The world's tropical reef ecosystems, and the people who depend on them, are increasingly 60 impacted by climate change [1][2][3][4][5][6][7] Reef, as well as the potential influence of water quality and fishing pressure on the severity of 71 bleaching. 72The geographic footprints of mass bleaching of corals on the Great Barrier Reef have varied 73 strikingly during three major events in 1998 , 2002 and 2016). In 1998, bleaching was 74 primarily coastal and most severe in the central and southern regions. In 2002, bleaching was 75 more widespread, and affected offshore reefs in the central region that had escaped in 1998 8 . 76In 2016, bleaching was even more extensive and much more severe, especially in the 77 northern, and to a lesser extent the central regions, where many coastal, mid-shelf and 78 offshore reefs were affected (Fig. 1a, b). In 2016, the proportion of reefs experiencing 79 extreme bleaching (>60% of corals bleached) was over four times higher compared to 1998 80 or 2002 (Fig. 1f) The severity and distinctive geographic footprints of bleaching in each of the three 88 years can be explained by differences in the magnitude and spatial distribution of sea-surface 89 temperature anomalies (Fig. 1a, b 102The geographic pattern of bleaching also demonstrates how marine heatwaves can be (Fig. 2a) (Fig. 1g). largely escaped bleaching in the two earlier events (Fig. 1a). Thirty-five percent of the reefs (Fig. 1b, e). We conclude that the overlap of disparate geographic bleaching at the scale of both individual reefs and the entire Great Barrier Reef (Fig. 1a, b). 134We found a similar strong relationship between the amount of bleaching measured 135 underwater, and the satellite-based estimates of heat exposure on individual reefs (Fig. 3). 136Low levels of bleaching was observed at some locations when DHW values were only 2-3 137 o C-weeks. Typically, 30-40% of corals bleached on reefs exposed to 4 o C-weeks, whereas an 138 average of 70-90% of corals bleached on reefs that experience 8 o C-weeks or more (Fig. 3). 139Resistance and adaptation to bleaching 140 Once we account for the amount of heat stress experienced on each reef, adding 141 chlorophyll-a, a proxy for water quality, to our statistical model yielded no support for the 142 hypothesis that good water quality confers resistance to bleaching 13 . Rather, the estimated 143 effect of chlorophyll-a was to significantly reduce the DHW threshold for bleaching 144 (Extended Data Table 1). However, despite the statistical significance, the effect in real terms 145 beyond heat stress alone is very small (Extended Data Fig. 1). Similarly, we found no effect 146 of the level of protection (in fished or protected zones) on bleaching (P > 0.1: Extended Data 147 Table 1). These results are consistent with the broad-scale pattern of severe bleaching in the 148 northern Great Barrier Reef, which affected hundreds of reefs across inshore-offshore 149 gradients in water quality, and regardless of their zoning (protection) status (Fig. 1a, b). 150Simila...

Reef-building corals possess a range of acclimatisation and adaptation mechanisms to respond to seawater temperature increases. In some corals, thermal tolerance increases through community composition changes of their dinoflagellate endosymbionts (Symbiodinium spp.), but this mechanism is believed to be limited to the Symbiodinium types already present in the coral tissue acquired during early life stages. Compelling evidence for symbiont switching, that is, the acquisition of novel Symbiodinium types from the environment, by adult coral colonies, is currently lacking. Using deep sequencing analysis of Symbiodinium rDNA internal transcribed spacer 2 (ITS2) PCR amplicons from two pocilloporid coral species, we show evidence consistent with de novo acquisition of Symbiodinium types from the environment by adult corals following two consecutive bleaching events. Most of these newly detected symbionts remained in the rare biosphere (background types occurring below 1% relative abundance), but one novel type reached a relative abundance of ~33%. Two de novo acquired Symbiodinium types belong to the thermally resistant clade D, suggesting that this switching may have been driven by consecutive thermal bleaching events. Our results are particularly important given the maternal mode of Symbiodinium transmission in the study species, which generally results in high symbiont specificity. These findings will cause a paradigm shift in our understanding of coral-Symbiodinium symbiosis flexibility and mechanisms of environmental acclimatisation in corals.

Algebraic multigrid methods for large, sparse linear systems are a necessity in many computational simulations, yet parallel algorithms for such solvers are generally decomposed into coarse-grained tasks suitable for distributed computers with traditional processing cores. However, accelerating multigrid methods on massively parallel throughput-oriented processors, such as graphics processing units, demands algorithms with abundant fine-grained parallelism. In this paper, we develop a parallel algebraic multigrid method which exposes substantial fine-grained parallelism in both the construction of the multigrid hierarchy as well as the cycling or solve stage. Our algorithms are expressed in terms of scalable parallel primitives that are efficiently implemented on the GPU. The resulting solver achieves an average speedup of 1.8× in the setup phase and 5.7× in the cycling phase when compared to a representative CPU implementation.

Climate change is driving global declines of marine habitat-forming species through physiological effects and through changes to ecological interactions, with projected trajectories for ocean warming and acidification likely to exacerbate such impacts in coming decades. Interactions between habitat-formers and their microbiomes are fundamental for host functioning and resilience, but how such relationships will change in future conditions is largely unknown. We investigated independent and interactive effects of warming and acidification on a large brown seaweed, the kelp Ecklonia radiata , and its associated microbiome in experimental mesocosms. Microbial communities were affected by warming and, during the first week, by acidification. During the second week, kelp developed disease-like symptoms previously observed in the field. The tissue of some kelp blistered, bleached and eventually degraded, particularly under the acidification treatments, affecting photosynthetic efficiency. Microbial communities differed between blistered and healthy kelp for all treatments, except for those under future conditions of warming and acidification, which after two weeks resembled assemblages associated with healthy hosts. This indicates that changes in the microbiome were not easily predictable as the severity of future climate scenarios increased. Future ocean conditions can change kelp microbiomes and may lead to host disease, with potentially cascading impacts on associated ecosystems.

Sparse matrix--matrix multiplication (SpGEMM) is a key operation in numerous areas from information to the physical sciences. Implementing SpGEMM efficiently on throughput-oriented processors, such as the graphics processing unit (GPU), requires the programmer to expose substantial fine-grained parallelism while conserving the limited off-chip memory bandwidth. Balancing these concerns, we decompose the SpGEMM operation into three highly parallel phases: expansion, sorting, and contraction, and introduce a set of complementary bandwidth-saving performance optimizations. Our implementation is fully general and our optimization strategy adaptively processes the SpGEMM workload row-wise to substantially improve performance by decreasing the work complexity and utilizing the memory hierarchy more effectively.

Environmental anomalies that trigger adverse physiological responses and mortality are occurring with increasing frequency due to climate change. At species' range peripheries, environmental anomalies are particularly concerning because species often exist at their environmental tolerance limits and may not be able to migrate to escape unfavourable conditions. Here, we investigated the bleaching response and mortality of 14 coral genera across high‐latitude eastern Australia during a global heat stress event in 2016. We evaluated whether the severity of assemblage‐scale and genus‐level bleaching responses was associated with cumulative heat stress and/or local environmental history, including long‐term mean temperatures during the hottest month of each year (SSTLTMAX), and annual fluctuations in water temperature (SSTVAR) and solar irradiance (PARZVAR). The most severely‐bleached genera included species that were either endemic to the region (Pocillopora aliciae) or rare in the tropics (e.g. Porites heronensis). Pocillopora spp., in particular, showed high rates of immediate mortality. Bleaching severity of Pocillopora was high where SSTLTMAX was low or PARZVAR was high, whereas bleaching severity of Porites was directly associated with cumulative heat stress. While many tropical Acropora species are extremely vulnerable to bleaching, the Acropora species common at high latitudes, such as A. glauca and A. solitaryensis, showed little incidence of bleaching and immediate mortality. Two other regionally‐abundant genera, Goniastrea and Turbinaria, were also largely unaffected by the thermal anomaly. The severity of assemblage‐scale bleaching responses was poorly explained by the environmental parameters we examined. Instead, the severity of assemblage‐scale bleaching was associated with local differences in species abundance and taxon‐specific bleaching responses. The marked taxonomic disparity in bleaching severity, coupled with high mortality of high‐latitude endemics, point to climate‐driven simplification of assemblage structures and progressive homogenisation of reef functions at these high‐latitude locations.

Extensive coral bleaching on the world's southernmost coral reef at Lord Howe Island, Australia The world's southernmost fringing coral reef and extensive high-latitude coral and reef assemblages occur at Lord Howe Island (LHI) (31°33¢S, 159°05¢E) (Harriott et al. 1995). More than 80 scleractinian species have been recorded from LHI reefs, and these corals dominate much of the reef benthos (Harriott et al. 1995; Harrison 2008). The first widespread coral bleaching event recorded at LHI occurred during the 1998 austral summer season when sea temperatures increased above 27°C (P. Harrison pers. obs.), but the bleaching had limited detectable impact on coral cover. During the 2010 summer season, sea temperatures around LHI were abnormally high and exceeded 28°C (~2-3°C above normal summer maximum), with an accumulated thermal stress of more than 19 degree heating weeks (http://coralreefwatch.noaa.gov). This thermal stress coincided with calm seas and high light penetration, resulting in the most extensive and severe coral bleaching event recorded at LHI to date (Fig. 1). Bleached and partially bleached coral cover exceeded 90% at Sylph's Hole and Comet's Hole in the lagoon during March 2010, with less extensive and patchy bleaching at other reef sites around LHI. Pocilloporid corals (Stylophora, Pocillopora and Seriatopora) and Montipora spp. bleached more extensively than other corals, with some Porites, Isopora and other acroporid and faviid colonies, and host sea anemones, observed with substantial or partial pigmentation loss at some sites. Some bleaching-related coral mortality was evident during March 2010, with up to 25% of corals at Comet's Hole having partial or complete bleaching-induced mortality. Rising sea temperatures are predicted to induce more frequent coral bleaching events in future, leading to range shifts in reef corals to higher-latitude regions (Greenstein and Pandolfi 2008). However, this severe coral bleaching event at LHI demonstrates that even the highest latitude coral reef assemblages are also susceptible to bleaching stressors, which could limit future reef development and predicted range shifts to higher latitudes. Isolated reefs such as those at LHI, which lie more than 1,000 km south of the Great Barrier Reef, are likely to be slower to recover from severe disturbances due to their geographic and genetic isolation from other reefs that could potentially supply allochthonous coral larvae for recruitment (Harrison 2008).

Since 2000, a disease displaying white-syndrome characteristics has been observed affecting corals from the genus Turbinaria in the Solitary Islands Marine Park, New South Wales, Australia. Recently termed Australian subtropical white syndrome, this disease is transmissible through direct contact and by a predatory vector, but transmission through the water column has not been observed. In aquarium experiments, progressive tissue loss, extending from the region where healthy Turbinaria mesenterina fragments were in direct contact with samples of diseased coral, was noted in 66% of treatments. No tissue loss occurred in any of the controls or when healthy fragments were not in direct contact with diseased corals. Field experiments confirmed that the disease was infectious through direct contact. Further experiments showed that the rate of tissue loss was significantly higher when corals were exposed to summer temperatures (26 • C). These results suggest that temperature increases predicted in most climate change models could lead to the loss of dominant coral species, displacing other organisms that rely on corals for food and shelter. Finally, the present study showed that removal of the disease margin provides a management tool to minimise coral tissue loss during an epizootic.