EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Partitioning the variance of species–area relation into sampling‐ versus scale‐dependent components

Youhua Chen, Tsung‐Jen Shen

Ecography · 16.11.2026

DOI: 10.1002/ecog.08712 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Ecography, online erschienen am 16.11.2026. Verfasst von Youhua Chen und 1 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The species–area relation (SAR) has been extensively studied over the past century, yet the sources of variation around an estimated SAR curve remain poorly understood. Here, we show that the sampling variance of the SAR can be decomposed into two additive components: 1) conditional variance, assuming that regional richness is fixed and represents uncertainty associated with local sampling; and 2) an additional component attributable to uncertainty in regional species richness, including uncertainty arising from rare species that remain undetected in the regional sample. We establish this decomposition mathematically and evaluate its behavior through numerical simulations. On this basis, empirical SAR observations can be classified as falling within the conditional confidence interval (CI), within the gap between the conditional and unconditional CIs, or outside the unconditional CI. The last outcome indicates variation not accounted for by the two sampling‐dependent components and may reflect scale‐dependent ecological processes or other sources of unexplained variation. Empirical applications showed that the relative contributions of local sampling uncertainty, regional richness uncertainty, and variation beyond these sampling‐dependent components differed among datasets, spatial scales, and sampling designs. For spatially explicit sampling of a tropical forest plot stem‐mapping dataset, observations were generally consistent with local sampling uncertainty when small sampling grains were freely distributed across space. In contrast, observations from a single contiguous sampling area more frequently fell outside the unconditional CI, a pattern that may reflect spatially structured ecological processes not represented by the sampling‐variance components. For nested sampling of amphibian diversity on the Chengdu Plain in SW China, both local sampling uncertainty and variation beyond the unconditional CI were frequently observed, whereas uncertainty in regional richness became more prominent when the local sampling area fraction exceeded 0.6. The resulting classifications also varied with the order of nested sampling: for the amphibian dataset, the large‐to‐small scheme accumulated species richness more slowly than the small‐to‐large scheme did. In conclusion, we develop a novel and readily applicable hierarchical variance‐partitioning framework to distinguish variation attributable to local sampling, uncertainty in regional richness, and variation not explained by these sampling‐dependent components along SAR curves.

Zitieren

APA

Chen, Y., & Shen, T. (2026). Partitioning the variance of species–area relation into sampling‐ versus scale‐dependent components. Ecography. https://doi.org/10.1002/ecog.08712

Vancouver

Chen Y, Shen T. Partitioning the variance of species–area relation into sampling‐ versus scale‐dependent components. Ecography. 2026. doi:10.1002/ecog.08712

BibTeX
@article{Chen2026,
  author = {Chen, Youhua and Shen, Tsung‐Jen},
  title = {Partitioning the variance of species–area relation into sampling‐ versus scale‐dependent components},
  journal = {Ecography},
  year = {2026},
  doi = {10.1002/ecog.08712}
}
Kostenlos starten →