XI E> HAHY

OF THE

UN IVLR5ITY

OF ILLINOIS

590.5
FI

v. 37-38

BIOLOQi

The person charging this material is re-
sponsible for its return on or before the
Latest Date stamped below.

Theft, mutilation and underlining of book.

are reasons for disciplinary action and may

"suit in dismissal from the Univers.ty.

UN.VERS.TY OP jmNO^UBRAR^TU^

L161— O-1096

Geographical Origins and Dispersions of

Termite Genera

Alfred E. Emerson1
Research Associate, Division of Insects

ACKNOWLEDGMENTS

This paper has been written in honor of Dr. Karl Patterson
Schmidt because of our long, close, personal friendship, our mutual
interest in zoogeography, taxonomy, ecology, and evolution, and
also because of the numerous collections of termites made by Dr.
Schmidt during his many zoological expeditions to various parts of
the world. We have long been associated in our professional work
at the American Museum of Natural History, Chicago Natural
History Museum, al\o! the University of Chicago. We have collected
together in the field, and we collaborated in writing a book (Allee,
et al. 1949). In 1916, near Ithaca, New York, Dr. Schmidt showed
me the first living termites I had ever seen, some years before I
started to specialize on the order in British Guiana under the
leadership of Dr. William Beebe.

Financial assistance in gathering the information contained in
this summary treatment has been generously given by the Dr.
Wallace C. and Clara A. Abbott Memorial Fund of the University
of Chicago, the New York Zoological Society, the John Simon
Guggenheim Foundation, and the Belgian American Educational
Foundation. I have also received generous collaboration from the
leading students of termites of the world during the twentieth
century and have exchanged specimens with many natural history
museums. Such a study as this could not have been accomplished
without international co-operation of many scientists and scientific
institutions. I am greatly indebted to my wife, Eleanor Fish
Emerson, for assistance in the preparation of the manuscript.

1 Professor, Department of Zoology, University of Chicago.

465

466 FIELDIANA: ZOOLOGY, VOLUME 37

INTRODUCTION

As the result of the catalogue of the termites of the world by
Snyder (1949) and the phylogenetic studies by Holmgren (1911,
1912), Hare (1937), and Ahmad (1950), termite systematics has
matured to the point where zoogeographical interpretations rest
upon a substantial foundation. Many new genera and species
doubtless await discovery, and some changes in the postulated
relationships may be expected. Ahmad (1950) verified far more of
the conclusions drawn by Holmgren (1911, 1912) than he refuted,
and one may guess that future studies of taxonomy and phylogeny
will verify from 80 to 90 per cent of the relationships indicated by
Ahmad. As knowledge of termite evolution has advanced, the
geographical relations of the genera have become far more consistent
and orderly.

The present paper deals with the distribution of the genera of
termites in the major zoogeographical regions. Many principles
of interest to ecologists, geographers, and evolutionists emerge from
the compiled data. Further refinement of interpretation may be
expected from the study of the distribution of species in their
ecological niches (see Allee and Schmidt, 1951). Spot maps of the
geographical records of each species in each genus have been prepared
by the author, but much rechecking of the taxonomy and distribution
needs to be done before these maps are ready for publication. In
the present study, the data are recorded in a tabular analysis
(Tables 1-8) that is designed to document a brief summary of
tentative conclusions (Emerson, 1952a) and to bring my tabular
analyses up to date (Emerson, 1928). Since the tables were con-
structed, Harris (1954) has described two new species of Procrypto-
termes and Amitermes from Socotra Island in the Ethiopian region.

Records of new genera and new species have been included in
the tables, thereby rendering the geographic conclusions more secure.
The new manuscript genera will be described in the near future.
Some genera will doubtless be subdivided into several genera with
further study, but most of the named genera can be considered as
well established. Complete collections available for study are a
necessity for accurate information. The collection of termites be-
longing to the American Museum of Natural History (Table 1)
and now in my custody contains more than 85 per cent of the known
named species and about 87 per cent of the known named and
unnamed species. More than 70 per cent of the named species of
the world are represented in this collection by primary type speci-

EMERSON: DISTRIBUTION OF TERMITE GENERA 467

mens (holotypes, paratypes, cotypes, neotypes). The collection, at
the end of 1954, contained 1,418 species, of which 818 were repre-
sented by all castes (Anoplotermes and Speculitermes without a
soldier caste are included), 86 by imago alone, 507 by soldier alone,
2 by worker alone, providing the species was described from the
worker, and 5 fossil species. In addition to the named species, the
collection also contains 375 new species awaiting description. A
high proportion of the genera and species has been verified, but
there are probably a number of errors that will be corrected within
the collection in the future in addition to the errors in the literature
that can be corrected only by re-examination of types or accurately
determined specimens. Collections of duplicates have been deposited
in Chicago Natural History Museum, the United States National
Museum, the Naturhistoriska Riksmuseet, the British Museum
(Natural History), and several regional institutions.

TERMS

Introduced species include those species presumably transported
through the agency of man. These may be recognized by their
occurrence in the constructions of man and by their species identity
on two sides of a natural barrier. Some twenty species of Kalo-
termitidae and Rhinotermitidae have been temporarily established
through human agency in areas outside their native habitat. These
are often of great economic importance. They give strong indica-
tions of ecological barriers to natural dispersal, both physical and
biotic. -In the tables introduced species are not included in the re-
gions to which they were transported but are listed only in their
native region. They need separate consideration and study.

Native species include species found in their native habitat or in
a man-modified habitat in close proximity to their natural occur-
rence. The tables include only the distribution of native species.

Cosmopolitan genera include those genera found in all of the major
zoogeographical regions. Kalotermes and Neotermes as now consti-
tuted are the only genera in this category. Neotermes is included,
although the only species found in the Nearctic region has its center
of distribution in the Neotropical region. Amitermes is not known
from the Papuan region but is otherwise a cosmopolitan genus.
Microcerotermes is cosmopolitan except for its absence in the Nearctic
region.

Tropicopolitan genera include those genera found in all of the major
zoogeographical regions with a tropical climate (without freezing

468 FIELDIANA: ZOOLOGY, VOLUME 37

temperatures in the termite microniche) and absent from the Palae-
arctic and Nearctic. Glyptotermes, Termes, and Nasutitermes are
the only such genera. They may occasionally penetrate subtropical
or warm temperate areas at high altitudes, or temperate extensions
in the southern portions of the tropical regions. Coptotermes is also
classified as a tropicopolitan genus, although it has one species that
extends beyond its tropical center into southern Japan. Neotermes
comes very close to this category, but it is not included because it
has one species extending into temperate Florida, one species in the
Madeira Islands, and one species in the Ryu-Kyu Islands.

Overlapping genera are those found in two or more major zoo-
geographical regions.

Endemic genera are those genera found only in a single major
zoogeographical region.

Relict genera are those that show primitive structure and are
considered to be ancient forms that have not evolved much or
rapidly (bradytelic) since their origin. Relict genera have few sur-
viving species but were once probably widespread, with numerous
species. The best example among termites is Mastotermes darwini-
ensis Froggatt, which belongs to a monotypic family but once was
a widespread genus, as attested by the fossil evidence (Table 3).
The endemic Archotermopsis and its related endemic genera within
the Termopsinae also belong to this category, but only the ancestral
genera were possibly widespread. Occasionally relict genera may
have survivors in several major regions. Examples are Stololermes
and Porotermes.

Semi-relict genera are those that indicate competitive reduction
in numbers but still have numerous species. They typically inhabit
several major regions, including Australia. Usually the numbers of
species do not indicate the region of origin, possibly because of
proportionately greater extinction of species in ancient areas of
abundance. Examples are Kalotermes, Neotermes, and Heterotermes.

Relatively primitive endemic genera are members of successful
specialized subfamilies that were unable to disperse from the region
of origin when climatic highways were available, probably because
of competitive biotic barriers. Genera in this category include the
more primitive members of the major subfamilies of Termitidae.
Examples are Protohamitermes to Eurytermes (Amitermitinae), Hop-
lognathotermes to Megagnathotermes (Termitinae), Acantholermes to
Synacanthotermes (Macrotermitinae), Syntermes to Rhynchotermes
(Nasutitermitinae), and Paracornitermes to Curvitermes (Nasutiter-
mitinae).

EMERSON: DISTRIBUTION OF TERMITE GENERA 469

Medium-specialized genera are those midway between the rela-
tively primitive endemic genera and the specialized endemic genera.
They are typically widespread and overlap several major regions.
These genera appear to have risen from the relatively primitive
genera in their respective subfamilies but, unlike their immediate
ancestors, they were able to penetrate the biotic barriers that con-
fined the more primitive forms. These genera often exhibit more
obviously specialized defensive adaptations in the soldier caste,
although their penetration of new regions may be partially due to
greater or more versatile adaptation to the physical environment as
well as to their biotic environment. Some of these genera show a
significant distributional difference in numbers of species. The
largest number is usually found in the region of origin as indicated
by the distribution of their somewhat more primitive prototypes.
However, because of ecological adjustments or for obscure reasons,
the largest number of species does not always indicate the region of
origin. Examples of medium-specialized genera are Glyptotermes,
Coptotermes, Schedorhinotermes, Microcerotermes, Amitermes, Termes,
Odontotermes, s. str., and Nasutitermes. It may also be concluded
that some ancestral medium-specialized genera now extinct were
replaced by specialized endemics. Modern relatives of these un-
known medium-specialized genera are Promirotermes, Paracapri-
termes, Capritermes, Angular iter mes, and Subulitermes. In some in-
stances an ancestral genus may be very close to or identical with
the modern relative. For instance, Schedorhinotermes is probably
very similar to the ancestral genus that gave rise to the neotropical
specialized genera, Rhinotermes, Dolichorhinotermes, and Acorhino-
termes, and might thus be considered a medium-specialized genus
that was eliminated in the Neotropical region by its own specialized
descendants.

Specialized endemic genera are those that arose from medium-
specialized genera. These genera are thought to be confined to one
region because of climatic or geological barriers at the time of their
origin. Examples are Eucryptotermes, Rhinotermes, Dolichorhino-
termes, Acorhinotermes, Hoplotermes, Ahamitermes, Globitermes,
Cephalotermes, Cylindrotermes, Drepanotermes, Gnathamitermes, Ortho-
gnatholermes, Cavitermes, Paracapritermes, Homallotermes, Quasi-
termes, Pericapritermes, Neocapritermes, Longipeditermes, Veloci-
termes, Constrictotermes, Tumulitermes, Angular itermes, Subulitermes,
Ceyloniter melius MS, Australitermes MS, Eutermellus, and Mala-
gasitermes MS.

470 FIELDIANA: ZOOLOGY, VOLUME 37

The Australian region includes both tropical and south teniperate
Australia, Tasmania, and New Zealand. For some animals, New
Zealand should probably be considered a separate region, but the
two endemic species of termites show an ancient faunal connection
with Australia, possibly during the Triassic or Jurassic. The termite
fauna of Australia becomes reduced in the temperate areas but the
zones are not sharp and all genera except Stolotermes and Porotermes
clearly are derived from the tropical fauna. Freezing temperatures
are usually local and not prolonged except in higher altitudes and
in Tasmania, and the little competition from other termites has not
prevented a southern extension of the northern tropical groups.

The Papuan region includes New Guinea, Aroe Islands, Moluccas,
Amboina (all east of Weber's Line), and the Pacific islands of
Melanesia, Micronesia, Polynesia, and Juan Fernandez (for con-
venience), but excludes New Zealand, Galapagos, and Cocos. The
single endemic termite on Juan Fernandez is tentatively placed in
Neotermes without clear affinities. The Hawaiian termites all show
derivation from the western oceanic islands and the Indomalayan
region but have probably all been introduced. A significant demar-
cation between an eastern oceanic and a western fauna with con-
tinental affinities occurs east of the Fiji and Samoan Islands.

The Indomalayan region has also been referred to as the Oriental
region. It includes the tropical Asian continent from western India,
southern China just north of Burma, and the southern coast of
China, through the tropical islands of the Pacific from Formosa,
the Philippines, and East Indies to Weber's Line west of the Mo-
luccas. The distribution of the Macrotermitinae most clearly ex-
hibits the limits of this region to the north and east. Timor and
Celebes are included, but Amboina, the Moluccas, and New Guinea
are excluded. The western boundary is the temperate steppes and
deserts of Afghanistan and Iran. Subtropical northern India and
Pakistan south of the Himalayas has an Indomalayan fauna clearly
related to the more southern tropics. Temperate altitudes in the
Himalayas, Assam, and Indochina have a Palaearctic fauna particu-
larly characterized by Archotermopsis, Hodotermopsis, and Relicu-
litermes.

The Palaearctic region includes Africa north of the Tropic of
Cancer and temperate Asia north and west of the Indomalayan
region. It contains the steppes and deserts of Iran, Afghanistan,
and the Near East. Europe and North Africa are clearly Palaearctic,
as are also the temperate highlands of Asia. A local tropical pocket

EMERSON: DISTRIBUTION OF TERMITE GENERA 471

in the Jordan Valley is included in the Palaearctic, although it has
elements of an Ethiopian flora and fauna, including a species of
Angulitermes.

The Ethiopian region includes tropical Africa and offshore islands
south from the Canaries, the Sahara Desert, Sudan, and southern
Arabia. Fernando Po, San Thom£, Socotra, Zanzibar, Pemba, and
Mafia have a typical Ethiopian fauna. The temperate Cape region
of South Africa is included as an extension of the tropical Ethiopian
region. The termites of temperate and subtropical South Africa
are derived from the tropical genera to the north, with the possible
exceptions of Stolotermes, Porotermes, and Microhodotermes.

The Malagasy region includes Madagascar, Europa Island, Co-
mores, Seychelles, Mauritius, Reunion, and, for convenience, Cocos-
Keeling. Cocos-Keeling would not be included in the Malagasy
region for most of its fauna, but a Malagasy species of Prorhinotermes
is found there. These Indian Ocean islands have a related termite
fauna that is distinct from that of the Ethiopian fauna from which
their major elements were derived. Because of both the presence
and absence of certain termite genera, the Malagasy region deserves
separate status. All its native species are endemic, and two mono-
typic endemic genera are known from Madagascar.

The Nearctic region includes temperate and subtropical North
America north from the highlands of Mexico and the lowlands north
of the Tropic of Cancer except the Bahamas and the southern tip
of Florida. A small amount of overlap between tropical and sub-
tropical species occurs in southern Florida and the Florida Keys.
The Bermuda fauna (all probably introduced) is southern Nearctic.
As in the case of the Palaearctic region, the genus Reticulilermes
marks the southern boundary of the Nearctic and extends northward
to about the 49° F. annual isothermal line. In southern Florida,
Reticulilermes extends to Key West.

The Neotropical region includes tropical North America from
the Tropic of Cancer south along the Mexican shores, the southern
tip of Florida from the vicinity of Miami south, Bahamas, West
Indies, Cocos Island, Galapagos Islands, and tropical Central and
South America. Temperate South America has a termite fauna
clearly derived from the Neotropical fauna with the exception of
Porotermes in Chile. The south temperate extensions of otherwise
tropical genera in southern Brazil, Uruguay, Argentina, and Chile
are the reason for including the whole of the area occupied by ter-
mites within the Neotropical region. In general, the northern

472 FIELDIANA: ZOOLOGY, VOLUME 37

faunal border is sharp and coincides with the northern distribution
of Nasutitermes and the southern distribution of Reticulitermes.
The boundary is sharp at- the altitude of 2,300 feet west of the city
of Vera Cruz in Mexico. There is some overlap of Neotropical and
Nearctic species in southern Florida and one species of Neotermes
with a center of distribution in the tropics reaches Seminole County,
Florida, about half the length of the state from the southern tip of
the mainland.

ECOLOGICAL ZOOGEOGRAPHY OF TERMITE GENERA

Termites are now a predominantly tropical order of insects.
Some species readily invade southern subtropical and warm tem-
perate regions such as southern Australia and Tasmania, southern
Africa, and southern South America, but these species, in most in-
stances, are clearly related to those in the northern tropical regions.
The demarcation line between the tropical and subtropical faunas
is much more sharp in the north on all continents under both dry
and humid conditions. A rapid reduction of species occurs at tem-
perate altitudes on tropical mountains, but a few otherwise tropical
genera such as Odontotermes may cross the boundary at high alti-
tudes.

Wholly temperate genera are few in number and all except
Gnathamitermes of southwestern United States and temperate Mexico
belong to the more primitive families below the Termitidae. It is
noteworthy that the surviving relicts of the primitive family Hodo-
termitidae are largely temperate in their distribution and the fossils
indicate Tertiary temperate habitats. In no case are the climatic
areas with cold winters and cool summers occupied by termites.
The 49° F. annual isothermal line of both hemispheres encloses
almost all native species.

Many genera are primarily tropical in distribution but may
have one or a few subtropical or warm temperate species. Such
genera are Procryptotermes, Cryptolermes, Neotermes, Calcaritermes,
Coptotermes, Heterotermes, Anoplotermes, Procornitermes, and Nasu-
titermes. The latter two genera invade the subtropics only in the
southern hemisphere.

Some genera are mainly tropical but have several species in
warm temperate regions, particularly in the northern hemisphere.
These include Kalotermes, Microcerotermes, Eremotermes, Amitermes,
and Tenuirostritermes.

EMERSON: DISTRIBUTION OF TERMITE GENERA 473

The greatest abundance of species and genera occurs in the
humid tropics, particularly in tropical rain-forests. The periodically
dry tropical and temperate grasslands have a less abundant termite
fauna than their more humid forested counterparts in the same
temperature zones. Tropical savannas, however, have a fairly large
number of species and genera that are confined to this biome. Some
typical genera adapted to grasslands, steppes, and deserts are
Paraneotermes (1 species), Hodotermes (2 species), Microhodotermes
(3 species), Anacanthotermes (3 species), Psammotermes (3 species),
Drepanotermes (2 species), Gnathamitermes (4 species), Synacantho-
termes (2 species), Allodontermes (6 species), Procornitermes (6
species), Paracornitermes (4 species), Trinervitermes (60 species),
and Coarctotermes (16 species).

A few genera are confined to arid regions. Anacanthotermes
and Psammotermes are examples. Psammotermes seems to have
maintained this adjustment since Eocene times, when it probably
became established in Madagascar. By a study of the time of
separation, it is possible to determine the ecological adaptations of
genera over long periods of geologic history.

Genera wholly confined to rain-forests are too numerous to list.
Some notable genera with species in both savanna and rain-forests
are Kalotermes, Procryptotermes, Glyptotermes, Heterotermes, Micro-
cerotermes, Amitermes (a typical genus of dry semi-deserts, steppes,
and savannas, but including a few species confined to rain-forests,
such as Amitermes excellens Silvestri in British Guiana and Brazil),
Apicotermes, Termes, Macrolermes, Odontotermes, Microtermes, Syn-
termes, Cornitermes, and Nasutitermes. The different species belong-
ing to each of these genera are confined to a climatic vegetational
type, with few exceptions.

It is obvious from the distribution patterns of termites that
temperature and moisture are the major physical factors that limit
their dispersal. These two factors largely determine the vegetation
types or biomes also, so that a high degree of correlation between
phytogeography and termite zoogeography is apparent. There
seems to be a tendency for termite species to be correlated with the
distribution of plant genera, and for termite genera to be correlated
with the distribution of plant families, while plant species are usually
much more local than are termite species. The reason for this type
of correlation may be that plants are more sensitive to small eco-
logical variations of a greater variety of factors than are termites,
which exert considerable social control over their microclimates

474 FIELDIANA: ZOOLOGY, VOLUME 37

within their nests and burrows so that they can tolerate greater
fluctuations in their extra-social environment.

Another important ecological factor that is often underempha-
sized by biogeographers is competition for limited necessities. The
role of competition in the dispersal of termites is only partially
understood even though its importance seems great. There seems
to be much unexploited food for termites in many habitats with a
rich fauna, so it is possible that competition for colonizing sites and
for nesting sites is greater than for the nutritive resources. This
may mean that, in many cases, competitive factors are greater
between species of termites than between termites and other animals.
Some circumstantial evidence for this hypothesis is to be found in
the restriction to the man-modified habitats of introduced species
of termites. Some twenty species, all belonging to the Kaloter-
mitidae and Rhinotermitidae, have been inadvertently transported
by man in his wood products and in the soil attached to living
plants. When these termites are introduced to regions climatically
similar to the areas from which they came, the species may become
established, often causing great economic damage to human con-
structions. Typically the introduced species are unable to invade
the native environment, but remain confined to the constructions
of man. It seems probable that competition with native species is
a biotic barrier to the spread of introduced species. Islands, which
have a depauperate termite fauna, seem to be more propitious for
the establishment of introduced termite species than are tropical
continental regions with a numerous and varied termite fauna.
Predators, parasites, and other phylogenetically unrelated organisms,
besides competition with related species, may be involved in the
production of biotic barriers. It will be noticed in the following
discussion that biotic factors are invoked to explain aspects of modern
distribution on a par with either geological or climatic factors.

The distribution of genera and higher categories indicates the
more important ecological factors, but many more subtle limitations
to dispersal may be detected by a study of each species. To cite a
rather extreme example, colonies of Termes inquilinus (Emerson) are
found only in the arboreal nests of Constrictoterm.es cavifrons Holm-
gren, and colonies of the closely related Termes fur (Silvestri) are
found only in the nests of Constrictotermes cyphergaster (Silvestri).
Obligatory inquilines are rare among termites, but they indicate the
complexity of ecological adjustment that sometimes determines
distribution. Future ecological zoogeographical studies of species
rather than genera, particularly by investigators with a thorough

EMERSON: DISTRIBUTION OF TERMITE GENERA 475

knowledge of local habitats, will add much detail to the broad
generalizations suggested in this paper and will correct errors that
almost certainly are present in a summary treatment of this type.

THE RELATIONSHIPS OF THE TERMITES OF ZOOGEOGRAPHICAL REGIONS

The generally recognized zoogeographical regions also exhibit
high correlations of their termite faunas. Each has been geographi-
cally or climatically isolated from the others at various times during
its Mesozoic and Tertiary history, and the termite faunas give
qualitative and quantitative indications of these periods of con-
tinuity and isolation.

The total number of termite species in each of the tropical
regions arranged in order of decreasing abundance duplicates the
order of land area from largest to smallest. This correlation is much
more obvious for the Rhinotermitidae and Termitidae than for the
more primitive families composed largely of relicts or semi-relicts.
Abundance of species is also roughly correlated with the variety of
habitats and ecological opportunity offered to termites during their
long history, together with the limitations imposed by interspecies
competitive relations between the termites themselves. This corre-
lation of abundance and area breaks down in the temperate regions
because of the influence of the temperature_ factor. Within the
tropical zone, a continental area with rain-forest, savanna, and
woodland will have more species than an area with a single vegeta-
tional type. The species decrease with increasing aridity and with
the agricultural and social activities of man. Although long-isolated
islands are likely to have endemic native species, small islands usually
have depauperate faunas, even when montane conditions offer a fair
variety of habitats.

In the following discussion of each region, I have attempted to
offer simple hypotheses to account for modern distributions of the
genera and higher categories. The basic data are listed in the tables
(Tables 2-8). For details of distribution and phylogeny, the reader
is referred to papers by Holmgren (1911, 1912), Hare (1937), Ander
(1942), Snyder (1949), Ahmad (1950), Ross (1953), and Emerson
(1925, 1928, 1933, 1942, 1945, 1947, 1950, 1952a, 1952b, 1953). The
suggested hypotheses rest upon circumstantial evidence, often with
data insufficient for much more than speculations. However, it
seems to me that the bulk of information and correlations is fairly
consistent with the hypotheses and that evidence from other anal-
yses, particularly from mammalian paleontology, phylogeny, and

476 FIELDIANA: ZOOLOGY, VOLUME 37

geography, is slowly accumulating to an impressive degree. Doubt-
less much future revision of the present tentative conclusions will
be made in the face of more abundant and more accurate data, but
statements of working hypotheses consistent with known facts form
a base for an advancing science of biogeography. The stated
conclusions are likely to be over-simplifications and some will
probably have to be replaced by more complex explanations. It is
hoped that this compilation of data and theory will stimulate others
to gather relevant information about other groups of organisms,
particularly when taxonomic and phylogenetic information is sound.

Australian region. — The Australian termite fauna is statistically
more closely related to the Ethiopian than to any other, but if the
two relict genera with possible antarctic dispersal are eliminated,
the data then indicate equal relations with the Papuan, Indo-
malayan, and Ethiopian regions. These data suggest that Australia
is a peripheral region with a Mesozoic fauna that probably arrived
from Indomalaya directly or through New Guinea.

Nearly half of the Australian genera belong to the families below
the Termitidae. All of these primitive genera may be presumed to
have originated in the Permian or Mesozoic and a number of relicts
and semi-relicts possibly arose in early or middle Mesozoic. Two,
Stolotermes and Porotermes, possibly reached Australia through a
southern antarctic land mass in Triassic times. Stolotermes is now
found in New Zealand, Tasmania, and eastern Australian temperate
and tropical habitats. Porotermes is in southeastern Australia and
Tasmania. Mastotermes could conceivably go back to the Permian,
but it doubtless came into Australia from the north, where the
family was probably cosmopolitan in the Mesozoic and Tertiary
periods. The single living species of Mastotermes occurs locally in
subterranean and somewhat dry habitats in both east and west
Australia north of the Tropic of Capricorn. It is absent from rain-
forests and areas once covered by rain-forests (Ratcliffe, Gay, and
Greaves, 1952).

The Kalotermitidae of Australia and New Zealand all belong to
widely spread genera that probably entered Australia in the Jurassic
or early Cretaceous from Eurasia through the ancient Papuan con-
nections. None of these genera indicate the region of origin except
Glyptotermes, which possibly originated in the Neotropical region.
Australian Glyptotermes are only found in the somewhat more humid
areas of the east, both north and south of the Tropic of Capricorn.
Of the four genera of Rhinotermitidae, all probably reached Australia
in early Cretaceous times from Indomalaya through the Papuan,

EMERSON: DISTRIBUTION OF TERMITE GENERA 477

and three give indications of having originated in the Indomalayan
region. The place of origin of Heterotermes is obscure. No relatively
primitive genera of Termitidae or its subfamilies occur in Australia.
Four are widespread genera that invaded Australia from the north
via Indomalaya in early Cretaceous. The origin of Microcerotermes
seems to have been in Africa. Nasutitermes surely came from South
America via a tropical Bering bridge. The origin of Amitermes is
obscure. Most species of Amitermes are found in Africa, but the
more primitive relatives are Indomalayan. The origin of Termes is
likewise obscure. Most species are Australian, but a large series of
more primitive relatives are exclusively Ethiopian. Nine specialized
endemics, all Termitidae, presumably arose in Australia in late
Cretaceous or Tertiary times from medium-specialized ancestry
invading from Indomalaya. There is no indication of any arrivals
in Australia since mid-Cretaceous, when the sea barrier at Weber's
Line was probably established. However, a dry climate or compe-
tition may have eliminated some of the genera that came in early
Cretaceous. It is noteworthy that the two most highly specialized
groups of termites, those with asymmetrical snapping mandibles in
the soldier and those with nasute soldiers, have endemic Australian
genera (Paracapritermes, Tumulitermes, Australitermes MS, etc.),
which indicate the dispersion of their immediate ancestral types by
mid-Cretaceous. (

Papuan region. — It might be expected that the Papuan fauna
would be more closely related to the Australian than to any other
region, but actually the relationships are closer to Indomalaya. In
part, this may be due to incomplete collections, but some of the
genera overlapping with the Indomalayan did not reach Australia
or were eliminated after reaching it. These include Prorhinotermes,
Capritermes, Grallatotermes, Hospitalilermes, and possibly Leucopi-
termes MS. Also, none of the endemic genera of Australia are re-
ported from the Papuan region. Some of these may ultimately be
found, but for the present the barrier seems fairly sharp. No endemic
genera are known from the Papuan region, and the statistical data
indicate that it is a peripheral region with a generic fauna originating
elsewhere, either in Indomalaya or reaching the Papuan region
through Indomalaya, with the exception of Rugitermes.

The Kalotermitidae are well represented, and four of the genera
with obscure origins possibly arose in the Jurassic and spread to the
Papuan region during the Jurassic or early Cretaceous. Glyptotermes
is possibly a late Jurassic or early Cretaceous invader through a
northern tropical connection with its Neotropical region of origin.

478 FIELDIANA: ZOOLOGY, VOLUME 37

Rugitermes has reached only the eastern oceanic islands. The genus
probably arose in the Neotropical region in early Tertiary times
and was able to reach the eastern edge of the Papuan region over
the seas by some unknown means. All other Papuan genera, in-
cluding those in the eastern oceanic islands, probably dispersed from
the western Papuan region and Indomalaya. All the genera of the
Rhinotermitidae are clearly extensions from Indomalaya during the
early Cretaceous. All except Heterotermes give numerical evidence
of their Indomalayan origin. All of the Termitidae doubtless came
from Indomalaya during the Cretaceous and probably most origi-
nated in Indomalaya. However, Microcerotermes indicates an Ethio-
pian origin, Nasutitermes came from the Neotropical region over
the Bering bridge, and Termes has an obscure geographical origin,
although its more primitive relatives center in Africa. Neotermes
and Prorhinotermes have the largest number of species in the Papuan
region, but this may well be due to island endemism rather than a
Papuan origin. Some Papuan genera are known only from New
Guinea, Moluccas, and Aroe Islands. These include Heterotermes,
Parrhinotermes, Termes, Capritermes, Grallatotermes, Hospitalitermes,
and Leucopitermes MS. Others may occur from New Guinea and
the Solomons east to Samoa. These include Schedorhinotermes,
Microcerotermes, and Nasutitermes. It is zoogeographically signifi-
cant that a highly specialized physogastric staphylinid termitophile
of the genus Thyreoxenus lives in the nests of Nasutitermes in the
Solomon Islands (Seevers, 1937). These termitophiles do not ac-
company the colonizing pair of termites but invade the mature
colonies of their hosts several years after the founding flight. This
means that the termites that flew over wide water-gaps and managed
to find mates after the flight would tend to leave the specialized
termitophiles behind. This contention is borne out in the case of
the island-hopping species of Nasutitermes in the Lesser and Greater
Antilles. It seems to me that the existence in the Solomons of a
Neotropical termitophile genus like Thyreoxenus in the nests of a
host genus, Nasutitermes, that originated in the Neotropics strongly
suggests tropical land continuities or short water-gaps in the past.
If Thyreoxenus is found in Indomalaya in the future, further pre-
sumptive evidence of Cretaceous land connections or contiguities
will be indicated.

Genera of termites reaching the eastern oceanic islands east of
Samoa from the west include Kalotermes, Procryptotermes, Crypto-
termes, Neotermes, Glyptotermes, and Coptotermes, while Rugitermes
reached the eastern islands from South America. Prorhinotermes

EMERSON: DISTRIBUTION OF TERMITE GENERA 479

has a distribution among the Papuan islands similar to Schedorhino-
termes, but inasmuch as it occurs in Cocos, Central America, and
the West Indies, it is to be expected in the eastern Polynesian
islands.

Indomalayan region. — It is noteworthy that the large tropical
continental regions have a smaller percentage of overlapping genera
and a larger percentage of endemic genera than the temperate or
peripheral and smaller regions (Tables 7, 8). This fact probably
indicates that the larger tropical continental regions are the centers
of origin of a larger proportion of existing genera, particularly
those genera that arose in Cretaceous or Tertiary times. The Indo-
malayan region stands somewhat below the Ethiopian or the Neo-
tropical regions in this respect. Nevertheless it stands high as either
a primary or a secondary center of termite evolution. The Indo-
malayan fauna is statistically closer to the Ethiopian and is almost
as close to the Papuan followed by the Neotropical. There are no
endemic genera of Mastotermitidae, Kalotermitidae, or Hodotermi-
tidae in the Indomalayan, nor is there a significantly larger number
of species of the overlapping genera of these families. However,
these indications of a center of origin tend to disappear with age, so
it is by no means impossible that important groups of termites
originated in this region during the early and middle Mesozoic but
have left no discernible evidence of these events. Whence examine
the data on the Rhinotermitidae and Termitidae, we ao find some
evidence of Indomalayan origins. Two of the seven Rhinotermitid
genera are endemic, and three of the five overlapping genera have
more species in Indomalaya. Of the remaining two genera, Hetero-
termes has an obscure place of origin and Prorhinotermes might
conceivably have originated in the Indomalayan in spite of the
larger number of Papuan species, possibly because of the essential
limitation of this genus to islands and speciation through island
isolation. The only other region with more endemic Rhinotermitids
is the Neotropical with four such genera, but the three in the sub-
family Rhinotermitinae are clearly specialized genera derived from
a stock very close to Schedorhinotermes, a genus that gives clear
indications of an Indomalayan origin. It is true, however, that the
two most primitive genera in the Psammotermitinae are not found
in Indomalaya. The absence of Psammotermes may be due to eco-
logical conditions and one cannot attach too much significance to
the absence of Glossotermes or a related primitive genus. One may
conclude, therefore, that present indications are that the Rhino-
termitidae had its main center of origin and dispersal in the Indo-

480 FIELDIANA: ZOOLOGY, VOLUME 37

malayan, and that this family probably arose from an extinct
ancestral group sharing the generalized and primitive mandibles of
the most primitive Hodotermitids and the ocelli that doubtless
occurred in primitive ancestors of termites and are retained in all
families except the Hodotermitidae.

The most generalized of the subfamilies of the Termitidae seems
to be the Amitermitinae. The evidence indicating a center of origin
is not as clear as in the other Termitid subfamilies but there is suffi-
cient to suggest an Indomalayan origin. Eleven genera, including
seven endemic genera, are found in Indomalaya — more than occur
in any other region. One of the endemics is Protohamitermes,
unfortunately only known from imagoes and workers of one species
from Sarawak. In spite of the paucity of material and information,
this genus shows striking affinity to the mandibular pattern of the
Rhinotermitidae and may be considered the most primitive genus
of the Termitidae. Of the four overlapping genera, Speculitermes
has more species in the Neotropical, Microcerotermes has more
species in the Ethiopian, Eremotermes has an equal number of known
species in the Palaearctic but probably originated in the Indoma-
layan, and Amitermes has the largest number of species (36) in the
Ethiopian, followed by Australia (27) and the Neotropical (11),
while only three occur in the Indomalayan. The more primitive
relatives of Amitermes are Indomalayan genera, so this inconsistency
in order of species abundance is possibly a result of speciation in
response to ecological conditions rather than an indication of the
center of origin and dispersal.

All of the six Indomalayan genera of Termitinae, including the
four endemic genera, are medium-specialized or specialized types.
The subfamily probably originated in the Ethiopian region and
invaded Indomalaya in early Cretaceous times. Termes seems to
have had a secondary center of speciation in Australia, but it possibly
originated in Africa. The highly specialized Capritermes seems to
have originated in the Indomalayan, where 32 species now occur,
and dispersed in the Cretaceous to New Guinea and through Africa
to Madagascar by Eocene times.

Fourteen genera of Nasuti termitinae occur in Indomalaya; nine
are endemic, and all are medium-specialized or specialized groups.
The primitive genera of this subfamily are exclusively Neotropical,
and it seems clear that the remote ancestors of the Indomalayan
genera came from the New World. The only extant genus still
found in the two regions is the tropicopolitan Nasutitermes, which

EMERSON: DISTRIBUTION OF TERMITE GENERA 481

must have dispersed in early Cretaceous over a tropical Bering
bridge. Consistency of species numbers gives evidence of the place
of origin and the direction of ancient dispersal. Corroborative
evidence is also found in the phylogeny and dispersal of the staphy-
linid termitophiles, a thorough study of which is soon to be published
by Dr. Charles Seevers. Somewhat more primitive genera than
Nasuiitermes are endemic in Indomalaya. Hirtitermes and Longi-
peditermes seem to have arisen from an extinct group that arrived
in Indomalaya at about the same time as Nasuiitermes. This stock
radiated to produce Grallalotermes and Hospitalitermes in early
Cretaceous times, when they probably spread to the Papuan region.
The same stock produced the endemic Bulbitermes and Lacessiti-
termes in Indomalaya, Coarctotermes in the Ethiopian and Malagasy
regions by Eocene times, and Constrictotermes in the Neotropics,
possibly from an ancestral genus that dispersed back to the New
World in the late Cretaceous. On a sub-branch of the subfamily,
Havilanditermes is closely related to Indomalayan Nasutitermes, and
Ceylonitermes shows a relationship to the Central American and Ne-
arctic Tenuirostrilermes that possibly arrived in the orient in Tertiary
times, and to Trinervitermes that probably had its origin in the
Ethiopian region when a tropical savanna was established in the Mio-
cene. Trinervitermes has five species in India and Ceylon, and one
questionably assigned species in Indochina. It appears most likely
that Trinervitermes invaded India from Africa in the Miocene or,
more probably, in the Pliocene, through a tropical savanna or steppe
corridor. Four genera of small termites on the Subulitermes branch
of nasutes occur in Indomalaya, all endemic except one that has a
tentatively included species in Amboina. These came from two
stocks from the Neotropics that also reached Australia and Africa,
and one reached Madagascar by Eocene times.

It appears from these considerations that Indomalaya was an
important secondary region of origin and dispersal of an original
Neotropical subfamily, and that several independent invasions of
Indomalaya occurred during early Cretaceous times.

The Indomalayan region has a larger proportion of original dry
and humid forest lands than any other tropical region, and this fact
together with the modification by dense populations of humans has
doubtless influenced its known termite fauna. Collections from still
unmodified habitats such as Borneo are very much needed.

Palaearctic region. — The Palaearctic termite fauna shows its
closest relations to the Ethiopian and next to the Indomalayan.
This indicates the origin of many of the genera in the contiguous

482 FIELDIANA: ZOOLOGY, VOLUME 37

southern tropics. The north temperate regions are characterized by
depauperate termite faunas, as might be expected for animals with
a tropical adjustment. However, a number of interesting relicts
occur in the Palaearctic, and in the case of the Hodotermitidae
these are possibly indicative of a temperate origin of this primitive
family. An alternative possibility is widespread competitive elimi-
nation by higher termites in the tropics. The abundance of species
of termites decreases north of the tropical border, and no records
are known north of 43° N. Lat. in Asia and 48° N. Lat. in Europe.
The Tertiary fossils indicate warm temperate conditions as far north
as 55° Lat., but in no case give evidence of a tropical climate. To
explain the present distribution of termites, it is necessary to pos-
tulate tropical climates as far north as the Bering land bridge at
about 60° Lat. during the Cretaceous and Jurassic periods. It is
hoped that isotope determinations of sea temperatures from Meso-
zoic fossils together with paleobotanical data will soon verify this
conjecture based upon the circumstantial evidence of termite
distribution.

Thirteen genera of termites are recorded from the Palaearctic,
of which over half are from the more primitive families below the
Termitidae. Eleven genera are overlapping and only two are
endemic. The endemics are the two very primitive relicts, Archo-
termopsis of the temperate Himalayas and Hodotermopsis of sub-
tropical altitudes of Indochina and subtropical latitudes of the
Ryu-Kyu Islands. In a number of particulars, including the denti-
tion of the imago mandibles, Archotermopsis is the most archaic
living termite. One fossil species is found in Eocene Baltic amber.
A more primitive genus, Termopsis, is also known from Baltic amber
as well as from other European deposits of Oligocene and Miocene
age. It is probable that the Termopsinae arose in temperate Eurasia,
possibly in early Mesozoic times, and in the Mesozoic reached North
America, where fossil and living derivative genera are known from
the Oligocene of Colorado. Microhodotermes, with two species in
North Africa and the Near East, also probably arose in the sub-
tropical Palaearctic and moved south through the highlands of
Africa to give rise to the single species in subtropical South Africa.
Anacanthotermes is a subtropical desert genus extending through
North Africa to the border of India, with one species in a dry area
of southern India. The genus presumably originated in the southern
Palaearctic steppes and deserts and dispersed eastward to tropical
Indomalaya. The Oligocene and Miocene fossil genus Ulmeriella
seems to belong to the Hodotermitinae with four Eurasian species

EMERSON: DISTRIBUTION OF TERMITE GENERA 483

and one known from the state of Washington. Possibly this extinct
genus also arose in the temperate Palaearctic and reached North
America in the Miocene or earlier.

The species of Palaearctic Kalotermes were probably derived from
the nearest tropical regions, although Kalotermes is found in many
subtropical latitudes over the world and has long had temperate
representatives, as evidenced by several European Tertiary fossils
tentatively assigned to this genus. Neotermes is essentially a tropical
genus but has a single Palaearctic species in the Madeira Islands
and one Indomalayan species that reaches Okinawa in the Ryu-Kyu
Islands near 26° N. Lat. Psammolermes in the North African desert
seems to be derived from Ethiopian ancestors. Coptotermes has one
species in southern Japan that is also known from tropical Formosa
and China and is doubtless descended from Indomalayan species.
Reticulitermes is the only Holarctic genus, and the nearly equal num-
bers of recorded species in Eurasia and in North America do not
indicate the center of origin. Species are known from Eocene Baltic
amber, so we may assume at least an early Tertiary history as a
temperate genus. The closest genus is the somewhat more primitive
Heterotermes with an obscure origin in the tropics, possibly in late
Jurassic or early Cretaceous. All records of Reticulitermes south of
the Tropic of Cancer are at temperate altitudes. Five species of M i-
crocerotermes occurring in the steppes and deserts of North Africa and
southern Asia were all probably derived from an ancient Ethiopian
center. Eremotermes in North Africa and the south Asiatic steppes
and deserts possibly had its origin in Indomalaya. Amitermes has
five species in North Africa and in the steppes and deserts of southern
Asia. For reasons explained elsewhere, it is difficult to state that
the Palaearctic species came from either Africa or India, in spite of
the much larger number of Ethiopian species. The single Palaearctic
species of the Termitinae is found near Jericho in the tropical Jordan
Valley of Palestine. This species is a northern extension of an
Ethiopian genus. Its presence indicates tropical conditions or at
least subtropical conditions in the Jordan Valley during the last
glaciation in Europe, and also indicates a continuous extension of
tropical climate through Egypt at the time of its establishment in
Tertiary or Pleistocene times.

During the Mesozoic and Tertiary, tropical climates must have
connected the Indomalayan and Ethiopian regions through what is
now southern Palaearctic Asia. The large majority of the genera
that moved along this corridor are typically savanna and steppe
types, but a few probably required more humid conditions. Nasu-

484 FIELDIANA: ZOOLOGY, VOLUME 37

titermes probably required forests in the Cretaceous, and the forest
types of Macrotermes were dispersed in mid-Tertiary, possibly in
the Miocene.

Ethiopian region. — As is to be expected, the Ethiopian termite
fauna as a whole is most closely related to the Indomalayan. Tropical
Africa is probably the major center of evolution of termite genera,
although it is not necessarily the most ancient center, because most
of the genera originating or dispersing from an Ethiopian center
belong to two advanced subfamilies of Termitidae, the Termitinae
and the Macrotermitinae, the first originating in the Cretaceous
and the second in post-Eocene Tertiary. The Ethiopian genera of
Kalotermitidae are all widely dispersed genera with only one, Glyp-
totermes, that gives any indication of the region of its origin, which
was possibly in the Neotropics.

The relict species of Stolotermes and Porotermes in temperate
South Africa may indicate an origin in a temperate Antarctic land
mass during the Triassic, as discussed elsewhere. Hodotermes is a
tropical steppe genus that invades the subtropical areas of South
Africa. It is known from only two species confined to Africa, but
all other genera of the Hodotermitidae seem to be adapted to tem-
perate climates and it may be guessed that Hodotermes is descended
from a Mesozoic Palaearctic ancestry or possibly from an extinct
tropical genus. The single species of Microhodotermes in subtropical
South Africa has two related species in the Palaearctic steppes and
deserts and presumably reached South Africa by way of highlands
through central Africa at some unknown geologic period. The single
southeast species of Psammotermes is related to a North African
temperate and tropical desert species and also to a southwestern
Madagascar steppe species, a fact that indicates a dispersion from
Africa at least by Eocene times. The Ethiopian Coptotermes
probably came from a center of origin in Indomalaya during the
Cretaceous. The single Ethiopian species of Heterotermes recorded
from Abyssinia is possibly an introduction from Central America.
Schedorhinotermes seems to have originated in Indomalaya in the
Cretaceous, but it may not have reached tropical Africa until post-
Eocene times, because it is unrecorded from Madagascar.

The Ethiopian Amitermitinae consist of seven genera, four of
which are monotypic endemic genera, of which only one, the rain-
forest Cephalotermes, has a fairly well-recorded distribution. Many
species of Anoplotermes are now known throughout tropical dry and
wet areas and extend into subtropical South Africa. This genus
has more species in the Neotropical region but strangely is unknown

EMERSON: DISTRIBUTION OF TERMITE GENERA 485

from either Indomalaya or Madagascar. Its origin and time of
dispersion must remain speculative, but an Indomalayan highway
during the Cretaceous seems most likely. Microcerotermes has the
largest number of species from Africa and is abundant in both humid
and dry tropical and subtropical climates. The numbers of species
are consistent in their order and indicate a northern dispersal from
Africa during the Cretaceous. Its more remote ancestors, however,
probably came into Africa from Indomalaya. Amitermes also has the
largest number of species in the Ethiopian region and is adjusted
to dry tropical and warm temperate climates. In Africa, the species
are absent from the rain-forests. The origin of the genus was
probably Cretaceous and was possibly Indomalayan, although the
numbers in each region are not consistent and the place of origin
is obscure.

The Termitinae have 34 endemic genera in the Ethiopian region
starting with the relatively primitive Hoplognathotermes. I found
eight new genera recently in the Belgian Congo. Fifteen genera
from Hoplognathotermes to Megagnathotermes have a relatively primi-
tive morphology, while 19 of the genera from Cubitermes to Ortho-
termes are relatively somewhat more specialized but still retain
biting mandibles in the soldier. Genera from both groups may be
either dry savanna or wet forest types and a few genera such as
Apicotermes and Cubitermes are found in a variety of climatic con-
ditions from Senegal to South Africa. It must be presumed that
primitive genera of Termitinae originated before the Cretaceous
dispersal of the medium-specialized genera with snapping mandibles
in the soldier. Some of the existing genera with biting mandibles in
the soldier may have originated in Tertiary times when the African
fauna was isolated, but it is difficult to understand why some rela-
tively primitive forms are not found in other regions, particularly
in Madagascar. The single genus, Spinitermes, in South America
with biting soldier mandibles is a possible exception to this distri-
bution pattern, but it needs more exact phylogenetic study. The
most primitive of those soldiers with symmetrical snapping man-
dibles is Promirotermes, now confined to the Ethiopian region. A
slightly more specialized genus, Termes, is tropicopolitan with the
largest number of species in Australia and otherwise not presenting
a consistent order of abundance. It must be assumed that this
genus spread to Australia and to South America during early Cre-
taceous times along with a stock close to Promirotermes that gave
rise to related genera now endemic to the Neotropical region. The
African and Palestinian Angulitermes also arose from such a stock

486 FIELDIANA: ZOOLOGY, VOLUME 37

or possibly from Termes itself. The genera with asymmetrical
snapping mandibles are also derived from a base close to Promiro-
termes, but the existing genera with the exception of Capritermes
are specialized endemics in every tropical region. Pericapritermes
is the Ethiopian representative of this group.

The fungus-growing termites or Macrotermitinae are clearly of
Ethiopian origin from some extinct group that shared the primitive
characters of both the Macrotermitinae and the Nasutitermitinae
and was itself probably derived from the primitive Amitermitinae.
All ten genera of the Macrotermitinae are Ethiopian, including the
primitive endemic rain-forest Acanihotermes and its close relative,
the rain-forest and savanna Pseudacanthotermes. The genera that
dispersed to the Indomalayan region, probably at the time of the
Miocene extension of grasslands, all have the greater number of
species in Africa. Some species of forest Macrotermes in Indomalaya
give slight evidence of forest continuity. The European fossil from
the Miocene assigned to Macrotermes needs generic verification. The
most specialized genus, Microtermes, a diminutive forest and savanna
genus, became established in Madagascar by some unknown means,
most probably during the post-Eocene Tertiary. The subfamily as
a whole seems to have originated in Oligocene times, subsequent to
the isolation of the Malagasy fauna in the Eocene. It is the only
subfamily of termites with an indicated Tertiary origin.

The Nasutitermitinae have ten Ethiopian genera. All are
medium-specialized or specialized genera and, so far as can be seen,
all have an Indomalayan ancestry in the Cretaceous and a more
remote Neotropical ancestry in the early Cretaceous. The genus
Nasutitermes arose in the Neotropics and moved across the tropical
Bering bridge to the tropical orient and thence to Africa. In Africa
the species remain in the forested areas and ecologically do not over-
lap with Trinervitermes locally. Grallatotermes arose in Indomalaya
and probably reached the east coast of Africa in the Cretaceous or
early Tertiary. Trinervitermes arose in Africa, probably with the
Miocene savannas. It is absent from the rain-forests but is abundant
from the northern tropical desert borders to the subtropical Cape. A
species has recently been reported from Aden in the Arabian penin-
sula. A few species seem to have dispersed to India and Ceylon in
the Miocene, or more probably in the Pliocene, and one that needs
verification is reported from Indochina. Coarctotermes is an African
and Malagasy genus that arose in the African dry environments in
the early Tertiary before the Eocene isolation of the fauna of Mada-
gascar. Its more primitive nasute ancestor came from Indomalaya,

EMERSON: DISTRIBUTION OF TERMITE GENERA 487

where it has several endemic relatives. Its ancestral genus may
have arrived in Africa at the same time as its relative, Grallatotermes,
but Coarctotermes is better adapted to dry conditions and is a more
successful group in Africa. Mimeutermes is a bizarre little genus
reported from Guinea and the Gold Coast in moderately dry climates.
It is the only African genus of the Paracornitermes-Subulitermes
branch with points on the reduced soldier mandibles. Its remote
ancestry came from the Neotropics in Cretaceous times. A some-
what more primitive relative has recently been found in Madagascar,
indicating the extinction or possibly the lack of discovery of other
relatives in Africa. Eutermellus and a group of related undescribed
nasute genera without pointed soldier mandibles are rain-forest
genera on the Subulitermes branch of the Nasutitermitinae. The
ancestors of this group also reached Africa through Indomalaya
from a remote origin in the Neotropical Cretaceous. The nasute
termites do not seem to be quite as successful in Africa as they are
in their Neotropical center of origin, while the Termitinae that
originated in Africa show decreasing success as they moved farther
from their native land. One may guess that these rough negative
correlations of abundance between these two subfamilies may be
the result of competition between specialized termites, with the
more ancient and locally better-adapted species maintaining them-
selves against the later invaders that were less well adapted to the
local conditions at the time of invasion. The nasute termites, which
possess specialized adaptive defense apparatus, are usually more
successful than any other equivalent group of tropical termites,
but in Africa the Termitinae seem to have the ascendancy. With
our meager information on termite competition, however, such
hypotheses are highly speculative.

Malagasy region. — The bulk of the termite fauna of the Malagasy
region shows clear affinity to the Ethiopian fauna and may be
presumed to have originated from Africa in the late Eocene (Beau-
fort, 1951). Moreau (1952) postulates an Eocene connection with
Africa that broke in late Miocene, but the termite and mammalian
faunas appear more consistent with a late Eocene separation. So
far as termites are concerned, there is no need to postulate dry land
connections, but only a faunal exchange over a much shorter water-
gap than now exists. All the termites in native habitats are specifi-
cally distinct, with the exception of one species of Cryptotermes that
was possibly introduced from Africa by early man. This species,
C. havilandi (Sjbstedt), is geographically incompatible with another
species of Cryptotermes in Madagascar soon to be described by Lena

488 FIELDIANA: ZOOLOGY, VOLUME 37

Moszkowski, and possibly had no close competitor in its dry wood
habitat in the comparatively moist regions of the island at the time
of introduction. Not only are all Madagascar species endemic, with
the exception of the introductions, but local endemism occurs in
many isolated islands such as Europa, Mauritius, Reunion, Sey-
chelles, etc. Also there are two endemic genera. Besides these
indications of long isolation from exchange with Africa, there are a
large number of genera in southeastern Africa that are absent from
Madagascar. The sharp faunal demarcation provides the reason
for giving full regional status to the Malagasy region.

No Mastotermitidae or Hodotermitidae are recorded. The Kalo-
termitid genera include about a third of all the species, a proportion
that indicates geographic and ecologic peripheral distribution. All
could have been directly derived from African congeners. A note-
worthy species among the Rhinotermitidae is Psammoterm.es voeltz-
kowi Wasmann, which retains the generic adjustment to arid habitats
characteristic of the genus in Africa. Coptotermes could easily have
come through Africa from the Orient. The single species of Prorhino-
termes is an exception to the usual derivation from Africa. It
probably arrived in Madagascar via the sea, and all the Indian
Ocean island records seem to be referable to a single species except
for an endemic species in Ceylon. P. canalifrons (Sjostedt) has
been found on Madagascar, Anjouan, Moroni, Mah£, Long Island,
Preslin, Amiranten, Poivre, Aldabra, Mauritius, Chagos Archipelago,
and Cocos-Keeling. This wide range for a single species indicates
continual or recent interbreeding, so that the sea seems to be no
barrier to dispersal in this case. No species of this genus is known
from Africa.

Nineteen species of Microcerotermes are now known from the
Malagasy, mostly from Madagascar itself. The recently described
genus Gibbotermes Cachan (1951) is not generically distinguishable
from Microcerotermes. The largest number of species of this genus
is recorded from the Ethiopian region and many are found in
South Africa. A single new species of Amitermes is now known,
adding Madagascar to the distribution of this widespread genus
with its adaptations to arid conditions. No primitive Termitinae
are known, but Termes has a single Madagascar species and Quasi-
termes is a specialized endemic genus with somewhat asymmetrical
snapping mandibles showing affinities to endemic genera in Australia,
Indomalaya, and the Neotropical region. It may be presumed that
the ancestral stock in Africa is now extinct or has been missed by col-
lectors. Capritermes with two species is an example of peripheral oc-

EMERSON: DISTRIBUTION OF TERMITE GENERA 489

currence of a genus otherwise known from many Indomalayan species
and a single species in New Guinea. Capritermes probably evolved
in the tropical orient in the Cretaceous period and spread in two
directions, with one branch moving through Africa to Madagascar
in the Cretaceous or early Tertiary but later becoming extinct in
Africa. Four or five species of Microtermes are found in Madagascar.
This zoogeographical problem is discussed under the section on
anomalies. Nasutitermes has ten species, indicating an early suc-
cessful invasion through Africa of this tropicopolitan genus. Coarc-
totermes with five species is otherwise found only in the Ethiopian
savannas and steppes. A new genus, Malagasitermes MS, is endemic
to Madagascar and is the only genus from the Subulitermes branch
of the Nasutitermitinae known from Malagasy. Its closest relative
is also a new genus, Leucopitermes MS, from Indomalaya and also
possibly from Amboina. A more remote relative is Mimeutermes
in Africa.

It seems to me that the Malagasy termite fauna is a segment of
the Eocene Ethiopian fauna with subsequent Tertiary evolution
of the endemic species and two endemic genera. The isolated region
during middle and late Tertiary received Microtermes and Prorhino-
termes over sea barriers and, since the establishment of man, has
acquired several introduced species. This postulated history of the
termite fauna fits well with the mammalian pattern, which also
indicates the isolation of an Eocene fauna with some sporadic
entrance of a few species across the strait and, of course, recent
introductions by man.

Nearctic region. — With the notable exception of Reliculitermes
and the relict Zootermopsis, the termite species and genera seem to
have their closest affinities to those of the Neotropical region. One
may postulate a tropical fauna that was extensive over North
America during the Mesozoic but receded southward into the tropics
during the whole of the Tertiary. The Bering land bridge to Eurasia
was the major route of dispersion in both directions during the
Mesozoic, but the cool temperate climate precluded termite spread
throughout the Tertiary for all except the temperate-adjusted
genera. Even the temperate genera may have dispersed in Mesozoic
times, but some could have moved over warm temperate highways
in the Tertiary. There is no indication that any termite used a
north Atlantic bridge during the Tertiary.

The number of species and genera diminishes rapidly with de-
crease in the annual temperature. A few termites reach southern

490 FIELDIANA: ZOOLOGY, VOLUME 37

Canada in Ontario and British Columbia. Most of the species are
found only in the warm temperate or subtropical habitats, and a
few genera may be confined to either the mesic Gulf states or to
the dry southwest.

No Mastotermitidae are now living in the Nearctic, but Masto-
termes is recorded from the Eocene of Tennessee. The only Kalo-
termitid genus with more than one species is Kalotermes, a genus
that is either zoogeographically or ecologically peripheral in its
present distribution. It may be successful in the warm portions of
the Nearctic region because of the lack of competition to which it
would succumb in the richer tropics. Procryptotermes has a single
southwest species, the only temperate species of the genus. Crypto-
termes has a single Florida species, also the only native temperate
species of the genus. Paraneotermes is a monotypic endemic genus
of Mexico and southwestern United States, but it seems to be con-
fined to damp wood even in the dry semi-deserts. It is descended
from Kalotermes. Neotermes, also descended from Kalotermes, has
one species in middle Florida and southward into the West Indies.
Calcaritermes has one Florida species probably dispersed from Central
America along the Gulf coast during a warm interglacial period,
possibly the last interglacial. Zootermopsis is an endemic Nearctic
genus with three species, two in coastal and montane Pacific Coast
states and Nevada, and one in Arizona in damp logs and large
stumps along watercourses. The genus is descended from a Palae-
arctic ancestry represented by relicts. An allied genus, Parotermes,
is known from the Oligocene of Colorado in strata that also preserved
a termite wing tentatively assigned to Zootermopsis. These termites
could have reached North America in Tertiary temperate climates,
but their primitive morphology and disjunct distribution make a
Mesozoic dispersal a possibility. The only other North American
Hodotermitid is a Miocene wing from Washington state assigned to
Ulmeriella. Two genera of Rhinotermitidae are Nearctic. Helero-
termes has two species in the southwestern United States and Mexico
in the subsoils and dead plants of arid habitats. Reticulitermes is a
Holarctic genus with six North American species distributed ac-
cording to soil temperature and moisture through mesic forests,
prairies, and semi-deserts. The time and place of origin of the genus
are obscure, but fossils are known from the Eocene of Europe and
the Oligocene of Colorado. Reticulitermes overlaps into the Neo-
tropical region only at the southern tip of Florida and along the
keys to Key West. In Mexico the distribution is sharply limited
at the border of the Neotropical. The northern limits of the genus

EMERSON: DISTRIBUTION OF TERMITE GENERA 491

coincide fairly well with the 49° F. annual isothermal line. Anop-
lotermes has one Nearctic species as far north as southern Texas.
Amitermes has several species in the arid grasslands and semi-deserts
of the southwest and Mexico. Gnathamitermes, with four species in
the arid southwest and Mexico, is an indigenous genus, probably
an offshoot from Amitermes. Tenuirostritermes is the only nasute
genus, probably originating in the Neotropics where most of its
species still occur in Central America.

Neotropical region. — Statistically the Neotropical termite fauna
is most closely related to that of the Ethiopian. With the removal
of Porotermes and Rugiiermes for reasons discussed elsewhere, the
closest affinity would be Indomalayan followed by the Ethiopian
and Australian. Next to the Ethiopian region, the Neotropical has
the largest number of termite genera and species and has been a
major center of origin and dispersion. All indications point to a
tropical climate with both humid and dry habitats well back into
the Mesozoic. The mammalian data and history indicate an Eocene
separation from North America that lasted into mid-Pliocene when
the Panamanian isthmus arose and allowed considerable movement
of animals both north and south. The termite fauna rather dia-
grammatically fits into the historical pattern of mammalian origin
and dispersion as outlined by Simpson, with two major exceptions.
The termites typically are tropical insects, while many groups of
mammals had a temperate origin and dispersal. And even without
corroborative fossils, many termite genera give indications of having
arisen and dispersed in the Mesozoic earlier than or contempora-
neous with early marsupial evolution. Only the more specialized
genera of termites show geographic patterns similar to those of
tropical families of placental mammals that arose in the early
Tertiary.

No Mastotermitidae are now known from the Neotropical region,
but fossil wings from the Pliocene of Minas Gerais in Brazil appear
to represent a new primitive genus of this family. Eight of the nine
genera of Kalotermitidae are known from the Neotropical region
and more than twice as many species are recorded as from any
other region. These data do not positively prove the origin of the
family in this region, but they do indicate a secondary if not a
primary center of origin. Kalotermes has 15 species, all found on
the shores or offshore islands, a fact that indicates competitive
elimination of this genus in regions of high competition. An illus-
tration of such competition is afforded by the occurrence of Kalo-
termes in the mangrove swamps of Panama and its absence in the

492 FIELDIANA: ZOOLOGY, VOLUME 37

rain-forests of the isthmus where more specialized Kalotermitidae
occur. The independence of these dry-wood termites from the soil
also reduces their competition with the soil termites of the Rhino-
termitidae and Termitidae. Procryptotermes needs further study-
before geographic generalizations can be made with confidence.
Cryptotermes, in spite of its specialized phragmotic soldier, shows an
ecological peripheral distribution in South and Central America,
but seems to succeed a little better than Kalotermes in highly com-
petitive regions. In Panama, the common introduced Cryptotermes
dudleyi Banks has not invaded the native habitats since 1890.
Possibly the native Cryptotermes create a biotic barrier. Neotermes
is a successful damp wood genus in interior as well as peripheral
areas. Rugitermes probably arose in the Neotropical region but has
managed to establish two endemic species in the eastern oceanic
islands of the Papuan region. Rugitermes is largely confined to
subtropical and tropical South America with an extension north to
Costa Rica, a distribution that indicates a post-Eocene origin in
South America and a sharply limited extension northward at the
time of the Pliocene emergence of the isthmus of Panama. This
limitation is not climatic and is more likely to be biotic in its nature.
Eucryptotermes is a highly specialized monotypic endemic genus
about which little is known beyond its occurrence in Santa Catarina
and its morphological relations with a Neotermes-like base. Its
phragmotic soldier is convergent with that of Cryptotermes and
Glyptotermes. Glyptotermes is a tropicopolitan genus, but the largest
number of species is Neotropical, followed in consistent order of
numbers with other regions along the lines of dispersion. This fact
may imply a somewhat later origin than the more primitive Kalo-
termitidae that do not show a significant order of abundance of
species. A late Jurassic or early Cretaceous dispersal would be the
most plausible. Glyptotermes occurs in subtropical Brazil and Ar-
gentina but is mainly tropical. Calcaritermes centers in Central
America with one species in Nearctic Florida and some in northern
South America. Recently one new species has been found on Una
Grande near Rio de Janeiro, casting some doubt on the hypothesis
that the genus arose in Central America in post-Eocene times and
dispersed into South America in the late Pliocene.

The only Hodotermitid in the region is the Chilean temperate
Porotermes, with a possible dispersal through Triassic Antarctica.

Glossotermes is tentatively placed at the beginning of the Rhino-
termitidae, but its origin and affinities are rather obscure. It is
known from a single soldier from the British Guiana rain-forest.

EMERSON: DISTRIBUTION OF TERMITE GENERA 493

Coptotermes seems to have come from Indomalaya, possibly in the
Cretaceous. Heterotermes does not give evidence of its phylogenetic
or geographic origin, but its dispersion also may have been Creta-
ceous or possibly Jurassic. Prorhinotermes is found fairly close to
the sea in Cocos, Central America, and in tropical Florida and
Cuba. The origin seems to have been in the western Pacific and
dispersal has followed sea routes rather than land routes. Possibly
it moved in floating logs through the early Pliocene seas connecting
the Atlantic and Pacific. Rhinotermes, Dolichorhinotermes, and
Acorhinotermes are a related series of genera derived from an Indo-
malayan stock close to Schedorhinotermes. The direction of the
ancestral dispersion seems to have been from the Old World to the
New World over the Cretaceous Bering bridge. All three genera
are South American except for the extension of Dolichorhinotermes
into Panama and the island-hopping of Rhinotermes marginalis (L.)
from South America up through the Lesser Antilles to Haiti over
water-gaps no greater than ninety miles and possibly less during the
Pleistocene glaciation. The patterns suggest a post-Eocene origin
of the genera of Rhinotermitinae in South America.

The Amitermitinae are represented by several medium-specialized
or specialized genera. Speculitermes is a soldierless termite with a
single recorded species in India and six in the Neotropical rain-forests
and savannas. The phylogeny and geographical origin of this group
are not well understood. Speculitermes may be the ancestral genus
of Anoplotermes. The latter genus has a very large number of species
in South and Central America. Anoplotermes has a most peculiar
distribution, with a single Nearctic species and a large Ethiopian
fauna. By its absence from Indomalaya and Madagascar, it is
unique among genera not relicts or semi-relicts that overlap between
Africa and South America. Any hypothesis as to its time or place
of origin must remain speculative until more information has been
gathered. Hoplotermes is a monotypic Mexican genus with its
closest affinity to the somewhat more primitive Indomalayan Eury-
termes. Microcerotermes indicates by the number of species that it
had a Cretaceous origin in Africa and dispersed through Indomalaya
to the New World over the Bering bridge. Cylindrotermes is an
endemic rain-forest genus of South America with a Pliocene extension
to Panama. It appears to have evolved from a stock related to the
Indomalayan Prohamitermes and the Ethiopian Cephalotermes. Ami-
termes has several Neotropical species, most of them found in the dry
savannas and deserts. However, two species are common in the
rain-forests, one in Panama and the other in Guiana and Brazil.

494 FIELDIANA: ZOOLOGY, VOLUME 37

The rain-shedding structures on trees constructed by the latter
species are remarkable behavior adaptations. Amitermes dispersed
in the Cretaceous, but its place of origin is obscured by its ecological
specialization. One species from southern South America has been
erroneously placed in Synhamitermes, but needs further study before
it is given a generic assignment.

Eleven genera of Termitinae occur in the Neotropics, eight of
which are either medium-specialized or specialized genera. Spini-
termes is the only genus with typical biting mandibles in the
soldier that is found outside of the Ethiopian region. The imago
mandibles are quite specialized and appear to show relationship to
the African Ophiotermes and Tuberculitermes, but this similarity may
be convergent. The imago mandibles of Orthognathotermes are more
primitive than those of Spinitermes, but the soldier has a peculiar type
of partially biting and partially snapping mandibles. The endemic
Genuotermes has a slightly more primitive soldier than Orthognatho-
termes, but its recently discovered worker mandibles are more
specialized. The phylogeny of these Neotropical genera needs to be
carefully studied before zoogeographical hypotheses can be made.
The group is remotely related to the relatively primitive Ethiopian
Termitinae, but so far no intermediates are known that could furnish
a clue to the time of origin and dispersal. Spinitermes and Genuo-
termes are confined to tropical South America, while Orthognatho-
termes reaches Panama to the north and the subtropical savannas of
Brazil and Paraguay to the south. This suggests that the ancestors
reached South America before the Eocene isolation and that the
three genera developed their modern generic distinctions during the
post-Eocene Tertiary. The Neotropical genera Dentispicotermes,
Spicotermes, Termes, Crepititermes, and Cavitermes have symmetrical
snapping mandibles in the soldier and seem to be a related group of
genera. All but Termes are endemic. From the imago mandible
of Spicotermes, it may be presumed that Spicotermes and the slightly
more primitive Dentispicotermes are both early offshoots of a basal
group related to the African Promirotermes and that the tropico-
politan Termes is somewhat more specialized. We may postulate,
therefore, that a somewhat more primitive group of termites than
Termes with symmetrical snapping mandibles in the soldier dispersed
from Africa at about the same time during the early Cretaceous and
reached South America well before the Eocene separation from
North America. Termes is the only related genus now found in the
tropical regions outside of the Ethiopian and Neotropical, but those
genera with asymmetrical snapping mandibles evolved from the

EMERSON: DISTRIBUTION OF TERMITE GENERA 495

same primitive stock related to Promirotermes and have established
endemic genera in every tropical region. Cornicapritermes is an
endemic genus with a frontal projection, a frontal gland, and strongly
asymmetrical mandibles in the soldier. Neocapritermes and Plani-
capritermes, without either a frontal projection or a gland, I now
consider to be the end product of the evolution of the Termitinae.
Both Ahmad (1950) and I (Emerson, 1950) placed them at the be-
ginning of the Termitinae because of their seemingly primitive
imago-worker mandibles and in spite of the soldier specializations,
but it would now appear that the so-called primitive imago-worker
mandibles are convergent to a type somewhat similar to Euhami-
termes and Cephalotermes in different respects. The Ethiopian
Pericapritermes indicates the reduction in size of the apical tooth
also characteristic of Neocapritermes, but the sharp angle between
the first and second marginal teeth of the right mandible is not
found in any other Termitinae.

The Nasutitermitinae originated in the Neotropical region by
early Cretaceous times and had already evolved the specialized
nasute soldier by the early or the mid-Cretaceous when nasutes
with a diphyletic reduction of the mandibles dispersed around the
tropical world. The ancestry of the Nasutitermitinae is not very
clear because of the extinction of the groups that connect this sub-
family with the other Termitidae. I gave evidence (Emerson, 1945)
for an ancestral relationship between the Nasutitermitinae and
Macrotermitinae and guessed that this primitive stock emerged
from the primitive Amitermitinae. This phylogeny is speculative.
After the origin of the Nasutitermitinae, however, the phylogenetic
and geographic history seem to be correlated. The most primitive
existing genus is Syntermes, with morphological indications of rela-
tionship to Procornitermes. From a base related to Procornitermes,
two branches split, one leading through stock giving rise to the
mandibulate genera of Cornitermes and Rhynchotermes on to nasute
genera of the Nasutitermes branch. The other branch split off from
a Procornitermes-\ike base and gave rise to the mandibulate genera,
Paracornitermes, Labiotermes, Armitermes, and Curvitermes. Nasute
genera emerged from this branch, the most primitive of which is
Angularitermes, followed by a number of genera found around the
tropical world. It is significant that the eight relatively primitive
mandibulate genera in both branches are wholly Neotropical. All
except Cornitermes, Rhynchotermes, and Armitermes are wholly South
American. Cornitermes reaches north to Costa Rica, and Rhyncho-
termes and Armitermes both reach Honduras. The Central American

496 FIELDIANA: ZOOLOGY, VOLUME 37

extensions are probably Pliocene or Pleistocene dispersions from
continental South America. All these genera are Tertiary relatives
of a basic Cretaceous mandibulate stock, and the series indicates
gradations in the double origin of the nasute soldier type in early
Cretaceous South America. Nasutitermes is representative of the
genera that spread throughout the tropical world before the mid-
Cretaceous isolation of Australia. It may be presumed that the
dispersion from the New World to the Old was across the Bering land
bridge under favorable moist tropical climates. The failure of the
relatively primitive mandibulate genera to move to the Old World
under the same favorable climatic conditions may be explained by
biotic barriers that could be penetrated only by the highly defended
nasute genera. In all probability several nasute genera moved at
the same time as Nasutitermes, because endemic genera in Indo-
malaya, somewhat more primitive than Nasutitermes, presumably
arose from a nasute base that is now extinct, and endemic nasute
genera of the Subulitermes branch, both with points and without
points on the vestigial mandibles, arose from extinct groups with a
South American origin. It is noteworthy that the order of abundance
of species in Nasutitermes is entirely consistent with this hypothesis.
One species of Nasutitermes extends from northern South America
through the Lesser Antilles to the Greater Antilles. Another species
of Nasutitermes occurs in northern South America, Central America,
and east of Yucatan through the Greater Antilles to the Virgin
Islands. The route of dispersal to the West Indies may have been
through Yucatan in this case, but the Lesser Antilles seem to have
been utilized for island-hopping dispersion to a greater degree by
termites. There is a tendency for these species of Nasutitermes to
leave their specialized termitophiles behind when they fly over wide
water-gaps, but more thorough collecting is necessary before strongly
supported conclusions can be drawn. Six endemic nasute genera
on the Nasutitermes branch are now Neotropical and all except
Constrictotermes presumably originated in the New World. Rotundi-
termes and Diversitermes are South American derivatives of a Nasu-
titermes-\ike base. Diversitermes occurs in tropical and south tem-
perate savannas. Velocitermes is a close relative of Diversitermes,
with both tropical savanna and rain-forest species. It extends to
Panama in Central America and has two endemic species in Haiti,
an indication of a post-Eocene origin and Tertiary dispersion by
way of the Lesser Antilles. Tenuirostritermes seems to have a center
of origin and dispersion in Central America. A few species have
extended north into Nearctic Mexico, Texas, and Arizona. One

EMERSON: DISTRIBUTION OF TERMITE GENERA 497

species in Venezuela may indicate a southern extension in the late
Pliocene. Ceylonitermes in Ceylon seems to be a fairly close relative
of Tenuirostritermes and may indicate a possible Tertiary extension
of a temperate-adapted genus. Parvitermes is a small genus confined
to the West Indies except for a tentatively included species from
Bolivia. Obtusitermes is a small genus in the forests of Panama,
Venezuela, Trinidad, and Cuba. Constrictotermes has slightly more
primitive relatives in the Indomalayan region, where much secondary
evolution and dispersion of nasute genera occurred. It is a possibility
that the ancestor of Constrictotermes moved into South America
from Indomalaya in late Cretaceous times after the isolation of
Australia in mid-Cretaceous and before the Eocene isolation of South
America. The present meager evidence for this hypothesis may be
supplemented in time by a study of the highly specialized termito-
philes known to inhabit the nests of these related genera.

Angularitermes is a primitive nasute now found in the rain-forests
of northern South America and is close to the base that presumably
dispersed around the world in the Cretaceous and gave rise to many
endemic genera in the tropical regions. In South America this
stock gave rise to several endemic genera, including Subulitermes
and Convexitermes, none of which are found north of Panama. These
genera may be considered post-Eocene derivatives of a basic early
Cretaceous stock with a Neotropical center of origin and dispersal.

The small monotypic and endemic subfamily of Serritermitinae
is confined to South America. It has a peculiar combination of
primitive characters such as large wing stubs in the imago and large
but unpigmented eyes in the soldier, together with such specialized
characters as the bizarre elongated and serrate soldier mandibles
and the extremely modified imago mandibles. The origin and rela-
tionships of this species must remain obscure until more data are
discovered.

ANOMALIES OF TERMITE DISTRIBUTION

Our knowledge of historical geology, paleoecology, taxonomy,
phylogeny, and evolutionary processes all seems to provide evidence
for a reasonable interpretation of the causes of contemporary world
distribution of termites as outlined in the preceding pages. However,
there are numerous instances that do not fit precisely into the
expected geographic order. These cases of anomalous distribution
are discussed below, together with possible explanations for the
discrepancies. The simplest explanations of the known facts may
be over-simplifications. The rule of parsimony may actually tend to

498 FIELDIANA: ZOOLOGY, VOLUME 37

obscure complicated events that determined the present distribution
of termites. It may be expected that the biogeography, paleon-
tology, and ecology of many other organisms will ultimately be so
correlated as to offer better interpretations of termite zoogeography.

Mastotermes: The lone surviving relict of this genus and the
family now is found only in tropical Australia. Australia is a con-
tinent containing numerous relicts that cannot compete against
more advanced animals, but there is a fairly abundant and advanced
competitive termite fauna in the area occupied by Mastotermes.
There is no simple explanation for the extinction of the Tertiary
representatives of the family over the world and the persistence of
the single relict in tropical Australia.

Procryptolermes: This genus is cosmopolitan, except for the ab-
sence of any records from the Palaearctic and Indomalayan regions.
Only one species occurs in the Nearctic, so the genus may have been
absent from the Palaearctic throughout its history. However, it
should have been present in the Indomalayan region during Mesozoic
and Tertiary times. Possibly it will be found there ultimately.

Rugitermes: This genus probably evolved from a Neotermes-like
stock in the Neotropical region during the post-Eocene Tertiary.
The occurrence of one endemic species in the Marquesas and Society
Islands and another in the Cook Islands indicates a westward
dispersion to these eastern oceanic islands that is not duplicated by
any other termite genus. Most if not all other native termites in
these islands originated in the western Papuan area. Kalotermi-
tidae could be more easily carried in the branches of a floating tree
than could the soil-dependent termites of the higher families.

Calcaritermes: The center of dispersal of this genus is Central
America. It may be presumed to have originated from Glyptotermes
in this area during the Tertiary separation of South and Central
America. During late Pliocene the genus could have invaded
northern South America, where it is now found. One new species,
however, has been discovered on Ilha Grande off the coast near Rio
de Janeiro, Brazil, a locality far south of the expected range, if the
hypothesis is correct. The single Nearctic species in the northern
half of Florida probably arrived from tropical Central America
during a warm interglacial period along the Gulf coast. Florida is
known to have been broken into low islands or submerged during
parts of the Pleistocene.

Slolotermes and Porotermes: These two relict genera belong to
different subfamilies and are not closely related, but they both

EMERSON: DISTRIBUTION OF TERMITE GENERA 499

represent about the same stage in evolution within the primitive
family of Hodotermitidae. Their occurrence at the southern tem-
perate extremities of the continental land masses may be explained
by the extinction of congeners dispersed through northern connec-
tions, but if so, one might expect some north temperate relicts. All
other genera of the Hodotermitidae clearly show north temperate
distribution or relationships. One wonders whether these genera in
the south temperate areas might be relicts of an Antarctic fauna.
Antarctic coal of Mesozoic age is reported. Porotermes and Stolo-
termes, because of their primitive structure, possibly could have
existed in Triassic times. Romer (1952) gives evidence from fossil
reptiles for his belief "in the existence of a southern inter-continental
connection between South America and South Africa in the Triassic."
The termites from such a postulated land mass also reached Aus-
tralia, Tasmania, and New Zealand. The hypothesis of a northern
dispersal is supported by P. J. Darlington, Jr. (personal communi-
cation), who cites the instance of Chiasognathinae (stag beetles) that
now occur in Australia, New Zealand, southern South America, and
South Africa. There are none in the north at present, but there is
one in Baltic amber. Ander (1942) also discusses other patterns
that parallel Stolotermes and Porotermes.

Hodotermitinae : One might expect relicts of this subfamily to
occur in Australia, but none are reported. When any group is
reduced to a few species, discontinuities and sporadic occurrence
result from extinctions in regions of earlier occupancy. The smaller
the number of surviving species, the more likely there will be unex-
plained inconsistencies. There is no record of any existing species
of this subfamily in the New World, but a Miocene wing of Ulmeriella
is known from the state of Washington, an indication of a more
widespread distribution of the group during the Tertiary.

Heterotermes: This genus is presumed to have arisen before the
Cretaceous separation of Australia from Indomalaya. One might
expect some records from central and southern Africa and possibly
also from Madagascar. The exclusion of this genus from the major
portion of the Ethiopian region seems difficult to explain, except by
competition with an ecological equivalent. Psammotermes is mu-
tually exclusive but is restricted to dry areas, while Heterotermes
may live in rain-forests. Heterotermes has been introduced into
Mauritius, Madagascar, and St. Helena, where it succeeds in the
climates of these islands. The single record from Abyssinia may
also be an introduction. The Galapagos records belong to a common
Central American species that may have been introduced.

500 FIELDIANA: ZOOLOGY, VOLUME 37

Prorhinotermes: The extensive island distribution of this genus
and its comparative rarity on continents seem to demand a distri-
bution by means of floating logs or coconuts during Tertiary times
or earlier and its exclusion by biotic barriers from richly faunated
continents. There is need for study of physiological tolerance to
marine transportation — a means to dispersal that seems unavailable
to practically all termites. Contemporary dispersal over seas is
indicated by the existence of some species in fairly widely separated
islands such as many islands in the Indian Ocean, but endemic
species on single islands or in small island groups indicate an earlier
dispersal and speciation before the advent of man. A hiatus of
records between the Samoan Islands and Cocos Island may be
filled with more complete collections.

Parrhinotermes: This genus, with a monomorphic soldier, is more
primitive than the Old World genus, Schedorhinotermes, with a
dimorphic soldier, which in turn antedated the New World Rhino-
termitinae with reduced mandibles in the minor soldier. Parrhino-
termes probably arose in Indomalaya and dispersed during the Cre-
taceous to New Guinea and Queensland. It must have been present
at the time of the origin and dispersal of Schedorhinotermes, but it
did not spread so far. We may presume biotic barriers that did not
prevent the better-defended Schedorhinotermes from extending its
range.

Schedorhinotermes: The center of origin of this genus was Indo-
malaya, with a Cretaceous dispersal eastward to the Papuan region,
including Palau and Santa Cruz, and south to Australia. The
westward movement to the Ethiopian region, however, seems to
have occurred later than the Eocene. Otherwise why is the genus
absent in Madagascar? An alternative explanation might be the
Tertiary or recent extinction of the genus in Madagascar for unknown
reasons. An ancestral relative did move to the New World during
the Cretaceous to give rise to the Neotropical Rhinotermitinae
during the post-Eocene isolation of South America.

Anoplotermes: This genus shares the loss of the soldier caste
with the somewhat more primitive Speculitermes. The large number
of species in the Neotropical and Ethiopian regions, and the lack of
any record in Indomalaya, where one species of Speculitermes is
known, present a unique pattern of distribution. These two genera
might be expected to have arisen in the Indomalayan region in the
late Cretaceous from more primitive Amitermitinae still represented
in the tropical orient. It is predicted that Anoplotermes will be

EMERSON: DISTRIBUTION OF TERMITE GENERA 501

found in Indomalaya with more concentrated collecting, but its
rarity or possible absence in this region seems to have no reasonable
explanation, and its absence from Madagascar is also difficult to
understand. It is a highly successful genus in the variety of habitats
offered in both the Neotropical and Ethiopian regions.

Amitermes: Except for the absence of any record from the Papuan
region, this genus is cosmopolitan. Its future discovery in the
savannas of southern New Guinea may be predicted.

Angulitermes: The occurrence of a new species of this otherwise
Ethiopian genus in a tropical pocket near Jericho in the Jordan
Valley in Palestine indicates a past connection of a tropical or sub-
tropical climate through Egypt to the Sudan. Such a climate could
have existed during the last interglacial warm period or earlier, but
the maintenance of subtropical or tropical conditions in the Jordan
Valley during the last glaciation is somewhat difficult to believe.

Capritermes: This genus has 32 species in Indomalaya, one species
in New Guinea, and two species in Madagascar. Its absence from
the Ethiopian region presents a problem. It may be postulated
that Capritermes was present in Africa in early Tertiary or late
Mesozoic times and later became extinct, possibly because of com-
petition with a related genus such as Pericapr Hermes.

Microtermes: All indications point to the origin of the Macro-
termitinae in the Ethiopian region later than the Eocene. No
explanation seems adequate to account for the existence of five
endemic species of the most specialized genus, Microtermes, in
Madagascar. It can only be supposed that some sort of unusual
dispersal occurred during the Tertiary, across a water barrier that
effectively barred all other termites. This problem is the most
baffling of all the cases of anomalous distribution.

Coarctotermes: This savanna and steppe genus is found in Africa
and Madagascar — an indication of an Eocene dispersal. Why is it
not also found in India, to which it could have moved with Tri-
nervitermes? It seems necessary to postulate an unknown biotic
barrier. Several related endemics in Indomalaya may have formed
a barrier.

Grallatotermes: If this genus originated in the Cretaceous in
Indomalaya and reached Mafia Island and Portuguese East Africa
by Eocene times, why is it not found in Madagascar? Possibly it
will be discovered in more complete collections.

Malagasitermes MS: This new genus includes a single species,
"Eutermes" Milloti Cachan (1949, pp. 189, 248). It belongs to the

502 FIELDIANA: ZOOLOGY, VOLUME 37

Paracornitermes-Subulitermes-Eutermellus branch of the Nasutiter-
mitinae. It has a point still present on the soldier mandible in some
cases. Its presence in Madagascar tends to substantiate the theory
of the origin of this large branch in the Neotropical region with a
Cretaceous dispersal westward to all tropical regions. The closest
allied genus, however, is Leucopitermes MS, erected to include
"Subulitermes" leucops Holmgren (1914, p. 252), from Sumatra,
Borneo, and Java. This genus also tentatively includes "Subuli-
termes" undecimus Kemner (1931, pp. 34, 36, 37), from Amboina.
The African genus Mimeutermes belongs to this group of genera but
is much more distant from Malagasitermes than is Leucopitermes.
We again note (see Capritermes above) a closer affinity between the
Indomalayan and Malagasy termites than the more obvious and
more easily explained affinity between the Ethiopian and Malagasy
termites on the one hand and the Ethiopian and Indomalayan on
the other. The simplest hypothesis would seem to be the extinction
or lack of discovery of the intermediate relatives in Africa. The
hypothesis of a land connection (Lemuria) across the Indian Ocean
does not seem to rest upon sufficient evidence.

The Malagasy Fauna: This fauna of 17 genera, two of which
are endemic, and 76 wholly endemic species shows closest affinity to
the Ethiopian fauna, followed by the Indomalayan. It is somewhat
strange that the presumed Eocene separation of the Malagasy and
African faunas did not include relict primitive genera of termites.
The subtropical Cape Province of South Africa harbors the relict
Stolotermes, Porotermes, and Microhodotermes, and the tropical
steppes and savannas have the relict Hodotermes. The lack of relict
Hodotermitidae in Madagascar can be explained only by the spotty
survival pattern of relict animals in general. The most primitive
termites in the Malagasy belong to Kalotermes. This genus is cosmo-
politan, although giving evidence of persistence only in ecologic
and geographic peripheral regions. It may be guessed that the
competitive elimination in regions with rich termite faunas results
from association with higher Kalotermitidae rather than the higher
families that live in other microniches. The two endemic genera in
Madagascar are highly specialized Termitinae and Nasutitermitinae.
The lack of relatively primitive endemic relicts of Nasutitermitinae
can easily be understood because of the remote origin of the sub-
family in the Neotropical region. But the lack of relatively primi-
tive genera of Termitinae that must have originated in the Ethiopian
region during the Cretaceous is hard to understand. The hypothesis
that such termites could not penetrate the biotic barrier composed

EMERSON: DISTRIBUTION OF TERMITE GENERA 503

of locally well-adapted genera seems reasonable when there is
an exchange between large faunas, but it is doubtful if a rich en-
demic termite fauna in Madagascar in the Eocene or earlier can
account for the absence of relatively primitive Termitinae, and
there is no indication that climatic factors could account for a
barrier as they may for some genera in dry Australia.

The Papuan Fauna: The Papuan termites in general reached
the region from Indomalaya in the Cretaceous and have produced
no specialized endemic genera. Nor are primitive relicts found in
the region. The climatic difference between New Guinea and Aus-
tralia may be responsible for the lack of southward extensions of
such moisture-adapted genera as Capritermes, Grallatotermes, and
Hospitalitermes into Australia. It is strange that the endemic
Australian genera are in no case found in New Guinea. There is no
reasonable explanation for the absence of at least some of the following
genera from the drier portions of southern New Guinea: Ahamitermes,
Drepanotermes, Paracapritermes, Protocapritermes, Tumulitermes, Oc-
casitermes, Macrosubulitermes MS, Occultitermes MS, Australitermes
MS. Of these only Tumulitermes is a markedly successful genus
with many species in Australia, and a genus with many species has
a better chance to spread to another region than one with a highly
restricted local range. However, a few of these genera should have
been able to invade New Guinea during the Pleistocene connections
with Australia. It is predicted that Amilermes and some of the
listed Australian endemics will ultimately be collected in the drier
habitats of New Guinea.

DISCUSSION

The suggested hypotheses to account for the known distributional
patterns may well be over-simplifications. Fluctuating climates and
land emergence and submergence through the millions of years
covered by the Mesozoic and Cenozoic eras, together with variations
in rate of evolution, vagility, and adaptations of both closely related
and remotely related animals, render diagrammatic simplicity and
consistent statistical order improbable. The fact that so much
order is apparent in the tables of termite distribution and that
reasonable hypotheses account for the order and most of the dis-
crepancies is rather surprising. The uniformity of the nutritive
adjustments, the weakness of the flight, the pairing subsequent to
the flight, the general bradytelic evolution, and the early origin
of the order are the probable reasons why termite origins and dis-
persal are still discernible from neozoological data. Tests, correc-

504 FIELDIANA: ZOOLOGY, VOLUME 37

tions, and augmentations of the hypotheses will follow (1) accurate
determination of all termite species and genera, (2) discovery of
fossil termites, particularly of Mesozoic tropical species (see Romer,
1945), (3) accurate study of phylogenetic relationships, (4) knowl-
edge of parallel distributions and phylogenies of plants and animals
intimately or broadly associated with termites (see Darlington,
1948), (5) the geologic substantiation of ancient conditions, (6) the
accurate determination of comparative vagility (see Simpson, 1952),
and (7) further advance in our understanding of evolutionary
processes.

CONCLUSIONS

Termite distributions in general present patterns that correlate
with paleogeology, paleontology, ecology, and phylogeny. Simple
explanatory hypotheses may stimulate further research and com-
parison within the developing science of biogeography.

*-S ^H (M U3 rH

t- O 00 <M r-t O

O O i-H O
K W N W

£.2

© tO O CO

■** © © •«* »-i

w

ca

g

'O tn

(D 0>

o

■ 2.

s

3

<1>

w

3

3

c

8

f*

u

0)

T3 0>

0

9 r?

<X> lO 00

oo t- oj oo

M iH 00 (M
N N H •*

i-H © © CO

w h oo n h o

(M i-H i-l CM

IN W I W

rH *»

H

o

3

c

C2
<

'-J3

>»

H

—

is

o

U

£~

0)

J3

O ■«* CO <£>

Oi OS <M «C

oo oo io oo

i-H t-H i-l CO

: : : : ::■:•: g -s

• <D 2J ' °< a.

111 I I si 11 J ll i

u "3 £ E 45 £ '-3 -S •- fc "S * * 5

a -S "8 I l^sz^

§ W W tf H

505

sapads
[issoj puB

3UIAI[ [BJOJ,

•«* t- CO -^<

<© «D Oi O l-H r-l

-* •>* c- us

m m n w

saioads
Suiai'i jbjoj,

00 CO <D i-H CM

00 H 10

lO lO t« o
rr ->* c- \a

CO CO M U5

Eh

sapads © 1-1 o o © © © i i-h
jissoj A\a^ !

sapads
lissoj pauiB^j

o o o o o o

rH iH CO O O O

—

saioads © © © © t- co © i ©
Suiaij Avaj^ <M I t*

saioads OrtH»«»|p i i-h

• TjiJ] COW N I «£>

SUIAIJ paUIB^J <M i-t N «0

lO CO U5 H H
lO lO Tj< t-
<M <M CO •**

n

E-

a

S3
CU

g
I

E

Bjaua3 [b^oj,

BjauaS (^iuo)
jissoj A\a^

BjauaS
Suiaij A\a^

BjauaS
Suiaij pauiB^j

H lO COH ■* t- O H

Oi-HOOOOO I i-l

BjauaS (a^uo) ^
jissoj patties

i* CO i-l o o

o © o o o c- o

o i-i o> oo oo o o

o> § 3

43 -c

■o .s « -o 2

S •§ -B 2 €

? i i j 2

9

c8 O J3 S O

« WDJhP

!■§

o

Eh

i-H «> O 00 i-H r-H
IN lO H Tf

o o o o o o

o o o o o o

CJ O O U3 o O

0) <v

0) 0)

jag

5 s s S 2 5

e C £> c« 9 C K

C <* e-i |5 z w fe

506

sapads

auao
-o^spu

auaaoijjj

1-H Oi 1— I 1— I Tj< 1— I 1—1 I— I Cg ^< 1— I 1— I ^(HHHIO

5D<M <M «-< ^H«C

^^H-H • I **

ix

<!

C

bX
C

auaooij^

auaooSjio

auaoog

UBiuiaaj

•-<*,-l T*

(M • ■ <M • • -<M (M i-H i-*

i-H -hhCO i— '

•<M .-• ,-( • T-H 1-H

o

•jauiy 'S

•jauiy *^

boujv

Bisy i-i

>> adojng • oo i-h ■ ■>* i-i i-h • cm ■<* i-h • -<* ^ • ■ eo -^i-h ^ ; ; "■*

CO Q)

— o . ~ w ^
„, " a* ?> _

2 «

8.S

go

|
c

- 5 i

hfi? in
■ Si C « 2

507

sapads Ava^j

sapads
paure^[

^WNHHNHt»0 \ 1-4

uoiSaj

UAVOUJJUf^

o [Boidoj^oa^

K

o
o
S3

DI^DJBa^

UBidojmg

<M i— I t-i iH i— I

HION • t- ■ -CO

o^tji • ia

CL> OipJBaBJBJ

<

8

C3

-opui
uBndBj

UBiiBJ^sny '-1

rH <M -^ • ,-H • • lO

I

3

2^

S e

o

• s • • • ■ s • •

• e • 05 ■ • ** • •

• C 05 0) • S 03 05

•a Wemoe^tf woo

508

sapads Aiajq

saioads

oooooooo

HNM50KINM05 I OS

OO!0i0OOOONMO«O I t>

UOlSaJ
UMOUJlUfl

jBoidoj^oa^
oipjBa^

UBidoxq^a

• i— I y-t <M i-H

opojeaBjBj

'■S UB^B[BUI

I -opui

PS

UBTldBJ

UBiiBJ^sny

• O t- 'HNt-iOt-

01 0>

II

*J

K M «i S*J-P

a»<5Soo
2 F c n o -tJ

11

<< n n c/2 Ph m S <3

OS

•3 £

i g i 1

£.2 £ E

5 o

pS

Jill

PSooHPh

S SR G

a

£

»j o £

£ O <D r< »-.

Ce*5 !s *- ?

e ^ "^ O C
•STr! c "^

h O h ■- ;

509

sapads Ava^q ©©--HCo<oo©©rHO©©.-ii-io©o.-<c--;©©o

SaiOadS paUIB W HNOlOOlMHMNtlNr-iOTfHHNM^N'^H
ir>ir>T i— I CO i— i [— U5 CO i-H CO CO ^l" C\l i— < i— < W3 i~i i— I CO •"* i— I C<1 ''3* i— 1

.8

"5

H

uoi8aj umou^uji
IBOidoj^oa^j

DI^0JB3^[

Xsb3b[Bj\[
UBidoiqia
opojBaBiBj ;» ;

UBAB[BUIOpUI "HO* ;^h CO ■ ;©"*"

UBndBj «e •

■ ■ <£> lO • iH ■© • • • • ITS • • • • i-H ■ ■ i-l I OS

■ • lO • . ,-!•■•. . . . . ,_| . . |c30

■ • -i-H oo •■** • i co

oj h ■ • ' I O

i-l I <M

i—i -OS • • CO • -i-Hi-l • • i-H • • 50 • • • I i— '

■ CO ■ • • ^f • • •• • • CO • • • CO

•i-H • COCO CO

UB;iBj;sny

8:2
IS

eigisk

25P

CIh

L SS 2 O J3

•ill

«3£

— -

* 1-5 S « 2 S s

0)

1-1 C" O S L *"

o j- -C ;> g (-.
^5 C 0> « >>rt

:- ;- -

§£^£

-r

3

v.

c £ c

- C c3 M
c3 2 53 cS J3 '«

a S 3 P 2 9

510

sapads Majq

HNOOOOHooNfC^oonaoHHHecooDiOMMOr-rH

sapads
paurej^;

iNHINft^OHOlOOCiJiOt-IMHOO

m°JL

MWHMHHHHHlOMlOINHlOOOTllHINHOOOOUJOOlOHrHrH

uoi3aj

IBOidoj^oa^

oi^ojBa^

•£sb2bJBJ^

UBidoiui'J cceoHMHHrtHHiomiONHiooo'tHiNHooaiioooiOHHH

^- Ol^DJBaBlBJ

s

UBiCBIBUI

©

-opui

•*

►J

n

ubihIbj

<j

H

UBijBJ^snv

8*

SOT OT
. 0) <u

j3 *» -e
{||"§-g

. OJ

<u <u », C ■g

1-} _i_a /i^ (11 W

3 S

c -

OJ.fa 3 3 O <D

a) a> i-,

o » 0
© fl §

a2xi

£ S S

t- C t-

o) bo <v

3 Mo

otS^OO

lag*

ki c e n

, <u C C aj

-^ 51 a> fl

i*a ft ■*-> P

oI2 « *?

oj35

• w 82

w S fl

S£fc

O 3-=

"5 2 «

i as fl
lOfeW

0) • OT c

SOT ft 0)
* £•" 3

a; c *-> T3 c So
Pi §&1

•fl.'fl M,Q ?«S

a « i? 3 £ 2

511

saioads .waisj •-'©<M'-i<-'©©<mcm^©co<-i©<-io©o©oc<io-»o©t}«o in

SdI39dS Hr-(OONr)IHlOO)rHHOOTllT-lNlHHr1HNXHH10(NOiH I CO

^oj,

NHOOHMTfHt-i-lNHrHlOHnHHHrHMOiHlOlONCOH |»Q

.-H lO t-H ec 1-H ■»*

uoiSej

UMOU^Ufl

|B3ldOJ^C9^

•ri<.-iC- -Nihh -»-ieO

jCsb3bibj\[
UBidoiq^a oj ^ oo »-i eo

■1-1 -CM • -(NlON

§ ubXbjbui o .

""' -OpUJ r-l - • • • • -co |«J

H

^ uBndBj :::::::::::,H::::::::::'"H::::|<N

H

uBjiBa^sny :::::::>:: : : : : : : :^- : : : : : I §3

X

ep qj «3

£2

isse

S3 m £
■3 oj C

53 in

2 £

o £

V

03 SR

•rv 5 S3 3
x o !2 C

- - S .

_ i- gq ■—

C ^= 53 *r. .2 P. P O. & o:

'■£ O £

• B P • E e

$ «> • C «- tn S3 £

S E »:£■- Cu «

OojOOM

2 £

£ ^
roiBi

53 ti ",

„, £ tn <W

£.^£ §3

»J ■> •!-

ai.fi£ s".-s fe

a oj oi s S o. os

8 o-c cexi g.H

" 3 o C

g § ««

512

sapads ms^
sapads pauiB^

uoiSaj UAvouijuf^

fBOidoj^oaj^

DpojBajs^
XsbSbjbj^
ueidoiqia

-

aipjBasiBj

UBi?B[BUIOpUJ

uBndBj
UBi[Bi;snv

« S

« M o>

cu •■■ =

« a +; fc

o _

js -

£ 3

CO 0)

<J Ph

C ft

2 £ o So

82|.

2 « £ *>

&£= 2

t/3 3 «lj Pl,

O

a S 5 ^

S is •?

£ o « w | .§

513

saioads Ma \r oooooooicocC'-ic-ooo^HC<i«5»-HC<i»-ioooeoiocoooooo

uoi3aj

UA\OUJJUf}

514

os o

SaiOadS AY9{^ rHOOi-ie<l©©©00©.-i.-ii-ii-

S9l39dS

ONaMOHNNNHHOOONO I H ,-1

p9uiB>j t-

IB^O J, HNftftNHNNNi- I i-H ^H ^H 1-1 CO i-H I © i-h

uoiSdJ

UAVOU?lU]^

IBDldOJ^Oa^ rH (^ OS OS (M

oipjB9^[ : : : : :

O i-l

xsb3bibj^ :::::::::::::::: I 55

UBidoiqi3

O^OJBaBIBjJ

ubvCbibui
-opui

uBndBj
uBijBj^sny

■ i-H 1-1 i-H CO i-H

i-l (M(M

CM • -i-H i-H

C£>lO

coe-a

CO

to no

■**OJ

o

o

CO CO

|«D

co o>

o

01

lOO

13

OJU5 I Tf

oo o>
to eo

s§

en ^

SCO
0)

q> £

— a)

95

3

CO Q)
h4W

I El

F ^^5

■" a] 5

-2 - s-
3 O eS
daOfc

icCQ

8 .-2
.8 S

I -co

lH C0(

3 £-

— —

Si

— -

J2 C

>».S
0> (-.

OO

row

%%\

w £ £

S 2 £ E

g g 0> OJ

«* a> ~ ~

S S? 33

3J3 g.2

pi r/i V v^

o 3

0<3

«-. o

t-

01

s

0-

■p

fi

0)

£

E

eu

a

<DT3

t-

o>

9

gg

>WH

ct)

c

4)

*j

E

0/

o

0; o

"Sis

3

C

H

CO L-

52 o

s*

E>

515

{oiuiapug

•,-H • © -i-i©©

■oo • • i-i -i-ti-ti-t

roiuiapug i1-"-1 : : : : :° : : .^ : :° : : | w

w

OOHOH

'H WCOHCO

•O -©©i-H©©© • -©Tj<COt-<X> -i-l

i

s

0

£

—

a:

PQ 1

a

I

8

O
bo

c

>
3

f Diiuapua
UBidoiq^g <

IB^OJ, • "5 • t-i i-l (M i-H i— I i-H • -r-iC-OOOO

roiuiapug :0<^ : ;©©©© : : i©0

ai^DJBaBiBj <

I IB^°iL ' <M <M • • CO i-H i-H i-H • • -COi-l

roiuiapua :° : : :° ^©-^©t-^o*

UB^Bp3UIOpUJ I

IBIOT ^ ' ' " '"", 'HHi- ir- I 00 i-H «£> CO ■**

(Dimapug :° : : :°° : :°°° :°

ubtkIbj <

IBIO T ' ^ hh -COH w ■ ^*

foiuiapug '-'© •©© • •©©
uBjiBJ^sny <

516

H

►J
«

S

o

r~

03 -u

CD CD

W CO

li

*&

o -o
'So S

0> C3

« w
8

>- ■-

° s

t>
o >>

o *

£ T3

r_, CD

n bo

c s

a; ca

■^ o

2'§>

CO CD

s

v-' «
bo «

C W

ca
2 fe

CD

C
3
O
bC

C

a

o

sauua^iuiuiBJo^

3BUpilUja;OJ0BJ^

aBuiiiuuaj,
aBui^iuua^iuiy

aBppiuijax
aBppiuijajouiicjj
aBppiuija^opojj

9Bpi}lUIJ910p35[

aBppiuua^o^sBp^

H O O

rH O O

CO c- -t

OOOOOOOOi-HOOOOOOO

OOOOOOOOOOOOOOi-HO

OOWOO)OJOO«CN HNOOON

oooooooot-ooooooo

OOOOCMUiOOOOOOOO-'t^H

OOOOOOOOOOOOrH^H^H^-i

OO^hOOOOOOOOOOOOO

CD

+->

c

to "t;

OS

CD

:s °

t-

b£

CD

CD

q

M

bj

a

bo tc

.S 5

>

a c

o

— CD

°» s

E

— (-.

$

oj

t- a;

O

gPH

a

H

o

H

517

IBDidaiioajN

CM •** -^ t-

o o oo oo
eo to

orpjBdft 3

CC 00 N lO OO 05 «

C-O CM CM Tj<-«1( OO CN CM

a

•s

cu

P5

Xsb3b[bj^ ^; S

UBidoiq;a £ S; N m ^ w °° £ £

S

opDJBaBjBj 55

CO 00 N lO OO COCO N U3

UBi?B|BUIOpUI Jg g

0

irendBj °°

00 i-l CM i-H •** CM OO OO

a)

>

UBiiwjsnv ££ ^°° ^S °° 25

cu

-

a

S3

E

—
E
3
7.

2.2

c .2

•M

C CD

bX

«*H

CD
-L>

«-l

rt

<_

O

cu

cS

0
0)

C5

t-l
CD

o
cu

L>

CU

C
cu

o
cu

bo

08 bfi

_-u

M

D,

b£

tx

cd
-l>

.IS c8

"o

-L>

S

OS

be

-L>

=

"2 c

a

G

0)

a

cu

c

CD

O. CU

o

CD

-L>

cu

1

cu

cu

5

Lt

o JJ

£ r

s

0
In

^3

Q

L

cu

a

s

"S.

CU

+j

CU

CU

c
a

PL,

§Ph

o

Ph

o

Pn

Ph

C

1

0

EH

z

W

518

IBOldOJ^Oa^I

c

to
.2

orpjsaN

c

o

s

S

o
O
.g

XsbSbibj^

>>

O

imdon^g

T3
C
c8

a

o

•a

0PDJB9BIBJ

<M 00 O Nt"H <M «C> 1-1 10000 CC(00 r-iWO 0>01N

«£>"<* O U3 00O Who ^fH
<N (M y-t co

t-OJO IOOJO

OS «C C<J

OOO t-HtHO NtDO CO«£>© M(Oh

<-H "t HtO i-HCM •<* ,-1 rH

1O00O HOO
CO HN

->*«>>-l Nt-O 5OU3C0 00<M<M

co «o i— i t-rfo coco©

i— I t- W H (M

lOOO t*<M© t-UO (M
<M <M i-H

OOON 50U50 T* ,-H ,-H U3050

cu C
-= .2

!«
M

UB^BIBUIOpUJ

ubmcIbj

(Mooo

rH Tj<

t>Tj<C<l WOOT NHO lOOOO <MrH
"3 ,-HCM t— I t- CO i->Oi

■ <£> U5(M ■** --H O (MUJO HiOO 1O00O <M .-H i-H

•i— teo co t-h i— i 1-hco co i— i c^i

I

Cm

c
o

•a

Ph

I

.8

c

I

s

o
U

O

uBiiBJ^sny

• Nt-O <M?DO U000O r* 00 <-H OOJO (D(CO O] i-t i
■i— I CD HM CO T-H.-H i— ( lO •>* rHIM

a
o

e

.££

eu u o

- = £
a> a> g

Sc.5

£ fe c
gCnO

c

£ 2
££
o S
°o c
c « o

«s§

c w ^

3 S3 c
»ChO

09

Cm

g
c

§

E
8c

•S|

cu O i-.

G ° 2

SB S
M— -

g °> S
>> rt

g

C G

o o

.si

g8o
g c £

A « £

§ S3 c

c
o

s

8S
c£

•S O G
« ° °

&.s|

||8

Ik

.2 fee

cu

^g-S

ag*'

ri V c
rtCnO

H

o 5

« s

o

£ co c

c
o

s
c£

_. ca o
5 w o !r «

c c E

■sSfS

E S

p o >>

8ChO

"8

c.S •- c.s
o >> o o >>

SChO

519

REFERENCES

Ahmad, Muzaffer

1950. The phylogeny of termite genera based on imago-worker mandibles.
Bull. Amer. Mus. Nat. Hist., 95: 37-86, 17 figs.

Allee, W. C, Emerson, A. E., Park, O., Park, T., and Schmidt, K. P.

1949. Principles of animal ecology, xii+837 pp., illus. Philadelphia, W. B.
Saunders Company.

Allee, W. C, and Schmidt, K. P.

1951. Ecological animal geography, xiii+715 pp., illus. New York, John
Wiley and Sons, Inc.

Ander, Kjell

1942. Die Insektenfauna des Baltischen Bernsteins nebst damit verkniipften
zoogeographischen Problemen. Lunds Univ. Arsskrift (N.F. Avd. 2), 38:
1-82.

Beaufort, L. F. de

1951. Zoogeography of the land and inland waters, viii+208 pp., 10 figs.
London, Sidgwick and Jackson, Ltd.

Cachan, P.

1949. Les termites de Madagascar. Mem. Inst. Sci. Madagascar, (ser. A),
3: 177-275.

1951. Les termites de Madagascar. Premier Suppl. Mem. Inst. Sci. Mada-
gascar, (ser. A), 5: 1-18.

Darlington, P. J., Jr.

1948. The geographical distribution of cold-blooded vertebrates. Quart. Rev.
Biol., 23: 1-26, 105-123.

Emerson, Alfred E.

1925. The termites of Kartabo, Bartica District, British Guiana. Zoologica,

6: 291-459, 70 figs., 2 pis.
1928. Termites of the Belgian Congo and the Cameroon. Bull. Amer. Mus.

Nat. Hist., 57: 401-574, 79 figs., 24 maps, 18 pis.
1933. A revision of the genera of fossil and recent Termopsinae (Isoptera).

Univ. Calif. Publ. Ent., 6: 165-196, 40 figs.
1942. The relations of a relict South African termite (Isoptera, Hodotermitidae,

Stolotermes) . Amer. Mus. Nov., no. 1187, pp. 1-12, 2 figs.
1945. The neotropical genus Syntermes (Isoptera: Termitidae). Bull. Amer.

Mus. Nat. Hist., 83: 427-471, 12 figs., 1 map.
1947. The imago of Stolotermes africanus Emerson. Jour. Ent. Soc. So. Africa,

9: 127-129.

1950. Five new genera of termites from South America and Madagascar
(Isoptera, Rhinotermitidae, Termitidae). Amer. Mus. Nov., no. 1444, pp.
1-15, 6 figs.

1952a. The biogeography of termites. Bull. Amer. Mus. Nat. Hist., 99:
217-225.

520

EMERSON: DISTRIBUTION OF TERMITE GENERA 521

1952b. The neotropical genera Procornitermes and Cornitermes (Isoptera,
Termitidae). Bull. Amer. Mus. Nat. Hist., 99: 475-540, 32 figs., 2 maps.

1953. The African genus Apicotermes (Isoptera: Termitidae). Ann. Mus.
Roy. Congo Beige, (ser. 8), Sci. Zool., 17: 99-121, 4 figs.

Hare, Laura

1937. Termite phylogeny as evidenced by soldier mandible development.
Ann. Ent. Soc. Amer., 37: 459-486, 13 pis.

Harris, W. V.

1954. Termites from Socotra (Isoptera). Ann. Mag. Nat. Hist., (ser. 12),
7: 493-496.

Holmgren, Nils

1911. Termitenstudien 2. Systematik der Termiten. Die Familien Masto-
termitidae, Protermitidae und Mesotermitidae. K. Svensk. Vet.-Akad.
Handl., 46, no. 6, 88 pp., 6 figs., 6 pis.

1912. Termitenstudien 3. Systematik der Termiten. Die Familie Meta-
termitidae. K. Svensk. Vet.-Akad. Handl., 48, no. 4, 166 pp., 88 figs., 4 pis.

1914. Wissenschaftliche Ergebnisse einer Forschungsreise nach Ostindien,
ausgefiihrt im Auftrage der Kgl. Preuss. Akad. Wissensch. Berlin von H. v.,
Buttel-Reepen. III. Termiten aus Sumatra, Java, Malacca und Ceylon.
Gesammelt von Herrn Prof. Dr. v. Buttel-Reepen in den Jahren 1911-1912.
Zool. Jahr. Abt. Syst., 36: 229-290, 1 pi.

Kemner, N. A.

1931. Die Termitenfauna von Amboina. Lunds Univ. Arsskrift (N.F. Avd. 2),
27: 1-53.

Moreau, R. E.

1952. Africa since the Mesozoic: with particular reference to certain biological
problems. Proc. Zool. Soc. London, 121: 869-913.

Ratcliffe, F. N., Gay, F. J., and Greaves, T.

1952. Australian termites. The biology, recognition, and economic importance
of the common species. 124 pp., 21 figs. Melbourne, Commonwealth Scientific
and Industrial Research Organization.

Romer, A. S.

1945. Vertebrate paleontology (2nd ed.). viii+687 pp., illus. Chicago. Univ.
Chicago Press.

1952. Discussion [of Colbert, E. H., The Mesozoic tetrapods of South Amer-
ica]. Bull. Amer. Mus. Nat. Hist., 99: 250-254.

Ross, H. H.

1953. On the origin and composition of the Nearctic insect fauna. Evolution,
7: 145-158.

See vers, C. H.

1937. New species of termitophilous Staphylinidae from tropical America and
the Solomon Islands. Ann. Ent. Soc. Amer., 30: 1-23, 20 figs., 3 pis.

Simpson, G. G.

1952. Probabilities of dispersal in geologic time. Bull. Amer. Mus. Nat.
Hist., 99: 163-176.

Snyder, T. E.

1949. Catalog of the termites (Isoptera) of the world. Smiths. Misc. Coll.,
Publ. 3953, 112: 1-490.