Climatic Change

, Volume 131, Issue 4, pp 607–620

Regional coral responses to climate disturbances and warming is predicted by multivariate stress model and not temperature threshold metrics

  • Timothy R. McClanahan
  • Joseph Maina
  • Mebrahtu Ateweberhan
Article

DOI: 10.1007/s10584-015-1399-x

Cite this article as:
McClanahan, T.R., Maina, J. & Ateweberhan, M. Climatic Change (2015) 131: 607. doi:10.1007/s10584-015-1399-x

Abstract

Oceanic environmental variables derived from satellites are increasingly being used to predict ecosystem states and climate impacts. Despite the concerted efforts to develop metrics and the urgency to inform policy, management plans, and actions, few metrics have been empirically tested with field data for testing their predictive ability, refinement, and eventual implementation as predictive tools. In this study, the abilities of three variations of a thermal threshold index and a multivariate stress model (MSM) were used to predict coral cover and community susceptibility to bleaching based on a compilation of field data from Indian Ocean reefs across the strong thermal anomaly of 1998. Field data included the relative abundance of coral taxa 10 years before the large-scale temperature anomaly, 2 years after (1999–2000), and during the post-bleaching recovery period (2001–2005) were tested against 1) a multivariate model based on 11 environmental variables used to predict stress or environmental exposure (MSM), 2) estimates of the time until the current mean maximum temperature becomes the mean summer condition (TtT), 3) the Cumulative Thermal Stress (CTS) for the full satellite record, and 4) the 1998 Annual Thermal Stress (1998 ATS). The MSM showed significant fit with the post-1998 cover and susceptibility of the coral community taxa (r2 = 0.50 and 0.31, respectively). Temperature threshold indices were highly variable and had relatively weak or no significant relationships with coral cover and susceptibility. The ecosystem response of coral reefs to climatic and other disturbances is more complex than predicted by models based largely on temperature anomalies and thresholds only. This implies heterogeneous environmental causes and responses to climate disturbances and warming and predictive models should consider a more comprehensive multiple parameter approach.

Copyright information

© Springer Science+Business Media Dordrecht 2015

Authors and Affiliations

  • Timothy R. McClanahan
    • 1
  • Joseph Maina
    • 1
    • 2
  • Mebrahtu Ateweberhan
    • 1
    • 3
  1. 1.Wildlife Conservation Society, Marine ProgramsBronxUSA
  2. 2.Australia Research Council Centre of Excellence for Environmental Decisions (CEED)University of QueenslandSt LuciaAustralia
  3. 3.School of Life SciencesUniversity of WarwickCoventryUK