The term pandemism also is in use, but not all authors are consistent in the sense in which they use the term; some speak of pandemism mainly in referring to diseases and pandemics, and some as a term intermediate between endemism and cosmopolitanism, in effect regarding pandemism as subcosmopolitanism. This means near cosmopolitanism, but with major gaps in the distribution, say, complete absence from Australia. Terminology varies, and there is some debate whether the true opposite of endemism is pandemism or cosmopolitism.
The term "cosmopolitan distribution" usually should not be taken literally, because it often is applied loosely in various contexts. Commonly the intention is not to include polar regions, extreme altitudes, oceans, deserts, or small, isolated islands. For example, the housefly is nearly as cosmopolitan as any animal species, but it is neither oceanic nor polar in its distribution. Similarly, the term "cosmopolitan weed" implies no more than that the plant in question occurs on all continents except Antarctica; it is not meant to suggest that the species is present in all regions of every continent.
Another concept in biogeography is that of oceanic cosmopolitanism and endemism. Though there is a temptation to regard the World Ocean as a medium without biological boundaries, this is far from reality; many physical and biological barriers interfere with either the spread or continued residence of many species. For example, temperature gradients prevent free migration of tropical species between the Atlantic and Indian-plus-Pacific oceans, even though there is open passage past continental masses such as the Americas and Africa/Eurasia. Again, as far as many species are concerned, the Southern Ocean and the Northern marine regions are completely isolated from each other by the intolerable temperatures of the tropical regions. In the light of such considerations, it is no surprise to find that endemism and cosmopolitanism are quite as marked in the oceans as on land.
Another aspect of cosmopolitanism is that of ecological limitations. A species that is apparently cosmopolitan because it occurs in all oceans, might in fact occupy only littoral zones, or only particular ranges of depths, or only estuaries for example. Analogously, terrestrial species might be present only in forests, or mountainous regions, or sandy arid regions or the like. Such distributions might be patchy, or extended, but narrow. Factors of such a nature are taken widely for granted, so they seldom are mentioned explicitly in mentioning cosmopolitan distributions.
Cosmopolitanism of a particular species or variety should not be confused with cosmopolitanism of higher taxa. For example, the family Myrmeleontidae is cosmopolitan in the sense that every continent except Antarctica is home to some indigenous species within the Myrmeleontidae, but nonetheless no one species, nor even genus, of the Myrmeleontidae is cosmopolitan. Conversely, partly as a result of human introduction of unnatural apiculture to the New World, Apis mellifera probably is the only cosmopolitan member of its family; the rest of the family Apidae have modest distributions.