Assessing the forecastability & forecast skill of models to predict sex ratios of Valeriana edulis

Abstract

Anthropogenic actions are rapidly changing ecological systems, and the species specific responses to this change are expected to have great impacts on ecological organization, natural resource abundance, and human health. Scientists urgently need accurate forecasts to help inform social, political, and environmental decisions. However, from past attempts at ecological forecasts, it is unclear whether existing data and approaches are sufficient to meet this challenge. This study addresses this knowledge gap by forecasting the population sex ratio of the long-lived, dioecious perennial ​Valeriana edulis in which biased sex ratios increase the likelihood of extinction​. ​We constructed forecasting models with varying biological complexity, trained each model with historic data, and assessed the forecast skill of each model in predicting the sex ratio observed in 16 populations in 2020. Despite the prior expectation that male frequency is increasing in populations over time, our linear trend forecasts did not produce a high forecast skill. In contrast, our climate-informed models produced forecasts with the highest skill, outperforming the null model for 14 out of 16 populations. Ultimately, these results indicate that sex ratios might be predictable given our current understanding of species biology. On a larger scale, this study presents an optimistic outlook for ecological forecasting to inform biodiversity policy and inform more sustainable management decisions under ongoing climate change.

Local Knowledge Graph (11 entities)

Related Works

Items connected by shared entities, co-authorship, citations, or semantic similarity.

References (11)

1 in Knowledge Hub, 10 external