Early Middle Ordovician evidence for land plants in Argentina (eastern Gondwana)


  • C. V. Rubinstein,

    1. Unidad de Paleopalinología, Departamento de Paleontología, Instituto Argentino de Nivología, Glaciología y Ciencias Ambientales, CCT CONICET – Mendoza, A. Ruiz Leal s/n, Parque General San Martín, 5500 Mendoza, Argentina
    Search for more papers by this author
  • P. Gerrienne,

    1. Paléobotanique, Paléopalynologie, Micropaléontologie (PPM), Department of Geology, University of Liège, B-18, Sart Tilman, BE–4000 Liège 1, Belgium
    Search for more papers by this author
  • G. S. de la Puente,

    1. Unidad de Paleopalinología, Departamento de Paleontología, Instituto Argentino de Nivología, Glaciología y Ciencias Ambientales, CCT CONICET – Mendoza, A. Ruiz Leal s/n, Parque General San Martín, 5500 Mendoza, Argentina
    Search for more papers by this author
  • R. A. Astini,

    1. Laboratorio de Análisis de Cuencas, Centro de Investigaciones en Ciencias de la Tierra, CONICET-Universidad Nacional de Córdoba. Pabellón Geología, Ciudad Universitaria, 2º Piso, Oficina 7, Av. Vélez Sársfield 1611, X5016GCA-Córdoba, Argentina
    Search for more papers by this author
  • P. Steemans

    1. Paléobotanique, Paléopalynologie, Micropaléontologie (PPM), Department of Geology, University of Liège, B-18, Sart Tilman, BE–4000 Liège 1, Belgium
    Search for more papers by this author

Author for correspondence:
Claudia V. Rubinstein
Tel: +54 261 5244217
Email: crubinstein@mendoza-conicet.gov.ar


  • The advent of embryophytes (land plants) is among the most important evolutionary breakthroughs in Earth history. It irreversibly changed climates and biogeochemical processes on a global scale; it allowed all eukaryotic terrestrial life to evolve and to invade nearly all continental environments. Before this work, the earliest unequivocal embryophyte traces were late Darriwilian (late Middle Ordovician; c. 463–461 million yr ago (Ma)) cryptospores from Saudi Arabia and from the Czech Republic (western Gondwana).
  • Here, we processed Dapingian (early Middle Ordovician, c. 473–471 Ma) palynological samples from Argentina (eastern Gondwana).
  • We discovered a diverse cryptospore assemblage, including naked and envelope-enclosed monads and tetrads, representing five genera.
  • Our discovery reinforces the earlier suggestion that embryophytes first evolved in Gondwana. It indicates that the terrestrialization of plants might have begun in the eastern part of Gondwana. The diversity of the Dapingian assemblage implies an earlier, Early Ordovician or even Cambrian, origin of embryophytes. Dapingian to Aeronian (Early Silurian) cryptospore assemblages are similar, suggesting that the rate of embryophyte evolution was extremely slow during the first c. 35–45 million yr of their diversification. The Argentinean cryptospores predate other cryptospore occurrences by c. 8–12 million yr, and are currently the earliest evidence of plants on land.


Embryophytes are a monophyletic group of photosynthetic eukaryotes that share the following characters: a diplobiontic life cycle; a terrestrial growth habit; a cuticle; antheridia; archegonia (Kenrick & Crane, 1997a; Karol et al., 2001). Another important character of embryophytes is the ability to synthesize sporopollenin, an extremely resistant polymer that is a major component of the outer cell wall (exine) of spores and pollen grains. The presence of sporopollenin in exine was presumably a decisive factor during the early terrestrialization of plants because it allows long-range dispersal of spores, and provides structural support and partial protection against UV light, desiccation and microorganism attack. The production of spores covered by sporopollenin is generally accepted to indicate unambiguous embryophyte affinities (Kenrick & Crane, 1997b; Wellman et al., 2003; Gensel, 2008). Sporopollenin is also found in Coleochaete (Charophyta; Delwiche et al., 1989), but in this case it is deposited on the fertilization product (the zygote), and not on the meiosis products (zoospores).

Colonization of the land by plants presumably occurred in a step-wise fashion starting during the Early Paleozoic with plants at a bryophyte, most likely liverwort, grade of organization (Steemans et al., 2009). It resulted in acceleration of weathering processes and in the formation of modern terrestrial environments, including structured soils and complex microbial communities (Algeo et al., 2001); it also profoundly affected carbon cycling, changed the atmosphere composition and irreversibly altered climates (Algeo et al., 2001; Berner, 2001; Graham et al., 2004; Beerling & Berner, 2005). Despite outstanding recent advances, the timing and locations of origination of all major embryophytes traits (alternation of generations, spores, cuticle and stoma, vascular tissue, leafs, wood, seeds, etc.) that lead to their dominant position in almost all terrestrial ecosystems are the subject of active controversies, especially concerning the early evolution of sporopollenin-coated propagules (Steemans et al., 2010). The propagules of the earliest embryophytes are cryptospores (i.e. spores that lack haptotypic features such as a trilete or a monolete mark) (Steemans, 2000). Cryptospores may be preserved in monad, dyad or tetrad configurations, with different types of contact areas between individual spores and diverse external features: cryptospore outer walls are smooth or ornamented; the cryptospores themselves are naked or enclosed within an envelope; the envelopes are smooth or ornamented. Before this work, the oldest uncontroversial records of embryophytes were late Darriwilian/Llanvirn (c. 463–461 million years ago (Ma)) cryptospores from the Czech Republic and from Saudi Arabia, on the western margin of the Gondwana paleocontinent (Vavrdová, 1990; Strother et al., 1996). Analysis of wall ultrastructure of slightly younger cryptospores from Oman suggests that they were produced by early liverworts (Wellman et al., 2003). In this paper, we report the discovery of a moderately diverse cryptospore assemblage from an early Middle Ordovician locality in Argentina (see the Supporting Information, Fig. S1).

Materials and Methods

Palynological samples were collected from the Rio Capillas outcrop, southern part of the Sierra de Zapla of the Sierras Subandinas (Subandean Ranges), Central Andean Basin, north-western Argentina (Fig. S2). The Ordovician section exposes in stratigraphic order the upper part of the Zanjón Formation, the Labrado Formation and the lower part of the Capillas Formation (Fig. 1). Based on sedimentological and ichnological data, the Zanjón Formation is interpreted as a succession of tidal parasequences that shallow up into heterolithic and muddy intervals and subaerial exposure features (Astini, 2003; Astini & Marengo, 2006). Accordingly, the Zanjón Formation is interpreted to represent mostly estuarine environments.

Figure 1.

 Middle Ordovician units exposed in the Río Capillas showing stratigraphy, lithology and location of samples. The biostratigraphic range of some important palynomorphs is provided. The sample studied here is from level 7999. A specimen of the chitinozoan biostratigraphic marker, Lagenochitina combazi, is illustrated. Bar, 50 μm. Lithology: (a) bioturbation, (b) siltstone/mudstone, (c) sandstone, (d) cross-bedded sandstone. Palynomorphs: (e) cryptospores, (f) chitinozoans, (g) acritarchs.

Five samples were collected from upper part of the Zanjón Formation and its transition to the Laja Morada Member of the Labrado Formation (Fig. 1) that are assigned an early to middle Dapingian age (early Middle Ordovician) on the basis of the presence of the chitinozoan biostratigraphic marker Lagenochitina combazi (Chen et al., 2009; F. Paris, pers. comm.) (Fig. 1; Note S1). This age is consistent with the presence of the acritarch genus Aremoricanium, which has a first appearance datum in the Dapingian (Brocke et al., 2000). Sample 7999 (Fig. 1) from the Zanjón Formation includes a moderately diverse cryptospore assemblage described here.

Samples were carefully cleaned before processing in order to prevent contamination from modern material. The samples were processed at the laboratory of the Paleopalynology Unit, IANIGLA, CCT CONICET – Mendoza, Argentina, using standard palynological HCl–HF–HCl acid maceration techniques. A duplicate sample was processed at the palynological laboratory of the University of Liège, Belgium. Each sample was crushed and c. 30 g were demineralized in HCl–HF. The residue was oxidized in HNO3 and KClO3 and sieved through a 12 μm mesh. The palynological characteristics of preparations from both laboratories were identical and included cryptospores.


The cryptospores are relatively well preserved, allowing the identification of several specimens at generic or specific level (Figs 2, S3). The assemblage comprised monads and tetrads. The assemblage includes the following taxa. (1) Chomotriletes? sp. (Fig. 2a,b; Note S2) is an ornamented hilate monad consisting of a subcircular cryptospore distally sculptured by convolute muri mimicking finger prints. The specimens are close to Chomotriletes? sp. A from the Late Silurian of the Urubu area, Brazil (Steemans et al., 2008). (2) Gneudnaspora (Laevolancis) divellomedia (Balme, 1988) or Laevolancis chibrikovae (Steemans et al., 2000) (Fig. 2c,d) is a monad with subcircular amb and an equatorial and distal patina delimitating a thin proximal laevigate hilum. The relatively poor preservation of the specimens does not allow a secure identification of the species. (3) Sphaerasacus glabellus (Steemans et al., 2000) (Fig. 2e,f) is a circular monad with a thick laevigate wall, enclosed within a thin and narrow envelope. Specimens are similar to those observed in the Late Ordovician (Wellman, 1996) and in the Llandovery (Steemans et al., 2000). (4) Tetrahedraletes cf. medinensis (Strother & Traverse, 1979) (Fig. 2g) is a tetrad of subcircular, laevigate spores, with a very thin exine. (5) A new genus (Fig. 2h) consists of a quadraedric tetrad of subcircular laevigate spores with equatorial crassitude, enclosed within a thin, loose, apparently smooth envelope. Identical tetrads have been found at a Llandovery locality from Paraguay (Mendlowicz Mauller, 2004) and at Middle to Late Ordovician localities from Saudi Arabia Commission Internationale de la Microflore du Paléozoïque (CIMP; http://www.cimp.ulg.ac.be/) working group on Saudi Arabia, work in progress).

Figure 2.

 Cryptospore plate from upper part of the Zanjón Formation. (a) Distal face of Chomotriletes? sp.; (b) proximal face of Chomotriletes? sp.; (c,d) Gneudnaspora (Laevolancis) divellomedia or Laevolancis chibrikovae; (e,f) Sphaerasacus glabellus; (g) Tetrahedraletes cf. medinensis; (h) new genus. Bar, 20 μm. (See the Supporting Information Fig. S3 for color plate.)


The cryptospore assemblage described here is of Dapingian age (c. 473–471 Ma), thus predating by c. 8–12 million yr the previous earliest known cryptospores from the Czech Republic (Vavrdová, 1990) and from Saudi Arabia (Strother et al., 1996) (Fig. S1). There is a growing body of evidence that terrestrial life originated on the Gondwana paleocontinent. The early cryptospore record (Vavrdová, 1990; Strother et al., 1996), the earliest plant fragments (Wellman et al., 2003) and the earliest trilete spores (Steemans et al., 2009), most probably produced by the earliest vascular plants or their immediate ancestors, were all described from localities on the western margin of Gondwana. On the basis of the discovery of this Argentinean cryptospore assemblage and according to the paleogeographic configuration (Fig. S4) of the Gondwana during Dapingian times (Scotese, 2003), we suggest that the early steps of the terrestrialization of plants could rather have occurred on the eastern margin of the continent, some six thousand kilometers more to the east, and that plants only subsequently reached the western part of Gondwana. Interestingly, localities from Argentina and from the Czech Republic/Saudi Arabia were all located at c. 35°S latitude (Scotese, 2003) during Middle Ordovician times (Fig. S4).

The classification of green organisms (green algae and embryophytes) is still actively debated, but it is generally accepted that embryophytes originated from charophycean green algae, with Charales being their most probable closest living relatives (Karol et al., 2001; Qiu et al., 2007), and that liverworts are the most basal lineage in embryophyte phylogeny (Qiu et al., 1998, 2007). The estimated time of divergence of Charales and embryophyte ranges from c. 805 Ma to c. 438.8 Ma, depending on the study considered (Magallón & Hilu, 2009). The cryptospores described here are c. 470 million yr old. Their occurrence may be considered a reliable estimate of the latest possible date for the origin of embryophytes. It also provides a solid calibration point for molecular evolutionary models used to estimate the divergence time of the major lineages of embryophytes. Our discovery of Dapingian cryptospores further supports the basalmost position of the liverworts in embryophyte phylogeny.

During Ordovician times, the Sierras Subandinas were the outermost region of the proto-Andean foreland basin, positioned at the eastern Gondwana margin. The Zanjón Formation that yielded the cryptospore assemblage described here has been interpreted as mostly representing estuarine environments (Astini, 2003; Astini & Marengo, 2006). Unfortunately, this does not give any information on the actual origin of the cryptospore-producing plants, as small particles such as cryptospores can be transported in water over very long distances (Traverse, 2007). Nevertheless, the presence of early cryptospores in nearshore or estuarine environments suggests that their mother plants did not live in remote upland areas disconnected from major hydrographical networks, as had previously been suggested for much younger, earliest Devonian cryptospores (Steemans, 2000).

The assemblage described here includes five cryptospore genera. Both from morphological and systematic points of view, this Dapingian assemblage is not different from younger cryptospore occurrences, including Aeronian (Early Silurian, c. 439–436 Ma) assemblages. This seems to indicate that the evolutionary rate of the earliest embryophytes was extremely low, or that selective pressures did not act on the morphology of propagules.

The Dapingian cryptospore assemblage from Argentina currently represents the earliest unequivocal fossil evidence of the presence of plants on emerged lands (Fig. S1). It suggests that the earliest steps of plants in terrestrial habitats may have occurred on the eastern margin of the Gondwana paleocontinent (Scotese, 2003) (Fig. S4). Although moderate, its diversity (five different genera) implies that the very origin of land colonization by plants is presumably older, and that it may have occurred during Early Ordovician times (488–472 Ma) or even Cambrian (c. 542–488 Ma) times (Strother & Beck, 2000).


We thank E. Javaux, M. Streel (University of Liège, Belgium) and C. M. Berry (Cardiff University, Wales) for critical reading of the manuscript. We also thank F. Paris (University of Rennes, France) for critical comments on the biostratigraphic value of the chitinozoan used in this paper. This work has been supported by FONCyT (PICT 2006 N° 01272), CONICET (PIP N° 11220090100921) and MINCYT-FNRS Project (Scientific Cooperation Program between Argentina and Belgium). This work is a contribution to the International Geoscience Program IGCP 503 ‘Ordovician Palaeogeography and Palaeoclimate’. PG is a F.R.S.-FNRS Research Associate; PS is a F.R.S.-FNRS Senior Research Associate.