Module 0: History & Foundations of Biogeography

Alexander von Humboldt’s 1802 ascent of Chimborazo (6,263 m, Ecuador) together with his botanist companion Aimé Bonpland produced the famous Tableau physique des Andes et pays voisins, published in the Essai sur la géographie des plantes (1807). The plate is commonly credited as the first quantitative biogeographic diagram: a vertical cross-section of the Andes showing named plant taxa plotted against altitude, with parallel columns for temperature, pressure, humidity, snowline, and the azimuth of visible stars.

The central empirical claim was that plant communities stratify by altitude in a sequence that recapitulates the latitudinal zonation from the equator to the pole. A traveller climbing Chimborazo passes through tropical lowland forest, premontane and cloud forest, subalpine puna and páramo, alpine tundra, and finally the nival zone of permanent snow—in roughly 6 km of vertical ascent rather than 8,000 km of horizontal travel.

The Lapse Rate and Why Altitude Mimics Latitude

The key physical driver is the environmental lapse rate \(\Gamma \approx 6.5\,\text{°C}/\text{km}\). Combined with the adiabatic cooling of rising moist air and Clausius–Clapeyron temperature dependence of saturation vapour pressure, this produces a stacked climate:

Humboldt’s instrumentation (a Ramsden sextant, a Fortin barometer, Deluc hygrometers, and a Saussure cyanometer for sky colour) allowed him to tabulate numerical values alongside his species lists—this is the move that makes the Tableau biogeography rather than merely catalogue.

Biomes as Integrated Climate Envelopes

A biome can be defined, following Holdridge (1947) and foreshadowed by Humboldt, as a region of parameter space \((T, P, E)\) with \(T\) mean annual temperature, \(P\) annual precipitation, and \(E\) potential evapotranspiration. Holdridge’s “life zones” are contiguous polygons on a log-log biotemperature–precipitation grid, and altitudinal transects trace a trajectory through this grid:

Charles Lyell’s Principles of Geology (1830–1833) supplied the temporal and mechanistic frame without which 19th-century biogeography could not operate. Three Lyellian ideas are foundational: (i) uniformitarianism—present processes acting over deep time explain the geological record; (ii) slow change in the distribution of land and sea, so that present distributions reflect ancient geography; (iii) extinction as a normal background process.

Darwin carried all three volumes aboard HMS Beagle (1831–1836). The Beagle voyage supplied Darwin with three biogeographic observations that shaped the Origin of Species (1859):

• The Galápagos finches and mockingbirds: island endemics resembling, but distinct from, South American mainland forms.
• The South American fossil record of Toxodon, Macrauchenia, and giant sloths—extinct relatives of still-living forms, demonstrating regional continuity of descent.
• The pattern of representative species across contiguous habitats (rheas south of the Rio Negro replacing those to the north) and across ocean gaps.

Darwin framed biogeography in Chapter XII–XIII of the Origin in terms of two principles that remain canonical:

The Ternate letter from Alfred Russel Wallace, written in February 1858 on the island of Ternate (Indonesian Maluku), arrived at Darwin’s Down House in June 1858 and forced the joint Darwin–Wallace presentation at the Linnean Society on 1 July 1858. Wallace’s independent derivation of natural selection is one of the most celebrated priority coincidences in science, and it is biogeographic: Wallace was in the middle of an eight-year field programme across the Malay Archipelago (1854–1862) precisely because the archipelago’s distributional peculiarities demanded explanation.

Philip Lutley Sclater, in his 1858 paper “On the general geographical distribution of the members of the class Aves” (Journal of the Proceedings of the Linnean Society), proposed six primary ornithological regions. His criterion was the distribution of passerine genera: a region is a set of contiguous areas that shares a distinctive fauna at the generic level. Sclater’s six:

• Palearctic—Eurasia and North Africa north of the Sahara.
• Nearctic—North America north of the Mexican plateau.
• Neotropical—South and Central America plus the Caribbean.
• Aethiopian (now Afrotropical)—sub-Saharan Africa plus southern Arabia.
• Indian (now Indomalayan)—tropical South and Southeast Asia.
• Australian—continental Australia, New Guinea, and adjacent islands.

Wallace, in The Geographical Distribution of Animals (1876, 2 volumes), endorsed Sclater’s scheme and extended it to all vertebrate classes, demonstrating that the same six regions emerged whether one tallied mammals, reptiles, or amphibians rather than birds. This multi-taxon congruence is what converted Sclater’s regions from an ornithological convenience into a general biogeographic claim.

Wallace explicitly noted that the boundaries are not equally sharp. The Nearctic and Palearctic share enough Holarctic fauna that some 20th-century authors (Heilprin, Darlington) would merge them. In contrast, the line separating Indomalayan and Australian faunas cuts sharply through the 30-km strait between Bali and Lombok—Wallace’s Line of 1859.

Jaccard Similarity as Quantitative Backbone

A modern restatement of the Sclater–Wallace argument uses the Jaccard coefficient, introduced for plant communities by Paul Jaccard in 1902. For two regions with faunas \(A\) and \(B\):

In a letter to Sclater (published 1859), and more fully in The Malay Archipelago (1869), Wallace argued that the fauna of Bali is essentially Asian (macaques, barbets, woodpeckers, squirrels), whereas the fauna of Lombok, 30 km east across the deep Lombok Strait, is Australian (cockatoos, megapodes, marsupials). No vertebrate Rubicon is sharper in the world.

The physical basis, unknown to Wallace but confirmed by 20th-century bathymetry, is that the Lombok Strait tracks the edge of the Sunda Shelf, a submerged continental platform connecting mainland Asia to Bali, Java, Sumatra, and Borneo. At Last Glacial Maximum sea level (\(\sim\)120 m lower), the Sunda Shelf was subaerial and faunal exchange was trivial up to Bali; east of Wallace’s Line the strait remained 250 m deep and impassable to non-flying mammals.

Later workers sub-divided the region:

• Huxley’s Line (1868): extends Wallace’s Line north between Borneo and the Philippines.
• Weber’s Line (1902): the line of faunal balance, east of Sulawesi, where Australian taxa begin to dominate.
• Lydekker’s Line (1896): the western edge of the Sahul Shelf, past which the fauna is fully Australian.

The region between Wallace’s and Lydekker’s Lines is called Wallacea. It includes Sulawesi, the Lesser Sundas, the Moluccas, and Timor, and is characterised by a chaotic mix of Asian and Australian elements that reflects a patchwork of deep-sea filter barriers.

While Sclater and Wallace were mapping regions, a parallel quantitative tradition was taking shape on the continent. Paul Jaccard (1902, Swiss Alps) introduced the coefficient that now bears his name to compare plant communities across elevation strata; it remains the workhorse of \(\beta\)-diversity analysis.

Olof Arrhenius (1921) working in Stockholm archipelago plots observed that the number of vascular plant species \(S\) in a sample plot of area \(A\) obeys an approximate power law:

The species–area power law is the single most-tested quantitative prediction in biogeography. Empirical \(z\) values cluster around 0.1 for nested subsamples of a continuous continental biota, 0.2–0.3 for oceanic islands, and 0.3–0.5 for habitat fragments. The conceptual reason these differ is developed in Module 2: it is the core of MacArthur & Wilson (1967).

These early results reveal the hidden hinge of classical biogeography: regions and species–area both reduce to statistical statements about how taxa are packed into space. Once one grants descent with modification, the rest is combinatorics plus geology.

Two 19th-century botanists transported Humboldtian thinking into the English- and French-speaking worlds, respectively, and into the Darwinian synthesis:

• Augustin Pyramus de Candolle (1778–1841) laid out the concept of botanical regions in his Essai élémentaire de géographie botanique (1820). He distinguished “habitation” (the geographical range of a taxon) from “station” (its local ecological setting)—a distinction that survives today as “range” versus “niche”. De Candolle divided the world into 20 botanical regions by endemism.
• Joseph Dalton Hooker (1817–1911), director of Kew, travelled to Antarctica (Erebus and Terror, 1839–1843), the Himalayas, and New Zealand, producing the Flora Antarctica, Flora Novae-Zelandiae, and Flora Tasmaniae. His introductory essay to the Flora Tasmaniae (1859) developed the idea of an austral flora shared among South America, southern Africa, Australia, and New Zealand—what we now explain by Gondwanan vicariance. Hooker was Darwin’s closest botanical correspondent and the first major scientific convert to natural selection.

Linnaeus (1707–1778) framed biogeography essentially statically: species had been created on Paradisus, a single mountain with altitudinal climate zones, and had dispersed to their current ranges after the Flood. This “Mount Ararat hypothesis” was the orthodoxy through the early 19th century. The Linnaean era produced enormous catalogues of species distributions but had no mechanism for change in range.

The Darwinian turn after 1859 replaced creation with descent and replaced dispersal-from-a-centre with an evolutionary history of centres, barriers, and migrations. Two strong predictions follow:

• Regional endemism: if taxa evolve in situ, long-isolated regions should accumulate unique lineages. This predicts Madagascar’s lemurs, New Zealand’s tuatara and moa, and Australia’s marsupials.
• Fossil successions track regions: South American Miocene mammals should resemble South American Pliocene mammals more than they resemble Old World Miocenes. This is strongly confirmed (Simpson 1980).

Joseph Le Conte, George Gaylord Simpson (Evolution and Geography, 1953), and Philip Darlington (Zoogeography, 1957) carried this framework through the 20th century. Brown & Gibson’s Biogeography (1983) is widely regarded as the start of the modern synthesis: the first textbook to integrate the classical regional tradition with island biogeography theory, vicariance biogeography, and the new molecular data.

James H. Brown and Arthur C. Gibson’s textbook Biogeography (Mosby, 1983; 2nd ed. with McCain, 1998) is usually taken as the inflection point at which 19th-century regional biogeography and 20th-century island biogeography merged into a single discipline. The book organised the field around four interacting components, each of which we address in a later module of this course:

• Historical biogeography—plate tectonics, vicariance, dispersal history, fossil record (Module 3 and Module 7).
• Ecological biogeography—climate control of biome boundaries, the latitudinal diversity gradient, species–energy theory (Module 5).
• Island biogeography—MacArthur–Wilson equilibrium theory, species–area, the General Dynamic Model of oceanic islands (Module 2).
• Phylogeography—molecular genealogies projected onto geography, coalescent theory, phylogenetic \(\beta\)-diversity (Module 6).

A key contribution of Brown & Gibson was to insist that these four components are testably different. An observed pattern of Madagascan lemur endemism can be “explained” by each: vicariance (Madagascar rifted from Africa in the Jurassic), ecological filtering (wet forest plus a specific food resource), island biogeography (species–area effects on a 587,000 km\(^2\) island), or phylogeographic history (late-Eocene oceanic raft from East Africa). Distinguishing these requires phylogenies and dated fossils—the kinds of data that only became abundant in the 1990s and 2000s.

Holt et al. (2013, Science 339:74–78) marks the latest quantitative rebuild of Sclater’s 1858 regions: they used species-level ranges and phylogenetic \(\beta\)-diversity for 21,037 mammals, birds, and amphibians to recover 11 realms with 20 subrealms. The six Sclater regions remain broadly recognisable, but the Madagascan, Oceanic, and Panamanian realms emerge as first-order entities. This analysis is the centrepiece of Module 1.

Classical biogeography made three empirical claims that have survived in essentially their original form:

• Biotas cluster into regions of distinct generic composition (Sclater 1858, vindicated by Holt 2013).
• Altitudinal and latitudinal zonation parallel one another (Humboldt 1807, quantified by modern lapse-rate meteorology).
• Species number scales with area as a power law (Arrhenius 1921, confirmed across taxa and biomes, and theoretically grounded by MacArthur & Wilson 1967).

Three claims did not survive:

• Land bridges everywhere: Darwin, Hooker, and especially American zoogeographers of the late 19th and early 20th centuries proposed transoceanic land connections to explain disjunct ranges. Plate tectonics (Wegener 1912, confirmed 1960s) dissolved most of these bridges: continents have moved and disjunctions often date to Gondwanan vicariance, not sunken Atlantises. (Module 3.)
• Centres of origin are dispositive: the notion that every taxon has a single centre of origin from which it spreads, central to Matthew (1915), was undermined by the recognition that lineages can have multiple coeval origins via vicariance and that “centres” often reflect present-day richness gradients rather than ancestral geography.
• Fixity of regions: Sclater and Wallace treated the six regions as stable over the Cenozoic. Molecular phylogeography since the 1990s (Avise 2000) shows regional boundaries have shifted with climate cycles, sea-level change, and mountain uplift on timescales of 10\(^4\)–10\(^7\) yr.

The modular structure of this course tracks the resolution of these three failures in turn.