Young fishes persist despite coral loss on the Great Barrier Reef

Unprecedented global bleaching events have led to extensive loss of corals. This is expected to lead to extensive losses of obligate coral-dependent fishes. Here, we use a novel, spatially-matched census approach to examine the nature of fish-coral dependency across two mass coral bleaching events. Despite a >40% loss of coral cover, and the ecological extinction of functionally important habitat-providing Acropora corals, we show that populations of obligate coral-dependent fishes, including Pomacentrus moluccensis, persisted and – critically – recruitment was maintained. Fishes used a wide range of alternate reef habitats, including other coral genera and dead coral substrata. Labile habitat associations of ‘obligate’ coral-dependent fishes suggest that recruitment may be sustained on future reefs that lack Acropora, following devastating climatic disturbances. This persistence without Acropora corals offers grounds for cautious optimism; for coral-dwelling fishes, corals may be a preferred habitat, not an obligate requirement.


Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.

n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one-or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted Give P values as exact values whenever suitable.

For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated Our web collection on statistics for biologists contains articles on many of the points above.

Software and code
Policy information about availability of computer code

Data collection
We quantified changes in coral cover using a software called photoQuad, which generates randomly stratified points over each 'planar perspective' photoquadrat.

Data analysis
All statistical modelling was performed in the software R, using the lme4 and glmmTMB packages.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A list of figures that have associated raw data -A description of any restrictions on data availability The data that support the findings of this study are available from the corresponding author upon request.

Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences
Behavioural & social sciences Ecological, evolutionary & environmental sciences 2 nature research | reporting summary

October 2018
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf Ecological, evolutionary & environmental sciences study design All studies must disclose on these points even when the disclosure is negative.

Study description
To quantify changes in the cover of live coral and coral-associated fishes, in response to mass coral bleaching, we compared data from a total of 132 photoquadrats, repeatedly sampled across 24 months. Specifically, in each field trip, we sampled 19 transect sites across the reefal system of Lizard Island. Transects varied in length depending on reef length, i.e. range 50-210 m. Within each transect, photographs of a 1 m2 quadrat were taken at approx. 5 m intervals, i.e. range per transect, 12-38 quadrats depending on transect length. Each specific quadrat location was photographed 4 times: 1) Jan. 2016 -before mass bleaching; 2) April 2016 -peak bleaching, 3) Oct. 2016 -6-mo post bleaching, 4) Jan. 2018 -24-mo since first sampling trip. Changes in coral cover and associated fish assemblage were then compared between sampling trips, in relation to the 2016-2017 mass bleaching events.

Research sample
Our research samples were replicate 1 m2 quadrats. In total, we analysed 132 individual 1 m2 quadrats and each specific quadrat location was sampled four times across 24-months. Within each quadrat, we 1) quantified the percent cover of benthic categories (e.g. live coral cover) and 2) identified and recorded all visible fishes, which consisted primarily of small, coral-associated fishes.
Although quantifying coral cover from quadrats is common practice, quantifying reef fishes from photographs is a novel approach and provides several distinct advantages, including 1) reduced diver disturbance to fish communities and 2) no time constraints when quantifying fishes.

Sampling strategy
Quadrats were spread across the entire lagoon and fringing reefs of the sampling location, Lizard Island. Quadrats were selected if they had greater than 20% coral cover (based on visual estimates) as the focus of the paper was on fish responses to coral loss. The number of quadrats was not set a priori but depended on the initial results. Of the 451 quadrats examined 132 had the requisite starting coral cover and were subsequently examined throughout the 24 month period.

Data collection
Our data collection procedure involved both a field and lab component. During the field component, we collected a series of photographs of a 1 m2 quadrat at 451 quadrat locations (132 of 451 were analysed). Specifically, at each quadrat location, we collected three images: 1) an undisturbed horizontal perspective photograph of the reef and coral-associated reef fishes (at a distance of 2 m), taken within seconds of reaching the site and prior to the placement of the quadrat, 2) a second horizontal perspective photograph with the 1 m2 quadrat in place, using the identical camera placement as in the first image, and 3) a planar perspective photograph (i.e. bird's-eye view) of the 1 m2 quadrat in place over the substratum. Images were subsequently examined back in the laboratory using standard computer-based image analysis. Coral cover was quantified using the software photoQuad, which generates 40 random stratified data points over the image, from which we calculated percent cover. All corals were quantified to species level when possible, categorized by state, i.e. living, bleached etc. To quantify reef fishes, photoquadrats were processed in Adobe Illustrator, by drawing an outline of the quadrat on the first photograph of the series (i.e. undisturbed), using the second photograph in the series as a reference. All visible reef fishes within the delineated 1 m2 (and 1.5 m above the quadrat) were recorded to species level and categorized as either adult or recruit. All field photographs were taken by R.S. or S.T. All images were processed by one person for consistency (S.T.).
Timing and spatial scale To quantify changes in coral cover and the associated fish assemblage, as a result of mass bleaching, we carefully timed our sampling trips across a 24-mo timeline. The first sampling trip was conducted in Jan. 2016, immediately before the 2016 mass bleaching event.
Data from this sampling trip would serve as a baseline. To quantify the extent of bleaching, we re-sampled the same sites in April 2016, at the peak of the mass bleaching event. All sites were once again sampled in Oct. 2016, 6-mo post bleaching and Jan. 2018, 24-mo post bleaching, in order to quantify both short and long-term changes in both coral and fish communities, in response to mass bleaching. The duration of each sampling trip was approx. 2 weeks. All samples were collected from reefs surrounding Lizard Island, which was near the GBR epicenter of the 2016 global mass bleaching event.

Data exclusions
Per sampling period, a total of 132 (out of 451) 1 m2 photoquadrats were analysed. Since we were explicitly examining the response of coral-associated reef fishes (i.e. fishes with a facultative or obligate so-called dependency on live coral) to mass bleaching, we only analysed quadrats with a minimum live coral cover of 20% (based on visual estimates) in the first sampling period. For analyses on benthic cover, points which were categorized as 'unidentifiable', were excluded from analyses. This number, however, was very low, i.e. 16 of 21,120 benthic points examined were excluded.

Reproducibility
Our study quantifies changes in both live coral cover and coral-associated fishes across the 2016 global mass bleaching event, a natural phenomenon that cannot be replicated. However, our results are directly comparable to other studies of mass bleaching events.
One potential issue in our sampling design is is quadrat placement, i.e. sampling the same exact site across all trips. The effect of random sampling, as well as, variation in quadrat placement, was assessed by examining 40 random points from the first sampling period versus the exact same 40 points from the third sampling period (n = 15 randomly selected quadrats). The results showed just a 1.4% difference, and hence, our method appears to provide a good indication of benthic changes among temporal samples with excellent reproducibility.

Randomization
The starting locations of each of the 19 surveyed transect sites were chosen haphazardly in the first sampling trip, i.e. we did not focus our efforts in areas of the reef which were more prone to bleaching or that exhibited higher bleaching rates in past. Transects were spread across the reefal system of Lizard Island.