Community Structure and Biogeography

Island biogeography equilibrium model

The MacArthur–Wilson equilibrium model explains island species richness as a balance between two rates. Immigration from a source pool adds species, and extinction on the island removes them. Immigration falls as richness rises because fewer arriving species are new to the island, while extinction rises as richness rises because more species share the same limited resources. Equilibrium richness is reached where the two rates are equal. At equilibrium, species keep being replaced, a process called turnover, but the total number stays the same. The basic model treats all species as ecologically equivalent, so the extinction rate depends on the number of species present and on island area, not on which species they are. Biotic interactions such as interspecific competition or intraguild predation are not explicit parameters of the basic model. Area is the key variable for extinction. Large islands carry larger populations that are less likely to vanish through chance events, so the extinction rate falls as area increases. For the same reason, small islands hold fewer species. Their small populations go extinct more easily, and they also contain fewer habitat types and so fewer niches. Speciation is slow on ecological time scales and is not faster on small islands, so it cannot explain the pattern. Large islands hold more species because they offer more habitat types, present a larger target for colonists, and sustain larger populations. Distance from the mainland changes immigration to a given island, but it does not by itself explain why richness scales with area. Richness is therefore highest on islands that combine large area with short distance to the source, since they receive the most immigrants and suffer the least extinction. Islands also have higher extinction rates than the mainland. A given sampled area therefore holds fewer species on an island than on the mainland. The island species–area curve lies below the mainland curve at small areas, is steeper, and converges with the mainland curve at large areas.

Rank abundance diversity

A rank-abundance plot ranks species from most to least abundant and plots abundance against rank, so each community appears as a curve. The number of points on the curve gives species richness, and the slope of the curve reflects evenness, the degree to which individuals are spread across species. A flat curve means abundance is shared evenly among many species. A steep curve with a few dominant species at the top means most individuals belong to those species. Diversity depends on both components. A community with more species and a flatter curve is more diverse than one with fewer species and a steeper curve, even when both have the same total number of individuals. Equal totals do not imply equal diversity, because the distribution of individuals among species matters. Evenness can be read directly from the slope. A flatter curve indicates greater evenness, while a steep curve shows that a few species dominate. A community spread over many species with a flat curve is therefore more even than one in which fewer species hold most of the individuals. When only richness is known and evenness is unknown, comparisons cannot be settled. The Shannon index is maximal when all species are equally abundant and falls as the distribution becomes more uneven. Thus a community with fewer species but near-perfect evenness can have a higher index than a richer community that is strongly uneven, and no ranking follows from richness alone.

Latitudinal diversity gradient

The latitudinal diversity gradient is the widespread pattern in which species richness declines from the equator toward the poles. Several hypotheses explain it. A larger total area in the tropics provides more habitat and more populations, which raises the number of species that can be maintained. Higher speciation rates in warm, productive environments, often linked to ecological specialisation, add species over time. Lower extinction rates in climatically stable regions allow species to accumulate. The distance between Earth and the Sun is not part of any accepted explanation. Low seasonality is one hypothesis. When resources are available throughout the year, species can specialise on different resources, which reduces competition and permits coexistence. Stable environments also keep populations large and less exposed to demographic stochasticity, so extinction is less frequent. Generation time works in the opposite direction from what is sometimes assumed. Short generation times speed evolutionary change and diversification, whereas long generation times slow them. Reduced predation is not an established mechanism of the aseasonality hypothesis. The same latitudinal trend appears along mountain ranges on different continents, where tree richness increases toward lower latitudes. Greater incident solar radiation at low latitudes supports higher primary productivity and therefore more niches for trees. Temperature falls toward higher latitudes, so it cannot explain higher richness there. Distance from the coast and trade-wind strength do not vary consistently with latitude across these regions and cannot account for the pattern.

Species–area relationship

The species–area relationship describes how species richness increases with area. It is commonly written as S = cA^{z}, where c is a constant, A is area, and z is the exponent. On an arithmetic plot, the curve's shape depends on z. When z lies between 0 and 1, the curve passes through the origin, rises steeply at small areas, and flattens as area grows, so it is concave down. A straight line requires z = 1, and a convex curve requires z > 1. A single power function with a positive exponent cannot produce a hump-shaped curve with a maximum. On log–log axes, a power law becomes a straight line, \log S = \log c + z\log A, and the slope equals z. If the slope is not 1, the untransformed relation is a power law and not a linear one. Other forms look different on log plots. An exponential relationship is straight only on a semi-log plot, with \log S plotted against A. A Michaelis–Menten function is a saturating curve and remains curved on log–log axes.

Range overlap and species richness

Range overlap refers to how many species' geographic ranges share a given location, and overlap determines local richness. If ranges are placed at random on a closed surface with no boundaries, any point has the same chance of falling inside a given range. That probability equals the range's area divided by the total area of the surface. Expected richness per unit area is therefore uniform, and no latitude has an intrinsic excess of species. Range shape and size do not by themselves create concentrations of species in particular regions. Peaks in richness require boundaries that constrain where ranges can fall. Overlap also matters for interactions. Where two closely related species' ranges overlap, competition can drive divergence in traits such as bill shape. This is character displacement. The divergence is confined to the zone of co-occurrence, while populations outside that zone remain similar. Competitive exclusion removes one species, convergent evolution makes unrelated lineages look alike, and allopatric speciation requires geographic isolation. None of these explains trait divergence between species that co-occur at one site.

Continental drift evidence

Continental drift theory, proposed by Wegener, holds that today's continents were once joined in a supercontinent and later separated. Its strongest biological evidence is disjunct distribution. The seed fern Glossopteris and marsupial mammals occur on southern continents that are now separated by oceans, which is explained if those landmasses were once connected. Natural selection, metapopulation theory, and island biogeography do not account for such intercontinental disjunctions. Molecular data can also reveal dispersal. In a haplotype network, each circle is a haplotype and each dash is a mutational step. Shallow networks, in which haplotypes are separated by few steps and shared among regions, indicate gene flow and high dispersal, even across a barrier such as an ocean. Deeper networks, with divergence among regions connected by open water, indicate restricted gene flow and low dispersal. When generation times are similar, mutations accumulate at comparable rates, so differences in network depth reflect dispersal rather than generation time.

Community assembly processes

Local communities are assembled from a regional species pool, and the species that establish at a site depend on processes acting locally. Local interspecific competition, stochastic demographic variation, and local predator–prey interactions can each make communities at different sites differ in composition, because they remove or retain different species at each place. A regionally uniform climate imposes similar conditions across sites and tends to make communities resemble one another. It is therefore the least likely cause of differences among local communities. Environmental filtering is a process in which abiotic conditions such as heat or drought allow only species with matching tolerances to persist. Co-occurring species therefore tend to share tolerance traits. Competition-based processes, such as limiting similarity and competitive exclusion, predict the opposite pattern, in which coexisting competitors differ in traits.

Secondary succession after shifting cultivation

Secondary succession occurs on land where soil and seed sources remain after disturbance, as on plots abandoned after shifting cultivation. Early stages are dominated by light-demanding pioneer species with low shade tolerance and little woody cover. As the plot regenerates, woody vegetation accumulates and the canopy closes, so shade-tolerant species establish later under the developing cover. The community therefore moves from the pioneer condition toward a state of high woody cover and high shade tolerance. When interpreting a graph of succession, the key is that both woody cover and shade tolerance increase with time.