C. F. Clements

Stability metrics are robust across data qualities but sensitive to community size.

The paper evaluates stability metrics for ecosystems using time series data, finding they are robust to data quality but sensitive to community size.

Early warning signals are hampered by a lack of critical transitions in empirical lake data

This study evaluates early warning signals for ecological regime shifts in lake data and finds few critical transitions, with mixed performance of prediction methods.

Restoring local climate refugia to enhance the capacity for dispersal-limited species to track climate change

This study simulates how restoring climate refugia can enhance species persistence under climate change by increasing spatial heterogeneity and reducing environmental stochasticity.

Planktonic functional diversity changes in synchrony with lake ecosystem state

The paper examines the relationship between planktonic functional diversity and lake ecosystem state across five lakes, finding that FD synchronizes with state metrics but is not a reliable early indicator.

Community structure determines the predictability of population collapse

This study shows that community structure affects the reliability of early warning signals for population collapse, with abundance-based signals being more reliable than trait-based ones in multi-species communities.

Body mass and latitude as global predictors of vertebrate populations exposure to multiple threats

Machine learning methods trained on simple models can predict critical transitions in complex natural systems

This paper trains a deep neural network on simple models to predict critical transitions in complex systems, showing high reliability in detecting and classifying catastrophic vs. non-catastrophic transitions in real-world data.

Effect of time-series length and resolution on abundance- and trait-based early warning signals of population declines

This study evaluates how time-series length and resolution affect the effectiveness of early warning signals for population decline, showing that trait data can improve forecasting accuracy.

Eco-evolutionary processes underlying early warning signals of population declines

This study shows that ecological and evolutionary factors influence the strength of early warning signals for population declines, and combining trait and abundance data improves prediction accuracy.

When do shifts in trait dynamics precede population declines?

The paper shows that shifts in fitness-related traits can precede population declines under medium-to-slow environmental change, influenced by predictability, reproductive rate, plasticity, and genetic variation.