Amphipoda from depths exceeding 6,000 meters revisited 60 years on

,


INTRODUCTION
The first compilation of amphipods living in the hadal zone (depths > 6,000 m) was simply entitled 'Amphipoda from depths exceeding 6000 meters' by Dahl (1959).Based on the findings of the Danish Galathea Expedition, this account reported that despite being a comparatively rich collection, including many new species and depth records, it established 'nothing essentially new' to the existing knowledge of amphipod morphology and taxonomy compared to the works of A. Schellenberg and J.M. Pirlot in the 1920s and 30s, respectively.Based on the 15 species from three genera from Galathea, Dahl (1959) concluded that they conformed well to his existing understanding of structural patterns of common abyssal species.The study ultimately covered 21 species by the inclusion of various descriptions from Swedish and former Soviet Union expeditions in the 1950s.It was noted that although these species belonged to 10 different families, eight of them belonged to the superfamily Lysianassidae.Dahl (1959) also noted a high degree of endemism, with only two of these species known at the time, to reside in more than one trench, an observation that continues to be a central focus for contemporary hadal biology.
Following the coining of the term 'hadal' by the leader of the Galathea Expedition (Bruun, 1956), another member of the expedition published the first concept of a 'hadal community' (Wolff, 1960).Amphipoda was listed as a dominant hadal group, alongside several others, namely holothurians, isopods, and polychaetes.The number of species known was 20-25, with 56% estimated to be endemic (Wolff, 1960).Just ten years later, Wolff (1970) re-evaluated these figures based on many other records supplied by Belyaev (1966) following the Vitjaz expeditions.He reported 17 hadal amphipod species, of which 14 (82.4%) were endemic, and declared that the amphipods were in the 'five significant abyssal-hadal groups' as defined by a high percentage of hadal species (alongside echinoids, polychaetes, bivalves, and holothurians).All these studies were based largely on 'rather ponderous samplers' (Hessler et al., 1978), such as bottom trawling and sediment grabs and dredges, which are now known not to be the optimum method in recovering highly mobile species, such as scavenging amphipods ( Jamieson et al. 2013a).
The relative ease and frequency of sampling and the scale and breadth of studies is such that hadal amphipods have now been used as a model taxon and have resulted in a disproportionate level of knowledge on this group compared to any other hadal group, including those identified by Wolff (1960Wolff ( , 1970) ) as being of 'great ecological significance' .Here, we revisit the Amphipoda from depths exceeding 6,000 meters' (sensu Dahl, 1959) and offer a review contrary to the concept of amphipods originally offering "\'nothing essentially new' , but rather have since provided some of the greatest insights into life at the deepest places in the ocean.
We review herein this important hadal group from how they appear in situ during sampling and hadal exploration, to how they survive the trench environments (pressure adaptation and trophic ecology), how their populations are structured (community structure and life history), inter-trench distribution (phylogenetics and connectivity), their usage in hadal microbiology, and how they provide a warning that the hadal zone is impacted by anthropogenic activity.We finally reflect on outstanding questions and how amphipods can continue to illuminate mysteries of the hadal zone.

BIODIVER SIT Y AND ENDEMISM
In the 50 years since Wolff (1970) reported 17 species of hadal species of amphipods, the number of known species now exceeds 100 (Table 1, Fig. 1).This expansion of known diversity can be attributed to a combination of increased sampling, both geographically and bathymetrically, the subsequently increased number of specimens for comparison and study, and the rise in the application of the integrative taxonomy, whereby morphological assessments are combined with DNA barcoding ( Jażdżewska & Mamos, 2019;Weston et al., 2021a, b).Based on the published and available records to date, amphipods are represented at hadal depths by four suborders, 16 superfamilies, and 28 families, of which 25% are in the superfamily Lysianassoidea (Table 1).The significant families, in terms of diversity and known locations, are Alicellidae, Pardaliscidae, Hirondellidae, and Scopelocheiridae, followed by others such as Eusiridae, Uristidae, Hyperiopsid, and Eurytheneidae, with the latter being quite conspicuous in some trenches, particularly in the upper trench depths (< 7,000 m).Strikingly, 83% of the known species are recorded only from one to two hadal features (Fig. 2).With the expansion of hadal exploration, this statistic has remained relatively constant, and it has given rise to the notion that the hadal zone is home to high levels of endemic diversity.This endemism is considered to be of species found to reside predominantly at hadal depths and/ or at one or several neighbouring hadal features.Yet, endemism or restricted distribution to one hadal feature is not absolute.Several species challenge this notion by being found in multiple trenches, specifically Halice aculeata Chevreux, 1912, Halice quarta Birstein & M. Vinogradov, 1955, Halice subquarta Birstein & M. Vinogradov, 1960, Hyperiopsis laticarpa Birstein & M. Vinogradov, 1955, Metaceradocoides vitjazi Birstein & Vinogradov, 1960 at four, and Hirondellea gigas (Birstein & Vinogradov, 1955) and Princaxelia abyssalis at eight Dahl, 1959 (Fig. 2, Table 2).Some species are recorded from multiple hadal feature among several oceans, specifically, Paralicella caperesca Shulenberger & Barnard, 1976 at six features, Alicella gigantea Chevreux 1899 at seven features, and Paralicella tenuipes Chevreux, 1908 at eight features.Two species are considered to have a global distribution, Hirondellea dubia Dahl, 1959 at eight features (Weston & Jamieson, 2022) and Bathycallisoma schellenbergi (Birstein & Vinogradova, 1958) at 16 features (Weston et al., 2022).With increased sampling, the true distribution of particular species is becoming better known and recent work can address questions of biogeography and population connectivity across the inherently disjunct hadal zone.
While early thoughts on species endemism concentrated on whether there was a distinct hadal fauna, Hirondellea gigas, for example, was noted to be endemic to hadal depths but in multiple trenches from the Northwest Pacific Ocean (Hessler et al., 1978).France (1993) examined the morphologies of H. gigas from the Mariana, Philippine, and Palau trenches (Pacific) and concluded that these geographically isolated populations may have reduced levels of gene flow causing them to diverge morphologically.Patterns emerged within specific genera, such as Hirondellea Chevreux, 1899 whereby the Peru-Chile Trench (Southeast Pacific) had three species (H.sonnei Kilgallen, 2014, H. wagner Kilgallen, 2014, and H. thurstoni Kilgallen, 2014, whereas multiple trenches in the Southwest Pacific only had H. dubia (Lacey et al., 2016;Wilson et al., 2018).Since then, H. dubia is being found in other more distant trenches, such as the Mariana ((Northwest Pacific), South Sandwich (Southern Ocean), and Puerto Rico trenches (North Atlantic), yet still at largely hadal depths (Weston & Jamieson 2022).Other families, such as Pardaliscidae contain four species of Princaxelia Dahl, 1959 at hadal depths, out of which, P. stephenseni Dahl, 1959 is also known from bathyal depths, P. magna Kamenskaya, 1977 is known only from the Yap Trench (West Pacific), P. jamiesoni Lörz, 2010 is known from the adjoined Japan and Izu-Bonin trenches (Northwest Pacific), yet P. abyssalis Dahl, 1959 is known from eight of the West Pacific trenches (Kamenskaya, 1981;1995).Another species, Bathycallisoma schellenbergi, is now known from 16 hadal features spanning four oceans (Kilgallen & Lowry 2015;Lacey et al., 2016;Jażdżewska & Mamos, 2019;Chan et al., 2020;Weston et al., 2021bWeston et al., , 2022;;Jamieson et al., 2022), and all but one record of 5,600 m in the New Hebrides Trench (South Pacific) (Lacey et al., 2016) have been hadal.Population genomics has uncovered that while B. schellenbergi does have a global distribution at the species level, populations are highly restricted to the individual features (Weston et al., 2022), with evidence for limited gene flow between neighbouring features on a shared convergence zone.
Assigning a meaningful value to levels of endemism is difficult due to many specimens having been described either from a single record or those with uncertain identification (see Table 1).Moreover, many records are not corroborated with molecular taxonomy, and therefore some species may have been misidentified, such as those highlighted by Ritchie et al. (2015) that await reclassification.Such was the case with Scopelocheiridae (Kilgallen & Lowry, 2015), or uncovered as a cryptic species complex, like Eurythenes S.I.Smith in Scudder, 1882 and Paralicella caperesca (Havermans et al., 2013;Jażdżewska et al., 2021).With these caveats in mind, of the 119 species listed in Table 1, 95 species (79.8%) are currently recorded as endemic to hadal depths.

IN SITU IMAGERY
Perhaps one of the most striking aspects of hadal amphipods is the visual imagery recorded when the bait is placed on the trench floor in front of a camera (Fig. 3).The first in situ images of hadal amphipods were taken in the Romanche Fracture Zone (Atlantic Ocean) at 7,500 m (Edgerton, et al., 1957;Cousteau, 1958).These two photographs showed amphipods active on the seafloor with other faunal groups, such as ophiuroids, anemones, and isopods.This in situ imagery gave a window into a dynamic hadal community.
The 1970s saw an increase in the use of baited cameras and traps, which revealed often very dense aggregations of deep-sea amphipods at simulated carrion falls (Paul, 1973;Shulenberger & Hessler, 1974).Hessler et al. (1978) first demonstrated the significance of amphipods at hadal depths by deploying a camera with traps to 9,600 and 9,800 m in the Philippine Trench (Pacific).These first glimpses demonstrated their ability to locate bait rapidly, aggregate in large numbers, consume the bait in a matter of hours then disperse, indicating an ecological function of rapidly dispersing organic matter.
One of the first dramatic features of these observations is the gigantism, or large body size, of some species.Hessler et al.Table 2.All known records of amphipods collected from hadal depths including general location, depth, and record reference.Location abbreviations: AT, Aleutian Trench; BT, Banda Trench (also known as Weber Basin); BVT, Bougainville Trench; DFZ, Diamantina Fracture Zone; IBT, Izu-Bonin Trench (also known as Izu-Ogasawara Trench); JT, Japan Trench; JVT, Java Trench (also known as Sunda Trench); KT, Kermadec Trench; KKT, Kuril-Kamchatka Trench; MST, Massau Trench; MT, Mariana Trench; NBT, New Britain Trench; NHT, New Hebrides Trench; NPAC, Pacific Ocean; PCT, Peru Chile Trench; PHT, Philippine Trench; PLT, Palau Trench; PRT, Puerto Rico Trench; SCBT, San Cristobal Trench; SCZT -Santa Cruz Trench, SOT, South Orkney Trench; SST, South Sandwich Trench; TT, Tonga Trench; VT(MT), Volcano Trench (part of Mariana Trench); WZFZ, Wallaby Zenith Fracture Zone; YT, Yap Trench.All names and species authorities are consistent with the accepted classifications available on the World Register of Marine Species (WoRMS; Horton et al., 2017;World Register of Marine Species (WoRMS Editorial Board), 2023).1972) published a record of a very large lysianassoid amphipod from the abyssal plains of the North Pacific (5,304 m) that was twice as big as any known species at that time, estimated to be 282 mm long.A similar, albeit smaller individual, was recorded by Lemche et al. (1976) at 6,758-6,776 m in the New Hebrides Trench.The Hessler et al. (1972) specimens were later identified as the 'supergiant' amphipod Alicella gigantea (Barnard & Ingram 1986), while the Lemche et al. (1976) individual was only 4 cm long and may have been Eurythenes sp. or a juvenile A. gigantea, now both known from that trench (Lacey et al., 2016).An impressively large A. gigantea was later filmed, and recovered, from hadal depths (6,265-7,000 m) in the Kermadec Trench (South Pacific) ( Jamieson et al., 2013b) and is now known to be present at seven hadal features across Pacific, Indian, and Atlantic oceans (Table 2).Characterised by their elongated, articulate body, larger body size, and behaviour that differs from the scavenging lysianassoids, the princaxelids (Pardaliscidae) comprise another easily identifiable hadal amphipod group (Lörz, 2010).
These amphipods predate on the scavenging lysianassoids that gather at bait ( Jamieson et al., 2012), and are present in most trenches studied so far (AJJ, personal observation).
As striking as some of the larger species are on camera, the large aggregations of more typical-sized lysianassoid species, < 20 mm, at the bait is perhaps the most impressive.Since the black and white still images of Hessler et al. (1978), technology has progressed to permit full colour high-definition video and high-resolution still recording of this activity ( Jamieson, 2015).The small body size of these species makes identifications and diversity assessments nearly impossible.The size of the aggregations also increases with time on the bottom, culminating in feeding frenzies that can fill the field of view of a camera making quantifying their numbers difficult (Fig. 3).The time of arrival and time of peak numbers decreases and increases respectively with depth, which results in impressive displays of bait consumption at the deepest parts of a trench (Figs. 4,5).Although imaging these responses is mostly qualitative and unquantifiable, it Bonnierella linearis J.L. Barnard, 1964PCT 6,342-6,328 Belyaev, 1989 SUPERFAMILY: Hadzioidea S. Karaman, 1943(Bousfield, 1983) Bathyceradocus stephenseni Pirlot, 1934BT 7,250-7,340 Belyaev, 1989BVT 6,920-7,652 Belyaev, 1989 Metaceradocoides vitjazi Birstein & Vinogradov, 1960IBT 8,900 Belyaev, 1989JT 6,600-7,370 Belyaev, 1989MT 8,215-8,225 Belyaev, 1989YT 7,190-7,250 Belyaev, 1989 Table 2. Continued demonstrates very effectively their voraciousness and efficiency in the redistribution of organic matter at these extreme depths.

FEEDING AND TROPHIC ECOLOGY
The success of the Amphipoda at hadal depths may also be a virtue of adaptations for a low food environment, as similar for amphipods inhabiting other rare food habitats like Lake Baikal (Mekhanikova, 2010) and subterranean systems (Hutchins et al., 2014).Many of the well-known hadal species are recovered by baited traps and therefore belong to the scavenging guild.
Although not all hadal amphipods are scavengers, with predator species like Princaxelia jamiesoni Lörz, 2010 and Rhachotropis saskia Lörz & Jażdżewska, 2018[Lörz et al., 2018], scavengers in low-food environments, such as the trenches, must adapt to detecting, intercepting, and rapidly consuming large carrion falls, which can be unpredictable both in space and time, while supplementing potentially long intermediate periods with alternative sources of nutrition (Dahl, 1979).Hadal amphipods exhibit effective strategies such as localizing potential food sources, feeding on large muscular food parcels, consuming large quantities of food in relatively short periods of time, storage of this energy for gradual utilization over extended periods of starvation and supplementing their diet with alternative food sources, such as wood (Kobayashi et al., 2012), that become available between large carrion falls.Chemosensory adaptations are evident in many key deepsea amphipods.For example, chemosensory stimulation is used to detect the odour plume emanating from a food fall (e.g., Tamburri & Barry, 1999).These amphipods typically have short and stout antennae, with an array of chemosensory setae on the ventral side of the first flagellar article, which is kept depressed to increase exposure to stimuli.When swimming, these sensors are prominently exposed to the surrounding body of water (Dahl, 1979).Some species sweep water over the proximal part of the antennule, mouthparts, and into the branchial region when beating their pleopods, presumably to maximise the chance of detecting chemical stimuli (Dahl, 1977).
Other methods of food detection may include sound and 'mechano-reception.' Such hydroacoustic stimuli could occur as a large carcass impacts the seafloor or during potentially 'noisy' feeding bouts (Klages et al., 2002).Smith & Baldwin (1984) estimated the spherical spreading of noises created by Eurythenes gryllus s.l.during feeding could potentially produce 15 dB which can be heard for up to 1 km.The augmentation of hydroacoustic stimuli may attract further visitors to the carrion-fall, although chemoreception appears the most likely the primary detection method.
Hadal amphipods are often observed in mass aggregations at baited cameras where bait carcasses are typically stripped of every visible shred of flesh within 10 h.This efficient consumption of food is a result of highly adapted mouth morphology (Fig. 6).The basic gammaridean amphipod has strongly and irregularly serrated incisors with well-developed lacinia mobilis on both mandibles (Dahl, 1979).When biting, the left incisor passes in front of the right incisor which, in turn, moves between the left lacinia mobilis, which then positions in front of the right one.This feeding motion, however, is slightly different for three common hadal amphipod genera (Eurythenes, Hirondellea, and Paralicella Chevreux, 1908).The right incisor slides in behind the left and the 'shape' of the bite is bowl-shaped, which allows these genera to remove larger pieces of food than those with a flattened mandible (Dahl, 1979).Furthermore, Eurythenes and Hirondellea gigas have distinctively shaped molars, such that when they are closed, they form an almost complete funnel from the mouth to the stomach, thought to aid in guiding larger food particles to the digestive tract.Amphipods may also be capable of characterizing, or 'tasting,' the food source upon contact, via chemical and physical gustatory seta on the gnathopods and pereopods (Kaufmann, 1994).This selectivity of food items, or anatomical parts of the food, that differ either chemically or texturally is supported by observations of selective feeding on livers and gonads during necrophagy (Scarratt, 1965), presumably to maximise energetic intake per unit feeding (Kaufmann, 1994).
The enhanced consumption rates are complemented further by their ability to ingest large volumes of food.The alimentary tracts of Eurythenes, Hirondellea, Paralicella, and Bathycallisoma Dahl, 1959 are adapted for the accumulation and storage of large volumes of food relative to their body size.The food is stored in the midgut, which can expand to fill the entire body cavity and are therefore capable of ingesting more food than their shallower water counterparts.Paralicella can extend ventrally to two or three times its body size (Shulenberger & Hessler, 1974;Thurston, 1979).These 'capacious guts' and high consumption rates, enable them to survive prolonged periods of starvation (Sainte-Marie, 1992) (Fig. 6).The diets of most hadal amphipods are not well known, but most of the better-known species are scavengers.The concept of obligate scavengers has been debated for some time (Britton & Morton, 1994;Kaiser & Moore, 1999), and it now seems evident that many hadal amphipods are indeed facultative scavengers.In low-food environments, obligate scavenging is inhibited by the unpredictable disparate nature of carrion-fall occurrence and thus an additional strategy is trophic plasticity, whereby carrion-based scavenging is supplemented with alternative food items.Blankenship & Levin (2007) reported high trophic plasticity of three lysianassoid amphipod species from the Tonga Trench, namely Eurythenes gryllus s.l., Bathycallisoma schellenbergi, and Hirondellea dubia.A combination of gut content metabarcoding and stable isotope analysis further indicated that these amphipods supplement necrophagy with detrivory, predation, and cannibalism.In the absence of carrion Eurythenes gryllus s.l., and Bathycallisoma schellenbergi, for example, were predatory and possibly detrivores, having digested tunicates, ascidians, pelagic salps, or larvaceans as well as other amphipods.Shi et al. (2018) found that Alicella gigantea from the New Britain Trench was dependent on high-quality organic matter such as carrion, whereas Hirondellea gigas from the Mariana Trench and Bathycallisoma schellenbergi from the Massau Trench favored more detritus and bacterial organic matter as supplementary food.
The diet and feeding strategies may also vary by age, depth, and sex.For example, Shi et al. (2020) demonstrated that Alicella gigantea from the New Britain Trench (South Pacific) consumed more bacteria and zooplankton debris in early developmental stages before shifting to carrion consumption, and later stages adopting a more diverse diet of benthic fauna, algae, and carrion.
Females were also found to be more reliant on carrion from higher trophic organisms than males.This strong dependency on surface-derived material, whatever that form may be, was further expanded by Tokuda et al. (2020) who when describing trophic interactions in the Mariana and Kermadec trenches found that amphipods consumed infauna, carrion, terrestrial plant debris, older material from turbidity flows, and perhaps even seep-derived organic material.Functional classification of gut bacteria from Hirondellea gigas also implies heterotrophic utilization of a variety of carbon sources (Zhang et al., 2018).These trophic linkages to the surface are perhaps best demonstrated by Kobayashi et al. (2012), who showed that Hirondellea gigas from 10,897 m in the Mariana Trench, have a unique digestive enzyme capable of digesting wood debris.These studies support the notion that amphipods are one of the most trophically diverse taxa in the marine environment (Nyssen et al., 2002).
In addition to facultative scavenging and trophic plasticity, there are further strategies to overcome what could be long periods before the next major feeding opportunity, metabolic reduction, and long-term energy cache.Energy management is demonstrated by drastically reducing metabolic activity during starvation periods (Smith & Baldwin, 1984).Tamburri & Barry (1999) demonstrated that an abyssal amphipod, Orchomene obtusus G.O. Sars, 1895, could survive without food for four to six weeks in laboratory conditions.This strategy is complemented by rapidly converting the food to lipids (triacylglycerols, wax esters, phospholipids, and diacylglycerol ether) for longterm energy storage (Lee et al., 2006).Upon sequencing the transcriptome of Hirondellea gigas, Lan et al. (2016) suggested that six key genes directly participating in energy metabolites are positively selected suggesting these relate to food utilization and efficient energy storage.This implies a novel genetic adaptation strategy for Hirondellea gigas, at least, to sustain starvation periods when food is scarce.Yayanos & Nevenzel (1978) found that H. gigas specimens from the Philippine Trench had appreciable stores of lipids in their bodies, presumably a large energy cache to cope with long periods of starvation.Lipid concentrations of the necrophagous Eurythenes gryllus s.l.(now E. atacamensis Weston & Espinosa-Leal, 2021[Weston et al., 2021a]) from the Atacama Trench (Eastern Pacific) accounted for 7-18% of the dry weight and were dominated by monounsaturated fatty acids, suggesting that these hadal amphipods are more dependent upon lipid reserves than similar shallower water amphipods (Perrone et al., 2003).
The high lipid content of amphipods gave rise to the 'rising particle hypothesis' (Yayanos & Nevenzel, 1978), which postulated that the high lipid content of benthic amphipods could generate particles that could rise into the water column.They found the lipid content of Hirondellea gigas from 9,600 m in the Philippine Trench to be as high as 26.1% dry weight.While these concentrations of lipids were likely in a solid phase at ambient deep-sea temperatures and pressures, they speculated that lipids could be released to the environment by several mechanisms, including natural decay and animal mortality, cannibalism, predation, and the debris generated during mass feeding bouts.They estimated that lipid particles from 5,000 m depth could reach the surface within a week.Such a mechanism, although never truly investigated, could explain why the largest amphipod, Alicella gigantea, estimated to be 340 mm in length, was an adult female regurgitated by an albatross on Laysan Island in the Hawaiian Islands (Harrison et al., 1983).Given all records of Alicella gigantea are ~5,000 m deep or more (except for one at 1,720 m; Barnard & Ingram, 1986) it is difficult to explain how else a mature adult ended up being consumed by an albatross.
The significance of the hadal Amphipoda goes beyond their ability to consume bait but also their consumption by predatory species, particularly in the upper trenches.Linley et al. (2017) documented the feeding strategies of fishes spanning the bathyal, abyssal, and hadal depths of three Pacific trenches and found that the bathyal and abyssal fish communities were almost exclusively necrophagous.The fish species (mostly Ophidiidae and Macrouridae) that inhabited the abyssal-hadal transition zone comprised predatory and generalist feeders; however, the characteristically hadal fishes, the snailfishes (Liparidae), were exclusively predators.Similar patterns are observed in the large crustacean decapods, most common at the lower abyssal and upper hadal depths.The Decapoda are largely represented by the Penaeidae ( Jamieson et al., 2009;Swan et al., 2021) whereby the shallower abyssal species Cerataspis c.f. monstrosus Gray, 1828 is largely necrophagous while the deeper abyssal and hadal Benthesicymus crenatus Spence Bate, 1881 appears to be entirely predatory.Scavenging in these groups therefore decreases with increasing depth, whereas predation appears to increase, and the preferred prey is the abundance of amphipods.Gerringer et al. (2017a) analysed the stomach contents of hadal snailfishes, from the Mariana and Kermadec trenches.Amphipods comprised 95.8% of the items (84.6% of the weight) of the stomach contents of Notoliparis kermadecensis (Nielsen, 1964) and 97.3% of the items (87.7% of the weight) in Pseudoliparis swirei Gerringer & Linley, 2017 [Gerringer et al. 2017b] from the Mariana Trench.All other sources (decapods, polychaetes, squids, and fish-carrion remnants) were only minor dietary components, which demonstrates the importance of amphipods as a source of food to larger hadal organisms, although they are limited to depths of ~8,100 and ~7,700 m, respectively in the case of fishes and decapods (Yancey et al., 2014;Swan et al., 2021).This means that beyond 8,000 m, hadal amphipods are free from larger predators, which also coincides with a large increase in baited trap yields and the aggregation density filmed by baited cameras.

PRESSURE ADAPTATION
High hydrostatic pressure is perhaps the most conspicuous environmental gradient of the hadal trenches and is a key physical influence on the evolution and distribution of both micro-and macro-organisms (Bartlett, 2002).Piezophily (high-pressure adaptation) is an essential prerequisite for survival at hadal depths as pressure increases by 1 atmosphere (atm) for every 10 m depth ( Jamieson et al., 2010).Hadal amphipods, therefore, occupy depths with ambient hydrostatic pressures from ~600-1,100 atm (~60-110 MPa).Some amphipods, particularly in Hirondelleidae, exhibit extraordinarily large depth ranges.Hirondellea dubia, for example, is known from 4,700-10,817 m (range 6, 117 m) (Weston & Jamieson, 2022).This pressure-tolerant characteristic and ease of capture have made amphipods the model taxa in many key pressure-related physiological studies (MacDonald & Gilchrist, 1980, 1982;Yayanos, 1978Yayanos, , 1981;;Downing et al., 2018).
During decompression (equating to vertical migrations), amphipods are 'relatively hardy in this respect' and the decompression tolerance is species-specific (MacDonald, 1997).Yayanos (2009) recalls how amphipods from the Mariana and Philippine trenches may be capable of significant decompression and that the pressure tolerance of amphipods increases with the depth of capture.Individuals of Paralicella caperesca from 5,900 m were recovered to the surface at ambient pressure (~600 atm) and decompressed to atmospheric pressure.Loss of locomotor activity occurred at 215 atm but was regained following recompression, thus suggesting that this species is capable of over 3,500 m of vertical migration (Yayanos, 1981).Other species such as Eurythenes gryllus s.l.from as deep as 4,000 and 5,900 m have been readily resuscitated by recompression following decompression paralysis (MacDonald & Gilchrist, 1980;Yayanos, 1981).MacDonald (1997) described the effects on shallow-water species when exposed to increasing pressure.In amphipods, relatively slow dorsally directed spasms of the longitudinal musculature are elicited at high pressure.At higher pressure still, a progressive immobilisation begins, eventually culminating in mortality.Some deep-water species, however, exhibit no hyperexcitability at pressures higher than their natural depth range (MacDonald, 1997).MacDonald & Gilchrist (1982) recovered amphipods in a pressure retaining trap at 394-442 atm (~4,000 m) and pressurised the specimens to 700 atm (~7,000 m) and found the amphipods did not convulse, albeit mild hyperexcitability was observed.The failure to convulse at higher pressures suggests that specimens from 4,000 m had a greater pressure tolerance than those captured at 2,700 m.Likewise, Yayanos (1981) demonstrated that hadal amphipods had a greater pressure tolerance than those from 5,800 m, which again suggested that for some amphipod species at least, the deeper their natural depth range, the greater their pressure tolerance.
These high pressures can have profound biochemical perturbations.For example, pressure can disrupt protein conformations and ligand binding and can encourage protein aggregate damage (Yancey, 2020).Downing et al. (2018) analyzed a variety of amphipod species spanning the intertidal to full ocean depths, including the Mariana and Kermadec trenches.They concluded that shallow osmolytes (i.e., taurine, glycine, and betaine) decreased with depth while potential piezolytes or pressure counteractants increased with depth.These piezolytes include trimethylamine oxide (TMAO), glycerophosphorylcholine, proline betaine (all methylamine chaperones), glycerophosphorylethanolamine, and scyllo-inositol, which is a polyol chaperone that differs from methylamines by avoiding damaging protein aggregations.Although TMAO is known for its piezolyte properties (Yancey et al., 2001), the other solutes found in hadal amphipods may explain how they can thrive at depths beyond other organisms, such as teleost fishes (Yancey et al., 2014), cephalopods ( Jamieson & Vecchione, 2020), and decapods (Swan et al., 2021), that appear restricted to maximum depths of ~7,000-8,200 m.
Other mechanisms in the adaptation to high pressure have been found within the exoskeletons of Hirondellea gigas (Kobayashi et al., 2019;Okada et al., 2022).One adaption suggested was an 'aluminum gel shield' found on the surface of the exoskeleton of H. gigas to prevent the dissolution of calcitic exoskeleton below the carbonate compensation depth (Kobayashi et al., 2019); however, this aluminum shield may be an artefact in the analysis of the signals in energy-dispersive X-ray spectroscopic analysis (Okada et al., 2022).Okada et al. (2022) found that contrary to an aluminum accumulation, that many parts of the body, most extensively the gastric ossicles and setae, are enriched with bromine.Further, to limit exoskeleton dissolution below the carbonate compensation depth, magnesium is particularly substituted for calcite.Together, these studies present a window into physical pressure adaptions.
How organisms have adapted to extreme environmental stressors, such as high pressure, is an important and enduring theme in evolutionary biology.Heat-shock proteins (HSPs) are a prominent family of cellular chaperones that are involved in the folding, assembly, and degradation of cellular proteins, cell-cycling, and signal transduction.Ritchie et al. (2018) characterised two key heat-shock protein genes (hsp70 and hsp90) in abyssal and hadal Lysianassoidea and Alicelloidea amphipods.Four phylogenetically distinct isoforms were resolved for both hsp70 and hsp90, with one isoform in each gene being exclusive to the hadal genus Hirondellea.They concluded that increased hydrostatic pressure was important environmental stress that shaped the adaptation of heat-shock protein genes in deep-sea amphipods.DNA damage can occur due to high hydrostatic pressure (Rothschild & Mancinelli, 2001), therefore, hadal organisms may require high frequencies of DNA repair (Dixon et al., 2004).Lan et al. (2017) observed a positive selection in the H. gigas Replication Factor A1 (RFA1) gene, known to aid in DNA replication, DNA repair, and chromosomal stability, implying a role in counteracting the negative effects of high hydrostatic pressure on DNA.

COMMUNIT Y STRUCTURE
Identification of patterns in community structure has been common practise in ecological research for decades.Until the turn of the last century, however, hadal sampling was sporadic and often lacked replication at a given depth, or sufficient sampling from multiple depths within a trench to conclude structural patterns.It was not until Blankenship et al. (2006) developed the methods of Hessler et al. (1978) in the Philippine Trench and repeated them across the entire hadal depth range of the Tonga Trench (central Pacific).Thirteen baited trap hauls between 5,155-10,787 m revealed that the now recognised three main species, all lysianassoids, exhibited distinct vertical zonation ranging from a small overlap of the abyssal Eurythenes spp. in the upper trench to the dominance of large numbers of Hirondellea dubia at the deepest depths, and populations of Bathycallisoma schellenbergi at the mid-depths, with the latter two species being endemic to hadal depths (Blankenship et al., 2006).Further, the H. dubia and B. schellenbergi populations showed the first indications of ontogenetic structuring with depth and were consistent across species (see below).
Such approach was then taken by Jamieson et al. (2011), who focussed on the community structure as it transitioned from the abyssal plains to the trenches.At seven sampling sites in the Kermadec Trench between 4,329 to 7,966 m, they recovered 14 species and demonstrated a marked, statistically significant compositional change (ecotone) across the abyssal-hadal boundary, shifting at depths between 6,007 and 6,890 m, indicating that there is an ecologically distinct amphipod community at hadal depths in this trench.These data were then compared to five stations sampled across similar depths in the Peru-Chile Trench (Eastern Pacific) to test the abyssal-hadal ecocline over two trenches on either side of the South Pacific (Fujii et al., 2013).The abyssal communities were very similar in that they were dominated by cosmopolitan species of Paralicella, Abyssorchomene De Broyer, 1984 and Eurythenes.Two very dissimilar groups were identified, the hadal Kermadec sites (6,890 to 7,966 m) dominated by Hirondellea dubia, and the hadal Peru-Chile sites (7,050 to 8,074 m).While the Kermadec hadal group was very similar to that in the Tonga Trench (Blankenship et al., 2006), the Peru-Chile hadal group was characterised by the presence of Eurythenes and Hirondellea, later described as Eurythenes atacamaensis (Weston et al., 2021a), andHirondellea sonnei, H. wagneri, andH. thurstoni (Kilgallen, 2014).Fujii et al. (2013) demonstrated vast abyssal cosmopolitanism in contrast to trench-specific hadal endemism and that the environmental gradients of depth and geographic isolation (or dispersal distance) best explained these variations in the amphipod assemblage structure.Lacey et al. (2016) expanded both the sampling rate and the bathymetric and geographic reach of this approach by analysing amphipod community composition from the Kermadec Trench (1,490-9,908 m), the New Hebrides Trench (2,000-6,948 m), and the Peru-Chile Trench (4,602-8,074 m).
While hadal ecological research has been focused largely on subduction trenches, non-subduction features like fracture zones and troughs are begin to be explored and studied to access how community structure may be influenced by topography.As an initial investigation of hadal fracture zones, Weston et al. (2021b) analysed the change in community composition from 4,932 m of the Zenith Plateau to 6,546 m at the flat-bed trough of the Wallaby-Zenith Fracture Zone (East Indian Ocean).Eleven species were identified across the abyssal-hadal transition zone, and the community composition gradually changed from an entirely abyssal community to a largely abyssal community with key hadal fauna at hadal depths.This pattern contrasted with the ecotone shift characteristic of subduction trenches, suggesting that the community is influenced not solely by pressure but also by the topography and depositional environments.The work by Weston et al. (2021b) highlights the importance of expanding expeditions to understand the ecosystem role and dynamics of other hadal depth geomorphology features.

LIFE HISTORY
As with all Malacostraca, amphipods are brooders, where females carry their fertilised eggs in an external brood pouch until the fully formed juveniles emerge.Beyond that, there is little understanding of the life cycle and life history of hadal amphipods.Research within the hadal zone has focused on accounting for patterns in the distribution of males, females, and juveniles (ontogenetics).The first account of ontogenetics in hadal amphipods was from Hessler et al. (1978), who identified seven to eight female instars and four male instars within the populations of Hirondellea gigas from 9,600-9,800 m in the Philippine Trench.These instars represent a particular stage of maturity based on primary and secondary sexual characters that exhibit a relatively constant growth ratio.An exception to the growth ratio was found in females, where it decreased markedly at the stage coinciding with an accumulation of reproductive products suggesting that females breed only once with a total fecundity of approximately 97 oocytes per female.These data, however, were from a limited depth range, representing the deepest population of what would turn out to be a species that typically spans the entire hadal depth range.
Vertical ontogenetic stratification across the hadal depth range was later found in the congener of Hirondellea gigas, H. dubia from the Tonga and Kermadec trenches, as well as in Bathycallisoma schellenbergi (Blankenship et al., 2006;Wilson et al., 2018).These two studies found that early instars tended to reside in the shallower depths of each species' depth range, with size and instar (as a proxy for age) increasing with increasing depth.This trend was later found by the expansion of the capture depth of Hirondellea gigas from additional samples in the Izu-Bonin Trench (Eustace et al., 2013), Bathycallisoma schellenbergi in the New Hebrides and Kermadec trenches (Lacey et al., 2018), and Eurythenes atacamensis in the Atacama Trench (Thurston et al., 2002;Eustace et al., 2016;Weston et al., 2021a).Together, these studies suggest that the increased juvenile presence around the upper trench could be driven by reduced competition and relaxation of pressure effects on metabolism.This in turn shows that females descend deeper with age, culminating in a stage of reproduction preparation at the deepest depths (also coinciding with the absence of males).
The vertical ontogenetic stratification has also been found to scale to the topography of the trench (Lacey et al., 2018), i.e., the patterns are the same but the actual depth across which it occurs depends on the depth of the trench.These observations suggests that the population structure is more affected by trench topography and how it may influence resource distribution and ecological interactions, rather than simply by the effects of hydrostatic pressure.Perhaps the most striking example of this condition is Bathycallisoma schellenbergi.This characteristically hadal amphipod has a typical depth range of 6,000 to 9,000 m in the trenches, with distinct vertical ontogenetic stratification (Blankenship et al., 2006;Lacey et al., 2016).Weston et al. (2021b), however, found them in a hadal feature that reached only 6,625 m in depth, the Wallaby-Zenith Fracture Zone.The species was found within a very narrow depth range of 6,537-6,546 m in the fracture zone, where the population showed no apparent stratification as it comprised a mix of both adults and juveniles.This distribution revealed that a population that normally spans over 3,000 m in depth range was residing in a shallower, albeit hadal, feature in the bottom 100 m, which suggests that while the minimum depth of occurrence could be driven by hydrostatic pressure, the maximum depth and ontogenetic structuring therein, are not, and are more likely linked to topography.

VISION AND COLOUR
Perhaps one of the most striking characteristics of freshly recovered amphipods is their eyes (Senna, 2009;Narahara-Nakano et al., 2018).The often vividly yellow or orange sessile compound eyes typically occupy a large surface area of the head, extending from the dorsal to the ventral side as an irregular band (Hallberg et al., 1980) (Fig. 7).Eyes are, however, quickly lost upon preservation in ethanol and their descriptions are therefore lacking in many species (see Stoddart & Lowry, 2004;Horton & Thurston, 2013).Eye morphology can be characterised by the constellation and consistency of several features.There are four main sub-categories of eyes: the ampeliscid, hyperiid, gammarid, and lysianassid types, with the latter being the most common at hadal depths (Hallberg et al., 1980).The lysianassid type has a reduced or no dioptric apparatus and hypertrophied (enlarged) rhabdoms, therefore differing from that of littoral species (Strauss, 1926as quoted in Hallberg et al., 1980).Few studies on the eye morphology of deep-water amphipods have been undertaken, but for Eurythenes gryllus s.l.there are at least no traces of the crystalline cone, except for some flattened cells distal to the rhabdom that may represent the remnants of the cone cells (Hallberg et al., 1980).While these organs do not appear capable of generating an image, they may provide crude directional information given that they extend dorsally to ventrally.The organs, however, appear to be adapted to detecting low ambient light intensities (Hallberg et al., 1980).
Like many crustaceans, amphipods are often orange-pinkred pigmentation except for species, such as Alicella gigantea, or genera, such a Princaxelia, that appear uniformly white.This colouration is a result of carotenoid pigments, which are synthesised by higher plants, mosses, bacteria, algae, and fungi (Gaillard et al., 2004) and acquired by amphipods through their diet, likely through phytoplankton-derived particulate organic matter (Repeta & Gagosian, 1984).These carotenoids serve an important function in deep and low-light environments in terms of camouflage.Given the quick attenuation of red wavelengths through water, a uniformly red exoskeleton provides valuable camouflage against predators and exposure by illumination from bioluminescent light ( Johnsen, 2005).Many species of amphipods vary across a spectrum of the white-orange-pink-red and sometimes deep red, with individuals of the same species from the same haul, or filmed in situ, being quite different (e.g., white vs. deep red) (Fig. 8).Earlier studies of the abyssal Eurythenes gryllus in the North Pacific indicated an increase in colour intensity with age and sexual maturation (Smith & Baldwin, 1984;Charmasson & Calmet, 1987).Similarly, Arctic populations of E. gryllus showed an increase in colour intensity correlated with body size (Thoen et al., 2011).The amount of pigmentation continues to increase as growth ceases or slows down at a certain size or stage, but variations in pigment composition are probably affected by both diet and moulting/growth patterns or other factors like metabolic pathways or reproduction (Thoen et al., 2011).The absence of this pigmentation in Alicella gigantea and Princaxelia spp.may be a result of a lack of predation with the former, and the latter consisting of predatory species.

MICROBIOME AND PAR A SITISM
The first study of bacteria associated with hadal Amphipoda came from a specimen of Hirondellea gigas from 10,476 m in the Mariana Trench (Yayanos et al., 1981).The specimen was incubated at close to ambient pressure and temperature where the bacterial strain 'MT-41' (later assigned to Colwellia; DeLong et al., 1997) was isolated and partially characterised and deemed an obligately piezophilic bacterium.The generation time of 'MT-41' was three times longer than a similar abyssal strain, 'CNPT-1.' The 'CNPT-1' abyssal strain did not grow at pressures beyond ~8,500 m, while the 'MT-41' strain ceased growing at pressures less than those found shallower than ~5,000 m.DeLong et al. (1997) also demonstrated that samples from 3,500 m at the rim of the Mariana Trench, just 28 km from the 10,476 m 'MT-41'sample site, also could not survive beyond 8,500 m.This observation suggested for the first time that the microbiome of the hadal bacteria may be as endemic to hadal depths as the amphipod itself.A similar strain, Colwellia marinimaniae strain 'MTCD1 T ' was more recently isolated using the same amphipod species from the Mariana Trench and was found to be the most piezophilic organism ever described, with a growth pressure range of 80-140 Mpa and an optimum of 120 Mpa at 6 ˚C (Kusube et al., 2017).Strains were also isolated from Hirondellea dubia from the Kermadec Trench (South Pacific) from a depth of 9,856 m (Lauro et al., 2007).Both isolates, 'KT27' and 'KT99,' , were piezophilic and represented the first psychropiezophiles from the Southern Hemisphere.The closest cultured relatives of 'KT27' and 'KT99' were the isolates Colwellia sp.strain 'MT-41' (Yayanos, 1986) and Shewanella sp.strain 'PT99' (DeLong et al., 1997), respectively.
Gut microbiotas are known to play a crucial role in organismal health, including pathogen defence, environmental adaptation, and digestion (Cheng et al., 2019), and are becoming increasingly more appreciated in their critical role in assisting the host to adapt to hadal environments (He et al., 2018;Zhang et al., 2018Zhang et al., , 2019)).Studies are emerging on the gut microbiota of hadal amphipods, again centred on Hirondellea gigas from the Mariana Trench.The most dominant bacterium was found to be Psychromonas sp., of which Zhang et al. (2018) successfully recovered a nearly complete genome (designated 'CDP1').This strain was found to have a reduced genome, consistent with the hypothesis that in some ecosystems the bacteria are subject to strong selection to minimise the material costs of growth (Giovannoni et al., 2005).The genome reduction in CDP1 was posited to reflect a long-term series of strategic adaptations within the gut of Hirondellea gigas through a long history of coevolution.
A comparative analysis of the gut microbiota between Hirondellea gigas from the Mariana and Japan trenches revealed significant differences in the composition of the gut microbes such as variation in abundance of Psychromonas, Propionibacterium, and Pseudoalteromonas species of bacteria (Zhang et al., 2019).The comparative metagenomics suggested that the gut microbial compositions of the two populations have undergone significant divergence and might be a result of selective factors, such as environmental variations and microbial interactions.Cheng et al. (2019) compared the gut microbial composition and diversity of two distinct species of amphipods, Hirondellea gigas and Halice sp. from the Mariana Trench.The genera 'Candidatus Hepatoplasma' dominated overall, followed by Psychromonas in Hirondellea gigas and Psychrobacter in Halice sp.These two hadal species shared the predominant symbiont of 'Ca.Hepatoplasma' but were classified into two different operational taxonomic units (OTU), suggesting host-specific colonization of symbionts.In terms of diversity and abundance, there was a significant divergence of gut microbiota found between the two species, both which came from the same environment.Comparison between the microbiomes of Alicella gigantea, Bathycallisoma schellenbergi, and Hirondellea gigas from the New Britain Trench further strengthens the evidence that gut microbiomes at hadal depths are species-specific (Chan et al., 2021).Whether the host-microbe associations are a result of vertical transmission (where microbes pass from the parents and not from the environment) or horizontal transmission (acquisition from the environment) is yet to be resolved (Zhang et al., 2018).
Records of parasitic nematodes at hadal depths are rare and mostly known from free-living adult stages living in the sediments (Leduc, 2015).Leduc & Wilson (2016), however, described the benthimermithid nematode Trophomera cf.marionensis from the body cavity of Hirondellea dubia from the Kermadec Trench at depths of 7,018 to 10,005 m.Although morphologically identical, the nematodes were about twice the length of Trophomera marionensis recorded from shallower areas.The prevalence of these parasites among the Hirondellea dubia populations was estimated to be < 1 %.The ecological role that these parasitic interactions play is largely unknown.

ANTHROPOGENIC SIGNATURES
The hadal zone is remote, understudied, and rarely visited, and as such one might easily assume it to be pristine and free from signatures of anthropogenic impacts.In recent years, however, hadal amphipods have been used in multiple studies as a model taxon in which to demonstrate the extent of anthropogenic contamination at extreme depths.As in shallower systems, amphipods function as biomonitors for anthropogenic influences, such as heavy metals.Reid et al. (2018) provided an initial baseline for examining trace-metal variability.Amphipods can also be used for a signal of contamination.The range of contamination sources is as alarmingly widespread as contamination levels are high.For example, Wang et al. (2019), for example, detected 'bomb 14 C' radiocarbon in Hirondellea gigas and Alicella gigantea, a radiocarbon resulting from thermonuclear tests during the 1950s and 1960s that doubled the amount of 14 C in the atmosphere (Hua & Barbetti, 2004).The bomb 14 C is mixed with ocean and terrestrial carbon pools that are then transported to the trenches via particulate organic matter (POM).Similarly, Oguri et al. (2013) demonstrated that the cesium ( 137 Cs) fallout following the 2011 Tohoku-Oki Earthquake and Tsunami induced Fukushima Daiichi nuclear disaster could reach the Japan Trench communities in < 100 d as it binds to, and descends, with POM.While it is not yet been recorded in amphipods, it appears 137 Cs is likely to be present, as it was shown to have penetrated both the POM and seafloor sediment.
Other anthropogenic pollutants include persistent organic pollutants (POPs), such as polychlorinated biphenyls (PCBs, used as dielectric fluid) and polybrominated diphenyl ethers (PBDEs, used as flame retardants).POPs are highly detrimental to organismal health through their endocrine disrupting properties (Rhind, 2012) and possess an inherent hydrophobicity therefore a high binding affinity to POM descending in the water column.They also have inherent lipophilicity, so these compounds readily bioaccumulate in organisms, with cumulative increases at each trophic level (Ballschmiter et al., 1983).Both man-made PCBs and PBDEs were detected in all amphipods studied across the entire depths range of both the Mariana and Kermadec trenches (Hirondellea gigas, H. dubia, and Bathycallisoma schellenbergi, respectively; Jamieson et al., 2017).The Mariana Trench amphipods had the highest levels of PCBs (147.3-905ng g −1 lw), which were fifty times more contaminated than crabs from paddy fields fed by the Liaohe River, one of the most polluted rivers in China, while the Kermadec Trench amphipods showed very high levels of PBDEs (13.75-31.02ng g −1 lw).In addition to these POPs, Cui et al. (2020) also detected high concentrations (up to 1343 ng g -1 lw) of chlorinated pesticides, such as dichlorodiphenyltrichloroethanes (DDT) and chlordanes as well as four brominated natural products (BNPs) that resembled methoxylated brominated diphenyl ethers or polybrominated biphenyls, from specimens of Hirondellea gigas, Bathycallisoma schellenbergi, and Alicellea gigantea from the Mariana, Mussau (Southwest Pacific), and New Britain trenches, respectively.
Monomethyl mercury (MMHg) is a globally distributed neurotoxic pollutant also known to biomagnify in marine food webs.Mercury isotope measurements in hadal amphipods have been used to demonstrate the extent to which surface-ocean-derived mercury has infiltrated the hadal zone (Sun et al., 2020).When surface-dwelling organisms and POM sink, large amounts of mercury are transferred to the underlying trench fauna, as demonstrated by high concentrations of MMHg in amphipods from both the Mariana and Kermadec trenches (1.49‰ ± 0.28, N = 13 and 1.57‰ ± 0.14, N = 12, respectively; Blum et al., 2020).
The occurrence of microplastics (< 5 mm in size) in the deep sea has seen a rapidly growing realisation (Van Cauwenberghe et al., 2013, Taylor et al., 2016).Peng et al. (2018) documented microplastic abundances in hadal sediments of the Mariana Trench varying from 200 to 2200 pieces per litre, values that are much higher than those in typical deep-sea sediments, making it available through detritivory (Blankenship & Levin, 2007).Jamieson et al. (2019) detected ingested microplastics in the hindguts of lysianassoid amphipod populations in six trenches around the Pacific Rim ( Japan, Izu-Bonin, Mariana, Kermadec, New Hebrides, and Peru-Chile trenches), at depths ranging from 7,000 m to 10,890 m.They found that 72% of individuals examined across all sites contained at least one microparticle.The number of microparticles ingested per individual across all trenches ranged from 1 to 8. The average microparticle per specimen varied per trench, from 0.9 ± 0.4 SD (New Hebrides Trench) to 3.3 ± 0.7 SD (Mariana Trench).
One species encapsulates the sober fact of the extent of marine plastic pollution in the hadal zone, Eurythenes plasticus Weston, 2020 [Weston et al., 2020].A new hadal amphipod species from 6,010 and 6949 m deep in the Mariana Trench was found to have ingested polyethylene terephthalate (PET) and was named 'plasticus' to highlight the extent of the plastic problem (Weston et al., 2020).Eurythenes plasticus was used in a campaign to raise awareness about the problem of marine plastics in collaboration between Newcastle University (UK), Worldwide Fund for Nature (Germany), and the advertising agency BBDO-Dusseldorf (Fig. 9).The 2020 campaign led to 120,000 signatures across 94 countries on a petition to end marine plastics, > 500,000 visitors to exhibitions in UK and German museums and amassed a global reach of 1.4 billion people.It was awarded the Guinness World first for the first new species already found to have been contaminated with plastic as of 2021.

HADAL A MPHIPODS -60 YE AR S ONWARDS
The ability to sample such high numbers and species of any hadal organismal group still makes the Amphipoda highly significant in the study of hadal ecosystems.While many questions have been answered, there is still a wealth of knowledge to be gained and questions that could be addressed over the next 60 years, ranging from taxonomy to the biology of individual species to global population and evolution.As new technology is developed and expeditions intensify, the work already done on the hadal Amphipoda should serve as a framework to be applied to other crustacean orders and invertebrate phyla.Such a multi-disciplinary approach to other taxa, exemplified by the work on hadal amphipods, will be required to disentangle trends across multiple hadal habitats, species, higher taxonomic ranks, and functional groups.

DIVER SIT Y AND TA XONOMY
Only a fraction of the hadal zone has been explored, and similarly, as this review demonstrates, only a fraction of specimens recovered from hadal depths have a scientific name.Specifically, 71 species are assigned to a valid genus, while at least another 40 records have uncertain identification (Table 1).This gap represents both a taxonomic impediment and highlights that much work is still required for fundamental taxonomy and the study of biodiversity in the hadal zone.This gap is the widest for pelagic and non-benthic scavenger amphipods, which is largely derived from the predominate use of landers.As more technologies, such as suction samplers on submersibles, come available and are deployed to hadal depths our understanding of the occurrence and distribution of non-scavenging fauna can be expected to expand.
Once specimens are on land, focus on fundamental taxonomy remains necessary for the progression of the understanding of hadal biology, particularly to provide identitifications, understand distributions, unravel biogeographic patterns, understand evolutionary histories and relationships, discover new biodiversity, and track long-term changes to anthropogenic stressors.One approach to supporting this progression is by applying an integrative taxonomic approach, of coupling morphological taxonomy with DNA barcoding, to gain robust identifications (see Page et al., 2005;Padial et al., 2010).This approach is being well applied to document diversity at hadal features (e.g., Jażdżewska & Mamos, 2019;Weston et al., 2021b), resolve identities within the cryptic genera of Eurythenes (e.g., Havermans et al., 2013;Weston et al., 2021a) and Paralicella (e.g., Jażdżewska et al., 2021), and illuminate cosmopolitan distributions within the species of Bathycallisoma schellenbergi (e.g., Weston et al., 2020) and Hirondellea dubia (e.g., Weston & Jamieson, 2022).While successful DNA barcoding is not possible for all specimens, potentially due to preservation type, age, substantial degradation during repair, or a myriad of other possibilities, future works should prioritise an integrative taxonomic approach to overcome taxonomic impediments and achieve robust identifications that provide the foundation for other questions.

Physiology and chronobiology
Marine organisms, much like their terrestrial counterparts, exhibit cyclic behaviour that has been the focus of chronobiological research for many years (Naylor, 1996, Last et al., 2009).The marine environment is subject to tidal (~12.4 h), daily (~24 h), lunar-day (~24.8 h), semi-lunar (~15 d), and lunar cycles (~28 d) that constitute only the main environmental cycles present at full ocean depth (Taira et al., 2004).While circadian clocks are principally governed by the light/dark cycle, tidal clocks are entrained by a suite of cycles associated with the predictable inundation and exposure of the shore.These tidal cycles (unlike circadian) can include hydrostatic pressure, among other parameters (Reid & Naylor, 1990).
On a cellular level, pressure changes of less than 20 kPa (equivalent to ≤ 2m depth), trigger perturbations in cell growth and/ or metabolism (MacDonald & Fraser, 1999).It is unclear how amphipods sense pressure.Identification of hydrostatic pressure receptors in marine organisms without a gas phase has proved elusive until it was found that thread hair receptors in the statocyst of decapod Crustacea responded to small changes in hydrostatic pressure (Fraser & MacDonald, 1994;Fraser, 2001;Fraser  & Takahata, 2002).It is now presumed that amphipods detect pressure using statocysts much in the same way as decapod crustaceans (Platvoet et al., 2006) and are therefore likely to detect minute variations in hydrostatic pressure.The location of a statocyst nevertheless remains elusive within hadal amphipods.
If the statocysts are sensitive enough to detect tidal fluctuations present at depth ( Jamieson et al., 2020), are the amphipods exhibiting chronobiological behaviour?Observations have shown that in some deep-sea organisms there is an apparent synchronisation of activity (Guennegan & Rannou, 1979;Wagner et al., 2007).Amphipod fluctuations at 4,855 m in the North Atlantic have been attributed to current and pressure changes, although these parameters were not measured (Thurston, 1979).Synchronization of an organism's endogenous biological oscillators, of tidal and/or daily to annual periodicities, is fundamental to optimal physiological performance, life history, and ultimately survival.How hadal amphipods can naturally detect and respond to small fluctuations such as tidal cycles or even large seismic events and perturbations is an important but unresolved issue.

Reproduction and life history
All studies that have examined the life histories of hadal amphipods across a wide depth range of a trench have identified a clear pattern of juveniles at the shallower depths and mature females at the deeper depths (Blankenship et al., 2006;Eustace et al., 2013;Lacey et al., 2018;Weston et al., 2021a).As the females descend deeper with age, culminating in the reproduction preparation stage at the deepest depths, they are seldomly recovered in baited traps as they likely no longer feed to prevent the expulsion of eggs.This results in the last stage of the females going unsampled using the most successful sampling methods for amphipods.This prevents a 'closed loop' analysis of the reproduction cycle that leaves large unresolved questions regarding how the juveniles end up at the shallower depth of the trench, which can be nearly 5,000 m shallower than the deepest point where the large females were last reported.Do these brooding females suddenly ascend back up 5,000 m to the top of the trench to release their brood, or is there a pelagic stage and they float up as larvae?Or do the juveniles hatch and make the ascent themselves?As challenging as all these options appear to be, some mechanism must exist to explain the gap in observations.
Other intriguing questions in the study of the life history are the age and life span of species and the timing of reproduction.While we know they make the depth-maturity descent during their lifetime, there are currently no estimates of how long this cycle takes to complete.Ingram & Hessler (1987) estimated that the Eurythenes gryllus s.l.(Lichtenstein in Mandt, 1822) age to maturity to be 7.9-12.8y for females and 3.5-5.6y for males based on instar analysis calculation; however, this remains unresolved at hadal depths.Gerringer et al. (2018) posited that the relatively young age of reproductively mature hadal snailfishes may be due to long-lived species more typical of the relatively stable aseismic abyssal plains being highly susceptible to disturbances in seismically active trenches.Perhaps the reason for a sudden shift in amphipod diversity in the upper trenches is related to population resilience to seismically induced disturbance events such as burial or smothering, following a mass wasting event (Oguri et al., 2013).

Ecology beyond the trenches
Although hadal amphipods are principally known for inhabiting large subduction trenches, other geomorphologies are hadal but not trenches.Jamieson & Stewart (2021) analysed bathymetric data from the Northwest Pacific where a large majority of hadal features occur and concluded that trenches only accounted for 32% of the 2.1 million km 2 of seafloor greater than 6,000 m.The remaining comprised basins and fracture zones, albeit generally shallower.Weston et al. (2021b) reported finding quintessentially hadal amphipods within the bottom 100 m of a flat-bottomed fracture zone that was only 6,625 m deep.Bathycallisoma schellenbergi is well documented from the larger subduction trenches (Weston et al., 2022) and typically exhibits ontogenetic stratification (Lacey et al., 2018), which it did not in this shallower fracture zone.Similarly, the hadal species Hirondellea gigas has been found in in the Sui Shin Hole, an equally shallow location in the Philippine Plate (AJJ, unpublished data).These observations suggest that these amphipods are indeed 'hadal' amphipods and not 'trench' amphipods, and much of their global footprint may well be in non-trench hadal zones such as basins and fracture zones that are relatively under-sampled.

Connectivity and speciation
The hadal zone represents an intriguing seafloor landscape, as inverted island-like habitats.Recent population genomic analysis is dually highlighting that the seafloor separating hadal features does pose reproductive barriers to population connectivity but also that the distribution of hadal fauna is more complex than one species per trench.For example, Weston et al. (2022) showed that Bathycallisoma schellenbergi has a cosmopolitan distribution at hadal depths, but populations were highly restricted to individual features with only limited gene flow between topographically connected features.Weston & Jamieson (2022), however, showed that Hirondellea dubia, while largely present in the Southwest Pacific, is not present in the Indian Ocean but is present in the Southern and North Atlantic Oceans, with evidence of co-inhabiting the Mariana Trench with the congener H. gigas.Hirondellea gigas, does not appear to exhibit such a global distribution and is so far known to be confined to the Northwest Pacific.All three species appear to not inhabit the Atacama Trench where three other Hirondellea species are known (Kilgallen, 2014) and a closely related potential cryptic linage of Bathycallisoma schellenbergi (Weston et al., 2022).These results represent three dominant hadal species from two genera, none of which exhibit the same biogeographic pattern.This suggests that there is more research to be done with regards to population connectivity and phylogeography which would benefit from considerations of the nontrench hadal features and life history stages described above and historical conditions of the deep ocean that may provide insight into how and when an individual hadal feature is being populated and are connected.

Omics and the hadal zone
The hadal zone represents an extreme point for understanding rules of the life and the final terminus for pathways in the deep ocean.With abundant diversity, recovery in high numbers, and shallower comparative species, hadal amphipods represent model taxa to determine the fundamentals of cellular organization, evolutionary adaptions, and protein functions required to thrive at full ocean depth.At present, the number of available 'omics resources are in its infancy, with only three mitogenomes (Lan et al., 2016;Li et al., 2019a, b) and three transcriptomes (Lan et al., 2017;Li et al., 2021) available.One of the limits to the number of genomic resources is the extreme estimated size of genome size hadal amphipods, ranging from 4-34 Gb (Ritchie et al., 2017b).With the exponential rise in next-generation sequencing technologies, particularly with long-read capabilities, coupled with the steady decline in sequencing costs and increase in bioinformatic capabilities, genomic, transcriptomic, and epigenetic research of hadal amphipods is primed for explosive growth over the next several decades.This will enable the continued discovery of hadal diversity, novel pressure adaption, and genetic resources to uncover how and when life evolved to the hadal zone.

CONCLUSIONS
Since their initial discovery in the hadal zone, amphipods have become a model taxon in understanding life at great depth.While Dahl (1959) mused that hadal amphipods established 'nothing essentially new,' they have since become central to contemporary hadal science, from anthropogenic contaminants to evolutionary genetics.Their extraordinary responses to artificial food-falls, and ease of capture in large numbers have put amphipods at the centre of many hadal science studies.This ecological trait has resulted in a disproportionately greater number of studies, discoveries, and insights than the others in the 'five significant abyssal-hadal groups' reported by Belyaev (1966).It is therefore arguable that the Amphipoda are any more important ecologically than other hadal groups but have been and will likely continue to be highly important scientifically.

Figure 1 .
Figure 1.Hadal amphipods.Alicella gigantea from 7094 m in the Mariana Trench (A).Bathycallisoma schellenbergi from 8,370 m in the Puerto Rico Trench (B).Stegocephalidae gen.sp.from 8,380 m in the Puerto Rico Trench (C).Eurythenes magellanicus (H.Milne Edwards, 1848) from 8094 m in the Mariana Trench (D).Hirondellea gigas from 10,936 m in Mariana Trench (E).Hyperiopsis laticarpa from 10,936 m in the Mariana Trench (F).Undescribed species of Stephonyx from 8380 m in the Puerto Rico Trench (G).Scale bars are 1 cm.

Figure 2 .
Figure 2. Most amphipod species have been found at one hadal feature, thus driving the concept of hadal endemism.

Figure 3 .
Figure 3. Large aggregation of Hirondellea dubia at 8,000 m in the Kermadec Trench after one hour on the seafloor (white reference cross is 50 cm) (A).Remains of the bait after 8 hours on the seafloor showing the efficiency with which amphipods can remove flesh from carrion (B).Alicella gigantea, a.k.a.'the supergiant,' feeding at 5,900 m in the Mariana Trench (C), and with large aggregations of Bathycallisoma schellenbergi in the San Cristóbal Trench (7,220 m) (D).Princaxelia cf.magna at 7,542 m in the Mariana Trench (E).

Figure 4 .
Figure 4.In situ images showing increasing numbers of scavenging amphipods with increasing depth, from the Tonga Trench (left) and the Peru-Chile Trench (right), all images are from three hours after landing on the seafloor.

Figure 5 .
Figure 5.Time course of bait consumption, exclusively by amphipods, mostly Hirondellea dubia at 7,000 m in the South Sandwich Trench.Scale bar = 20 cm.

Figure 6 .
Figure 6.Feeding adaptations of hadal Bathycallisoma schellenbergi: The expansion of the capacious gut following feeding (A), the feeding morphologies (B), and the mandibles as seen using a low vacuum environmental scanning electron microscope (SEM) (C-D).

Figure 8 .
Figure 8. Colour variation in freshly caught amphipods: Eurythenes atacamensis from 8,074 m in the Atacama Trench (A-D), Bathycallisoma schellenbergi from 7,000 m in the South Sandwich Trench (E, F), and Hirondellea gigas from 10,900 m in the Mariana Trench (G, H).

Table 1 .
Summary of records of hadal amphipod to family level, including the number of records and locations.Numbers in parentheses represent uncertain identifications.