Visual representation of a network of co-occurrence and the corresponding network of biotic interactions
A new study led by researchers at Miguel B. Araújo Lab has pioneered a method for interpreting the complex network of species interactions within ecosystems. Their findings, published today in Nature Ecology and Evolution, offer a significant breakthrough in ecological research, showcasing a novel way of inferring the relationships between organisms based on how often they occur together in nature.
“The study meticulously analysed ten different bipartite networks of actual biotic interactions and compared them with inferred interaction networks obtained by examining patterns of species co-occurrences”, explains Nuria Galiana, the first author of the study who is a Marie Curie Research fellow at the National Museum of Natural Sciences, CSIC. “We discovered that only about 20% of species co-occurrences hint at true interactions. A key finding is the transformation of degree distributions from exponential patterns in simple co-occurrence networks to power laws in networks of biotic interactions. This transformation is attributed to the intricate combination of species’ biotic requirements and their environmental preferences, which can now be predicted through analysis of the frequency of their co-occurrences”.
“In the study, super-generalist species, which have a wide range of environmental tolerance and varied diets, were identified as central architects in the structure of ecological communities. These adaptable species engage with many other species, facilitating a network structure that is both robust and complex”, concludes Nuria Galiana.
“The study goes beyond mere observational analysis by demonstrating how co-occurrence frequency can be used to precisely predict fundamental aspects of the actual interaction network. This represents a leap forward in our ability to understand how ecological communities assemble and function. The implications for conservation are profound, suggesting that large-scale strategies could be designed more effectively by leveraging this new understanding” proposes Miguel Araújo.
However, the authors also point for the importance of spatial scale in such analyses, indicating that broader scales might necessitate more careful adjustments to co-occurrence data to align it with interaction networks.
While previous studies have often relied on expert opinion or traits to infer interactions due to the difficulty of direct measurement, the principles uncovered in this research promise to refine these methods, making accurate predictions more accessible and reducing the need for extensive interaction data. This innovative framework, tested on bipartite networks, opens future research avenues to apply these insights to more complex network types, like food webs, enhancing our grasp of the intricate web of life and informing more nuanced conservation efforts.