Quantifying spatial gradients in coral reef benthic communities using multivariate dispersion

Alice Lawrence, Adel Heenan, Gareth J. Williams

Research output: Contribution to journalArticlepeer-review

49 Downloads (Pure)

Abstract

Tropical coral reefs are dynamic, disturbance-driven ecosystems that are heterogeneous across space and time, partly owing to gradients in cross-scale human impacts and natural environmental factors. Localized management interventions that strive to maintain the long-term persistence and function of coral reefs need to be informed by how and why reef habitats vary. Using the 'multivariate dispersion' metric, a statistical approach to measure ecological community variability, we quantified spatial gradients in coral reef benthic communities around Tutuila Island in American Samoa, central South Pacific. Benthic communities with low, medium and high dispersion each had distinct and consistent underlying benthic community characteristics. Low dispersion sites were consistently characterized by high hard coral cover, medium dispersion sites were generally dominated by crustose coralline algae, while high dispersion sites were dominated by turf and fleshy coralline algae. Variability in hard coral and turf algal cover explained 42% of the underlying variation in benthic community dispersion across sites, while site-level gradients in human impacts and environmental factors did not correlate well with variations in benthic community dispersion. The metric should be further tested on temporal data to determine whether it can summarize complex community changes in response to and following acute disturbance.

Original languageEnglish
Article number241254
JournalRoyal Society Open Science
Volume12
Issue number4
DOIs
Publication statusPublished - 2 Apr 2025

Keywords

  • American Samoa
  • benthic heterogeneity
  • betadisper
  • community variability
  • coral life history
  • coral species

Fingerprint

Dive into the research topics of 'Quantifying spatial gradients in coral reef benthic communities using multivariate dispersion'. Together they form a unique fingerprint.

Cite this