Low oxygen levels caused by Noctiluca scintillans bloom kills corals in Gulf of Mannar, India
Though the time and place of the origin of this bloom is unknown, the presumable causes of it were high temperatures, abundant nutrients, low tidal amplitude, and little current. According to fishermen, these bioluminescent blooms were first seen about 15 nautical miles offshore of the Mandapam coast between India and Sri Lanka on 6th September, and subsequently moved towards the shore (Fig. 2). Bloom of N. scintillans in 2008 was reported to affect all the marine organisms including corals in GoM12. On 14th September, our preliminary assessment revealed that corals in Shingle and Krusadai islands were possibly affected by the bloom. A great multitude of N. scintillans cells were found settled on corals and other benthic organisms in the affected areas. A greenish settlement was observable on live coral colonies and other benthic organisms including macro algae, coralline algae and sponges etc.(Fig. S2). Settling of N. scintillans on benthic organisms has been reported to cause significant damage to the reef organisms through asphyxiation12. At Shingle Island, the area of significant impact was about 8.1 hectares on the shoreward side of the Island (79°14′14.38″E, 9°14′44.23″N) at depths between 1 and 3 m (Fig. 3). At Krusadai Island, an area of 2.1 hectares in the shoreward side was found affected by the bloom (79°13′20.78″E, 9°15′00.88″N) at depths between 1 and 2 m. The rest of the reef areas in both of these islands were healthy without any impact. The settled cells of N. scintillans were found to be washed ashore during subsequent surveys. In addition to dead fishes, a multitude of benthic communities such as crustaceans, mollusks and echinoderms were also found dead on the bottom in the impacted areas. Surveys between 15 and 18th September 2019 confirmed that corals in other islands (Pullivasal, Poomarichan, Manoliputti, Manoli and Hare) were in good health, and without any noticeable impact due to the bloom. Shingle and Krusadai islands occur closest to the mainland, and the concentrated bloom appeared to get trapped by currents between the mainland shore and islands.
Figure 2
(a) Green tide of Noctiluca scintillans in the Gulf of Mannar; (b) image of N.scintillans cells; size of the grid is 1 mm2 (N. scintillans exhibits bioluminescence when disturbed).
Full size image
Figure 3
Map showing the affected islands in the Mandapam group shown in Fig. 1. Base map was prepared by digitizing the georeferred Toposheet of Survey of India (http://www.surveyofindia.gov.in/) and field data using Open source GIS software (QGIS 3.10.6; https://qgis.org/en/site/forusers/download.html).
Full size image
On 14th September, coral mortality was not observed in the affected areas though the colonies were observed to be disturbed by the settling N. scintillans cells. Low dissolved oxygen levels have been reported to be the primary cause of benthic mortality during algal blooms22. Dissolved oxygen levels were 1.48 mg l−1 at Shingle Island and 2.02 mg l−1 at Krusadai Island in the affected areas. This compares to ‘normal’ levels for coral reefs of 5–8 mg l−1, and Haas et al.11 found that dissolved oxygen content less than 4 mg l−1 is detrimental to acroporid corals. Moreover, branching coral forms have been reported to be more susceptible to hypoxic episodes than spherical or massive forms5. Corals are routinely exposed to fluctuations in oxygen levels at the tissue level due to photosynthesis and respiration processes of endosymbionts7, but are negatively impacted when (sub-) lethal thresholds of hypoxia exposure are exceeded1,5,11. Lethal hypoxia thresholds appear to differ considerably between coral species, ranging between 0.5 and 4 mg O2 l−11,5,11, while sub-lethal hypoxia thresholds for corals are almost entirely unknown5.
Seawater temperature can significantly impact dissolved oxygen levels23,24. Water temperature was 29.9 and 29.8º C (Table 1) at Shingle and Krusadai islands respectively and these levels are marginally higher than the normal levels for this particular time of the year. Apart from the summer months (April to June), temperature levels in GoM do not go higher than 29º C20. The concentration of N. scintillans was 43.4 × 105 and 27.3 × 105 cells l−1 at Shingle and Krusadai Islands respectively; pH and TDS were also high in the affected area (Table 1). Dissolved oxygen levels in other sites of these two islands and in other five islands were higher than 5 mg l−1.
Table 1 Environmental characterization at the affected sites in Shingle and Krusadai Islands.
Full size table
During the next assessment on 17th of September 2019, severe coral mortality was observed at the affected sites. At Shingle Island, overall coral colony density was 134.25 (SE ± 3.28) no.100 m−2 (n = 537) within ten 20 m belt transects which is dominated by Acropora (64%) followed by Montipora (15%). Out of total 537 colonies, 33.52% (n = 180) were found dead (Fig. 4), which include 34.5 (SE ± 1.05) no.100 m−2 (n = 138) of Acropora, 7.75 (SE ± 0.75) no.100 m−2 (n = 31) of Montipora and 2.75 (SE ± 0.35) no.100 m−2 (n = 11) of Pocillopora. The death of coral colonies was so rapid that the coral tissue was intact on the colony surface and still had its natural colour (Fig. 5). When wafted with water by hand or with scuba air, the tissue peeled off exposing the skeleton (Supplementary video). Other observed genera such as Dipsastraea, Favites, Porites, Hydnophora, Goniastrea, Echinopora, Turbinaria, Platygyra, Goniopora and Symphyllia in the same site were all alive (Fig. S3), though with excess mucus production. This may be explained by differential lethal thresholds for oxygen levels at species and growth form levels5,19. At Krusadai Island, the overall coral density on 17th September was 66 (SE ± 2.54) no.100 m−2 (n = 132), dominated by Acropora. Among the counted colonies, 6 (SE ± 1.03) no.100 m−2 of Acropora were found recently dead while mortality was not observed in other available genera such as Montipora, Pocillopora, Dipsastraea, Favites, Porites and Turbinaria. Dissolved oxygen levels had increased to 3.78 mg l−1 at Shingle Island and to 4.02 mg l−1 at Krusadai Island at the affected sites and the water had started to become clear. The concentration of N. scintillans had reduced to 1.63 × 103 cells l−1 and 0.88 × 103 cells l−1 at Shingle and Krusadai Islands, respectively (Table 1).
Figure 4
Density of live and dead colonies of affected coral genera (Acropora, Montipora and Pocillopora) in Shingle Island, by date; the green line indicates the drastic decline of Acropora density between 17.09.2019 and 27.09.2019.
Full size image
Figure 5
Rapid mortality of corals presumably due to low oxygen levels caused by Noctiluca scintillans; (a, b) Acropora; (c) Montipora; (d) Pocillopora.
Full size image
Assessment on 27th September 2019 at the impacted area in Shingle Island, showed that the overall density of coral colonies within ten 20 m transects was 135.75 (SE ± 2.82) no.100 m−2 (n = 543) and of them 70.35% (n = 382) of colonies belonging to Acropora, Montipora and Pocillopora were found dead revealing that the impact of algal bloom was more severe than expected (Fig. 4). It was almost two weeks since the corals had died and hence secondary algae had started colonizing the dead colonies. On the same day at the impacted area of Krusadai Island, overall coral density within five belt transects was 65.5 (SE ± 1.83) no.100 m−2 (n = 131), of which 9.09% (n = 12) of colonies belonging to Acropora were found dead. By 27th September, dissolved oxygen levels had increased to 6.02 and 5.73 mg l−1 respectively at the affected areas of Shingle and Krusadai islands (Table 1). N. scintillans cells were absent in all the sites indicating the end of bloom. On 04th October 2019, the overall coral colony density within 20 m belt transects was 138 (SE ± 2.08) no.100 m−2 (n = 552) and of them 71.23% (n = 393) colonies belonging to Acropora, Montipora and Pocillopora were found dead at the area of impact in Shingle Island (Fig. 4). No further mortality was witnessed in the affected area of Krusadai Island. Secondary algae have completely overgrown the dead coral colonies making the reef look green (Fig. S4). Dissolved oxygen levels were reasonably high at 7.13 and 7.24 mg l−1 respectively at Shingle and Krusadai Islands during this time (Table 1).
Coral mortality due to algal bloom and consequent hypoxia has rarely been reported12,13,25. The present study reports that the impact of blooms can be severe on corals. Different coral species respond differently to low oxygen levels according to their respiration and photosynthesis5,26. Thus, low oxygen levels can orchestrate the coral mortality by affecting coral’s productivity and respiration7. Further, fast growing corals such as Acropora and Pocillopora have been reported to be more susceptible to low oxygen levels11,13,27. Fast growing coral species have faster metabolism rates28 and hence metabolic oxygen requirements are higher11,29. Thus, the mortality of fast growing species in the present study was presumably due to the low oxygen levels induced by N.scintillans bloom.
Bleaching episodes in 2010 and 2016 had also caused significant mortality to these fast growing species in GoM19,20. Corals in GoM start to bleach when water temperature exceeds 30º C and the temperature levels during this bloom period ranged between 28.4 and 29.9º C. Though bleaching was not observed, heat stress might also have played its role in coral mortality along with low oxygen levels as the temperature level almost reached 30º C. Similar temperature levels were reported during the bloom of N. scintillans in 2008 in GoM12.
Corals in Gulf of Mannnar are still recovering from the 2016 bleaching episode20 and hence the present decline is significant. Phase shifts on coral reefs are predominantly associated with shifts from hard coral-dominated communities to macroalgae-dominated ones30. Space competition between corals and other organisms such as algae and sponges has been reported to negatively impact the corals of GoM after the 2016 bleaching event20,31. At present, secondary algae have completely occupied the dead coral colonies, which will affect the coral recovery by hindering the attachment of new coral recruits during the next spawning season32. Recent studies suggest hypoxia increases coral susceptibility to bleaching27, and may increase disease prevalence and algal proliferation7. Thus algal blooms add to the existing array of threats to corals of GoM that needs to be understood more with further focused research.
On account of the problems related to climate change, there has been a steady and severe decline of coral reefs in the past two decades. Bleaching and diseases have been reported to cause mass coral mortalities within a very short time. The observations of the present study alert us to possible mass mortality due to short-term hypoxic condition caused by algal blooms. Algal blooms and hypoxic conditions are predicted to occur more frequently in the future due to climate change14. Hence, it is likely that shallow water coral reefs will be affected more frequently by temporary low oxygen levels caused by algal blooms. More studies are, however, required to understand the mechanism of coral mortality due to algal blooms, impacts on community composition and the potential for subsequent recovery. More
