Until now, it was believed that plants of the grape family arrived at the European continent less than 23 million years ago. A study on fossil plants draws a new scenario on the dispersal of the ancestors of grape plants and reveals that these species were already on the territory of Europe some 41 million years ago.

The paper describes a new fossil species of the same family, Nekemias mucronata, which allows us to better understand the evolutionary history of this plant group, which inhabited Europe between 40 and 23 million years ago.

This study, published in the Journal of Systematics and Evolution, is led by researcher Aixa Tosal, from the Faculty of Earth Sciences and the Biodiversity Research Institute (IRBio) of the University of Barcelona. The article is also signed by Alba Vicente, from the Biodiversity Research Institute (IRBio) and the Catalan Institute of Palaeontology Miquel Crusafont (ICP), and Thomas Denk, from the Swedish Museum of Natural History (Stockholm).

A new ancestor of the grape family

The grape family (Vitaceae) is made up of some 950 species, and is divided into five tribes (in botany, this is an intermediate taxonomic classification between the family and the genus). One of these tribes is the Viteae, made up of 200 species, including the grape vine plant (Vitis vinifera), which is of great global economic interest. The new paper published in the JSE focuses on studying the tribe Ampelopsideae, made up of 47 species.

“Our study changes the paradigms accepted until now and shows that the Ampelopsis and Nekemias lineages of the Ampelopsideae tribe were already present in Europe and Central Asia during the middle Eocene (between 47 and 37 million years ago). This indicates that this dispersal was approximately 20 million years earlier than previously estimated,” says Tosal, first author of the study and member of the UB’s Department of Earth and Ocean Dynamics.

“In particular, we show that a lineage now restricted to North America already existed in Europe and Central Asia, thanks to the discovery of the fossil species Nekemias mucronata, which is very similar to the present-day North American Nekemias arborea. Nekemias mucronata cohabited with Ampelopsis hibschii, the closest relative of today’s Ampelopsis orientalis,” explains Tosal.

In contrast, the latter has had a different dispersal from N. mucronata, as this lineage is now endemic to the eastern Mediterranean.

“This study helps us to better understand the evolution of the Ampelopsideae tribe during the second dispersal pulse, especially in Europe and Central Asia, which took place during the Palaeogene,” says Tosal.

Nekemias mucronata lived from the late Eocene to the late Oligocene (37–23 million years ago). It seems that it was able to grow in a broad range of climates, from regions with low winter temperatures (-4.6 °C in cold periods)—such as those found in Kazakhstan during the Oligocene (33–23) million years ago—to regions with warm mean annual temperatures—such as those of the Oligocene in the Iberian Peninsula—or even in climates with intermediate temperatures such as those recorded in the center of the European continent.

“N. mucronata was also not overly demanding in terms of rainfall. It could grow in areas with abundant rainfall and low rainfall seasonality; for example, in Central Europe during the Oligocene, or the Iberian Peninsula or Greece during the same time,” says ICP researcher Vicente.

“This fossil species had a compound leaf, a peculiarity shared with some species of the vine family. Although it is difficult to confirm the number of leaflets of the compound leaf, it would have consisted of at least three. We have been able to recognize common patterns between the apical and lateral leaflets, which allows us to distinguish them from other fossil species of the vine family in Eurasia,” he adds.

“What makes Nekemias mucronata unique is the presence of a mucro at the tip of the leaflet teeth, which gives the species its name. The straight shape of the base of the apical leaflet is also quite distinctive, as all other Eurasian fossil species are buckled (with an invagination near the petiole),” says Vicente.

Dispersal of Ampelopsideae across the Atlantic Bridge or the Bering Strait

To date, the oldest record of the grape family has been found in the Upper Cretaceous deposits of India (75–65 Ma). The earliest record of the plant lineage in the Americas is from the Upper Eocene, around 39.4 million years ago, and at about the same time in Europe and Central Asia the Ampelopsis and Nekemias lineages are already found.

How did these species disperse in the past? These tribes diverged between the Upper Cretaceous and the Upper Eocene and, although there are still many unknowns, it seems that they dispersed and evolved quite rapidly.

According to current data, which are consistent with the molecular clock technique, “the Ampelopsideae could have followed two cluster routes or a mixture of both. The first proposed route follows the North Atlantic isthmus. That is, the family appeared in India, then moved on to central Asia and Europe during the middle Eocene (between 47 and 37 million years ago), and finally moved on to the Americas via Greenland,” says Denk.

“Another possible route suggests that, once the Vitaceae family appeared in India, the Ampelopsideae tribe dispersed eastward from Asia during the middle Eocene (47–37 million years ago) and quickly moved to the Americas via the Bering Strait, and from there to Europe along the North Atlantic isthmus,” Denk says.

Although the dispersal of these two species does not seem to be linked to climate, it is possible that the increase in aridity during the Oligocene in the Iberian Peninsula and southern Europe explains the extinction (27–23 million years ago) of the last population of N. mucronata found in the Iberian Peninsula. In parallel, Ampelopsis hibschii was restricted to the Balkan area and finally became extinct about 15 million years ago.

“However, there are still many unanswered questions about the early dispersal phases (from the Late Cretaceous to the Palaeogene). For this reason, we would like to continue studying this family, and perhaps we will be able to unravel what happened during their early cluster phases, which occurred between 66 and 41 million years ago,” the team concludes.

As computing technology advances, we have shifted from using large, single-chip processors to systems made up of smaller, specialized chips called “chiplets.” These chiplets work together to boost processing power and efficiency.

This transition is crucial because we’ve reached the physical limits of how many transistors can fit on a single chip. As transistors shrink, problems such as overheating and power inefficiency become more severe.[¹] Using multiple chiplets in one system can increase computing power without facing these physical constraints.

The challenge of communication between chiplets

Traditionally, communication within a chip has been managed by a system called Network-on-Chip (NoC), which acts like a data highway. This method becomes inefficient as systems get more complex, especially with multiple chiplets. Data has to travel farther across more grid points, slowing communication and increasing energy consumption.

When we scale this approach to various chiplets, we create what’s known as Network-in-Package (NiP). However, the same issues—delays, energy inefficiency, and limited scalability—still exist as wired connections dominate data transfer.

To solve these problems, researchers are exploring wireless communication at the chip level. Instead of relying on wires, chiplets could communicate wirelessly using tiny antennas.

Terahertz (THz) frequencies, electromagnetic waves between infrared and microwave, offer high-speed data transfer, making them ideal for this application. However, THz signals are highly noise-sensitive, disrupting communication and making it harder to decode transmitted data.

Floquet engineering: Improving signal detection

Our research addresses this issue with Floquet engineering, a technique from quantum physics that helps control electron behavior in a material when exposed to high-frequency signals.[^2,3,4] This technique makes the system more responsive to certain frequencies, improving the detection and decoding of THz wireless signals, even in noisy conditions.

We applied this method to a two-dimensional semiconductor quantum well (2DSQW)—a very thin layer of semiconductor material that restricts electron movement to two dimensions. This setup enhances the system’s ability to detect THz signals, even when noise interference is high. Our research is published in the IEEE Journal on Selected Areas in Communications.

Dual-signaling architecture for more accurate communication

To further improve noise handling, we developed a dual-signaling architecture, where two receivers work together to monitor signals. This setup allows the system to adjust a key parameter, called reference voltage, based on the noise levels detected. This real-time adjustment significantly improves signal decoding accuracy.

Our simulations showed that this dual-signaling system reduces error rates compared to traditional single-receiver systems, ensuring reliable communication in noisy environments—a critical requirement for chip-scale wireless communication.

By overcoming the challenges of noise and signal degradation, our dual-signaling technique marks a key advancement in developing high-speed, noise-resistant wireless communication for chiplets. This innovation brings us closer to creating more efficient, scalable, and adaptable computing systems for the technologies of tomorrow.

This story is part of Science X Dialog, where researchers can report findings from their published research articles. Visit this page for information about Science X Dialog and how to participate.