The study of biodiversity in the context of climate change requires an interdisciplinary effort. The islands are biodiversity hot spots, in many cases whose architecture of interactions is unknown, and they represent ideal natural laboratories for the study of ecological complexity due to their limited geography, Indeed, their study has revealed fundamental concepts. Understanding the fragility or robustness of island ecosystems in the face of anthropogenic activities is crucial for their conservation. At the same time, island ecological complexity contains ingredients, such as the distribution of biodiversity, or the variability of ecological interactions, that are attractive and suitable to be analyzed from the perspective of complex systems. Based on the effort of building insular food webs of several islands and with different degree of anthropogenic disturbance, the objective of the subproject is the development of a theoretical framework for the study of the emergence and sustainability of insular ecosystems and their diversity, as well as their response to disturbances due to climate change. Representing the complexity of ecological architecture requires the use of multilayer networks, simplicial complexes and/or hypergraphs. Multilayer networks allow the representation of weighted and different types of interactions, while complex simplistic and hypergraphs are useful to associate more than two aspects with each other, which is the case of the data that will be collected in the associated subproject. To examine the emergence of island ecosystems, we will integrate various levels of biological organization involved in the dynamics of interactions between individuals and species. In particular, we will focus on the modelling of the possible dynamics involving interactions between genes, traits, individuals and species. Models based on the integration of biological processes will allow us to explore the emergence of networks of interaction between species and their possible temporal evolution. The results of the models will be compared with the data obtained in different islands, as well as in their degree of affectation by anthropogenic activity. Finally, we will study scenarios that simulate disturbances due to climate change, both punctual disruptive events (extreme weather events) and constant ones in time (e.g., temperature increase, invasive species). The project also contemplates the task of automating data collection based on image analysis with automated learning techniques. This technology will allow the generation of a volume of data with ecological value much higher than the current state. The complementarity between the two subprojects, the interdisciplinary and multi-disciplinary nature of the groups that make up the project, the collaboration of highly prestigious international groups offer an opportunity to create synergies and train a new generation of researchers with interdisciplinary training.