AMERICAN MUSEUM Novitates

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024

Number 3097, 31 pp., 84 figures May 19, 1994

Electron Microscopic Studies of Mummified Tissues in Amber Fossils

DAVID GRIMALDI,! ELIZABETH BONWICH,? MICHAEL DELANNOY,? AND STEPHEN DOBERSTEIN?*

ABSTRACT

The degree and consistency of fine and ultra- structural preservation of primarily soft tissue re- mains preserved in amber fossils were examined, using scanning (SEM) and transmission electron microscopy (TEM). 16 insects of various taxa and structure were studied with the SEM, as well as 4 plant specimens, which were from two chemically distinct ambers: 25—30 million year old (myo) am- ber from the Dominican Republic, and ca. 40 myo amber from the Baltic region. A new technique was used for “‘exhuming” a specimen in order to examine the in-situ preservation of whole organs and tissues. Remarkable preservation found in both ambers confirmed earlier reports based on a few specimens. Preservation of all soft tissues in Do- minican amber insects appears to be more con- sistent than in Baltic amber insects, several of which were virtually hollow “‘casts.””» Membranous struc-

tures preserved in the insects include air sacs, un- collapsed tracheae, and various portions of the gut, as well as the brain and bundles of muscle fibers in their original origins and insertions. Very little shrinkage was observed. Specialized pockets, the mycangia, in wood-boring beetles (Platypodidae) still possessed spores and conidiophores of their symbiotic fungus. For plants, columnar cells of leaf mesophyll were found in their original positions and sizes, pollen grains retained the exine sculp- turing, and mats of epidermal cells were preserved in anthers.

At the ultrastructural level, both flight muscle and brain tissues show substantial degradation of cytoplasmic components due to dehydration. Al- though sarcomeres are easily identifiable in native muscle samples, the sarcomeric repeats disappear upon hydration, indicating that the repeated struc-

' Associate Curator, Department of Entomology, American Museum of Natural History.

? Student, Department of Biological Sciences, Barnard College of Columbia University, New York, NY.

> Department of Cell Biology and Anatomy, Johns Hopkins School of Medicine, Baltimore, MD 21224.

*4 Johns Hopkins School of Medicine; presently: Postdoctoral Fellow, Department of Molecular and Cell Biology, 519 LSA, University of California, Berkeley, CA 94720.

2 AMERICAN MUSEUM NOVITATES

tures are probably composed of inorganic salts de- posited on thick filaments during dehydration. Membranous structures are generally better pre- served than proteinaceous ones. Flight muscle mi- tochondria are particularly well preserved with tracheoles and T tubules also identifiable. Axon tracts in the central nervous system can be distin- guished from cytoplasmic regions, and parallel

NO. 3097

strips of membranes surrounding cytoplasm are abundant in rehydrated brain samples.

Mode of tissue preservation appears to be a rap- id and thorough fixation and dehydration, sufh- cient for preserving DNA in amber more consis- tently than any other kind of fossil. Pollen was not found viable, but the possibility remains that vi- able bacterial and fungal spores occur in amber.

INTRODUCTION

In general reviews of taphonomy, reference is often made to Conservat Lagerstatten, or fossil deposits of exceptional preservation (Allison and Briggs, 1991; Briggs, 1991). These include, for example, the Eocene oil shale of Messel, Germany; the fine-grained Lower Cretaceous limestone of Ceara, Brazil (Grimaldi, 1990; Maisey, 1991; Martill, 1988); and the intricately preserved carbo- naceous flowers from the Turonian (mid- Cretaceous) of New Jersey (Crepet et al., 1992). For deposits like the shale and lime- stone, impressions of soft tissues may remain (“extraordinary”’ preservation, sensu Briggs [1991]), but these are only low molecular weight residues and carbonized films. The flowers from the Raritan-Magothy Forma- tion of New Jersey are actually charcoalified: turned to pure carbon after forest fires swept over them while buried in leaf litter. A sed- imentary deposit with unique preservation are the arthropods from the middle Devo- nian shale of Gilboa, New York (ca. 378 mil- lion years old): sockets of setae, sensilla, and other microscopic details are preserved, ap- parently as the original cuticle (Shear et al., 1984; Schawaller et al., 1991).

Three rarely-discussed modes of preser- vation retain features of organisms more life- like than any other kinds of fossils: freezing, dehydration, and preservation in amber. All apparently prevent hydrolysis of complex molecules, by either suspending water as a solid or removing it. Very often frozen or- ganisms, such as the carcasses of mammals in cold deserts, are partially dehydrated, due to the sublimation of ice crystals. Frozen and dehydrated organisms rarely extend past the Holocene or Pleistocene; thus, their paleon- tological value is limited. Yet, no preserva- tion is more celebrated than the highly ritu- alized burials of the ancient Egyptians,

particularly those from the 21st dynasty (ca. 1050 sBc)(David and Tapp, 1984), although exceptional mummification occurs in other ancient cultures around the world (Cockburn and Cockburn, 1980; Hansen et al., 1991). Egyptian cadavers were first eviscerated, then washed in wine, and the thoracic cavity stuffed with incense, cassia and other spices, and crushed myrrh (a fragrant, resinous gum from Commiphora [Burseraceae]). The cadavers were also stored for about 70 days in dry natron (hydrated sodium carbonate, Na,CO,- 10H,0, a natural deposit of salt lakes). Linen wrappings around the body were likewise im- pregnated with resin, or with pulverized am- ber. The success of Egyptian mummification was revealed by histological (Sandison, 1955; Zimmerman, 1973) and ultrastructural stud- ies (Curry et al., 1979; Lewin, 1967; Riddle, 1980). Cells, albeit shrunken, contained nu- clei and nuclear membranes, although mi- tochondria were not readily observable (Lew- in, 1967). DNA, too, has been extracted and sequenced from Egyptian mummies (Paabo, 1985).

How the Egyptians began the use of amber for mummification is intriguing, especially since amber does not occur in Egypt. Amber is the highly polymerized, fossil form of tree resins; hundreds of deposits occur around the world, varying in age, botanical origin (Lan- genheim, 1969), and the kind and quality of small organismal inclusions in it. Very an- cient amber (ca. 125 million years old) occurs in Lebanon, Jordan, and Israel (Schlee and Dietrich, 1970; Bandel and Vavra, 1981; Nis- senbaum, 1975), but the early Egyptians probably traded for it with the Phoenicians, who may have acquired it from more exten- sive deposits in Sicily. Amber and resin were certainly not as important as preservative agents of the mummies as was the natron,

1994

but their use was perhaps inspired by the re- markable external preservation of small or- ganisms, like insects, that the Egyptians prob- ably saw in the amber. They certainly were unaware of the great antiquity of amber, and of the preserved details of internal tissues. Here, we report the preservation of soft, in- ternal insect and plant tissues fossilized in lower Tertiary ambers, with unexpectedly consistent, lifelike fidelity: a process of nat- ural mummification, with finer preservation than is found in human mummies, and per- haps the ultimate in Conservat Lagerstatten.

ACKNOWLEDGMENTS

It is a pleasure to acknowledge the help of several people who were instrumental in bringing this study to fruition. At the AMNH, Peling Melville provided a great deal of time and expertise to the scanning electron micro- graphs; and Jacquie Beckett produced prints from the SEM negatives. Jake Brodzinsky was a very helpful source of study specimens. We thank Drs. Rob DeSalle (AMNH), George Eickwort (Cornell University), and Jean Lan- genheim (University of California, Santa Cruz) for their critical review of the manu- script; and Drs. Thomas D. Pollard and Da- vid B. Weishample of the Johns Hopkins School of Medicine, who provided support and stimulating discussions.

Costs for the work at the AMNH were payed for by an NSF grant to Melanie Stiass- ny, which sponsored Elizabeth Bonwich in the REU program in Comparative and Evo- lutionary Biology. The personal generosity of Dr. Herbert Axelrod is gratefully appreciated, which allowed for the purchase of the spec- imens used here and numerous other rare ones which will never, ever be marred.

ABBREVIATIONS

an antenna as air sacs at atrium (of spiracle)

b brain cm _ cibarial muscle cr crop

ef eye facets es esophagus fl flagellomere ge genitalia

GRIMALDI ET AL.: AMBER FOSSILS 3

gl glossa

g gut

h head

Imd_ longitudinal median dorsal muscle my mycangium

oc ocelli

oim oblique intersegmental muscle olm oblique lateral muscle

pe pedicel

ph phragma

pr pyloric region th thorax

tr trachea

ve ventriculus

PREVIOUS STUDIES

There have been several reports on the preservation of tissues in amber fossil ar- thropods, each of which is based on only one or two specimens. Kornilowitsch (1903) was the first to report the fine structure of tissue preserved in amber fossil insects. He thin- sectioned the legs of Diptera and Neuroptera in Baltic amber and found remarkable stri- ations in the muscles. Mierzejewski (1976a) was the first to apply electron microscopy to tissues “‘exhumed”’ from amber fossils. He reported “optic cells, pigment cells, or crys- talline cones” in the facets of a dolichopodid fly preserved in Eocene Baltic amber, as well as the minute tracheae that deliver oxygen to these and other insect cells. (Unfortunately, the SEM in his plate II, fig. 2, offers little resolution; secondary iris cells in the middle of each optic cell do not appear to be pre- served). Elsewhere (Mierzejewski, 1976b) il- lustrated with the SEM an entire book lung from a Baltic amber spider: these are an in- vaginated series of thin parallel plates with a thin cuticular covering. The reports by Poin- ar and Hess (1982, 1985) and Poinar (1992) are based on a single sample of tissue which adhered to the inner wall of the abdomen of a fungus gnat (Diptera: Mycetophilidae), also in Baltic amber. Using transmission electron microscopy (TEM), they reported nuclei, “lipid droplets” (histochemical tests for lip- ids were not done), mitochondria and cristae, and apparent endoplasmic reticulum in epi- dermal cells. Muscle banding was not well preserved. They considered such preserva- tion to be anomalous, since it was believed that the resin would need to be in direct con-

4 AMERICAN MUSEUM NOVITATES

tact with tissues, which might occur if the body wall was traumatically opened. The possibility was mentioned that “‘sugars and terpines [sic] present in the original resin” could have dehydrated the tissues. Poinar (1992) also presented a TEM of tissue from the abdomen of a braconid wasp in Canadian Cretaceous amber (ca. 80 myo), in which “folded membraneous [sic] structures adja- cent to the vacuolated cytoplasm” (p. 270) were identified.

Schliiter (1989) presented SEMs of a ter- mite and possible ant head preserved in Cen- omanian amber from France (ca. 100 myo). Fine structure of the epicuticle that was ob- served included microtrichiae, and reticula- tions on the original cuticle. The preservation of soft tissues was not examined. He also found minute crystals of marcasite in some specimens, which is often associated with the iron-rich sediments of ancient deltas where amber deposits usually occur. Ground water with dissolved minerals seeps into fine cracks and permeates an inclusion where crystalli- zation later occurs. In fact, Baroni-Urbani and Graeser (1987) described with SEMs a pyritized cast of an ant in Baltic amber. Enough detail was preserved to determine that the cuticular microsculpturing of this ex- tinct species differed from that of any living species. The work of these authors implied, though, that electron microscopy of the cu- ticle of amber insects would depend on pyr- itization, since this mineral would conduct electrons well.

Studies by Henwood (1992a, b) used SEM and TEM on three insects preserved in Oligo- Miocene amber from the Dominican Repub- lic. In the TEM study, the flight muscle of an empidid (““dance’’) fly was cross-sectioned, revealing the hexagonal ultrastructure of my- ofibrils and the mitochondria densely packed among them, even showing the intricate fold- ing of the internal cristae. Longitudinal sec- tions of the flight muscle, which could have shown additional ultrastructural detail, were not made. In the SEM study, Henwood stud- ied two beetle specimens, a cantharid (“‘sol- dier beetle”’) and a nitidulid (“‘fungus beetle’’). Nitidulids are as heavily sclerotized as most other beetles (the cantharid less so), and both these specimens were completely intact, in- dicating that the viscous resin does not need

NO. 3097

to contact the internal organs directly. Some tissues were found in their original positions, but shrunken to about 50% of the original size. Most interesting was the observation of a proventriculus in the nitidulid, which is posterior to the esophagus and is very slightly sclerotized (virtually membranous). Blunt, heavily sclerotized teeth lining the interior serve as a grinding mill and food filter, and the entire structure was remarkably intact. Unfortunately, Henwood’s method of cutting entirely through the specimen with a saw abraded much of the internal remains, and she concluded by acknowledging the need for an alternative method of extraction.

The present study expands upon the pre- vious ones by using many more specimens of insects of various taxa, sizes, internal anat- omy, and degree of cuticle sclerotization as well as several plant specimens. The insects derive from two chemically distinct ambers, which allows additional taphonomic com- parisons. This approach more fully addresses the question of consistency of soft tissue pres- ervation in amber fossils. Also, we used a procedure for examining whole tissues and organs that was minimally destructive, leav- ing the internal structures more intact. The remarkable tissue preservation is apparently due (at least partially) to dehydration, which is an ideal condition as well for long-term storage of pollen (Shivanna et al., 1991). Thus, we examined possible viability of Hymenaea pollen in Dominican amber, which is the tree that gave rise to this amber (Langenheim, 1969; Hueber and Langenheim, 1986) and whose anthers are a relatively common in- clusion.

MATERIALS AND METHODS

Amber fossilized insects and plant parts were acquired from several sources. Baltic amber was acquired from Jorge Wiinderlich (StauSenhardt, Germany), and Dominican amber from Jacob Brodzinksy (Santo Do- mingo, Dominican Republic), Manuel Perez (Orland, FL), and several other dealers in Santo Domingo and Santiago. Specimens were selected according to several criteria, most important of which was an abundant supply of individuals. Particular effort was made to use only common species (e.g., where

1994

a series of at least 20 specimens existed in the AMNH collection and which could be easily replaced). Secondly, selected speci- mens had no fine cracks between the surface and the inclusion, which might have allowed invasion of the amber seal by bacteria, mois- ture, or other elements of decomposition. Third, specimens represented a great taxo- nomic diversity, which generally correspond- ed with great variation in basic body form, habits, relative amount of muscle mass, etc.... Fourth, species of significantly dif- ferent sizes were chosen, since the proportion of muscle mass to body surface area would apparently be a significant factor in mum- mification. Lastly, species were chosen based on differences in the degree of sclerotization and thickness of cuticle, beetles being the most heavily sclerotized and termites being the least.

The following specimens were examined: For Dominican amber—three stingless bees (Proplebeia dominicana [Wille], Apidae: Me- liponini), two termites (Reticulitermes sp., Termitidae), three platypodid beetles, two fungus gnats (Mycetophila sp.: Mycetophili- dae), two scuttle flies (Megaselia sp.: Phori- dae), one leaflet of Hymenaea (Legumino- sae), and seven anthers of this tree. For Baltic amber—two dolichopodid flies, and two my- cetophilid fungus gnats.

Internal organs of the insect inclusions and the pollen contents of stamens were exposed using the following method. Each specimen was first photographed with light microgra- phy using a Zeiss SV-8. Amber was ground to within 3-4 mm on all sides of the inclu- sion. A groove ca. 1.5 mm thick was circum- scribed around the mid-sagittal line of the specimen (while observing under 10 mag- nification), using a 2 in. diameter emory wheel on a motorized flexible shaft (Dremel, Inc.). The groove formed a circle less than 1 mm from the inclusion. Fine powder from the cutting was blown away with compressed air. A sharp, pointed X-Acto blade was used to carefully score the internal edge of the groove, even closer to the inclusion, until slight lev- erage between the two halves ofamber caused the piece to split, generally along the axis of the groove and through the middle of the inclusion. Very little particulate debris from

GRIMALDI ET AL.: AMBER FOSSILS 5

the cutting contaminated the preparation and no abrasion of the internal parts occurred.

For observation with the SEM, the intact end of the amber piece was mounted to a stub using liberal amounts of silver paint. A 5A coating of gold-palladium was applied. Amber has excellent insulative properties, as best seen when it gathers static charges upon rubbing (hence the Greek name for amber, elektron). Thorough grounding will reduce electron charging, although additional pre- cautions were made using low voltage (2-3 kV). Use of a Zeiss DSM (Digital Scanning Microscope) 950 with electron-collection en- hancement features allowed study at such low kVs with little loss of resolution.

To observe the ultrastructure of tissues and cells with the TEM, two specimens of the extinct Dominican amber bee, Proplebeia dominicana, were freshly opened (this is one of the most common inclusions in this am- ber). Bundles of the longitudinal median dor- sal muscles were carefully lifted from the tho- rax using hand-sharpened watchmaker’s forceps, with the tips first washed in alcohol and dried. Brain tissue was removed from the head in the same way. Tissue specimens were immediately stored in dry Eppendorf tubes and express shipped from New York to Baltimore (1 day delivery). They were not frozen or refrigerated. Numerous possible “controls” could be used to compare to the amber specimens: air dried and freshly fixed specimens; and specimens stored for varying amounts of time in tree saps, various fresh resins, Canada Balsam, etc. ... That is a fu- ture project beyond the scope of the present one; it would address the chemical factors responsible for the preservation we report here.

Samples were prepared for transmission electron microscopy by two protocols. Some were embedded directly in LR White resin (Newman et al., 1983; Polysciences Inc., Warrington, PA) by immersing the tissue in unpolymerized LR White for at least 8 hours. A second set of tissue samples was rehydrated and dehydrated before embedding. These samples were rehydrated overnight at room temperature in Ruffer’s solution (30% etha- nol, 1% NaCO, in distilled water [Lewin, 1967]). Following rehydration, the samples were dehydrated by progressive 60 minute

6 AMERICAN MUSEUM NOVITATES

incubations in 30, 50, 75, and 90% ethanol in distilled water. The final dehydration was carried out by three 20 minute incubations in 100% ethanol followed by one 20 minute incubation in 50% ethanol/50% LR White. The samples were then infiltrated in 100% LR White for at least 8 hours. Each set of embedded samples was polymerized at 50°C overnight in fresh resin in gelatin capsules. Ultrathin (90 nm) sections were cut from the polymerized blocks on a Reichert-Jung Ul- tracut E microtome using a Diatome 35° an- gle compression-free knife, then transferred to Formvar-coated EM grids and allowed to air dry. The sections were stained with 1% OsO, and uranyl acetate. All samples were viewed and photographed on a Zeiss TEM 10A transmission electron microscope at 60 kV.

To test for pollen viability we attempted to induce pollen tube growth on the following medium: | ml. of ‘‘mother’” solution (0.1 mg/ml boric acid, 0.3 mg/ml CaNQO,, 0.2 mg/ ml MgSO,, 0.1 mg/ml KNO, dissolved in sdH,O), which was added to: 9 ml sdH,O, 9 g sucrose, 0.8 g gelatin. Pollen controls were from fresh petunia flowers. Negative controls were made by microwaving the petunia pol- len on high for 2 minutes, which rendered the pollen inviable. Four anthers of Hymen- aea in Dominican amber were cut open and the tissues scraped out using an ethanol- cleaned pin. Controls and experimental sam- ples were incubated on the culture medium for 72 hours at room temperature and ex- amined for pollen tube growth. Experimental samples (from amber) were tested in an area of the lab separate from the controls, to re- duce any possible contamination.

Mounted specimens were retained on SEM stubs and are deposited in the amber fossil collection of the AMNH, should future ex- amination be of interest. Each has a catalog number, to which we refer in the text and the figures. The following references were con- sulted for interpreting morphological and cy- tological structures; terminology derives from these references: Bold (1973); Chapman (1982); Smith (1968); Snodgrass (1925, 1935). The SEM plates are arranged according to the specimen and type of organism, even though the following discussion is presented by types

NO. 3097

of tissues. TEM plates are presented last, but ultrastructure is discussed under sections concerning the appropriate tissues.

THE AMBERS

Amber from the Dominican Republic and the Baltic were chosen because of the differ- ences in age and chemistry, the latter feature of which is well documented. Dominican am- ber was unquestionably formed from an ex- tinct species of the living genus Hymenaea (Leguminosae) (Langenheim, 1969; Hueber and Langenheim, 1986). Although cited often by some as being Eocene in age (e.g., Poinar, 1992), the available evidence on stratigraphy suggests an origin around the Oligo-Miocene boundary for the oldest deposits, and Hen- wood (1992b) cautioned against an Eocene age. Hymenaea trees today produce copious amounts of resin, containing arrays of ses- quiterpene hydrocarbons and diterpenoid resin acids (Langenheim, 1981; 1990).

Baltic amber, by contrast, derives from a conifer, possibly a pine (Pinus) or close rel- ative, but probably an araucarian (Gough and Mills, 1972; Mills et al., 1984). It has a par- ticularly high concentration of resin acids, such as succinic and communic acids (Gough and Mills, 1972). The chemical differences between Dominican and Baltic amber are readily observable via the external preser- vation of the insects in them. Insects in Do- minican amber are typically perfectly pre- served, whereas Baltic amber specimens often have a milky coating on the body. This milk- iness is due to a froth of microscopic bubbles (Mierzejewski, 1978), presumably the prod- ucts of microbial decomposition and/or au- tolysis of internal tissues.

Much older amber fossils from the Turo- nian (mid-Cretaceous, ca. 90-94 myo) of New Jersey and the Neocomian of Lebanon (ca. 125 myo) are in the AMNH collections, but were not used for these tissue studies. The paleontological value of this material is great- er than that of the Tertiary material, and it is difficult (or impossible) to replace. Any studies causing partial or entire destruction of a Cretaceous specimen should be done only after exhaustive morphological study on a large series of specimens of one species.

1994

INSECT PRESERVATION

Cuticular Structures. Preservation of cu- ticular structures (external as well as invagi- nated into the body) would not seem sur- prising in amber fossil insects, given the intricately preserved external detail of in- sects. The natural positions of structures, and the detail of their preservation, however, were unexpected. In the stingless bee (specimen B1), the row of imbricate plates on the glossa (or tongue) was found, with a fringe of fine hairs along its left edge—the hairs function in imbibing nectar (fig. 6). Also seen in the stingless bees (B1, B3) was a curious detail of the mesothoracic phragma (a paddle- shaped apodeme). It possessed a geometric pattern of eight cells, each having a finer pat- tern of roughly hexagonal cells on it (figs. 17, 18). These must be imprints of the muscle bundles, fibers, and myofibrils. Similar geo- metric patterns are observed on insect egg chorions, which are the imprints of the fol- licle cells (Hinton, 1981); and the epicuticle of various insects bears the hexagonal im- prints of epidermal cells. It has always been assumed that the original cuticle was intact in amber fossil insects, but the degree of pre- served detail and molecular composition has not been examined. In one of the scuttlefly specimens, the reverse and obverse images of the antennae showed that actual sensilla and setulae can be preserved not just as im- prints (figs. 26, 27). This specimen clearly shows how the actual material is separated from and smaller than the amber “cast.” Since it is unlikely that the cuticular surface would shrink, even during dehydration, it is more likely that the cast surface represents expan- sion of the amber, probably due to polymer- ization of the original resin. How much orig- inal chitin remains in the “amberized”’ insect cuticle has yet to be determined. Insect cu- ticle is composed of 25-40% chitin, which is a B-pleated polysaccharide sheet intercalated with proteins. Miller et al. (1993) found that the cuticular remains of Pleistocene beetles buried in sediments (ca. 15,000 years old) contained half the amount of chitin that would be expected from fresh material.

Fine structural details of the original cuticle can also be seen by the microtrichiae pre-

GRIMALDI ET AL.:

AMBER FOSSILS 7

served on the wing membrane of a dolicho- podid fly in Baltic amber (fig. 56). A row of six minute campaniform sensilla occur on the wing vein of a mycetophilid midge, also in Baltic amber (fig. 61).

Musculature. Insects have a complex mus- culature, possessing approximately twice the number of muscles as do mammals (Snod- grass, 1935). Much of this musculature is in the thorax and serves to power the legs and wings. Commonly seen was preservation of the longitudinal median dorsal muscles of the thorax; in stingless bee specimen B1 (fig. 4) it is seen as closely joined bundles forming a large sheet occupying most of the thorax. This muscle mass is completely intact. Smaller muscles are also well preserved, with their insertions and origins intact. A bundle of 10— 11 muscles is attached to the mesothoracic phragma and postnotal wall of stingless bee B1 (fig. 10). In B3, the slender oblique lat- erals, oblique intersegmentals, and portions of longitudinal median dorsal muscles were seen (fig. 17). Small muscle fibers attached to the membranous sucking pump (pharynx) of stingless bee B1 showed transverse striations at higher magnification (fig. 8). A compact bundle of small muscles was found dislodged from the mostly empty thorax of a myceto- philid midge in Baltic amber (fig. 55). Not all muscles appeared so intact: some thoracic muscles of platypodid beetle P1 were very fibrous and shredded. Shrinkage of muscles was not discerned in any specimen that pos- sessed them, although the muscle and other soft tissue was always a very dark red or tar black. Causes of the discoloration are un- known.

At the ultrastructural level, native Prople- beia dominicana flight muscle is easily iden- tifiable, with striking patterns of repeated sar- comeres present in longitudinal sections of all myofibrils (fig. 69). The Z-and M-lines are present as spaces between the electron-dense remains of the thick filaments in the A-band. Higher magnification reveals almost no fi- brillar appearance to the A-bands, suggesting that the electron-dense material present there is composed primarily of inorganic salts de- posited on thick filaments during dehydra- tion (figs. 70-71). Strips of electron-dense mi- tochondria separate the myofibrils. The

8 AMERICAN MUSEUM NOVITATES

mitochondria have well preserved, densely packed cristae characteristic of insect flight muscle (fig. 72).

Membranous structures in flight muscle are generally better preserved than proteinaceous ones. In addition to mitochondria, tracheoles are present, and the membranes of epithelial cells lining these tracheoles are apparent (fig. 73). In some sections, T-tubules are pre- served (fig. 74). Extensive membranes are of- ten found superficial to the muscle cells them- selves (fig. 75).

Samples that were rehydrated prior to em- bedding are essentially composed of strips of mitochondria, separated by spaces with al- most no electron-dense material (fig. 76). The material composing the sarcomeres is almost completely extracted by the rehydration step. We suspect that the extracted material is composed of inorganic salts, with perhaps some remaining proteins. It is possible that fixation with glutaraldehyde during rehydra- tion might preserve some proteinaceous structures like sarcomeres while extracting inorganic salts.

Several flight muscle samples dissected from Proplebeia dominicana had an addi- tional tissue structure attached to the surface of the muscle bundle itself (fig. 77). We sus- pect this tissue to have connected the flight muscle to the axillary sclerites of the wing, since cuticle is often found attached (fig. 78). This tissue consists primarily of amorphous fibrillar material (probably collagen fibers) interposed with small electron-dense parti- cles (figs. 79-80). We do not know the origin of these particles, which may be nuclei, cells that have been extensively distorted by the rehydration process, or precipitated inorgan- ic salts. In any case, the fibrillar matrix is apparently quite well preserved, and is an attractive candidate tissue for isolation of fossil polypeptides.

Nervous Tissue. Nervous tissue is notori- ously difficult to preserve well. It was very surprising, therefore, to find several speci- mens with the brain (protocerebrum) intact (stingless bee B1 [figs. 5, 9]; phorid fly [fig. 25], platypodid beetle, P1 [fig. 30]), and my- cetophilid midges [fig. 59]). In the stingless bee and platypodid the tissue was loosely fi- brous, indicating loss of some interstitial ma- terial, although there was virtually no shrink-

NO. 3097

age of the entire structure. In the other specimens this structure was dense and amor- phous.

Brain tissue removed from a specimen of Proplebeia dominicana was embedded and sectioned for TEM and was rather poorly pre- served, as expected. Some histological fea- tures are still observable, however. Most ar- eas of the sections are covered with a dense outer layer of electron-dense salts (fig. 81). Notable exceptions to this are regions of con- voluted membranes, which we suspect to be the remains of the major axon tracts of the central nervous system (fig. 81).

In brain tissue that was rehydrated before embedding, some other interesting details be- came apparent after extraction of the inor- ganic salts. In some areas, extensive tracts of parallel membranes surround patches of elec- tron-dense cytoplasmic remains (figs. 82, 83). We believe these to be membranes of oli- godendrocytes wrapping large axons. Anoth- er possibility is that these membranes are ar- tifacts occasionally seen in TEM of necrotic tissue (Commonly called “myelin figures’’), caused by extraction of lipids from dying cells. Cytoplasm of neural cells is generally poorly preserved, with almost no cytoplasmic de- tails. The putative axon tracts observed in brain sections are also preserved after rehy- dration, although the membranes appear swollen and distended compared to native samples (fig. 84).

Membranous Structures. These include portions or most of the digestive, excretory, and reproductive systems, as well as the dor- sal aorta. In several specimens, but seen best in B1 (figs. 11, 12), were the folds of fine membranous air sacs, which occupy the pos- terodorsal half of the thorax in bees. Unlike the other parts of the respiratory system (tra- cheae and tracheoles), these membranes have no trusswork of minute, chitinous rings (taen- idia). Excellent examples of preserved tra- cheae are seen in a Baltic amber mycetophilid (figs. 67, 68). The tracheae are hardly col- lapsed. Cross sections of tracheoles were seen in TEM thin sections of Proplebeia bee flight muscle from Dominican amber (fig. 73).

Portions of digestive tracts of several spec- imens were remarkably intact. In one my- cetophilid the ventriculus possessed a geo- metric pattern of folds on the outer surface,

1994

each fold forming a cavity with a papilla in it (figs. 21, 22). The papilla is probably a crypt, perhaps of regenerative cells. A portion of the thin esophagus was seen in one termite (specimen T1, not figured); and a more ex- tensive portion of the foregut, including the crop, was found in stingless bee B2 (figs. 16, 17). In one platypodid beetle, an extensive portion of the gut in the abdomen was ob- served, which included the entire midgut and hindgut (fig. 38). The region where the fine malpighian tubules occur had an odd pres- ervation, which was an amorphous arrange- ment of stacked mounds; holes; and flat, hex- agonal crystals (fig. 41). Other parts of the midgut showed unusual formations of deep, fine pleats (fig. 39). The detail of membra- nous tissue preservation is best seen from an unidentified structure from a termite thorax (fig. 24). The tissue is very thin with one mar- gin frayed, showing loose fibers.

Other organisms. Intimately associated with arthropods in amber are a host of or- ganisms, such as parasites and symbiotic mi- crobes, as well as evidence of other ecological relationships. It is not uncommon, for ex- ample, to find stingless bees in amber with clumps of pollen adhering to the hairs on the hind legs and abdomen. One bee specimen (B1) was examined that had a clump of pollen on the ventral side of the abdomen. The exine was intact enough to observe that two dis- tinctly different pollen types were collected (fig. 13), indicating that the bee had been vis- iting two kinds of flowers. The exine of the most common pollen type is very finely pit- ted and. quite similar to that of living Hy- menaea (Langenheim and Lee, 1974) as well as to the amber Hymenaea (fig. 48); the other type has coarse sculpturing (fig. 14).

Some wood-boring beetles feed not on the wood, but on the fungus cultures that they inoculate in the galleries. Ambrosia beetles (families Scolytidae and Platypodidae) are quite common in the Dominican amber; no doubt they were living in the amber tree itself, since amber pieces are occasionally found with beetle galleries in them. The beetles transmit the fungus via specialized struc- tures, the mycangia (Batra, 1963). Mycangia are pockets of invaginated cuticle that occur in various insects, mostly beetles, that harbor inoculum of symbiotic fungi. The beetles feed

GRIMALDI ET AL.: AMBER FOSSILS 9

on the fungi and also serve as the main dis- persal agent. Mycangia vary tremendously in the Coleoptera, with structures occurring on the mandibles, head, elytra, abdomen, and thorax (Crowson, 1981). Perhaps no fungal- beetle symbiotic relationship is more spe- cialized and intimate than that of ambrosia beetles and their ambrosia fungus (subclass Hemiascomycetidae). The various ambrosia fungi are known only from the beetle’s my- cangia and galleries and are specific to the species of beetle, not the host tree (Batra, 1963).

Specificity of the ambrosia fungus to the beetle may be due to the apparent glandular nourishment that the fungus receives in the mycangium. In fact, the shape and location of saclike mycangia in the Scolytidae and Pla- typodidae are often species-specific (Batra, 1963; Francke-Grosmann, 1956). Batra clas- sified five categories of mycangia in these bee- tles based on their location on the body.

In an unidentified platypodid beetle (P2), two large ventral mycangia were found on the insect’s left side (mycangia occur in pairs; see fig. 33). The anterior one was largest, ap- proximately 200 um long, drop-shaped, and was lying between the meso- and metathorax (fig. 34). The posterior mycangium was round and lying between the metathorax and ab- dominal sternite 1 (figs. 35). The size and general location correspond with other re- ports based on living species (cf. figs. 5, 16, and 17 in Batra [1963]). Both mycangia were replete with spores and conidiophores (fig. 36). Viability of the spores has not been test- ed.

PLANT PRESERVATION

Leaves typically have a heavily cutinized (waxy) epidermis, beneath which there is a layer of mesophyll. The mesophyll is the pho- tosynthetic layer of cells, and comprises of a spongy layer of squamous cells and a palisade layer of columnar cells. In a small leaf of Hymenaea in Dominican amber, an entire surface was exposed, which was charcoal- black, finely cracked, and crumbling (fig. 44). Close examination, however, revealed that the palisade layer of mesophyll was intact in places: columnar cells were easily seen and showed no distortion (fig. 45).

10 AMERICAN MUSEUM NOVITATES

The anthers of Hymenaea are, not surpis- ingly, rather common in Dominican amber. According to Jean Langenheim (personal commun.), during the pollination period the area beneath a Hymenaea tree is covered with dehisced stamens (the anthers do not fall from the filaments). Seven specimens were ex- humed: four for testing pollen viability, three for study under the SEM. In five of the an- thers the material inside had a black, gummy appearance, but was actually brittle and very dry (e.g., figs. 46, 47). Under the SEM this material was amorphous, and presumably is a dried film lying over pollen and tissue. In anther specimen Al most of the pollen was not clearly visible because it was obscured by this apparent film. Some pollen grains that were dispersed into the amber were found to have the exine less obscured (this anther was chosen for study, in fact, because it was de- hiscent and apparently mature). The exine of the pollen of the extinct Dominican amber tree, Hymenaea protera, has a surface of dense, fine pits very similar to that of Hy- menaea courbaril and H. verrucosum (fig. 48; cf. fig. 4 in Langenheim and Lee [1974]). This is also very similar to the pollen found on stingless bee specimen B-1 (fig. 14). In the other two anthers (1 dehiscent, the other not; see figs. 49-51) there was completely different preservation. Internal tissues were not a tarry black, but chalky and brown. Extensive mats of dense, fine, hairlike structures filled most of the anther, but no pollen grains were ap- parent. These structures are probably colum- nar epithelial cells. In the test for viability, pollen tubes grew on all the positive control samples. As expected, no tubes were found in the negative controls or in the experimen- tal samples removed from the Hymenaea an- thers in Dominican amber. Despite the de- hydration properties of amber, enzymes and other labile molecules in the pollen probably undergo autolysis.

CONCLUSIONS

The observations reported here and else- where reveal tissues, cells, and cellular ultra- structure in amber insects and plants, with a startling lifelike fidelity. In two of the four small (< 4 mm body length) Baltic amber flies that were examined here, the thoracic

NO. 3097

and abdominal cavities were largely vacant of tissue, but internal tissues and organs of the other Baltic amber flies were as intact as is typically found in the Dominican amber insects. Thus, the preservative qualities of ambers are not equivalent, although the tis- sue which did remain in the Baltic amber flies was as well preserved under the SEM as that in Dominican amber flies (TEM of Baltic am- ber fly muscle, for example, was not done in our study). In the insects with preserved tis- sues, the organs showed little or no shrinkage, contrary to the observations of Henwood (1992b) who reported up to 50% shrinkage of muscle from a Dominican amber fly. The general lack of shrinkage or autolysis, and preservation of such delicate structures as air sac membranes and brain tissue, indicate a very rapid mummification, which tempts ex- planations on the possible chemical process involved.

The best scenario we can provide to ex- plain such apparently rapid and thorough cel- lular fixation is that the most volatile, low molecular weight fractions (mono- and ses- quiterpenes) in the original resin readily dif- fused through intact body walls and perfused the tissues. For insects, the intersegmental membranes would be an important area of diffusion (even though they are thinly cutic- ular and possess a waxy layer, the interseg- mental membranes are the thinnest parts on an arthropod). These volatile fractions must have replaced the cellular water. This can be seen in many amber fossilized insects: often there is a transparent, light brown ‘‘halo” around them, which must be aqueous com- ponents of the body fluid sequestered by the surrounding resin. Monosaccharides, alco- hols, aldehydes, and esters also occur in res- ins (Langenheim, 1990), and a few investi- gators have implicated a role for at least some of these in “‘amberization.”’ Currently, stud- ies have begun to examine the infiltration of the volatile sesquiterpene hydrocarbons and some oxygenated forms into insect tissue. The role of monosaccharides and other com- pounds is probably much less significant than that of terpenes, because they occur in much lower concentrations, and (with the exception of alcohols) would perfuse tissues more slow- ly than volatile terpenes. Henwood (1992a), for example, mocked a sap flux with maple

1994

syrup, in which flies were preserved and their muscles later thin sectioned. Muscles in the syrup-preserved flies were not nearly as well preserved as the muscles of amber insects reported here and by Henwood, and the sugar concentration of the syrup is much higher than any found in natural plant exudates, let alone in resins.

The ultrastructural results prompt ques- tions as to the actual molecular preservation that has occurred. What proteins, if any, are present? In what state might they be? Im- munological evidence suggests preservation of protein tertiary structure in brachipod shells that are up to 4 million years old (but no older) (Collins et al., 1991), and glycoproteins from an 80 million year old mollusk shell were reported (Weiner et al., 1976). The den- sity of mollusk and brachiopod shells may uniquely provide for the preservation of such macromolecules, particularly since the mode of fossilization in marine sediments would seem less than ideal for protein preservation. Contrary to this are the plant fossils from the fine, dense, anoxic clays of the mid-Miocene of Clarkia, Idaho. Chloroplast preservation is reported from leaves found there (Niklas et al., 1985), as well as two cases of DNA, from a magnolia (Golenberg et al. 1990; Go- lenberg, 1991) anda bald cypress (Taxodium; see Soltis et al., 1992). However, Logan et al. (1993) found that preservation of biomole- cules is highly selective in Clarkia fossils, with no evidence of polysaccharides, polyesters, or proteins, but only lignins and an aliphatic biopolymer present. Logan et al. questioned the DNA results from the Clarkia plant fossils and, indeed, DNA preservation in them is hardly consistent: among “hundreds” of ex- tractions, these two published examples and several unpublished ones are the only suc- cesses (P. and D. Soltis, personal commun., 1992; E. Golenberg, personal commun. 1993). Quality of protein preservation is dependent on how one analyzes the molecule. Amino acid sequencing might reveal short chains of a-amino acid units, a result of peptide bond hydrolysis, and it would be useful only if hy- drolysis occurs preferentially in some bonds.

GRIMALDI ET AL.:

AMBER FOSSILS 11

Or, side groups such as amino and carboxyl groups could be lost. If analytical methods are dependent on tertiary structure for de- tecting protein preservation, one would also be measuring the degree to which H bonding and crosslinking are still intact (which varies with the kind of protein). Ambler and Daniel (1991) reviewed successful extractions of protein from ancient materials.

There is little doubt that amber will pre- serve DNA more consistently than any other kind of fossil. Four examples of DNA from insect and plant tissues in amber have been published, three of them from Dominican amber: 1. A large, primitive termite, Mas- totermes electrodominicus (DeSalle et al., 1992); 2. A stingless bee, Proplebeia domin- icana (Cano et al., 1992); 3. A leaf of the amber tree, Hymenaea protera (Cano et al., 1993b: an unrefereed report in which se- quences were not presented); and 4. A ne- monychid weevil in 125 million year old Leb- anese amber (Cano et al., 1993a). Three other, unpublished successes are known thus far, all concerning Dominican amber: a drosophilid fruit fly (Y. Shirota, Hirosaki University); an anisopodid woodgnat, Valeseguya disjuncta (DeSalle and Grimaldi, unpubl.: AMNH); and several chrysomelid beetles (B. Farrell, per- sonal commun.: Univ. Colorado). Such con- sistent preservation of DNA must be attrib- utable to the unique chemistry of resins. The fact that the primary structure of DNA re- mains reasonably intact in amber fossils (ap- proximately 250 base pair segments on av- erage), indicates an ability of the amber to preserve the phosphate-ester bond between the deoxyribose sugar units. This bond is the one most susceptible to hydrolysis (Eglinton and Logan, 1991; Lindahl, 1993), and points mostly to the dehydration properties of the resin.

What proteins remain, and in what state, is the subject for future studies. Regardless of this, all evidence still indicates that amber very consistently and exquisitely mummified the small organisms that became entombed in the ancient resin.

12 AMERICAN MUSEUM NOVITATES NO. 3097

Figs. 1-6. Stingless bee, Proplebeia dominicana (B1) in Dominican amber. 1: Intact specimen, show- ing cut circumscribed around mid-saggital line. 2: Freshly opened specimen (light micrograph). 3, 4: Opposite halves of entire bee. 5: Detail of head. 6: Detail of glossa. Scales: 3,4: 500 um; 5: 100 um; 6: 20 «wm.

1994 GRIMALDI ET AL.: AMBER FOSSILS 13

Figs. 7-12. Details of stingless bee (B1). 7: Sucking pump with attached muscles. 8: Detail of muscle (note transverse striae). 9: Protocerebrum (brain). 10: Bundle of small muscles in thorax. 11: Thorax. 12: Detail of thorax, showing membranous air sacs. Scales: 7: 50 nm; 8: 10 wm; 9: 50 wm; 10: 20 um; 11: 200 wm; 12: 100 um.

14 AMERICAN MUSEUM NOVITATES NO. 3097

Figs. 13-18. Stingless bees, P. dominicana, in Dominican amber. 13: Clump of pollen from abdomen of B1. 14: Detail of one pollen grain. 15, 16: Specimen B2. 15: Head and thorax, lateral view. 16: Detail of esophagus and crop seen in 15. 17, 18: Specimen B3. 17: Transverse section through thorax. 18: Cell imprints in cuticle of phragma seen in 17. Scales: 16, 17: 200 um; 18: 5 um.

1994 GRIMALDI ET AL.: AMBER FOSSILS 15

Figs. 19-22. Female fungus gnat (Mycetophila sp.: Mycetophilidae) in Dominican amber (M1). 19: Habitus of intact specimen. 20: Lateral view of entire, opened specimen. 21: Portion of the membranous ventriculus. 22: Detail of ventriculus. Scales: 20: 500 wm; 21: 20 um; 22: 5 wm.

16 AMERICAN MUSEUM NOVITATES NO. 3097

Figs. 23-27. Insects in Dominican amber. 23, 24: Isoptera (termite), Reticulitermes sp. (T1). 23: Thorax, showing unidentifiable muscle bundles. 24: Detail of the edge of a sheet of membrane. 25-27: Fly, Megaselia sp. (Phoridae). 25: Head. 26, 27: Antenna, opposite halves. Note the actual structures and corresponding impressions. Scales: 23: 200 um; 24: 10 um; 25: 50 wm; 26, 27: 20 um.

1994 GRIMALDI ET AL.: AMBER FOSSILS 17

Figs. 28-31. Platypodid beetle in Dominican amber (P1). 28: Entire, intact specimen. 29: Entire

specimen, opened. 30: Protocerebrum (brain). 31: Detail of fibrous tissue in or on brain. Identity of the tissue is uncertain—it does not resemble nervous tissue. Scales: 29: 500 um; 30: 100 wm; 31: 20 um.

18 AMERICAN MUSEUM NOVITATES NO. 3097

Figs. 32-36. Platypodid beetle in Dominican amber (P2). 32: Entire, intact specimen. 33: Entire specimen, opened. 34: Detail of head and part of thorax, showing anterior mycangium. 35: Detail of posterior mycangium. 36: Detail of fungal spores and conidia in mycangium. Scales: 33: 200 um; 34: 100 wm; 35: 50 wm; 36: 10 um.

1994 GRIMALDI ET AL.: AMBER FOSSILS 19

ie

Figs. 37-42: Platypodid beetle in Dominican amber (P3). 37: Entire beetle (lateral), opened. 38: Apex of abdomen separated from amber cast. 39: Wall of ventriculus, showing group of fine, deep pleats. 40: Gut. 41: Area near pylorus, showing amorphous structures and some crystals, but not malpighian tubules. 42: Detail of amorphous structures near pylorus. Scales: 37: 500 wm; 38: 100 wm; 39: 10 wm; 40: 200 um; 41: 50 wm; 42: 10 wm.

20 AMERICAN MUSEUM NOVITATES NO. 3097

1994 GRIMALDI ET AL.: AMBER FOSSILS 21

Figs. 46-51. Anthers of Hymenaea protera in Dominican amber. 46, 47: Opposite halves of specimen Al. This specimen was clearly dehiscent and dispersing its pollen. 48: Detail of pollen released near edge of anther. The pollen was not viable (see text). 49: Portion of extensive matting from specimen A2, which was an immature anther. 50, 51: opposite halves of anther A2. Scales: 46, 47: 500; 48: um; 49: 10 um; 50, 51: 200 um.

22 AMERICAN MUSEUM NOVITATES NO. 3097

a

Figs. 52-56. Dolichopodid fly in Baltic amber (D1). 52, 53: Opposite halves of whole specimen, opened. The bright area at the ventral part of the thorax is the milky impurity, which is due to numerous fine bubbles. The thorax was largely empty. 54: Head, partly exposed showing original cuticle, and globule within (presumably the dried remains of liquified decomposition). 55: Bundle of fine muscles dislodged from the thorax. 56: Detail of original microtrichiae on wing. Scales: 52, 53: 200 um; 54: 100 um; 55: 50 wm; 56: 10 um.

1994 GRIMALDI ET AL.: AMBER FOSSILS 23

Figs. 57-62. Mycetophilid fungus gnat (Mycetophila sp.) in Baltic amber (M-B1). 57, 58: Opposite halves of entire, opened specimen. The thorax is largely hollow. 59: Detail of head. 60: Detail of bases of mid and hind legs. Original cuticle is present, but no remnants of soft tissue remain. 61: Detail of wing membrane pressed into surface of amber, showing line of sensilla on vein (arrows). 62: Male genitalia, showing internal sclerites and apodemes, but not muscle bundles. Scales: 57, 58: 500 um; 59: 50 wm; 60: 100 um; 61: 20 um; 62: 100 wm.

24 AMERICAN MUSEUM NOVITATES NO. 3097

mia

imen, opposite halves. 65: Detail of thorax. 66: Detail of head. 67: Detail of thorax, showing tracheae and atrium. 68: Detail of

another trachea. Scales: 63, 64: 500 um; 65: 200 um; 66: 50 wm; 67: 100 wm; 68: 50 um.

1994 GRIMALDI ET AL.: AMBER FOSSILS 25

ay

Figs. 69-72. TEM thin sections of Proplebeia flight muscle from Dominican amber. 69: Myofibrils with sarcomeric repeats (2200 x). 70, 71: Detail of fig. 69. (70: 16,000 x; 71: 35,000 x). There is a lack of fibrillar material in the dark banded areas, suggesting replacement by inorganic salts. 72: Detail of

mitochondria tightly packed among myofibrils. Finger print patterns are the internal cristae of the mitochondria.

26 AMERICAN MUSEUM NOVITATES NO. 3097

ae

Figs. 73-76. TEM thin sections of Proplebeia flight muscle and associated tissue in Dominican amber. 73: Tracheoles and associated epithelial cells lying between myofibrils (2500 x ). 74: T-tubules (11,000 x). 75: Extensively folded membranes superficial to muscle cells (10,000 x). 76: Rehydrated muscle tissue (52,000 x). Note that the sarcomeres are almost completely extracted, suggesting these are preserved mostly as inorganic salts.

1994 GRIMALDI ET AL.: AMBER FOSSILS 27

Figs. 77-80. TEM thin sections of a connective tissue attached to flight muscle of Proplebeia in Dominican amber. 77: Whole section (10,000 x). 78: Portion with cuticle attached (32,000 x). 79, 80: Details of tissue, with amorphous fibrillar material (collagen fibers?) and electron-dense structures, which are possibly distorted cells, nuclei, or salt crystals (79: 12,000 x; 80: 30,000 x).

28 AMERICAN MUSEUM NOVITATES NO. 3097

Figs. 81-84. TEM thin sections of brain tissue from Proplebeia in Dominican amber. 81: Whole section, showing nervous tissue embedded in a thick layer of electron-dense salts (3500 x ). 82-84: Brain tissue after rehydration, which extracts the inorganic salts. 82, 83: Extensively laminated membranes, perhaps of oligodendrocytes surrounding large axons (82: 22,000 x; 83: 80,000 x). 84: Swollen mem- branes surrounding putative axon tracts (3500 x).

1994

GRIMALDI ET AL.: AMBER FOSSILS 29

REFERENCES

Allison, P. A., and D. E. G. Briggs

1991. The taphonomy of soft-bodied animals. pp. 120-140, Zn S.K. Donovan, The processes of fossilization. New York: Columbia Univ. Press.

Ambler, R. P., and M. Daniel 1991. Proteins and molecular paleontology.

Philos. Trans. R. Soc. London (B) 333: 381-389.

Bandel, K., and N. Vavra

1981. Ein fossiles Harz aus der Unterkreide Jordaniens. Neues Jahrb. Geol. Palaont. Mk. 1: 19-33.

Baroni-Urbani, C. and S. Graeser

1987. REM-Analysen an einer pyritisierten Ameise aus Baltischen Bernstein. Stutt- gart Beitr. Naturk. (B) 133: 16 pp.

Batra, L. R.

1963. Ecology of ambrosia fungi and their dis- semination by beetles. Trans. Kansas Acad. Sci. 66: 213-236.

Bold, H. C.

1973. Morphology of plants, 3rd ed. New York: Harper and Row. 668 pp.

Briggs, D. E. G.

1991. Extraordinary fossils. Am. Sci. 79: 130- 141.

Briggs, D. E. G., and A. J. Kear

1993. Fossilization of soft tissue in the labo- ratory. Science 259: 1439-1442.

Cano, R. J., H. N. Poinar, and G. O. Poinar, Jr.

1992. Isolation and partial characterisation of DNA from the bee Proplebeia domini- cana (Apidae: Hymenoptera) in 25-40 million year old amber. Med. Sci. Res. 20: 249-251.

Cano, R. J., H. N. Poinar, N. J. Pieniazek, A. Acra, and G. O. Poinar, Jr. 1993a. Amplification and sequencing of DNA

from a 120-135-million-year-old wee- vil. Nature 363: 536-538.

Cano, R. J., H. Poinar, and G. O. Poinar, Jr.

1993b. DNA from an extinct plant. Nature 363: 677.

Chapman, R. F.

1982. The insects: structure and function. Cambridge, MA: Harvard Univ. Press.

Cockburn, A., and E. Cockburn

1980. Mummies, disease, and ancient cul- tures. Cambridge: Cambridge Univ. Press, 340 pp.

Collins, M. J., G. Muyzer, P. Westbroek, G. B. Curry, P. A. Sandberg, S. J. Xu, R. Quinn, and D. MacKinnon 1991. Preservation of fossil biopolymeric

structures: Conclusive immunological

evidence. Geochim. Cosmochim. Acta 55: 2253-2257.

Crepet, W., K. C. Nixon, E. M. Friis, and J. V. Freudenstein 1992. Oldest fossil flowers of hamamelida-

ceous affinity, from the Late Cretaceous of New Jersey. Proc. Nat. Acad. Sci. 89: 8986-8989.

Crowson, R. A.

1981. The biology of the Coleoptera. New York: Academic Press.

Curry, A., C. Anfield, and E. Tapp 1979. Electron microscopy of the Manchester

mummies. In Manchester Museum Mummy Project, pp. 103-111. Man- chester: Manchester Univ. Press.

David, R., and E. Tapp (eds.)

1984. Evidence embalmed: modern medicine and the mummies of ancient Egypt. Manchester: Manchester Univ. Press.

DeSalle, R., J. Gatesy, W. Wheeler, and D. Gri- maldi 1992. DNA sequences from a fossil termite in

Oligo-Miocene amber and their phylo- genetic implications. Science 257: 1933- 1936.

Eglinton, G. and G. A. Logan

1991. Molecular preservation. Philos. Trans. R. Soc. London (B) 333: 315-328.

Francke-Grosmann, H.

1956. Hautdriisen als Trager der Pilzsymbiose bei Ambrosiakafern. Z. Morphol. Oek- ol. Tiere 45: 275-308.

Golenberg, E. M., D. E. Giannasi, M. T. Clegg, C. J. Smiley, M. Durbin, D. Henderson, and G. Zurawski 1990. Chloroplast DNA from a Miocene Mag-

nolia species. Nature 344: 656-658.

Golenberg, E. M.

1991. Amplification and analysis of Miocene plant fossil DNA. Philos. Trans. R. Soc. London (B) 333: 419-427.

Gough, L. J. and J. S. Mills 1972. The composition of succinite (Baltic

amber). Nature 239: 527-528.

Grimaldi, D. (ed.)

1990. Insects from the Santana Formation, Lower Cretaceous, of Brazil. Bull. Am. Mus. Nat. Hist. 195: 191 pp.

Hansen, J. P., J. Meldgaard, and J. Nordquist (eds.) 1991. The Greenland mummies. Washington,

D.C.: Smithsonian Institution Press.

Henwood, A.

1992a. Exceptional preservation of dipteran flight muscle and the taphonomy of in- sects in amber. Palaios 7: 203~212.

30 AMERICAN MUSEUM NOVITATES

1992b. Soft-part preservation of beetles in Ter- tiary amber from the Dominican Re- public. Palaeontology 35: 901-912.

Hinton, H. E.

1981. Biology of insect eggs. vol. 1-3. Oxford:

Pergammon Press. Hueber, F. M. and J. L. Langenheim

1986. Dominican amber tree had African an-

cestor. Geotimes 31: 8-10. Kornilowitsch, N.

1903. Has the structure of striated muscle been retained in amber fossils? Naturf. Ge- sell. zu Dorpat 13: 198-206 [in Rus- sian].

Langenheim, J. H. 1969. Amber: a botanical inquiry. Science 163: 1157-1169. Terpenoids in the Leguminosae. Jn R. M. Polhill and P. H. Raven (eds.), Ad- vances in Legume Systematics. Kew: Proc. Int. Conf. Legume Systematics. 1990. Plant resins. Am. Sci. 78: 16—24. Langenheim, J. H., and Y.-T. Lee

1974. Reinstatement of the genus Hymenaea (Leguminosae: Caesalpinioideae) in Af- rica. Brittonia 26: 3-21.

1981.

Lewin, P. K. 1967. Paleo-electron microscopy of mummi- fied tissue. Nature 213: 416-417. Lindahl, T. 1993. Instability and decay of the primary

structure of DNA. Nature 362: 709-714. Logan, G. A., J. J. Boon, and G. Eglinton 1993. Structural biopolymer preservation in Miocene leaf fossils from the Clarkia site, northern Idaho. Proc. Nat. Acad. Sci. 90: 2246-2250.

Maisey, J. G. 1991. Santana fossils: an illustrated atlas. Neptune City, NJ: T.F.H. Publications, 459 pp. Martill, D. M.

1988. Preservation of fish in the Cretaceous Santana formation of Brazil. Palaeon- tology 31: 1-18. Mierzejewski, P. 1976a. On application of scanning electron mi- croscope to the study of organic inclu- sions from the Baltic amber. Ann. Geol. Soc. Poland 46: 291-295. 1976b. Scanning electron microscope studies on the fossilization of Baltic amber spiders (preliminary note). Ann. Med. Sect. Polish Acad. Sci. 21: 6. Electron microscopy study on the milky impurities covering arthropod inclu- sions in the Baltic amber. Prace Muz. Ziemi 28: 81-84.

1978.

NO. 3097

Miller, R. F., M.-F. Voss-Foucart, C. Toussaint, and C. Jeuniaux 1993. Chitin preservation in Quaternary Co- leoptera: preliminary results. Palaeo- geogr., Paleoclimatol., Palaeoecol. 103: 133-140. Mills, J. S., R. White, and L. J. Gough 1984. Thechemical composition of Baltic am- ber. Chem. Geol. 47: 15-39. Newman, G. R., B. Jasani, and E. D. Williams 1983. Asimple post-embedding system for the rapid detection of tissue antigens under the electron microscope. Histochem. J. 15: 543-555. Niklas, K. J., R. M. Brown, Jr., and R. Santos 1985. Ultrastructural states of preservation in Clarkia angiosperm leaf tissues: impli- cations on modes of fossilization. In C. J. Smiley (ed.), Late Cenozoic history of the Pacific Northwest, pp. 143-160. San Francisco: Pac. Div. Am. Assoc. Adv. Science. Nissenbaum, A. 1975. Lower Cretaceous amber from Israel. Naturwissenschaften 62: 341-342. Paabo, S. 1985. Molecular cloning of ancient Egyptian mummy DNA. Nature 314: 644-645. Poinar, G. O., Jr. 1992. Life in amber. Palo Alto, CA: Stanford Univ. Press. 350 pp. Poinar, G. O., Jr., and R. Hess 1982. Ultrastructure of 40-million-year-old in- sect tissue. Science 215: 1241-1242. Preservative qualities of recent and fos- sil resins: electron micrograph studies on tissue preserved in Baltic amber. J. Baltic Stud. 16: 222-230. Riddle, J. M. 1980. A survey of ancient specimens by elec- tron microscopy. Jn A. and E. Cockburn (eds.), Mummies, disease, and ancient Cultures, pp. 274-287. New York: Cambridge Univ. Press. Sandison, A. T. 1955. The histological examination of mum- mified materials. Stain Technol. 30: 277-283. Schawaller, W., W. A. Shear, and P. M. Bonamo 1991. The first Paleozoic Pseudoscorpions (Arachnida, Pseudoscorpionida). Am. Mus. Novitates 3009: 17 pp. Schlee, D., and H.-G. Dietrich 1970. Insektenfuhrender Bernstein aus der Unterkreide des Lebanon. Neues Jahrb. Geol. Palaeontol. Monatsh. 1: 40—50. Schliiter, T. 1989. Neue Daten uber harzkonservierte Ar-

1985.

1994

thropoden aus dem Cenomanium NW- Frankreichs. Doc. Naturae 56: 59-70 + 6 plates.

Shear, W. A., P. M. Bonamo, J. D. Grierson, W. D. Ian Rolfe, E. L. Smith, and R.A. Norton 1984. Early land animals in North America;

evidence from Devonian age arthro- pods from Gilboa, New York. Science 224: 492-494.

Shivanna, K. R., H. F. Linskens, and M. Cresti

1991. Pollen viability and pollen vigor. Theor. Appl. Genet. 81: 38-42. Smith, DS. 1968. Insect cells, their structure and function.

Edinburgh: Oliver and Boyd. Snodgrass, R. E. 1925. Anatomy and physiology of the hon- eybee. New York: McGraw-Hill. 327 pp.

GRIMALDI ET AL.: AMBER FOSSILS 31

1935. Principles of insect morphology. New York: McGraw-Hill. Soltis, P. S., D. E. Soltis, and C. J. Smiley 1992. AnrbcL sequence from a Miocene 7ax- odium (bald cypress). Proc. Nat. Acad. Sci. 89: 449-451. Weiner, S., H. A. Lowenstam, and L. Hood 1976. Characterization of 80-million-year-old mollusk shell proteins. Proc. Nat. Acad. Sci. 73: 2541-2545. Zimmerman, M. R. 1973. Blood cells preserved ina mummy 2000 years old. Science 180: 303-304.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available

free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, N.Y. 10024.

This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).