Theoretical and data-driven models in Ecology

Life on Earth has evolved over billions of years, resulting in a rich diversity
of species and ecosystems that provide essential services for human survival and
well-being. However, this biodiversity is rapidly declining due to human
activities such as habitat destruction, climate change, invasive species and emerging
diseases. These interconnected drivers are causing widespread loss of species
and degradation of ecosystems, threatening global ecological stability and
human prosperity. Addressing this crisis requires an interdisciplinary approach to
understand and mitigate its impacts, ensuring the preservation of biodiversity
and the sustainability of human societies.

In this thesis we develop theoretical and data-driven methods to address
pressing issues in Ecology and Conservation Biology through the lens of Complex
Systems and interdisciplinary research. We address a range of contemporary
challenges related to biodiversity loss driven by climate change and emerging
diseases. These challenges include the spread of diseases, ocean acidification,
and the decline of critical ecosystems such as coral reefs and seagrass meadows.
We rely on a combination of theoretical models, computational simulations,
and advanced data analysis techniques to gain a deeper understanding of these
complex ecological phenomena.

In the first two parts of this thesis we develop mathematical models of
disease spread to fill knowledge gaps in the transmission dynamics of marine
and vector-borne plant diseases. We focus on two case studies: the Mass
Mortality Event (MME) of Pinna nobilis and the vector-borne plant diseases
caused by the bacterium Xylella fastidiosa. We investigate the role of key factors
such as temperature or pathogen mobility in the transmission of the MME and
the impact of the non-periodic seasonal abundance of insect vectors on the
spread of plant diseases. These models provide insights into the mechanisms
driving the dynamics of these diseases and the potential for their control and
management.

In the third part, we apply this knowledge to develop a novel theoretical
framework to predict the potential distribution of vector-borne plant diseases
based on environmental and climatic factors. We demonstrate the utility of this
model by predicting the risk of Pierce’s disease of grapevines, caused by Xylella
fastidiosa, under current and future climate scenarios. Our methodology
represents a significant advancement in the field of disease biogeography, providing
a way to integrate the inherent complex ecological interactions of diseases to
predict their potential establishment based on environmental conditions.

Finally, in the fourth part of this thesis, we develop and apply data-driven
methods to monitor and assess the health of coastal marine ecosystems. We
present an innovative framework to reconstruct missing data from ocean pH
time-series using deep learning techniques, which enhance our ability to
monitor ocean acidification accurately. Additionally, we employ machine learning and
satellite imagery to map and evaluate the condition of seagrass meadows, offer-
ing a scalable and cost-effective approach to ecosystem monitoring. Moreover,
we conduct a global analysis of the spatial properties of coral reefs using remote
sensing data, uncovering universal patterns in reef size distribution and
geometry. These insights are crucial for developing targeted conservation strategies
to protect these vulnerable ecosystems.

This thesis underscores the importance of interdisciplinary research,
integrating ecological theory, complex systems’ science, and artificial intelligence
to tackle ecological challenges. The findings contribute to the development
of effective conservation strategies, aiming to mitigate the impacts of climate
change and emergent diseases on biodiversity. Ultimately, this work supports
efforts to preserve the integrity of ecosystems and ensure the sustainability of
human societies in the face of ongoing environmental changes.