High-resolution modeling of thermal thresholds and environmental influences on coral bleaching for local and regional reef management

Abstract

Coral reefs are one of the world’s most threatened ecosystems, with global and local stressors contributing to their decline. Excessive sea-surface temperatures (SSTs) can cause coral bleaching, resulting in coral death and decreases in coral cover. A SST threshold of 1 °C over the climatological maximum is widely used to predict coral bleaching. In this study, we refined thermal indices predicting coral bleaching at high-spatial resolution (1 km) by statistically optimizing thermal thresholds, as well as considering other environmental influences on bleaching such as ultraviolet (UV) radiation, water turbidity, and cooling effects. We used a coral bleaching dataset derived from the web-based monitoring system Sango Map Project, at scales appropriate for the local and regional conservation of Japanese coral reefs. We recorded coral bleaching events in the years 2004–2016 in Japan. We revealed the influence of multiple factors on the ability to predict coral bleaching, including selection of thermal indices, statistical optimization of thermal thresholds, quantification of multiple environmental influences, and use of multiple modeling methods (generalized linear models and random forests). After optimization, differences in predictive ability among thermal indices were negligible. Thermal index, UV radiation, water turbidity, and cooling effects were important predictors of the occurrence of coral bleaching. Predictions based on the best model revealed that coral reefs in Japan have experienced recent and widespread bleaching. A practical method to reduce bleaching frequency by screening UV radiation was also demonstrated in this paper.

Bleaching and post-bleaching mortality of Acropora corals on a heat-susceptible reef in 2016

“…The SST was measured at 8:30 on the surface of a seawater intake well at the farming center. DHW (°C-weeks), which measures the cumulative effect of thermal stress (Liu, Strong & Skirving, 2003) based on MMM max (the mean of the maximum monthly SST from each year in the time period of the climatology; (Donner, 2009; Donner, 2011) over 12 weeks, was calculated for 1998 and 2016 following Kumagai & Yamano (2018): where i : day, HS max i : HotSpots max i = SST i -MMM max . The mean of the monthly means of the SST in July and August, when SST is highest in the year, from 1988 to 1997 was employed as MMM max , and α = 1 °C was used, following previous studies (Liu et al, 2006).…”

Coral bleaching patterns are the outcome of complex biological and environmental networking

“…This is likely also true for the regulatory factors, notably oxygen availability, where eutrophication events that drive hypoxia are occurring against the backdrop of ocean warming‐driven deoxygenation (e.g., Altieri et al, ). Consequently, while environmental models describing reef trajectories are becoming increasingly sophisticated (e.g., Baird et al, ; Ellis et al, in press; Kumagai, Yamano, & Committee Sango‐Map‐Project, ; Wolff et al, ), we now need to urgently develop these to account for how net bleaching outcomes reflect dose dependencies within the entire environmental network (Figure ), and in turn the affect the inherent underlying metabolic network(s) (Figure ). This is no small task but central to guiding more informed management decisions and interventions based on what will bleach, where and when.…”

“…However, in doing so, management (re‐) prioritization must be careful that efforts to minimize exposure to one stressor does not increase exposure to another (see Bruno et al, ), returning us a central issue: how can MPA (re‐) planning be effectively achieved without understanding the complex environmental network that governs bleaching susceptibility? Realizing such a goal clearly rests on rapidly improving capacity to monitor reef environment condition, but also applying these data to more advanced network models that can track how changing reef environments (Figure ; see also, Ellis et al, in press) trigger alternate metabolic cascades and hence bleaching outcomes (Figure ; Baird et al, ; Kumagai, Yamano, & Committee Sango‐Map‐Project, ).…”

“…The most recent global back‐to‐back (2015–2017) bleaching events have re‐affirmed the fragility of coral reef ecosystems to climate change and associated local social–environmental stressors (e.g., Darling et al, ; Hughes et al, , ; Lapointe et al, )—for many, these events were a confronting first for how rapid and destructive bleaching‐driven mortality occurs. Advances in bleaching forecasting (e.g., Heron, Maynard, Hooidonk, & Eakin, ; Kumagai, Yamano, & Committee Sango‐Map‐Project, ; van Hooidonk et al, ) meant that the most recent events particularly provided new capacity for research communities to capture mass bleaching as it unfolded, transforming empirical knowledge of bleaching patterns in nature as well as improving understanding of the core biological and ecological responses at play (Hughes et al, , ; McClanahan et al, ; see also Eakin et al, ). We are, as a result, at an important, and exceptionally time‐sensitive, turning point in our understanding of coral bleaching and how we move forward.…”

Effects of moderate thermal anomalies on Acropora corals around Sesoko Island, Okinawa

“…Following these criteria, DHW between 4–8°C-week are defined as representing a moderate thermal anomaly in this study. DHW has been widely used to quantify bleaching thresholds and to asses thermal stress variability on a large spatial scale (≥hundreds of km) [38–40]. Small-scale thermal disparity and consequent differential bleaching responses have been observed [33,41].…”