Land plants arrived tens of millions of years later than we thought, spore fossils reveal

Land plants may have transformed the planet considerably later than molecular clock studies have long suggested. A new analysis, published in Nature Plants, reconstructs how the characteristics of plant spores evolved and uses that reconstruction to recalibrate the timescale of early land plant diversification. The study, led by Jack Hooper, Annabel M. Worth, Charles H. Wellman, James W. Clark and Philip C. J. Donoghue, concludes that the common ancestor of all living land plants originated in the Late Ordovician, between 457 and 450 million years ago, roughly 55 million years later than the equivalent estimate from a previous analysis of the same molecular dataset.

The problem the researchers set out to solve is a long-standing one. The earliest phase of land plant evolution is documented by an abundant fossil record of sporopollenin-walled spores, but that record has been difficult to interpret because the sequence in which plant spore characteristics evolved was unresolved. Different fossils have been assigned to different positions on the tree, and the resulting divergence time estimates have varied enormously, from the Mesoproterozoic to the Early Ordovician, with recent studies converging on a late Cambrian to Early Ordovician origin for the land plant crown group.

To pin down the pattern of spore evolution, the team compiled a dataset of 27 discrete characters describing spore morphology, spore ontogeny and the life history of the parent sporophyte. These characters were scored for 27 extant and 11 extinct genera, chosen to represent the principal clades of bryophytes and tracheophytes and their algal outgroups. Ancestral state estimation was then performed on three timetrees: one containing only extant taxa, and two others that also included fossil genera and differed in the phylogenetic placement of enigmatic early fossils known as eophytes, which were treated either as stem-embryophytes or as stem-tracheophytes.

Because there is little consensus about which ancestral state reconstruction methods are most accurate, the researchers deliberately ran four competing approaches: maximum parsimony, a maximum likelihood equal-rates model, a maximum likelihood all-rates-different model, and Bayesian Markov chain Monte Carlo-based stochastic character mapping. The team used the equal-rates model as a benchmark, noting that a recent simulation study found it to be the most accurate in the greatest number of simulations, while the all-rates-different model frequently overestimated transition rates between character states.

The central result was remarkably consistent. Across all analytical models and all three phylogenetic arrangements, the ancestral embryophyte was inferred to have released monad spores derived from tetrads, which were unenveloped and bore a trilete aperture, with a sporoderm consisting of three layers: an outer exospore, an inner exospore of granular or amorphous sporopollenin, and an endospore. Sporopollenin deposited on white-line-centred lamellae was recovered as the ancestral mode of exospore formation. The ancestral sporophyte was inferred to have been homosporous, to have undergone monoplastidic meiosis, and to have possessed a parietal or secretory tapetum that contributed to exospore formation.

This finding overturns a traditional interpretation. Trilete spores, which develop in tetrahedral tetrads and are marked by a distinctive Y-shaped scar, have long been considered a defining feature of vascular plants. The new analysis instead infers that trilete spores were produced by the crown-embryophyte, the ancestor of all living land plants. As a character of total-group Embryophyta, the trilete spore must therefore have evolved among stem-embryophytes, and fossil cryptospores, the permanently fused tetrads and dyads that dominate the earliest spore record, cannot belong to the crown group. Because cryptospores have previously been used to constrain divergence time estimates within the embryophyte clade, the authors argue that this calibration approach requires reappraisal.

On that basis, the team defined new fossil calibrations. The divergence of Zygnematophyceae and Embryophyta is revised to a minimum of 449.65 million years ago, based on the earliest unequivocal trilete spore, Ambitisporites avitus, from the Qusaiba-1 core in the Qasim Formation of northern Saudi Arabia. The divergence of Bryophyta and Tracheophyta is constrained to between 459.1 and 426.9 million years ago, with the maximum based on the oldest possible trilete spores from the Dalby Limestone of central Sweden, which the authors did not use as a minimum constraint because they were concerned the trilete mark in the figured specimens might be a compression artefact, and the minimum based on Cooksonia cf. pertoni from the Homerian Cloncannon Formation of County Tipperary, Ireland. The divergence of Lycopodiophyta and Euphyllophyta is constrained to 431.2 to 420.7 million years ago using the same Cooksonia record and the earliest secure record of crown-Tracheophyta, Zosterophyllum sp. from Ludlow-age beds on Bathurst Island in Arctic Canada.

A relaxed molecular clock analysis using these revised calibrations produced a markedly younger and more precise timescale. Crown-embryophytes are estimated to have originated within a narrow 7-million-year interval in the Late Ordovician, between 457 and 450 million years ago, while both crown-tracheophytes, at 432 to 428 million years ago, and crown-bryophytes, at 445 to 429 million years ago, diverged in the early Silurian. The divergence estimates for the euphyllophyte, lycophyte, setaphyte, liverwort and moss crown clades were also younger and more precise than in previous analyses. Crucially, the new dates bring molecular clock estimates into closer accord with the fossil record itself. Earlier estimates of a Cambro-Ordovician origin implied a gap of up to 80 million years between the genetic divergence of the lineage and the first appearance of unequivocal land plants as fossils; under the new timescale, that mismatch shrinks to as little as 19 million years. The divergence of crown-tracheophytes is now estimated to have been coincident with the earliest known specimens of Cooksonia, a stem-tracheophyte.

The revised timescale also has consequences for models of the early land plants’ impact on the Earth system. The initial diversification of land plants has been implicated in Ordovician climate shifts: modelling of the weathering potential of the moss Physcomitrium patens suggested that diversifying non-vascular plants could have amplified silicate weathering enough to draw atmospheric carbon dioxide down to levels consistent with the Late Ordovician glaciations associated with a marine mass extinction. Land plant diversification has likewise been invoked in the rise of Palaeozoic atmospheric oxygen to modern levels by 400 million years ago through increased global carbon burial. But those models assumed a diversification of non-vascular land plants from around 470 million years ago, based on the diversity of regular tetrad cryptospores. The new estimate, placing crown-embryophyte diversification at 457 to 450 million years ago and the appearance of crown-tracheophytes at 428 to 432 million years ago, is discordant with models that assumed crown-tracheophytes diversified from the first appearance of unequivocal trilete spores around 450 million years ago, and the authors suggest the modelling assumptions may require reassessment.

The authors are careful to flag the uncertainties that remain. The phylogenetic position of the eophytes, early fossils such as Partitatheca and Lenticulatheca, has yet to be formally tested in a cladistic analysis, and some character histories differed between models; for example, the all-rates-different model inferred an enveloped ancestral spore, a result the authors attribute to the poorly evidenced placement of Partitatheca, an enveloped-dyad-producing eophyte, close to the embryophyte node. Not every detail of spore evolution was uniform across analyses either: mosses were recovered as having independently gained persistent spore mother cell walls and a perispore, liverworts lost the aperture independently, and cellular elaters were gained along the stem-lineage, while polyplastidic meiosis and an amoeboid tapetum were estimated as ancestral to euphyllophytes. The team acknowledges that this will not be the final attempt to estimate the timescale of early land plant evolution, since future studies will explore different assumptions and interpret the underlying data in different ways. But by integrating a formal interpretation of the spore fossil record, which dominates the first tens of millions of years of land plant fossil history, the study offers a recalibrated timeline in which land plants emerged in the Late Ordovician and vascular plants followed in the early Silurian, with further work needed to fully understand what that later arrival meant for the evolution of the Earth system.

Subject of Research: The evolution of early land plant spores and its use in recalibrating the molecular timescale of land plant diversification

Article Title: Reconstructing spore evolution recalibrates the timescale of land plant diversification

Article References: Hooper, J., Worth, A. M., Wellman, C. H., Clark, J. W., & Donoghue, P. C. J. (2026). Reconstructing spore evolution recalibrates the timescale of land plant diversification. Nature Plants. https://doi.org/10.1038/s41477-026-02432-3

Image Credits: AI Generated

DOI: 10.1038/s41477-026-02432-3

Keywords: land plants, spores, molecular clock, Ordovician, Silurian, embryophytes, trilete spores, cryptospores, fossil calibrations, ancestral state reconstruction, bryophytes, tracheophytes