Submersible- and lander-observed community patterns in the Mariana and New Britain trenches: Influence of productivity and depth on epibenthic and scavenging communities

Abstract

Deep-sea trenches remain one of the least explored ocean ecosystems due to the unique challenges of sampling at great depths. Five submersible dives conducted using the DEEPSEA CHALLENGER submersible generated video of undisturbed deep-sea communities at bathyal (994 m), abyssal (3755 m), and hadal (8228 m) depths in the New Britain Trench, bathyal depths near the Ulithi atoll (1192 m), and hadal depths in the Mariana Trench Challenger Deep (10908 m). The New Britain Trench is overlain by waters with higher net primary productivity (∼3-fold) than the Mariana Trench and nearby Ulithi, and receives substantially more allochthonous input from terrestrial sources, based on the presence of terrestrial debris in submersible video footage. Comparisons between trenches addressed how differences in productivity regime influence benthic and demersal deep-sea community structure. In addition, the scavenger community was studied using paired lander deployments to the New Britain (8233 m) and Mariana Trenches (10918 m). Differences in allochthonous input were reflected in epibenthic community abundance, biodiversity, and lifestyle representation. More productive locations were characterized by higher faunal abundances (∼2-fold) at both bathyal and hadal depths. In contrast, biodiversity trends showed a unimodal pattern with more food-rich areas exhibiting reduced bathyal diversity and elevated hadal diversity. Hadal scavenging communities exhibited similar higher abundance but also ∼3-fold higher species richness in the more food-rich New Britain Trench compared to the Mariana Trench. High species and phylum-level diversity observed in the New Britain Trench suggest that trench environments may foster higher megafaunal biodiversity than surrounding abyssal depths if food is not limiting. However, the absence of fish at hadal depths suggests that certain groups do have physiological depth limits. Submersible video footage allowed novel in situ observation of holothurian orientation, jellyfish feeding behavior as well as lifestyle preferences for substrate, seafloor and overlying water. This study documents previously undescribed species in the New Britain Trench, including an ulmariid scyphozoan (8233 m) and an acrocirrid polychaete (994 m), and reports the first observation of an abundant population of elpidiid holothurians in the Mariana Trench (10908 m). It also provides the first megafaunal community analysis of the world's deepest epibenthic community in the Mariana Trench Challenger Deep, which was composed of elpidiid holothurians, amphipods, and xenophyophores.

Full text

Submersible- and lander-observed community patterns in the Mariana
and New Britain trenches: Influence of productivity and depth
on epibenthic and scavenging communities
Natalya D. Gallo
a,
n
, James Cameron
b
, Kevin Hardy
c
, Patricia Fryer
d
, Douglas H. Bartlett
e
,
Lisa A. Levin
a
a
Center for Marine Biodiversity and Conservation, Scripps Institution of Oceanography, University of California San Diego, La Jolla, CA 92093-0218 USA
b
Blue Planet Marine Research Foundation, 16255 Ventura Blvd. Suite 525, Encino, CA 91436 USA
c
Global Ocean Design, 7955 Silverton Ave., Suite 1208, San Diego, CA 92126 USA
d
SOEST/HIGP University Hawaii at Manoa, POST Bldg. #503 1680 East-West Rd. Honolulu, HI 96822 USA
e
Marine Biology Research Division, Center for Marine Biotechnology and Biomedicine, Scripps Institution of Oceanography, University of California, San
Diego, La Jolla, CA 92093-0202 USA
article info
Article history:
Received 11 April 2014
Received in revised form
24 December 2014
Accepted 29 December 2014
Available online 14 January 2015
Keywords:
Hadal zone
Human occupied vehicle
Baited lander
Trench ecology
Megafauna
Submersible
Biodiversity
Benthos
Challenger Deep
abstract
Deep-sea trenches remain one of the least explored ocean ecosystems due to the unique challenges of
sampling at great depths. Five submersible dives conducted using the DEEPSEA CHALLENGER submersible
generated video of undisturbed deep-sea communities at bathyal (994 m), abyssal (3755 m), and hadal
(8228 m) depths in the New Britain Trench, bathyal depths near the Ulithi atoll (1192 m), and hadal
depths in the Mariana Trench Challenger Deep (10908 m). The New Britain Trench is overlain by waters
with higher net primary productivity (!3-fold) than the Mariana Trench and nearby Ulithi, and receives
substantially more allochthonous input from terrestrial sources, based on the presence of terrestrial
debris in submersible video footage. Comparisons between trenches addressed how differences in
productivity regime influence benthic and demersal deep-sea community structure. In addition, the
scavenger community was studied using paired lander deployments to the New Britain (8233 m) and
Mariana (10918 m) trenches. Differences in allochthonous input were reflected in epibenthic community
abundance, biodiversity, and lifestyle representation. More productive locations were characterized by
higher faunal abundances (!2-fold) at both bathyal and hadal depths. In contrast, biodiversity trends
showed a unimodal pattern with more food-rich areas exhibiting reduced bathyal diversity and elevated
hadal diversity. Hadal scavenging communities exhibited similar higher abundance but also !3-fold
higher species richness in the more food-rich New Britain Trench compared to the Mariana Trench. High
species- and phylum-level diversity observed in the New Britain Trench suggest that trench environ-
ments may foster higher megafaunal biodiversity than surrounding abyssal depths if food is not limiting.
However, the absence of fish at our hadal sites suggests that certain groups do have physiological depth
limits. Submersible video footage allowed novel in situ observation of holothurian orientation, jellyfish
feeding behavior as well as lifestyle preferences for substrate, seafloor and overlying water. This study
documents previously unreported species in the New Britain Trench, including an ulmariid scyphozoan
(8233 m) and an acrocirrid polychaete (994 m), and reports the first observation of an abundant
population of elpidiid holothurians in the Mariana Trench (10908 m). It also provides the first
megafaunal community analysis of the world's deepest epibenthic community in the Mariana Trench
Challenger Deep, which was composed of elpidiid holothurians, amphipods, and xenophyophores.
&2015 Elsevier Ltd. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Historically, information about the ecology of the deep sea has
been gained through the specimens recovered (frequently damaged)
from bottom trawls and grab samples (Belayev, 1989), and through
seafloor photographs taken by underwater cameras (Heezen and
Hollister, 1971;Lemche et al., 1976). The advent of new imaging
technologies (Solan et al., 2003)andtheuseofsubmersiblesallowed
us to progress from snapshot views of the deep sea to a more holistic
study of undisturbed deep-sea communities. Few deep submergence
vehicles (DSVs) can reach full ocean depths, but those that can allow
unique access to the world's deepest ecosystems. Previously published
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/dsri
Deep-Sea Research I
http://dx.doi.org/10.1016/j.dsr.2014.12.012
0967-0637/&2015 Elsevier Ltd. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
n
Corresponding author.
E-mail address: ndgallo@ucsd.edu (N.D. Gallo).
Deep-Sea Research I 99 (2015) 119–133

studies on trench biology using submersible-obtained video surveys
have focused on either the ecology of individual taxa such as
amphipods (Hessler et al., 1978)oronhadalchemosyntheticcommu-
nities (Fujikura et al., 1999; Ohara et al., 2012). It is now possible to use
DSV-obtained video surveys and a comparative approach to analyze
deep-sea trench megafaunal community structure. The wealth of HD
video footage and still images collected by the DEEPSEA CHALLENGE
Expedition (2012) to the New Britain Trench and the Mariana Trench
made such an analysis possible.
Deep-sea trenches compose the majority of the hadal zone,
defined as being deeper than 6500 m (Watling et al., 2013). They
represent the deepest biozone in the world but make-up only
0.24% of the entire ocean (Jamieson, 2011). Trenches are unique
deep-sea ecosystems and remain one of the least understood
habitats on Earth (Jamieson et al., 2010). They are characterized
by high levels of species endemism (Wolff, 1960), high hydrostatic
pressure, low temperatures, the absence of light, sufficient oxygen
content, high sedimentation rates, and limited food supply
(Jamieson, 2011). However, recent studies suggest they may trap
more particulate organic matter (POM) than previously appre-
ciated (Glud et al., 2013). Deep-sea trenches occur at tectonic
convergence zones and 26 trenches deeper than 6500 m have
been described worldwide, with the majority occurring in the
Pacific(Jamieson, 2011).
The study of deep-sea trenches has a long history (Heezen and
Hollister, 1971; Belyaev, 1989; Gage and Tyler, 1991) with much of
the foundational knowledge attributed to the extensive sampling
efforts of the Danish Galathea and the Soviet Vitiaz expeditions
during the mid-1900s. Studies of underwater photographs of trench
benthic communities (Lemche et al., 1976; Belyaev, 1989) have also
added to our knowledge of trench fauna. These studies revealed
that metazoan life was present in all sampled trenches and that
species endemism in trenches was high, with the total degree of
endemism for benthic metazoans increasing with depth (Belyaev,
1989). Due to these early sampling efforts, several important trends
were established that shape how we understand deep-sea com-
munities today. These include recognition that megafaunal biodi-
versity declines rapidly from 2 to 6 km, a gradual transition zone is
reached between the abyssal and hadal zone at 6 to 7 km (Belyaev,
1989), and then diversity decreases much more slowly in the hadal
zone below 7 km (Vinogradova, 1962). Wolff (1960, 1970) identified
several distinctive features that characterize hadal communities
including a) dominance of certain groups like the actinians, poly-
chaetes, isopods, amphipods, echiurids, and holothurians, b) lower
representation of non-holothurian echinoderms, c) insignificance or
lack of fish and decapod crustaceans and d) mass-occurrence of
holothurians at maximal trench depths. Deposit-feeding holothur-
ians in the genus Elpidia and scavenging lysianassoid amphipods in
the genus Hirondellea are recognized as common trench-floor
inhabitants (Jamieson et al., 2010).
The deep sea is an organic-carbon limited system with a high
fraction of refractory compounds reaching the deep-sea floor (Gage,
2003; Jamieson et al., 2010). However, recent biogeochemical/respira-
tion studies (Glud et al., 2013)andthehighabundanceofdeposit-
feeding organisms found in trenches (Belyaev, 1989)suggestthat
trenches may differ from the surrounding abyssal plain by being regions
of resource accumulation. Many authors have noted the importance of
the overlying primary productivity regime and the amount of organic
matter exiting the euphotic zone in influencing the density and
composition of different trench communities (Wolff, 1960; Longhurst,
1995; Jamieson, 2011). Additional food sources for deep-sea trench
communities include allochthonous marine and terrestrial sources due
to the proximity of trenches to land (Gage, 2003). Sinking carcasses of
euphausiids from the upper water column are an important food source
for abyssal ophiuroid communities in the Orkney Trench (Sokolova,
1994). Allochthonous organic input from nearby landmasses also
positively influence the faunal abundance in trenches (Belyaev, 1989),
with a greater quantity of animals corresponding to trenches where
abundant plant debris is present, such as the Philippine Trench, where
coconut husks and bamboo have been recovered (Bruun, 1956).
The Challenger Deep in the Mariana Trench (MT) is the deepest
spot in the ocean (Nakanishi and Hashimoto, 2011), with pressures
reaching 1100 bar or approximately 1.1 tonnes per cm
2
(Jamieson,
2011). The Mariana Trench is overlain by oligotrophic waters with
annual rates of primary production estimated to be !59 g C m
"2
y
"1
(Jamieson et al., 2009b). Maximum bottom currents (8.1 cm s
"1
)
occur at the deepest point of the trench but are of short duration,
with typical current velocities beingo1.5 cm s
"1
for 22.9–63.8% of
the time (Tai ra et al., 2004).
Submersible exploration of the Challenger Deep has a rich history,
starting with the successful descent of the Trieste in 1960, piloted by
Don Walsh and Jacques Piccard. Following that historic descent, no
manned submersible dives occurred over the next 50 years, and
exploration was based on the descents of the remotely operated
vehicles, ROV Kaiko in the 1990s and early 2000s and the hybrid ROV
(HROV) Nereus in 2009. These expeditions offered new insight into
some of the specificorganismsthatliveatthebottomofthe
Challenger Deep. Sediment cores obtained by Kaiko in the Challenger
Deep revealed high-density assemblages of non-calcareous forami-
nifera (Todo e t al., 2005), including a number of new taxa (Gooday
et al., 2008). Researchers studying Kaiko images reported very sparse
life on the seabed (Barry and Hashimoto, 2009)andHROVNereus
test dives revealed a seabed dominated by small amphipods and
scarce polychaete worms, with a single small holothurian observed
(Bowen et al., 2009;P.Fryerpers.comm.,2014).In2009,adrop
camera deployed by National Geographic and Scripps Institution of
Oceanography researchers recovered imagery from the Sirena Deep
at 9970 m in the MT, revealing many xenophyophores and a
rhopalonematid jellyfish (unpublished observation). However, no
quantitative megafaunal community analyses have been published
from these expeditions.
The New Britain Trench (NBT) is a 840 km-long curved trench in
the northern Solomon Sea, close to the landmass of Papua New
Guinea (Davies et al., 1987). The deepest point is the Planet Deep at
9140 m (Davies et al., 1987). The New Britain Trench has received
limited biological attention (Heezen and Hollister, 1971; Lemche
et al., 1976) with the majority of published studies focusing instead
on its geology (Tiffin et al., 1984; Davies et al., 1987), since the sharp
bend of the trench has one of the highest rates of seismic activity in
the world (Tiffin et al., 1984). Wolff (1960) notes that the New
Britain Trench is particularly difficult to trawl due to bottom
configuration, which likely also contributed to the lack of previously
collected data. There is also confusion with nomenclature within
the deep-sea biological literature in that the eastern component of
the trench near Bougainville Island has occasionally been referred
to as the Bougainville Trench even though there is no shallow sill
separating it from the western part, which is referred to as the New
Britain Trench (Tiffin et al., 1984). To avoid confusion, we will refer
to the entire trench as the New Britain Trench, but believe that
published historical trawl results from the Bougainville Trench
(Belyaev, 1989) can be used for comparison with the findings of
this study. The New Britain Trench presents an interesting contrast
to the Mariana Trench because it receives more allochthonous input
from both terrestrial and marine sources.
In this study, we utilized the images and videos obtained by the
DEEPSEA CHALLENGE submersible and the landers to provide a novel
look at the ecology of deep-sea ecosystems. This paper focuses on
megafauna, operationally defined as organisms readily visible in
photographs (Solan et al., 2003)orvideo.Benthicmegafaunalassem-
blages, composed of benthic and demersal community members,
were characterized with respect to their relative abundance, composi-
tion, diversity, and lifestyles. The variable productivity regimes
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133120

associated with the locations and depths of the DEEPSEA CHALLENGER
submersible dives and lander deployments allowed us to test the
following hypotheses: (1) food availability as represented by allochtho-
nous marine and terrestrial input to the sea floor is reflected in hadal
community structure, (2) differences in food availability have similar
effects on community abundance, biodiversity, and lifestyle represen-
tation at bathyal as at hadal depths, and (3) hadal epibenthic and
scavenging communities show similar functional responses to food
availability. Based on previous literature we hypothesized that sites
with higher organic carbon input would have higher organismal
abundance, higher percentage of demersal fauna in the epibenthic
community, and higher biodiversity.
2. Materials and methods
2.1. DEEPSEA CHALLENGE Expedition
In early 2012, the DEEPSEA CHALLENGER, a single occupant
submersible, undertook a series of dives leading up to a full-ocean-
depth dive to the Challenger Deep (CD) in the Mariana Trench
(MT). The DEEPSEA CHALLENGE Expedition consisted of thirteen
submersible dives. Five of these were selected for analysis based
on their value in addressing questions about trench ecology. Three
dives of progressively increasing depth (994 m, 3755 m, 8228 m)
were undertaken in the New Britain Trench (NBT) by the DEEPSEA
CHALLENGER (Table 1,Fig. 1). Following their success, the DEEPSEA
CHALLENGER dove to full ocean depth in the MT CD, and reached a
maximum depth of 10908 m (Table 1,Fig. 1). One additional dive
was conducted near the Ulithi atoll at bathyal depths (1192 m)
(Table 1,Fig. 1). Additionally, baited landers capable of descending
to full-ocean depth were deployed on this expedition in the NBT
(8233 m) and MT (10918 m) (Table 2,Fig. 1).
Unlike traditional submersibles built with a horizontal structural
plan, the innovative vertical attitude of the DEEPSEA CHALLEN-
GER (Fig. 2)wasdesignedtomaximizeitsdescentandascentrates
thereby increasing time available for seafloor exploration. The bulk of
the 7.3 m tall submersible is made up of a now-patented syntactic
foam. The foam provides flotation and a strong structural core
designed to counter the buoyancy and extreme pressure demands of
operating from sea surface to 1100 bar pressure at the bottom of the
Mariana Trench. The pilot sits upright in a 1.09 m diameter, 6.35 cm
thick, steel sphere attached to the bottom of the foam beam (Fig. 2).
Beneath the pilot sphere is an array of scientificequipmentincluding
push-core sediment samplers, a payload bay with space for a suction
sampler, and a hydraulic manipulator arm. The submersible has both a
Seabird CTD and a ParoscientificDigiquartzpressuresensor,and
depths were derived using the UNESCO pressure to depth equat-
ion (Fofonoff and Millard, 1983). Inside the sphere, a Red Epic camera,
mounted directly in the small viewport, captured IMAX-quality ‘5K-
raw’images. External to the sub were four cameras, each one-tenth
the size of previous deep ocean HD cameras–a3Dpairontheboom
arm, and wide-angle and macro cameras on the manipulator arm. Like
the syntactic foam, the cameras and their titanium housings, the
batteries, thrusters and LED lights were all designed specifically for
the submersible (Hardy et al., 2014b).
Baited landers were also developed and deployed to allow for
additional comparisons of hadal scavenging communities. The DEEP-
SEA CHALLENGE “landers”are untethered, unmanned vehicles (4.27 m
tall by 0.76 m wide by 0.91 m deep) (Fig. 2)thatfreefallfromthesea
surface to the seafloor, slowed by water drag to an acceptable terminal
velocity. Syntactic foam provided primary buoyancy. Large hollow
glass spheres were used for instrument housings and supplemental
buoyancy. The landers are metacentrically stable in all conditions. They
can remain in situ for great lengths of time with samplers and sensors,
until acoustically commanded to release their iron anchor weight and
rise to the surface. The DSC lander was equipped with camera and
light arms, Niskin water samplers, sediment corers, and a drop arm
with two baited animal traps (Hardy et al., 2014a).
2.2. Sampling design of submersible-obtained video and
identification of megafauna
All available camera footage from each dive was examined. Mega-
fauna that could clearly be seen with the naked eye in the Red Epic or
boom camera footage were recorded and still images were extracted
from the video. The best representative image of each observed taxon
was then compiled into a key for each of the dives. With the help of
deep-sea taxonomic experts, these still images were then used to
identify observed organisms to the lowest taxonomic level possible.
Due to the limitations of using imagery to identify species, counts of
these taxa represent minimum species numbers (as in Fodrie et al.,
2009)andcertaintaxamayencompassseveralcrypticspecies.
Identified taxa were quantified at each dive site using 2-minute
samples of the Red Epic camera footage, chosen because it provided
continuous coverage of each dive, had the highest resolution, and
reduced the camera-specificbiasbetweendives.Bottomfootagefrom
each dive was split into 10-minute sections spanning the entire time at
the bottom. The first 2 minutes of each 10-minute segment was
extracted as a separate clip using Final Cut Pro and was used for
quantification of megafauna, as well as terrestrial plant detritus,
lebensspuren (animal-generated structures in sediments), large proto-
zoans, and the presence/absence of hard substrate. Two-minute
samples were excluded from the analysis if it was apparent that the
location overlapped with the prior 2-minute sample, resulting in
o20% bottom time quantified for certain dives (Table 1). Since
these overlapping samples were excluded, we assume no overlap
between 2-minute samples and consider them to be replicates for the
specificdive.However,becauseonlyonetransectwasmadeatagiven
depth in each trench, the 2-minute samples are technically
pseudoreplicates.
2.3. Quantification of fauna from submersible dives
Megafauna were quantified within the 2-minute samples to the
lowest taxonomic level possible. For some smaller taxa like amphi-
pods, it was not possible to distinguish species. Even for taxa where
Table 1
Location, depth range, bottom time and information for two-minute video clips analyzed from the five submersible dives in the W. Pacific Ocean. NBT¼New Britain Trench,
MT CD¼Mariana Trench Challenger Deep. Depth range represents the maximum and minimum depths traversed along the seafloor during each dive and the error margin is
73 m.
Dive Locality Depth range (m) Total bottom time # of 2 min samples bottom time quantified Date
D04 NBT 5135
0
54″S, 151136
0
53″E 884–994 4 h 30 min 17 13% 23-Feb-12
D05 NBT 5149
0
48″S, 151142
0
36″E3712–3755 4 h 24 min 25 19% 28-Feb-12
D08 NBT 5152
0
48″S, 152122
0
48″E 7984–8228 3 h 04 min 18 20% 7-Mar-12
D10 MT CD 11122
0
12″N, 142135
0
24″E 10876–10908 2 h 34 min 15 19% 26-Mar-12
D11 Ulithi 9152
0
48″N, 139133
0
36″E 1130–1192 4 h 10 min 20 16% 1-Apr-12
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 121

species-level identifications were made –they may include cryptic
species. All distinct taxa observed during the sampled video segments
of the five dives are listed in Appendix A and a key used for
identification is available as Appendix B.Useofthekey(Appendix B)
increased consistency of counts between 2-minute samples and across
dives. Small epiphytic organisms such as zoanthids observed growing
on the stalks of sponges at the Ulithi dive site were not counted
because they were too small to consistently identify and count.
Echiuran-generated lebenspurren were counted by the number of
star-shaped feeding traces (Ohta, 1984)observed,butwerenot
included in the community analyses because one star-shaped trace
cannot be assumed to equal one live echiuran (Bett et al., 1995).
Fig. 1. Map displaying locations for the five DEEPSEA CHALLENGER submersible dives (D04, D05, D08, D10, D11) and the two lander drops (DOV1-7 and DOV1-12). Upper
panel map shows the Mariana Trench and nearby Ulithi atoll and bottom panel map shows the New Britain Trench, near the island of New Britain. Depths indicate maximum
depth reached during each dive.
Table 2
Location, depth, bottom time and information for still images analyzed from two trench lander deployments in the W. Pacific Ocean. NBT¼New Britain Trench, MT
CD¼Mariana Trench Challenger Deep. Pressure was determined using an RBR pressure sensor and depths were derived using the UNESCO pressure to depth
equation (Fofonoff and Millard, 1983).
Dive Locality Depth (m) Bottom time Images Images quantified Date
DOV1-7 NBT 5153
0
18″S, 152121
0
21″E 8233 "7h 1 image every 5-6 min 70 3-Mar-12
DOV1-12 MT CD 11122
0
8″N, 142125
0
58″E 10918 "5.5 h 2 images every 10 min 68 3-Apr-12
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133122

Xenophophores were also counted but not included in the community
analyses, since we do not know which contain living protoplasm.
Counts of echiuran lebenspurren and xenophyophores are reported in
Appendix A.Terrigenousplantmaterialvisibleontheseafloor surface
was also quantified in each 2-minute sample. Because the submersible
speed, orientation, and altitude were variable throughout the dive,
quantification per 2-minute sample is the most consistent measure of
abundance possible for this study.
For subsequent analysis, species recorded in the 2-minute samples
were also grouped using an intermediate taxonomic designation and
into a phylum-level designation (both specified in Appendix A).
Community and diversity analyses were conducted at species, inter-
mediate, and phylum-level taxonomic groupings to ensure that the
conclusions of the study were robust despite the challenges of species-
level identification. Epibenthic community composition was also
analyzed across dive sites by lifestyle (demersal or benthic). For the
lifestyle analysis, demersal organisms were characterized as all those
that were either observed in the water column (while the submersible
was near the bottom) or swimming up from the seafloor for an
extended period of time (designation specified in Appendix A),
however some may be bentho-pelagic. Most holothurians observed
were characterized as benthic due to the fact that they were
consistently observed on the seafloor. Enypniastes eximia was char-
acterized as demersal for the lifestyle analysis, but is acknowledged to
be bentho-pelagic.
Diversity at dive sites was assessed using several metrics assuming
that all species numbers represent minimums, and true diversity is
likely higher. Shannon-Wiener diversity index (H
0
)(log
2
), Pielou's
evenness index (J
0
), Berger Parker index for rank-1 dominance, total
species richness per site, and rarefaction diversity E(S
100
)were
calculated using the distinct taxa identified (Appendix A). Diversity
metrics were calculated using all species observed during a dive across
all 2-minute samples. All diversity indices were calculated in PRIMER 6
(Clarke and Gorley, 2006). Diversity was also assessed at the phylum
level using these same metrics. This taxonomic level is less sensitive to
errors in species-specificidentification, but also may provide informa-
tion about higher-order evolutionary adaptations to extreme pressure
and food limitation.
2.4. Quantification of scavengers from lander images
The scavenging community of the NBT and MT CD was character-
ized and compared to the epibenthic community using lander images
obtained from one autonomous baited lander deployment each in the
NBT (8233 m) and the MT CD (10918 m) (Table 2). The NBT lander
was deployed !2.8 km from the submersible dive site and used
chicken as bait, and the MT CD lander was deployed !17.2 km a wa y
from the submersible dive site and used skipjack tuna as bait.
Photographs from the landers were visually inspected and all blurred
images were removed. For the NBT deployment, one photograph was
quantified every !5-6 minutes. For the MT CD deployment, two
photographs were quantified every !10 minutes. All scaven gin g
fauna approaching the lander and visible in the images were counted
for the entire deployment time. Relative amphipod abundance was
assessed by counting the number of visible amphipods in each frame
throughout the durationofthedeployment.
2.5. Statistical analyses
Univariate analyses were performed using JMP Pro 11.0. Differences
in average abundance, biodiversity (Shannon's H
0
log
2
), and percent
benthic fauna in the epibenthic community were computed using
counts from the 2-minute samples and tested for significant difference
between the two upper bathyal sites (NBT 1 km VS Ulithi 1.1 km) and
the two hadal sites (NBT 8.2 km VS MT CD 10.9 km). To increase
robustness of the data set, counts for 2-minute samples were first
bootstrapped 100 times, and the bootstrapped dataset was tested for
normality and equal variance. Welch's ANOVA was used to test for
significant difference between the means when the data were normally
distributed but variances were unequal. When the dataset did not meet
the assumptions of normally distributed data, as was the case for the
abundance analysis for NBT 8.2 km and MT CD 10.9 km, the non-
parametric Median and Kolmogorov Smironov tests were used.
All multivariate statistical analyses were conducted using PRIMER 6
(Clarke and Gorley, 2006). Total minimum species observed in the 2-
minute samples were included in the biodiversity estimates and
rarefaction curves were created with untransformed data. Differences
Fig. 2. Illustration of the DEEPSEA CHALLENGER on the left and a photograph of the DEEPSEA CHALLENGE lander descending through the water column on the right. Photo by
Charlie Arneson, used with permission.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 123

in community composition at dive sites were assessed using non-
parametric multidimensional scaling plots (MDS) of Bray-Curtis simi-
larity matrices, generated from fourth root-transformed abundance
data. Additional MDS plots were generated using untransformed
counts and presence/absence to assess the relative importance of
abundance in generating the observed patterns. ANOSIM was used
to conduct pairwise tests of significance and SIMPER was used to
evaluate within- and between-assemblage similarities and to assess
contributions of specifictaxa.
3. Results
3.1. Variability in allocthonous organic input and hard substrate
The Mariana Trench underlies oligotrophic waters that have a
reported low annual primary production rate of 59 g C m
!2
y
!1
(Jamieson et al., 2009b). In contrast, the New Britain Trench is located
beneath less oligotrophic waters. Data on net primary production
(NPP) are not published for the northern Solomon Sea for the waters
overlying the NBT, so an approximation was calculated using the
standard Vertically Generalized Production Model (VGPM, Behrenfeld
and Falkowski, 1997)usingMODIS-AquaChldatadownloadedfrom
the Oregon State University Primary Productivity website based on
methods described in Kahru et al. (2009).Thecalculatedten-year
average (2003–2012) of NPP for the waters overlying the NBT
(latitude !5.667 to !6, longitude 152 to 152.333) was "115 g C
m
!2
y
!1
.Forconsistencyofcomparison,thesameanalysiswas
performed for the Mariana Trench and the Ulithi location. The
calculated ten-year NPP average (20 03–2012) for the waters overlying
the Mariana Trench (latitude 11.200–11.600, longitude 142.000–
142.950) was "39 g C m
!2
y
!1
,orapproximatelyonethirdthe
value of the NPP calculated for the NBT (Fig. 3). The ten-year NPP
average for Ulithi (latitude 9.517–10.167, lo ngi tud e 139. 00 0–139 .56 7)
was "53 g C m
!2
y
!1
.UlithiNPPwashigherthantheMariana
Trench NPP, but still considerably lower than the NBT NPP (Fig. 3).
The NBT also receives substantial allochthonous organic input
from the nearby island of New Britain. This was evidenced
at bathyal (1 km), abyssal (3.7 km), and hadal (8.2 km) depths in
the NBT by the presence of leaves, sticks, palm fronds, and
coconuts (Fig. 4); this has previously been documented by
Lemche et al. (1976). Organisms were frequently seen interacting
with this plant material, but direct consumption was never
observed. Terrestrial detritus declined in the NBT with depth
following a power function of y¼58122x
!1.15
with a high coeffi-
cient of determination (R
2
¼0.994). This attenuation was likely due
both to distance from land, as well as increasing depth. In contrast,
the Ulithi bathyal (1.1 km) site and the MT CD hadal (10.9 km)
site are situated far from large land masses. This is reflected both
in the limited quantity of terrestrial detritus observed at Ulithi
and the absence of terrestrial detritus observed in the MT CD
(Fig. 4).
While this analysis focuses on the influence of allochthonous food
input including surface NPP (Fig. 3)andterrestrialorganicdetritus
(Fig. 4)onstructuringcommunitiesinthedeepsea,itdoesnotinclude
large carrion falls or chemosynthetic food sources. It should be noted
that several large bones, likely from marine mammals, were observed
Fig. 3. Differences in net primary productivity (NPP) (mg C m
!2
d
!1
) between the
three areas visited by the DEEPSEA CHALLENGE Expedition: the New Britain Trench
(NBT) (red), Ulithi (green), and the Mariana Trench (MT) (orange). Figure shows
NPP monthly averages over a ten-year period (2003–2012) approximated using the
standard Vertically Generalized Production Model using MODIS-Aqua Chl data
downloaded from the Oregon State University Primary Productivity website. (For
interpretation of the references to color in this figure, the reader is referred to the
web version of this article).
Fig. 4. Observations of terrestrial organic detritus on the seafloor. Mean counts of individual pieces of terrestrial organic detritus observed in the 2-minute video samples
from the five dives in the New Britain Trench (NBT), Ulithi, and Mariana Trench Challenger Deep (MT CD). Error bars are standard error of the mean. Right panel shows
examples of terrestrial organic detritus, including sticks, leaves, palm fronds, and coconuts, observed at different depths within the NBT. All images in the row correspond to
the indicated depth.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133124

at 8.2 km in the NBT and were colonized by benthic organisms,
suggesting that carrion falls are an additional contributing food source.
Along with differences in marine and terrestrial allochthonous
organic input, the dive sites differed based on the presence and
abundance of hard substrate. Rocky outcrops and hard substrates
were observed in 100% of the 2-minute samples at the Ulithi
bathyal (1.1 km) site, compared to only 24% of the bathyal (1 km)
NBT samples, 4% of samples at the abyssal (3.7 km) NBT site, and
44% at the hadal (8.2 km) NBT site. The hard substrates observed at
the hadal (8.2 km) NBT site were composed of pillow basalts and
steep rocky ledges. No hard substrates were observed during the
MT CD hadal (10.9 km) dive.
3.2. Epibenthic community composition
The epibenthic communities at each of the dive sites differed
significantly from each other (Figs. 5 and 6)(ANOSIM,GlobalR¼0.79,
po0.01), and each had different dominant taxa (Appendix A,Fig. 6).
Because MDS patterns generated from untransformed counts and
from presence/absence (data not shown) show the same community
differences as the fourth-root transformed counts (Fig. 5A), and all
communities differ significantly from each other (ANOSIM, po0.01),
we infer that differences in community structure between dive
locations are related primarily to community composition rather
than abundance patterns. Among the five dive sites, the MT CD hadal
(10.9 km) site had the greatest community homogeneity with 61%
average similarity in community composition among the 2-minute
samples at the species level and 75% at the phylum level. In contrast,
epibenthic communities at the two bathyal sites and at the NBT
hadal (8.2 km) site were more heterogeneous (Fig. 5). Patchiness of
the abiotic environment at the two bathyal sites and at the NBT hadal
(8.2 km) site likely increased community heterogeneity. Our findings
are robust independent of the challenges of species identification
because community patterns at a higher taxonomic grouping (phy-
lum-level) show the same community-specificpatternsasthoseat
the species-level (Fig. 5)(Somerfield and Clarke, 1995).
The Ulithi bathyal (1.1 km) community included sessile animals
(sponges and corals) with amphipods and fish contributing to the
demersal assemblage. In contrast, the NBT bathyal (1 km) commu-
nity contained more mobile fauna, including decapods, amphipods,
and fishes, with crinoids dominating the sessile fauna (Fig. 6).
Comparisons of the two bathyal sites revealed significant differ-
ences in community composition (po0.01), with 83% average
dissimilarity at the species level and 44% average dissimilarity at
the phylum level. The abundance of decapods and holothurians at
the NBT (1 km) site and the abundance of sponges, hexacorals, and
octocorals at the Ulithi (1.1 km) site (Fig. 6) contributed 62% of the
community dissimilarity between the two bathyal sites (SIMPER).
Pairwise comparisons between bathyal (1 km), abyssal (3.7 km),
and hadal (8.2 km) dive sites within the NBT showed that these
communities differed significantly from each other (ANOSIM,
po0.01). The average dissimilarity between the bathyal (1 km) and
abyssal (3.7 km) NBT sites was 70% at the species level and 46% at the
phylum level. The NBT abyssal (3.7 km) and hadal (8.2 km) dive sites
also had high dissimilarity (87% at the species level and 59% at the
phylum level). The abundance of holothurians, actinarians, and
polychaetes at the hadal (8.2 km) site contributed to the high degree
of dissimilarity between the abyssal (3.7 km) and hadal (8.2 km)
communities.
The hadal epibenthic communities studied here consisted of
both benthic species observed consistently resting or movingon the
sea floor and demersal species that swim or drift just above the sea
floor. Among the 560 benthic community members observed at the
hadal NBT (8.2 km) site and the hadal MT CD (10.9 km) site, 44%
were the cnidarian Galatheanthemum sp. and 43% were holothur-
ians; enteropneusts, small caymenostellid asteroids, actinarians,
and unknown infaunal organisms accounted for another 13% of
the observations. Of the 334 demersal observations at the two hadal
sites, the most common taxa were scale worms (Polynoidae 46%),
amphipods (26%) and a number of unidentified gelatinous swim-
ming forms, likely belonging to the phylum Cnidaria (27%).
Common taxa at bathyal depths (sponges, corals, crinoids and
vertebrates) were absent in the hadal video imagery (Fig. 6).
Similar to the two bathyal communities, the two hadal com-
munities differed significantly from each other with 85% average
dissimilarity at the species level and 64% average dissimilarity at
the phylum level. Both hadal communities were characterized by
the presence of elpidiid holothurians and amphipods (Fig. 6). The
presence of polychaetes, actinarians, enteropneusts, and unidenti-
fied gelatinous swimming forms at the NBT hadal (8.2 km) site
contributed to the high degree of dissimilarity between these two
hadal communities. Xenophophores were commonly observed in
the hadal MT CD community and counts are given in Appendix A,
however since we do not know which contain living protoplasm,
they were not included in the community analysis.
3.3. Community abundance, biodiversity, and lifestyle representation
The Ulithi and MT CD sites were considered to be more food
limited than the NBT sites, and thus we compared the organismal
abundance, biodiversity, and lifestyle representation of the benthic
communities at bathyal and hadal depths. The NBT bathyal (1 km)
Fig. 5. Multidimensional scaling plot based on a Bray-Curtis similarity matrix of fourth root-transformed abundance data at the species level (A) and at the phylum level
(B) for megafauna at the five dives sites. Each point represents a 2-minute sample and icon shape and color indicate dive location and depth. For A, 2D stress is 0.16, ANOSIM
(Global R¼0.83, po0.01), and for B, 2D stress is 0.20, ANOSIM (Global R¼0.73, po0.01). NBT¼New Britain Trench, MT CD¼Mariana Trench Challenger Deep.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 125

dive site had higher organismal abundance than the Ulithi bathyal
(1.1 km) dive site (F
1,147
¼239.39, po0.0001) (Fig. 7A) and the NBT
hadal (8.2 km) dive site had higher organismal abundance than the
MT CD hadal (10.9 km) dive site (Median Test, K-S Test, po0.0001)
(Fig. 7B). Rarefaction biodiversity did not follow the same trend
(Fig. 8), and the more oligotrophic Ulithi bathyal (1.1 km) community
was characterized by higher biodiversity than the NBT bathyal (1 km)
dive site (F
1,131
¼142.10, po0.0001) (Fig. 7C), whereas the more
oligotrophic MT CD hadal (10.9 km) community had lower diversity
than the NBT hadal (8.2 km) site (F
1,179
¼175 5. 57, po0.0001)
(Fig. 7D). Comparison of lifestyle representation reveals that the more
oligotrophic Ulithi bathyal (1.1 km) community was composed of a
higher proportion of benthic fauna compared to the NBT bathyal
(1 km) community (F
1,196
¼5218.98, po0.0001) (Fig. 7E). In contrast,
the NBT hadal (8.2 km) community was composed of a higher
proportion of benthic fauna compared to the more oligotrophic MT
CD hadal (10.9 km) community (F
1,198
¼274.14, po0.0001) (Fig. 7F).
When xenophyophore test counts were included in this analysis
(results not shown), the pattern for the hadal sites was reversed, with
higher benthic lifestyle representation in the MT CD (10.9 km) site
than the NBT (8.2 km) site, but the patterns of the abundance and
diversity analyses were unchanged.
3.4. Biodiversity trends with depth
For the megafaunal analysis at the species level, there was no clear
trend in biodiversity with depth (Table 3,Fig. 8). Within the NBT, the
bathyal (1 km) site had the highest biodiversity, with 35 species
observed and an H
0
of 3.50. Biodiversity was lowest in the NBT at
the abyssal (3.7 km) site (H
0
¼2.06), but then increased again at the
hadal (8.2 km) site (H
0
¼2.64). Across the five sites visited, the MT CD
hadal (10.9 km) site had the lowest biodiversity as revealed by all
diversity metrics (Table 3,Fig. 8).
Analysis of biodiversity at the phylum level revealed different
trends in biodiversity with depth (Fig. 8B). Within the NBT,
phylum-level biodiversity declined slightly between the NBT bath-
yal (1 km) (H
0
¼1.54) and abyssal (3.7 km) sites (H
0
¼1.08), but this
decline was smaller than that observed for species-level biodiver-
sity (Fig. 8A). Surprisingly, the NBT hadal (8.2 km) site had very
high phylum-level biodiversity (H
0
¼1.93; Fig. 8B) with represen-
tatives of 6 phyla observed in the hadal community (Cnidaria,
Ctenophora, Annelida, Arthropoda, Echinodermata, and Hemichor-
data) and high evenness (J
0
¼0.74) (Table 3).
3.5. Observation of potentially new taxa
Several previously unreported taxa were seen in the DEEPSEA
CHALLENGE dive and lander imagery. A pelagic polychaete in the
family Acrocirridae, genus Teut hidodrilus (squidworm) was observed at
1kmintheNBT(Fig. 9AandB).Thehadaltrench-floor community in
the NBT revealed three new taxa and may extend the depth limit for
caymenostellid asteroids to "8.2 km. The largest organisms present in
the NBT hadal community were ulmarid jellyfish that had previously
not been described from this trench or from similar depths elsewhere.
The observed species tentatively belongs to the subfamily Poraliinae
and appeared to be feeding on the sediment surface in the NBT
(Fig. 9C). Three individuals were observed during the intensively
surveyed 36 minutes of bottom time (2 min samples), and nine were
observed during the whole 3-hour bottom time. Crustaceans thought
to be either penaid shrimp or mysids were also observed in the lander
imagery from the NBT at 8.2 km (Fig. 9D). They were observed
approaching the baited lander but did not approach or feed on the
bait. Twenty-two individuals were observed in 70 images taken by the
baited lander over a 7-hour deployment time, but it is not possible to
know if these were all distinct individuals or the same individuals
returning over that time period. The NBT trench floor ("8.2 km) also
Fig. 6. Comparison of total epibenthic community composition based on intermediate taxonomic grouping at each of the five dive sites visited. Percent community
composition is based on summed totals of each group counted using all 2-minute samples from each dive. NBT¼New Britain Trench, MT CD¼Mariana Trench Challenger
Deep. Additional taxa observed but not shown in this figure, due to their minor contribution to the community, include: Ulithi 1.1 km unidentified anthozoan (5%), NBT
3.7 km cephalopod ( o1%), and NBT 8.2 km ctenophore (o1%) and enteropneust (4%). Counts of echiuran lebenspurren and xenophyophore tests are included in this figure,
however, are not included in the univariate and multivariate community analyses.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133126

had large pieces of wood debris, as well as marine mammal bones,
which were colonized by what appeared to be small caymenostellid
asteroids (Fig. 9EandF).
The submersible dive to the Challenger Deep (10.9 km) revealed
sixty-five individual elpidiid holothurians at depths between 10876–
109 08 m in t he 2- minute video samples. T hes e hol oth uri ans are
thought to be a previously unreported species and had elongate semi-
transparent bodies, extended oral tentacles (between 3–5commonly
observed in the videos), and were always observed on top of the
sediment with no evidence of burrows, fecal coils, and rarely any
evidence of trails. Whenever more than one holothurian was observed
in a single frame, the holothurians were always oriented in the same
direction. We tentatively suggest that these individuals belong to
Peniagone (Fig. 9GandH),agenusthatisknowntoexhibitorientation
behavior.
3.6. New Britain Trench and Challenger Deep hadal scavenging
communities
Amphipods were the dominant fauna visible in the autonomous
baited lander images. The NBT trench-floor scavenging community
(8233 m) was more diverse and had more abundant scavengers than
Fig. 7. Differences in abundance (A and B), diversity (C and D), and percent of seafloor benthic versus demersal lifestyles (E and F) in the epibenthic communities between
two bathyal (A, C, and E) and two hadal (B, D, and F) sites. NBT ¼New Britain Trench, MT ¼Mariana Trench Challenger Deep. % benthic refers to the percent of fauna in the
community consistently observed on the sea floor. Of the two bathyal sites, Ulithi 1.1 km is considered more food-limited and, of the two hadal sites, MT CD 10.9 km is
considered more food-limited. Data shown have been bootstrapped 100 times and all differences were determined to be statistically significant (po0.0001). Box and
whisker plots show median, upper and lower quartiles, maximum and minimums, and outliers for each bootstrapped dataset.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 127

the MT CD hadal scavenging community (10918 m). The abundance of
amphipods in each image was greater in the NBT for the majority of
the lander deployment until about 325 minutes after landing, at which
point the number of amphipods attending the bait leveled off (Fig. 10).
In contrast, the abundance of amphipods attending the bait in the MT
CD showed no asymptote throughout the duration of the deployment
(Fig. 10). Lander imagery and unpublished molecular sequencing
results from lander-collected amphipods (Cytochrome oxidase subunit
1, G. R ou se, unp ub l. ) reveale d 5 am ph ip od species in t he N BT,
including Alicella gigantea,whereasonly2amphipodspecieswere
detected in the MT CD. Image analysis also revealed the presence of
holothurians and crustaceans (decapods or mysids) in the NBT hadal
scavenging community, which were not present in the MT CD hadal
scavenging community. Results from image analysis combined with
CO1 sequencing (G. Rouse, unpubl.) suggest that the NBT hadal site,
with seven distinct taxa, had a more diverse scavenging community
than the MT CD community with only two distinct taxa.
4. Discussion
4.1. Strengths and challenges of video-derived submersible data
Recent advances in technology have enhanced hadal research by
providing time lapse or continuous photographs and video that
offer behavior information (Jamieson et al., 2009a, 2009b, 2011b).
However, in trenches these have often been associated with landers,
which have a fixed domain of study. This study pairs manned
submersible footage from the DEEPSEA CHALLENGER with still
images from baited landers to provide a holistic description of
hadal epibenthic communities using video imagery. The video
imagery offers information about ecological patterns that might
not be evident in trawl samples, as well as information about spatial
heterogeneity that is not evident in imagery from fixed locations.
While the use of video for analyzing community patterns has
many strengths, one of the principal challenges is identifying
species consistently and accurately. To reduce bias resulting from
the difficulties of species identification we (1) extracted from the
video the best view of each distinct taxon observed and sent it to
deep-sea experts to obtain the lowest taxonomic identification
possible (Appendix A) and (2) conducted analyses of community
composition and diversity at several taxonomic levels, including
an intermediate taxonomic grouping and a phylum-level grouping
(specified in Appendix A). Similar patterns in community compo-
sition, community heterogeneity, and community diversity at the
phylum-, intermediate-, and species-level taxonomic grouping,
suggest that the conclusions of this study are robust and not
biased by the challenges of taxonomic resolution from video
imagery (Somerfield and Clarke, 1995). All diversity indices likely
represent species minima, as there may be small, cryptic indivi-
duals that cannot confidently be identified to species level from
video. Specifically, diversity of the amphipods, demersal annelids,
and demersal gelatinous taxa are almost certainly underestimated
by this analysis, whereas larger and more easily differentiated
megafauna such as fish, echinoderms, and some cnidarians repre-
sent more accurate diversity estimates. Although highly mobile
species may have avoided the submersible, most individuals
observed in the video did not seem to react to the presence of
the submersible, and deep-submergence vehicles have previously
been used to study mobile, deep-sea animals (Barham et al., 1967).
A strength of using video footage was that it allowed us to
witness the feeding behaviors of the recently recognized ulmariid
Fig. 8. Rarefaction curves for distinct taxa (A) and distinct phyla (B) observed for the five different dive sites visited on the DEEPSEA CHALLENGE expedition. NBT¼New
Britain Trench, MT CD¼Mariana Trench Challenger Deep. (For interpretation of the color key in this figure, the reader is referred to the web version of this article).
Table 3
Epibenthic diversity indices for phylum and species-level epibenthos for the 5 dive sites.
Ulithi 1.1 km NBT 1 km NBT 3.7 km NBT 8.2 km MT CD 10.9 km
Phylum Species Phylum Species Phylum Species Phylum Species Phylum Species
Shannon’sH
0
(log
2
)2.09 4.28 1.54 3.50 1.08 2.07 1.93 2.6 4 1.00 1.62
Pielou's J
0
0.81 0.82 0.66 0.68 0.42 0.53 0.74 0.66 1.00 0.70
Berger Parker 0.44 0.15 0.65 0.35 0.78 0.63 0.40 0.32 0.52 0.50
E(S
100
)5.89 27.01 4.84 23.34 5.21 11.3 5.13 10.18 2 4.95
Total No. Species 6 37 5 35 6 15 6 16 2 5
Total No. Individuals 194 221 272 272 234 264 667 763 124 131
Diversity indices were calculated using the sum of all quantified individual species or phyla using all 2-minute samples from each dive. Counts of xenophyophore tests and
echiuran-generated traces were not included. All diversity indices represent minimum values, as cryptic species may exist.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133128

jellyfish in the NBT and the elpidiid holothurians in the MT CD
(Fig. 9). The ulmariid jellyfish at 8.2 km in the NBT was observed
feeding on particulates on the sediment surface by skimming the
sediment, leaving the sediment surface when disturbed by the
submersible. Previous studies have described the important role
jellyfish play both as members of epipelagic, deep-sea communities
(Miyake et al., 2002; Lindsay and Pagès, 2010), and as vectors of
carbon transport to the deep sea following jellyfish blooms in
surface waters (Sweetman and Chapman, 2011). This study reveals
that jellyfish are also important members of the hadal community
in productive trench environments.
In the MT CD at 10.9 km, video footage allowed us to observe that
the elpidiids were oriented uniformly (Fig. 9G), suggesting they are
utilizing currents for feeding. This is consistent with previous
observations of elpidiid holothurians in the genus Peniagone utilizing
bottom currents for orientation and/or feeding (Ohta, 1985; Okada
and Ohta, 1993). While this behavior is not novel, this is the first
description of an abundant population of epibenthic elpidiids at the
bottom of the CD. Observations of elpidiid behavior in the Peru-Chile
Trench made with time-lapse video (Jamieson et al., 2011a)provide
another example of the value of video for increasing our knowledge
of the behaviors of hadal species.
4.2. Epibenthic community patterns in the bathyal compared to the
hadal zone
We tested the hypothesis that increased food has similar effects
on community structure at bathyal as at hadal depths by examin-
ing nearby bathyal epibenthic communities. This hypothesis was
supported for abundance (Fig. 7A, B), but not for Shannon species
diversity (Fig. 7C, D) and representation of benthic lifestyles
(Fig. 7E, F). The presence of hard substrate at the Ulithi bathyal
(1.1 km) site and the NBT hadal (8.2 km) site may also have
Fig. 9. Images of epibenthic and benthopelagic taxa, some of which are believed to be previously unreported, that were observed during the DEEPSEA CHALLENGE
submersible and lander dives. A and B: Teuthidodrilus polychaete species from 1 km in the NBT. C. Large hadal ulmarid cnidarian observed at 8.2 km in the NBT. D. Crustacean
(decapod or mysid) observed during baited lander deployments at 8.2 km in the NBT. E and F. Caymenostellid asteroids observed on wood debris and bones at 8.2 km in the
NBT. G and H. Epidiid holothurians observed at 10.9 km in the MT CD. NBT¼New Britain Trench, MT CD ¼Mariana Trench Challenger Deep.
Fig. 10. Number of amphipods observed in each frame in the hadal scavenging
community of the New Britain Trench (NBT) and Mariana Trench Challenger Deep
(MT CD) over the deployment time of the lander.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 129

contributed to the higher diversity and higher proportion of
benthic fauna observed at these sites.
The diversity results are consistent with a unimodal response
of deep-sea diversity to food availability (Levin et al., 2001),
wherein more food gives rise to higher diversity at hadal depths,
but decreased diversity at bathyal depths, where ambient food
supply is higher. Most data for deep-sea diversity trends with
depth in the benthos exhibit a unimodal pattern with maximum
values at mid slope depths (Rex and Etter, 2010; Menot et al.,
2010). However, few extend to full-ocean depth. The results
presented here indicate that diversity below abyssal depths will
vary significantly depending on overlying productivity regime and
proximity to sources of terrestrial organic matter (Fig. 7C and D).
This is consistent with current understanding that food supply is a
primary driver of hadal community structure (Bruun, 1956;
Jamieson, 2011). Another interesting observation is that phylum-
level biodiversity may actually increase with depth (Fig. 8). This
finding should be tested in other trenches, since the conclusion is
based on three depths sampled in the NBT.
4.3. Similar responses in hadal scavenging and epibenthic
communities
The use of paired baited lander deployments and submersible dives
to the bottom of the NBT and MT CD, allowed us to assess if hadal
epibenthic communities showed similar patterns of abundance and
diversity as scavenging communities. Both communities exhibited
greater abundance (Fig. 10)andhigherdiversityintheNBTthanthe
MT CD at hadal depths. These trends were largely driven by amphipods
in the scavenging assemblage, but also reflect the presence of elpidiid
holothurians and decapod or mysid crustaceans at 8.2 km in the NBT.
Lander imagery reveals a more complex food web in the hadal
scavenging community of the NBT; crustaceans (decapods or mysids)
and one amphipod species that attended the bait without feeding may
be carnivores that consume the bait-attending scavengers. The
observed amphipods may belong to the genus Princaxelia,whichare
carnivorous bait-attending amphipods found at hadal depths in Pacific
Ocean trenches (Kamenskaya, 1984; Jamieson et al., 2011a, 2011b). In
contrast, the MT CD scavenging community was only composed of two
amphipod species, which both appeared to be scavengers.
This study reports higher diversity in the amphipod scavenging
community at these depths than has been described in previous
papers in the Phillipine Trench (Hessler et al., 1978), the Kermadec
Trench (Jamieson et al., 2011c), and in the Tonga Trench (Blankenship
et al., 2006). In the Tonga and Kermadec trenches, 4 species of
lysianassid amphipods were found to partition the trench vertically
with the younger stages often occurring at shallower depths (Blanke-
nship et al., 2006). In the NBT –5speciesco-occurringat8.2kmmay
be a record for hadal depths, and suggests a more complex food web
in the NBT.
4.4. Influence of geological differences on community patterns
Previous studies have explored the effects of habitat hetero-
geneity due to geological characteristics on macrobenthic com-
munity structure and species richness (e.g., Fodrie et al., 2009; De
Leo et al., 2014). The Puerto Rico, Kermadec, Tonga, Peru-Chile,
New Hebrides, West Solomon, and the New Britain trenches are
known to have masses of rubble, talus slopes, and fragmented
outcrops (Heezen and Hollister, 1971), which increase habitat
heterogeneity in these trenches. While our analysis focused on
differences in food supply as the main factor giving rise to the
differences in community patterns observed, habitat heterogene-
ity due to geological differences likely also influenced our results.
One line of evidence for this is that the sites with the greatest
community heterogeneity across 2-minute samples (Ulithi 1.1 km
and NBT 8.2 km) (Fig. 5) were also sites that had rocky substrates
observed in more than 40% of the samples. Thus, the lower
diversity observed at the NBT abyssal (3.7 km) site and the MT
CD hadal (10.9 km) site may additionally be due to the homo-
geneous, fine silty substrate observed at these sites. Even in the
MT CD, evidence of community patchiness and habitat hetero-
geneity were also apparent, with higher densities of xenophyo-
phore tests observed where there were shallow sediment troughs.
The NBT and MT subduction zones are similar with respect to
the degree of seismic activity (see Fryer et al., 2003 for the MT;
Yon ish ima et al ., 20 05 a nd Be nz et al. , 2010 for the NBT) but they
differ in the type and degree of deformation, which could yield
distinct ecological disturbance regimes for the biota. The forearc of
the NBT slopes relatively gently from sea level to !7000 m depth,
then steepens within 5.5 km of the trench axis. There are a few
narrow channels offshore of river mouths that broaden into swales to
depths of 4,000 m in the forearc area, but little in the way of features
that suggest significant faulting or wide-spread crustal deformation.
By contrast with the NBT forearc, that of the southern MT area is
highly deformed, with numerous fault lineaments and multiple fault-
controlled canyons that feed sediment from the inner forearc slope
into the trench axis (Fryer et al., 2003). Thus in addition to the
southern MT axis having essentially no allochthonous organic input,
it likely suffers from a greater frequency of resurfacing. This greater
disturbance may reduce animal densities, inhibit subsurface,
bioturbation-dependent lifestyles, and reduce fitness of benthic taxa.
The 2.7 km depth difference between the two hadal sites
compared in this study (MT CD 10.9 km and NBT 8.2 km) may
also have contributed to the lower abundances, lower diversity
and altered lifestyles documented in the MT CD. Epibenthic
diversity (Vinogradova, 1962) and scavenging amphipod diversity
(Blankenship et al., 2006) is known to decrease with increasing
depth in the hadal zone and certain groups such as decapod
crustaceans and fishes are absent below 8.3 km (Jamieson et al.,
2009a, 2009b; Fujii et al., 2010; this study), possibly due to
physiological limitations (Yancey et al., 2014). It would be bene-
ficial to conduct an additional study in a trench of similar bottom
depth to the NBT, that is overlain by oligotrophic waters (similar to
the MT CD), in order to tease apart the influence of depth,
compared to the influence of allochthonous input, on observed
differences in the epibenthic and scavenging communities.
4.5. Observations of importance
Despite limitations on level of identification, this study contri-
butes polychaete, crustacean, cnidarian, and echinoderm observa-
tions of note for their depth records or evolutionary significance. The
genus Teuthi dodrilu s was first described in 2010 (Osborn et al., 2011)
and is currently monotypic with the only known species (Teuthido-
drilus samae)occurringintheWesternCelebesSeaat2000–2900 m;
asimilarmidwaterpolychaetehadpreviouslybeenobservedoffof
western India at 1500 m by the SERPENT project (http://archive.
serpentproject.com/231/)butwasnevercollected.Basedondiffer-
ences in appearance, depth of observation, and distance from the
Celebes Sea, the individual observed at 1 km in the NBT (Fig. 9Aand
B) likely represents a new species in this genus, and may be of
particular interest for future study due to its location. This genus is of
evolutionary interest because it is thought to be transitional between
benthic and pelagic polychaetes and is a sister group to the ‘bomb’-
bearing clade (Osborn and Rouse, 2011).
Prior to the recent discovery of decapod crustaceans at hadal
depths in the Japan and Kermadec Trenches (Jamieson et al., 2009a),
it was thought that decapod crustaceans were unable to survive at
hadal depths. If our NBT crustacean observations (Fig. 9D) were
actually a decapod crustacean, they would extend the maximum
depth of decapod crustaceans from 7703 m to 8233 m. Whether
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133130

decapod or mysid, their abundance suggests that they may be an
important part of the food web in carbon-rich hadal trenches.
Prior to this expedition, the deepest known holothurian in the
published literature was Myriotrochus brunni, which was collected
by trawl from 10710 m in the Mariana Trench (Wolff, 1970;
Belyaev, 1989). Our observation of elpidiid holothurians in the
MT CD at 10.9 km (Fig. 9G and H) extends the depth range for
elipidiids to maximum ocean depth, making them the deepest
known deuterostome taxon. Taken together these findings make
clear that the hadal zones remain a fertile area for extending the
known depth ranges of invertebrate life forms.
We saw largely soft-bodied taxa at hadal depths, with few
organisms having calcareous or siliceous skeletons. The prevalence
of soft-bodied organisms at hadal depths is hypothesized to be due
to the difficulties of biomineralization below the calcium compen-
sation depth (CCD) at 4000–5000 m (Jamieson et al., 2010). Despite
this, some organisms with calcareous skeletons do exist at hadal
depths, such as serpulid polychaetes that inhabit calcareous tubes
(Kupriyanova et al., 2014) and limpets reported from wood and
seagrass in the Puerto Rico Trench (Leal and Harasweych, 1999). At
8.2 km in the NBT, we observed caymenostellid asteroids and
holothurians in the genus Elpidia, which have calcareous endoske-
letons or microspicules. The unique mineral structure of echino-
derm skeletons is composed of high-magnesium calcite and confers
high strength properties with minimal amounts of material (Weber
et al., 1969). This structure may allow for hadal echinoderms, such
as holothurians, to produce skeletal structure below the CCD.
The complete absence of fish in both the images from the baited
landers and the submersible videos from both the bottom of the
NBT and the MT, supports the hypothesis that fish have a physio-
logical depth limit (Yancey et al., 2014). While our results suggest
that the MT food web would be less likely to support fish taxa, the
high abundance and diversity of potential prey items in the NBT
suggests that the absence of fish at the base of the NBT is not due to
food limitation. During the dive of the Trieste to the Challenger
Deep, Jacques Piccard and Don Walsh observed a “flatfish”at
10916 m through the viewport of the submersible. This would be
the deepest fish sighting known to man, but this sighting has been
denounced as erroneous by the scientific community (Wolff, 1961;
Jamieson and Yancey, 2012). The deepest published fish observation
is the liparid Pseudoliparis amblystomopsis, observed at 7703 m in
the Japan Trench (Jamieson et al., 2009a, 2009b; Fujii et al., 2010).
Deep-sea bony fishes utilize the osmolyte trimethylamine N-oxide
(TMAO) to stabilize protein structure against distortion by hydro-
static pressure (Kelly and Yancey, 1999; Samorette et al., 2007). It is
thought that deep-sea fishes become isosmotic at 8200 m due to
TMAO accumulation, and thereby are physiologically excluded from
living at greater depths (Yancey et al., 2014). The absence of fish
observed at 8.2 km in the NBT and at 10.9 km in the MT (Fig. 6)
supports this teleost depth limit. However, during manuscript proof
preparation the media reported lander-based (unpublished) obser-
vations of snailfish living at 8145 m. (http://schmidtocean.org/story/
show/3584).
4.6. Heterogeneity of deep-sea epibenthic communities
Each of the sites examined had a distinct appearance with different
faunal dominance (Fig. 6,Appendix A). While echiuran-generated
lebensspuren were not included in the quantitative analyses, they
were abundant at the abyssal (3.7 km) NBT site (Fig. 6). Other regions
with high echiuran density have also been described including the
Kaikoura canyon of New Zealand (De Leo et al., 2010)andthebaseof
the southern Chile margin (T. Shank, A. Thurber, L. Levin, unpubl.).
These observations suggest that echiurans may dominate in abyssal
mid-slope regions that receive considerable allochthonous input.
At the base of the slope, the trench floor of the NBT at 8.2 km was
dominated by hadal anemones in the genus Galatheanthemum,elpidiid
holothurians in the genus Elpidia,andenteropneusts(Fig. 6). The
observed community composition in the NBT was similar to that
observed in eutrophic trenches such as the Puerto Rico Trench (George
and Higgins, 1979). In contrast, amphipods andelipidiidholothurians
dominated the oligotrophic MT CD megafaunal community, and
protozoan xenophyophore tests were also abundant (Fig. 6). The
dominant taxa present at our hadal sites have been reported from
other trenches (Belyaev, 1989; Blankenship-Williams and Levin, 2009),
but we did not expect to see such a large number of holothurians in the
MT CD. High densities of deposit-feeding holothurians and amphipods
in the trench axis may reflect accumulation of food (Belyaev, 1989;
Jamieson et al., 2010).
5. Conclusions
In summary, the combined DEEPSEA CHALLENGE dives and lander
deployments offer a rare glimpse into the seafloor and demersal
assemblages of two very different trenches. Our observations support
akeyroleforallochthonousproductivityinshapingabundanceand
diversity and the relative importance of demersal lifestyles. High
putative species- and phylum-level diversity observed in the New
Britain Trench suggest that trench environments may foster higher
megafaunal biodiversity than surrounding abyssal depths if food is not
limiting. As hypothesized, hadal and bathyal assemblage abundances
responded similarly to greater food availability, but diversity and
lifestyle representation did not. Also (as hypothesized), scavenging
and non-scavenging benthic assemblages exhibited similar responses
to food availability. Since climate change and near-trench deep-sea
mining efforts may alter allochthonous input to trenches, changes in
trench community structure, abundance, and biodiversity should be
considered and studied. In this study, possible new species were seen,
novel behaviors were observed, and new depth records set. Although
the lander recovered amphipods, most of the taxa we observed were
not sampled; retrieval would potentially have allowed the description of
new species and the confirmed identification of others. We suggest
multiple observation and sampling approaches are needed to maximize
knowledge of these extreme and mysterious trench communities.
Acknowledgements
We would like to thank: Mati Kahru for providing support in
calculating net primary production (NPP) for the waters overlying
the New Britain Trench, Ulithi, and the Mariana Trench; Phil
Alderslade, David Billett, Robert Carney, Harim Cha, Jeffery Drazen,
Andrew Heyward, Daphne Fautin, Andrey Gebruk, Dhugal Lindsay,
Alexander Mironov, Tina Molodtsova, Karen Osborn, Greg Rouse,
Nadezhda Sanamyan, Timothy Shank, Mindi Summers, and two
additional anonymous specialists for their help with organismal
identification; Lynn Waterhouse and Guillermo Mendoza for
providing statistical advice; Ralph Pace and Cody Gallo for image
processing and enhancement; Christina Symons for generously
sharing cruise metadata and for producing Figure 1; Ron Allum
and the many DEEPSEA CHALLENGE Expedition team members
who made the recovery of the video data possible; and all
crewmembers of R/V Mermaid Sapphire, the Spirit of New Guinea
and S/S Barakuda for expedition support. We thank four anon-
ymous reviewers who provided helpful comments on earlier
versions of the manuscript; National Geographic and Rolex for
their support of the expedition; the Blue Planet Marine Research
Foundation for financial support of the analysis of the video data;
the Avatar Alliance Foundation for supporting open access pub-
lication; and the National Science Foundation IGERT grant No. NSF
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 131

DGE 0903551 for graduate assistance for NDG. This material is
based upon work supported by the National Science Foundation
Graduate Research Fellowship under Grant No. DGE-1144086. Any
opinion, findings, and conclusions or recommendations expressed
in this material are those of the authors and do not necessarily
reflect the views of the National Science Foundation.
Appendix A
All quantified taxa from the five DEEPSEA CHALLENGE submersible
dives. Counts indicate the total number of individuals observed during
each dive using all 2-minute video samples. Column titled “B/D”
indicates which taxa were characterized as benthic (B) or demersal
(D) for the lifestyle analysis. Columns titled “IntTaxGroup”and
“Phylum”indicate intermediate taxonomic and phylum classifications
used for the community and biodiversity analyses. NBT¼New Britain
Trench, MT CD¼Mariana Trench Challenger Deep.
Appendix B
Image-based key with all taxa quantified from the five DEEPSEA
CHALLENGE submersible dives. Order and names correspond to
names and counts in Appendix A. Images represent the best still of
each taxon extracted from the DEEPSEA CHALLENGE videos.
Appendix C. Supporting information
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.dsr.2014.12.012.
References
Barham, E.G., Ayer, N.J., Boyce, R.E., 1967. Macrobenthos of the San Diego Trough:
photographic census and observations from bathyscaphe, Trieste. Deep-Sea Res
14, 773–784.
Barry, J.P., Hashimoto, J., 2009. Revisiting the Challenger Deep using the ROV Kaiko.
Marine Technology Society Journal 43, 1–2.
Behrenfeld, M.J., Falkowski, P.G., 1997. Photosynthetic rates derived from satellite-
based chlorophyll concentrations. Limnol. Oceanogr. 42, 1–20.
Belyaev, G.M., 1989. Deep-Sea Ocean Trenches and Their Fauna. Nauka Publishing
House, Moscow p. 385 (Translated to English by Scripps Institution of
Oceanography, USA, 2004).
Benz, H.M., Herman, M., Tarr, A.C., Hayes, G.P., Furlong, K.P., Villaseñor, A., Dart, R.L.,
Rhea, S., 2010. Seismicity of the earth 1900–2010 New Guinea and vicinity.
USGA Open-File Report 2010–1083-H.
Bett, B.J., Rice, A.L., Thurston, M.H., 1995. A quantitative photographic survey of
‘spoke-burrow’type lebensspuren on the Cape Verde abyssal plain. Int. Revue
ges. Hydrobiol 80, 153–170.
Blankenship-Williams, L.E., Levin, L.A., 2009. Living deep: a synopsis of hadal trench
ecology. Marine Technical Society Journal 43 (5), 137–143.
Blankenship, L., Yayanos, A., Cadien, D., Levin, L., 2006. Vertical zonation patterns of
scavenging amphipods from the hadal zone of the Tonga and Kermadec
trenches. Deep-Sea Res 53, 48–61.
Bowen, A.D., Yoerger, D.R., Taylor, C., McCabe, R., Howland, J., Gomez-Ibanez, D.,
Kinsey, J.C., Heintz, M., McDonald, G., Peters, D.B., Bailey, J., Bors, E., Shank, T.,
Whitcomb, L.L., Martin, S.C., Webster, S.E., Jakuba, M.V., Fletcher, B., Young, C.,
Buescher, J., Fryer, P., Hulme, S., 2009. Field trials of the Nereus hybrid
underwater robotic vehicle in the Challenger Deep of the Mariana Trench.
Oceans 2009 Proceedings.
Bruun, A.F., 1956. The abyssal fauna: its ecology, distribution, and origin. Nature
177, 1105–1108.
Clarke, K.R., Gorley, R.N., 2006. PRIMER v6: User Manual/Tutorial. PRIMER-E,
Plymouth.
Davies, H.L., Keene, J.B., Hashimoto, K., Joshima, M., Stuart, J.E., Tiffin, D.L., 1987.
Bathymetry and canyons off the Western Solomon Sea. Geo-Marine Letters 6,
181–191.
De Leo, F.C., Smith, C.R., Rowder, A.A., Bowden, D.A., Clark, M.R., 2010. Submarine
canyons: hotspots of benthic biomass and productivity in the deep sea. Proc. R.
Soc. B, http://dx.doi.org/10.1098/rspb.2010.0462.
De Leo, F.C., Vetter, E.W., Smith, C.R., Rowden, A.A., McGranaghan, M., 2014. Spatial
scale-dependent habitat heterogeneity influences submarine canyon macro-
faunal abundance and diversity off the Main and Northwest Hawaiian Islands.
Deep-Sea Res. II 104, 267–290.
Fodrie, F.J., Levin, L.A., Rathburn, A.E., 2009. High densities and depth-associated
changes of epibenthic megafauna along the Aleutian margin from 20 00–
4200 m. J. Mar. Biol. Ass. U. K 89 (8), 1517–1527.
Fofonoff, N.P., Millard, R.C., 1983. Algorithms for computation of fundamental
properties of seawater. UNESCO Technical Papers in Marine Science 44, 1–53.
Fryer, P., Becker, N., Appelgate, B., Martinez, F., Edwards, M., Fryer, G., 2003. Why is
the Challenger Deep so deep? EPSL 211 (3-4), 259–269.
Fujikura, K., Kojima, S., Tamaki, K., Maki, Y., Hunt, J., Okutani, T., 1999. The deepest
chemosynthesis-based community yet discovered from the hadal zone, 7326 m
deep, in the Japan Trench. Mar. Ecol. Prog. Ser. 190, 17–26.
Fujii, T., Jamieson, A.J., Solan, M., Bagley, P.M., Priede, I.G., 2010. A large aggregation
of liparids at 7703 meters and a reappraisal of the abundance and diversity of
hadal fish. BioSci 60 (7), 506–515.
Gage, J.D., 2003. Food inputs, utilization, carbon flow and energetics (Chapter 11).
In: Tyler, P.A. (Ed.), Ecosystems of the Deep Oceans. Elsevier Science, Nether-
lands, p. 582.
Gage, J.D., Tyler, P.A., 1991. Deep-sea biology: A natural history of organisms of the
deep-sea floor. Cambridge University Press, New York p. 504 (pgs).
George, R.Y., Higgins, R.P., 1979. Eutrophic hadal benthic community in the Puerto
Rico Trench. Ambio Special Report 6, 51–58.
Glud, R.N., Wenzhofer, F., Middelboe, M., Oguri, K., Turnewitsch, R., Canfield, D.E.,
Kitazato, H., 2013. High rates of microbial carbon turnover in sediments in the
deepest oceanic trench on Earth. Nature Geoscience 6, 284–288.
Gooday, A.J., Todo, Y., Uematsu, K., Kitzato, H., 2008. New organic-walled For-
aminifera (Protista) from the ocean's deepest point, the Challenger Deep
(western Pacific Ocean). Zoological Journal of the Linnean Society 153 (3),
399–423.
Hardy, K., Sutphen, B., Cameron, J., 2014a. Technology of the DEEPSEA CHALLENGE
Expedition (Part 1 of 2: The Landers) (July 2014). Ocean News and Technology
〈http://www.oceannews.com/technology-of-the-deepsea-challenge-expedition〉.
Hardy, K., Sutphen, B., Cameron, J., 2014b. Technology of the DEEPSEA CHALLENGE
expedition (Part 2 of 3: DEEPSEA CHALLENGER) (July 2014). Ocean News and
Technology 〈http://www.oceannews.com/technology-of-the-deepsea-challen
ge&-expedition-part-2-of-3-deepsea-challenger〉.
Heezen, B.C., Hollister, C.D., 1971. The face of the deep. Oxford University Press,
New York p. 659 (pgs).
Hessler, R.R., Ingram, C.L., Yayanos, A.A., Burnett, B.R., 1978. Scavenging amphipods
from the floor of the Philippine Trench. Deep-Sea Res. I 25, 1029–1047.
Jamieson, A.J., 2011. Ecology of deep oceans: hadal trenches. In: eLS. John Wiley &
Sons, Ltd, Chichester. http://dx.doi.org/10.1002/9780470015902.a0023606.
Jamieson, A.J., Fujii, T., Mayor, D.J., Solan, M., Priede, I.G., 2010. Hadal trenches: the
ecology of the deepest places on Earth. Trends Ecol. Evol. 25, 190–197 .
Jamieson, A.J., Fujii, T., Solan, M., Matsumoto, A.K., Bagley, P.M., Priede, I.G., 2009a.
First findings of decapod crustacea in the hadal zone. Deep-Sea Res 56,
641–647.
Jamieson, A.J., Fujii, T., Solan, M., Matsumoto, A.K., Bagley, P.M., Priede, I.G., 2009b.
Liparid and macrourid fishes of the hadal zone: in situ observations of activity
and feeding behaviour. Proc. R. Soc. B 276, 1037–1045.
Jamieson, A.J., Gebruk, A., Fujii, T., Solan, M., 2011a. Functional effects of the hadal
sea cucumber Elpidia atakama (Echinodermata: Holothuroidea, Elasipodida)
reflect small-scale patterns of resource availability. Mar. Bio 158, 2695–2703.
Jamieson, A.J., Lorz, A.-N., Fujii, T., Priede, I.G., 2011b. In situ observations of trophic
behaviour and locomotion of Princaxelia amphipods (Crustacea: Pardaliscidae)
at hadal depths in four West Pacific trenches. Journal of the Marine Biological
Association of the United Kingdom 92, 143–150.
Jamieson, A.J., Kilgallen, N.M., Rowden, A.A., Fujii, T., Lorz, A.-N., Kitazawa, K.,
Priede, I.G., 2011c. Bait-attending fauna of the Kermadec Trench, SW Pacific
Ocean: Evidence for an ecotone across the abyssal-hadal transition zone. Deep-
Sea Res 58, 49–62.
Jamieson, A.J., Yancey, P.H., 2012. Validity of the Trieste flatfish: Dispelling the myth.
Biol. Bull. 222, 171–175.
Kahru, M., Kudela, R., Manzano-Sarabia, M., Mitchell, B.G., 2009. Trends in primary
production in the California Current detected with satellite data. J. Geophys.
Res. 114, C02004.
Kamenskaya, O.E., 1984. Ecological classification of deep-sea amphipods, 119. Trudy
Institute of Oceanology, Russia, pp. 154–160 (In Russian).
Kelly, R.H., Yancey, P.H., 1999. High contents of trimethylamine oxide correlating
with depth in deep-sea teleost fishes, skates, and decapod crustaceans. Biol.
Bull. 196, 18–25.
Kupriyanova, E.K., Vinn, O., Taylor, P.D., Shopf, J.W., Kudryavtsev, A.B., Bailey-Brock, J.,
2014. Serpulids living deep: calcareous tubeworms beyond the abyss. Deep-Sea Res
90, 91–104.
Leal, J.H., Harasweych, M.G., 1999. Deepest Atlantic molluscs: hadal limpets
(Mollusca, Gastropoda, Cocculiniformia) from the northern boundary of the
Caribbean Plate. Invertebr. Biol. 118 (2), 116–136.
Lemche, H., Hansen, B., Madsen, F.J., Tendal, O.S., Wolff, T., 1976. Hadal life as
analyzed from photographs. Videnskabelige Meddelelser fra Dansk Naturhis-
torisk Forening 139, 263–336.
Levin, L.A., Etter, R.J., Rex, M.A., Gooday, A.J., Smith, C.R., Pineda, J., Stuart, C.T.,
Hessler, R.R., Pawson, D., 2001. Environmental influences on regional deep-sea
species diversity. Annu. Rev. Ecol Syst. 32, 51–93.
Lindsay, D., Pagès, F., 2010. Voragonema tatsunoko (Trachymedusae: Rhopalonema-
tidae), a new species of benthopelagic medusa, host to the hyperiid amphipod
Mimonectes spandli (Physosomata: Mimonectidae). Zootaxa 2671, 31–39.
Longhurst, A., 1995. Seasonal cycles of pelagic production and consumption. Prog.
Oceanogr. 36, 77–167.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133132

Menot, L., Sibuet, M., Carney, R.S., Levin, L.A., Rowe, G.T., Billett, D.S.M., Poore, G.,
Kitazato, H., Vanreusel, A., Galéron, J., Lavrado, H.P., Sellanes, J., Ingole, B.,
Krylova, E.M., 2010. New perceptions of continental margin biodiversity. In:
McIntyre, A.D. (Ed.), Life in the World's Oceans: Diversity, Distribution, and
Abundance. Wiley-Blackwell.
Miyake, H., Lindsay, D.J., Hunt, J.C., Hamatsu, T., 2002. Scyphomedusa Aurelia
limbata (Brandt, 1838) found in deep waters off Kushiro, Hokkaido. Northern
Japan. Plankton Biol. Ecol 49 (1), 44–46.
Nakanishi, M., Hashimoto, J., 2011. A precise bathymetric map of the world's
deepest seafloor, Challenger Deep in the Mariana Trench. Marine Geophysical
Research 32, 455–463.
Ohara, Y., Reagan, M.K., Fujikura, K., Watanabe, H., Michibayashi, K., Ishii, T., Stern, R.J.,
Pujana, I., Martinez, F., Girard, G., Ribeiro, J., Brounce, M., Komori, N., Kino, M., 2012.
Aserpentinite-hostedecosystemintheSouthernMarianaForearc.Proc.Natl.Acad.
Sci 109, 2831–2835.
Ohta, S., 1984. Star-shaped feeding traces produced by echiuran worms on the
deep-sea floor of the Bay of Bengal. Deep-Sea Res 31 (12), 1415–1432.
Ohta, S., 1985. Photographic observations of the swimming behavior of the deep-
sea pelagothuriid holothurian Enypniastes (Elasipoda, Holothurioidea). J. Ocea-
nographic Society of Japan 41, 121–133.
Okada, H., Ohta, S., 1993. Photographic evidence of variable bottom-current activity
in the Suruga and Sagami bays, central Japan. Sedimentary Geology 82,
221–237.
Osborn, K.J., Madin, L.P., Rouse, G.W., 2011. The remarkable squidworm is an
example of discoveries that await in deep-pelagic habitats. Biol. Lett 7,
449–453.
Osborn, K.J., Rouse, G.W., 2011. Phylogenetics of Acrocirridae and Flabelligeridae
(Cirratuliformia, Annelida). Zoologica Scripta 40, 204–219.
Rex, M.A., Etter, R.J., 2010. Deep-Sea Biodiversity: Pattern and Scale. Harvard
University Press p. 354.
Samerotte, A.L., Drazen, J.C., Brand, G.L., Seibel, B.A., Yancey, P.H., 2007. Correlation
of trimethylamine oxide and habitat depth within and among species of teleost
fish: an analysis of causation. Physiol. Biochem. Zool. 80, 197–208.
Sokolova, M.N., 1994. Euphausiid “dead body rain”as a source of food for abyssal
benthos. Deep-Sea Res 41 (4), 741–746.
Solan, M., Germano, J.D., Rhoads, D.C., Smith, C., Michaud, E., Parry, D., Wenzhofer, F.,
Kennedy, B., Henriques, C ., Battle, E., C arey, D., Iocco, L., Valente, R., Watson, J.,
Rosenberg, R., 2003. Towards a greater understanding of pattern, scale and process
in marine benthic systems: a picture is worth a thousand worms. J. Exp. Mar. Biol.
Ecol. 285-286, 313–338.
Somerfield, P.J., Clarke, K.R., 1995. Taxonomic levels, in marine community studies,
revisted. Mar. Ecol. Prog. Ser. 127, 113–119.
Sweetman, A.K., Chapman, A., 2011. First observations of jelly-falls at the seafloor in
a deep-sea fjord. Deep-Sea Res 58, 1206–1211.
Taira, K., Kitagawa, S., Yamashiro, T., Yanagimoto, D., 2004. Deep and bottom
currents in the Challenger Deep, Mariana Trench, measured with super-deep
current meters. J. Oceanogr. 60, 919–926.
Tiffin, D.L., Honza, E., Keene, J., 1984. A geological and geophysical investigation of
the Western Solomon Sea, Trobriand Basin, and adjacent areas –a cruise report
of the R/V Natsushima. Cruise Report No. 92.
Todo, Y., Kitazato, H., Hashimoto, J., Gooday, A.J., 2005. Simple foraminifera flourish
at the ocean's deepest point. Science 307, 689.
Vinogradova, N.G., 1962. Vertical zonation in the distribution of deep-sea benthic
fauna in the ocean. Deep-Sea Res 8, 245–250.
Watling, L., Guinotte, J., Clark, M.R., Smith, C.R., 2013. A proposed biogeography of
the deep ocean floor. Prog. Oceanogr. 111, 91–112 .
Weber, J., Greer, R., Voight, B., White, E., Roy, R., 1969. Unusual strength properties
of echinoderm calcite related to structure. J. Ultrastruct. Res. 26 (5-6), 355–366.
Wolff, T., 1960. The hadal community, an introduction. Deep-Sea Res 6, 95–124.
Wolff, T., 1961. The deepest recorded fishes. Nature 190, 283–284.
Wolff, T., 1970. The concept of the hadal or ultra-abyssal fauna. Deep-Sea Res 17,
983–1003.
Yancey, P.H., Gerringer, M.E., Drazen, J.C., Rowden, A.A., Jamieson, A., 2014. Marine
fish may be biochemically constrained from inhabiting the deepest ocean
depths. PNAS, http://dx.doi.org/10.1073/pnas.1322003111.
Yonishima, S., Mochizuki, K., Araki, E., Hino, R., Shinohara, M., Suyehiro, K., 2005.
Subduction of the woodlark Basin at New Britain Trench Solomon Islands
region. Tectonophysics 397, 225–239.
N.D. Gallo et al. / Deep-Sea Research I 99 (2015) 119–133 133