INTRODUCTION

The nomenclature of (sub)fossil resins is problematic (Vavra, 2009), and the distinction between amber and its subfossilized precursor, copal, has not been clearly defined. Generally speaking, copal is much younger and less polymerized (and hence softer) than amber. Many paleontologists consider copal too young to be of interest and, as a result, little research has focused on this material. However, Penney and Preziosi (2010, 2013), Penney and others (2012b), and Penney and Green (2012) highlighted the potential value (at many different levels) of subfossils in copal.

Both amber and copal are renowned for preserving insects and other inclusions with lifelike fidelity (Penney, 2010), including, in some cases, at the subcellular level (Koller, Schmitt, & Tischendorf, 2005). However, the exact mode of preservation in amber is unknown, although it is commonly referred to as a kind of mummification process resulting from rapid fixation and dehydration of anything that became trapped in the original resin secretion. The process will presumably vary, albeit possibly only slightly, in different ambers as the result of differences in the chemistry of the resin secreted from the amber producing trees, which originated from various different families (e.g., Langenheim, 1995). Certainly, the process is not uniform for all inclusions. For example, the recent application of X-ray computed tomography in studies of fossils in amber has demonstrated that, in some instances, internal organs are preserved, e.g., in the case of a strepsipteran preserved in Eocene Baltic amber (Pohl & others, 2010), whereas digital dissection of a spider preserved in Eocene French amber revealed that nothing substantial was preserved internally (Penney & others, 2007).

Although the application of computed tomography and synchrotron scanning to amber and copal inclusions has aided the study of these fossils considerably by narrowing the divide between paleontological and neontological taxonomy (Penney & others, 2007, 2011, 2012a, 2012d; Bosselaers & others, 2010; Pohl & others, 2010; Soriano & others, 2010; Dunlop & others, 2011, 2012) and by allowing us to study unique paleoethology (Dunlop & others, 2012; Penney & others, 2012c), they are still considered by many neontologists to be taxonomically subequal to extant forms. These techniques are time consuming, very expensive, require specialist technical expertise, and are restricted in their availability. It would be greatly beneficial if fossils could be extracted from their (sub)fossilized resin matrix so that they could be studied alongside extant forms using the same techniques (including electron microscopy; e.g., Azar & others, 1999) and taxonomic characters for their identification. Extraction of fossils from (sub)fossilized resins would also facilitate studies of their molecular paleobiology and taphonomic physiochemical changes resulting from diagenetic processes, in addition to more accurate quantitative studies of Quaternary tropical forest biodiversity.

Previous successful studies to extract inclusions by dissolving amber include those of Azar (1997), Azar and others (1999), and Mazur and others (2012). Azar (1997) was able to dissolve upper Neocomian—basal lower Aptian (ca. 135 Ma) amber from Lebanon. Various solvents (e.g., ethanol, butanol, acetone, toluene) were tried, but only chloroform gave satisfactory results, yielding articulated but fragile fragments of insect cuticle, including heads, abdomens, wings, and genitalia. Mazur and others (2012) found xylene, toluene, chloroform, orange oil, and turpentine oil useful for dissolving inclusions out of Eocene (ca. 52 Ma) Cambay amber from India. In both, the aforementioned studies, the extracted inclusions were very fragile, no doubt because of their great antiquity. Here, we apply a similar technique to inclusions preserved in Dominican amber, Baltic amber, and Colombian copal, and describe a new species of stingless bee from the last deposit.

Stingless bees (tribe Meliponini) are one of only two highly eusocial bees, the other being the well-studied honeybee (tribe Apini). Unlike Apini, with only 11 species in the single genus Apis, stingless bees form a large and diverse taxon that consists of 60 genera, many of which are poorly known (Rasmussen & Cameron, 2010). They are found in abundance in warm humid forests around the globe and have left an imprint in the fossil record spanning most of the Cenozoic. Poinar (1999) proposed that the demise of stingless bees known from Dominican amber but absent from the Greater Antilles today resulted from a cool period associated with increased aridity during the Plio-Pleistocene, although Peñalver and Grimaldi (2006) suggested that the insularization of Hispaniola was probably a more important factor. Hence, fossils and subfossils of this group of organisms have the potential to be informative about past biogeographical processes and the comparative extinction resistance of island versus continental lineages. The relatively young age of many copals means that inclusions may belong to extant species, even though they may have not yet been described in the scientific literature. The possibility that a copal inclusion may belong either to an extant or an extinct species highlights the importance of considering both neontological and paleontological data when describing new taxa from copal-producing regions (Penney, Ono, & Selden, 2005; Azar, Nel, & Waller, 2009).

MATERIALS AND METHODS

Our samples included one specimen of Miocene Dominican amber (ca. 16 Ma) containing a flat-footed beetle (Coleoptera), one specimen of Baltic amber (ca. 44–49 Ma) containing a small fly (Diptera), and a specimen of sub-Recent Colombian copal (post-WWII) (radiocarbon dated at the University of Arizona AMS facility) containing stingless bees (Apidae; Meliponini; Trigonisca sp.) (Fig. 1.1). A second Colombian copal specimen, also containing the same species of stingless bee, was sent for dating at the Arizona AMS facility. This came back with an age of 10,612 ± 62 years, representing the oldest formally dated Colombian copal sample. The specimens were trimmed to a small workable size using a High-Tech diamond trim saw, and then each specimen was further shaved down using a scalpel into a small cube of approximately 0.4 g (∼4 mm³), with the inclusion situated in the middle. Microphotographs were assembled from a stacked series of digital images recorded by a Nikon Coolpix 4500 camera mounted on a Leica M10 stereomicroscope with 0.63× and 1.6× planapochromatic objectives (Green, 2005).

Using a laminar flow hood, each specimen was placed with 5 ml of chloroform into a glass tube with a screw-top lid, and then placed in a water bath (with the rocking motion switched off) at 40° C and left for 48 hours, during which time they were checked periodically.

RESULTS

After 48 hours, the copal sample appeared to have dissolved to a greater extent than either of the amber samples. However, there was a highly viscous, nondissolved fraction floating across the surface of the chloroform. The inclusion had dissolved out fully intact and was positioned immediately below this fraction. When the tube was shaken gently, the inclusion floated freely in the liquid below the viscous layer. For all intents and purposes, it resembled a Recent entomological specimen preserved in alcohol. The amber samples had dissolved to a degree, but in a much less consistent manner. Some parts had softened and come apart to form separate gooey masses, and there were also small, hard fragments of nondissolved amber. The insect inclusions had disintegrated, not unexpectedly, leaving only tiny black fragments as evidence of their previous existence. Leaving the specimens in the water bath for several days longer did not result in any significant changes in the amber samples. Thus, the amber samples were set aside as unworkable in terms of their taxonomic value, whereas the bee was successfully extracted from the copal and identified as a new species, following dissection in alcohol under a stereomicroscope.

SYSTEMATIC PALEONTOLOGY

Figure 1.

Trigonisca ameliae Penney n. sp. in Quaternary copal from Colombia: 1, holotype (NHM II 3059 [1]), scale bar = 1.0 mm; 2, paratype 1 (NHM II 3059 [2]) showing wing venation; 3, paratype 2 (NHM II 3059 [3]); 4, close-up of holotype; 5, close-up of holotype showing lack of setae on head and antennae; 6, paratype 3 showing left mandible dissected from the head after extraction of the inclusion using chloroform (with trace outline of T. moratoi mandible for comparison) (new).

Figure 2.

Wing venation of Trigonisca ameliae Penney n. sp. in Quaternary copal from Colombia, as drawn from paratype 1 (NHM II 3059 [2]) (see Fig. 1.2) (new).

DISCUSSION

Given that the attempts to dissolve specimens out of amber resulted in total destruction of the inclusion beyond recognition, with only particulate fragments remaining, we do not recommend this approach. Our studied amber inclusions were no doubt typical of most fossil arthropods preserved in amber in consisting of only hollow spaces lined with a thin layer of diagenetically altered cuticle. However, our findings contrast with those of Azar (1997), who was able to recover articulated elements of fossilized insect bodies from Lebanese amber. Maybe this is due to differences in the resin chemistry or maybe Azar (1997) used much thinner samples than we did. It should be noted that Cambay amber from India derives from a dammar-like resin, and so is only weakly polymerized and cross-linked (Rust & others, 2010), compared to the majority of other ambers of Cenozoic age (e.g., Baltic and Dominican), so the fact that Mazur and others (2012) were able to dissolve it in organic solvents is not particularly unexpected. An alternative method of accessing (but not completely extracting) inclusions fossilized in amber is to crack the amber open using liquid nitrogen, as employed by Stankiewicz and others (1998), or by cutting the specimen to a small size, cutting a groove around the inclusion and then gently splitting the amber apart (Grimaldi & others, 1994).

However, the ability to extract inclusions from copal is potentially useful for various areas of paleobiological research. Unfortunately, in the extraction of the bee from the chloroform it was necessary to pull it through the viscous nondissolved fraction, and it became coated with the sticky residue, which started to harden and became difficult to work with as it cooled in air. Various attempts were made to remove this, such as washing it with absolute alcohol as an alternative solvent, also in a hot water bath, but without success. A second specimen from the same piece of copal kept in a water bath at 40° C for six days appeared to have dissolved completely, and the bee sank to the bottom of the tube when it was shaken slightly. There was barely any sticky film over the fluid surface, and it was possible to pipette the inclusion out of the fluid without any sticky coating. This specimen was extracted in a dedicated ancient DNA clean room for destructive DNA sampling, so it was not possible to take any images of it. The ability to fully dissolve the resin would be highly beneficial for future studies, and subsequent research by resin chemists would be particularly welcomed.

Nonetheless, we were able to recover the stingless bee from the first sample, articulated and in its entirety, which could then be dissected as if it were a recently caught specimen, albeit with some hindrance from the sticky coating. We were able to examine important taxonomic features, such as the mandibular dentition, that were impossible to see in the individuals preserved inside the copal matrix, resulting in the new species described herein. It is worth noting that the extracted specimen seemed to be a little more fragile than might be expected in recently preserved extant material, and so such specimens should be treated with particular care. At this stage, the long-term fate of such extracted specimens is unknown, but we would expect them to survive reasonably well if preserved in alcohol.