In this study, we develop innovative methods for quantifying fish production enhancement on restored shellfish reefs using automated underwater video analysis and computer vision technology. Our research demonstrates that restored oyster reefs significantly increase local fish production, with automated video analysis providing accurate, cost-effective monitoring compared to traditional sampling methods. By combining ecological restoration assessment with cutting-edge machine learning techniques, we establish new standards for measuring restoration success and provide scalable tools for adaptive management of marine habitat restoration projects.

In this comprehensive global study, we examine how climate effects on belowground tea litter decomposition vary depending on ecosystem type and organic matter composition across diverse wetland environments worldwide. Using standardized decomposition experiments across 180+ sites spanning multiple continents, we reveal that climate sensitivity of decomposition processes is strongly mediated by local ecosystem characteristics and substrate quality. Our findings demonstrate that wetland type and organic matter chemistry create distinct climate-decomposition relationships, challenging simple global models of carbon cycling.

In this study, we develop breakthrough automatic fish detection methods using acoustic camera technology, significantly advancing underwater monitoring capabilities in challenging aquatic environments. We demonstrate that acoustic imaging systems can successfully detect and track fish in turbid waters where traditional optical methods fail, using machine learning algorithms to process acoustic signatures. Our innovative approach provides reliable fish detection with 85% accuracy across diverse species and environmental conditions, offering new possibilities for non-invasive monitoring in rivers, estuaries, and coastal waters where visibility is limited.

In this research, we successfully adapt automated fish detection and classification frameworks for deployment on remotely operated vehicles (ROVs) operating in shallow marine environments. We address key challenges in mobile underwater surveys, including variable lighting conditions, camera movement, and real-time processing constraints. By integrating advanced computer vision algorithms with ROV technology, we demonstrate improved accuracy and efficiency in fish monitoring compared to traditional static camera systems, providing a scalable solution for comprehensive marine biodiversity assessments in dynamic underwater environments.

We developed an automated fish counting system for baited underwater video stations (BRUVS) that achieved 88% accuracy in counting Australasian snapper across varying densities. Our deep learning approach successfully addressed the challenge of fish obscuration in high-density scenarios, providing a cost-effective alternative to manual video analysis. This automated method offers substantial potential for improving the efficiency and statistical rigor of fishery-independent monitoring programs used in stock assessment.

In this research, we examine how different types of disturbances interact with habitat connectivity to influence ecosystem resilience across marine environments. Using experimental and observational approaches, we demonstrate that the relationship between connectivity and resilience is highly dependent on disturbance type, with pulse disturbances (e.g., storms) showing different connectivity-resilience relationships compared to press disturbances (e.g., chronic sedimentation). Our findings reveal that well-connected habitats can either enhance or reduce resilience depending on whether connectivity facilitates recovery or spreads disturbance effects, providing critical insights for adaptive ecosystem management under changing environmental conditions.

In this continental-scale study, we examine how ecosystem type drives variation in tea litter decomposition rates and associated prokaryotic microbiome communities across diverse freshwater and coastal wetland environments. Using standardized tea bag methodology across multiple sites, we demonstrate that ecosystem characteristics (salinity, temperature, oxygen levels) fundamentally shape both decomposition processes and microbial community structure. Our findings reveal distinct microbial signatures associated with different wetland types, providing insights into biogeochemical cycling and carbon processing in wetland ecosystems.

We examined the differential importance of deep versus shallow seagrass habitats in supporting nekton assemblages within the Great Barrier Reef ecosystem. Using comprehensive field surveys, we reveal distinct community structures and species compositions between depth zones, with shallow seagrass beds supporting higher abundance and diversity of mobile fauna. Our findings demonstrate that depth gradients create unique ecological niches within seagrass meadows, highlighting the importance of protecting both shallow and deep seagrass habitats to maintain the full spectrum of marine biodiversity and ecosystem functioning in tropical marine environments.

We examined the fundamental role of saltmarsh grass ecosystems in supporting fishery food webs across subtropical Australian estuarine environments. Using stable isotope analysis and food web modeling, we demonstrate that saltmarsh vegetation provides critical basal production that supports commercially important fish species through complex trophic pathways. Our findings reveal that saltmarsh-derived organic matter forms the foundation of estuarine productivity, with direct implications for fisheries management and coastal habitat conservation.

In this research, we demonstrate how habitat connectivity enhances herbivory patterns and ecosystem functioning in disturbed marine environments, revealing the adaptive benefits of well-connected ecological networks. Using laboratory experiments across connectivity gradients, we show that well-connected habitats maintain higher and more consistent herbivory rates even under disturbance conditions, leading to improved primary productivity and ecosystem stability. Our findings provide compelling evidence that connectivity acts as an insurance mechanism against disturbance impacts, enabling ecosystems to maintain critical ecological processes through enhanced species movement and resource flows between habitat patches.

We investigated how habitat structural complexity shapes food web architecture and trophic interactions in Great Barrier Reef seagrass meadows. Using stable isotope analysis and size spectra analyses, we demonstrate that increased habitat complexity supports more diverse and interconnected food webs, with complex habitats facilitating higher trophic diversity and more specialized feeding relationships. Our findings reveal that structural complexity acts as a fundamental driver of ecosystem function in seagrass systems, influencing energy flow pathways and community stability, with important implications for seagrass conservation and restoration strategies.

In this study, we examine functional changes in reef systems under ocean warming conditions, focusing on the asymmetrical effects of altered grazing behavior by crustacean mesograzers on primary producer communities. Using controlled temperature experiments, we demonstrate that warming disproportionately affects grazing intensity and selectivity, leading to shifts in algal community composition and reef function. Our findings reveal that climate change impacts on key herbivore species can cascade through reef food webs, fundamentally altering ecosystem processes and highlighting the vulnerability of reef systems to warming-induced changes in species interactions.

In this research, we investigate how herbivory patterns influence the delivery of critical ecosystem services in tropical seagrass ecosystems, including carbon sequestration, coastal protection, and fisheries support. We demonstrate that moderate grazing pressure enhances ecosystem service delivery by maintaining seagrass productivity and diversity, while excessive herbivory can compromise service provision. Our findings reveal complex trade-offs between different services under varying grazing regimes, providing evidence-based guidance for managing herbivore populations to optimize the multiple benefits that seagrass ecosystems provide to coastal communities and global climate regulation.