The Dispersal of Land Animals

by William Diller Matthew (1930)

Editor Charles H. Smith's Note: This work, one of Matthew's last, confirms that his understanding of the relationship of evolution to global dispersal episodes did not change very much over the length of his later career. Original pagination indicated within double brackets. Citation: Scientia (Bologna) 48 (1930): 33-42.

     [[p. 33]] The present distribution of land animals is conditioned partly by climate and environment, partly by the configuration of the land areas of the globe. Few animals are cosmopolitan; for the most part those of the tropics are different from those of temperate and arctic zones, those of desert and plain from those of forest and swamp, while the oceans and mountain barriers that separate the continental regions mark out a third line of division. On the basis of these faunal distinctions the world has been divided into zoological regions and sub-regions, each characterized by its peculiar fauna. The boundaries and status of these regions have been much disputed because the barriers that confine one class of animals are not equally or similarly effective with other classes, and the faunal divisions, associations and boundaries vary accordingly. The regions most generally accepted are based primarily upon the distribution of terrestrial mammals. They conform less exactly to the distribution of birds and of lower vertebrates, or of the various invertebrate groups. Broadly speaking, geography has played the most important part in mammalian distribution. Climate and environment become relatively more and more important as we go down the scale.

     If we examine the configuration of the present land areas, not on the misleading Mercator projection but on a globe or hemisphere-maps, it will appear that the northern continents form essentially one great land mass, divided only by the narrow and shallow strait between Alaska and Siberia. From it project the three southern continents as isolated peninsulas, South America and Africa connected by narrow isthmuses, Australia isolated save for a chain of islands. Antarctica is wholly isolated, with a wide stretch of deep ocean all [[p. 34]] around it. Besides these ocean barriers, two other barriers have been of prime importance in limiting distribution. The great mountain barrier of the Himalayas separates southern Asia and the associated islands as a distinct region. The great desert barrier of the Sahara cuts off central and southern Africa from the northern littoral.

     Another physiographic feature of prime importance is the continental shelf. The continents are essentially broad platforms, raised in some areas into mountains or plateaux but averaging only a few hundred feet above sea-level, and with some portions of the platform submerged a few hundred feet below sea-level. The borders of the platform are everywhere sharply defined by the continental shelf, and beyond them the ocean suddenly deepens to several thousand feet, with a comparatively level floor of an average depth of two miles.

     Most of the islands of the ocean lie upon these submerged banks, within the continental shelf and connected with the mainland by water of less than two hundred metres depth. Of the islands that lie beyond the shelf a few are of large size, most of them small volcanic or coral islands. The continents and the larger oceanic islands are substantially in isostatic balance throughout. That is to say, they are underlain by lighter rocks than those beneath the oceans, and are essentially floating at their present sea-level. The smaller islands are uncompensated so far as known.

     The writer believes that the continental platforms and ocean basins are for the most part of immense antiquity. Large parts of them have been submerged in shallow water, re-elevated, eroded, and their materials shifted about, with a general tendency to deposit sediments within or upon the margin of the shelf, while the soluble parts of the rocks, lime, magnesia, iron, are more uniformly distributed wherever life or chemical agencies may precipitate them from the ocean water. The epicontinental and littoral sediments, dominantly silica and potash-soda-alumina silicates, consolidate into the lighter acid rocks; the lime-magnesia-iron, dominant in the oceanic sediments, with a smaller increment of cosmic dust, accumulate far more slowly and consolidate into heavier basic rocks. Hence isostasy results in a general tendency to build out the margins of the continents into steep-fronted deltas, and to raise elevated strips and mountain chains along their borders; and on the other hand to deepen the ocean basins [[p. 35]] progressively and to differentiate progressively the << sial >> platform from the << sima >> deep.

     The more stable land areas have been refuges and centres of dispersal for terrestrial life since the geologic record began, and their relationships have been one of the two varying factors of dispersal and distribution.

     Modern geologic knowledge disproves the old concept of a progressively cooling earth with its attendant climatic phenomena through the successive geological periods. It indicates instead an alternation of (a) climatic conditions not unlike those that prevail today, with (b) periods of more uniform climate all over the earth. The zonar, widely varying climates were associated with uplifted mountain regions and general elevation of the continental platforms, the uniform climates with reduction of the mountains and plateaux to vast peneplains by prolonged erosion, and flooding of the low-lying portions of the continental platforms by the ocean. The contrasted temperature zones result in more active atmospheric circulation, and with the high mountain ranges result in sharp contrasts of rainfall in different regions, extensive desert areas contrasting with areas of heavy rainfall; the more uniform climates with low relief of continents would result in more uniformly distributed rainfall averaging high because of the smaller land areas and larger ocean area of evaporation. Periods of extreme elevation culminate in glacial epochs, when great ice caps cover considerable regions of the land. From such an epoch the world is today but half emerged. Greenland and Antarctica are still buried deep beneath the ice. Another great glacial period of at least equal severity, with no less elevation and extension of the continents, followed the Coal Period, and separated the Palaeozoic from the Mesozoic era. A third, probably of no less intensity, preceded the Palaeozoic era. Other epochs of uplift, sometimes accompanied by evidences of glaciation, precede or intervene between these three great landmarks of geologic time; but they are less extreme or less clearly proved to be widespread or universal. Some of them may be local, not involving a general alteration of climate or geography.

     The alternate periods of uniform climate are marked by wide extension of limestone formations over the shallow continental seas, followed often by extensive coal formations accumulated in vast swamps under a climate still moist and mild.

     [[p. 36]] These general conditions of climate and geography have been everywhere modified by local geologic movements, and it is only the great outstanding extremes that are generally recognized as universal. Nevertheless the correspondence in character and succession of formations in different parts of the globe indicates that to a great extent each of the accepted geologic periods represents a climatic cycle of this kind. It begins and closes with an epoch of uplift, and includes a far longer period of quiescence, with progressive flooding and subsequent slow emergence of the continental platforms.

     The fossil record of marine life is most complete during the epochs of submergence, more broken and scanty during epochs of emergence. The reverse might seem to hold true for terrestrial life; but in fact the limited extent and great thickness and unfavorable conditions for preservation of organic remains in the formations of the emergent phases, together with the fact that so large a part of them were deposited at the extreme margin of the continental shelf and hence now beyond our reach, makes the record of terrestrial life relatively poor, even in this phase.

     Every species of animal is an adaptation more or less perfected to the needs of its particular mode of life. It has to compete with rivals and maintain itself against enemies and adverse features of environment. If the environment remains constant the species will in time perfect itself along the line in which it is specializing. If the environment changes, whether the change be in the physical surroundings or in the associated fauna, the species must either adapt itself to the new environment, seek elsewhere for a continuance of the old, or become extinct. Evolution, migration and extinction are thus conditioned by a changing environment.

     The periods of continental uplift and zonar climates have compelled a great amount of change in terrestrial faunas. Some animals adapted themselves to new and more varied environment and conditions of life. A great adaptive radiation took place. Others migrated to regions where their old environment still prevailed, and, competing with the autochthonic fauna, introduced new conditions of survival which induced profound evolutionary changes in both native and invading faunas. Others again, unable to adapt themselves successfully or to find a refuge, became extinct, and their disappearance from the fauna left gaps in the interrelationship which were filled by a further series of adaptive readjustments.

     [[p. 37]] It is obvious that the first effects of the oncoming zonar climates would be felt in the regions nearer the poles. The fauna must be adapted to the severe conditions initiated there and spreading gradually over the temperate regions; while the tropical forest regions would serve as a refuge for less progressive types. The partially adverse environment and severe competition in the northern regions will tend to speed up selection and progressive evolution, and in result the polar regions will be centres of dispersal of higher, more dominant types. Mountain regions play a nearly similar rôle, and plains and steppes one partly similar.

     The extension of the continents to their utmost borders, and connection of the outlying ones with the central mass, provides a greater area for terrestrial evolution and tends toward the production of cosmopolitan faunas.

     Hence, in the zonar phase of the climatic cycle we expect to see new and dominant types evolved in the northern land masses, and thence invading the southern continents, where they would meet with a corresponding group or groups of dominant types evolved in the south and pressing northward. If the general relations of the continents were in the main as they are today, the southern groups, while corresponding to those of the north, would be relatively backward and weak because evolved in smaller and isolated areas. If Professor Wegner's hypothesis of continental drift or the very widely accepted views of Gondwana Land were true, then the southern dominant types should be equal to or stronger than those of the north, and the history of dispersal ought to be along the lines so conditioned. Among mammals, for instance, the notoungulates, the marsupial carnivores, the edentates should have been the source of the existing mammals of the northern world, and the artiodactyls, perissodactyls, carnivora, etc., should have been now extinct or surviving only as a few rare types in tropical forests.

     The great mountain ridges and deserts of the Old World, lying athwart the lines of polar dispersal, have served as barriers more effective or less so according to circumstances. In the New World the mountain ranges and desert strips lie radial from the northern and southern centres, so that the faunas in both North and South America are more uniformly distributed as to latitude.

     (It is to be observed that it is not the barrier that limits [[p. 38]] distribution in many instances, but the sharp diversity of climate and environment conditioned by or associated with that barrier. Eastward of the Himalayas the Oriental grades into the Palæarctic fauna conformant to the gradation of climate. The boundary between Nearctic and Neotropical faunas lies not at the long winding, narrow isthmus of Panama, but where the Mexican plateau breaks down into the Central American lowlands).

     The broad lines of dispersal of land animals are obviously in accord with the foregoing principles. Among mammals it will hardly be questioned that the highest, dominant types and groups, those of latest evolution, are found chiefly in cold or temperate climates, in the northern world, in mountain and in plain faunas; while primitive survivals are chiefly tropical or southern, and forest-living. Among lower vertebrates and the various groups of land invertebrates the same general arrangement appears to hold true in each group. Advanced progressive types are either confined to the northern world or have become cosmopolitan. Primitive survivals are found in tropical forests or in southern continents or in islands, especially tropical or southern islands.

     Turning to the fossil record of land animals we find the inferences from modern distribution confirmed whenever the record is adequate. It makes clear however that most of the present distribution of mammals is the result of dispersal and migration during the Tertiary period. To a certain extent this is superposed upon an older dispersal, apparently Mesozoic, but so imperfectly documented that the direct evidence is of no great value. To this older dispersal we owe probably the Monotremes of Australasia, and perhaps the first arrival of certain marsupials and placentals in the southern continents; but the adaptive radiation of the several groups of marsupials in the southern continents, of edentates and notoungulates in South America, of hyracoids in Africa, took place during the Tertiary period coördinately with the adaptive radiation of most of the placental orders of mammals in the northern world, and the latter progressively invaded the tropical and southern continents as these became connected with the northern land mass.

     While the geologic record of the lower land vertebrates and invertebrates is far less complete, it conforms apparently to the same principles, but it would appear that a more [[p. 39]] considerable part of their present distribution is due to pre-Tertiary dispersals; although it would seem that even with these the Tertiary adaptation and dispersal has been responsible for all the details and much of the major group distribution. Our knowledge of pre-Tertiary land faunas is relatively scanty, and mostly limited to certain restricted facies, so that the problem is more obscure and very liable to misinterpretation, especially with somewhat superficial acquaintance with the data.

     Turning from generalities to detailed examination of a few sample groups, we find that the distribution of man conforms to these principles. He is today cosmopolitan, but the dominant races are native to the northern world, have largely invaded the south temperate and tropical regions, and the tropical forests are most resistant to their colonization. In these tropical forests survive more primitive human stages, also the anthropoid apes, the old world and new world monkeys, and in the more outlying regions of forest the still more primitive lemuroids. Monkeys occur in the later Tertiary of Holarctica, lemuroids in the older Tertiary, but it is not demonstrated by the evidence that Palæarctica was the centre of dispersal of the primates, although the data may be so interpreted.

     The Tertiary record shows that the Canidæ, now cosmopolitan, evolved in North America, that the Felidæ now cosmopolitan, evolved somewhere in Holarctica, that the Ursidæ and Mustelidæ were likewise of Holarctic origin, that the Procyonidæ, now chiefly Neotropical, came from North America, and the Viverridæ, now chiefly Oriental and African, evolved in Palæarctica, and that South America lacked true carnivora until the Pliocene, Africa until the Miocene, Australia still lacks them; but each southern continent evolved in their place parallel groups of carnivorous marsupials, which became extinct when the true carnivores invaded that continent.

     Among the Artiodactyl ungulates the dominant modern group of ruminants is shown to have dispersed from Palæarctica; relatively primitive survivals formerly inhabiting Holarctica are the tragulines of the Oriental region, Hyæmoschus of the African forests, the giraffe and okapi. The camels originated in North America, spread to South America and Asia and North Africa, and have become extinct in their original home. The pigs originated in Palæarctica and spread to the Oriental and African region, the Peccaries in North [[p. 40]] America and have invaded South America and are now Neotropical. The Perissodactyl ungulates have a corresponding history. They originated in Holarctica, and the most primitive living group, the tapirs, are now oriental and neotropical, although formerly Holarctic; the rhinoceroses, formerly Holarctic, are now Ethiopian and Oriental. The horses, the most specialized, originated apparently in North America, became cosmopolitan, and still survive in Africa and Asia. One might go through the list of the various mammalian groups, finding, wherever the evidence is adequate, that the great majority, and nearly all the dominant modern types, are of northern origin.

     On the other hand, certain groups of mammals are shown to have undergone a great adaptive radiation and expansion in the southern continents, although the evidence goes to show that the primary types were originally invaders from the north. The epoch of uplift at the end of the Cretaceous enabled them to reach the southern continents, and during the isolation of the earlier Tertiary they expanded into faunas of varied adaptations, only to be overwhelmed by later northern invasions. In South America the edentates, notoungulates, litopterna and the marsupial carnivores seem to have been autochthonous since the end of the Cretaceous, and, except for a few edentates, have wholly disappeared; the hystricomorph rodents and platyrhine monkeys appear to have arrived about the middle of the Tertiary, and a large proportion of these still survive. In Africa the fossil record is less complete, but the Hyracoids, Arsinoitheres and perhaps the Proboscidea would seem to represent the authochthonous early Tertiary fauna, while a more continuous connection with Palæarctica gave rise to a succession of invasions of the northern faunas. Australia, on the other hand, more isolated, received no mammals except the marsupials, of very ancient arrival and adaptively radiated into a wide diversity of type, a few stray rodents and bats of much later arrival, and the dingo, a probable companion of the invasion by man. In absence of a great northern invasion, the marsupial radiation still survives in Australia. New Zealand, still more remote was never reached by mammals, and certain birds partly took their place. Unfortunately we have no record of the Tertiary land fauna of Australasia and New Zealand, save for a single marsupial, Wynyardia, from the Tertiary of Tasmania. It is related to the modern phalangers.

     [[p. 41]] Antarctica today affords no opportunity for the existence of a terrestrial fauna. In the Tertiary and in preceding periods of mild climate it probably was habitable. It is probable that it was heavily forested, with a flora related to the contemporary flora of the southerly continents and islands. If any land connection with South America, Australia or New Zealand existed, the flora would be closely related to that of the connected region; if not, it should be as diverse from any of them as they are from each other. A land connection again would involve a fauna largely identical, continued isolation a fauna largely in common so far as those animals are concerned which could be transported by storms or other accidents, but not likely to include mammals. Birds would be present, and in absence of mammals would probably evolve terrestrial adaptations. Bats might do so, but this seems less probable.

     I regard the existing penguins as probably a terrestrial bird readapted to their present life. It is of interest that the sole fossil evidence that we have of the Tertiary fauna of Antarctica consists of the Tertiary penguins of Seymour Island. Some of these are of gigantic size. A study of their adaptive differences from modern penguins might be illuminating.

     The distribution of the lower vertebrates conforms in its broader lines to that of the mammals. None of them, however, are so much controlled by land connections. This is even more clearly the case with most invertebrates. Birds, bats and insects can fly. Small animals or their eggs and spawn may be transported by storms or on masses of floating vegetation far more easily than mammals. While they may not normally surmount ocean barriers, yet their general presence on even the smallest and most remote of oceanic islands sufficiently proves that they do so occasionally. It is not within reason to suppose that all oceanic islands have had a mainland connection. Few of them appear to be of great geological antiquity, and their faunas are scanty and incomplete in proportion to their isolation and small area. The larger and older geologically, and the nearer to the mainland, the greater the chances of colonization and the longer the time for adaptive radiation within the island area, which is often indicated by the submarine topography as formerly larger than the present and more connected up. The distinction is [[p. 42]] clear between the strictly continental faunas of all islands within the continental shelf and the insular faunas, incomplete and peculiar in almost uniform accord with the relations above indicated. A few minor and partial exceptions (certain Mediterranean and East Indian islands) are in regions of exceptional disturbance and upheaval indicated both by the Tertiary geology of the land and the extreme irregularity of the sea bottom.

     The principles above outlined are for the most part not new. They have been held by the more cautious and conservative geologists and zoölogists for many years. They have always been opposed by others of a more radical and speculative trend of mind, to whom the fascination of daring and sensational conclusions is irresistible. In the face of this opposition the doctrine of the permanence of the continents and oceans has never been popular, nor has the interpretation of distribution in terms of existing and well known agencies aided by the cumulative effects of time been generally approved.

     There are, however, two principles of modern geology which were not known to the older writers, and which form an essential part of this interpretation. The first is that all the continents and probably all the older oceanic islands are in approximate isostatic balance (within a few hundred feet), but the small volcanic islands are not (and tend therefore to sink back when the force that uplifted them is spent). This affords a sound reason for the permanency of continental platforms and ocean basins, and strongly supports the arguments of Dana and others. The second principle is that of cyclic alternations of climate broadly corresponding to the geologic periods and eras, in place of the old concept of a gradually cooling earth. The effect of this is to modify the concepts of dispersal presented by Wallace and others, substituting a definite sequence of diverse successive phases of evolution and dispersal conditioned by the diverse phases of the climatic-diastrophic cycle, for the older concept of a continuously progressive change. It is apparent that most of modern distribution is a result of Tertiary dispersal, but that this is superposed upon one or more older cycles of dispersal.

*                 *                 *                 *                 *

Return to Home