Is local biodiversity declining or not? A summary of the debate over analysis of species richness time trends

Recently, a debate has developed over how biodiversity is changing across the planet. While most researchers agree species extinctions are increasing globally due to human activity, some now argue that species richness at local scales is not declining as many biologists have claimed. This argument stems from recent syntheses of time-series data that suggest species richness is decreasing in some locations, increasing in others, but not changing on average. Critics of these syntheses (like us) have argued there are serious limitations of existing time-series datasets and their analyses that preclude meaningful conclusions about local biodiversity change. Specifically, authors of these syntheses have failed to account for several primary drivers of biodiversity change, have relied on data poor time-series that lack baselines needed to detect change, and have unreasonably extrapolated conclusions. Here we summarize the history of this debate, as well as key papers and exchanges that have helped clarify new issues and ideas. To resolve the debate, we suggest future researchers be more clear about the hypotheses of biodiversity change being tested, focus less on amassing large datasets, and more on amassing high-quality datasets that provide unambiguous tests of the hypotheses. Researchers should also keep track of the contributions that native versus non-native species make to biodiversity time trends, as these have different implications for conservation. Lastly, we suggest researchers be aware of pros and cons of using different types of data (e.g., time-series, spatial comparisons), taking care to resolve divergent results among sources to allow broader conclusions about biodiversity change. •Recent analyses of time-series data suggest biodiversity is not declining at local scales as often claimed.•We argue existing time-series are inappropriate for testing hypotheses about biodiversity change.•Datasets are low quality, lack spatial representation, and fail to account for key drivers of change.•Failure to consider these problems has led to misinterpretations and unwarranted extrapolations.•New analyses must use higher-quality datasets to test improved hypotheses about diversity change.