Terrestrialisation, i.e. the establishment of complex continental ecosystems, stands as one of the most significant evolutionary events in the history of life on Earth. It started in the Ordovician (ca. 460 Myr) with the invasion of land by plants and continued into the Devonian (ca. 400 Myr) with the rise of several new groups of animals such as continental arthropods and tetrapods (four-legged vertebrates). The origin of tetrapods predates the early Middle Devonian (395 Ma) but their complete terrestrialisation is only recorded by the Carboniferous. As a result, the Late Devonian represents the ideal period to understand processes related to tetrapod terrestrialisation.
The taxonomic richness, the phylogenetic relationships, and the palaeoenvironmental settings of Devonian tetrapods are starting to be better understood. On the contrary, the ecological dimension of their transition to land remains elusive. A plethora of reasons for why tetrapods left the water has been proposed, e.g. lowered oxygen concentration in the water, drying environment, and increased competition for food or space in the water linked to a greater availability of food on land. Among all these reasons, the large amount of predation or aggression in the water has been proposed too. Tetrapods obviously did not evolve in an ecological vacuum and faunal interactions (i.e. their trophic positions, their possible competitors) might have been critical drivers for their later invasion of land. Thus, an overall understanding of the terrestrialisation process of tetrapods requires robust inferences of their ecological roles within the Late Devonian aquatic vertebrate faunas. This will help, in turn, to decipher the factors driving the tetrapod terrestrialisation process, notably testing for the amount of predation pressure applied upon them and the extent of their ecological niches in the water.
The EU-funded ‘Trophic Networks at the dawn of tetrapod Terrestrialisation’ (TNT) project proposed to tackle these major issues by investigating the trophic relationships within eight key Devonian (Famennian) vertebrate localities worldwide containing early tetrapods together with a well-known vertebrate assemblage. This project is inherently interdisciplinary, blending calcium isotope analyses with quantitative analyses of ecomorphological characters to obtain reliable inferences on the ecology of Late Devonian vertebrates within different palaeoenvironments and palaeoclimates. Because Recent trophic networks provide a comparative basis for understanding Past ecosystems, two Recent localities with fossil analogues (giant salamander and lungfish as analogues of Devonian tetrapods and lungfish, respectively) were planned to be studied in the scope of the TNT project in order to measure the yet understudied calcium isotope variability of freshwater organisms.
In summary, the main objectives of the project were:
1) To validate the use of calcium isotopes for reconstructing trophic networks of modern freshwater ecosystems. Two localities were targeted: the Edward Lake in Uganda with Recent lungfish analogue of Devonian lungfish (Protopterus aethiopicus) and the Biwa Lake in Japan with Recent tetrapod analogue of Devonian tetrapod (Andrias japonicus).
2) To reconstruct the trophic networks of Late Devonian vertebrates in tetrapod-bearing localities (marine and freshwater) thanks to calcium isotope geochemistry and ecomorphological analyses.
3) To test the hypothesis that predation pressure drove tetrapod terrestrialisation or not.
For this project, bone and tooth tissues of 160 ‘fishes’, but also mollusc shells, and water samples from the Lake Edward were analysed for their ratio in calcium isotopes. The results demonstrated that the use of calcium isotopes for reconstructing trophic networks of Recent freshwater ecosystems works perfectly and that there are clear clusters that appear according to the ecological/trophic niches. These results gave the necessary framework to use such method in freshwater fossil localities and to test the hypothesis that predation pressure drove tetrapod terrestrialisation or not. One Late Devonian locality (Strud in Belgium) was tested in this framework. The study showed that no trophic signal can be highlighted, because bone samples are prone to extensive diagenesis that hide the calcium isotope signature. This finding highlights the fact that fossil tissues more resistant to the diagenesis i.e. enamel, must be used in the context of calcium geochemical analyses in the deep-time fossil record. Also, as diagenetic processes, vary from one fossil locality to another, it is necessary to carry on these analyses on various Devonian tetrapod-bearing localities to understand the terrestrialisation process.
