Global-scale sustainability transformations are urgently required to address catastrophic climate change and biodiversity loss. In practice, however, the conservation, restoration and management actions to combat these crises will be applied at local scales: to individual forests, fields, lakes, and grasslands. Practitioners, policymakers and ecologists have long-recognised the futility of silver bullet, 'one-size-fits-all' strategies, and the importance of targeting and tailoring actions to specific or individual contexts, and developing action plans for specific species, or individual sites. For example, the UK government's ambitious tree planting initiatives emphasise the importance of 'the right tree in the right place' to achieving Net Zero, because the wrong trees planted in the wrong place may fail to establish, or worse still, lead to net carbon emissions that persist for decades. However, despite the need for context-specific information for individual sites, ecology has traditionally focused on average effects, rather than individual effects, undermining the ultimate goal of applied ecology: the application of ecological science to the real-world.

To do science that can truly inform on-the-ground policies, ecology can capitalise on methods developed by the rapidly advancing field of 'precision medicine', which similarly rests on the premise that the effect of a given drug or treatment may be different for each individual patient, necessitating personalised healthcare plans. The application of precision medicine methodologies, which require 'big data', is now possible due to the increasing availability of vast quantities of ecological data from multiple sources, including remote sensors, citizen scientists and environmental monitoring networks. In this exploratory project, we will adapt the use of precision medicine approaches to applied ecological problems and develop novel approaches that help guide the next generation of applied ecologists into the big-data era. We will develop frameworks and tools for sampling and modelling big data in a way that yields accurate and practicable knowledge that can inform on-the-ground actions.

This project will achieve this using a 'virtual ecology' approach using an existing computer model - 'iLand' - that simulates real, forested landscapes in Europe, the U.S.A and Japan. First, we will use iLand to virtually implement forest restoration actions to individual forests across these landscapes. The forests in these landscapes will then be subjected to virtual sampling, thus mimicking the collection of data by ecologists, citizen scientists and remote sensing technologies. For example, we will mimic surveyors that measure forest biodiversity as part of a national environmental monitoring campaigns. We will then apply different modelling methods to analyse the virtually sampled data, including methods developed in the field of precision medicine as well as methods that are commonly used in ecology. By comparing the results of these analyses against the 'true' simulated data, we will be able to evaluate how different sampling and modelling decisions influence the accuracy of our results. Specifically, we will assess the accuracy of predictions individual forest responses to restoration actions across a range of environmental conditions (i.e., elevation, climate, soil types). We will use this understanding to develop guidelines for designing future environmental monitoring surveys, and modelling of big data, that maximises the accuracy of these predictions.

By shifting applied ecology's focus away from average effects, towards actionable, individual-level effects, our guidelines will enable policymakers to target and prioritise restoration actions to areas where desired effects are greatest, thus maximising efficacy of limited resources for addressing climate change and biodiversity loss.