Worldwide Distribution and Depth Limits Of

Journal of Crustacean Biology Advance Access published 4 February 2021 Journal of Crustacean Biology The Crustacean Society Journal of Crustacean Biology (2021) 1–13. doi:10.1093/jcbiol/ruaa102 Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Worldwide distribution and depth limits of decapod crustaceans (Penaeoidea, Oplophoroidea) across the abyssal-hadal transition zone of eleven subduction trenches and five additional deep-sea features

Jackson A. Swan1, Alan J. Jamieson1, , Thomas D. Linley1, and Paul H. Yancey2 1School of Natural and Environmental Sciences, Newcastle University, Newcastle Upon Tyne, NE1 7RU, UK, and 2Biology Department, Whitman College, Walla Walla, WA 99362, U.S.A

Correspondence: Alan J Jamieson; e-mail: [email protected] (Received 6 November 2020; accepted 4 January 2021)

ABSTRACT Decapod crustaceans are conspicuous members of marine benthic communities to at least 7,700 m deep. To assess the bathymetric extent of this taxonomic group, baited landers were deployed to across the abyssal-hadal transition zone of 11 subduction trenches spanning the Pacific, Atlantic, Southern, and Indian oceans and additional sites. Decapods were dominated by penaeid shrimps (superfamily Penaeoidea), in particular Benthesicymus Spence Bate, 1881 and Cerataspis Gray, 1828, with the former being found deeper. Benthesicymus cf. crenatus Spence Bate, 1881 was observed in the Kermadec, Mariana, New Hebrides, Puerto Rico, Peru-Chile, Tonga, San Cristobal, and Santa Cruz trenches, plus the South Fiji Basin and the Wallaby- Zenith Fracture Zone. They were not recorded in the Abaco Canyon, Agulhas Fracture Zone, Java Trench, or any of the polar locations. Cerataspis cf. monstrosus Gray, 1828 was present in the Kermadec, Mariana, New Hebrides, Puerto Rico, and Java trenches, the Abaco Canyon, Agulhas Fracture Zone, Wallaby-Zenith Fracture Zone and the South Fiji Basin, but absent from the Tonga, San Cristobal and Santa Cruz trenches. Hymenopenaeus nereus (Faxon, 1893) was only recorded in the Peru-Chile Trench. Unidentified species belonging to superfamily Oplophoroidea were observed to a maximum depth of 6,931 m. Decapods are thus are primarily represented at hadal depths by penaeoid shrimps, consistently present at tropical and temperate latitudes to ~7,700 m, while absent from equivalent depths in polar regions. Their maximum depth may be limited due to hydrostatic pressure, while potentially affected by temperature and oxygen in some instances. Muscle samples of three specimens from 6,000 m (Mariana and Kermadec trenches) were found to have high levels of trimethylamine N-oxide (TMAO; 260 mmol kg–1), the major piezolyte, a protectant against hydrostatic pressure, in other deep-sea organisms. We speculate that physiological limits to TMAO concentration may prevent them from inhabiting the greatest hadal depths. Key Words: abyssal zone, deep-sea fauna, hadal zone, piezolytes, trimethylamine N-oxide (TMAO), subduction trenches

INTRODUCTION that shrimps belonging to superfamilies Penaeoidea (suborder Dendrobranchiata) and Oplophoroidea (infraorder Caridea) Based on the first major trawling efforts at hadal depths were present in the trenches of the Western Pacific (mainly in the 1950s (Wolff, 1960, 1970), members of Decapoda the Japan and Kermadec trenches) to at least 7,703 m and were long thought to have no representatives greater than 6,890 m, respectively. This conclusion was, however, based 5,700 m (Herring, 2002; Blankenship & Levin, 2007), des- on just eight deployments of a baited camera spanning three pite some anecdotal evidence to the contrary (Pérès, 1965; localities in which decapods were observed in the shallowest Hessler et al., 1978). Jamieson et al. (2009a) demonstrated five deployments.

The study of Jamieson et al. (2009a) was followed by sev- depth in bony fishes, though as osmoregulators hypo-osmotic to eral other studies that showed that the dominant Penaeoidea seawater, there are no ‘shallow’ osmolytes to replace (Gillett et al., was Benthesicymus crenatus Spence Bate, 1881 and a species of 1997; Samerotte et al., 2007; Yancey et al., 2014). TMAO was pos- Oplophoroidea thought to belong to genus Acanthephyra A. Milne- tulated to play a major role in pressure adaptation in all these Edwards, 1881. The former was clearly attracted to the baited taxa, protecting macromolecules from pressure perturbations. camera by the elevated abundance of lysianassoid amphipods Moreover, bony fishes appear to be limited to depths below 8,400

which they were observed to prey upon, while the latter appeared m, with a limit to the impact of TMAO on osmoregulatory physi- Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021 to be incidental observations as they showed no obvious interest ology possibly creating that limit (Yancey et al., 2014, Linley et al., in the bait or associated fauna (Jamieson et al., 2011, 2012, 2013; 2016; Downing et al., 2018). Linley et al., 2017). The last ten years have seen an increase in biological observations at hadal depths (Jamieson, 2018), yet megafaunal studies tend MATERIALS AND METHODS to focus on either amphipods (e.g., Fujii et al., 2013; Ritchie et al., 2015; Lacey et al., 2016) or fishes (e.g., Linley et al., 2016, 2017). Sampling effort The technological limitations of working at such extreme depths has resulted in a lack of exploratory vehicles rated to 11,000 m Baited camera landers were deployed in every ocean at a number and the near-cessation of bottom trawling at hadal depths since of different geomorphological features (Table 1, Fig. 1). In the the 1950s. As a result, free-fall baited camera and trap landers North Pacific, the primary site was the Mariana Trench, where have become the dominant methodology at hadal depths. The 37 deployments were made between 4,040 and 10,925 m over bias towards fish and amphipod studies mirrors the ease in which three expeditions in 2014, 2017, and 2019. The primary site in these groups are imaged and captured using baited traps and cam- the southwestern Pacific was the Kermadec Trench, where 44 de- eras, respectively. In the pursuit of these more robust quantitative ployments were made between 3,946 and 9,281 m spanning five studies, however, there are several other taxonomic groups that expeditions between 2009 and 2014. Six deployments were made frequent the scientific vehicles that are often cryptic (e.g., Isopoda; between 6,848 and 10,823 m in the neighbouring Tonga Trench Jamieson et al., 2012), infrequent (e.g., Cephalopoda; Jamieson in 2019. Deployments were made in 2013 and 2015 between & Vecchione, 2020) or geographically limited (e.g., Gastropoda; 4,101 and 6,948 m in the New Hebrides Trench east of New Aguzzi et al., 2012). Drawing conclusions on the maximum depth Caledonia (1,600 km north of Kermadec and Tonga trenches), and geographical expanse at such extreme depths is difficult rela- as well as another four deployments between 4,078 and 4,100 tive to amphipods (Downing et al., 2018) and fishes Yancey( et al., m on the South Fiji Basin that connects the New Hebrides and 2014). Kermadec trenches. Five deployments were made further north Decapod crustaceans represents one such overlooked taxo- in the San Cristobal and Santa Cruz trenches off the Solomon nomic group that regularly appears in baited camera datasets. Islands in 2019, in and between the two trenches from 5,906 to The presence of decapods is often limited to a small number of 7,431 m (given the proximity and overlapping sampling, these images and their diversity appears low, but they are consistent in two trenches are hereafter referred to as the Solomon trenches). appearing in images from upper hadal depths. This has meant Two expeditions (2010 and 2018) made 21 deployments between that recent studies tend only to briefly mention the Decapoda 4,051 and 8,074 m in the Peru-Chile Trench, southeastern Pacific, in passing (e.g., Jamieson et al., 2011, 2013; Linley et al., 2017; with the majority in the southern sector, often called the Atacama Jamieson & Vecchione, 2020) and their significance is perhaps Trench. In the Indian Ocean, seven deployments were made in being under-appreciated. the Wallaby-Zenith Fracture Zone (west of Australia) between In the ten years since the first unequivocal finding of hadal 4,730 and 6,537 m in 2017, and a further seven deployments be- decapods, hundreds more lander deployments have been made tween 6,146 and 7,176 m in the Java (Sunda) Trench, south of to the great depths (Jamieson, 2018). In addition to the increase in Indonesia, in 2019. In the Atlantic Ocean, 12 deployments were the number of scientific observations being made at hadal depths, made between 4,040 and 8,370 m in the Puerto Rico Trench on the recent Five Deeps Expedition (Jamieson et al., 2019) provided two expeditions (2018 and 2019), another two deployments in the a platform in which to perform these investigations across multiple Abaco Canyon off the Bahamas in 2018 (4,900 m), and one in the sites spanning every ocean in less than a year (Stewart & Jamieson, Agulhas Fracture Zone southwest of South Africa in 2019 (5,493 2019). m). In the Southern Ocean, three deployments were made in the We collated all baited-camera records from the abyssal-hadal South Shetland Trench between 2,875 and 5,200 m in 2015, and transition zone from the Five Deeps Expedition (Jamieson et al., seven deployments in the South Sandwich Trench between 6,044 2019), the HADEEP projects (Jamieson et al., 2009b), HADES-M and 8,266 m in 2019. In the Arctic Ocean, four deployments were and HADES-K projects (Linley et al., 2017), and various other ex- made between 4,128 and 5,591 m in the Molloy Hole west of peditions on several research vessels to multiple locations in every Svalbard in the Fram Strait in 2019. ocean, to provide a more comprehensive and definitive resolution to the diversity and distribution of decapods. We also measured Equipment organic osmolytes that might serve as piezolytes (solutes that pro- tect macromolecules from hydrostatic pressure effects) in muscles We used various designs of baited camera landers (Table 2). Hadal- of three captured specimens from the Kermadec and Mariana Lander A was used in 2009 (Jamieson et al., 2009c); Hadal-Lander trenches in order to explore the apparent depth limit of decapods. B between 2010 and 2011 (Jamieson et al., 2011); Hadal Lander C Marine invertebrates are osmoconformers that use high levels and the Abyssal-Lander between 2013 and the first 2017 exped- of organic osmolytes intracellularly in part to maintain osmotic ition (Linley et al., 2016). The second 2017 expedition used a series balance with seawater. In very shallow invertebrates, these are typ- of different landers based on a modified Abyssal-Lander called ically taurine, glycine, betaine, and alanine (Kelly & Yancey 1999; Gonzo (Weston et al., 2021). The first 2018 expedition used two Downing et al., 2018). Previous studies on sea anemones, decapod identical landers that were essentially compact versions of Hadal- crustaceans, squids (to 3,000 m) and amphipods, including those Lander C called Bad Spoon and Bad Ape. From 2018 to 2019 a in the oceans’ greatest depths, found that the typical ‘shallow’ fleet of three landers, Skaff, Closp, and Flere were used (Jamieson osmolytes decline while the piezolyte trimethylamine N-oxide et al., 2019). (TMAO) increases with depth (Kelly & Yancey, 1999; Yancey et al., All cameras were baited with mackerel or tuna. Temperature 2014; Downing et al., 2018). TMAO also increases linearly with and depth were recorded by the Hadal-Landers with an SBE-39

2 MAXIMUM-DEPTH LIMIT OF DEEP-SEA DECAPODS Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021 Sampling depth range (m) Sampling depth range 4,128–5,591 4,830–7,210 6,848–10,823 5,906–7,431 4,040–10,925 6,146–7,176 5,493 6,044–8,266 4,040–8,370 4,900 4,051–8,064 4,730–6,537 8,000–10,890 2,875–5,200 4,703–10,890 3,975–9,005 4,101–6,948 4,078–4,100 3,946–6,552 6,116–8,631 6,979–9,281 4,602–8,074 5,172–7,561 4 7 6 5 7 7 1 7 5 1 7 3 3 4 8 3 4 5 6 Number of stations 16 15 27 25 11˚S/165˚E 11˚S/56˚E 11˚N/142˚E 11˚N/142˚E 11˚N/142˚E Feature position Feature 23˚S/174˚W 42˚S/10˚E 55˚S/26˚E 23˚S/71˚W 22˚S/122˚E 60˚S/58˚W 32˚S/177˚W 20˚S/168˚E 25˚S/171˚E 32˚S/177˚W 32˚S/177˚W 32˚S/177˚W 23˚S/71˚W 32˚S/177˚W 19˚N/67˚W 79˚N/2˚E 26˚N/76˚W 19 ˚N/67 ˚W Molloy Hole (MOL) Molloy Puerto Rico Trench (PRT) Trench Rico Puerto Tonga Trench (TON) Trench Tonga Solomon trenches (SOL) Solomon trenches Mariana Trench (MT) Mariana Trench Java Trench (JAV) Trench Java Agulhas Fracture Zone (AFZ) Fracture Agulhas South Sandwich Trench (SAND) Trench South Sandwich Puerto Rico Trench (PRT) Trench Rico Puerto Abaco Canyon (ABA) Abaco Canyon Peru-Chile Trench (PCT) Trench Peru-Chile Wallaby-Zenith Fracture Zone (WZFZ) Fracture Wallaby-Zenith Mariana Trench (MT) Mariana Trench South Shetland Trench (SHET) Trench South Shetland Mariana Trench (MT) Mariana Trench Kermadec Trench (KT) Trench Kermadec New Hebrides Trench (NHT) Trench Hebrides New South Fiji Basin (SFB) South Fiji Kermadec Trench (KT) Trench Kermadec Kermadec Trench (KT) Trench Kermadec Kermadec Trench (KT) Trench Kermadec Peru-Chile Trench (PCT) Trench Peru-Chile Kermadec Trench (KT) Trench Kermadec Feature Arctic North Atlantic South Pacific South Pacific North Pacific Indian South Atlantic Southern North Atlantic North Atlantic South Pacific Indian North Pacific Southern North Pacific South Pacific South Pacific South Pacific South Pacific South Pacific South Pacific South Pacific South Pacific Ocean Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV Pressure Drop Pressure DSSV RV Sonne SO261 RV Sonne SO258 RV SY1615 Shinyo-Maru BIO Hesperides PharmaDEEP FK141109 RV Falkor RV TN309 G. Thompson Thomas RV KAH1310 Kaharoa RV KAH1310 Kaharoa RV KAH1301 Kaharoa RV KAH1202 Kaharoa RV KAH1109 Kaharoa RV Sonne SO209 RV RV KAH0910 Kaharoa Vessel and cruise number and cruise Vessel Chronological order of expeditions that provided data included this study. Note that only deployments of only deployments Note that of order included. transition or deeper are Chronological the abyssal-hadal represent baited camera landers that included this study. data provided that expeditions 2019 2019 2019 2019 2019 2019 2019 2019 2018 2018 2018 2017 2017 2015 2014 2014 2013 2013 2013 2012 2011 2010 2009 Table 1. Table Year

3 J. A. SWAN ET AL. Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Figure 1. Sampling locations. ABA, Abaco Canyon; AFZ, Agulhas Fracture Zone; JAV, Java Trench; KT, Kermadec Trench; MAR, Mariana Trench; MOL; Molloy Hole; NHT, New Hebrides Trench; PCT, Peru-Chile Trench; PRT, Puerto Rico Trench; SAND, South Sandwich Trench; SFB, South Fiji Basin; SHET, South Shetland Trench; SOL, Solomon trenches (San Cristobal and Santa Cruz); TON, Tonga Trench; WZFZ, Wallaby Zenith Fracture Zone.

PT sensor (Sea-Bird Scientific, Bellevue, WA, USA) or an acoustic highly variable data and were collected over multiple locations current meter with conductivity, temperature, and depth (CTD) with varying number of deployments per site using different sensors (SeaGuard, Aanderaa Instruments, Bergen, Norway) on equipment, Nmax and Tarr data are grouped and presented geo- the Abyssal-Lander and Gonzo; Bad Ape and Bad Spoon used graphically by ocean (North and South Pacific, Atlantic, Indian). compact temperature and pressure sensors (RBR-Duet; RBR, Ottawa, ON, Canada). Depth and temperature were recorded by Biochemical analyses Skaff, Flere, and Closp using CTD probes (SBE 49 FastCAT, Sea- Bird Scientific, Bellevue, WA, USA). Onboard the ships, white muscle samples were rapidly dissected from freshly captured shrimp, and (for Mariana Trench specimens) hemolymph was obtained with a 1-ml syringe from under the Image analysis dorsal carapace. Muscle and hemolymph samples were stored at –80˚C until shipped to Whitman College, Walla Walla, WA, USA Each minute of the footage was manually analysed using VLC on dry ice. There, frozen samples were processed as previously de- media player (VideoLAN, France). ‘Bait-attending’ species were scribed (Kelly & Yancey, 1999). In brief, they were homogenised in defined as species observed to directly interact with the baited 7% (vol/vol) perchloric acid to remove proteins by precipitation, experiment, either by feeding at the bait itself, or preying upon followed by centrifugation. Then osmolyte concentrations of the other scavengers that were feeding on the bait. The observations supernatants were measured with appropriate standard as previ- of that species were assumed to be incidental if no direct interest ously described (Kelly & Yancey, 1999; Downing et al., 2018): for in the bait or other feeding organisms was observed. Individual TMAO by using an iron-EDTA reagent and a colorimetric reac- decapods seen in the videos and images were identified and the tion with picric acid; for other osmolytes by high-pressure liquid time of each sighting recorded. To provide an impression of chromatography with separation on a SugarPak I column (Waters, population density, beyond simply presence or absence, the max- Milford, MA, USA) and detection by refractive index. imum simultaneous number (Nmax) visible in a single image or within a 1-min video sequence was noted, as was the first arrival time (T ) in minutes. Higher N and low T values are indi- arr max arr RESULTS cative of greater local density. These values are, however, given to illustrate relative differences in population density between The first notable result was the absence of bait-attending deployments and not true population density estimates. The decapods at depths exceeding 4,000 m in both the Arctic behaviour of decapods at simulated food falls is not as predict- and Southern (Antarctica) oceans. Decapods, however, were able or consistent as in the case of the more commonly studied observed at all other sites. There were three main bait- taxa. Arrival of decapods and how long they remain at the bait attending peneaoid shrimps: Benthesicymus cf. crenatus Spence is greatly influenced by the presence (or absence) of various fish Bate, 1881 (Benthesicymidae), Cerataspis cf. monstrosus Gray, species (macrourids, ophidiids, and liparids) that occupy discrete 1828 (Aristeidae), and Hymenopenaeus nereus (Faxon, 1893) and different depths ranges within the depth range of the deca- (Solenoceridae). These were all captured at least once (Fig. 2) pods (Linley et al., 2016, 2017). As animal interactions result in and observed in situ (Fig. 3A–D). Oplophoroid prawns were

4 MAXIMUM-DEPTH LIMIT OF DEEP-SEA DECAPODS

Table 2. Lander vehicles and details of camera type and settings used in this study.

Lander Depth rating (m) Year Camera type Typical setting

Hadal-Lander A 11,000 2009 3CCD video 1 min on, 4 min off Hadal-Lander B 11,000 2010–2011 5 Megapixel still 1 min interval

Hadal-Lander C 11,000 2013–2017 3CCD video 1 min on, 4min off Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021 Abyssal Lander 6,000 2013–2017 5 Megapixel still 1 min interval Gonzo 6,000 2017 HD video 30 sec on, 1 min off Bad Spoon/Bad Ape 11,000 2018–2019 HD video 30 sec on, 1 min off Skaff/Closp/Flere 11,000 2018–2019 HD video Continuous observed at various sites to a maximum depth of 6,931 m and the same individuals were re-entering later in the deployment. were not bait-attending (Fig. 3E, F). There were several other Although primarily a predator, B. cf. crenatus was observed to oc- miscellaneous decapod species observed in the footage, albeit casionally scavenge pieces of bait that had been torn off by larger infrequently and sporadically, but were too small to identify scavengers but was not observed to feed on the bait directly. with confidence. The maximum number of B. cf. crenatus observed at any one time was nine, at 6,142 m in the Mariana Trench. Solitary shrimp Penaeoidea were generally observed, typically reaching maximum numbers of three or less with a few exceptionally high numbers between 6,000 Individuals of Benthesicymus cf. crenatus were observed in the and 7,500 m. Their arrival time tended to be less than one hour, Kermadec Trench (5,100–6,979 m), Mariana (4,523–7,716 m), with longer arrival times at the both the shallow and deeper ex- New Hebrides (5,215–6,898 m), Puerto Rico (5,360–6,954 m), tremes of their depth range. Tonga (6,848–7,273 m), San Cristobal (6,013–7,220 m), and Santa Cerataspis cf. monstrosus, which has been cited in most of the lit- Cruz trenches (5,906–7,431 m), South Fiji Basin (4,078–4,100 m), erature as Plesiopenaeus armatus (Spence Bate, 1881) (Aristeidae), was and the Wallaby-Zenith Fracture Zone (6,084 m) (Fig. 4A). The present in the Kermadec (4,061–6,474 m), Mariana (4,040–6,010 species was not present in the Abaco Canyon, Agulhas Fracture m), New Hebrides (4,100–6,190 m), Puerto Rico (4,040–5,880 m), Zone, Java Trench, Peru-Chile Trench, and the polar locations. and Java (5,760–6439 m) trenches, as well as in the Abaco Canyon The specimens observed at 7,716 m in the Mariana Trench are (4,900 m) and the Agulhas (5,493 m) and Wallaby-Zenith fracture also the deepest recorded decapod to date, albeit extending the zones (4,767–6,084 m), and the South Fiji Basin (4,078–4,100 m). known depth range by just 13 m. They were absent from the Peru-Chile, Tonga, San Cristobal, and The species is assumed to be the benthopelagic B. crenatus, based Santa Cruz trenches, and in the polar locations (Fig. 5A). on external morphology, primarily the short, serrated rostrum and The species is assumed to be the free-swimming C. cf. monstrosus almost transparent scaphocerites and a darker area on the upper based on external morphology, primarily the long, elongated ros- rear carapace. A single specimen collected on the 2014 exped- trum, opaque-red scaphocerites, and a deep, homogeneous red ition to the Kermadec Trench was confirmed to be B. crenatus (J.C. body colour. One specimen was obtained by baited trap at 4,100 Drazen, personal communication). m in the South Fiji Basin and was confirmed as C. cf. monstrosus (S. Benthesicymus cf. crenatus. was observed in 49 of 176 deployments Ahyong, Australian Museum). (27.8%). There were 143 deployment made between 4,000 m Cerataspis cf. monstrosus was observed in 24 of our 176 deploy- and the maximum known depth for the species (7,414 m). Within ments (13.6%). There were 94 deployment made between 4,000 this range, the 43 positive deployments accounted for a 34% en- m and the maximum known depth for the species (6,474 m). counter rate. If their absence from the polar regions is taken into Within this range, the 24 positive deployments accounted for a account, their encounter rate was 37%. 25.5% encounter rate. If their absence from the polar regions and The vertical spread of deployments managed to capture the some of the South Pacific trenches is taken into account, their en- whole depth range of B. cf. crenatus indicating that it is prevalent counter rate was 31.2%. between ~4500 m and 7700 m, with the exception of some, al- The maximum number of individuals observed at any one time beit rare, observations on the South Fiji Basin. The numbers of was nine, at 4,139 m in the New Hebrides Trench. They were individuals simultaneously observed (Nmax) generally increased typically observed to be solitary but could, albeit briefly, reach a from one at ~4,000 m (coinciding with very long arrival times) maximum number of two or three in some deployments (Fig. 5B). to 5 and 6 between 6,000 and 7,000 m (with the exception of The time of first arrival did not appear to correlate to depth or one high number recorded at 6,142 m in the Mariana Trench). location (Fig. 5C).

Arrival times (Tarr) from 4,000 to 7,000 m were generally less than When bait attending, C. cf. monstrosus was observed to be pri- 100 min. Maximum numbers dropped to one and arrival times marily a scavenger but occasionally predatory at specific sites. exceeded 120 min beyond 7,000 m, m (Fig. 4B, C). In the New Hebrides Trench, for example, it was observed to When bait attending, B. cf. crenatus exhibited predatory behav- be a scavenger, circling the bait reducing distance, as if homing iour, being observed to prey on scavenging lysianassoid amphipods in on its precise location. It would eventually make contact with swarming around the bait. They also occasionally attempted to the bait and begin examining it with its pereopods. Once a suit- remove smaller ‘supergiant’ amphipods (Alicella gigantea Chevreux, able feeding location was found, individuals would lower its man- 1899), albeit unsuccessfully. Benthesicymus cf. crenatus was observed dibles to the bait and attempt to rip off a piece of flesh. Tail flip using its pereopods to search the bait for members of the feeding responses would often accompany this behaviour. On many oc- swarm that were small enough to overpower and remove. If casions, individuals would examine the bait and then leave, po- one amphipod was found it would grip the prey and either pry tentially returning later in the deployment to attempt to find a it free with its pereopods or vigorously kick its pleopods until the suitable feeding location. In deployments where C. cf. monstrosus amphipod became dislodged. The shrimp would then swim away behaved as a predator (Kermadec and Mariana trenches, and with the amphipod, presumably to feed. If the initial seeking was Wallaby-Zenith Fracture Zone), the behaviour was similar to that unsuccessful, shrimp would leave the field of view and presumably where amphipods would be grabbed and taken away.

5 J. A. SWAN ET AL. Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Figure 2. Carapace and rostrum sketches compared to captured specimens of the three dominant hadal decapods: B. cf. crenatus, C. cf. monstrosus, and H. nereus.

The only other conspicuous species recorded was the benthic (6,000–6,474 m), Mariana (6,010–6,931 m), New Hebrides Hymenopenaeus nereus (Solenoceridae), which was locally abundant (4,100–6,190 m), Puerto Rico (4,830–6,356 m), and Java trenches but exclusive to the Peru-Chile Trench (4,051–8,074 m; Fig. 5A). (5,760–6,439 m), as well as the Wallaby-Zenith Fracture Zone The identification was confirmed based on external morphology, (4,767–6,084 m). They were absent from the Peru-Chile, Tonga, primarily the short, sharp, and serrated rostrum, lighter orange San Cristobal. and Santa Cruz trenches, and the polar regions body colour, and the morphology of the pereopods in several (Fig. 6). samples obtained by baited trap. The identification of the oplophoroid to species level was not Hymenopenaeus nereus appeared to be a predator when bait at- possible due to the small size of individuals. Two specimens re- tending, targeting scavenging amphipods gathering at the bait. covered by baited trap from the New Hebrides Trench at 5,180 m Individuals were often observed climbing on the bait or stationery were nevertheless identified as Acanthephyra tenuipes (Spence Bate, on the sediment next to it. They would then either remove smaller 1881) and Heterogenys microphthalma (Smith, 1885). amphipods from the bait or intercept them as they swam to- These two oplophorids swam with an unusual lateral orienta- wards or around the bait. Once an amphipod was caught, shrimp tion and were observed occasionally passing, mostly solitary, by would remain in the vicinity of the bait to feed, as opposed to the bait. Very occasionally they were observed interacting with the leaving the immediate area as observed in B. cf. crenatus and C. cf. bait, albeit briefly, before swimming away. Their feeding strategy monstrosus. Individuals of H. nereus would remain around the bait could not be determined by the video footage. hunting in this way until disturbed by the approach of a fish (e.g., Another oplophorid was observed at 4,040 m in the Mariana macrourids). The low number of observations, as the result of Trench and at 4,100 m in the New Hebrides Trench and South only being recorded in the Peru-Chile Trench, offers little insight Fiji Basin. It was large, deep red and with long and distinctive into depth-related trends in maximum numbers or arrivals time white antennae. It was tentatively identified as Hymenodora spp. (T.- (Fig. 5D, E). Y. Chan, personal communication) (Fig. 3F). Two other species of penaeids were observed. The first, from 5,880 m in the Puerto Rico Trench, was larger, with less ridged Osmolyte/piezolyte contents lateral body movements and a lighter orange body colour than in B. cf. crenatus (Fig. 3D). It visited the bait just once, for 11 min, and Three individuals of B. crenatus captured near 6,000 m in was not seen at another depth or location. A small benthesicymid baited landers, two from the Mariana Trench and one from the was observed at abyssal depths in the South Fiji Basin and New Kermadec Trench were analysed for organic osmolytes in muscle. Hebrides Trench but was too small to identify from video. A spe- Hemolymph was tested in the Mariana specimens (Table 3). By cimen was recovered by baited trap and identified as Benthesicymus far the dominant muscle osmolyte was TMAO (62–66% of total howensis (Dall, 2001). measured contents), followed by alanine (16–23%) and glycine (9–13%), with much lower contents of ß-alanine (1.2–1.7%), Oplophoroidea betaine (trimethyl-glycine; 0.9–1.7%), glycerophosphorylcholine (GPC) (0.7–0.9%) and scyllo-inositol (0.1–0.2%). The osmolyte Oplophoroids resembling species of Acanthephyra and Heterogenys taurine, typically high in shallow-water marine invertebrates, was Chace, 1986 (Fig. 3E) were observed in the Kermadec not detected. In hemolymph, only TMAO was detected, and at

6 MAXIMUM-DEPTH LIMIT OF DEEP-SEA DECAPODS Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Figure 3. In situ images of Penaeoidea: Benthesicymus cf. crenatus (A), Cerataspis cf. monstrosus (B), Hymenopenaeus nereus (C), unidentified penaeid from the Puerto Rico Trench (D), aff. Acanthephyra or Heterogenys sp. (E), and Hymenodora sp. from the Mariana and New Hebrides trenches (F). only very low levels compared to muscle, indicating that the or- unremarkable in abyssal-plain species (Ebbe et al., 2010), with the ganic osmolytes are primarily intracellular. caveat that future sampling may reveal greater diversity.

Population-level depth limits DISCUSSION Bait-attending penaeids were absent from the Arctic and Antarctic These observations reveal that the decapods of the abyssal-hadal (Southern Ocean) locations at depths greater than 4,000 m. The transition zone are represented primarily by C. cf. monstrosus and B. same may be true for the oplophorids, but as they could be inci- cf. crenatus, with the exception being the Peru-Chile Trench, were dental observations rather than observations made while attracted C. monstrosus appears to be replaced by H. nereus. Only a few bait- to bait, they are likely overlooked in studies using bait. attending species have thus become adapted to the hadal zone, Oplophorids showed a similar biogeographical pattern to C. cf. and they appear to be limited to the upper part of that zone (< monstrosus in that they were absent from the polar regions and the ~7,700 m). The observed depth limits are even shallower for Peru-Chile Trench. In this survey, they share a similar, albeit nar- populations in some of the localities (see below). rower, depth range as B. crenatus within the abyssal-hadal transition The identification of H. nereus, based on samples obtained zone. Compared to other dominant deep-sea taxa of caridean simultaneous to the video observations, is a confident one. The shrimps, oplophorids have been shown to have lower feeding identifications of C. cf. monstrosus and B. cf. crenatus are, however, rates (Cartes, 1993). A lower metabolic rate would explain why based on the identification of specimens from only the Mariana oplophorids are one of the only caridean families to be observed at and Kermadec trenches, respectively. It is possible that not all ob- depths greater than 6,000 m as they are already adapted for oligo- servations belong to these two species, particularly with regard to trophic environments. Such adaptations would also explain why the different feeding behaviours observed inC. cf. monstrosus. If oplophorids exist in relatively higher abundance in the extremely morphological differences are present, they are beyond the reso- nutrient-poor areas, such as the New Hebrides Trench, where lution of the cameras. Such interpretation assumes that C. cf. there is an absence of scavenging fishes (Linley et al., 2017) and monstrosus and B. cf. crenatus are pan-oceanic, cosmopolitan species, reduced competition with more energetically expensive decapods.

7 J. A. SWAN ET AL. Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Figure 4. Vertical distribution of Benthesicymus cf. crenatus, across all sites. Grey triangles represent baited camera deployments where the species was absent and black dots where present. The solid horizontal lines indicate the maximum water depth in each location. See Table 1 for abbreviations of locations.

Maximum numbers (Nmax) (B) and first arrival times (Tarr) (C) at each deployment where B. cf. crenatus was present in the North and South Pacific and the Atlantic and Indian oceans.

Our study did not resolve the true minimum depth for C. cf. whose maximum depth happens to extend slightly into the upper monstrosus as it was largely present at the shallower deployments hadal depths. (~4,000 m) at the most studied sites. The maximum depth to The absence of C. cf. monstrosus in the San Cristobal, Santa date is 6,474 m from our study. In contrast, B. cf. crenatus appears Cruz, and Tonga trenches may therefore be the result of sam- to have a deeper depth range, with minimum depth throughout pling taking place beyond their depth range in those locations, as the sites varying between 4,500 and 6,000 m. Their absence at they were found in the Kermadec and New Hebrides trenches. shallower deployments suggests that their entire depth range was Their absence from the Peru-Chile Trench is nevertheless con- captured. The maximum depth of B. cf. crenatus in this study fusing, given their seemingly large biogeographical distribution. was 7,716 m in the Mariana Trench, which is just 13 m deeper Furthermore, they are known from depths of ~4,000–4,200 m in than the previous depth record of 7,703 m in the Japan Trench the Peru Basin immediately west of our study (Drazen et al., 2019) (Jamieson et al., 2009a). Benthesicymus cf. crenatus appears to be and further still on the Clarion-Clipperton Fracture Zone (CCFZ) characteristic of the abyssal-hadal transition zone, whereas C. (Leitner et al., 2017). Hymenopenaeus nereus is also known from the cf. monstrosus appears to be an abyssal species (as does H. nereus) CCFZ (Hendrickx & Wicksten, 2016).

8 MAXIMUM-DEPTH LIMIT OF DEEP-SEA DECAPODS Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Figure 5. Vertical distribution of Cerataspis cf. monstrosus across all sites (A). Grey triangles represent baited camera deployments where the species was absent, and black dots where present. The solid horizontal lines indicate the maximum water depth in each location. The data for the Peru-Chile Trench (PCT) are not for C. cf. monstrosus but rather Hymenopenaeus nereus. See Table 1 for abbreviations of locations. Maximum numbers (Nmax) (B) and first arrival times (Tarr)

(C) at each deployment where C. cf. monstrosus was present in the North and South Pacific, and the Atlantic and Indian oceans. Maximum numbers (Nmax) (D) and first arrival times (Tarr) (E) at each deployment where H. nereus was present in the South Pacific.

9 J. A. SWAN ET AL. Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021

Figure 6. Vertical distribution of Oplophoroidea across all sites. Grey triangles represent baited camera deployments where oplophoroids were absent, black dots where present. The solid horizontal lines indicate the maximum water depth in each location. See Table 1 for abbreviations of locations.

Another peculiarity in the Peru-Chile Trench is the relatively significantly controlled the distribution of decapods in the deep shallow maximum depth of B. cf. crenatus. This species is also Eastern Pacific through physiological exclusion of swimming spe- known from the abyssal Peru Basin and CCFZ (Drazen et al., 2019 cies from the lower oxygenated waters beneath oxygen minimum and Leitner et al., 2017, respectively) but does not extend as far zones (OMZs). We also found that oplophorids, which are also con- into this trench is it does in the other trenches in our study. Low sidered free-swimming, were entirely absent from the Peru-Chile temperatures are known to inhibit pressure tolerance (Cossins Trench, further supporting the effect of dissolved oxygen in the dis- & MacDonald, 1989; Childress, 1995) and the maximum depth tribution of deep-water shrimps. Future behavioural observations of a given species should be deeper in warmer trench settings of H. nereus could indicate that they are predominantly benthic and (Brown & Thatje, 2011). Their absence from the polar sites may seldom swim. The reduced swimming behaviour is likely what al- be explained by low (sub-zero) ambient temperatures. Their max- lows this species to inhabit areas of lower O2 concentration. imum depth increases from 6,987–7,273 m in the south Pacific trenches (Kermadec and Tonga) to 7,703–7,716 m in the north Class-level depth limit Pacific trenches (Japan and Mariana), where the mean temperature of the abyssal-hadal transition zone increases from 1.21–1.26 The decapods have a shallower maximum depth than other mem- to 1.67–1.68 ˚C, in the south and north Pacific, respectively. bers of Malacostraca known from hadal depths: Cumacea to Sampling in the next warmest trenches, the Santa Cruz and San 8,042 m, Mysida to 8,720 m, Tanaidacea to 9,174 m, Isopoda to Cristobal trenches were unfortunately not deep enough to resolve 10,730 m, and Amphipoda to 10,925 m (Jamieson, 2015). Even maximum depth, but the trend of deeper-warmer would suggest among other classes of Arthropoda known from hadal depths, that with a mean abyssal-hadal transition zone temperature of the Decapoda are still a ’shallow’ group: Cirripedia to 7,880 m, 1.97 ˚C, the Peru-Chile Trench would see a maximum depth of Copepoda to 8,500 m, and Ostracoda to 9,500 m (Jamieson, B. cf. crenatus close to or exceeding that of the Mariana and Japan 2015). trenches. Their depth range is instead ~1,500 m shallower than Exclusion of the Decapoda from the greatest hadal depths may depth and temperature would predict, suggesting another environ- not be directly due to hydrostatic pressure. Food limitation or low mental factor is affecting their distribution. temperature are possibilities, potentially also explaining the ab- The absence of oplophorids and C. cf. monstrosus, the presence sence from moderate depths in polar regions. It is also possible of B. cf. crenatus at shallower depths, and their sudden replace- that deeper living decapods are not bait-attending; penaeoids such ment by H. nereus could be a result of low oxygen. Glud et al. (2021) as benthesicymids are believed to be active predators that scour the sediment for small crustaceans, rather than relying on bait falls found that the bottom-water O2 concentration in Atacama Trench (~160 μmol l–1) was somewhat lower than in the Kermadec Trench (Papiol & Hendrickx, 2016b). (~210 μmol l–1), which is comparable to other trenches were oxygen If exclusion from greater depths is due, at least in part, to pres- data are available. Bottom-water oxygen in the Mariana Trench has sure, piezolyte concentration may play a role in determining what been reported as ~182–185 μmol l–1 between 6,000 and 10,900 m the maximum depth is. TMAO is by far the major piezolyte in (Glud et al., 2013; Wenzhöfer et al., 2016; Kawagucci et al., 2018; the decapods analyzed, just as it is in other taxa including hadal Luo et al., 2018). Decapod crustaceans are highly sensitive to de- snailfishes (Liparidae) (Yancey et al., 2014; Linley et al., 2016) creases in dissolved oxygen (DO) (Hendrickx & Wicksten, 2016). and amphipods (Downing et al., 2018). As osmoconformers with This sensitivity is possibly a function of their swimming capabil- seawater, invertebrates having depth limits are unlikely to suffer ities as ‘free-swimming’ decapods (such as C. cf. monstrosus), which from osmotic balance difficulties, since piezolytes simply replace have been shown to possess higher oxygen-minimum thresholds ‘shallow’ osmolytes like taurine (see below). This contrasts with associated with greater swimming capabilities (Childress et al., bony fishes in which sufficient TMAO levels for pressure counter- 1990; Maynou & Cartes 1998; Vaquer-Sunyer & Duarte, 2008). action in the greatest trench depths would create a hyperosmotic Papiol & Hendrickx (2016a) demonstrated that dissolved oxygen stress in these normally hypoosmotic animals (Yancey et al., 2014). 10 MAXIMUM-DEPTH LIMIT OF DEEP-SEA DECAPODS

An inadequate food supply involving TMAO is a possible limiting factor among decapods if TMAO itself or necessary pre- SUM 418 406 390 405 ± 14

n/a n/a cursors (e.g., lipids with trimethylamine head groups) are mainly derived from the diet, as is the case in some animal groups (Seibel & Walsh, 2002). Interspecific competition at bait falls could be a major limiting factor given the increased presence of competing

necrophagivores such as amphipods at depths greater than 7,700 Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021 m. This scenario could involve an inability to accumulate sufficient TMAO, not just a reduction in general nutrition. The main 0.56 0.89 0.81 0.75 ± 0.17 0.19 nd nd

Scyllo-inositol competition among decapods for access to scavenging amphipods are the liparid fishes (snailfishes). Liparids tend to occupy similar upper trench depths, but are known to extend slightly deeper, to 8,145 m, albeit in diminishing numbers (Linley et al., 2017). The largest number of liparids, however, coincides with the largest adults that occur within the same depth ranges as the decapods, 1.42 ß-alanine 5.21 5.27 6.81 5.76 ± 0.91 5.76 ± nd nd therefore if the decapods move deeper, nearer 8,000 m, competition would be eased, but they do not. Furthermore, if they move deeper still, to depths greater than 8,200 m, there would be no competition for amphipod prey at all, plus in the absence of any Glycine 11. 9 other predators, the deeper amphipod populations are far more 39.5 54.3 50.8 48.2 ± 7.7 trace trace abundant (Jamieson et al., 2009d). This suggests that a depth limit other than diet and competition is controlling the maximum depth of decapods. Oxygen concentrations are a possible limiting factor with re- Alanine 96.8 67.3 80.7 81.6 ± 15 81.6 ± 20.2 trace trace fresh weight; nd, no data. weight; fresh spect to TMAO. Free O2 is needed to convert trimethylamine, a –1 breakdown product of many common lipids as well as betaine, into TMAO, by TMA-oxidase (Seibel & Walsh, 2002). It is, how-

ever, not known if free O2 is ever a limiting factor for this process over the ranges of ambient oxygen in our study sites. 1.21 Betaine 4.67 3.54 6.50 4.90± 1.49 nd nd Another possibility is that TMAO itself is toxic above ~350 mM, the level found in fishes at their greatest depth of ~8,000 m. TMAO in the absence of a perturbant such as pressure can have negative overstabilising effects on macromolecules (Yancey & Siebenaller, 2015), which however, has not yet been tested. It is also possible that a very high concentration of TMAO is not toxic per se, but simply cannot adequately counteract the strong pressure effects below ~8,000 m. If so, TMAO might be needed at such high concentrations that it would make individuals hyper-osmotic to seawater, a potential physiological stress as 3.32 2.88 3.49 3.23 ± 0.3 0.80 nd nd

Glycerophosphorylcholine Glycerophosphorylcholine (GPC) in the case of hadal fishes (Yancey et al., 2014). If so, other adaptations to pressure are needed at the greatest depths. Perhaps decapods, similar to bony fishes but unlike amphipod crustaceans and other crustacean taxa noted earlier, have not evolved such mech- anisms. TMAO levels in amphipods do increase linearly with depth from 0 to ~11,000 m, but at depths greater than ~4,000 m, other potential piezolytes show increasing contributions with depth to the osmolyte pool along with further declines in concen- 12.9 64.3 25.2 Trimethylamine Trimethylamine (TMAO) N-oxide 268 272 241 260 ± 17 trations of ‘shallow’ osmolytes. These potential piezolytes, most of which have not been found in hadal fishes, include GPC, proline betaine, and scyllo-inositol, all of which have protective properties that might aid with pressure adaptation (Downing et al., 2018). These solutes are in different proportions and proline betaine is absent entirely in decapods. Whole amphipods, Abyssorchomene Tissue White muscle White White muscle White White muscle White Mean ± SD % of total % of total Hemolymph Hemolymph musculosus (Stebbing, 1888) (N = 4) caught in the same trap as the Mariana Trench decapods from 6,068 m possessed the same per- centage of TMAO within the osmolyte pool (64% in both groups) (see Table 3 for decapods), as did the non-piezolyte glycine (~12% in both groups). The non-piezolyte alanine, however, was much Depth (m) 6,068 6,068 6,061

6,068 6,068 lower (1.9% versus 20%), while the other organic osmolytes were much higher in amphipods compared to decapods: 4.1% versus

0.8% for GPC, 12.6% versus 1.2% for betaine, 2.9% versus 0.2% for scyllo-inositol, and 1.8% versus 0% for proline betaine. All are known stabilisers against perturbants of macromolecules, though Primary organic osmolytes in muscle and hemolymph from three specimens of in mmol kg three and hemolymph from Primary organic osmolytes in muscle . Values crenatus Benthesicymus cf. the properties of piezolytics have not yet been confirmed for most of them. Regardless, it may be that amphipods are able to accumulate non-TMAO piezolytes at the greatest hadal depths while Specimen number, Specimen number, trench Table 3. Table 200101 Mariana 200101 200102 Mariana 200102 100138 Kermadec Kermadec 100138

200101 Mariana 200101 200102 Mariana 200102 decapods (and fishes) cannot.

11 J. A. SWAN ET AL.

Resolving the nuances in population-specific depth ranges re- Downing, A.B., Wallace, G.T. & Yancey, P.H. 2018. Organic osmolytes quires a re-think of how deep benthopelagic decapods are sam- of amphipods from littoral to hadal zones: Increases with depth in pled. Typical baited caged fish traps are highly ineffective in trimethylamine N-oxide, scyllo-inositol and other potential pressure sampling decapods and therefore negates the comparative physio- counteractants. Deep Sea Research Part I, 138:1–10. logical studies needed to unequivocally resolve, or enhance our Drazen, J. C., Leitner, A.B., Morningstar, S., Marcon, Y., Greinert, J. & Purser, A. 2019. Observations of deep-sea fishes and mobile scaven- understanding of, the drivers of vertical zonation. gers from the abyssal DISCOL experimental mining area. Biogeosciences,

16: 3133–3146. Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021 Ebbe, B., Billett, D.S., Brandt, A., Ellingsen, K., Glover, A., Keller, S., ACKNOWLEDGMENTS Malyutina, M., Martínez Arbizu, P., Molodtsova, T., Rex, M. & Smith, C.R. 2010. Diversity of abyssal marine life. In: Life in the world’s We thank Tin-Yam Chan, National Taiwan Ocean University oceans: Diversity, distribution, and abundance, (A. McIntyre, ed.), pp. 139– and Shane Ahyong, Australian Museum for taxonomic assist- 160. Wiley-Blackwell, Chichester, UK. Faxon, W. 1893. Reports on the dredging operations off the west coast of ance. We thank the captain, crew, and scientific parties of the fol- Central America to the Galapagos, to the west coast of Mexico, and in lowing expeditions: RV Sonne SO209, SO258, SO261, RV Kaharoa the Gulf of California, in charge of Alexander Agassiz, carried on by KAH0910, KAH1109, KAH1202, KAH1301, KAH1310, RV the U.S. Fish Commission steamer “Albatross” … Preliminary descrip- Thomas G. Thompson TN309, RV Falkor FK141109, RV Hesperides tions of new species of Crustacea. Bulletin of the Museum of Comparative (PharmaDeep) and the DSSV Pressure Drop (Five Deeps Expedition). Zoology at Harvard College, 24: 149–220. Fujii, T., Kilgallen, N.M., Rowden, A.A. & Jamieson, A.J. 2013. Amphipod We especially thank Shane Eigler (Triton Submarines, US), community structure across abyssal to hadal depths in the Peru- Heather Stewart (British Geological Survey), and Johanna Weston Chile and the Kermadec Trenches. Marine Ecology Progress Series, 492: and Nick Cuomo (Newcastle University) for their assistance in 125–138. deploying the landers at sea. The HADEEP project was funded Gillett, M.B., Suko, J.R., Santoso, F.O. & Yancey, P.H. 1997. Elevated levels of trimethylamine oxide in muscles of deep-sea gadiform tele- by the Nippon Foundation (Japan) (grant 2009765188) and the osts: A high-pressure adaptation? Journal of Experimental Zoology, 279: Natural Environmental Research Council (UK) (NE/E007171/1), 386–391. HADEEP II-IV by Total Foundation (France) through the projects Glud, R.N., Berg, P., Thamdrup, B., Larsen, M., Stewart, H.A., ‘Multi-disciplinary investigations of the deepest scavengers on Earth’ Jamieson, A.J., Glud, A., Oguri, K., Sanei, H., Rowden, A.A. & (2010–2012) and ‘Trench Connection’ (2013–2015),’ Hades-K’ Wenzhöfer, F. 2021. Hadal trenches are dynamic hotspots for early diagenesis in the deep-sea. Nature Communications Earth and Environment [in by the National Science Foundation (USA) (OCE 1130712), press]. ‘Hades-M’ by Schmidt Ocean Institute (USA) (cruise FK141109), Glud, R.N., Wenzhöfer, F., Middelboe, M., Oguri, K., Turnewitsch, R., ‘PharmaDeep’ by European Union Research Infrastructures pro- Canfield, D.E. & Kitazato, H. 2013. High rates of microbial carbon ject ´EuroFleets 2´ under the 7th Framework Programme of the turnover in sediments in the deepest oceanic trench on Earth. Nature Geoscience, 6: 284–288. European Commission (grant 312762), HADES–ERC Advanced Gray, J.E. 1828. Spicilegia Zoologica; or, original figures and short systematic descrip- Grant ‘Benthic diagenesis and microbiology of hadal trenches’ tions of new and unfigured animals, Vol. 1. pp. 1–8. Treuttel, Wurtz & Co. (grant 669947), the ‘Hadal Zones of our Overseas Territories’ by and W. Wood, London. the Darwin Initiative funded by the UK Government (DPLUS093,) Hendrickx, M.E. & Wicksten, M.K. 2016. New records of decapod crust- and the Five Deeps and Ring of Fire expeditions, funded by Victor aceans in the eastern Pacific. Crustaceana, 89: 603–610. Herring. P.J. 2002. The biology of the deep ocean. Oxford University Press, Oxford. Vescovo of Caladan Oceanic LLC (USA). The authors declare no Hessler, R.R., Ingram, C.L., Yayanos, A.A. & Burnett, B.R. 1978. conflict of interest. Scavenging amphipods from the floor of the Philippine Trench. Deep- Sea Research, 25: 1029–1047. REFERENCES Jamieson, A.J. 2015. The hadal zone: life in the deepest oceans. Cambridge University Press, Cambridge, UK. Aguzzi, J., Jamieson, A.J., Fujii, T., Sbragaglia, V., Costa, C., Menesatti, P. Jamieson, A.J. 2018. A contemporary perspective on hadal science. Deep- & Fujiwara, Y. 2012. Shifting feeding behaviour of deep-sea buccinid Sea Research Part II, 155: 4–10. gastropods at natural and simulated food falls. Marine Ecology Progress Jamieson, A.J. & Vecchione, M. 2020. First in situ observation of Series, 458: 247–253. Cephalopoda at hadal depths (Octopoda: Opisthoteuthidae: Blankenship, L.E. & Levin, L.A. 2007. Extreme food webs: Foraging strat- Grimpoteuthis sp.). Marine Biology, 167: 82 [doi: 10.1007/ egies and diets of scavenging amphipods from the ocean’s deepest 5 s00227-020-03701-1]. kilometres. Limnology and Oceanography, 52: 1685–1697. Jamieson, A.J., Fujii, T., Solan, M., Matsumoto, A.K., Bagley, P.M. & Brown, A. & Thatje, S. 2011. Respiratory response of the deep-sea Priede, I.G. 2009a. First findings of decapod crustacea in the hadal- amphipod Stephonyx biscayensis indicates bathymetric range limitation zone. Deep-Sea Research Part I, 56: 641––647. by temperature and hydrostatic pressure. PLoS ONE, 6: e28562 [doi: Jamieson, A.J., Fujii, T., Solan, M. & Priede, I.G. 2009b. HADEEP: Free- 10.1371/journal.pone.0028562]. falling landers to the deepest places on Earth. Marine Technology Society Cartes, J.E. 1993. Feeding habits of oplophorid shrimps in the deep Journal, 43: 151–159. western Mediterranean. Journal of the Marine Biological Association of the Jamieson, A.J., Fujii, T., Solan, M., Matsumoto, A.K., Bagley, P.M. & United Kingdom, 73: 193–206. Priede, I.G. 2009d. Liparid and Macrourid fishes of the hadal zone: Chace, F.A. Jr. 1986. The caridean shrimps (Crustacea: Decapoda) of In situ observations of activity and feeding behaviour. Proceedings of the the Albatross Philippine Expedition, 1907–1910, Part 4: Families Royal Society Part B, 276: 1037–1045. Oplophoridae and Nematocarcinidae. Smithsonian Contributions to Jamieson, A.J., Kilgallen, N.M., Rowden, A.A., Fujii, T., Horton, T., Zoology, 432: 1–82. Lörz, A.-N., Kitazawa, K. & Priede, I.G. 2011. Bait-attending fauna of Childress, J.J. 1995. Are there physiological and biochemical adaptations the Kermadec Trench, SW Pacific Ocean: evidence for an ecotone across of metabolism in deep-sea animals? Trends in Ecology and Evolution, 10: the abyssal-hadal transition zone. Deep-Sea Research Part I, 58: 49–62. 30–36. Jamieson, A.J., Lacey, N.C., Lörz, A.-N., Rowden, A.A. & Piertney, S.B. Childress, J.J., Cowles, D.L., Favuzzi, J.A. & Mickel, T.J. 1990. Metabolic 2013. The supergiant amphipod Alicella gigantea (Crustacea: Alicellidae) rates of benthic deep-sea decapod crustaceans decline with increasing from hadal depths in the Kermadec Trench, SW Pacific Ocean. Deep- depth primarily due to the decline in temperature. Deep Sea Research Part Sea Research Part II, 92: 107–113. A, 37: 929–949. Jamieson, A.J., Fujii, T. & Priede, I.G. 2012. Locomotory activity and Cossins, A.R. & Macdonald, A.G. 1989. The adaptation of biological feeding strategy of the hadal munnopsid isopod Rectisura cf. herculea membranes to temperature and pressure: fish from the deep and cold. (Crustacea: Asellota) in the Japan Trench. Journal of Experimental Journal of Bioenergetics and Biomembranes, 21: 115–135. Biology, 215: 3010–3017.

12 MAXIMUM-DEPTH LIMIT OF DEEP-SEA DECAPODS

Jamieson, A.J., Ramsey, J. & Lahey, P. 2019. Hadal manned submersible. Pérès, J.M. 1965. Apercuw sur les résultats de deux plongées effectuées Sea Technology, 60: 22–24. dans le ravin de Puerto-Rico par le bathyscaphe Archimède. Deep Sea Jamieson, A.J., Solan, M. & Fujii, T. 2009c. Imaging deep-sea life beyond Research, 12: 883–891. the abyssal zone. Sea Technology, 50: 41–46. Ritchie, H., Jamieson, A.J., Kilgallen, N.M. & Piertney, S.B. 2015. Kawagucci, S., Makabe, A., Kodama, T., Matsui, Y., Yoshikawa, C., Phylogenetic relationships among hadal amphipods of the Superfamily Ono, E., Wakita, M., Nunoura, T., Uchida, H. & Yokokawa, T. 2018. Lysianassoidea: Implications for taxonomic revision and biogeographic Hadal water biogeochemistry over the Izu-Ogasawara Trench ob- distribution. Deep-Sea Research Part I, 105: 119–131.

served with a full-depth CTD-CMS. Ocean Science, 14: 575–588. Samerotte, A.L., Drazen, J.C., Brand, G.L., Seibel, B.A. & Yancey, P.H. Downloaded from https://academic.oup.com/jcb/advance-article/doi/10.1093/jcbiol/ruaa102/6128500 by University of Newcastle user on 09 February 2021 Kelly, R.H. & Yancey, P.H. 1999. High contents of trimethylamine oxide 2007. Correlation of trimethylamine oxide and habitat depth within correlating with depth in deep-sea teleost fishes, skates, and decapod and among species of teleost fish: an analysis of causation. Physiological crustaceans. Biological Bulletin, 196: 18–25. and Biochemical Zoology, 80: 197–208. Lacey, N.C., Rowden, A.A., Clarke, M., Kilgallen, N.M., Linley, T.D., Seibel, B.A. & Walsh, P.J. 2002. Trimethylamine oxide accumulation Mayor, D.J. & Jamieson, A.J. 2016. Community structure and diversity in marine animals: relationship to acylglycerol storage. Journal of of scavenging amphipods from bathyal to hadal depths in three South Experimental Biology, 205: 297–306. Pacific trenches. Deep Sea Research Part I, 111: 121–137. Smith, S.I. 1885. On some new or little known decapod Crustacea, from Leitner, A.B., Neuheimer, A.B., Donlon, E., Smith, C.R. & Drazen, J.C. recent Fish Commission dredgings off the east coast of the United 2017. Environmental and bathymetric influences on abyssal bait- States. Proceedings of the United States National Museum. 7: 493–511. attending communities of the Clarion Clipperton Zone. Deep-Sea Spence Bate, C.S. 1881. On the Penaeidae. Annals and Magazine of Natural Research Part I, 125: 65–80. History, Series 5, 8: 169–196. Linley, T.D., Gerringer, M.E. Yancey, P.H., Drazen, J.C. Weinstock, C.L. Stebbing, T.R.R. 1888. Report on the Amphipoda collected by H.M.S. & Jamieson, A.J. 2016. Fishes of the hadal zone including new spe- Challenger during the years 1873–1876. In: Report on the Scientific Results cies, in situ observations and depth records of hadal snailfishes. Deep-Sea of the Voyage of H.M.S. Challenger during the years 1873–76, Zoology, Vol. Research Part I, 114: 99–110. 29 (part 67): i-xxiv, 1–1737, pls. 1–212. Linley, T.D., Stewart, A., McMillan, P., Clark, M., Gerringer, M.E., Vaquer-Sunyer, R. & Duarte, C.M. 2008. Thresholds of hypoxia for Drazen, J.C., Fujii, T. & Jamieson, A.J. 2017. Bait attending fishes of marine biodiversity. Proceedings of the National Academy of Sciences of the the abyssal zone and hadal boundary: community structure, functional United States of America, 105: 15452–15457. groups and species distribution in the Kermadec, New Hebrides and Wenzhöfer, F., Oguri, K., Middelboe, M., Turnewitsch, R., Toyofuku, T., Mariana trenches. Deep-Sea Research Part I, 121: 38–53. Kitazato, H. & Glud, R.N. 2016. Benthic carbon mineralization in Luo, M., Glud, R.N., Pan, B., Wenzhöfer, F., Xu, Y., Lin, G. & Chen, D. hadal trenches: Assessment by in situ O2 microprofile measurements. 2018. Benthic carbon mineralization in hadal trenches: insights from Deep-Sea Research Part I, 116: 276–286. in situ determination of benthic oxygen consumption. Geophysical Weston, J.N.J., Peart, R.A., Stewart, H.A., Ritchie, H., Piertney, S.B., Research Letters, 45: 2752–2760. Linley, T.D. & Jamieson, A.J. 2021 Community structure of scavenging Maynou, F. & Cartes, J.E. 1998. Daily ration estimates and comparative amphipods in the Wallaby-Zenith Fracture Zone: Extending the hadal study of food consumption in nine species of deep-water decapod paradigm beyond subduction zones. Marine Biology [doi: 10.1007/ crustaceans of the NW Mediterranean. Marine Ecology Progress Series, s00227-020-03798-4]. 171: 221–231. Wolff, T. 1960. The hadal community, an introduction. Deep Sea Research, Milne-Edwards, A. 1881. Description de quelques crustacés macroures 6: 95–124. provenant des grandes profondeurs de la Mer des Antilles. Annales des Wolff, T. 1970. The concept of the hadal or ultra-abyssal fauna. Deep Sea Sciences Naturelles, Série 6, 11: 1–15. Research, 17: 983–1003. Papiol, V. & Hendrickx, M.E. 2016a. Community structure of deep-water Yancey, P.H. & Siebenaller, J.F. 2015. Co-evolution of proteins and solutions: decapod crustaceans below the oxygen minimum zone in the protein adaptation versus cytoprotective micromolecules and their roles south-east Gulf of California and analysis of environmental drivers. in marine organisms. Journal of Experimental Biology, 218: 1880–1896. Marine and Freshwater Research, 67: 1862–1879. Yancey, P.H., Gerringer, M.E., Drazen. J.C., Rowden, A.A. & Papiol, V. & Hendrickx, M.E. 2016b. Ecology of Benthesicymus tanneri Jamieson, A.J. 2014. Marine fish may be biochemically constrained (Dendrobranchiata: Benthesicymidae) from the Mexican Pacific slope. from inhabiting deepest ocean depths. Proceedings of the National Academy Journal of Crustacean Biology, 36: 50–60. of Sciences of the United States of America. 111: 4461–4465.