Amber is probably the mode of preservation that provides the most fossils of micro-organisms. The first studies mentioning this kind of inclusions were published in the middle of the nineteenth century (Berkeley, 1848, Göppert and Berendt, 1845 and Göppert and Menge, 1886). During more than a century after, only few works were published (most of them concerning Baltic amber). However at the end of the twentieth century, studies of amber micro-organisms increased, and many publications have been published during the last two decades. These works showed that all the groups of micro-organisms that we have identified in modern environment could have been preserved in amber (Table 1, Table 2, Table 3, Table 4, Table 5, Table 6 and Table 7). In the following paragraphs, details about all these discoveries are given, with a special focus on mid-Cretaceous amber of France.

Bacteria (Bacteria excepted Actinobacteria and Cyanobacteria: Table 1; Actinobacteria: Table 2; Cyanobacteria: Table 3) have been first mentioned in amber by Galippe (1920) when he identified some Microzymas that he isolated from some amber samples and that he cultivated in his laboratory (Galippe's cultivated bacteria were most probably contaminants). Few bacteria were found in amber until the nineties when studies of DNA from amber developed. At that time, dozens of bacteria had been found in various ambers (e.g. Greenblatt et al., 1999, Greenblatt et al., 2004 and Veiga-Crespo et al., 2004). However, most of these studies were criticised as many scientists thought that these bacteria were only contaminants (Austin et al., 1997; Hebsgaard et al., 2005 and Orlando, 2005). At present, many doubts exist on the origin of these bacteria. As shown in Table 1, Table 2 and Table 3, few bacteria have been clearly identified on a morphological basis only, in part because there is insufficient morphology for visual identification.

Concerning the mid-Cretaceous amber of France, the most abundant micro-organisms correspond to prokaryotes. Rod-shaped bacteria have been preserved in many amber pieces whilst sheathed bacteria (similar to the genera Leptotrichites and Sphaerotilus) are only preserved as tiny cortices around some amber pieces. Actinobacteria (Fig. 1.1) are also very abundant and they are usually preserved as spherical colonies of 50-75 μm in diameter (Girard et al., 2009b). This shape shows that these Actinobacteria were able to grow in the resin before its solidification. The most abundant prokaryotes are Cyanobacteria, and more precisely the species Palaeocolteronema cenomanensis (fig. 1.2). This taxon created 1-2 cm thick cortices around some amber pieces. In these cortices, the filaments are centripetally orientated, showing that P. cenomanensis grew in the resin before its fossilisation. The most unexpected discovery was the presence of fossil phycocyanin pigments in some filaments of P. cenomanensis (Girard et al., 2009a).

Fungi (Table 4) represent the group of amber micro-organisms that have been the most intensively studied. They were the first micro-organisms found in the nineteenth century. However some of them seem to have been identified as fungi (according to the classification system of that day), whereas they were only Actinobacteria. For example the species Penecillites curtipes was described in 1848 (Berkeley, 1848) but the description and illustrations given by the author indicate that this inclusion is probably an Actinobacteria. The study of the amber fungi really started in the eighties and nineties. During these two decades, several authors studied fungi (Table 4) that have been preserved in amber of various origins (Baltic Sea, Dominican Republic, Mexico, USA, France, Germany...) and of various ages (Cretaceous, Eocene, Oligocene, Miocene...). These discoveries greatly enriched the fungi fossil record, providing many fossil Ascomycota and only few Basidiomycota. As for the bacteria, study of fossil DNA from amber provided few evidences of the presence of fungi in amber (Greenblatt et al., 1999), but these results are still debated.

Eukaryotes are less abundant than prokaryotes in French mid-Cretaceous amber, but they are more diversified. Fungi are the most abundant eukaryotes. However most of them are not identifiable as just fragments of the mycelium are preserved. Only few could be identified. They mainly correspond to Ascomycota such as specimen close to the genera Metacapnodium and Aspergillus, a cladosporioid fungi (Fig. 1.3) and possible Endomycetaceae (Girard, 2010). An extraordinary discovery was the one of a carnivorous fungus. A dimorphic fungus forming mycelia and yeast stages has been identified as a carnivorous fungus (Deuteromycotina) based on the occurrence of hyphal rings as trapping devices (Schmidt et al., 2008). More recently, bracket fungi (Basidiomycota) have been reported from faecal pellets preserved in mid-Cretaceous amber of France (Schmidt et al., 2010b).

Non-photosynthetic protists from amber were not mentioned in the literature until the 1940s (Legg, 1942) when inclusions similar to the genus Paramecium were found in Cretaceous amber of Canada. Then protists were known in few ambers (Table 5). Recently the Triassic amber of Italy yielded a diverse assemblage of ciliates and amoebae (Schmidt et al., 2006). The amber that yielded the most abundant and diversified assemblage is the Cenomanian amber of Schliersee. This amber has been intensively studied as it was supposed to be Triassic in age, but a recent study proved that it is Cenomanian (Schmidt et al., 2001). Because of its supposed age, this amber has been intensively studied and at least seven genera of ciliates and five of amoebae have been identified (Poinar et al., 1993a, Poinar et al., 1993b, Schmidt et al., 2001, Schmidt et al., 2004 and Schönborn et al., 1999). Concerning the other ambers, only sporadic discoveries have been made, but fossils to several lineages of non-photosynthetic protists were discovered.

Protists in French amber mostly correspond to testate amoebae. Most of them belong to the Arcellinida (close to the genera Arcella (Fig. 1.4), Hyalosphenia (Fig. 1.5), Centropyxis, Cryptodifflugia and Leptochlamys) (Girard, 2010 and Schmidt et al., 2010a). Only a specimen of a euglyphid has been found, close to the genus Assulina. Other protists are less abundant. Only few ciliates (Fig. 1.6) have been identified (specimens close to the genera Brachonella, Grossglockneria, Astylozoon and other Spirotrichia). Flagellates have not been clearly identified.

During the last ten years, several publications also described amber protists (Ascaso et al., 2003, Ascaso et al., 1997, Kedves and Párdutz, 2002, Martín-González et al., 2008, Martín-González et al., 2009a, Martín-González et al., 2009b and Veiga-Crespo et al., 2007). In these publications, some ovoid micro-inclusions that look like protozoans (mostly ciliates, flagellates and amoebae according to the studies) are described and illustrated. However, some characteristics of these inclusions indicate that they are probably not protists. Most of these inclusions have many delicate internal vacuoles (sometimes several dozens) that have been interpreted as the internal vacuoles of the protists. However this abundance of vacuoles is inconsistent with the poor preservation of protists and is probably due to a phenomenon of “vacuolisation” during amber polymerisation. In addition, despite resembling “membranous vacuoles”, they do not include more robust structures such as nuclei, cytoskeleton, flagella or organelles. Similar inclusions are preserved in mid-Cretaceous French ambers in which they can be very abundant. These inclusions show important variation in size and form in a single amber piece. This characteristic is not really compatible with the trapping of a natural population of protists. Other studies are needed to really confirm or infirm the biological origin of all these inclusions, but, according to our experience, they resemble typical non-biological inclusions found in amber, inclusions that surely cannot be assigned to protists (Girard, 2010). We propose that as a precaution, these vacuolated structures should not be considered genuine fossilised protists, as they may not be of biological origin. We suggest setting these aside when reconstructing time lines based on the fossil record, until there is convincing evidence of their biological origin.

Photosynthetic protists (Table 6) were found in Baltic amber for the first time in 1848 (Berkeley, 1848) and the oldest evidence is from the Carboniferous amber of Scotland (Smith, 1896). However, since these first discoveries, only few additional specimens of probable photosynthetic species have been found, mostly in Cretaceous ambers (from Germany, France, Myanmar and Spain) (Table 6). Some of the specimens are doubtful (especially specimen similar to euglenids) as they are very similar to most of the amber micro-inclusions that have a cell shape whilst they do not have a microbial origin. Some photosynthetic protists have also been preserved in mid-Cretaceous amber from France (Girard, 2010). They mostly correspond to unicellular green algae close to the genera Myrmecia (Girard et al., 2009b), Quadrigula (Fig. 1.7), Enallax (Girard, 2009) and Chlorcorona. Filamentous green algae are rare and badly preserved in mid-Cretaceous French amber. Only a specimen close to the genus Trentepohlia and a possible Zygnematales have been found. Few micro-inclusions have been assigned to possible badly preserved euglenids (Girard, 2010).

Lastly, few micro-inclusions have been described as Incertae Sedis (Table 7). These inclusions of unknown origin are surprisingly quite rare as amber can preserve many strange inclusions. However only few authors have dared to describe microfossils as Incertae Sedis. Mid-Cretaceous amber from France has also preserved Incertae Sedis. They can have a fusiform shape or be spherical. They do not exhibit the particular vacuoles of non-microbial micro-inclusions already discussed above (Girard, 2010). It is why, without any precise assignment, these micro-inclusions are called Incertae Sedis as they probably do not belong to well preserved micro-organisms.

Amber studies raise the problem of the identification of micro-organism species. Modern species of micro-organisms, from bacteria to protists, are now mostly defined on molecular and/or genetic methods (Fenchel and Finlay, 2006). Such techniques cannot be used for microfossils from amber. Several attempts to extract ancient microbial DNA from amber have been tried since the beginning of the nineties. Hamamoto and Horikoshi (1994) were the first to announce the extraction of ancient microbial DNA from amber (more precisely from Baltic amber). In the following years, many authors published on the discovery of ancient DNA from various ambers (Baltic, Dominican, Israel, Mexican, Burmese ambers) (Alharbi, 2008, Cano and Borucki, 1995, Cano et al., 1994, Greenblatt et al., 1999, Greenblatt et al., 2004, Lambert et al., 1998, Veiga-Crespo et al., 2004 and Veiga-Crespo et al., 2007). However all these studies have been criticized and some authors highlighted some weaknesses in the studies about ancient DNA from amber (Austin et al., 1997; Hebsgaard et al., 2005 and Orlando, 2005). Priest et al. (1995) questioned a possible contamination of the samples by modern DNA. Austin et al. (1997) raised the problem of the reproducibility of the extraction of ancient DNA from amber. They highlighted that nobody is able to reproduce the DNA sequence extracted from amber and copal. On this observation, Austin et al. (1997) argued that this lack of reproducibility is the proof that ancient DNA does not exist in amber. Parducci and Bennett (2005) commented on this argument and concluded that it is probably the main reason for the weak development of the research on ancient DNA. This argument shows that the problem of reproducibility was not resolved in 2005. Hebsgaard et al. (2005) explained this lack of reproducibility by inadequate experimental setups for the study of ancient DNA, providing insufficient authenticity of results. Amber had been thought to be the perfect medium for DNA preservation, but Orlando (2005), in a review of ancient DNA research, argued that the only way to get very old DNA is to preserve it in a dark and cold medium (such as Siberian permafrosts). Nobody studies in detail the fate of DNA in plant resin as the polymerisation processes transform resin to amber. At the moment, there is no real proof, according to us, for ancient DNA preserved in amber.

This short review of the study of ancient DNA from amber shows that many doubts exist around the results and the study of DNA from amber cannot be used as a tool for identifying fossil micro-organisms from amber. Thus there exists only one technique that can be used: optical morphological identification. Indeed, the identification of morphological structures remains the only way to identify microfossils. However it raises issues regarding the morphological identification of micro-organisms. Regarding bacteria, it is clear that no precise affinity can be found only on morphological characteristics. Waggoner (1994b) stated that the attribution of a fossil Actinobacteria to a modern group is tentative. Such statement can be expanded to almost all groups of bacteria. Only rare cases allowed more precise identification of bacteria to genus on morphological characters (Breton and Tostain, 2005, Girard et al., 2009a and Schmidt and Schäfer, 2005). When studying bacteria from amber, one might remember that identification of bacteria is tentatively and any taxonomical attribution might be doubtful.

Regarding the eukaryotes, the problem is only a little simpler. Fungi have many diverse morphological characteristics that allow a more precise identification, especially on the identification of spores and reproductive structures. The main problem remains the fragmentation of fungal remains in amber. Most of the time, it consists only of isolated fragments of fungal hyphae. Only rare cases provide hyphae, spore and other fungal structures preserved in the same amber piece and allow precise identification (e.g. Girard, 2010, Schmidt et al., 2008 and Schmidt et al., 2010a). Regarding protists, the problem is more complex. Modern species have been traditionally identified on morphological characters. However the development of new techniques, and more precisely of genetic techniques, has provided doubts about the validity of such morphological identification. However Fenchel and Finlay (2006) argued that the identification of protists with genetic methods (such as barcoding system) is also problematic as the results greatly depends of the choice of the genes selected. Using the case of the species Cyclidium glaucoma (a ciliate), Fenchel and Finlay (2006) argued that morphological, physiological and ecological properties of protists are good characters to identify protist species. Such methods can thus appear perfect to identify micro-organisms from amber. However, both Adl and Gupta (2006) and Foissner (2006) provided evidence that the Fenchel and Finlay (2006) argument was simplistic, and that they had dismissed much of the existing evidence regarding the inability of morphological characters to distinguish among species of protists. Furthermore Foissner et al. (1999) showed that, depending on the resin the preservation of some protists can greatly differ from one amber to another. Many external and internal characters are quickly lost after embedding of the micro-organisms in the resin. This is problematic for the identification of amber microfossils as most of the diagnostic features that allow the determination can be lost early in the fossilisation processes due to poor fixation. One might keep in mind this problem when studying amber microfossils. We therefore propose that species identification for fossilised micro-organisms is tentative because all tools used to identify modern species cannot be used. Only a morphotypic resemblance can be determined. This morphological identification is susceptible to error due to poor preservation of the original morphology.

Recent discoveries of micro-organisms in various ambers (for more details, see Girard, 2010) led some authors to propose that morphological stasis exists in different microbial lineages. This concept was formally formulated by Martín-Gonzáles et al. (2008) in their article entitled “Morphological stasis of protists in Lower Cretaceous amber”. These authors argued that their observation confirmed a conclusion of Fenchel and Finlay (2006), i.e. that “microbial phenotypes are ancient in term of the geological time-scale and have been maintained through stabilizing selection”.

A quick review of the literature allows one to see that the fossil record of micro-organisms in amber seems to support this concept. In the tables (Table 1, Table 2, Table 3, Table 4, Table 5, Table 6 and Table 7) of this article, the modern genera and species that have been found in amber are underlined. Schliersee amber (Cenomanian, Germany) is probably the best example. Six taxa of photosynthetic protists have been identified in this amber. Only one corresponds to a new genus (Dörfelt and Schäfer, 2000). Four green algae were attributed to the extant genera Chlamydomonas, Chloromonas, Choricystis and Chlorella and an euglenid was placed in the modern genus Euglena (Schönborn et al., 1999). Two fungal taxa have also been found, one belonging to a new genus, one to the extant genus Pithomyces. Sixteen taxa of amoebae and ciliates have also been found in Schliersee amber (Poinar et al., 1993b, Schmidt et al., 2004 and Schönborn et al., 1999). This assemblage comprises only one new genus (Triassamoeba) and three new species (Hyalosphenia baurei, Paramecium triassicum, Triassamoeba alpha) whilst thirteen taxa correspond to extant species.

Schliersee amber is not unique. Most of the other ambers have preserved micro-organisms that have been attributed to modern taxa (such as the genera Euglena, Paramecium, Centropyxis). These discoveries seem to corroborate the concept of morphological stasis. These apparent stases can result from different factors. First many micro-organisms have broad distribution (for some a worldwide distribution) as mentioned by Fenchel and Finlay (2006). Second these authors argued protists have large population size and slow population replacement, so that phenotypes can persist for millions of years.

However, even though the concept of morphological stasis can be “sexy”, some doubt can be erected concerning its foundation. The fossil record on which it has been formulated needs to be revised. First, some amber micro-inclusions have been described to be modern taxa of eukaryotes whilst their biological origin is still doubtful. Many of them could be non-biological inclusions only. That means many supposed fossils supporting that concept are not real fossils. Reinvestigations of all these inclusions will help to clarify this point. Second the preservation of the micro-inclusions in amber can greatly influence the identification of microfossils. For example, many testate amoebae have been found in French mid-Cretaceous amber (Girard, 2010). Most of them have been attributed to modern genera (such as Arcella, Centropyxis, Hyalosphenia), but their poor preservation state did not allow further identification. All the very well preserved specimens clearly show differences from modern taxa (such as the species Centropyxis perforata and Leptochlamys galippei that differ from the modern Centropyxis and Leptochlamys by the presence of pores on their test (Schmidt et al., 2010a)). Thus, with the current fossil record of eukaryotes from amber, it is difficult to corroborate or infirm the concept of morphological stasis.

What about prokaryotes? Even though the concept of morphological stasis was erected for eukaryotes, prokaryotes from amber seem also to support it. Some of the morphologically identified prokaryotes have been attributed to modern taxa (Waggoner, 1993, Waggoner, 1994a, Waggoner, 1994b and Waggoner, 1996). However there is little morphological diversity among prokaryotes. They are identified using molecular techniques, and their diversity is evidenced in their metabolism and DNA sequence, not by morphology. Molecular techniques can be used when dealing with amber microfossils since we are dealing with fossilised remains, and the morphological identification of amber prokaryotes is very uncertain. It is probably impossible to correctly assign most bacteria to genera based on morphological similarity. Several authors proposed that prokaryotic species cannot exist stably, due to frequent uptake of environmental DNA and due to lateral transfers of genes (Doolittle, 1999, Ochman et al., 2000 and Vishwanath et al., 2004).

The fossil record of protists has been greatly improved by recent studies of microfossils. Most of the specimens have been attributed to modern genera and sometimes to modern species. The lack of DNA in fossilised amber specimens prevents identification of specimens beyond morphological similarity with extant species. The discoveries from amber support the concept of morphological stasis as formulated by Martín-Gonzáles et al. (2008) although with some caution as we indicated. This concept corroborates the conclusion of Fenchel and Finlay (2006), that microbial phenotypes are geologically ancient, having been maintained through stabilising selection.

However, doubts about the real origin of several of these amber micro-organisms have been formulated (Girard, 2010 and Schmidt et al., 2010a), especially regarding the vacuolated inclusions that have not biological origin according to us. It seems that many supposed protists from amber can only be non-biological protist-like inclusions mimicking protists. These possible wrongfully identified amber protists disturb our knowledge about microbe fossil record and do not allow us to corroborate or infirm any hypothesis about the evolution of terrestrial protists.

Re-investigation of these possible protists from amber is absolutely necessary in order to clarify the fossil record of the different terrestrial protists. The concept of morphological stasis can have great repercussion in the study of amber microfossils. If it is corroborated, it can raise the problem of the definition of a fossil microbial species. Indeed, even the systematic affinities of these amber micro-inclusions are confirmed, can we really imagine that species have been maintained for several dozens of millions of years? It has been demonstrated that lateral transfers of genes are very common for microbes (especially for prokaryotes (Doolittle, 1999, Ochman et al., 2000 and Vishwanath et al., 2004)). Such phenomena seem to be incompatible with the fact that species were maintained for millions of years. This problem raises, for terrestrial microbes, the problem of how to define a species, on the basis of morphology alone. Data obtained on modern taxa seems to demonstrate that morphological studies are also inappropriate for microbes. Indeed the monophyly of modern microbes could not be defined by their morphology most of the time, but resolved by molecular analyses (Roger and Hug, 2006 and Simpson and Roger, 2004).

It is difficult (even impossible) to verify the monophyly of fossils species with their modern relatives. To get a better definition of a fossil microbe species, the increase of common works between molecularists and morphologists is an absolute necessity.