Laurie Burnham.
Survey of Social Insects in the Fossil Record.
Psyche 85:85-134, 1978.

Full text (searchable PDF, 4772K)
Durable link: http://psyche.entclub.org/85/85-085.html

The following unprocessed text is extracted from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.

SURVEY OF SOCIAL INSECTS
IN THE FOSSIL RECORD*
BY LAURIE BURNHAM
Museum of Comparative Zoology,
Harvard University,
Cambridge, Massachusetts 021 38, U.S.A.
Biologists have long been intrigued by the complex social systems of various insects. Despite a voluminous literature dealing with the evolution of these systems, immense gaps remain in our understand- ing of insect sociality. Several theories have been proposed to explain the evolution of social behavior in certain groups of insects (e.g., Hamilton, 1964), but none consider this problem with respect to geological time. The present paper does so by examining the fossil record for clues not only on the antiquity of sociality, but also on the nature of these early social insects. Included in this survey are those insects recognized as eusocial: the Isoptera, and three super- families of the Hymenoptera: Vespoidea, Formicoidea, and Apoidea.
ISOPTERA
The termites are remarkable in two regards: 1) as a group, they are fully eusocial, exhibiting a wide range of behavioral modifica- tions and sophistications, and 2) their record in the geological past, although sparse, is highly indicative of an Early Mesozoic origin. This latter point is of particular significance if one considers sociality among insects as a pinnacle of evolutionary success. Wilson (1971, p. 1) states that "[insect societies] best exemplify the full sweep of ascending levels of organization, from molecule to society." The possibility that termites evolved a social organization as far back in geological time as the Jurassic (roughly 190 million years ago) is of great interest, particularly when attempting to draw parallels with the evolution of sociality in the Hymenoptera, a group phylogenetically very remote from the termites. *Manuscript received by the editor July 7, 1978. 85




================================================================================

86 Psyche [March
Five of the six families' of termites recognized by Emerson (1955) have a fossil record extending at least as far back as the Tertiary. In 1967, Cretatermes carpenter! (Hodotermitidae) was found in an Upper Cretaceous deposit in Labrador (Fig. l), a discovery which immediately placed the origin of the Isoptera no later than the Mesozoic - an extension of 45 million years from previously known specimens. In addition, the advanced phylogenetic position of Cretatermes provides evidence for a much earlier origin of the order than has formerly been recognized (Emerson, 1967). An examination of various fossil localities reveals a widespread termite fauna during the Tertiary Period (Table 1). The Termitidae are found in Miocene deposits of California and Germany; the Rhinotermitidae, Hodotermitidae, and Kalotermitidae are found at various Tertiary deposits throughout the United States and Europe; and the Mastotermitidae have the most widespread Cenozoic distribution of all, having been found at localities in the United States, Europe, South America, and Australia. This latter finding is highly intriguing because the family Mastotermitidae today has but one species, Mastotermes darwiniensis, which is restricted to north- ern Australia.* Emerson (1955) postulates that this widespread 'The sixth family is the Serritermitidae - an aberrant taxon known from only one species.
*A look at past climatic shifts provides additional insight into the redistribution of the termites, particularly with respect to the Mastotermitidae, now solely restricted to Australia. Reconstructions of paleo-climatic patterns may be made fairly accurately on the basis of floral analyses (Reid and Chandler, 1933). The presence of Sequoia stumps in the Florissant Shales of Colorado provides evidence for warmer tempera- tures during the Oligocene (Emerson, 1969). Tiffney (1977) postulates on the basis of fossil angiosperm assemblages that temperatures in New England during the Oligocene were much more equable than at present - the temperatures ranging from 26OC to 9OC in contrast to today's 21å¡ to -lOå¡C Furthermore, extended frosts and hard freezes were unknown. In the more tropical climate of the Oligocene, 'colony activities were presumably carried out year round in a relatively warm, moist environment, explaining the widespread distribution of the Mastotermitidae during the Lower to Middle Tertiary. By the Late Miocene or Early Pliocene, the earth's climate began shifting towards cooler temperatures with the rising level of the continental land masses and increasingly large polar ice caps. My hypothesis is that, unable to adapt to an increasingly colder climate, and possibly to a concomitant change in predator pressures, the Mastotermitidae began to die out during the Tertiary. And, because at this time the Termitidae were undergoing tremendously successful radiation in Africa and South America, the Mastotermitidae became geographically restricted to northern Australia, represented today by only one relict species, Mastotermes darwiniensis.




================================================================================

19781 Burnham - Social Insects in Fossil Record 87 Figure 1.
Cretafermes carpenters Emerson from lower part of Upper Cretaceous of Labrador. Note humeral suture at wing base. Original photograph of holotype in Prince~on Museum. Length of wing, 7.5 rnm, geographical distribution provides strong evidence to support a Mesozoic origin of the order. He argues (1975) that the breakup of the united land mass Pangaea in the Permian or Lower Triassic must have occurred subsequently to the origin of the Isoptera to explain their distribution in the southern and northern continental land masses and that all five families must have been present in the Late Mesozoic to explain their diversity and distribution by the Tertiary.
In 1971 he looked at a variety of primitive and derived characters of each family and analyzed the geographical distribution of the groups, using plate tectonics to provide the following estimates on the geological origin of the families:
Mastotermitidae - possibly Early Mesozoic. Hodotermitidae - Triassic, or Early Jurassic before the breakup of southern continents.
Kalotermitidae - mid-Jurassic, or Lower Cretaceous, before the separation of Africa and South America.
Rhinotermitidae - Late Jurassic, Early Cretaceous. Termitidae - Cretaceous.
Because termites are such poor fliers and do not mate until the adults have cast their wings, he considers water gaps of more than 50 miles capable of preventing termite dispersal. While I am supportive of the theory that places great importance on the role of a unified land mass in animal dispersal, I do not agree that this can effectively be used to date the origin of the Isoptera.



================================================================================

TABLE 1 ISOPTERA IN THE FOSSIL RECORD.
Geological Age
CRETACEOUS
Hodotermitidae
* Cretatermes carpenteri Emerson
EOCENE
Mastotermitidae
*Blattotermes wheeleri \Collins
*Zdomastotermes myslicus Ha up t
Kalotermitidae
Neotermes grassei Pi ton
Hodotermitidae
Termopsis mallaszi Pongracz
OLIGOCENE
Mastotermitidae
*Miotermes insignis (Heer)
*Miotermes spectabilis (Heer)
Mastotermes bournemouthensis von Rosen
Mastotermes heeri (Goppert)
Mastotermes balheri von Rosen
Kalotermitidae
* Prokalotermes hageni (Scudder)
*Electrotermes giradi (Giebel)
*Electrotermes affinis (Hagen)
Kalotermes rhenanus Hagen
*Eotermes grandaeva Statz
*Proelectrotermes berendti (Pictet)
Locality
Labrador, Canada
Tennessee, U.S.A.
Geiseltal, Germany
Menat, France
Hungary
Oeningen, Germany
Oeningen, Germany
England
Schlesien, Germany
England
Florissant, Colorado
Baltic Amber
Baltic Amber
Rott, Germany
Rott, Germany
Baltic Amber
References
Emerson, 1967
Emerson, 1965
Emerson, 1965
Emerson, 1969
Snyder, 1949
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, I969
Emerson, 1969
Emerson, 1969
Emerson, I969
Emerson, I969
Emerson, I969




================================================================================

Hodotermitidae
Archotermopsis tornquisti von Rosen
Termopsis bremii Heer
* Parotermes insignis Scudder
* Parotermes scudderi Cockerel1
* Ulmeriella bauckhorni Meunier
* Ulmeriella cockerelli Martynov
Rhinotermitidae
*Reticulitermes minimus (Snyder)
Reticulitermes fossarum (Scudder)
Reticulitermes antiquus (Germar)
Reticulitermes creedei Snyder
* Parastylotermes robustus (Rosen)
MIOCENE
Mastotermitidae
*Spargotermes costalimai Emerson
Mastotermes vetustus Pongracz
Mastotermes minor Pongracz
Mastotermes haidingeri (Heer)
Mastotermes croaticus von Rosen
*Miotermes procerus (Heer)
* Miotermes randeckenensis von Rosen
* Pliotermes hungaricus Pongracz
Kalotermitidae
Cryptotermes ryshkoffi Pierce
Kalotermes swinhoei (Cockerell)
Kalotermes tristis (Cockerell)
Kalotermes nigrit us Snyder
Baltic Amber
Baltic Amber
Florissant, Colorado
Florissant, Colorado
Rott, Germany
Siberia, U.S.S.R.
Baltic Amber
Florissant, Colorado
Baltic Amber
Creede, Colorado
Baltic Amber
Brazil
Radoboj, Croatia
Radoboj, Croatia
Radoboj, Croatia
Radoboj, Croatia
Radoboj, Croatia
Wiirttemberg, Germany
Radoboj, Croatia
Calico, California
Burma
Burma
Chiapas, Mexico
Snyder, 1949
Snyder, 1949
Snyder, 1949
Cockerell, 19 13
Emerson, 1968
Emerson, 1968
Emerson, 197 1
Emerson, 197 1
Emerson, 197 1
Emerson, 197 1
Emerson, 197 1
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1965
Emerson, 1969
Emerson, 1969
Emerson, 1969
Snyder, 1960




================================================================================

Geological Age
TABLE 1. (CONCLUDED)
Locality
MIOCENE (continued)
Hodotermitidae
* Ulmeriella latahensis Snyder
* Ulmeriella martynovi Zeuner
Rhinotermitidae
Heterotermes primaevus Snyder
Reticulitermes hartungi (Heer)
Reticulitermes laurae Pierce
* Parastylotermes calico Pierce
*Parastylotermes washingtonensis (Snyder) Termitidae
Gnathamitermes magnoculus rousei Pierce
Macrotermes pristinus (Charpentier)
Latah, Washington
Biebrich, Germany
Chiapas, Mexico
Radoboj, Croatia
Calico, California
Calico, California
Latah, Washington
Calico, California
. Radoboj, Croatia
References
Emerson, 1968
Emerson, 1968
Emerson, 197 1
Emerson, 1971
Emerson, 197 1
Emerson, 1971
Emerson, 197 1
Pierce, 1958
Snyder, 1949
*Extinct genera.




================================================================================

19781 Burnham - Social Insects in Fossil Record 9 1 Simpson (1952) has made some insightful remarks on the matter. He contests the premise that if a given group of organisms requires a land connection, then disjunctive areas occupied by the group must have been once connected by continuous land. His contention is that there is no group of organisms that cannot be dispersed over water. Given a probability of only one chance in a million that an organism can cross a stretch of water, when geological time is considered the chance that the event will actually take place (over tens of millions of years) becomes significantly greater. It is further argued that successful colonization is dependent on successful invasion and the ability of the intruder to compete with existing species. Chances for survival are much higher when there are numerous, simultaneous arrivals of individuals. In my opinion, the termites support such reasoning, and this can be argued in several ways. Firstly, termites are relatively light- bodied, winged insects. Studies by Simberloff and Wilson (1969) and Glick (1933) on the repopulation of an island by wind trans- ported insects strongly support the possibility that termites are capable of being carried considerable distances in the upper atmos- phere. Furthermore, because termites swarm in such large numbers prior to reproduction, a reasonable possibility exists that they will be dispersed to a new habitat as either a group or at least as a malelfemale pair. A wind current strong enough to blow one individual into the upper atmosphere should be equally capable of carrying multiple individuals, and, according to windflow, of trans- porting them in the same directional pathway. Secondly, termites are ideally suited to dispersal over large bodies of water via floating logs. The more primitive families construct their extensive nesting colonies in wood and logs; as a consequence, it is entirely plausible that a dead tree falling into a body of circulating water could be carried extended distances. Furthermore, this mode of transportation provides the termites with a source of food during their sojourn, and travel en rnasse obviates the prob- lems of reproduction upon arrival. In addition, as Simpson points out, the larger the number of individuals, the more likely it is that they will be successful competitors in the new habitat. I am not presenting this as evidence that the termites did not evolve while the earth's land masses were still contiguous, but am merely pointing out the problems in arguing that land dispersal was essential for termites.




================================================================================

92 Psyche [March
The Isoptera exhibit strong affinities to the Blattodea; evidence linking the two groups to a common ancestor is well marked between the Mastotermitidae, an archaic termite family, and the Cryptocercidae, a family of generalized cockroaches. This theory of common ancestry is supported by several comparative morphologi- cal and behavioral studies (Emerson, 1965; McKittrick, 1965; Ahmad, 1950; Cleveland, 1934; Hill, 1925). McKittrick (1965) goes so far as to incorporate both groups into the Dictyoptera, an order which also includes the Mantodea. The gut fauna, female genitalic structures, anal expansion of the hind wing, morphology of the proventriculus, and deposition of eggs in ootheca-like masses are much alike in Mastotermes and Cryptocercus. Furthermore, both groups inhabit similar habitats. As a consequence, termites have often been referred to as merely social cockroaches. This degree of relatedness becomes immediately interesting in view of the extensive geological record of the cockroaches.
Fossil cockroaches are first found in deposits from the Upper Carboniferous, which makes them among the oldest insects known. Furthermore, they comprise 80 percent of the fossil insect fauna during that period (Carpenter, 1930) - an indication that they have not only existed, but have flourished, for three hundred million years. If the similarities between termites and cockroaches are indeed the result of monophyletic, rather than convergent or parallel evolution, one might speculate on a much earlier origin for the Isoptera than is shown by the fossil record. McKittrick (1965) admits that the flagellate gut fauna essential for cellulose digestion in both groups may have arisen independ- ently in each; however, she believes that the similarities in two important morphological characters, the female genitalia and the dental belt of the proventriculus, represent primitive characters and are therefore indicative of a common origin for Mastotermes and Cryptocercus. On the other hand, Tillyard (1926, 1936), Cleveland (1934), Imms (1 9 19), Carpenter (personal communication), among others, believe that the termites were derived from more ancient stock and may have evolved during the Late Paleozoic. Hamilton (1978) supports the view that social termites arose from "roach-like ancestors" in the habitat of dead phloem, and suggests that the invasion of Cryptocercus into the same typeof habitat was inde- pendent of the ancestral termite. The possibility of termite bbevolu-



================================================================================

19781 Burnham - Social Insects in Fossil Record 93 tion under bark" seems immensely feasible; not only is isolation (and, hence, inbreeding) possible, but selective pressures leading to dependence on a cellulose diet would also be high. It seems an excellent explanation for the early separation of the termites and cockroaches from a common protorthopteran (protoblattoid) an- cestor as long ago as the Late Paleozoic. More definite conclusions on the origin of the Isoptera must wait until termites or termite-like insects have been found in pre-Cretaceous strata. HYMENOPTERA
The Hymenoptera belong to the major subdivision of the Insecta known as the Endopterygota. There are no clues elucidating the nature or precise age of the earliest endopterygote insects, but the fossil record does provide insight into the history of the group as a whole. Representatives of two endopterygote orders, Neuroptera and Mecoptera, are found as far back as the Early Permian, some 280 million years ago. This occurrence suggests an origin of the Endopterygota approximately 100 million years after the origin of the true insects.3
The earliest known Hymenoptera have been found in Triassic beds of Central Asia (Rasnitsyn, 1964) and Australia (Riek, 1955). These fossils establish a minimum age for the order of about 220 million years. All the specimens known from this period belong to the suborder Symphyta, and surprisingly enough belong to the existing family Xyelidae.
A major advance in the evolution of the Hymenoptera occurred with the development of a constriction between the first and second abdominal segments; this presumably had the selective advantage of increasing the flexibility of the abdomen, important for both oviposition and defense. Hymenoptera which possess this adapta- tion, a diagnostic character of the suborder Apocrita, are first known from Upper Jurassic deposits of Central Asia (Rasnitsyn, 1975, 1977). These specimens have been assigned to the more primitive division of the Apocrita known as the Terebrantia or The oldest known insects, found in Upper Carboniferous deposits, comprise 11 orders and include the Apterygota (Thysanura), Paleoptera and Exopterygota. It should be noted that here the use of the term insect does not include the Collembola, Protura or Diplura.




================================================================================

94 Psyche [March
Parasitica; the other division within this suborder is the Aculeata.4 Members of the latter are characterized by modifications of the ovipositor that have enabled its use not only for oviposition, but also as a transport vessel for defensive and prey-paralyzing com- pounds. This structure unquestionably plays an important role in colony defense and might provide an explanation for the restriction of eusociality within the Hymenoptera to the Aculeata. The oldest known aculeate hymenopteron, Cretavus sibericus, was discovered in an Upper Cretaceous (Cenomanian) deposit in Siberia in 1957. Although placed by Sharov (1962) in an extinct superfamily Cretavidea, related to the Scolioidea, it has recently been transferred to the existing family Mutillidae by Rasnitsyn (1977, p. 109). Since 1967, species representing 10 families and 19 genera of aculeate Hymenoptera have been found in Upper Cretaceous deposits in Central Asia (Rasnitsyn, 1977) (Table 2). Evans (1966) believes that such diversity by the Late Cretaceous is indicative of an earlier origin and postulates that the group may have evolved during the Jurassic. However, it must be pointed out that the Cretaceous is one of the longer periods in the earth's history, having a duration of roughly 70 million years, and may have been of sufficient length to account for such diversification.
VESPOIDEA
Included in this group are the three families considered to be "true wasps": The Masaridae and Eumenidae, both of which are solitary, and the Vespidae, where one finds behavioral modifications ranging from subsocial to highly advanced eusocial (P-ichards, 1953, 197 1). It is the Vespidae, by virtue of their sociality, with which I am primarily concerned in this paper.
There are many gaps in our record of the early social wasps and of the Vespoidea in general. Most striking, perhaps, about the fossil record of the wasps is their lack of representation (see Table 3). The "The classification of the Aculeata has recently undergone a major revision by D. J. Brothers (1975), in which the seven previously recognized superfamilies (Bethyloidea, Scolioidea, Pompiloidea, Formicoidea, Vespoidea, Sphecoidea, and Apoidea) are now combined into three: the Bethyloidea, Sphecoidea (subdivided into the Spheci- formes and Apiformes), and Vespoidea (subdivided into the Vespiformes and Formiciformes). However, since this revised classification has not been generally accepted in its entirety, I am employing here the more conventional classification (sensu Riek, 1970; Richards, 1971).