Despite efforts from scientists and regulators, biodiversity is declining at an alarming rate. Unless we find transformative solutions to preserve biodiversity, future generations may not be able to enjoy nature's services. We have developed a conceptual framework that establishes the links between biodiversity dynamics and abiotic change through time and space using artificial intelligence. Here, we apply this framework to a freshwater ecosystem with a known history of human impact and study 100 years of community-level biodiversity, climate change and chemical pollution trends. We apply explainable network models with multimodal learning to community-level functional biodiversity measured with multilocus metabarcoding, to establish correlations with biocides and climate change records. We observed that the freshwater community assemblage and functionality changed over time without returning to its original state, even if the lake partially recovered in recent times. Insecticides and fungicides, combined with extreme temperature events and precipitation, explained up to 90% of the functional biodiversity changes. The community-level biodiversity approach used here reliably explained freshwater ecosystem shifts. These shifts were not observed when using traditional quality indices (e.g. Trophic Diatom Index). Our study advocates the use of high-throughput systemic approaches on long-term trends over species-focused ecological surveys to identify the environmental factors that cause loss of biodiversity and disrupt ecosystem functions.
Keywords: computational biology; ecology; freshwater; functional biodiversity; machine learning; multilocus metabarcoding; none; sedaDNA; systems biology.
Over long periods of time, environmental changes – such as chemical pollution and climate change – affect the diversity of organisms that live in an ecosystem, known as ‘biodiversity’. Understanding the impact of these changes is challenging because they can happen slowly, their effect is only measurable after years, and historical records are limited. This can make it difficult to determine when specific changes happened, what might have driven them and what impact they might be having. One way to measure changes in biodiversity over time is by analysing traces of DNA shed by organisms. Plants, animals, and bacteria living in lakes leave behind genetic material that gets trapped and buried in the sediment at the bottom of lakes. Similarly, biocides – substances used to kill or control populations of living organisms – that run-off into lakes leach into the sediment and can be measured years later. Therefore, this sediment holds a record of life and environmental impacts in the lake over past centuries. Eastwood, Zhou et al. wanted to understand the relationship between environmental changes (such as the use of biocides and climate change) and shifts in lake biodiversity. To do so, the researchers studied a lake community that had experienced major environmental impacts over the last century (including nutrient pollution, chemical pollution and climate change), but which appeared to improve over the last few years of the 20th century. Using machine learning to find connections over time between biodiversity and non-living environmental changes, Eastwood, Zhou et al. showed that, despite apparent recovery in water quality, the biodiversity of the lake was not restored to its original state. A combination of climate factors (such as rainfall levels and extreme temperatures) and biocide application (particularly insecticides and fungicides) explained up to 90% of the biodiversity changes that occurred in the lake. These changes had not been identified before using traditional techniques. The functional roles microorganisms played in the ecosystem (such as degradation and nitrogen metabolism) were also altered, suggesting that loss of biodiversity may lead to loss of ecosystem functions. The findings described by Eastwood, Zhou et al. can be used by environmental regulators to identify species or ecosystems at risk from environmental change and prioritise them for intervention. The approach can also be used to identify which chemicals pose the greatest threat to biodiversity. Additionally, the use of environmental DNA from sediment can provide rich historical biodiversity data, which can be used to train artificial intelligence-based models to improve predictions of how ecosystems will respond to complex environmental changes.
© 2023, Eastwood, Zhou et al.