Bulletin of the Geological Society of Finland, Vol. 80, 2008, pp. 105–124 Insect frass in Baltic amber Matti Nuorteva1) and Kari A. Kinnunen2)* 1) Alkutie 28 D, FI-00660 Helsinki, Finland 2) Geological Survey of Finland, P.O. Box 96, FI-02151 Espoo, Finland Abstract Inclusions of wood debris loosened from pine-like trees are abundant in Baltic amber of Eocene and Oligocene age. The possibilities to find insect frass and excrement among wood debris are outlined and some examples are given. Comparison with the frass and excrement produced by present-day insects provide a possibility to identify insects even though their fossils are lacking. This information can be used to characterize former forest environments. Amber forests may have also covered Southern Finland, and this possibility is discussed. Furthermore, the presence of wood debris may be utilized to recognize amber fakes, which is important for both gem trade and paleontology. It is proposed that databas- es and identification keys of frass and excrement should be constructed. Keywords: amber, fossils, inclusions, Insecta, wood, fecal pellets, identification, Poland, Lithuania *Corresponding author e-mail: [email protected] 1. Introduction The largest known deposits of fossil tree resin, amber, Pinites succinifer (Goeppert & Berendt, 1945) Con- are located along the southern coasts of the Baltic re- wents pine or pines. It resembled present-day pines gion. It is known as Baltic amber as a gemstone and like Pinus silvestris L. species. Morphological find- in the geological literature. Baltic amber was formed ings and pollen data both point to the family Pinace- as a resin during the Eocene epoch about 30 – 50 Ma ae (Weitschat & Wichard, 1998). However, Zher- ago (Ander, 1942; Krzemi|nska & Krzemi|nski, 1992; ikhin (2002) notes that many species currently liv- Grimaldi, 1996; Ross, 1998). Large amounts of res- ing on pines have not been identified in Baltic am- in were preserved after it amberized into hard frag- ber. Moreover, the chemical composition of the res- ments, and when it was repeatedly transported into in indicates Agathis (Araucariaceae) and Cedrus trees new sedimentary environments. The original loca- (Poinar, 1992; Krumbiegel & Krumbiegel, 2005) or tion of the amber trees, however, has not been pre- genus Pseudolarix (Grimaldi, 1996; Grimaldi & En- served in the geological record. The environment of gel, 2005). the amber forests can only be approximated on the Beck (1993) has comprehensively studied the basis of the plant and animal remains preserved in it. chemical composition of Baltic amber, and concludes It is not known for certain, which tree or trees that it relatively closely resembles the resin of modern were the resin-segregating ones. It may have been Agathis species. According to Beck (1993), the wood 106 M. Nuorteva and K. A. Kinnunen anatomy of the Araucariaceae, which was only recent- years (Krzemi|nska & Krzemi|nski, 1992). In Baltic ly divided from the Pinaceae family, is rather similar amber, practically all the insects correspond to exist- to that of Agathis. These trees are nowadays extinct, ing families, about half to existing genera, and a few and the chemical composition of their resin is close close to still existing species (Larsson 1978, Poinar to that of modern East-Asian species. Some insects 1992). However, Wunderlich (1986) estimates that living in trees are especially faithful in selecting their 75 % of the spider fauna in Baltic amber consists of host trees. When more amber insects have been stud- extinct genera. The estimate from his latest studies ied this will provide new information on the distinc- (Wunderlich, 2004) on the fauna in Samland am- tiveness of these ancient trees. ber is 88%. The large amount of amber preserved to the pres- The insect finds show that the insect fauna of the ent day shows that the resin flow was intense, which ancient amber pines resembled that of present pines may indicate disturbances in the life cycle of the trees. (Larsson 1978). Only the largest species are lacking However, the causes that led to these disturbances are from amber. No fossils have been found in Baltic am- unknown. One can speculate on climatological causes ber that could represent needle-eating sawflies (Hy- or on injuries caused by insects or pathogens. menoptera: Pamhilididae, Diprionidae) or moths When the composition of this entomofauna is re- (Lepidoptera: Geometridae, Lymantriidae, Noctu- solved, it will be possible to understand the circum- idae, Lasiocampidae, Thaumatopoeidae, Sphingi- stances that contributed to the destruction of am- dae). However, their larvae probably lived in the tree ber pines. Several studies, e.g. Ander (1942), Ba- crowns. Only two fossil species of Diprionidae saw- chofen-Echt (1949), Larsson (1978), Poinar (1992), flies have been described (Bachofen-Echt, 1949). The Krzemi|nska & Krzemi|nski (1992), Weitschat & species that belonged to the above-mentioned orders Wichard (1998) and Zherikhin (2002), have focused of insects and which consumed needles in the cano- on the entomofauna together with the amber data. pies, may have reduced the vitality of the trees. However, knowledge of the actual species that caused Surprisingly few descriptions exist on the particu- the damage, is still inadequate. larly species-rich family of beetles, and of flat-headed There is no information available on specific borers (Buprestidae) only about ten species (Spahr, pathogenic insects that could have been disastrous to 1981; Zherikhin, 2002). Similarly, large long horned trees. The data on present-day insects which dam- beetles (Cerambycidae) have only seldom been ad- age forests, imply that their activities may have led hered in resin, except for some of their small larvae to the death of amber trees. When speculating on (Larsson 1978). Thus only about twenty families have the causes of the resin flow in damaged trees, how- been published (Spahr, 1981; Poinar, 1992). Impres- ever, one should remember that the largest destruc- sive individuals belonging to these two families may tive insects were not trapped in the resin, because in- attract forgers, and erroneous records are likely. sects larger than 1 cm are rare in amber (Schlee, 1990; Nowadays the bark beetles (Scolytidae) are signif- Krzemi|nska & Krzemi|nski, 1992; Poinar, 1992). Per- icant destroyers of trees. Many families and species haps large, powerful insects did not stick to the resin, of bark beetles have been observed in Baltic amber or perhaps birds and other predators ate them. Sim- (Spahr, 1981; Keilbach, 1982). On the other hand, ilarly, complete spider fossils are usually below 1 cm the number of individuals is small, although these (Wunderlich, 1986). insects may attack dying trees and therefore they The living styles of present-day insects on pines should be more frequent (Schedl, 1947, 1967; Ba- (Pinus) are reasonably well known. This insect fauna chofen-Echt, 1949; Larsson, 1978; Krzemi|nska & is relatively similar to those present in the Eocene and Krzemi|nski, 1992). Those species which nowadays Oligocene. The evolution of insects and their parent attack relatively healthy trees are not known in am- trees has probably been synchronous for millions of ber. The trees in amber forests may have developed Insect frass in Baltic amber 107 resistance against predators widespread in recent trees 2. Baltic amber forests (Zherikhin, 2002), or perhaps our knowledge of the insects in amber forests is still too deficient. In order to better understand the insect fauna of pine- Moreover, insect predators and parasitoids pro- like amber forests, it is important to collate the pub- vide some suggestions on their host animals, al- lished facts on these forests, their locality, vegetation, though their host insects are absent as fossils. Al- climate and composition of the insect fauna. though one cannot find the bodies of the insects themselves, it is possible to obtain information on 2.1. Provenance of Baltic amber the presence of pests by identifying their waste, ex- crement and frass. While eating, however, these for- Besides the Baltic countries, Baltic amber is also found mer destructive insects would have produced plenty in England, Denmark, Germany and Russia up to the of excrementd and wood dust. These particles prob- Ural Mountains. According to Bachofen-Echt (1949), able adhered to the resin. Different pest genera and the amber-producing forests were located in Fennos- species have distinct waste products, which can be candia. It is generally believed, that amber was formed morphologically identified. For example, Eckstein in Central Scandinavia, Finland, Estonia, and partly in (1939), Nolte (1939) and Becker (1949) have de- Russia (Ander, 1942; Kremi|nska & Kremi|nski, 1992; scribed how to identify conifer-eating insects on the Grozonkowski, 2004; Selden & Nudds, 2004; Krum- basis of their frass. biegel & Krumbiegel, 2005). The known deposits of The waste products of insects in amber have not Baltic amber are all secondary and have been trans- aroused much research interest (Weidner, 1956; Zher- ported by glaciers, rivers and littoral processes. The ikhin, 2003). Only Weidner (1956) has described ter- low specific gravity (1.05 − 1.09) and therefore the mite excrement in Baltic amber. In addition, Grimal- buoyancy of amber enabled it to easily drift especially di & Engel (2005) have published one photo de- in saline seawater (1.03) over very long distances. picting mineralized excrement of a termite from Eo- It is now known that from the late Oligocene to cene age wood in Queensland, Australia. Frass and the early Miocene (see Overeem et al., 2001) the Eri- fecal pellets have been found in Dominicean amber danos fluvio-deltaic system transported pieces of am- (Grimaldi, 1996). Wunderlich (2004) has described ber from northern Fennoscandia to the Gdansk del- dissected remains of preys proced by spiders in Bal- ta (Krumbiegel & Krumbiegel, 2005). This river sys- tic amber. Arillo (2007) has recently reviewed studies, tem occupied the area of the present Baltic Sea and, which depict insects paleobehaviours. Arillo (2007) in its later stage, reached as far as the North Sea (Fig. suggests that defecation may result while insects are 1). In the Holocene, storms redistributed pieces of struggling against the sticky resin. However, morpho- amber from the sediments in Poland and Kaliningrad logically distinctive excrement and frass are abundant to the west and east along the shoreline of the south- and visible in many pieces of amber (Schubert, 1939; ern Baltic (e.g. Reinicke, 1995; Botheroyd & Both- Schlee, 1990; Ross, 1998; Pieli|nska, 2001; photo of a eroyd, 2004; Wichard & Weitschat, 2005). The Eri- piece of excrement in Dahlström & Brost, 1995). By danos River flowed from the Gulf of Bothnia along studying and comparing them with recent examples, the Baltic basin to the Gdansk delta. These ancient it should be possible to infer the former presence of river channels were leveled down during subsequent some insect genera. glacial activity. The Eridanos system was huge and This study is focused on the possibilities of iden- comparable in size to the present Orinoco delta in tifying from Baltic amber wood particles and excre- southern America. ment produced by insects. The study is based on the The huge Gdansk delta and the amber occurrences characteristics of some modern debris and on com- in Samland (now in the Kaliningrad area) are the are- parison with similar finds in Baltic amber. as from where our research material originates. 108 M. Nuorteva and K. A. Kinnunen Eocene the climate cooled down. This led to changes in fauna and flora (Weitschat & Wichard, 1998; Wi- chard & Weitschat, 2005). 2.3. Vegetation Recognizable plant fossils are much rarer in amber than insect finds. As a result, information about the vegetation is more deficient, although the living hab- its of the fossilized insects may indicate the former presence of certain plant species. It is important to know the abundance of tree species in amber forests (Ander, 1942; Bachofen-Echt, 1949; Larsson, 1978; Krzemi|nska & Krzemi|nski, 1992; Grimaldi, 1996; Weitshat & Wichard, 1998; Kosmowska-Cerano- wicts, 2001; Krumbiegl & Krumbiegl, 2005; Wi- chard & Weitschat, 2005). In addition to pine-like trees, other conifers were also present. There were nu- merous species of oak especially among the decidu- ous trees. Tiny hairs on the leaves and buds of oaks are typical of Baltic amber (Bachofen-Echt, 1949; Krzemi|nska & Krzemi|nski, 1992; Wichard & Weitschat, 2005). Fig. 1. Eridanos fluvio-deltaic river system (Amber riv- Some species, for example various palm trees as well er) of the Cenozoic age and its Gdansk delta along the as Castanea and Mangolia, which are more souther- contemporary coastline. The main commercial amber deposits were originally located in these deltaic sed- ly today, were present in amber forests. Today, similar iments and later redeposited to the glacial sediments. forests primarily grow in North America and in Pale- The amber material of the present study was mainly arctic China (Ander, 1942). No forests that are com- from this delta area. The ancient amber forests were lo- cated in this river area mainly in southern and central pletely identical with the ancient amber forests are Fennoscandia. Compiled from Overeem et al. (2001), known anywhere. Botheroyd & Botheroyd (2004) and Kosmowska-Cer- Plant and animal remains in Baltic amber indi- anowicz (2004). cate that pine-like amber trees grew in or near aquat- ic environments. The abundance of water bug fos- 2.2. Climate sils in amber clearly verifies this (Wichard & Weis- At the end of the Cretaceous the Baltic region was chat, 1996). The resin of pine-like trees growing in positioned about 10° – 12° further south than its mires was preserved in moist sediments, where it was present position (Larsson, 1978). At that time, the not oxidized. The aquatic environment has likewise climate in the Baltic region was predominantly sub- protected the amber pieces against the action of for- tropical. It is possible that the differences in tempera- est fires. Several species of coniferous tree most like- ture along the north-south axis were smaller than to- ly grew on drier land too, but their resin was oxi- day, and thus tropical and subtropical life forms had dized and destroyed (Weischat & Wichard, 1998). a more extensive distribution (Weitschat & Wichard, It should be noted that the Trichoptera and Diptera 1998; Grimaldi & Engel, 2005). Thermophilic spe- fauna found in amber consist of species living in flow- cies were common in amber forests, but in the end of ing as well as in standing water. This suggests that the Insect frass in Baltic amber 109 amber forests grew in a mountaneous landscape (An- the western market in the 1990’s. At that time, one of der, 1942; Bachofen-Echt, 1949; Weitschat & Wi- us (K.K.) had the opportunity to examine large am- chard, 1998). ber parcels for inclusions. He bought pieces of am- ber with interesting inclusions for his private collec- tion. This private collection (K.K.) consists of 1085 2.4. Insects amber pieces with a diameter ranging from 1 to 8 The abundant, well-preserved insect inclusions make cm (21 amber pieces are in the size range 4 – 8 cm). it possible to compare them with present-day insects. Altogether 43 doubly polished slabs have been pre- The fauna which was specifically associated with the pared from selected samples for microscope examina- living amber trees at that time has since disappeared tion and photographic documentation. In these sam- and, according to Larsson (1978), on the whole its ples insects were rare (only a few percent of the sam- genera are only distantly related to present forms. The ples contained them), but wood debris was surpris- insect fauna differs from the European one, although ingly common (Figs. 2 and 5). Wood fragments were there are some similarities. Tropical insects can ex- visible in most lower grade, but otherwise transpar- ist together with holoarctic fauna, although the sub- ent amber pieces, most of which originated from the tropical component is usually the dominant one (An- Gdansk area. der, 1942; Larsson, 1978; Poinar, 1992; Szadziewski Some of these microscopic wood fragments resem- & Sontag, 2001). When one compares the insects in bled debris produced by certain wood boring insects amber to present insects and their known living hab- (noticed by K.K. in the beginning of the 1990’s). its, it is clear that there are similarities between the Their geometrical shape resembled the sawdust pro- living habits. duced during the carving or drilling of wood ma- Insect fauna associated with resin-excreting trees terial. The amber containing these inclusions were is obviously the most numerous. The insects were shown to the other author (M.N.), who is a profes- trapped in resin while it was still fluid, and hence sor of forest entomology. M.N. became keenly in- night-active and hibernal-active insects are rare as terested in the topic. His expertise in insect habitats fossils (Krzemi|nska & Krzemi|nski, 1992). It is not in trees (e.g. Nuorteva, 1956), subfossils (Koponen known whether some of the species known as fos- & Nuorteva, 1973) and excrement (Nuorteva, 1972) sils from Baltic amber are still living. However, all the and in the identification of wood markings, was fun- orders and main part of the families and genus are damental to this study. still known today, although some of them are pres- The amber specimens were first studied under a ently living outside Europe. The cooling of the cli- stereomicroscope and morphologically classified. mate presumably made some insects to move south Some surfaces were oiled in order to obtain a sharp into warmer conditions (Poinar, 1992; Wichard & view of the the inclusions. Tens of sections enclosing Weitschat, 2005). Near relatives of some insects, now wood fragments were prepared from pieces of am- extinct in Baltic countries, live for example in Spain, ber. These sections were polished on both sides. The northern and southern America, South Africa, South- thickness of the sections was approximately 3 mm. east Asia, and Australia (Ander, 1942; Larsson, 1978; They were studied under a stereomicroscope, polariz- Krzemi|nska & Krzemi|nski, 1992; Grimaldi, 1996; ing microscope and then photographed. Our amber Weitschat & Wichard, 1998; Ross 1998). material mainly consisted of pieces from 1 to 3 cm in size. Only 21 larger specimens were studied. The size 3. Material and methods range is close to the typical size of amber in Holocene sediments. Large quantities of Baltic amber of varying quality Our research material mainly consisted of com- from Poland, Lithuania and Kaliningrad appeared on mercial pieces of Baltic amber obtained from Poland 110 M. Nuorteva and K. A. Kinnunen Fig. 2. Dark layer of wood debris in Baltic amber, most probably from Gdansk area, Poland. Length 32 mm. In private collection (K.K.). Photo: K.K. and Lithuania. Most of them most probably originate information about these redeposition processes, and from Jantarnyi, Kaliningrad. In addition, 26 pieces of not about the primary habitat of amber forests. amber from the paleontological collections in Geo- logical Museum of Finnish Museum of Natural His- 4. Genuineness of the amber tory, University of Helsinki, were studied. Their lo- cality was given as Samland. The 170 pieces of Baltic Amber fakes are problematic and sometimes difficult amber from the entomological collection in Zoolog- to recognize and pose a serious problem in research ical Museum of Finnish Museum of Natural Histo- and trade. Forgers are not interested in small pieces ry, University of Helsinki, the precise origin of which of minor value. However, gem-grade and museum- was not known, were also studied. Additionally, 34 quality pieces are more problematic, and we also ex- amber pieces purchased by M.N. were included in amined a number of them in our study. As pieces of the study. fake amber with large insect inclusions are common The museum ambers in Geological Museum were on the market, methods for validating their authen- commonly hydrated and covered with a network of ticity are important. microfractures (cf. Shashoua, 2002). Microscopic The inclusion of fossils, insects or other, which is study of these damaged pieces was problematic. The extinct today, would definitely prove the authenticity. samples were coated with castor oil in order to obtain However, most insects can only be determined to the photographs. The museum personnel granted per- genera level, and this is therefore of limited use in this mission for this procedure, because no damage was connection. Likewise, most insect fossils are deformed caused to the specimens. and incomplete making their identification difficult. The exact locality of most of our amber specimens For this reason, we have used other methods. from the Baltic area was unknown. We think, how- Inclusions of wood debris are characteristic of Bal- ever, that this has not had any significant influence tic amber and they provide a simple way to test the on our conclusions. Amber pieces have been repeat- genuineness of specimens. We would like to add wood edly redeposited throughout their geological history. fragments to the list of inclusions, together with py- The exact localities – if known – would only provide rite grains and tiny hairs probably detached from the Insect frass in Baltic amber 111 imports from southern Baltic areas. The finds in Nau- vo and Borstö were pebbles found on the shore of Airisto Island. They were described in 1700’s in the records of Turku University. These samples were most probably destroyed in 1827 in the great fire of Turku. The amber pebbles may have been attached to algal drift and thereby transported from the south to their present location. Or they may have been transported in ballast material used in sailing ships for centuries. This interpretation is supported by the fact that Pale- ogene (66 − 23 Ma) sedimentary rocks and strata are not known in Finland (cf. Kohonen & Rämö 2005). From the Comb Ceramic Stone Age (ca 5000 BC Fig. 3. Stress anisotropism between amber layers. Dark – 3200 BC) onwards, amber was imported into Fin- particles are composed of wood debris. Double pol- ished section of Baltic amber. In private collection (K.K.). land mainly as finished products: pendants, beads and Width of the picture area is 4 mm. Transmitted crossed small sculptures. The finds are abundant, especially in polarized light with lambda-plate. Photo: K.K. the red-ochre tombs dated to the Comb Ceramic cul- ture in southern and middle Finland (Äyräpää, 1945, male flowers of oak trees (see Bachofen-Echt, 1949; 1960). Well-preserved small amber objects with hu- Ross, 1998), that are considered as diagnostic of gen- man forms have been found on the lake bottom in uine Baltic amber. In addition, the wave-like layering front of Astuvansalmi rock painting cliffs (Grönha- of the amber (Fig. 3), which is intimately connected gen, 1991). Uino (2005) has reviewed the Finnish lit- to the inclusions and described in detail in the follow- erature on archaeological amber when describing the ing, seems to be an important diagnostic tool. most recent archaeological finds of amber in Rävåsen, Other amber tests have also been proposed. They Ostrobothnia. include differences in fluorescence under ultraviolet radiation and the absence of stress-induced polari- 6. Petrography of amber zation figures around the inclusions. We tried these tests but, in our opinion, they did not give reliable The morphology of pieces of amber may indicate answers about amber authenticity. their location and orientation in trees. This could narrow the potential insect candidates that produced 5. Amber in Finland the frass. Most lumps of amber are, however, round- ed and fractured and their surface is weathered. This To our knowledge, amber has not been found in situ limits the possibility to interpret their original loca- in Finland; although amber forests most probably tion in trees. Pieces of amber are easily transported in grew in the area of present southern and central Fin- flowing water, because of their low density. Although land (see Ander, 1942; Bachofen-Echt, 1949; Kati- in wet conditions they remain almost completely un- nas, 1983; Kosmowska-Ceranowicz, 2004; Wichard weathered, the long drift distances have mechanical- & Weitschat, 2005). ly shaped their forms through abrasion. Most of our There are three records of amber finds in Finland: amber samples can be classified as superficial ambers at Ingarskila in Inkoo (Holmberg, 1857), and at Nau- formed outside the trunk. Slab-like, interior ambers vo and at Borstö in Airisto (Laitakari, 1967). Accord- formed in the trunk have only rarely been found. ing to a number of Finnish archaeologists, the am- Most of our ambers in the superficial category re- ber from Ingarskila most probably represent stone age sembled trunk amber with a very few drop-like ici- 112 M. Nuorteva and K. A. Kinnunen cle forms (see Katinas, 1983). The morphology of the The flow layer observations showed that the wood side facing the trunk or twigs was typically better pre- debris was specific to each resin layer. Some layers served and, in a few specimens, it still contained rem- showed strong enrichments. A few pieces of amber nants of wood material. had layers with wood fragments of different mean di- We could not identify the species of wood frag- ameter. These findings may indicate that the wood ments in our samples. Earlier published studies on debris was transported, together with the resin, to the ambers contained larger wood fragments. These frag- site of solidification. Hair fibres detached from oak ments have been identified as pieces of Pinus suc- leaves may indicate eolian transport, because they ciniferum (see review in Gübelin, 1978). were also found as inclusions, mainly in clear amber. The wood fragments in our samples showed a simi- However, wood fragments were rare in amber pieces lar size range, regardless of the diameter of the host am- containing oak leaf fibres. ber. Feret’s diameter (the largest diameter present) was The wood debris was also entirely embedded in measured by image analysis using Image Tool software. individual resin layers, and not concentrated in the The diameter of the wood fragments ranged from 0.6 contact surfaces. The same applies to aerial contam- – 1.8 mm. The mean diameter was ca. 1 mm. ination because bubbles were embedded in the indi- In most cases the wood fragments were random- vidual flow layers. On the contrary, the infrequent in- ly scattered throughout the amber. In a few cases they sect inclusions were typically located on or near the were concentrated in layers, and in pile-like structures contact surfaces. in the layers (Fig. 2). The base of the amber piec- es (the surface facing the trunk) almost always con- 7. Possible frass producers tained more wood fragments. 7.1. Needle eating insects On partly fractured pieces of amber, flow layers were visible as undulating, wave-like surfaces. Some Many larvae of butterfly species eat pine needles over pieces of amber were fractured along the contact sur- large areas. The same applies to the larvae of some face between successive resin layers. In some of the sawflies. Nolte (1939) has described the excrement polished amber sections, the contact surfaces were produced by insects that eat pine needles. Using seen as bending lines in reflected light on their sur- Nolte’s procedure, Larsson & Tenow (1980) meas- face. In transmitted light the flow layers were observ- ured the mass of the excrement and other waste frag- able when the doubly polished slab was swivelled to a ments produced by some insect groups. The total position where the contact surfaces were aligned par- amount of faeces produced was in the order of 11 kg allel to the microscope axis. Crossed polarized light dry weight/ha, and of green litter 1.5 kg/ha. This is showed that the contact surfaces of the flow layers a high figure compared to the small size of the par- were under minor stress (Fig. 3). In most cases their ticles. For this reason, we consider that they should thickness varied within a few millimetres. In a few also occur in amber. Some beetles (adults and larvae) cases, however, it was possible to determine the top also eat pine needles, but their excrement may be dif- direction of the flow from deformed bubbles and ant ficult to identify. inclusions located close to the contact surface. In gen- Aphides (Homoptera: Aphidoidea) that sucked eral, the flow layers showed that the resin surfaces tree sap from needles and branches are common were convex upwards. This observation can be used in amber (Heie, 1967; Larsson, 1978; Zherikhin, in determining the original alignment of the lump of 2002). When occurring in huge colonies, they may resin. The orientation of the wood fragments in the significantly reduce the vitality of trees. Aphids pro- flow layers was mostly random. The viscosity of the duce plenty of honeydew, which fall as small drop- resin probably prevented the flow from having an ori- lets from the crowns of deciduous and coniferous entating effect. trees where they live. It is likely that traces of this sac- Insect frass in Baltic amber 113 chariferous, sticky material may have been trapped Some recent Denroctonus bark beetle species in- in amber. trude into living trees and may create abundant flow of resin. Plenty of wood dust would adhere to the resin. Adult beetles know to avoid resin and thereby 7.2. Species living on tree bark avoid becoming trapped. It is noteworthy that they Ants (Hymenoptera: Formicidae) exploit the honey- have not been met in Baltic amber (Larsson, 1978). dew of aphids. This may explain the commonness of Wood dust mixed with amber should therefore be ant fossils in amber. They also prey on small inverte- searched for. Members of this bark beetle, which brates. Their excrement is a fluid material. live on Pinaceae conifers, existed already at the time Parasitical hymenoptera are likewise common when Baltic amber was formed. An engraving made in amber (cf. Larsson, 1978). Many are associat- by Denroctonus bark beetle has been identified on a ed with specific host animals. Because recent ex- petrified, middle Eocene (45 Ma) specimen of Larix amples are reasonably well known, it is possible to altoborealis log from the Canadian High Arctic (La- make conclusions about the host animals on the bandeira et al., 2001). basis of the presence of specific parasites found in The frass of the larvae consists of fragments that amber. became detached when the galleries were built and Likewise the prey of some coleoptera and their of excrement that passed through the alimentary ca- commensals living in the galleries of bark beetles has nal. This larval frass usually remains below the bark been preserved in amber (cf. Larsson, 1978). and only drops out when the bark is splitting off. In Lycoriidae adults belonging to flies (the order Dip- such cases, the resin would most probably have al- tera) are common in amber (Ander, 1942; Evenhuis, ready hardened and the adherence of frass particles 1994). Most of their larvae lived in humus material. would be only accidental. An exception to this is the Some larvae of the genus Sciara live as huge swarms present-day timberman (Acanthocinus aedilis L., Ce- on dying trees and eat their phloem. It is possible that rambycidae), which forms frass accumulations, in the small, round excrement of these larvae could be which the bark splits off while the tree still is rela- found in Baltic amber. tively vital. This kind of debris could also be found Adults of dipterous flies of the Dolichopodidae in Baltic amber. family are abundant in Baltic amber (Larsson, 1978; Evenhuis, 1994). One genus (Medetera) of this fami- 7.4. Insects gnawing sapwood ly laid their eggs in the galleries of bark beetles, where the larvae operate as predators. However, they do not Many larvae of beetles groove their channels partly leave any recognizable excrement. in the sapwood. Some larvae of the snout beetle (Pis- sodes, Curculionidae), and of other beetles, excavate a cavity or pit in the wood material when encapsulat- 7.3. Beetles penetrating under the bark ing, and then cover it with a wood splinter. Some lar- The phloem layer below the bark is rich in nutrients vae of Cerambycidae species (e.q. Acanthocinus aedi- and it attracts many beetles to breed. The adult bark lis L. and Rhagium inquisitor L.) may surround their beetles (Scolytidae) push dust when they drill through pits with a crown like structure composed of wood the bark and when they excavate egg galleries in the splinters. bark. This dust is fairly fine-grained, but its mass may Some larvae dig their channels into the duramen be large. It is probable that some of them may have (e.g. Monochanus). Abundant accumulations of fra- been included in amber. Thus, the characterization of ss may lead to bursting of the bark, and the frass is wood dust may aid in verifying the former presence thereby expelled in late summer. Likewise, it is pos- of bark beetles. sible to find frass removed by beetle larvae in its sec- 114 M. Nuorteva and K. A. Kinnunen ondary environment. This may happen e.g. when Literature on the frass and excrement produced moth larvae (Lepidoptera, Psychidae) attach wood by insects living in pines is scanty. Becker’s (1949) frass to their cases while moving along the trunk (cf. study is the most detailed one. Some information can the photos in Weitschat & Wichard ,1998). be found in Vité (1952, 1953) and Eckstein (1939), who also give data on other insects. Nolte (1939) de- scribes excrement produced by insects living on the 7.5. Insects digging burrows in tree trunks needles of firs and pines. Some information can be Many genera of insects are known to bore their cav- found on butterflies and sawflies in the literature. ities in dead tree logs, but fluidal amber resin is no Brauns (1954) has described the excrement of lar- longer present in this stage. Their frass could be in- vae of Diptera. cluded only accidentally in amber. However, car- In practice, only adult beetles and termites pro- penter ants (Camponotus) do excavate their galler- duce solid excrement. This is because the adults of ies in living trees. They do not eat the wood mate- most other species consume very little solid nutrition. rial, but instead push it out as frass. Adults of Cam- The size of the excrement is similar, although there is ponotus have been found in Baltic amber (Keilbach, variation in the size of the adults. On the other hand, 1982). the size of the excrement produced by larvae depends Termites (the order Isoptera) are insects that live on their size. The difference in the size of excrement as colonies in subtropical and tropical areas. At the between adults and larvae require further study. present time their most northern occurrences are in The size of the excrement depends on the growth Central and Southern Europe. Some dig their nests stage of the larva. Identification of the small piec- in dead trees or those in a poor condition. Mostly es of excrement produced by young larvae is usual- they eat mushrooms that avoid direct sunlight. Ter- ly impossible because one should know the largest mites may have been common in amber forests, but excrement size of the individual species. Larvae eat individuals from the swarms have mainly become ad- in order to grow and therefore abundant excrement hered in the resin (Bachofen-Echt, 1949; Larsson, is produced. Food flow through the alimentary ca- 1978; Poinar, 1992). The excrement of termites is nal is rapid, and even the cell structure of the plant easily identified and has been described from Baltic tissue they have eaten is usually still identifiable. A amber (Weidner, 1956). It should be noted, however, thin film (peritrophic membrane) usually envelops that termites are more frequent in Dominican than in the excrement. The excrement produced by insects Baltic amber (Schlee, 1990). that eat wood material is always among the frass loos- It is possible that resin may have dropped onto de- ened from the wood material. cayed wood material colonized by insects. This may The shape of the excrement is often diagnostic. also be one way in which frass has become included Some butterfly larvae, which eat pine needles, typical- in amber. Typical fossils are the adult beetles of Ano- ly have longitudinal, transverse grooves on their ex- biidae and Mordellidae, which produce a great deal crement (Plate 1, B, see also Müller, 1992). Some hy- of frass inside trees (Bachofen-Echt, 1949; Larsson, menoptera larvae similarly eat pine needles. The ex- 1978). crement of Diprionidae pine sawflies are frequently in the form of parallelograms (Plate 1, A). They have 8. Frass produced by insects living on been described in Nolte (1939), Eckstein (1939) and Escherich (1942). present-day pines Adult beetles and especially larvae do not eat all Our examples of frass and excrement produced by in- the wood and phloem particles they have detached sects are from Finland, because our aim here is only from their galleries. The fragments loosened from the to demonstrate their diverse shape. wood material are usually angular in shape, light-col-