The second half of my lecture on continents has been posted on YouTube and can be found here. It begins with the geological conception of the terms “continent” and “continental,” and then pivots to show why the continental scheme is of little if any use for understating either biogeography or human geography. The next section looks at the alternative world-regional system of global division. The lecture concludes with a brief reexamination of the underlying problem of geographical ignorance.
Canada’s western separatists often refer disparagingly to “the Laurentian elite” and “Laurentian Canada,” terms that are not familiar to most readers in the United States. These terms derive from the region around the Saint Lawrence River, Canada’s political, economic, and demographic core, but they refer more specifically to the Canadian establishment, or governing class. As the Wikipedia article on the “the Laurentian elite” notes, this group has a “bias toward federal power,” which is one of the main reasons why is deeply distrusted by western regionalists. It is closely associated with the center-left Liberal Party.
I could not find a map of “Laurentian Canada,” as its geographical bounds are vague and unspecified. It could conceivably cover the entire “Quebec City-Windsor Corridor,” a well-mapped region that contains roughly half the country’s population (see the map below). ChatGPT, however, tells me that “Laurentian Canada” refers “particularly to southern Quebec and eastern Ontario,” a region focused on Ottawa, Toronto, and Montreal. The chatbot Grok also specifies the inclusion of Quebec City. I am dubious of this claim, however, as Quebec City strikes me as too provincial and too francophone to merit inclusion. It is also uncertain whether this informal region encompasses southwestern Ontario; cities such as Windsor strike me as too peripheral to be included. I have therefore impressionistically mapped a limited “Laurentian Canada” within the larger “Quebec City-Windsor Corridor” (the second map posted below).
Quebec City–Windsor Corridor map
Laurentian Canada map
The Saint Lawrence River, and larger Saint Lawrence corridor, is tremendously important in the history of Canada and for the country’s self-image. In the U.S., however, the river and its basin are generally overlooked, even though a large portion of the watershed is in the United States. For more than a hundred miles in northern New York, moreover, the river forms the border between the two countries. But consider its portrayal in a visually appealing but conceptually muddled National Geographic map of the “Watersheds of the United States,” the subtitle of which asks, “Where do the U.S. watersheds drain?” A casual interpretation of this map might lead to the conclusion that the rivers of northern New York drain into the “Great Lakes,” as this watershed (depicted in orange), is not mapped as reaching the sea. Bizarrely, a few limited areas in Canadian that drain either into the Great Lakes or directly into the Saint Lawrence are mapped as part of the same drainage basin, yet almost all of the northern portion of the watershed goes unmapped.
Watershed of the U.S. map
The general exclusion of the Saint Lawrence River from the U.S. geographical imagination stems in part from a heritage of nationalist thinking. The Great Lakes figure prominently, as four of them are partly in the U.S. and the fifth, Michigan, is entirely American. But the Saint Lawrence? It seems to be regarded as a Canadian river, and therefore of relatively little significance. Contributing to this partial erasure of the Saint Lawrence is the fact that its upper reaches, bordering New York, is something of a demographic and economic backwater. The two largest cities along this portion of the river, Cornwall and Brockville in Ontario, have modest populations (47,800 and 22,100 respectively), and those on the U.S. side are smaller still.
The reduced significance of the upper Saint Lawrence is partly rooted in physical geography and history. Before the development of canals, traveling upstream from Montreal would take one into Lake Ontario but not much farther, given the profound barrier of Niagara Falls. To reach the upper Great Lakes, and hence the interior of the continent, travelers would instead voyage up the Ottawa River and then make a short portage to gain access to Georgian Bay in Lake Huron (see the map below).
Ottawa-River Route to Interior North America
But regardless of such occlusion in the U.S. geographical imagination, the Saint Lawrence is one of the great rivers of the world, the tenth largest by average discharge. In terms of flow, the Mississippi and Saint Lawrence are in a league of their own as far as North America rivers are concerned, with average discharges into the sea of 21,300 and 17,600 cubic meters per second respectively. The next largest, the Mackenzie, flowing into the Arctic Ocean, has an average discharge of 9,800 m³/s, whereas that of the mighty Columbia is “only” 7,400 m³/s (see the map posted below).
Largest Drainage Basins of North America map
The relatively short length of the Saint Lawrence has probably contributed to its relative obscurity in the United States. But its actual length is the subject of debate. Does it include the massive estuary, the largest in the world according to some sources? Although most watershed maps of the Saint Lawrence include the estuary, sometimes even up to Anticosti Island (see the Wikipedia map posted below), I have covered only a small portion of it on my own map (the second figure posted above). Equally vexing is whether the Great Lakes, and the short rivers that flow between them, can be counted as part of the same river. Vanishingly few sources do so, but the Wikipedia article on the Saint Lawrence includes all options in reckoning its length:
The two previous GeoCurrents posts examined the biological significance of continents by looking at the distribution of animals. It is time now to turn out attention to plants.
One of the most influential divisions of the world into “floristic kingdoms” is that of botanist Ronald Good, found in his 1947 book The Geography of Flowering Plants. As can be seen in the map below, Good’s floristic regionalization scheme has important similarities with, as well as differences from, Alfred Russel Wallace’s pioneering zoogeographical map. The Americas, for example, are similarly divided not at the isthmus of Panama but rather across Mexico. In the most striking divergence, northern North America does not stand alone but is instead connected with northern Eurasia and North Africa in the massive Holarctic kingdom. Most of Africa is likewise linked to southern Asia, along with Australia’s Cape York Peninsula, to form the “Paleotropical” kingdom. The rest of Australia combines with most of New Zealand to form another kingdom. Southwestern Africa, in contrast, forms its own diminutive Cape Kingdom, which is distinguished from the rest of Sub-Saharan Africa by its Mediterranean climate. Good also differentiated an Antarctic floristic kingdom, which has no counterpart in any zoogeographical system that I have seen. Unfortunately, this region is not easily discerned on the Wikipedia map that I have posted, despite its prominent label. It can be seen, however, in the pale color used for southern New Zealand and the far southern tip of South America. All told, Good’s map of flora departs more sharply from the architecture of continents scheme than do similar maps of fauna.
Good’s Map of Floristic Kingdoms
In the 1970s, Good’s scheme was modified and expanded by the Armenia botanist Armen Leonovich Takhtajan. Unfortunately, the maps that I have found of Takhtajan’s floristic regions are difficult to parse, as can be seen in the one reproduced below. Suffice it to say that he identified six floristic kingdoms, which he subdivided into 35 “regions” and many more “provinces.” Takhtajan’s kingdoms are essentially the same as those of Good, although some of their boundaries differ. Takhtajan, for example, maps an expansive Antarctic Kingdom encompassing most of Chile, Argentina, and New Zealand, as well as a host of islands scattered over a vast swath of the southern hemisphere. Despite its name, the region barely touches Antarctica itself, reaching only the northern extension of the Antarctic Peninsula. This restriction is hardly surprising, as the northern Antarctic Peninsula is the only part of the continent that supports any vascular plants, and it has only two (Antarctic hair grass, Deschampsia antarctic, and Antarctic pearlwort, Colobanthus quitensis).
Takhtajan’s Floristic Regions Map
The existence of an Antarctic floristic kingdom that excludes Antarctica might seem odd, but it makes sense in terms of geological history. When this biogeographical region began to form, Antarctica itself was both vegetated and closely linked to other landmasses, including South America and Zealandia (the submerged continent that is now mostly limited to New Zealand). The map posted below shows the arrangements of the major landmasses in the early Cenozoic Era some 60 million years ago, almost 30 million years before the formation of the Antarctic ice cap.
The Architecture of Continents 60 Million Years Ago
The Antarctic floristic region therefore does partially reflect continental divisions, but not those of the present day. Despite the passage of many millions of years, the similarities of the plant life found across this far-flung “kingdom” are striking. As noted in a Wikipedia article:
According to Ronald Good, about 50 genera of vascular plants are common in the Antarctic floristic kingdom, including Nothofagus [the southern beeches, dominant in many forests] and Dicksonia. Takhtajan also made note of hundreds of other vascular plant genera scattered and isolated on islands of the Southern Ocean, including Calandrinia feltonii of the Falkland Islands, Pringlea antiscorbutica of the Kerguelen Islands, and the megaherb genera of the New Zealand Subantarctic Islands.
Other botanists have developed their own floristic regionalization schemes, not all which include an Antarctic region. In 2001, for example, C. B. Cox deleted both the Antarctic and Cape kingdoms while dividing the vast Paleotropical region on partly continental lines, distinguishing an African from an Indo-Pacific kingdom. His map thus has closer affinities to Wallace’s zoogeographical map than it does with the floristic schemes of Good and Takhtajan. But Cox still deviated from the architecture of continents by placing northern North America and northern Eurasia in the same floristic zone.
Floristic Regions of Cox
The most useful global map of floristic regions that I have found is the one produced by Encyclopedia Britannica (posted below). This map largely follows the work of Good and Takhtajan. It divides plant “kingdoms” into “subkingdoms,” “regions,” and “divisions,” and clearly depicts the floristic affinities of widely scattered islands.
Britannica Floristic Regions Map
The “Macaronesian division” of the Boreal (Holarctic) Kingdom, is of particular interest. Macaronesia covers several islands and archipelagos in the northeastern Atlantic: the Azores, the Canary Islands, the Cape Verde, and Madeira. Many of these islands are – or were – noted for their laurisilva, or laurel forests, which require high humidity as well as stable and mild temperatures. Such conditions were once far more widespread than they are today. As noted in a Wikipedia article:
Laurel-leaved forests, called laurisilva, once covered most of the Azores, Madeira, and parts of the Canaries at an altitude of between 400 and 1,200 metres (1,300 and 3,900 ft), the eastern Canaries and Cape Verde being too dry. These forests resemble the ancient forests that covered the Mediterranean Basin and northwestern Africa before the cooling and drying of the ice ages. Trees of the genera Apollonias, Clethra, Dracaena, Ocotea, Persea, and Picconia, which are found in the Macaronesian laurel forests, are also known, from fossil evidence, to have flourished around the Mediterranean before the ice ages.
The Britannica map also depicts New Caledonia as forming its own division, part of the Indo-African (or Paleotropical) Kingdom. Although New Caledonia is part of the mostly submerged continent of Zealandia, it does not have New Zealand’s Antarctic floristic affinities. Evidently, the piece of continental crust on which New Caledonia sits was formerly submerged, with the modern island group emerging roughly 30 million years ago due to complex tectonic process. What make the flora of New Caledonia unique is not its ancient Antarctic origin, but rather its adaptation to soils infused with metal, especially nickel and chromium. As explained in a Wikipedia article:
The native flora [of New Caledonia] evolved many extremophile species that thrive in environments sufficiently toxic to put invasive plants at a disadvantage. Many areas, mainly on Grand Terre have some very high concentrations of metalliferous rocks. Their mineral content is poorly suited to most foreign species of plants. Ultramafic rocks also contain elevated amounts of chromium and nickel which may be toxic to plants. As a result, a distinctive type of vegetation develops on these soils.
Not surprisingly, the economy of New Caledonia, a dependency of France, is heavily dependent of nickel mining and smelting. Roughly 10 percent of the world’s nickel reserves are found in New Caledonia.
An important 2013 article in published in Science, “An Update of Wallace’s Zoogeographic Regions of the World,” redivides the world’s faunal regions. The study is based on a sophisticated statistical and phylogenetic analysis of 21,037 species of amphibians, non-pelagic birds and terrestrial mammals. In the end, the authors split the world into 11 zoogeographic realms, almost twice as many as Alfred Russel Wallace demarcated in 1876. They also subdivide some of these “realms” into smaller zoogeographical “regions,” resulting in 20 separate divisions. (See the three maps posted below.)
Alfred Russel Wallace’s Zoogeographic Regions Map
Zoogeographic Realms Map
Zoogeographic Regions Map
These updated zoogeographic maps stray further from the continental architecture than Wallace’s map of 1876. Let us begin with the first map, which depicts the larger faunal “realms.” Here the Eurasian Palearctic zone extends across most of the tundra zone of far northern North America, reflecting linkages that were established during glacial periods when the sea level was lower than it is today. But in the Eastern Hemisphere, the Palearctic zone is more restricted than it is in Wallace’s scheme, as North Africa and most the Middle East are placed instead in a separate Saharo-Arabian realm, while Japan, Tibet, and eastern China are also given a realm of their own. The new map retains Wallace’s “Oriental” faunal zone in South and Southeast Asia, but pushes its boundary farther to the southeast, reaching the zoogeographic line of Weber rather than that of Wallace. New Guinea is removed from Australia and appended instead to an “Oceanian” realm, while New Zealand is linked to Australia. Finally, Madagascar is separated from Africa and placed in its own “Madagascaran” realm. All told, these revisions further differentiate the zoogeographic map from that of the continents.
The more finely divided second map deviates further still from the continental model. Here North America is split into four pieces, although the division between the Panamanian and Amazonian regions does come close to the continental border between North and South America. Both South America and Sub-Saharan Africa are divided into two faunal regions. More important, Eurasia vanishes entirely, split among no fewer than eight zoogeographical regions. Intriguingly, the separation of the Indo-Malayan region from the larger Oriental realm almost exactly follows the conventional geographical division of Southeast Asia into Mainland and Insular (island) sub-regions. Note also the “Papua-Melanesian” outliers in northeastern Australia.
It is not surprising that these new maps place Madagascar in its own faunal realm and region. Although Wallace had slotted Madagascar, along with the Mascarene Islands and Seychelles, in a distinctive faunal subregion, he still placed it within thhise larger “Ethiopian” region (Sub-Saharan Africa, essentially). I have always found this classification perplexing, as the fauna of Madagascar is very distinctive. The landmass containing Madagascar, India, and Seychelles rifted away from Africa some 130-160 million years ago, while Madagascar began to split from India around 85 million years ago. As a result of this long separation, Madagascar developed its own faunal assemblage. As another map in the “Update of Wallace’s Zoogeography” article reveals (posted below), Madagascar has the second highest level of “evolutionary uniqueness” among the world’s zoogeographical regions, following only Australia. Such uniqueness is especially pronounced among mammals. As described in a Wikipedia article:
The mammalian fauna of Madagascar is highly distinctive and largely endemic. The extant nonmarine, nonchiropteran [non-bat] taxa constitute (as of June 2014) 168 species, 40 genera and 9 families; of these, besides a probably introduced shrew, endemic taxa make up all the species, all the genera, and all but one of the families. This endemic terrestrial fauna, consisting of lemurs, tenrecs, nesomyine rodents and euplerid carnivorans, is thought to have colonized the island from Africa via four (or five, if aye-ayes arrived separately) rafting events. The other historic terrestrial or semiterrestrial mammal group, the extinct hippopotamuses, is thought to have colonized the island possibly several times, perhaps via swimming
Zoogeographic Uniqueness map
The fauna of Madagascar, moreover, had been far more distinctive than it is today prior to the initial human settlement of the island roughly 2000 years ago. As the Wikipedia article cited above also notes:
Earlier in the Holocene, Madagascar had a number of megafaunal mammals: giant lemurs such as Archaeoindris which at over 200 kg was comparable in mass to the largest gorillas, as well as the hippopotamuses. The island also hosted flightless elephant birds weighing up to 700 kg, the largest known birds of all time. All of these went extinct following the first appearance of humans about 2000 years ago. Today, the largest surviving native mammals of the island, such as the indri and fossa, have weights only approaching 10 kg. Most if not all of the 29 listed extinct species are believed to have died out in prehistoric times; none of these are known to have survived into the post-European contact period.
As a final comment, the “Update of Wallace’s Zoogeographic Regions” article includes separate maps of amphibian, avian, and mammalian regions. Intriguingly, North America (minus the Caribbean) forms its own amphibian region, as do Madagascar, New Guinea, and the Philippines. It is probably not coincidental that North America, and especially the southeastern United States, has by far the world’s greatest salamander diversity (see the map below). The map of mammalian regions is also noteworthy, as it is here that the relationship between continents and zoogeographical regions collapses most completely.
Amphibian, Bird, & Mammal Zoogeographic Regions map
If the division of the terrestrial world into continents is partially but misleadingly rooted in geology, as recent GeoCurrents posts have argued, we must also ask whether it reflects the distribution of animal and plant life. Are continents, in other words, entities of biological significance? To answer this question, it is useful to begin with the work of Alfred Russel Wallace (1823–1913), the “father of zoogeography.” Wallace was one of the world’s greatest naturalists, deserving far greater recognition than he receives. Most impressively, he derived the theory of evolution through natural selection independently from Charles Darwin. Indeed, Darwin only decided to publish The Origin of Species after he received a letter from Wallace that outlined the same key ideas.
In this work, Wallace distinguished six great zoogeographical regions, or faunal realms, which he subdivided into smaller units. Slightly modified maps of Wallace’s scheme are still commonly encountered, remaining a cornerstone of zoogeography. To be sure, other authors have devised their own regionalization systems, one of which will be examined in a forthcoming post. But it still behooves us to compare Wallace’s pioneering map with the familiar depiction of continents to see the extent to which they coincide.
Alfred Russel Wallace’s Zoogeographical Regions Map
Map of Zoogeographical Regions Based on Wallace
The simple answer is that Wallace’s scheme accords reasonably well with the standard six-continent model – which combines Europe and Asia to form Eurasia – but with a few important off-sets and one significant addition. Wallace’s Nearctic region focused on North America, for example, is distinct from his Neotropical region focused on South America. The division is not at the continental boundary along the Isthmus of Panama, however, but is rather located well within North America, at the highland/lowland transition zone in southern and central Mexico. Similarly, Wallace’s “Ethiopian” region includes southern Arabia but excludes the part of Africa located north of the Sahara, which is instead appended to the Eurasian Palearctic region. The barrier separating these two faunal regions is thus one of aridity rather that one of water. Eurasia is itself divided, with the Indian subcontinent, along with most of Southeast Asia and southern China, forming its own “Oriental” region. Here the division is both topographic, marked by the Himalayas and other imposing highlands, and climatic, with the tropics separated from the temperate zone. Wallace’s scheme also extends the Australian zoogeographical region well beyond the continent of Australia. Not surprisingly, it includes New Guinea, which sits on the same chunk of continental crust. But it also encompasses New Zealand, which is part of the mostly submerged continent of Zealandia, as well as eastern Indonesia and the larger archipelagos of Melanesia. To sum it up, one might say that Wallacian zoogeography is partially but imperfectly lodged in the architecture of continents.
For most of the Cenozoic Era (the “age of mammals,” 66 million years ago to the present), the linkage between continental landmasses and distinct faunal assemblages was closer than it is today. South America, most notably, had been an isolated “island” with a unique set of animals for millions of years. It started to rift away from Africa some 130 to 110 million years ago, during the Cretaceous period, and did not gain a connection to North America until roughly 2.7 million years ago. The subsequent linkage of the two American landmasses resulted in the “Great American Biotic Interchange” (or GABI). In this exchange, many South American species moved north just as many North American species moved south. The result was a degree of homogenization of the faunas of the two continents. The main zoological division also moved north into southern Mexico, as it came to be based more on climate than on the legacy of physical separation.
South America as an Island 60 Million Years Ago Map
As is mentioned above, several studies have devised their own schemes of zoological regions, most of which remain relatively close to that of Wallace. One of the most detailed of these revisions is found in a 2013 article modestly entitled “An Update of Wallace’s Zoogeographic Regions of the World.” As we shall see in the next GeoCurrents post, the maps in this impressive article depict a much more intricate situation, as well as one that strays further from the continental model.
On the surface, there is little to distinguish small islands on continental fragments from oceanic islands that have no continental connections. Careful observation, however, can reveal some interesting differences. The Seychelles microcontinent in the Indian Ocean is an intriguing case in point. The islands sitting on this chunk of continental crust constitute a sovereign state, also called Seychelles. Seychelles is a small country of only 176 square miles (457 km2) spread out over 115 islands, fewer than a dozen of which are inhabited. Roughly 90 percent of its 118,000 inhabitants live on Mahé, the largest island in the archipelago, which is only 60.7 square miles in extent. A relatively prosperous country – with the highest per capita GDP in Africa, the continent to which it is conventionally appended – Seychelles relies heavily on tourism, much of which is up-scale.
Seychelles Micro-Continent Map
I doubt that many tourists notice anything distinctive about the physical geography of Seychelles. To be sure, its mountainous main islands are nothing like the low coral atolls of the Maldives, another tourism-dependent Indian Ocean country. But they might seem reminiscent of the volcanic high islands of the Pacific, such as Tahiti or those of the Hawaiian archipelago. But a geologically knowledgeable traveler would note the exposed granite of the larger Seychelles islands, which are very different from the volcanic rocks found in elevated oceanic islands. Granite erodes in a different manner than volcanic rocks, often giving rise to distinctive landscapes (see the illustration below).
Granite Rocks of Seychelles
The Seychelles’ granite core is old, having formed during Precambrian times roughly 750 million years ago. The area in which they formed was then part of the supercontinent of Gondwana. A large segment of Gondwana, which included Australia, Antarctica, India, Madagascar, and the Seychelles, began to rift away from the landmass roughly 145 million years ago. Madagascar-Seychelles-India later separated from Australia-Antarctica, and, later still, Madagascar broke away from this landmass. The Seychelles micro-continent finally began to rift away from India around 66 million years ago, roughly the time the dinosaurs died out and the Earth transitioned from the Mesozoic to the Cenozoic era.
Geological History of Seychelles Map
As an isolated microcontinent that separated from larger land masses many millions of years ago, Seychelles has no indigenous land mammals. It does, however, have several amphibians and reptiles whose ancestors were present before it split away from the larger lands. Of these, the most interesting to my mind are eight species of caecilians, all of which are endemic to the archipelago.*
Caecilians constitute one of the three living orders of amphibians, the Gymnophiona (the others are the frogs & toads [Anura] and salamanders [Urodela]). Caecilians are a relatively diverse lot, divided into ten extant families with 44 genera and 215 species. They can be quite abundant, with hundreds or even thousands of individuals living per hectare in undisturbed tropical rainforests. Although most caecilians are relatively small, Caecilia thompsoni can reach five pounds in weight and five feet in length.
But despite their abundance in certain habitats, caecilians are rarely encountered and have been relatively little studied. Although a few species are fully aquatic, most live underground in moist soil. Although they have vestigial eyes, burrowing caecilians are effectively blind and several species have a superficial resemblance to earthworms. But, like other amphibians, caecilians are vertebrates and are thus much more closely related to humans than to earthworms. Many have thick and hardened skulls, which they use to push through the soil as they hunt invertebrates. All species have teeth, which can be sharp and of fearsome appearance. Some people with whom I shared the photos posted below recoiled, viewing caecilians as the stuff of nightmares.
Caecilian Teeth
As the map posted below shows, caecilians live only in humid tropical zones. Their center of diversity is the Amazon Basin, with southwestern India forming a secondary center. Seychelles is also significant, with eight species and three genera. In the Americas, the range of caecilian does not extend north of southern Mexico, but individuals of the fully aquatic species Typhlonectes natans, commonly called the “rubber eel,” have recently been found in southern Florida. Presumable, they were released into the wild by people who had kept them in aquariums.
Caecilian Range and Diversity Map
One of the most intriguing aspect of caecilians is the care that some species give to their young, a trait that is relatively unusual among amphibians. Some caecilians, belonging to at least two families, nourish their young through “maternal dermatophagy.” As explained by Wikipedia:
[They feed their] young by developing an outer layer of skin, high in fat and other nutrients, which the young peel off with modified teeth. This allows them to grow by up to 10 times their own weight in a week. The skin is consumed every three days, the time it takes for a new layer to grow, and the young have only been observed to eat it at night.
Caecilian Maternal Care
Scientists have recently discovered that some female caecilians produce a substance very similar to milk, which is also used to feed their young. Milk is produced only by mammals, although pigeons, flamingos, and male emperor penguins feed their young with a somewhat similar substance, known as “crop milk.” Oddly, the nourishing caecilian secretions seem to be more similar to those of mammals than of birds. The discovery of amphibian “nursing” is recounted in an All That’s Interesting article:
[Researchers] … noticed that ringed caecilian young kept themselves close to the mother’s cloacal opening, the rear channel that serves urinary, digestive, and reproductive purposes. “The babies’ heads were close to the female’s cloacal opening all the time,” Jared told Scientific American. “Some even put their heads inside and seemed very excited.” The team decided to compare the intestines, bladders, cloacae, and oviducts of female ringed caecilians with and without offspring. Within the mothers’ oviducts, they found large glands full of fatty acid and a sugar-rich milky substance. The makeup of this “milk” was very similar to that of mammals. Researchers noted that it likely supplemented the protein-rich skin that the babies fed on once a week. Additionally, researchers discovered that the offspring wriggled near the cloacal opening and emitted high-pitched sounds to encourage the mother to release her milk. This phenomenon could occur up to six times a day. Such begging behavior has never been observed in amphibians before.
Another other interesting but sad fact about caecilians must be mentioned. In the folklore of several regions of India, these secretive and innocuous amphibians are feared and despised, incorrectly regarded as venomous. As a result, they are often killed whenever encountered, reportedly by being doused with salt or kerosene. Ironically, in India cobras – which kill several thousand Indians every year – are generally respected and even revered, largely because of their association with Lord Shiva.
Finally, it may be instructive to compare the geographical ranges of caecilians and moles (mammals in the family Talpidae). These highly dissimilar animals occupy similar ecological niches, living underground and burrowing to pursue earthworms and other soil invertebrates. Significantly, their ranges do not seem to overlap. Moles are found in humid (or seasonally humid) temperate areas in North America, Europe, and Asia, and do not extend into the tropics. Is it possible that they have been excluded from such areas by competition with caecilians? One place where their ranges might overlap is the southeastern Himalayas. The chatbot Grok 3, however, claims that adaptation to different microhabitats prevents moles and caecilians from living in the same places within this larger region:
In northeastern India (e.g., Assam or Arunachal Pradesh), caecilians are found at lower elevations in wet, forested areas, while Talpidae moles may inhabit higher elevations in the same region’s montane zones. However, their microhabitats rarely intersect. Caecilians favor saturated, organic-rich soils or streamside environments where their moist skin and amphibian life cycle are supported. Talpidae moles, being mammals, prefer drier, looser soils for tunneling and are less dependent on constant moisture. Elevation plays a key role: caecilians are scarce above 2,000 meters due to colder temperatures, while moles thrive at these higher altitudes.
As the previous GeoCurrents post argued, the division of the terrestrial world into a handful of continents derives in part from the division of the Earth’s crust into tectonic plates. But there is another geological factor equally pertinent to the concept of “continent.” It refers not to individual landmasses but rather to a particular kind of crust – the outermost rocky layer of the planet – regardless of whether it extends above sea level. If continents are defined based on continental crust, an entire “continent” could conceivably be well below sea level.
As the diagram posted below show, the Earth has two kinds of crust: oceanic and continental. Oceanic crust is relatively thin and dense, rich in iron and magnesium. Less-dense continental crust, rich in aluminum silicates, is much thicker. Being light and thick, continental crust generally extends above sea level. But the continental shelves that skirt most coastlines of continents are below sea level even though they are composed of continental crust. Several isolated fragments of continual crust are almost entirely submerged. The largest of these inundated “continents” is Zealandia, which covers approximately 4.9 million km2 (1.9 million sq mi). Only about seven percent of Zealandia is currently above sea level, mostly the two main islands of New Zealand and New Caledonia. Once connected to the massive southern continent of Gondwana, Zealandia began to rift away roughly 85 million years ago. As this occurred, its crust stretched and thinned, resulting in a lower elevation than the major continents.
Oceanic and Continental Crust Diagram
Topography of Zealandia Map
Although oceanic crust is generally well below sea level, exceptions occur. Extensive volcanic eruptions associated with hotspots can generate exceptionally thick oceanic crust, creating sizable island. Important examples include the Hawaiian Islands, the Mascarene Islands (Reunion and Mauritius) in the Indian Ocean, and, to some extent, Iceland. Iceland is a complicated case, however, as it is essentially oceanic but does contain a few fragments of continental crust. Kerguelen, in the southern Indian Ocean, is another such “composite” island. It is associated with a hotspot and as a result is mostly covered by rock of oceanic origin. But these rocks sit above a continental fragment that rifted away from Gondwana roughly 100 million years ago. As a result, Kerguelen is depicted as a microcontinent in the map posted below.
Areas of Continental Crust Map
During glacial periods, when the sea dropped due to the massive amount of water in continental ice caps, most continental shelves were above sea level. The continents were therefore larger than they are today, with a much closer correspondence between continental crust and dry land. Zealandia, however, had subsided to such a degree that even at the Last Glacial Maximum (LGM) around 20,000 years ago, only around ten percent of it extended above sea level. But that was enough to join the islands of New Zealand together, forming a significantly larger landmass than what exists today.
The Earth at Glacial Maximum 20,000 Years Ago Map
New Zealand at Glacial Maximum Map
As I was not able to find an adequate map of continental crust, I made my own, which must be regarded as a crude approximation (posted above). Several difficult judgement calls were necessary, as many island chains located tectonically active areas are composed of complex mixtures of oceanic and continental crust. Most, however, are dominated by oceanic crust, and as a result do not appear on this map (examples include the Andaman and Nicobar islands and the Kuril and Aleutian archipelagos). The Ryukyu Archipelgo, on the other hand, is composed mostly of continental crust and thus appear on this map as a narrow peninsula extending south from Japan.
Needless to say, it is difficult to derive the standard seven-continent model from this map of continental crust. A six-continent model, however, can be construed, albeit one quite different from the conventional six-continent model. Its constituent elements would be North America-Eurasia, Africa, South America, Antarctica, Australia, and Zealandia.
Tectonic plates are the basic building blocks of the Earth’s lithosphere, its outermost rocky layer. As these large segments of crust slowly move, landmasses and sea expanses are gradually rearranged. The current configuration of tectonic plates shows a tight connection with the architecture of continents: North America is on the North American plate; South America is on South American plate; Africa is on the African plate; Australia is on the Australian plate; and Antarctica is on the Antarctic plate. All these plates include large stretches of the sea, but they are anchored on their eponymous continents. The only exceptions are Europe and Asia, which are both located on the Eurasian plate. These correspondences indicate that the six-continent model – in which “Eurasia” supplants “Europe” and “Asia” –most closely represents the geological foundation of the Earth’s major landmasses. The six-continent model has the additional advantage of remaining faithful to the implicit definition of continents: large landmasses that are separated – or almost separated – from other large landmasses by stretches of the sea.
Relationship between Continents and Tectonic Plates Map
A closer inspection of the tectonic map, however, reveals that the connection between continents and plates is not as tight as it seems. Note that the Eurasian plate does not include either the Indian subcontinent or the Arabian Peninsula, both of which occupy their own plates. By geological criteria, these regions might be considered continents in their own right, although they obviously lack any maritime separation from Eurasia. The actual tectonic situation, however, is more complicated than that. The Indian plate, for example, is sometimes considered to be part of a much larger Indo-Australian plate, which would seemingly imply that India and Australia are part of the same “tectonic continent.” The current consensus, however, is that India and Australia were formerly on a single plate that has either split into two separate plates or is in the process of gradually splitting. In a tectonically informed view, all continents – and oceans – are but temporary entities.
Disconnections between Continents and Tectonic Plates map
Two Depictions of the Indo-Australian Plate map
Other disconnections between continents and tectonic plates are easily located. The North American Plate, for example, includes a huge part of eastern Siberia in Asia (or Eurasia), as well as northern Japan. The same plate does not, however, include much of western California or any part of Baja California in Mexico, which are instead located on the Pacific plate (see the map posted above for these and other discrepancies). More important, the African plate does not include Madagascar and most East Africa, which are instead on the Somali plate. Most global maps of plate tectonics, however, do not depict the Somali Plate. They presumably do so either for simplification or because the Somali plate is relatively “new,” currently in the process of rifting away from Africa.
Detailed maps of tectonic plates reveal an even more intricate situation. An excellent Wikipedia map, posted below, shows that most of eastern Asia is not necessarily on the Eurasian Plate, but sits instead on the Amur, Yangtze, and Sunda plates. Although the Amur and Yangtze plates can be defined as subdivisions of the Eurasian Plate, the Sunda Plate is more distinct. As its Wikipedia article notes, “The Sunda plate was formerly considered a part of the Eurasian plate, but GPS measurements have confirmed its independent movement at 10 mm/yr eastward relative to Eurasia.” Yet as detailed as it is, this map leaves out several important “micro-plates.” Another Wikipedia map, for example, shows that the Somali plate is itself divided into smaller segments. One of these, the Victoria plate, located between the two arms the East African Rift, is not moving in a single direction, but is rather rotating in a counterclockwise manner.
Detailed Wikipedia Map of Tectonic Plates
East Africa Tectonic Plates Map
A tectonic map that includes all micro-plates can become extraordinarily intricate, as the next post will explore.
Today’s post is another geo-quiz, which asks the reader to name an important geographical feature. Similar quizzes will be posted later this week. This quiz unfolds gradually, adding lines frame by frame until the feature becomes obvious and is named.
Screenshot
Another line of the same type is added.
Screenshot
Many lines of the same type are added and are colored blue.
Screenshot
Straight east-west and north-south lines are added.
Screenshot
Another line of the same type as the east-west and north-south lines is added.
Today’s post provides labels for the water-body shapes that were posted on May 13. This “answer key” has two parts. First, each shape is replicated along with a hint in the form of a riddle. After each of these “geo-riddles,” the answer is given on a map that provides additional geographical context.
(Riddle 3# is rather obscure; it asked whether is is sound to put south on the top of the map.)
The previous GeoCurrents post argued that even geographically informed people have a difficult time recognizing the shapes of seas and other large water bodies, due largely to our intrinsic tendency to prioritize land over water. But this tendency does not always come into play. The Mediterranean Sea, for example, is easily recognizable. But this is because we readily discern the large peninsulas that jut into it, rather than the sea itself. Lacking such peninsulas, the Baltic Sea is less seen (see the second image below).
Screenshot
Screenshot
The shape of one sea is particularly difficult to identify, even though it is very large (1,583,000 km2; 611,200 sq mi) and clearly defined. It is also the site of a significant and seemingly unresolvable territorial dispute between two powerful countries. But given its peripheral location for both of those countries, as well as the low population density of the lands that (nearly) surround it, it tends to slip off our conceptual maps. The four final shapes (below) reveal this sea in sequential stages.
The geographical imagination tends to prioritize land over water, which is hardly surprising. Water is generally in the cartographic background, conceptualized as something of a void. As a result, people have a difficult time recognizing the shapes of seas, bays, gulfs, large lakes, and other major bodies of water, even those located in areas that they know well. This seems to be true even for people with extensive geographical knowledge. To test this proposition, I have traced out the shapes of a number of well-known water bodies and used them as a quiz with a few friends and family members. None of them has done very well. Readers can test themselves with eight of these images posted below. It is important to note that these simple maps are all drawn at different scales.
At the end of the article, a few hints are provided. Answers will be given later this week.
Screenshot
Screenshot
Screenshot
Screenshot
Screenshot
Screenshot
Screenshot
Screenshot
Hints: All these water bodies are in North America. Straight lines indicate opening to the ocean or some other water body. If you are perplexed, focus on the shapes of islands and coasts rather than that of the water body itself.
A peninsula, Wikipedia tells us, “is a landform that extends from a mainland and is surrounded by water on most sides.Any body of water can count, the article claims, including rivers: “A piece of land on a very tight river bend or one between two rivers is sometimes said to form a peninsula. …” In practice, however, the inclusion of rivers occurs only at the local scale; a large extent of land bracketed by rivers is almost never called a peninsula, and for good reasons. Consider, for example, Mesopotamia, which is almost entirely “surrounded” by the Tigris and Euphrates rivers. Does it “extend from a mainland?” Is it ever depicted as a peninsula? An internet search returned exactly one instance of this usage, an article by Turkish authors on the “Vernacular Domed Houses of Harran, Turkey.” Perhaps they are not familiar with the vernacular definition of the world “peninsula” in English. Following such usage, ChatGPT insists that “Mesopotamia” is not a peninsula.
Screenshot
ChatGPT 3.5: “Mesopotamia is not a peninsula. It’s actually considered one of the world’s earliest civilizations, located in the region between the Tigris and Euphrates rivers, which is in present-day Iraq and parts of Syria, Turkey, and Iran. While it’s bordered by water, it’s not surrounded by water on three sides like a peninsula would be. Instead, it’s known as a river valley civilization due to its location between these two major rivers.
(Unfortunately, the chatbot’s answer is not fully coherent: “Mesopotamia” was the location of “one of the world’s earliest civilizations,” not the civilization itself, and it was definitely “surrounded by water on three sides,” at least if one counts flowing water in the definition of the term.)
But in Great Lakes region straddling the border between the United States and Canada the vernacular English definition of the word “peninsula” vanishes. Three large peninsulas are distinguished in this area: Michigan’s Upper Peninsula, Michigan’s Lower Peninsula, and Ontario Peninsula. All three appear prominently on political maps, but this is essentially an artifact of the division between the two countries. A non-political depiction of large water bodies in this region reveals only one large peninsula, the Lower Peninsula of Michigan, and it is not as large as it is conventionally imagined. The Upper Peninsula of Michigan and Ontario Peninsula are evident only if exaggerates the size of rivers and counts them as water bodies separating peninsulas from mainland areas. But if we were to follow this precedent, we would have to count hundreds of peninsulas that are never given that designation. The term “mainland” would also essentially lose all meaning.
Screenshot
One can, of course, object to these claims by noting that geographical terms are often based on convention rather that strict definition. As the Infoplease article on peninsulas aptly states, “The definition of a peninsula can be a bit arbitrary, and has as much to do with convention and politics as any geographical rules.” I therefore have no problem with casually referring to northwestern Michigan as the “Upper Peninsula,” and I will continue to do so myself. But it is still important to draw attention to geographical conventions that defy geographical definitions.
Screenshot
The prime example of a “geographical convention that defies geographical definitions” is portraying Europe as a continent. As Infoplease further notes, “On a map Europe may look like a peninsula that extends westward from the larger landmass of Eurasia, but historians and geographers have treated it as a separate continent for centuries.” Careful geographers and historians, however, no longer regard Europe as a continent. They insist that it is best regarded, like South Asia, as a subcontinent of Eurasia.
