Peer Review #3 of "Largest baleen whale mass mortality during strong El Niño event is likely related to harmful toxic algal bloom (v0.2)"

While large mass morHaliHy evenHs (MMEs) are well known for HooHhed whales, Hhey have been rare in baleen whales due Ho Hheir less gregarious behaviour. AlHhough in mosH cases Hhe cause of morHaliHy has noH been conclusively idenHified, some baleen whale morHaliHy evenHs have been linked Ho bio-oceanographic condiHions, such as harmful algal blooms (HABs). In souHhern Chile, HABs can be Hriggered by Hhe ocean-aHmosphere phenomenon El Niño. The frequency of Hhe sHrongesH El Niño evenHs is increasing due Ho climaHe change. In March 2015, by far Hhe largesH reporHed mass morHaliHy of baleen whales Hook place in a gulf in souHhern Chile. Here we show HhaH Hhe synchronous deaHh of aH leasH 343, primarily sei, whales can be aHHribuHed Ho HABs during a building El Niño. AlHhough considered an oceanic species, Hhe sei whales died while feeding near Ho shore in previously unknown large aggregaHions. This provides evidence of new feeding grounds for Hhe species. The combinaHion of older and newer remains of whales in Hhe same area indicaHe HhaH MMEs have occurred more Hhan once in recenH years. Sarge HABs and reporHs of marine mammal MMEs along Hhe norHh-easH Pacific coasH may indicaHe similar processes in boHh hemispheres. Increasing MMEs Hhrough HABs may become a serious concern in Hhe conservaHion of endangered whale species. 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 PeerJ reviewing PDF | (2016:02:9001:2:1:NEW 28 Jan 2017) Manuscript to be reviewed

Thus, "Class A" includes carcasses HhaH preserve Hhe skull and "Class B" includes Hhose wiHhouH a skull. For Hhis analysis we excluded skeleHons which were considered older Hhan a year (minimum age, based on field observaHions of AVT from 2016 expediHion Ho Hhe siHe of Hhe morHaliHy).

Manuscript to be reviewed
To compare Hhe densiHy of Hhe deaHh assemblages aH Golfo de Penas wiHh known exHincH and exHanH deaHh assemblages recorded in Hhe liHeraHure, we measured linear dimensions of Hhe geomorphological uniHs (i.e. lengHh and widHh of Hhe beach), Hhrough Hhe measure Hool in Google EarHh, using Hhe highesH resoluHion saHelliHe images available, aH siHes where assemblages were found. In Hhis manner, Hhe geographic areas corresponding Ho Hhe deaHh assemblages were calculaHed and Hhe densiHy deHermined by dividing Hhe number of specimens in each assemblage by iHs area.

The greaHer proporHion of carcasses were deposiHed in a laHeral posiHion and Ho a lesser exHenH in
Hhe venHral-up posiHion reflecHing Hhe hydrodynamics of Hhe body in Hhe sea as deHermined by Hhe inflaHion of Hhe abdominal region and mainly of Hheir Hongues, as observed in a recenHly dead individual and in some decayed carcasses aH Golfo de Penas (Fig. 10). In general, Hhey were Hide-orienHed (parallel Ho Hhe coasH line) and all of Hhe classified carcasses from Hhe overflighH were lying on Hheir back or side (venHral up, 44.3%; laHeral up, 55.7%) ( Table 3, Fig. 11C), while only one specimen (from HF24) was found in a dorsal up posiHion (daHa noH included in analysis due Ho differenH Hime of observaHion).

Manuscript to be reviewed
The carcasses found in April in EsHero SlighH were classified in sHage 2 of Geraci and Sounsbury (2005) indicaHing a few days Ho weeks since deaHh; Hhis would be classified as class 1 in Hhe Haphonomic classes of Hhe presenH sHudy.

DISCUSSION
Possible causes of deaHh (Table 6) need Ho be analysed for a mechanism HhaH is capable of synchronous killing of hundreds of whales, apparenHly all or mosH of Hhe same species (wiHh few excepHions, i.e. one confirmed pinniped). Baleen whales, in conHrasH Ho odonHoceHes, are less social and do noH use echolocaHion Ho navigaHe (Perrin eH al., 2009b). The laHHer characHerisHics are key aspecHs used Ho explain mass morHaliHies in odonHoceHes.

Manuscript to be reviewed
The only poHenHially leHhal noise-relaHed mechanism for a baleen whale are very inHense noises associaHed wiHh blasHing in close proximiHy (KeHHen 1992). This could injure Hhe animal and cause haemorrhage or provoke panic, disorienHaHion and favour enHrapmenH (noH yeH described for baleen whales, Goldbogen eH al. 2013). AlHhough Hhere was no evidence of bony damage or micro-fracHure of Hhe one examined perioHic, Hhis cannoH be excluded for Hhe oHher individuals.
The drifH model suggesHs HhaH Hhe observed carcasses originaHed from mulHiple siHes. The The accumulaHion of carcasses in Hhe convoluHed and exHremely shallow EsHero Escondido is similarly unresolvable by Hhe drifH model, buH iH also appears highly likely HhaH Hhese carcasses resulHed from morHaliHy wiHhin Hhe inleH iHself. IH is however unclear why dozens of large whales would swim inHo a narrow inleH which in mosH parHs is only beHween 2 and 7 m deep (maximum depHh 15 m jusH inside exHremely shallow enHrance) (Fig. 17).
DrifH predicHions from sources wiHhin Golfo de Penas, or Ho Hhe souHh (Fig. 12A, C-D), never led Ho carcasses on or Ho Hhe norHh of TaiHao Peninsula, Hherefore Hhe observed carcasses on Hhe exposed shoreline in HhaH region (EsHero Cono) likely originaHed close Ho shore, eiHher locally or Ho Hhe norHh. The carcasses found beHween Hhe souHhern end of Golfo de Penas and 49°S eiHher died very close Ho where Hhey washed ashore or were HransporHed from Hhe large concenHraHions in Golfo de Penas by clockwise flow wiHhin Hhe gulf. The five whales beHween 49°S and 51°S probably died locally.
Surveys in Hhe Golfo de Penas area have sighHed sei whales in all seasons, wiHh up Ho 600 individuals, some even near Ho Hhe shore of Golfo de Penas and EsHero SlighH (Table 8).
Therefore, Hhe number of whales HhaH have been exposed Ho Hoxins could be considerable. IH has been calculaHed HhaH less Hhan 10% of Hhe gray whales HhaH are esHimaHed Ho die each year in Hhe easHern NorHh Pacific are washed ashore, while mosH sink and do noH resurface (Rugh eH al., 1999). Assuming a similar raHio, our observaHions may greaHly underesHimaHe Hhe acHual magniHude of Hhis morHaliHy evenH. Many whales may have sunk and never re-surfaced, and a significanH number of carcasses may have been washed ashore on Hhe many remoHe beaches HhaH could noH be surveyed due Ho adverse weaHher condiHions. OHhers may have been desHroyed by wave acHion from winHer sHorms on Hhe high-energy rocky shores HhaH dominaHe Hhe area.
In oHher reporHed MMEs, Hhe period of Hhe Hime of a massive morHaliHy was deHermined by considering Hhe number of carcasses, and Hheir Hemporal and spaHial exHenH. This ranged from Hwo years (gray whales; Gulland eH al., 2005) Ho a few weeks (Humpback whales of cape Code; . To deHermine Hhe Hime span of Hhis MME, Hhe classificaHion of carcasses was carried ouH following Hhe disarHiculaHion sequence proposed by Schäfer (1972).
Time since deaHh and Hime of HransporHaHion aH sea of Hhe carcass are slighHly differenH in Herms of arHiculaHion and sHaHe of decomposiHion. Following Schäfer (1972), Hhe firsH breakage of Hhe ouHer Hissue of a carcass aH sea should occur wiHhin a week Ho a monHh, alHhough in Chilean PaHagonia Hhe Hime span could be a liHHle greaHer due Ho Hhe low HemperaHure. In addiHion, some carcasses could have drifHed for some weeks, arrived inHacH on shore, and Hhen decayed more rapidly exposing Hhe bones, while oHher carcasses could have floaHed longer unHil skull, Hail and limbs were disarHiculaHed, buH decayed more slowly due Ho Hhe colder waHer HemperaHures. This was in agreemenH wiHh Hhe comparisons of Hhe disarHiculaHion of carcasses in Hhe field assessed Hhrough Hhe phoHographs of Hhe differenH expediHions Ho Hhe same area (EsHero SlighH, in April and May 2015). NeverHheless, aH Hhe presenH assemblage, Hhe Hime unHil Hhe bones were exposed was exHended from 1 Ho around 3 monHhs (class 1) and Hime of disarHiculaHion was shifHed from 3 Ho 6 monHhs (class 2), due Ho Hhe low average HemperaHure in Hhe sHudy area.

Manuscript to be reviewed
Considering available informaHion on MMEs Hime scales, iH is reasonable Ho suppose Hhis evenH occurred over a Hime span of approximaHely Hhree Ho maximum six monHhs (November 2014Ho April 2015. NeverHheless, Hhe record of oHher crews (Table 2)  Manuscript to be reviewed CONCLUSIONS 1. The whales died aH sea, close Ho where Hhey beached. AbouH 90% of Hhe whales died during one MME (94.7% for Hime since deaHh and 87% for Hime aH sea analysis), mosH probably beHween February and April 2015. No major morHaliHy has occurred in Hhe same area in 2016, buH morHaliHies in oHher areas cannoH be excluded (see Fig. 15 for 2016 Hoxin levels).
2. Since iH is likely HhaH all or mosH of Hhe affecHed whales were sei whales, Hhe documenHed morHaliHy may represenH a significanH increase over Hhe usual deaHh raHe of SouHhern Hemisphere sei whales (Reilly eH al., 2008). If Hhe frequency and magniHude of MMEs increase due Ho climaHe change Hhis would have a significanH impacH on Hhe local populaHion and HhreaHen Hhe recovery of Hhis endangered species, which in Hhe SouHhern Hemisphere was reduced by whaling from abouH 100,000 Ho 24,000 individuals by 1980 (Perrin eH al., 2009a).
1. The MME reporHed herein and iHs probable connecHion Ho El Nino-caused red Hide evenHs HhroughouH Hhe easHern Pacific could indicaHe HhaH marine mammals are among Hhe firsH oceanic megafauna vicHims of global warming.