with bird watching. A few years later, I summarized my field notes from this trip in a brief paper in Evolution (Ricklefs, 1970), in which I related the habitat distribution and abundance of birds on the island to their geographical distribution and taxonomic differentiation, i.e. taxon cycle stage, adapting the scheme used by Greenslade (1968) for Solomon Island birds. From these observations, I concluded that

… the general pattern of the taxon cycle of birds inhabiting Jamaica suggests that recent immigrants to the island are abundant and widespread with respect to habitat, exhibiting ‘ecological release’. Among older populations, there is a trend towards more restricted distributions, often to montane habitats, and possibly lower population densities.

(Ricklefs, 1970, p. 477)

Thus, local ecological distributions of populations and their broader biogeographical distributions appeared to be connected as species moved through stages of the taxon cycle.

MacArthur left the University of Pennsylvania in 1965 for a professorship at Princeton University, and I finished my doctorate on avian growth under the mentorship of W. John Smith, a behavioural biologist. After a wonderful post-doctoral year immersed in tropical diversity at the Smithsonian Tropical Research Institute in Panama, I returned to the University of Pennsylvania as an assistant professor of biology. During a field season in Arizona in 1969, I met George W. Cox, who was then a professor of biology at San Diego State University. We discovered a mutual interest in biogeography, and a year later George came to my lab on sabbatical leave, during which time he conducted detailed censuses of birds across nine matched habitats on the islands of St. Kitts, St. Lucia and Jamaica, as well as in Trinidad and Panama, representing continental areas.

This collaboration resulted in a series of publications on West Indian birds. One of these (Ricklefs & Cox, 1972) developed the idea of the taxon cycle more thoroughly and introduced the hypothesis that the expansion and contraction of population distributions, both geographically and ecologically, reflected coevolutionary outcomes of the relationships of species with antagonist populations (e.g. predators and pathogens), which we called ‘counteradaptation’, combined with competitive pressure from new colonists to islands, which were clearly in expanding phases of the taxon cycle. A second paper (Ricklefs & Cox, 1978) confirmed in detail the relationship between local ecology and geographical distribution through the taxon cycle, and a third dealt with density compensation among bird populations in the West Indies (Cox & Ricklefs, 1977).

George and I were influenced in our thinking by a number of key papers that had been published during the previous decade or so. Of course, Wilson’s taxon cycle papers (Wilson, 1959, 1961) topped the list. William Brown’s paper on centrifugal speciation (Brown, 1957) introduced the idea of species formation in isolated peripheral populations left behind by a contracting range after an episode of expansion. Brown envisioned multiple phases of expansion and contraction, a continental version of the taxon cycle. These cycles could be driven by cyclical changes in climate, but George and I noted that because related species – presumably having similar ecological requirements – might be in stages II or IV at the same time (Ricklefs & Cox, 1978), extrinsic factors were unlikely to be responsible.

The solution to the individualistic nature of the taxon cycle occurred to us, in part, through the work of David Pimentel, at Cornell University, on coevolutionary relationships between flies and their parasitoids (Pimentel, 1961, 1968). Pimentel had shown in laboratory experiments that a host population could rapidly evolve resistance to parasitoids and achieve a new population equilibrium through a mechanism he called ‘genetic feedback’. We simply adapted this idea by suggesting that while colonizing species, almost by definition highly productive populations, were adapting to their new environments, multiple island residents, including antagonists, were also adapting to them. The result of this ‘counteradaptation’ was a reduction in population productivity leading to habitat contraction and reduced abundance, followed in time by extinction, failing a restart of the cycle. A new phase of expansion might begin when a population became so uncommon that it no longer formed a large part of the adaptive landscape of antagonist species, and selection for antagonistic adaptations towards these species declined.

George and I were not the only ones to have formulated scenarios linking adaptation to the expansion and contraction of populations. I have mentioned Matthew, Darlington and Wilson, among others. However, one author particularly stands out in my mind, both for his sense of natural history and for his ingenuity – as well as the strongly negative reception his ideas received in the scientific community. This was John C. Willis, author of Age and Area (Willis, 1922). He outlined his main thesis in an earlier paper on population distributions in the flora of Ceylon [Sri Lanka] (Willis, 1915). Willis noticed that, compared to species with wide geographical distributions, species endemic to Ceylon tended to be uncommon and narrowly distributed – often confined to high elevations on mountains (Fig. 2). To explain these observations, Willis reasoned that: species arose locally; endemic species in Ceylon, especially those with narrow distributions, were young species; and because these young species were restricted ecologically and with small populations, they were poorly adapted to their local environments: ‘endemic species have not in any way been developed to suit local conditions’ (Willis, 1915, p. 324).

Figure 2.

Willis’s (1915) classification of species endemic to Ceylon (809 species), species also distributed on peninsular India (492 species), and widespread species (1508 species), as very common (VC), common (C), relatively common (RC), relatively rare (RR), rare (R) and very rare (VR). The distributions are significantly heterogeneous (= 371, d.f. = 10, < 10−6). Based on Willis (1915, table III, p. 310).

Willis further emphasized what he regarded as the poor connection between adaptation and environment by pointing out that many narrowly endemic species thrive outside their restricted ranges when distributed by man:

We have been in the habit of looking upon endemic species with very small range as if they were suited to very special local conditions, and yet some of these, when planted elsewhere, grow with remarkable success. A very notable example is that of Cupressus macrocarpa [Monterey cypress], confined in nature to a small area in California, and yet proving to be one of the most successful conifers in all the sub-tropical regions of the world.

(Willis, 1915, pp. 324–325)

Willis supposed that species arise by random ‘macromutations’ that are independent of the characteristics of the environment in which they occurred, an idea that had several proponents at the time:

It is probable that species first appear as one or a few individuals by a sudden mutation, in a very restricted area. … Species do not, so far as we can tell, appear in any sort of advantageous response to local conditions, which are the only conditions that matter when they first appear.

(Willis, 1915, p. 341)

Over time, however, these new populations adapted to local conditions and began to grow and spread. Willis therefore related larger area to greater age. He also believed that species and genera tended to persist once formed, and that the most widespread, species-rich genera were the oldest and the source of all younger taxa. Willis’s ideas were genetically and evolutionarily naïve, and they were amply criticized at the time (Pearson, 1923; Gleason, 1924; Schonland, 1924). Willis made the fatal mistake of taking on Darwin in suggesting that newly formed species were not well adapted to their environments, and his insights relating geographical and ecological distribution, as well as the idea that geographical ranges were evolutionarily dynamic, were lost. Willis did, however, go on to become a pre-eminent botanist.

George Cox’s and my ideas about taxon cycles were not all that well received, either. For example, E. C. Pielou, in her book on biogeography (Pielou, 1979), suggested that variation in distribution could be explained by accidents of dispersal related to storms; the West Indies are ravaged by violent hurricanes, after all, and there was no need to invoke evolutionary mechanisms over long periods. Pregill & Olson (1981) similarly related species distributions to external environmental drivers – Pleistocene climate cycles, in their case – and dismissed the taxon-cycle concept, believing along with MacArthur & Wilson (1967) and David Lack (1976), among others, that adaptation to the local physical environment and to local habitats should only increase with time. To quote Pregill & Olson (p. 91):

Ecological doctrine and good sense revolt at the idea that a species with a long history of adaptation to a particular environment would be at a competitive disadvantage with newly arriving colonists … The concept of ‘counteradaptation’ is an artificial construct needed to explain a nonexistent phenomenon—the taxon cycle. The patterns of distribution that constitute the ‘stages’ of the taxon cycle are more reasonably interpreted in terms of the effects that the cycling of habitats had on species during the alternate wet and dry periods of the Pleistocene, in combination with the varying dispersal abilities of the individual species.

The key, both for Willis’s idea of age and area and for the taxon cycle concept, was the age of populations. Which populations were old, i.e. phylogeographically or phylogenetically distant from others, and which were young, was crucial. Taxonomic differentiation seemed a reasonable proxy for age, but without direct confirmation, it was clear that the taxon cycle would be a hard sell. Faced with this reality, and with no way to get around it at the time, I put aside island biogeography and spent several years working on the comparative biology of seabirds and other life history projects (e.g. Ricklefs, 1983, 1991, 1993, 1996). Meanwhile, throughout the 1980s and into the 1990s, community ecology became more experimental in response to disaffection with what some perceived as poorly supported claims of competitively based niche partitioning and ambiguities about the interpretation of patterns (Connor & Simberloff, 1979; Simberloff, 1980; Simberloff & Boecklen, 1981). As before, history and geography were not part of the diversity equation; local communities and regional species pools were governed by different processes. The open community perspective of Gleason (1926) and Whittaker (1953, 1967), linking local assemblages to regional distributions, was not considered relevant in many circles.

Several developments during the 1970s and 1980s helped to change the focus of biogeography and led to the emergence of macroecology. The relative abundance of species had been a central topic in ecology for decades, but interest in relating local abundance to distribution – Willis’s central theme – began to increase, particularly with a paper by Sam McNaughton and Larry Wolf (McNaughton & Wolf, 1970) entitled ‘Dominance and the niche in ecological systems’. This theme would be repeated through the development of macroecology (Bock & Ricklefs, 1983; Brown, 1984, 1995, 1999; Gaston & Blackburn, 2000; Gaston, 2003). Except for an early paper documenting the absence of a relationship between local abundance and ecological distribution in eastern North American birds (Ricklefs, 1972), I was not closely involved in these developments. I shall return to this theme later.

Then, in 1987, Ruth Patrick, a distinguished limnologist at the Academy of Natural Sciences in Philadelphia, asked me to write an article for Science on community ecology, perhaps in response to what she regarded as the sad state of the discipline at the time. The resulting paper (Ricklefs, 1987) may have surprised Ruth, who was a devotee of G. E. Hutchinson and Robert MacArthur, because of its emphasis on the connection between local diversity and regional processes. Other prominent ecologists to whom I sent a draft of the manuscript were not encouraging, even to the point of suggesting that I withdraw it prior to publication. With a few exceptions, including such individuals as Jim Karr (1976), David Pearson (1977), John Wiens (1977, 1984), John Lawton (1982) and Howard Cornell (1985a,b), community ecologists still championed the idea that local diversity reflected the ability of populations to coexist locally. Competition and other species interactions limited membership in communities; accordingly, differences in diversity reflected the different outcomes of interactions for coexistence under different physical conditions. In this regard, I continued to be impressed by differences in diversity in seemingly similar physical environments, the opposite of ‘community convergence’ (e.g. Cody, 1974; Cody & Mooney, 1978) and what Roger Latham and I referred to as ‘diversity anomalies’ (Latham & Ricklefs, 1993a,b; Ricklefs & Latham, 1993), and by the relationship between local and regional species richness, evident in George Cox’s research in the Caribbean (Cox & Ricklefs, 1977) and elsewhere (see Harrison & Cornell, 2008). Local diversity was not simply a matter of local determination.

At about the same time, I met Dolph Schluter (University of British Columbia) at an Ecological Society of America meeting, and, having similar ideas about the origin and maintenance of communities, we decided to organize a symposium that would integrate local, regional and historical perspectives. The symposium grew into an edited volume (Ricklefs & Schluter, 1993) that emphasized the importance of placing local communities in their geographical and historical contexts. Although many natural scientists, including most of the contributors to that volume, accepted this basic premise, Dolph and I like to believe that the book ‘legitimized’ regional/historical thinking in community ecology and helped to encourage additional work in this area.

A second chance encounter, with Eldredge (Biff) Bermingham at the American Association for the Advancement of Science (AAAS) meeting in 1989, promised to solve the problem of the relative ages of West Indian bird populations. Biff had been a doctoral student with John Avise at the University of Georgia and was just moving to a position as staff scientist with the Smithsonian Tropical Research Institute in Panama. Although he had worked with fish, he was persuaded that molecular phylogeographical studies of West Indian birds might provide interesting insights into historical biogeography on a regional basis, which was also his passion. We launched our first collecting expedition, sponsored by the National Geographic Society, to the Lesser Antilles in 1991, followed by 16 other field expeditions with many co-workers over the following 20 years.

Beginning with analysis of restriction fragment length polymorphisms (RFLPs) and, later, sequencing of mitochrondrial DNA, we found phylogeographical patterns consistent with the taxon cycle hypothesis. In particular, species ordered in the sequential stages I (recent colonists) to IV (single-island endemics) (see Fig. 1) exhibited increasing age (genetic divergence) of individual island populations; moreover, the ecological extent and abundance of island populations was inversely related to age (Seutin et al., 1994; Ricklefs & Bermingham, 1999, 2001, 2002). George Cox and I had shown in our 1972 paper (Ricklefs & Cox, 1972) that the slopes of species–area relationships of West Indian birds increased with taxon cycle stage, suggesting that extinction of island populations was more prevalent over time on smaller islands; the phylogeographical data confirmed this pattern (Ricklefs & Bermingham, 2004b). Being able to age populations in the West Indies as independent evolutionary units also made it possible to evaluate parameters of MacArthur and Wilson’s equilibrium theory (Ricklefs & Bermingham, 2001, 2008) and to estimate extinction rates of island populations (Ricklefs & Bermingham, 2008). Loose ends remain to be tied down (Johnson et al., 2000; Cherry et al., 2002; Ricklefs & Bermingham, 2004a; Dexter, 2010) but, finally, history, geography and ecology were linked, and the taxon cycle concept was upheld. All we needed was a mechanism to drive the cycle.

Taxon cycles need not be limited to archipelagoes. George Cox and I felt that similar phases of expansion and contraction characterized mainland biotas, as well, recalling the ideas of Matthew, Willis and others; the discrete nature of islands simply made the historical relationships between island populations more evident and allowed one to perceive taxon cycle stages. Macroecology has made researchers keenly aware of the tremendous variation in species abundance and distributional extent within continental regions, which can be directly linked to the taxon cycle concept. An observation that has impressed me in this regard is that range and abundance appear to be highly labile: most of the variance in these population characteristics can be attributed to differences between species within the same genus (Gaston, 1998; Webb & Gaston, 2000; Scheuerlein & Ricklefs, 2004; Ricklefs, 2010a), as we also noted in West Indian birds (Ricklefs & Cox, 1972) – and as, incidentally, was commented upon by Willis (1915, table IX) with regard to the plants of Ceylon.

From the perspective of community ecology, variance in abundance and distribution might be related to variation in the niche space of each species. Such phenomena as density compensation in response to variation in species richness (Cox & Ricklefs, 1977; Wright, 1980) are consistent with resource limitation of communities. It also occurred to me that clades with higher species richness might exert more pressure on available resources, leading to reduced local abundance or habitat distribution in such groups. However, my own analyses have consistently failed to show an inverse relationship between distribution/abundance and clade (family-level taxon) size: trees within the Barro Colorado Island 50-ha plot (Ricklefs, 2010a); birds within South America (Ricklefs, 2009); birds in eastern North America (Ricklefs, 2011); birds on a census plot in Amazonian Peru (Fig. 3). Willis (1915, pp. 312–313) had noticed the same pattern among plants in Ceylon:

Figure 3.

 Average densities (± SD) of bird populations in a census plot in Manu, Peru, as a function of the number of co-occurring species in taxonomic families. After data and taxonomy in Terborgh et al. (1990).

The variation in rarity between the different families or groups of families of Ceylon endemics is small, and goes to show that no one family has any particular advantage over another, and that there is no case that can be pointed to as showing, that any of the larger families is being driven to the wall in the struggle for existence.

It is sad that so many babies were thrown out with Willis’s bathwater!

High variance among closely related species means that differences in range and abundance cannot be related to evolutionarily conserved traits, which vary primarily at higher levels in the taxonomic hierarchy and evolve slowly. Such traits include adaptations for feeding and manoeuvring among habitat substrates (Miles & Ricklefs, 1984; Ricklefs & Miles, 1994), perhaps even basic physiological relationships to the environment (Ricklefs, 2009). Nor is variation in distribution easily related to external drivers such as climate, which would affect many related species in the same way [contra Pielou (1979) and Pregill & Olson (1981), cited above]. Likely influences on populations that could vary among closely related species include specialized pathogens whose influence on population productivity is sufficient to alter competitive relationships broadly within geographical regions.

Enough examples of introduced pathogens impacting both human and wildlife populations exist to convince one of the possibility that pathogens influence distributions (see, e.g. Ricklefs, 2010b). Less is known, however, about the role of pathogens in native biota. Suspecting that host–pathogen relationships might drive taxon cycles in West Indian birds, Bermingham and I made blood smears and preserved blood in lysis buffer from the beginning of our Caribbean fieldwork in 1991. Our objective was primarily to survey the prevalence of haemosporidian (malaria) blood parasites in West Indian bird populations. Although molecular markers were not yet available at that time, analyses of blood smears showed that parasite prevalence exhibited a significant host effect and a significant host-species × island statistical interaction, suggesting both host specialization and independent coevolutionary outcomes in isolated island populations (Apanius et al., 2000). This was later confirmed by genetic screening and identification of individual haemosporidian parasite lineages (Fallon et al., 2003). In general, host specialization is well documented among parasites and pathogens (e.g. Poulin, 2007).

I feel somewhat uncomfortable postulating unseen forces to explain fundamental patterns in ecological systems (even though I recall that A. R. Wallace had a keen interest in spiritualism and supernatural phenomena!); biogeographical ecology is replete with uncertainty as it is. Yet, perhaps pathogens will help us to resolve some of this uncertainty. The basic thesis, initially inspired by observations on West Indian birds, is this: (1) species are more or less equal competitors over broad regions of ecological space; (2) small variations in population productivity (i.e. births minus deaths) can lead to large changes in the size and distribution of a population; and (3) to a large extent, especially among closely related species, variations in distribution reflect species-specific coevolutionary outcomes of relationships with specialized pathogens.

Since the 1950s, ecology has been dominated by the paradigm of diversity being limited by species packing and the filling of available niche space, although environmental filtering of species has been recognized increasingly (Shmida & Wilson, 1985; Zobel, 1992; Weiher & Keddy, 1999; Webb et al., 2002; Silvertown et al., 2006; Cavender-Bares et al., 2009). The evidence for species packing playing a constraining role in diversity is unclear to me. Certainly competition is a potent force, and resources limit the overall productivity of populations and local assemblages. Species are also specialized, both with respect to the kinds of resources exploited by individuals and the environmental distributions of populations. But do these observations necessarily imply that competition and niche packing limit local diversity?

Species in expanding phases of the taxon cycle appear to colonize islands with little hindrance, suggesting that they can enter the available niche space. Many invasive species behave in the same manner (Sax & Brown, 2000; Sax, 2002; Sax et al., 2005; Qian & Ricklefs, 2006). Interactions among species within limited niche space presumably generate the density compensation observed on islands, but I see no reason that more species could not be squeezed into the niche space in mainland areas given suitable pressure of species production. Diversity anomalies observed between regions (Latham & Ricklefs, 1993a; Ricklefs & Latham, 1993; Qian & Ricklefs, 2000) and the frequently strong correlation between local and regional diversity (Cornell, 1985a; Ricklefs, 1987; Cornell & Lawton, 1992; Hugueny et al., 1997) suggest that local niche packing is flexible even within continental regions, and that diversity measured on any scale represents the tension between species production within the region and local species sorting through environmental selection and competition (Weiher & Keddy, 1999; McGill et al., 2006).

Within a region, similar environments tend to support a similar spectrum of species. Geographically distant localities might harbour different species in the same genus, suggesting that each of those species is suitably adapted to conditions within the broader region. Pairs of closely related species, one widespread and the other restricted, are relatively common and produce the high species-within-genus component of variance in distribution and abundance (Ricklefs, 2010b, 2011). Given this apparent ambivalence in the composition of a particular assemblage, individual species across a broad taxonomic spectrum might have roughly equivalent potential for maintaining populations over large portions of regional niche space, and many species would be roughly equivalent competitors over this space (Holt, 2006). In Lotka–Volterra competition theory, competitive equivalence implies that two species have similar carrying capacities and that interspecific competition coefficients are close to one. Under approximate competitive equivalence, whether one species or the other dominates would depend on small differences in either carrying capacity or the competition coefficient. Possibly, decisive differences might be caused by pathogens. Within the range of variation in population parameters caused by pathogens, these species would appear as though they are competitively equivalent, that is, neutral, although the outcome of interactions would be predominately deterministic.

Biogeographical analysis of West Indian birds revealed the strong influence of large-scale factors on local species richness, and emphasized the dynamic, idiosyncratic nature of species distributions. Willis (1915, p. 327) had similar insights almost 100 years ago, when he wrote:

… on the average the commonness of species depends on their age, and is independent of their local adaptation. The latter will in each individual case have a determining effect in the degree of commonness ultimately reached, but we have no ground whatever for making statements to the effect that a given species is common and is therefore well adapted. Probably mere chance has also a great deal to do with the commonness of a given species in any particular place.

Although direct evidence is lacking, pathogen interactions (not age per se) are plausible drivers of phases of population expansion and contraction. Extended to regional–continental settings, these observations imply that individual species are broadly competent ecologically, both within and between habitats, and that the distribution and abundance of species potentially reflect interactions with specialized pathogens.

The intersection of biogeography and community ecology, exemplified here by the distributions and local ecology of West Indian birds, provides a context for some conclusions about distribution and abundance more broadly. These conclusions will also serve as a set of my predictions as of 2011 about future directions in biogeography and macroecology. More than 40 years of living with the taxon cycle have given me courage to stick my neck out on views that many will regard as controversial or perhaps even foolish. Time will tell.

  • 1Related species are approximately equivalent ecologically. When it comes to exploiting environmental resources, one tree is practically as good as another tree, one sparrow is as good as another sparrow (Hubbell, 2001). I do not mean to disavow either that species are specialized or that they compete for resources, but only to suggest that geographical and habitat distributions of many species underestimate their ecological and geographical potential. One might say that the realized niche is poorly related to the fundamental niche.
  • 2Community dynamics appear quasi-neutral, but are not neutral. If species are closely matched ecologically and distributions do not strongly reflect local adaptation to common features of the environment, but rather individualistic factors, then distributions of similar species will appear to be idiosyncratic and outwardly random, rather than the result of interactions among specialized taxa.
  • 3Distribution and abundance are determined primarily by pathogens. If related species are ecologically indifferent with respect to each other, variations in distribution and abundance over large, relatively homogeneous regions are plausibly related to specialized pathogens, which might decouple evolved host-habitat/environment relationships.
  • 4Evolutionary dynamics create phases of expansion and contraction. Host species and their pathogens are engaged in coevolutionary dynamics involving virulence and resistance that can tip the balance in the demography of the host between expanding and contracting phases. These phases might depend on random mutations of the host or pathogen, which could account for the long intervals between phases of host expansion.
  • 5Thus, distribution and abundance of host populations are deterministic, but highly labile. Host–pathogen evolutionary dynamics are rapid compared with the evolution of adaptations for resource utilization and habitat utilization, and so the variance in distribution and abundance among populations resides primarily in differences among closely related species.
  • 6Local assemblages are readily invaded. If species are relatively indifferent ecologically to the presence of others of the same trophic level, then exotic species should invade local assemblages with little hindrance (Lawton, 1982; Sax & Brown, 2000; Sax et al., 2005), provided that they can tolerate local pathogens (Torchin et al., 2003; Ricklefs & Bermingham, 2007).
  • 7Ecological communities are regional and are not saturated. Clearly, presence and abundance locally are determined by the distributions of species within larger regions (Ricklefs, 2008). Thus, any concept of community must include interactions between populations within the entire region.
  • 8Host populations are self-limited by pathogens much of the time. Pathogen impacts on host populations are density-dependent and their linked population dynamics tend to prevent both ecological dominance and extinction of host species (Dobson & Crawley, 1994), although most species are relatively rare most of the time (Whittaker, 1965; Hubbell, 2001; Pitman et al., 2001).
  • 9Pathogens constitute many of the axes of multidimensional ecological space. According to competition theory, n species can readily coexist in a space with n−1 dimensions (MacArthur & Levins, 1967; Tilman, 1982; Chase & Leibold, 2003; Clark, 2010), following upon Hutchinson’s (1957) concept of the multidimensional niche. To the extent that pathogens are specialized, and many are (Poulin, 2007), they would constitute independent axes of niche space and permit the coexistence of practically as many species of host as there are specialized pathogens. The general requirement of strong self-limitation for the coexistence of competing populations is satisfied by specialized pathogens.
  • 10Pathogens diversify with their hosts, and so diversity is not self-limiting. One of the remarkable correlates of the control of host populations by specialized pathogens is that the diversity of these pathogens increases in direct proportion to the diversity of hosts (Poulin & Morand, 2000; Poulin, 2007). Accordingly, host diversity potentially is unlimited, subject to the ability of rare host populations to avoid extinction, and both regional and local diversity are determined primarily by large-scale processes responsible for species production (Ricklefs, 2007).

These conclusions, and predictions about potentially fruitful directions for future research, emphasize the need for integrating biogeography, macroecology, population biology, host–pathogen interactions and epidemiology. Although the pathogen effects discussed here are plausible in the context of what we know about host–pathogen relationships and emerging infectious diseases (e.g. Ricklefs & Bermingham, 2007; Ricklefs, 2010b), the impacts of pathogens on the coexistence, distribution and abundance of competing species in regional communities are poorly known. Pathogens are difficult to study in nature, although new experimental approaches are beginning to access natural host–pathogen interactions (e.g. Klironomos, 2003). Clearly, the pathogen perspective deserves consideration. Species in early stages of the taxon cycle might share with widespread and abundant species in continental regions the fortuitous and inescapably transient capacity to resist those pathogens that might otherwise control their populations. Thus, I believe that the taxon cycle can provide a general model for distribution and abundance in natural systems, emphasizing the population as the primary unit in community ecology and the region as the setting for community interactions.

Of course, the idea that pathogens might be important is not new, and it is fitting to give J. C. Willis (1915, p. 326) the last word:

We are using the word adaptation here to include all the relationships of the species to its environment, e.g., its liability to any diseases that may exist there. This last is a very important point. Some experiments which I have been carrying on upon the struggle for existence among closely crowded plants go to show that it is disease which kills the bulk of the losers, and that these are not necessarily the weakest but often well-grown plants.


I am grateful to the International Biogeography Society for honouring me with the Alfred Russel Wallace Award. My work on biogeography and community ecology has been generously supported by the National Geographic Society, Smithsonian Institution, National Science Foundation, and the Curators of the University of Missouri. The Alexander von Humboldt Foundation supported me during a recent sabbatical leave. For insightful and constructive comments on the manuscript, I thank Mark Lomolino, Michael Patten and Susanne Renner. None of this work would have been possible without the close colleagueship and friendship of George Cox and Eldredge Bermingham, as well as many of our associates and students.


Robert E. Ricklefs is Curators’ Professor of Biology at the University of Missouri–St. Louis. His interests include the historical biogeography of West Indian birds as a model system for regional faunal development and evolution. He has also worked on the generation and maintenance of regional patterns of species richness, and the connection between regional/historical processes and local assemblages of species. His current research includes the application of molecular assays to survey the host and geographical distribution of avian haemosporidian (malaria) parasites.

Editor: Robert Whittaker

This invited paper was occasioned by Robert E. Ricklefs’ recent receipt of the Alfred Russel Wallace Award, from the International Biogeography Society, for career contributions to biogeography.