The interdependencies among organisms in ecological systems govern ecological function, responses to ecosystem change and the provision of ecosystem services. Complex multi-species interactions form ecological networks that increase community stability, where it is known that higher diversity promotes dynamic stability and counters pressures on the ecosystem. Understanding these ecological networks and dynamics in response to multiple chronic and acute environmental disturbances is key to designing conservation interventions, but network dynamics remain largely unknown in tropical marine systems. The Wallacea region is not only the epicenter of global coral reef diversity, it's coral reefs are also the most important source of food and income for millions of people. The region is globally important for sustaining globally threatened reef biodiversity, because it is one among few reefs that have not yet experienced large-scale degradation due to climate stress events. However, the regions' reefs are greatly transformed by local stressors, most notably terrestrial derived sediment run-off and fisheries overexploitation. Management to mitigate these local stressors is underway, but is severely hampered by a pervasive lack of data and understanding how these stressors affect biodiversity and its interactions throughout the region. Given the limited local resources, time, and expertise, presently the lack of information to inform management will not be overcome for most reefs across Wallacea. This project addresses three aims, which require addressing to overcome these challenges. Firstly, we will employ traditional visual observation techniques for reef corals, macro-invertebrates and fishes to determine how ecological networks differ between largely unimpacted, sedimented or reduced water quality, and fished sites. Ecological networks will be quantified based on field data using novel/cutting edge ecological modelling approaches and network theory. This aim focuses on easily observable non-cryptic taxa initially, and will form the basis of our comparative assessment of eDNA which has great potential but remains unvalidated in coral reef systems. Our second aim will validate eDNA's utility in describing reef biodiversity dynamics and test for the first time how networks constructed from eDNA compare to those from ecological surveys. Outcomes will provide unique insights and the potential for a step change in reef monitoring approaches by expanding opportunities to intensify spatial and temporal sampling, alongside providing confidence limits associated with ecological networks and their metrics that have been constructed using these novel eDNA data. Finally, having assessed the application of eDNA techniques, we plan to map the coral reef biodiversity of Wallacea with extensive eDNA sampling from remote regions. We will expand eDNA sampling to 10 further sites within Wallacea to provide novel data at unprecedented spatial and ecological complexity scales, enabling us to model and predict change in response to anthropogenic pressures at scales most relevant to Indonesian marine spatial planning. Despite high functional redundancy in the mega-diverse coral reef ecosystems, the loss of species will shift community composition and functionality with wide-ranging ecological, social and economic implications. For conservation efforts in complex systems where cryptic species are under-sampled but play important functional roles, there is a real-world need to know true biodiversity and network interactions. We will develop and test novel approaches to tackle this challenge to inform sustainable development in the face of the coral reef crisis. The response dynamics in ecological networks inform community robustness and this information is key to management decision making.

Unmanned Aerial Vehicles (UAVs) are a step changing technology that allows for fast, low-risk and low- cost environmental sensing with a world-wide market value of $89bn over the next ten years (as estimated by the 2013 Teal Group forecast). However, current UAVs are greatly limited by being able to operate in air only, preventing them to move effectively on the ground or in water. Hybrid air-water mobility would address the societal need for remote water sampling in inaccessible areas such as during floods or after nuclear accidents. For example, UAVs could quickly provide water samples following a nuclear accident to better coordinate response efforts. While some UAVs can land on water, no technologies are available that allow them to both dive and fly, due to dramatic design trade-offs that have to be solved for movement in both air and water, and due to the absence of high-power propulsion systems that would allow a transition from underwater to air. Expanding on my emerging leader position in aerial robotics, I will develop a novel generation of hybrid flying robots, called Aquatic Micro Aerial Vehicles (AquaMAVs). These robots offer the revolutionary mission capability to fly to a site of interest, dive into the water to take a water sample and provide video footage and retake flight to return to the base station.The engineering approach will build on the analysis of aerial-aquatic animals such as flying fish, diving birds and gliding squid, and the implementation of their key bio-inspired design principles using the best of robotic engineering. This First Grant Scheme application, leveraged by major strategic investment from Imperial College to the AquaMAV research area and a recent donation of £1.25m for a new flight arena, will allow me to establish scientific leadership in bio-inspired aerial robotics. It will build the basis to deliver the key building blocks for aerial-aquatic robots requiring innovation in propulsion, adaptive morphologies and autonomy that allow for hybrid mobility. The final goal of this first grant is to demonstrate novel propulsion systems and wing folding mechanisms that are tailored to AquaMAV requirements. Following this grant the systems will be integrated in a fully featured AquaMAV that can fly, dive into water, dive while taking water samples and transition back to propeller driven flight. Besides the application oriented impact, this project will also deliver key scientific insights on locomotion modalities for robotics across varied terrain. Having partnered with five industrial partners, including the National Nuclear Labs and Shell, will provide the pathways to industrial impact and continued research support building on the results from this grant. Major corporate interest in this project is demonstrated by committed resources of more than £65k of in kind benefits from industrial partners and two CASE studentships for higher Technology Readiness Levels of AquaMAV. This grant will provide the bridge in research staff resources that will allow the demonstration of AquaMAV principles and build the basis for future impact. Aerial-aquatic robot mobility is widely unexplored and this project will initiate a novel and high impact area in academia as demonstrated by support from leaders in robotics and biology, including support from Harvard University. Once realised, AquaMAVs will be a major technological stepping stone for new robotic applications, such as search and rescue in flooded buildings, for water sampling during oil spills or after nuclear accidents, and for low cost oceanographic data collection, all of which are areas of vital national importance to the U.K.

Deforestation and forest degradation are causing widespread loss of tropical biodiversity, profoundly impacting ecosystem functioning as well as stocks of natural resources and ecosystem assets (natural capital). The greatest reductions in diversity are experienced as forests are converted to permanent agriculture, a process that disrupts the delivery of important ecosystem services such as pollination and pest control. In contrast, the impacts from well-managed smallholder agriculture are less extreme, as the associated land parcels are typically embedded within landscape mosaics comprising fallows and forest remnants. Wallacea is currently emerging as a new developmental frontier in Indonesia and a target for agribusiness and extractive industries. A particularly understudied part of the Asian tropics, it has an exceptionally distinctive vertebrate diversity which forms the second highest level of endemism in the world, making the region a global priority for both conservation and ecosystem service provision. In addition, land-use history and current trajectories remain poorly understood, with the region notably omitted from recent deforestation baselines for this very reason. In fact, the diverse history of the Wallacea archipelago is acknowledged as a major source of uncertainty when applying land-use change models developed from elsewhere in Southeast Asia, as well as predicting the impacts of future environmental change. Given that further forest degradation and agricultural conversion are expected in Wallacea, the future prospects for natural capital in the region depend to a large extent on how we manage human-modified landscapes. Bringing together an interdisciplinary team from British and Indonesian universities, with four NGO partners active in Wallacea, this project will deliver the science needed to understand tensions in land-use and the responses of biodiversity to environmental change in Wallacea. We propose a novel and ambitious study of biodiversity responses to recent and historical land-cover change across multiple landscapes in little-studied islands, so that evaluations of current land-use policies and predictions of future environmental scenarios will be evidence-based and realistic. More specifically, spatial trajectories of land-cover change will be generated for each landscape, drawing from publicly-available remote-sensed data and local land-use plans. This will enable us to hindcast forest cover back to Wallace's time and forecast to key target years for international policy commitments (e.g. 2030 for the UN Sustainable Development Goals and 2050s for the UN Framework on Climate Change). Significantly, we will generate new biodiversity data from across land-cover gradients in forests, agroforests and intensive farmland (e.g. cocoa, oil palm, coffee), model community responses to past, present and future forest cover, and apply state-of-the-art genomics methods to assess genetic and evolutionary responses to land-cover change for several important conservation flagship species (NERC-Ristekdikti programme's Goals 1 and 2). Focusing on terrestrial vertebrates, the fauna that environmental policies aiming at safeguarding biodiversity are typically focused upon, we will track Alfred Russel Wallace's journey through Sulawesi and the Moluccas (Maluku). Finally, with collaboration from Project Partners with additional expertise in Wallacea, we will evaluate the impact of current land-use policies on ecosystem assets and dependent human beneficiaries, drawing on our land-cover, biodiversity and additional carbon and socio-economic data (Programme's Goals 2 and 3). Because many provinces in Indonesia are yet to implement newly-required spatial planning processes, our joint environmental research has an unprecedented opportunity to inform local development.