Recovery pathways for lake ecosystems

Lead Research Organisation: University of Southampton


Global water demand has increased by 600% over the past 100 years and is predicted to grow significantly over the next few decades. Under current and future scenarios, aquatic ecosystems are under threat from multiple pressures, most notably human population growth with its often-conflicting demands for water, food and energy security, leading to declining freshwater quantity and quality. More than 30% of global biodiversity loss is due to pollution of water resources and aquatic ecosystems. Here, we focus on the contribution these challenges are having on freshwater lakes, where nutrient enrichment has been recognised as the key pressure. Across the world, increasing numbers of lakes are collapsing into polluted states. For example, in 2005, all 14,000 UK lakes >1 ha in size were risk-assessed for ecological status, of which over half failed the 'good' threshold, being judged at risk and likely to require some remediation. Most recently, in 2020, new data from the Environment Agency revealed that only 14% of lakes were classed as in good ecological health in England. Concomitantly, in a study of 273 Chinese lakes, amongst those with human impacts, only 22% showed relatively stable and resilient states. It is clear that such trajectories are unsustainable in terms of global freshwater use and functioning. Many countries have recognised the urgency of the problem, and are enacting local and global policies (e.g. EU, UN) for freshwater lake restoration. Despite substantial investment, current restoration projects are often based on short-term assumptions, and hence can fail due to the lack of knowledge of mid- to long-term recovery pathways after system collapse.

Our project addresses this challenge, to deliver a step change in our understanding of pathways to lake ecosystem recovery. Much of the difficulty in restoring lake ecosystems to 'good' (resilient) clear water states is due to hysteresis (the dependence of the state of a system on its history), keeping the unwanted turbid (polluted) state dominant by recycling of nutrients without any further additions. Consequently, simply removing the pollutant source (when possible or desirable) does not necessarily lead to lake ecosystem recovery. Our project will tackle this challenge by quantifying the balance of factors that influence the recovery of freshwater lake ecosystems from undesirable states. While controls on nutrients (diffuse and/or point-source) are often critical, the recycling rates, and internal feedback loops, in addition to other multidimensional stressors such as climate change, can also be important factors to consider. In our project, we will review current global lake datasets, and undertake detailed empirical work on a subset of lakes that have undergone full or partial recovery. In both datasets (i.e. global lakes, and our subset with new empirical analyses across all trophic levels), we will use newly developed indicators of ecological structure to assess which recovery pathway lakes can take (i.e. linear, step change, or hysteretic), and how long such recovery may take. We will use this knowledge to build systems-based models that can simulate the multidimensional causes of lake degradation, ecosystem collapse, and recoveries in a range of freshwater lake systems. The models will be able to simulate different future lake states, depending on inputs and associated management decisions. We will work with a wide range of stakeholders to run and test the models, so that ultimately, we can determine the best course of action for recovery with minimal hysteresis.

In summary, the step-change outcome of our project will be an ability to model an improved prediction of different lake recovery pathways, the relative roles of differing driver combinations, and respondent ecosystem trophic level and overall resilience. Our project will contribute to meeting one of the major challenges facing humanity: the sustainable management of Earth's freshwater resources.


10 25 50