Fig 1.
Whitsunday Islands, track of STC Debbie, and survey sites.
Maps showing the location of the Whitsunday Islands on the east coast of Australia and the track of STC Debbie (A) and the 43 survey sites around the Whitsunday Islands surveyed between 2012–2023 (B). Maps were generated in R using the ozmaps and sf packages, which use publicly available geospatial data from Australian government sources.
Fig 2.
Mean percent cover (±SE) of all hard corals, massive and encrusting corals and complex corals in the Whitsunday Islands. Tick marks indicate the beginning of each year, and the placement of points between tick marks indicate the time of year of the survey. The vertical red line and cyclone symbol represents the timing of STC Debbie on 25th March 2017. Standard error was calculated for each year but error bars are not visible due to very low values (<1%).
Fig 3.
GLMM model outputs comparing coral cover among 3 time periods.
Model output for GLMMs testing differences in percent cover among three time periods for complex corals (A) and massive and encrusting corals (B). Estimated marginal means and 95% confidence intervals are plotted for each time period, with raw data as grey points and data distribution as coloured violin plots. 95% CIs are too small to be visualised. Pairwise comparisons with 95% confidence intervals are presented as contrast estimates of coral cover change between time periods. The grey dashed line in each contrast panel represents an interpretation of the contrast significance where this difference is considered significant if the confidence interval does not cross zero. Time periods are: Pre-cyclone (the 3 years prior to STC Debbie), Post-cyclone 1 (2017 and 2018) and Post-cyclone 2 (2022 and 2023).
Fig 4.
GLMM model outputs comparing damselfish species richness and density among 3 time periods.
Model output for GLMMs testing differences in damselfish species richness (A-C) and density (D-F) among three time periods. Estimated marginal means and 95% confidence intervals are plotted for each time period (A and D), with raw data as grey points and data distribution as coloured violin plots (B and E). 95% CIs are too small to be visualised. Pairwise comparisons with 95% confidence intervals are presented as contrast ratios between time periods (C and F). The grey dashed line in each contrast panel represents an interpretation of the contrast significance where this difference is considered significant if the confidence interval does not cross 1.
Fig 5.
GLMM model outputs comparing butterflyfish species richness and density among 3 time periods.
Model output for GLMMs testing differences in butterflyfish species richness (A-C) and density (D-F) among three time periods. Estimated marginal means and 95% confidence intervals are plotted for each time period, with raw data as grey points and data distribution as coloured violin plots. 95% CIs were too small to be visualised. Pairwise comparisons with 95% confidence intervals are presented as contrast ratios between time periods. The grey dashed line in each contrast panel represents an interpretation of the contrast significance where this difference is considered significant if the confidence interval does not cross 1.
Table 1.
Summary of pairwise contrasts between years. Pairwise contrasts between pairs of years from 2016 onwards, and between 2016 and 2023 (i.e., the overall change between the survey before STC Debbie and the most recent survey) for the 10 most abundant damselfish species and Chaetodon aureofasciatus. For significant contrasts, the percent change from one year to the next is given, with + or – indicating increases or declines in density. Non-significant contrasts are indicated by NS.
Fig 6.
Species level patterns for the 10 most abundant damselfishes.
Mean density (marginal means + /- SE) per survey year for each of the 10 most abundant damselfish species (A-J). Panels for each species are arranged in order of decreasing abundance.
Fig 7.
Species level patterns for the 3 most abundant butterflyfishes.
Mean density (marginal means + /- SE) per survey year for each of the 3 most abundant butterflyfish species (A-C). Panels for each species are arranged in order of decreasing abundance.
Fig 8.
Damselfish distance-based redundancy analysis.
The results of distance-based redundancy analysis (db-RDA) showing the relationship between damselfish community structure and benthic variables, and grouped by Pre-cyclone (2012−16), Post-cyclone 1 (2017−18) and Post-cyclone 2 (2022−24) time periods.
Fig 9.
Butterflyfish distance-based redundancy analysis.
The results of distance-based redundancy analysis (db-RDA) showing the relationship between butterflyfish community structure and benthic variables, and grouped by Pre-cyclone (2012−16), Post-cyclone 1 (2017−18) and Post-cyclone 2 (2022−24) time periods.