Doctoral defence: Madli Jõks“Biodiversity drivers in oceanic archipelagos and habitat fragments, explored by agent-based simulation models”

On 26 May at 10:15 Madli Jõks will defend her doctoral thesis “Biodiversity drivers in oceanic archipelagos and habitat fragments, explored by agent-based simulation models” for obtaining the degree of Doctor of Philosophy (in Botany and Mycology).

Supervisor:
Prof. Meelis Pärtel, University of Tartu

Opponent:
Dr. Juliano Sarmento Cabral, University of Birmingham (United Kingdom)

Summary
Oceanic islands and habitat fragments within landscapes are both valuable ecological study objects, characterized by clear physical boundaries, immigration and extinction dynamics of well-defined populations, and high conservational priority. However, despite the apparent similarity, there is a crucial difference between these systems. Unlike real islands, isolation of habitat patches is not solely affected by their spatial arrangement, but also by organisms’ response to the surrounding landscape, or the ‘functional connectivity’. In this thesis, I used agent-based simulation models to explore the biodiversity drivers in oceanic archipelagos (Hawaii, Galápagos, Canary Islands, Cape Verde, Azores) and Estonian old-forest fragments, that are both characterized by spatial and functional isolation. I found that in oceanic archipelagos, modern habitat diversity and archipelago configuration, as well as geologic and eustatic history have affected biodiversity development. All studied archipelagos demonstrated the importance of habitat diversity, while the effect of modern archipelago configuration was significant only in those where archipelago history was strongly reflected in the present setting, with geologic and eustatic history being the more important determinants elsewhere. In Estonian forest study, I found that secondary forests on former agricultural land are not functionally connecting old-forest fragments in terms of ground layer vegetation, epiphytes, and several groups of soil fungi. Furthermore, I found that their ability to exploit younger forests is primarily limited by the availability of microhabitats. In conclusion, agent-based models allow us to gain crucial mechanistic knowledge about the biodiversity drivers in oceanic archipelagos and habitat fragments and to predict and avoid biodiversity losses in these isolated and valuable systems.