Chitinozoans are an enigmatic group of organic-walled marine microfossils with an abundant Lower Ordovician to Upper Devonian fossil record. They achieved maximum species diversity during the Middle Ordovician–Silurian, and their history during the group’s acme is well documented. Nevertheless, information about their origin and early evolution is sparse. Here we report three phosphatized flask-shaped vesicles recovered from the Cambrian Stage 5 (∼510 m.y. old) Duyun fauna of southern China as the earliest known chitinozoans, Eisenackitina? sp., extending the record of the microfossil group back by at least 20 m.y. Their exceptional occurrence within an Orsten-type Lagerstätte might imply a benthic mode of life. This discovery is significant in that it supplies critical data for evaluation of the basic morphology of Cambrian chitinozoans and assessment of their early evolution.


Since their first discovery ∼80 yr ago (Eisenack, 1931), chitinozoans have become an important fossil group; they have a rich record in Paleozoic rocks of Ordovician to Devonian age (Paris et al., 1999; Grahn and Paris, 2011). Their temporal and spatial occurrences are used for biostratigraphy (Webby et al., 2004) and for the assessment of paleoenvironments and paleoclimates (Vandenbroucke et al., 2010). The organic-walled vesicles of chitinozoans are usually preserved as isolated individuals, but may also be found in chains or in aggregate masses, sometimes associated with organic envelopes (cocoons) (Kozlowski, 1963; Gabbott et al., 1998). Theories on the biological affinity of chitinozoans have long attracted attention. The current consensus is that these microfossils are the egg cases of small, entirely soft-bodied marine metazoans (Paris and Nõlvak, 1999); this is supported by the discovery of three-dimensional coiled catenary structures, interpreted as immature eggs, within the decayed body of a parent organism (Paris and Nõlvak, 1999). Some tear- and flask-shaped microfossils from the late Precambrian Chuar Group of the Grand Canyon (Arizona, United States) were initially regarded to be the oldest chitinozoans (Bloeser et al., 1977). Subsequently, on the basis of evidence from increasingly documented Precambrian material, these essentially vase-shaped microfossils were suggested to be Protozoa (Schopf, 1992), most closely related to the testate amoebae (Porter and Knoll, 2000; Porter et al., 2003). To date, the origin, early evolution, and systematic position of chitinozoans have remained unclear. Thus, the discovery of chitinozoans from Cambrian Stage 5, reported here, offers fresh insight into the nature and early history of these organisms.


Approximately 600 kg of limestone nodules, collected from the Gaotai Formation at the outcrop (see Fig. DR1 in the GSA Data Repository1) near Balang village in Duyun, Guizhou, China (for locality details, see Zhang et al., 2011, their figure 1), were digested in acetic acid of low concentration (5%), and the insoluble residue was taken out of the acid solution every day. All phosphatized specimens were picked from the residue under a stereomicroscope (at 20× magnification), and then images were taken (secondary electron: 10–20 kV) using scanning electron microscopy (model FEI Quanta 200).


Despite being based solely on morphological characters, the biological significance of which is largely unknown, a Linnaean classification is adopted for chitinozoans. Here we follow the most recent generic and suprageneric classification of Paris et al. (1999).


Family Desmochitindae Eisenack, 1931, emend. Paris, 1981

Subfamily Eisenackitininae Paris, 1981

Genus Eisenackitina Jansonius, 1964, restrict. Paris, 1981

Eisenackitina? sp. (Figs. 1A–1I)


Key Laboratory for Paleobiology, Yunnan University. Specimen YKLP 12001 (Figs. 1A–1C) is a nearly complete, three-dimensionally preserved vesicle; YKLP 12002 has a broken aperture but a funnel-like copula (Figs. 1D–1F); YKLP 12003 is an internal mold of a chamber (Figs. 1G–1I).


Cambrian Stage 5 (Middle Cambrian), Gaotai Formation, Balang section in Duyun, Guizhou Province, southern China. The specimens co-occur with numerous lingulid brachiopod valves, bradoriid arthropod carapaces, rare scalidophoran embryos (Zhang et al., 2011), various small shelly fossils, and exoskeletons of the eodiscoid trilobite Pagetides qianensis (Zhang and Clarkson, 2012), which, as a key index fossil, confirms the chronostratigraphic age for the fossil assemblage.


Vesicle flask-shaped and claviform, with an ovoid chamber, 520–670 μm long and 230–325 μm wide, exhibiting a length/width ratio of ∼2:1, and tapering gradually from the basal region toward the aperture. Body chamber is subspherical to ovoid, displaying a convex and rounded base. No flexure, neck indistinct, but ∼1/3–1/2 of the total length of the vesicle, gently tapering forward and ending with a flaring collarette with a sinuate rim around the aperture. Surface of test wall slightly spiny to glabrous with many small, simple spines (length 3–8 μm) that are mostly concentrated on the neck. The spiny ornamentation fades away on the chamber. Well-developed copula on the base with some tiny holes (perforations) scattered around the base (Figs. 1C, 1F, and 1I). The fossils are phosphatized (Fig. DR2 in the Data Repository), as is to be expected with Orsten-style preservation (Maas et al., 2006).


Individual chitinozoans are organic-walled, urn-, flask-, tube-, or bottle-shaped vesicles that are distinctly polar (they have an aperture on one side and the other side is closed) and are radially symmetrical around their longitudinal axis. They have a chamber and a neck (largely missing in desmochitinids), are limited to between 50 μm and 2 mm in size (most commonly between 100 μm and 1000 μm) and their aperture can be sealed with a plug (Paris et al., 1999). A wide variety of chamber morphologies and surface ornamentation exists. They often exhibit structures that reflect their original chained nature (e.g., a copula). The Duyun specimens conform to most of these characteristics, although the plug has not been preserved (it is often missing in younger specimens) and the phosphatization process obscures the original test composition. In the Duyun specimens, the copula is hollow but contains some residue near the vesicle bottom (Figs. 1E and 1H). Therefore, we cannot prove the presence of a membrane between the copula and chamber, but evidence of noncommunication between the inside of the vesicle and the outside at the aboral pole, an essential trait in chitinozoans, is provided by differential preservation of the chamber (filled and/or internal mold) and copula (hollow) in Figures 1G and 1H.

More specifically, the Duyun chitinozoans display some diagnostic characters of the genus Eisenackitina within the family Desmochitinidae and the order Operculatifera, especially the spiny ornamentation and the overall vesicle shape. This assignment is tentative, as the vesicle shape is a little elongate (indistinct neck). An apparent basal “spine” is not typical of this genus but is likely an artifact (YKLP 12001; Fig. 1A). The material also resembles the usually smooth-walled genera Linochitina and Urnochitina, which are other desmochitinids with a short copula. Chitinozoan classification at the order level is based on the nature of the vesicle seal, either a prosome for the order Prosomatifera or an operculum for the order Operculatifera. This plug is a structure that is not preserved in the Duyun specimens (Fig. 1B), so although the overall shape suggests that the specimens should be within the Operculatifera, they could also be within the Prosomatifera. Some conochitinids and belonechitinids (Order Prosomatifera, Family Conochitinidae) have a very well developed mucron, similar to the copula structure seen here, and could therefore be a potential genus for these specimens within this order (the basal structure of the Cambrian specimens is too small to be an Eremochitina copula). The exact systematic position of the Duyun chitinozoans within the Chitinozoa thus remains somewhat circumspect, with the available morphological evidence suggesting they are in between Eisenackitina and Belonechitina, and this is symptomatic of most form-taxonomic classifications. It is interesting that some of the characters (copula, urn and/or claviform shape, spines) recognized are not generally considered to be basal (Paris et al., 1999), as they are not known from the earliest Ordovician species, but from stratigraphically younger taxa.


These Eisenackitina? specimens also superficially resemble some tintinnid loricae, which are flask shaped with a spherical chamber, a caudal appendage, and an outward-turned flaring collar around the aperture (Colom, 1948), but the average size of typical tintinnids, <200 μm in length, is smaller than that of Eisenackitina? sp. More notably, tintinnids differ from chitinozoans in accumulating mineral or sediment particles on the surface of the lorica: there is no indication of this in the Duyun material. The putative tintinnids from the Neoproterozoic of Mongolia (Bosak et al., 2011) have simpler morphology, expressed as alveolar spheroids.

Some foraminifera also resemble the Duyun Eisenackitina? sp. in size and shape. Their fossil record extends back to the early Cambrian (Culver, 1991; Streng et al., 2005), or possibly to the Ediacaran (Hua et al., 2010). Cambrian foraminifera are all unilocular and agglutinated (Culver, 1991; Streng et al., 2005), using mineral or sediment particles to construct their tests, unlike chitinozoans. There are no known foraminifera that display the wide aperture with flaring collarette or the copula apparent in Eisenackitina? sp.

Vase-shaped microfossils are well known from the late Precambrian (Bloeser et al., 1977; Porter and Knoll, 2000; Porter et al., 2003) and early Cambrian of China (Cao, 1998), but appear to be absent from younger strata. With their unilocular vesicles, some vase-shaped microfossils are superficially similar to the Duyun Eisenackitina? sp., which is larger than most vase-shaped microfossils, and differs in having some typical chitinozoan characteristics, i.e., a spiny surface, a flaring collarette with a sinuate rim, and a copula on the base.


The Duyun chitinozoans, like the co-occurring scalidophoran embryos (Zhang et al., 2011), are extremely rare but well preserved in carbonate nodules. This can be explained in either or both of two ways: (1) under certain exceptional conditions (i.e., the exquisite phosphatization), chitinozoans have the same preservation potential as the coexisting embryos; or (2) like other fossils from Orsten-type Lagerstätte, the earliest chitinozoan animals may have lived near the sediment flocculent layer (Waloszek, 1993) with a benthic life style. As small populations with restricted occurrences the Cambrian chitinozoans left little fossil record until their later colonization of pelagic niches (Grahn and Paris, 2011) and preservation in organic-rich pelagic mudstones. Alternatively, the early chitinozoans may already have adopted a floating mode of life, with the small number of chitinozoans preserved in the Duyun fauna representing chance material drifting to the seabed from the overlying water column; this hypothesis, however, does not explain why such pelagic microfossils are not more widely recovered from Cambrian strata. In this context, our hypothesis of a benthic antecedent for planktonic chitinozoans (Fig. 2) would be consistent with many other organism groups that have colonized the water column, including ostracods, graptolites, trilobites, decapods, and chaetognaths (Rigby and Milsom, 2000).

Previously known chitinozoans have an organic wall, which is not made out of chitin, but consists of a kerogen network dominated by aromatic units and with few aliphatics (Jacob et al., 2007). The three-dimensionally phosphatized Duyun chitinozoans show little sign of carbon (Fig. DR2), but display the typical preservation associated with Orsten-style deposits (Maas et al., 2006). In Orsten-style preservation the tissues of small organisms are rapidly impregnated or coated with phosphate. This process occurs early and is responsible for the exceptional preservation of many Cambrian soft-bodied faunas. The chitinozoans, although well known to be unaffected by many other diagenetic and geochemical processes, were presumably phosphatized in the same manner as all other small organisms in the Duyun fauna, which had various original compositions (see the Data Repository). Younger, Ordovician–Silurian chitinozoans are not usually recovered as phosphatized vesicles, but by that time, the typically Cambrian Orsten-style preservation had disappeared.

Evolutionary changes in chitinozoan tests have been reported to include size decrease, shape change toward the development of a distinct neck with a spherical chamber, and acquisition of surface ornament (spines, carinae, processes) (Paris et al., 1999; Armstrong and Brasier, 2005). Paris (1981) predicted that early chitinozoans were large; this is consistent with the Cambrian material described here, and with some Early Ordovician representatives. Moreover, the Cambrian chitinozoans had some characteristic ornament and structures associated with chitinozoans of the Ordovician. Thus, these early representatives of Chitinozoa, as with the initial radiation of many metazoan clades during the Cambrian explosion, appear to show rapid innovation of some advanced traits (Lee et al., 2011). In addition, the newly discovered chitinozoan roots in the Cambrian explosion, together with their well-known radiation during the Great Ordovician Biodiversification Event (GOBE), might provide evidence for the hypothesis that there is a link between these two major early Paleozoic biological events, seen as clearly separate events by some, or as part of one larger event by others (Servais et al., 2010). The stratigraphic range of chitinozoans as a group has long been used (along with their paleoecology) to exclude a number of potential biological affinities; this will now have to be revised.

The minute circular perforations (ranging in diameter from ∼3.0 μm to 4.7 μm) concentrated around the base of Eisenackitina? sp. (Figs. 1C, 1F, and 1I) are comparable to the regular perforations or borings (Eisenack, 1968; Wrona, 1980) in later chitinozoans, interpreted to be formed by parasitic heterotrophic bacteria or fungi. Microborings have also been reported in other Cambrian Stage 3 phosphatic or phosphatized fossils, where they show selective infestation; this implies a heterotrophic mode of life for the infesting microorganisms (Zhang and Pratt, 2008). Likewise, these perforations in Eisenackitina? sp. may attest to an early origin and long history for the interrelationship between the parasites and their chitinozoan hosts.


Eisenackitina? sp. from Cambrian Stage 5 is regarded here as the earliest known representative of the Chitinozoa. Its discovery extends the established chronological age of the earliest Chitinozoa from 488 Ma (Grahn and Paris, 2011) to at least 510 Ma. This establishes that the Chitinozoa, like the majority of metazoan groups, have a fossil record extending back to the Cambrian, their origins being a part of the Cambrian evolutionary radiation. Although chitinozoans diversified during the GOBE, their occurrence in the Cambrian indicates an earlier and intriguing evolutionary story (Fig. 2).

Supported by the National Natural Science Foundation of China (40972012), the Natural Science Foundation of Yunnan Province, China (2008CC005), and the Agence Nationale de la Recherche, France (project RALI [Rise of Animal Life]). We thank F. Paris and three anonymous referees for their critical and constructive comments, J.W. Schopf for valuable suggestions, T. Lan and J.-B. Hou for help in scanning electron microscope examination, and H.Q. Zhang for sample preparation.

1GSA Data Repository item 2013047, Figure DR1 (stratigraphy of the Balang outcrop) and Figure DR2 (SEM-EDX analysis of the chemical composition of Cambrian fossils from the Balang outcrop), is available online at www.geosociety.org/pubs/ft2013.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.