Ancient Holly-Like Pollen from Brazil Rewrites the Story of an Iconic Plant Lineage

Deep in the sedimentary rocks of the Araripe Basin in northeastern Brazil, a region already world-famous for its exquisitely preserved Cretaceous fossils, palynologists have uncovered something far stranger than another ancient fish or insect. Embedded in mudstones of the Rio da Batateira Formation, deposited roughly 118 to 113 million years ago during the Aptian stage, are tiny fossil pollen grains assigned to the genus Gemmatricolpites. Under the scanning electron microscope these grains reveal a wall architecture so strikingly similar to that of modern holly, the genus Ilex, that researchers are now asking an electrifying question: did an ancient, now-extinct lineage of the holly family flourish in the Cretaceous tropics of North Gondwana, or are we looking at one of the most convincing cases of convergent evolution in the plant fossil record?

The stakes in this debate are considerable. Ilex, with approximately 600 species of evergreen and deciduous trees and shrubs, is one of the largest woody genera among the dioecious flowering plants and the sole genus of the family Aquifoliaceae. Today its members thrive in humid tropical to warm temperate habitats, with centres of diversity in East Asia and South America. Yet the evolutionary origin of this familiar genus has long been shrouded in uncertainty. Fossils assigned to Ilex are frequently considered ambiguous, and their occurrence in the rock record often predates the ages inferred from molecular clock studies of the holly crown group. The new analyses, published in the journal Fossil Record, combine light microscopy, scanning electron microscopy and transmission electron microscopy to place the Cretaceous pollen under an unprecedented level of scrutiny.

What the researchers found in the Brazilian material is remarkable in its detail. The Gemmatricolpites grains are small, with polar axes measuring only 11.8 to 16.2 micrometres, and are prolate to spheroidal in shape. Their apertures are tricolpate, meaning three elongate colpi run parallel to the polar axis, or in about a quarter of the examined grains, tricryptoaperturate, in which the apertural regions are inconspicuous and marked only by smaller sculptural elements on the foot layer. The ektexine, the outer wall layer, is atectate and clavate to baculate, studded with mushroom-shaped clavae and rod-like baculae of varying sizes that decrease towards the apertures. Crucially, scanning electron microscopy reveals that the tops of these clavae and baculae carry an irregular microrugulate ornamentation, a pattern of ridges and grooves that is invisible under light microscopy alone.

This combination of features is not merely similar to modern holly pollen; it is essentially the holly pollen signature. Among all flowering plants, the atectate, conspicuously clavate ektexine of Ilex, often combined with baculae and gemmae, is unique and instantly recognizable. Within the order Aquifoliales, which also includes the families Helwingiaceae, Phyllonomaceae, Cardiopteridaceae and Stemonuraceae, only Ilex displays this distinctive wall sculpture. Transmission electron microscopy of the fossil grains showed a thick ektexine comprising a semitectum, a columellate infratectum and a remarkably thick continuous foot layer, together with a bilayered endexine that thickens towards the apertures. Ultrastructural studies of extant Ilex paraguariensis, the mate plant, reveal a pollen wall of strikingly similar construction, with all the same layers present in the same arrangement, albeit thinner in the living species.

The differences, however, are equally telling. Modern Ilex pollen is tricolporate, possessing both elongate colpi in the ektexine and circular pores in the endexine, whereas the Brazilian fossils are purely tricolpate or tricryptoaperturate. The fossil grains are also roughly half the size of the Eocene to Miocene Ilex-like pollen that the team examined from localities across Europe and South China, including the Paleocene-Eocene Thermal Maximum deposits of St. Pankraz in Austria, the Early Eocene Climate Optimum clays of Krappfeld, the middle Eocene Changchang Formation on Hainan Island, and Oligocene to Miocene sites in Germany and Austria. The authors suggest that this doubling in size through time might reflect a whole genome duplication event in the Aquifoliaceae during the Late Cretaceous, a hypothesis consistent with molecular evidence for genome history in the family, though they caution that the relationship between pollen size and genome size is not unequivocal.

Apertural variation through an evolutionary line might sound like a fatal objection, but the researchers marshal compelling evidence that it is not. Several plant genera are known to display variable apertural conditions even within a single species. In the family Krameriaceae, for example, pollen of the genus Krameria ranges from acolpate with endoapertures through cryptoaperturate, colporate, colpate to sulcate conditions, and these polymorphous apertures have been interpreted as non-homologous structures. The Arecaceae, the palm family, display an equally bewildering array of apertural types, from the common monosulcate condition to disulcate, trichotomosulcate, zonosulcate, diporate, triporate and inaperturate grains, all of which are also found in Eocene strata. Within Aquifoliales themselves, the occasional tricryptoaperturate condition persists today in the genera Gonocaryum and Cardiopteris. Apertural condition, in other words, may be a less reliable taxonomic anchor than the exine sculpture and ultrastructure, which show remarkable morphological stasis across more than 100 million years.

Tracing the fossil occurrences of Gemmatricolpites and related forms reveals a fascinating biogeographic pattern. The oldest records come from the Aptian of the Araripe Basin and, probably, from the Aptian of Gabon, where a comparable clavate tricolpate type has been reported. Albian Gemmatricolpites grains are known from the Basque-Cantabrian Basin of Spain, and by the Late Cretaceous the genus had achieved a nearly cosmopolitan distribution, with records from Minnesota, Delaware, Nigeria, Malaysia, Argentina, South Africa and Egypt. Tricolpate Ilexpollenites pollen and Ilex seeds appear in Upper Cretaceous strata of Antarctica, New Zealand, Australia, California and Germany. The authors interpret this pattern as witnessing the origin of an ancient Aquifoliales lineage in North Gondwana, around Brazil and Gabon during the Barremian to Aptian, followed by lateral migration out of the palaeotropics during the Cretaceous, ultimately supporting the hypothesis that the campanulids and their early diverging Aquifoliales clade are of Gondwanan origin.

The molecular evidence, intriguingly, is converging on the same story. Time-calibrated phylogenies of the asterids suggest an Early Cretaceous diversification, with a stem node age for Aquifoliales of approximately 121 million years and a crown node age around 113 million years, ages that sit comfortably within the Aptian assignment of the Rio da Batateira Formation. More recent phylogenetic studies of Aquifoliaceae incorporating greater numbers of Chinese Ilex species suggest a Late Cretaceous stem age of about 84.7 million years and a crown group age in the early Eocene at roughly 50.8 million years. The Eocene is precisely when the fossil record shows both ancient tricolpate and modern tricolporate Ilex-like pollen co-occurring in Europe and South China, before the tricolporate condition took over exclusively from the Oligocene onwards. The fossil and molecular timelines, once seen as irreconcilable, now appear to describe a coherent evolutionary narrative.

Not all Gemmatricolpites grains reported in the literature can be confidently affiliated with this ancient holly lineage, the researchers acknowledge. Many occurrences rest on poor light microscopy images, poorly preserved specimens or sculpture that bears no real resemblance to Aquifoliaceae. But the Brazilian material, investigated with the full arsenal of modern palynological technique, is different. The mixed pollen clumps found in some samples, containing Gemmatricolpites grains alongside unidentified monosulcate pollen, suggest the plants were visited regularly by pollinators or pollen-eating insects, perhaps leaving behind corbiculae or frass, and imply that the Gemmatricolpites producers were abundant, entomophilous and prolific pollen producers, just as hollies are today.

Whether the verdict ultimately lands on extinct lineage or convergent evolution, the study demonstrates how scanning and transmission electron microscopy can transform ambiguous palynological fossils into rich sources of evolutionary information. The unique atectate, clavate to baculate sculpture with microrugulate ornamentation, preserved with such fidelity in the Aptian rocks of the Araripe Basin, points towards a long-term morphological stasis of the pollen wall that spans the entire history of the holly family. If the interpretation is correct, the humid Cretaceous palaeotropics of North Gondwana nurtured an ancient lineage of Aquifoliales whose descendants, after a genome duplication and a global migration, gave rise to the hollies we know today, while their more archaic tricolpate relatives faded into extinction during the Eocene. The pollen grains, so small they were once missed entirely by sieved preparations, have turned out to be messengers from the deep roots of one of the plant kingdom’s most iconic genera.

Subject of Research: Gemmatricolpites R.L. Pierce pollen from the Lower Cretaceous of the Araripe Basin (NE Brazil) – an extinct lineage of Aquifoliaceae or a product of convergent evolution?

Article Title: Gemmatricolpites R.L. Pierce pollen from the Lower Cretaceous of the Araripe Basin (NE Brazil) – an extinct lineage of Aquifoliaceae or a product of convergent evolution?

Article References: Hofmann, C.-C., Ulrich, S., Coiro, M., Heimhofer, U., & Seyfullah, L. J. (2026). Gemmatricolpites R.L. Pierce pollen from the Lower Cretaceous of the Araripe Basin (NE Brazil) – an extinct lineage of Aquifoliaceae or a product of convergent evolution?. Fossil Record, 29(2), 451-469. https://doi.org/10.3897/fr.29.196612

Image Credits: AI Generated

DOI: 10.3897/fr.29.196612

Keywords: Gemmatricolpites, Pierce, pollen, Lower, Cretaceous, Araripe, Basin, Brazil, extinct, lineage, Aquifoliaceae, product