Atmos. Chem. Phys., 21, 5883–5903, 2021 https://doi.org/10.5194/acp-21-5883-2021 © Author(s) 2021. This work is distributed under the Creative Commons Attribution 4.0 License. Coral-reef-derived dimethyl sulfide and the climatic impact of the loss of coral reefs Sonya L. Fiddes1,2,3,a, Matthew T. Woodhouse3, Todd P. Lane4, and Robyn Schofield4 1Australian-German Climate and Energy College, University of Melbourne, Parkville, Australia 2ARC Centre of Excellence for Climate System Science, School of Earth Sciences, University of Melbourne, Parkville, Australia 3Climate Science Centre, Oceans and Atmosphere, Commonwealth Scientific and Industrial Research Organisation, Aspendale, Australia 4ARC Centre of Excellence for Climate Extremes, School of Earth Sciences, University of Melbourne, Parkville, Australia anow at: Australian Antarctic Program Partnership, Institute of Marine and Antarctic Studies, University of Tasmania, Hobart, Australia Correspondence: Sonya L. Fiddes (sonya.fi[email protected]) Received: 9 October 2020 – Discussion started: 27 October 2020 Revised: 3 March 2021 – Accepted: 17 March 2021 – Published: 20 April 2021 Abstract. Dimethyl sulfide (DMS) is a naturally occurring sponses are found to have no robust effect on regional climate aerosol precursor gas which plays an important role in the via direct and indirect aerosol effects. This work emphasises global sulfur budget, aerosol formation and climate. While the complexities of the aerosol–climate system, and the limi- DMS is produced predominantly by phytoplankton, recent tations of current modelling capabilities are highlighted, in observational literature has suggested that corals and their particular surrounding convective responses to changes in symbionts produce a comparable amount of DMS, which aerosol. In conclusion, we find no robust evidence that coral- is unaccounted for in models. It has further been hypothe- reef-derived DMS influences global and regional climate. sised that the coral reef source of DMS may modulate re- gional climate. This hypothesis presents a particular concern given the current threat to coral reefs under anthropogenic climate change. In this paper, a global climate model with 1 Introduction online chemistry and aerosol is used to explore the influence of coral-reef-derived DMS on atmospheric composition and Marine organisms (phytoplankton, algae) are known to pro- climate. A simple representation of coral-reef-derived DMS duce the chemical dimethyl sulfoniopropionate (DMSP). In is developed and added to a common DMS surface water cli- the ocean, DMSP experiences enzymatic cleavage, forming matology, resulting in an additional flux of 0.3 Tg yr−1 S, or dimethyl sulfide (DMS; Yoch, 2002; see Fig.1, point 1), 1.7 % of the global sulfur flux from DMS. By comparing the which can then be released into the atmosphere (2). At- differences between both nudged and free-running ensem- mospheric DMS (DMSa) can undergo a series of chemical ble simulations with and without coral-reef-derived DMS, reactions to become a sulfate aerosol (3). When in suffi- the influence of coral-reef-derived DMS on regional climate ciently large abundance (4), these sulfate aerosols can im- is quantified. In the Maritime Continent–Australian region, pact aerosol loading and cloud properties, altering the ra- where the highest density of coral reefs exists, a small de- diation budget directly (5) and indirectly (6) and having a crease in nucleation- and Aitken-mode aerosol number con- cooling effect (7). This effect has been hypothesised to form centration and mass is found when coral reef DMS emis- a short-term bioregulatory negative feedback system, known sions are removed from the system. However, these small re- as the CLAW (Charlson, Lovelock, Andreae and Warren) hy- pothesis (Charlson et al., 1987), whereby marine organisms Published by Copernicus Publications on behalf of the European Geosciences Union. 5884 S. L. Fiddes et al.: Influence of coral reefs on climate can alter their environment when stressed. This hypothesis (Swan et al., 2012; Fischer and Jones, 2012; Hopkins et al., remains unproven, and arguments against it cite the com- 2016; Swan et al., 2017). Of interest to this study are the plexity and non-linearity of the DMS–climate system (Quinn findings from Hopkins et al.(2016), where the effect of tidal and Bates, 2011; Green and Hatton, 2014). Nevertheless, at exposure on three Indo-Pacific coral species was studied in longer timescales, global modelling studies have shown that laboratory experiments. From their results, Hopkins et al. −2 −1 marine-derived DMS plays an important role in maintaining (2016) extrapolate a fluxDMS of 9–35 µmol m d over the current large-scale climate (Thomas et al., 2010; Wood- coral reefs from Acropora cf. horrida, while an additional 5 house et al., 2010; Gabric et al., 2013; Mahajan et al., 2015), and 8 µmol m−2 d−1 can be estimated from two other species providing global cooling via direct and indirect aerosol ef- in their experiments (P. cylindrica and S. hystrix.). These es- fects of up to 0.45 ◦C(Fiddes et al., 2018) when compared timates are equivalent to a total of 709–1548 µg m−2 d−1 of to a world in which no marine DMS exists. Our previous sulfur (if all species are present) and in this work is further paper (Fiddes et al., 2018) describes these studies, DMS sur- extrapolated to global coral reef coverage (approximately face water climatologies and flux parameterisations in more 284 300 km2), giving 0.074–0.16 Tg yr−1 of sulfur. Whilst detail. these extrapolations are highly speculative in terms of arti- Many global DMS–climate modelling studies have also ficial laboratory experiments, estimated exposure time and considered DMS under future scenarios (Bopp et al., 2004; coverage of coral reefs, and they account for just three Indo- Gabric et al., 2004; Kloster et al., 2007; Cameron-Smith Pacific species of coral and only DMS produced during tidal et al., 2011; Six et al., 2013; Grandey and Wang, 2015; stress, the Hopkins et al.(2016) estimations were the first to Schwinger et al., 2017). However, considering our under- attempt to quantify the large-scale flux of coral-reef-derived standing of DMS in the current climate remains uncertain, DMS. the aforementioned studies do not provide a clear consen- Following these observed results, numerous studies have sus on how DMS production may respond to a warming cli- made links to coral DMS, aerosol formation, cloud cover mate. With this in mind, better knowledge of current sources and/or sea surface temperatures (SSTs) (Modini et al., 2009; of DMS is important to further our understanding of DMS– Deschaseaux et al., 2012; Leahy et al., 2013; Swan et al., climate interaction now and into the future. 2017; Jones et al., 2017; Cropp et al., 2018; Jackson et al., One such source of DMS that is currently unaccounted 2018, 2020b). Jones(2013), Jones et al.(2017) and Cropp for in climate modelling is coral reefs. Recent studies have et al.(2018) further suggest that coral reefs participate in shown that corals, coral symbionts, and coral by-products bioregulatory feedback as suggested by the CLAW hypothe- (e.g. mucus) produce large amounts of DMSP (Broad- ses. Most of these studies do not explicitly account for the bent et al., 2002; Broadbent and Jones, 2004; Jones and complexity of the DMS–climate system and its significant Trevena, 2005; Jones et al., 2007; Burdett et al., 2015; non-linearities (see Thomas et al., 2011; Quinn and Bates, Jackson et al., 2020a). Of note for this work, Jones et al. 2011; Green and Hatton, 2014; Fiddes et al., 2018). In addi- (2018) have summarised reports of fluxDMS values of 0– tion, deducing a climatic impact of one aerosol species using 4906 µg m−2 d−1 and a mean of 205 µg m−2 d−1 in summer observations alone is fraught with co-varying and confound- and 0.6–481 µg m−2 d−1 with a mean of 77 µg m−2 d−1 over ing influences from other aerosol species. These complexities winter over the Great Barrier Reef (GBR). Jones et al.(2018) can only be addressed through modelling studies; however also suggest that total emissions from the GBR are equivalent no modelling study has included coral-reef-derived DMS as to 0.02 Tg yr−1 of sulfur, noting that total global sulfur flux a source of sulfur to date. from DMS is estimated to be between 9–35 Tg yr−1 (Belviso To add urgency to this problem, coral reef ecosystems et al., 2004; Elliott, 2009; Woodhouse et al., 2010; Tesdal globally are facing dire risk due to anthropogenic climate et al., 2016; Fiddes et al., 2018) and that DMS makes up change (Hughes et al., 2017, 2018). The IPCC special re- approximately one-fifth of the global sulfur budget (Sheng port on climate change (IPCC, 2018) states that under 1.5 ◦C et al., 2015). The Jones et al.(2018) coral reef flux estimation warming, 70 %–90 % of coral reefs will be extinct. The risk has been made from measurements both over coral reefs and to coral reefs is twofold; increasing sea surface tempera- in the GBR lagoon and also includes an estimate of the addi- tures are causing more frequent mass coral bleaching events tional flux from tropical cyclones. The tropical cyclone emis- (Hughes et al., 2017; King et al., 2017), whilst increas- sion has been calculated (not observed) using the Liss and ing ocean acidification is causing reduced calcification and Merlivat(1986) flux parameterisation, taking into account growth of coral species (Hoegh-Guldberg et al., 2017; Mag- average wind speeds of tropical cyclones and accounting for nan et al., 2016). Whilst the death of global coral reefs due approximately five cyclone days per year in the region.