Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
PROCEEDINGS
OF
Proc. R. Soc. B (2012) 279, 1017-1026
doi: 10.1098/rspb.2011.1496
Published online 7 Septem ber 2011
'
THE ROYAL
SOCIETY
A large population of king crabs in Palmer
Deep on th e w est Antarctic Peninsula shelf
and potential invasive im pacts
C ra ig R . S m i t h 1 *, L a u r a J. G r a n g e 1, D a v id L. H o n ig 2,
L ie v e n N a u d ts 3^ , B r u c e H u b e r 4, L io n e l G u id i1 a n d E u g e n e D o m a c k 5
1Department o f Oceanography, University o f H aw aii a t M anoa, 1000 Pope Road, Honolulu, H I 96822, U SA
2M arine Laboratory, Nicholas School o f the Environment, D uke University, 135 M arine Lab Road,
Beaufort, N C 28516, U SA
3Renard Centre o f M arine Geology, G hent University, Krijgslaan 281 s8, G ent 9000, Belgium
4D ivision o f Ocean and Climate Physics, Lam ont-D oherty Earth Observatory o f Columbia University,
Palisades, N Y 10964, U SA
5Department o f Geosciences, H am ilton College, 198 College H ill Road, Clinton, N ew York, N Y 13323, U SA
L ith o d id crabs (and o th e r sk eleton-crush in g p red ato rs) m ay have b e e n excluded from cold A n tarctic
co n tin en tal shelf w aters for m o re th a n 14 M yr. T h e w est A n tarctic P en in su la shelf is w arm in g rapidly
an d has b e e n hypothesized to be soon in v ad ed b y lithodids. A rem o tely o p erated vehicle survey in
P alm er D eep, a b asin 120 km on to th e A n tarctic shelf, revealed a large, rep ro d u ctiv e p o p u la tio n o f lith o ­
dids, providing the first evidence th a t king crabs have crossed th e A n tarctic shelf. D N A seq u en cin g an d
m orphology indicate th e lith o d id is Neolithodes yaldw yni A hyong & D aw son, previously re p o rte d only
from Ross Sea w aters. We estim ate a N . yaldw yni p o p u la tio n d ensity o f 10 600 k m ~ 2 an d a p o p u la tio n
size o f 1.55 X IO6 in P alm er D ee p , a density sim ilar to lith o d id p o p u latio n s o f co m m ercial in terest
aro u n d A laska an d S o u th G eorgia. T h e lith o d id o cc u rre d at d ep th s o f m o re th a n 850 m an d te m p eratu re s
o f m o re th a n 1.4°C in P alm er D eep , an d was n o t fo u n d in extensive surveys o f th e co ld er shelf a t d ep th s
o f 4 3 0 - 7 2 5 m . W here N . yaldw yni o ccu rred , crab traces w ere a b u n d a n t, m egafaunal diversity red u ced
an d ech in o d erm s absent, suggesting th a t th e crabs have m ajo r ecological im pacts. A n tarctic P eninsula
shelf w aters are w arm in g at approxim ately 0 .0 1°C y r-1 ; if N . ya ld w yn i is cu rren tly lim ited b y cold te m p ­
eratures, it could spread u p o n to th e shelf (4 0 0 -6 0 0 m d ep th s) w ith in 1 - 2 decades. T h e P alm er D ee p N.
yaldw yni p o p u la tio n provides an im p o rta n t m o d el for th e p o te n tia l invasive im p acts o f cru sh in g p red ato rs
on vulnerable A ntarctic shelf ecosystem s.
K e y w o rd s: lithodids; invasion; A n tarctic; clim ate w arm ing; b io tu rb a tio n ; biodiversity loss
1. INTRODUCTION
It has recently b een argued th a t lithodid crabs, an d m any
other large durophagous (crushing) p redators, have b een
excluded from th e cold w aters o f th e high A ntarctic conti­
n ental shelf (less th a n 600 m depths) for m ore th a n 14 M y r
[1,2]. It is hypothesized th a t the L ithodidae have failed to
colonize th e A ntarctic shelf due to low er th erm al lim its o f
0 .4 -2 ° C for adults and/or larvae [ 1 - 3 ] . It is fu rth e r specu­
lated th a t th e absence for m illions o f years o f lithodids an d
other durophagous pred ato rs (e.g. sharks, brachyuran
crabs, a variety o f bony fish) from A ntarctic shelf eco­
systems has allowed th e developm ent o f diverse epifaunal
com m unities th a t are likely to b e highly vulnerable to
invasion by crushing pred ato rs [1,2,4].
M a rin e ecosystem s o f the w est A n tarctic P eninsula
(WAP) are w arm in g rapidly, w ith sea surface te m p era­
tu res o n th e WAP shelf rising b y approxim ately 1°C
since 1950 [ 5 -7 ] . S u ch w arm in g m ay allow the en try of
*A u th o r fo r c o rre sp o n d e n c e (craigsm i@ haw aii.edu ).
^P resent address: M U M M , Royal Belgian Institute o f N atu ral Sciences,
3 d e e n 23ste L inieregim entsplein, B -8400 O ostende, Belgium .
E le ctro n ic su p p le m e n ta ry m aterial is available a t h ttp ://d x .d o i.o rg /
1 0 .1 0 9 8 /rsp b .2 0 1 1 .1 4 9 6 o r via h ttp ://rsp b .ro y also ciety p u b lish in g .o rg .
Received 15 Ju ly 2011
Accepted 18 A u g u st 2011
te m p eratu re -lim ited taxa o n to th e A n tarctic shelf [1]. A t
least 14 species o f lithodids are k now n from th e S o u th ern
O cean n o rth o f 60° S, a n d significant n u m b e rs o f lith o ­
dids have recently b ee n rec o rd e d at shelf an d slope
d ep th s from A n tarctic Islands as far so u th as 67.5° S in
th e B ellinghausen Sea an d 75° S in th e Ross Sea
[2,8,9]. In ad d itio n , tw o lith o d id species (Neolithodes
capensis an d Paralomis birsteini) have b e e n re p o rte d from
th e W AP co n tin en tal slope itself, in clu d in g 13 specim ens
(and one juvenile) o f P. birsteini at d ep th s o f 1 1 2 3 1304 m n o rth w est o f A delaide Islan d (figure 1) [2,4,9].
T h ese rec en t reco rd s o f N . capensis an d P. birsteini are
h y p o th esized to in d icate an ongoing lith o d id invasion
up th e WAP slope an d o n to th e shelf effected b y seafloor
im m ig ratio n o f ad u lts or dispersal o f d em ersal larvae
[1,2]. H ow ever, th e rep ro d u ctiv e status an d p o p u la tio n
densities o f N. capensis, P birsteini an d o th e r lith o d id s on
th e W AP slope are u n k n o w n (no ovigerous fem ales have
b ee n collected o n th e W AP slope), a n d th e ecosystem
im pacts o f these an d o th e r lith o d id species rem ain
u n d o c u m e n te d in A n tarctic w aters.
H ere, we re p o rt th e p resen ce o f an ap p aren tly large,
rep ro d u ctiv e p o p u la tio n o f th e lith o d id crab Neolithodes
yaldw yni in P alm er D eep , a 1440 m deep b asin on the
1017
T h is jo u rn a l is © 2011 T h e R oyal S ociety
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
1018
C . R. S m ith et al.
K ing crabs crossing the Antarctic S h elf
(b)
( a)
60° S
-5000 -4000 -5000 -2000 -1000
70°W
King George island
68°W
66°W
0
64°W
63° S
63° S
64° S
64° S
65° S
65° S
62° S
Palm er Deep
64° S
Bellingshausen
Sea
i
66° S
Adelaide
Island
W eddell Sea
68° S
66° S
66° S
70° S
80° W
72°W
64° W
56°W
70°W
68°W
Palmer
_Deep
66°W
64°W
(c)
48° S
56.6°
52°S
56.8°
56°S
57.0°
65° S
4° S
65°W
48°W
36°W
24°W
12°W
64°W
Figure 1. (a) West A ntarctic Peninsula region, including Palm er D eep and stations where ROV surveys, seafloor imaging sur­
veys and bottom trawls were conducted betw een 1998 and 2010. C ontours are in m etres. Red dots, locations where Thajte et al.
[4] observed Paralomis birsteini at depths of 1 1 2 3 -1 3 0 4 m. G reen dots, locations where Thatje et al. [4] failed to find lithodids
during ROV surveys (depths approx. 5 0 0 -6 0 0 m ). Blue dots, station locations AA, B, E, F and G of trawls and yoyo camera
surveys conducted in 2 0 0 8 -2 0 0 9 during F O O D B A N C S2 cruises, at depths of 5 7 8 -5 9 7 m , with no crabs detected (see elec­
tronic supplem entary m aterial, table S2). Boxes, locations of trawls and/or yoyo camera surveys conducted in 2010 during the
LARISSA cruise at depths of 4 3 0 -7 2 5 m , with no crabs detected (see electronic supplem entary m aterial, table S2). 1, Barilari
Bay; 2, Flandres Bay; 3, Andvord Bay; 4, H ughes Bay; 5, Lockyer Station and Admiralty Sound. (b) Swath and G EO SA T sea­
floor bathym etry (depths in metres) in the vicinity of Palm er D eep (box), illustrating the typical depth (5 0 0 -6 0 0 m) of the
WAP shelf, and the sill depth (approx. 450 m) of Palm er D eep (bathym etric data from [10]). (c) Swath bathym etry of
Palm er D eep showing ROV transect (solid black line indicates portion of transect m ore th an 850 m , red line indicates portion
of transect less than 850 m). Red vertical lines indicate positions of crabs (double length indicates crab pairs), yellow line indi­
cates location of collected individual, black line indicates dead lithodid. Red stars, locations of seabed photographs from 1998.
in n e r WAP co n tin en tal shelf 120 km from th e op en slope.
T h is is th e first evidence th a t lithodids can cross th e A n t­
arctic co n tin en tal shelf, an d th e first re p o rt o f N . yaldw yni
b ey o n d th e Ross Sea. We p re se n t d ata on th e d e p th distri­
b u tio n , ecosystem effects an d evidence o f rep ro d u ctio n
for N . yaldw yni, b ased o n a rem otely o p erated vehicle
(ROV) survey a n d crab collection. O u r d ata indicate
th a t this king crab has m ajo r im pacts on sed im en t tex tu re,
b io tu rb a tio n an d diversity o f epib en th o s in P alm er D eep
below 950 m , a n d th a t lithodids are likely to have b een
p rese n t in P alm er D ee p at least since 1998.
2. STUDY A REA , MATERIALS AND M ETHODS
(a) S tu d y a rea
P alm er D eep is a m ud-floored basin located approxim ately
120 km southeast o f the WAP continental shelf-slope break
Proc. R. Soc. B (2012)
w ith a m axim um d ep th o f approxim ately 1440 m (figure 1)
[10]. T h e basin is approxim ately 14 x 8 km an d is
em bed d ed in a continental shelf w ith a m ean regional
d ep th of 4 0 0 - 6 0 0 m (figure 1) [11]. S edim entation rates
in P alm er D eep are 0 .1 - 0 .4 cm yr 1 [12], an d oxygen con­
centrations exceed 4 m l 1 1 at all d epths (Palm er L T E R
1 9 9 4 -2 0 0 8 , unpu b lish ed data). A ny w ater m ass or plank­
tonic larvae crossing the WAP shelf from the S o u th ern
O cean into P alm er D eep m u st pass over sill depths less
th a n or equal to 450 m on the shelf (figure 1) [10].
W ater m ass characteristics o f the WAP shelf an d P alm er
D eep are com plex. T h e to p approxim ately 200 m o f the
w ater colu m n is characterized b y A ntarctic Surface W ater
(AASW ), w hich w arm s seasonally n ear th e surface b u t
rem ains colder th a n 1°C below 1 0 0 m [6,7,13,14].
Below approxim ately 200 m d ep th , th e WAP shelf an d
P alm er D eep are influenced by U p p e r C ircu m p o lar D eep
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
K in g crabs crossing the Antarctic S h elf
C . R. S m ith et al.
1019
W ater (U C D W ), w hich floods an d m ixes eastw ard onto
th e shelf driven by the A ntarctic C ircum polar C u rren t
[6,14,15]. T h e U C D W is typically w arm er (m ore th a n
1.0°C) an d m ore saline th a n the deep A A SW im m ediately
above [13]. W arm ing since 1990 in th e W AP region is
rep o rted to have increased th e h ea t co n ten t o f the
U C D W influencing the WAP shelf in th e P alm er D eep
region, potentially w arm ing the u p p er 300 m o f the
U C D W by approxim ately 0.7°C [7].
(b) R e m o te ly o p e r a te d veh icle s u r v e y
We u sed Genesis, a Cherokee (Sub A tlantic) rem otely o per­
ated vehicle (ROV) operated by the R enard C en tre o f
M arin e Geology, deployed from th e R V IB N. B. Palmer to
co n d u c t a 5.2 h survey along th e floor o f P alm er D eep at
approxim ately 1419 m an d up its eastern slope to a d ep th
o f 600 m (figure 1; see electronic supplem entary m aterial,
table S I). M etre-scale positioning o f th e RO V was achieved
w ith an IX S E A u ltra-sh o rt baseline positioning (U SBL)
system. T h e RO V carried forw ard-looking colour digital
video an d digital still cam eras o n p an -an d -tilt m ountings,
parallel laser scalar beam s (10 cm separation), a stan d ­
alone m em ory C T D (Sea & S un Technology C T D 9 0 M ;
accuracy 0.005°C ), a sm all m an ip u lato r arm an d a sam ple
drawer. T h e RO V was low ered to the P alm er D eep floor,
an d th e n driven eastw ard at approxim ately 0.5 knots an d
an altitude o f 0 .5 - 1 m (figure 1). T h e RO V generally
m oved in a straight line, occasionally stopping to obtain
video an d still im ages o f m egafauna an d b o tto m features,
to collect m egafaunal specim ens an d to obtain video o f
crab activities. A single lithodid crab was collected by the
m an ip u lato r and recovered in th e RO V draw er (figure 2;
see electronic supplem entary m aterial, table S I).
(c) D e n sity e s tim a te s o f c ra b s a n d o th e r m eg a fa u n a
T h e ab u n d an ce a n d 9 5 p e r cent confidence lim its (C L ) of
lithodids an d o th e r m egafauna along th e R O V tran sect
w ere evaluated using line-transect sam pling m eth o d s
[ 1 6 , 1 7 ] an d th e softw are D i s t a n c e v . 6 . 2 . L in e-tran sect
sam pling is widely u sed for estim ating the ab u n d an ce of
biological p opulations along linear transects, su ch as o u r
RO V survey [ 1 7 ] . F o r line-tran sect analysis, th e digital
RO V video w as replayed, an d lithodids an d o th e r identifi­
able m egafauna (m ore th a n or equal to 5 cm in m in im u m
dim ension) w ere co u n ted as th e RO V m oved in a straight
line at an altitude o f approxim ately 0 . 5 m w ith good visi­
bility; no counts were m ad e w hen the R O V deviated from
a linear course or w hen sedim ent was stirred u p by RO V
thru sters. L ith o d id positions along th e tran se ct are b ased
on G P S fixes from th e RO V U S B L system. T h e distance
o f each m egafaunal individual from th e centreline o f the
transect, as the anim al cam e into RO V view, was m easu red
using the laser scale. T h e RO V field o f view (average
2 5 0 cm w idth) w as also m easu red w ith th e laser scale.
E rrors in distance m easures from th e centreline w ere esti­
m a ted to be + 5 cm , b ased o n replicate m easurem ents.
We th e n u sed th e pro g ram D i s t a n c e v . 6 . 2 to select the
b est m o d el (from four alternatives) for th e d istrib u tio n of
anim als from th e tran se ct centreline b ased on th e Akaike
in form ation criterion [ 1 7 ] : th e uniform key with cosine
adjustments m odel best fitted th e crab data; th e h azard rate
simple polynomial m odel best fitted th e n o n -crab m ega­
fauna. F o r estim ates b ased on lin e-transect sam pling to
Proc. R. Soc. B (2012)
Figure 2. (a) ROV photograph of two N. yaldwyni at 1411 m
in Palm er Deep. C rab in foreground is foraging in the sedi­
m ent with its chelipeds. N ote the crab traces (gashes and
punctures) covering the sedim ent surface. (b) D orsal view
of the collected specim en of N. yaldwyni, (c) Ventral view
of N. yaldwyni with brood pouch closed and (d ) brood
pouch open to expose developing eggs.
b e u n b iased , th e follow ing assum ptions m u st b e m e t
[17]: (i) objects o n th e survey centreline can b e d etected
w ith a p robability o f 1.0, (ii) objects do n o t m ove in
response to th e RO V p rio r to m e asu rem en t an d (iii) dis­
tances are m easu red w ith little o r n o error. T h ese
assu m p tio n s are reasonably well m e t in o u r dataset.
T h e ap p ro x im ate ab u n d a n ce o f lith o d id b io tu rb a tio n
traces on th e seab ed was also evaluated along th e RO V
survey using th e laser scale. T h e ab u n d a n ce o f crab b io ­
tu rb a tio n features w as classified as h ig h (m ore th a n 100
crab traces m ~ 2, w ith m an y traces overlapping), m o d erate
(n u m ero u s isolated traces visible) or low /absent (few to
no lith o d id traces recognizable).
To d eterm in e w h eth e r th e diversity o f ep ifau n a on
drop stones differed in d ep th zones w ith a b u n d a n t lith o ­
dids (i.e. at d ep th s m o re th a n 950 m ) co m p ared w ith
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
1020
C . R. S m ith et al.
K ing crabs crossing the Antarctic S h elf
d ep th s w ith o u t lith o d id crabs (less th a n 850 m ), w e eval­
u a te d species density (i.e. n u m b e r o f species p e r u n it
area) on drop stones resting on th e seafloor. U sin g the
laser scale, we estim ated the p la n areas o f drop stones vis­
ible as th e R O V m oved along its survey line an d co u n ted
th e n u m b e r o f distinguishable epifaunal species o n each
dro p stone. T h e overlapping d istrib u tio n o f anim als on
densely covered drop stones p rec lu d ed co u n ts o f individ­
u al ab u n d a n ce . We only in c lu d ed species co u n ts for drop
stones ranging in size from 200 to 1000 cm 2, to avoid
biases resulting from th e effects o f d ro p -sto n e size (i.e.
h ab itat-islan d area) on species diversity. Species richness
p e r drop stone was n orm alized to 500 cm 2, th e approxi­
m ate m ean size o f drop stones studied. D ifferences in
m e an dro p -sto n e size an d species density b etw een d ep th
zones w ere assessed w ith t-tests, an d differences in the
p ro p o rtio n o f ‘clean’ drop stones (devoid o f co u n tab le
epifauna) w ere evaluated w ith F ish e r’s exact test. S tatisti­
cal tests w ere c o n d u c te d in M i n i t a b v . 15. F o r sedim entdw elling epibenthic m egafauna, diversity com pariso n s
w ere m ade w ith rarefaction diversity curves for to tal
m eg afauna from m o re th a n 950 m a n d less th a n 850 m ,
u sin g th e softw are P r i m e r v. 6 .
(d) S e a b e d p h o to g r a p h s
Still p h o to g rap h s o f the P alm er D eep seafloor were
o b ta in ed in 1998 from th e R V L . M . Gould d u rin g
cruise L M G 9 8 -0 2 , w ith a vertically dow nlooking B enthos
372 C a m e ra an d S trobe system deployed o n a C T D
fram e w ith a 3 m b o tto m -c o n ta c t sw itch. T h e b o tto m
area p h o to g ra p h e d in each fram e (5.3 m 2) w as calculated
from cam era view angles an d elevation.
(e) H is to r ic a l h y d ro g ra p h ic d a ta
W ater-colum n profiles o f in situ te m p eratu re an d salinity
for th e P alm er D ee p vicinity from 1 9 8 0 -2 0 0 9 were
o b ta in ed from th e N atio n al O ceanographic D a ta
C e n tre, from th e P alm er L T E R P ro g ram (courtesy of
D. M a rtin so n an d R. Iannuzzi), an d u n p u b lish ed (b u t
calibrated) data from cruise L M G 9 8 -0 2 .
(f) D N A id e n tific a tio n o f th e lith o d id
A 650 b ase-p air region o f the m ito ch o n d rial gene for
cy to chrom e oxidase c su b u n it I (CO I) was seq u en ced to
identify th e lith o d id collected in P alm er D eep using
accep ted sequencing p rotocols (details pro v id ed in elec­
tro n ic su p p le m e n tary m aterials). A consensus sequence
w as p ro d u c e d in S e q u e n c h e r v. 4.7 an d co m p ared against
G en B an k an d th e B arcode o f Life databases.
3. R E SU LT S
(a) L ith o d id a b u n d a n c e , d e p th d is tr ib u tio n a n d
e s tim a te d p o p u la tio n s iz e
T h e RO V survey ran g ed from 1419 to 600 m d ep th s,
along a to ta l distance o f approxim ately 3.3 km (figure 1;
see electronic su p p le m e n tary m aterial, table S I).
D u rin g th e survey, we observed 42 live lith o d id crabs
an d one d ead specim en (or exuvium ) along the 2.0 2 km
o f th e survey b etw een 1419 an d 850 m d ep th s (i.e.
w ith in th e d e p th zone w here crab traces w ere presen t;
figures 1 - 3 ; see discussion o f traces below ). All lithodids
o cc u rre d o n the flat, m u d d y seafloor. E ig h teen crabs were
ex am ined in close-up video or digital still im ages (e.g.
Proc. R. Soc. B (2012)
figure 2a) an d th ey closely resem b led th e collected speci­
m e n in m o rp h o lo g y an d size (w ith carapace w id th s of
approx. 8 - 1 0 cm ); th e rem ain in g 24 crabs also resem b led
th e collected specim en, b ased on visible features. T h u s,
all 42 o bserved live lith o d id s ap p e ar to b e in th e genus
Neolithodes, a n d m o st are p ro b ab ly a single species, id e n t­
ified below as N . yaldw yni (figure 2). C rab s w ere
d istrib u ted b ro ad ly along o u r tran se ct below a d e p th o f
950 m , o ccu rrin g m ostly as single individuals, b u t th ree
tim es in pairs w ith in 2 m o f one an o th er. We estim ate
crab a b u n d a n ce w ith in th e 8 5 0 - 1 4 1 9 m d e p th range to
b e 10 600 k m -2 (95% C L s 8 0 3 0 - 1 3 900 k m -2 ), b ased
on lin e -tran sect sam p lin g [16] (table 1). T h e area
o f P alm er D eep b elow 950 m (i.e. th e d ep th zone w here
crabs a n d crab traces w ere m o st a b u n d a n t) is 146 km 2;
if we assum e th a t th e m e an a b u n d a n ce (95% C L ) of
crabs p e r km 2 on o u r tran se ct are representative o f th e
entire P alm er D eep area below 950 m , th e n this yields a
N. yaldw yni p o p u la tio n size o f 1.55 x IO6 individuals
w ith in P alm er D eep (95% C L o f 1.17 to 2 .04 x IO6).
(b) L ith o d id id e n tifica tio n
A single ovigerous fem ale crab (carapace w id th =
103.6 m m , carapace length = 117.3 m m , po sto rb ital cara­
pace length = 103.2 m m ), carrying nearly m atu re eggs
an d larvae, was collected at a d ep th o f 1154 m (figure 2;
see electronic supplem entary m aterial, table S I). A 650
nucleotide region o f th e C O I gene for this fem ale specim en
was identical to C O I sequences o f tw o N. yaldw yni speci­
m ens identified by S. A hyong, a co-describer o f th e species
[18] (see electronic supplem entary m aterial, figure S I).
T h e P alm er D eep N. yaldw yni C O I sequences are sub stan ­
tially distinct from those o f th e closely related species
N. brodiei, N. asperrimus an d N. duhameli (see electronic
supplem entary m aterial, figure S I). M orphologically, the
collected specim en keys o u t to N. yaldw yni [19] an d m atches
th e diagnosis o f Aí yaldw yni [18,19]. We are therefore co n ­
fident th a t this gravid fem ale is N . yaldwyni.
(c) L ith o d id lo c o m o tio n , fe e d in g a n d b io tu rb a tio n
N u m e ro u s king crabs w ere o bserved w alking across th e
P alm er D eep seafloor, creating p u n c tu re m arks w ith
leg tips a n d gashes in th e sed im en t as leg tips w ere
dragged th ro u g h th e sed im en t (figures 2 a n d 3; see elec­
tro n ic su p p le m e n tary m aterial, videos S I an d S2).
P u n c tu re m arks from leg tips w ere approxim ately 1 cm
in d ia m eter an d gashes w ere up to approxim ately 1 cm
w ide, approxim ately 0.5 cm deep a n d u p to 20 cm
long. In ad d itio n , crabs w ere o bserved to forage by
repeatedly p ro b in g th e sed im en t to 3 - 4 cm d epths
w ith o p en chelipeds, closing th e chelipeds w hile th e
tips w ere su b m erg ed in th e sed im en t, an d th e n scooping
th e c a p tu re d sed im en t b o lu s in to o p en m o u th p a rts for
pro cessin g (figure 2a; see electronic su p p lem en tary
m aterial, videos S I an d S2). T h is p ro b in g an d scooping
activity created sed im en t p u n c tu re m arks an d clum ps 2 3 cm across (figure 3). M u c h o f th e sed im en t scooped to
th e m o u th p a rts fell th ro u g h th e m o u th p a rts d u rin g p ro ­
cessing (see electronic su p p le m e n tary m aterial, video
S I). T h is lo co m o tio n an d foraging activity created a
dense fabric o f gashes, p u n c tu re s an d sed im en t clum ps,
p ro d u c in g a characteristic gash ed /lu m p y tex tu re to th e
s e d im e n t-w a te r interface (figures 2a an d 3). T h e crab
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
K in g crabs crossing the Antarctic S h elf
C . R. S m ith et al.
1021
Figure 3. (a) ROV photograph of N. yaldwyni traces in seafloor sediments at 1418 m depth in Palm er D eep. Brackets, gashes in
sedim ent form ed by N. yaldwyni walking. Arrows, paired puncture marks and clumps form ed during lithodid foraging in the
sedim ent with chelipeds (see figure 2a; see electronic supplem entary material, videos SI and S2). N ote that fresh traces cross
over older traces. (b) Palm er D eep seafloor at 836 m depth, illustrating the absence of N. yaldwyni traces, (c) Vertical p hoto­
graph of Palm er D eep seafloor at 1425 m in 1998 (southern red star in figure le). Brackets, apparent lithodid walking
gashes (straight and V-shaped); arrows, puncture marks and sedim ent clumps similar to those form ed during lithodid foraging.
(d) Exam ple of a drop stone from the N. yaldwyni depth zone (more th an 950 m) with only hexactinellid sponges and serpulid
w orm tubes visible (1360 m ). N ote crab traces in the sediments around the stone, (e) Exam ple of a drop stone from above the
N. yaldwyni depth zone (less than 850 m ), with brisingid asteroids, ophiuroids, an echinoid, hexactinellid sponges and serpulid
w orm tubes visible (811 m).
Table 1. Estim ated population densities (and 95% CLs) for N. yaldwyni and other (non-crab) sediment-dwelling epibenthic
m egafauna along the ROV transect in Palm er Deep. Neolithodes yaldwyni density was calculated over the depth interval where
crabs were observed and crab bioturbation traces evident (8 5 0 -1 4 1 9 m). Densities of other sedim ent dwelling megafauna
were calculated for depths of more than 950 m (N. yaldwyni present and crab traces abundant) and less than 850 m
(N. yaldwyni and crab traces absent) to evaluate lithodid influence.
taxon
depth zone (m)
density (km 2)
Neolithodes yaldwyni
cerianthid anemones
cerianthid anemones
non-crab megabenthos w ithout cerianthid anemones
non-crab megabenthos w ithout cerianthid anemones
>850
<850
>950
<850
>950
10 600
15 000
47 800
4670
4290
(8 5 0 -1 4 1 9 )
(6 0 0 -8 5 0 )
(9 5 0 -1 4 1 9 )
(6 0 0 -8 5 0 )
(9 5 0 -1 4 1 9 )
lower 95%
C L (km -2 )
upper 95%
C L (km -2 )
8030
3710
43 300
13 900
60 500
52 700
a
a
3070
6010
aS am p le size to o sm all to calcu late C L s.
loco m o tio n
an d
foraging
activities
in
P alm er
D eep ap p eared to substantially m ix sedim ents to d ep th s
o f 3 - 4 cm .
Proc. R. Soc. B (2012)
T h e ab u n d a n ce o f lith o d id w alking an d feeding traces
was h ig h (approx. 1 0 0 - 3 0 0 lith o d id traces p e r m 2) an d
th e gash ed /lu m p y sed im en t tex tu re w as co n tin u o u s from
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
1022
C . R. S m ith et al.
K ing crabs crossing the Antarctic S h elf
d ep th s o f 1419 to approxim ately 950 m along th e tran se ct
(figures 2a an d 3). F ro m depths o f approxim ately 950 m
to approxim ately 850 m , the lith o d id trace ab u n d an ce
g radually dim in ish ed from high to low /absent u n til there
w as n o evidence o f crab activity o n the sed im en t shallow er
th a n approxim ately 850 m (figure 3). A t dep th s o f less
th a n 850 m , th e texture o f soft sedim ents was relatively
sm o oth, an d characterized b y w orm tubes an d sm all b io ­
genic m o u n d s (features m issing at d ep th s m o re th a n
950 m ), indicatin g th a t w alking an d foraging N . yaldw yni
w ere rare o r ab sen t shallow er th a n approxim ately 850 m
in P alm er D ee p (figure 3). A t dep th s o f 8 5 0 - 6 0 0 m in
P alm er D eep, sed im en t tex tu re an d b io tu rb a tio n features
w ere sim ilar to tho se o f th e m u d d y seafloor observed over
w ide areas o f th e W AP shelf at d ep th s o f 4 3 0 - 7 2 5 m at
latitu d es from 63 to 68° S, w here lith o d id crabs have
n o t b ee n rec o rd ed d u rin g extensive b o tto m p h o to g rap h ic
surveys an d traw l sam pling (figure 1; see electronic
su p p le m e n tary m aterial, table S2; [20]; C . R. S. S m ith,
L. J. G ran g e & C . C lark 2 0 0 8 - 2 0 1 0 , u n p u b lish ed d ata).
All th ree p h o to g rap h s o f th e P alm er D eep floor from
1998 (cruise L M G 9 8 -0 2 ), tak en at tw o locations
1 - 4 km from th e R O V tran se ct (figure 1), show a sedi­
m e n t surface covered w ith a gashed/lum py texture
(figure 3). T h is sed im en t surface tex tu re is essentially
id en tical to th a t along the RO V tran se ct b etw een dep th s
o f 1419 an d 950 m (figure 3), an d suggests th a t foraging
lith o dids w ere p rese n t at approxim ately 1400 m d ep th s in
P alm er D eep in 1998, 12 years p rio r to o u r RO V survey.
A lth ough no crabs were observed in the p h otograp h s, at
th e density o f crabs estim ated from o u r RO V survey
(approx. 0.01 m 2), the probability o f seeing a lithodid
in th ree p hotographs covering a total area o f approxim ately
16 m 2 was very sm all (less th a n 2% ).
(d) A b u n d a n c e a n d d iv e r s ity o f o th e r b en th ic
m eg a fa u n a
T h e ab u n d a n ce o f sedim ent-dw elling, epib en th ic m ega­
fau n a co u n tab le in the R O V video w as approxim ately
th reefold g reater w ithin the crab zone at d ep th s m ore
th a n 950 m th a n at dep th s w ith o u t crabs (i.e. less th a n
850 m ; table 1); how ever, 90 p e r ce n t o f m egafaunal
ab u n d a n ce at d ep th s o f m o re th a n 950 m consisted o f a
single species o f cerian th id anem one, w hile cerian th id s
co n stitu ted 76 p e r cent o f ab u n d a n ce at dep th s o f less
th a n 850 m . T h e a b u n d a n ce o f n o n -ce ria n th id , sedi­
m en t-dw elling m egafauna was th u s roughly sim ilar at
d ep th s o f m o re th a n 950 m an d less th a n 850 m
(approx. 4290 k m -2 a n d approx. 4670 k m -2 , resp ect­
ively; table 1). R arefaction diversity o f the sedim entdw elling m egafauna was substantially higher at dep th s
o f less th a n 850 m th a n at d ep th s o f m o re th a n 950 m
(figure 4). T h is was a co n sequence o f b o th greater tax o ­
n o m ic richness (curve endpoints) an d m u c h g reater
evenness (initial curve slopes) at th e shallow er dep th s
(less th a n 850 m ) w here N. yaldw yni was absent. E chinod erm s w ere ab sen t from th e sedim ent-dw elling epibenthic
m eg afauna at dep th s o f m ore th a n 950 m , w hile fo u r
species o f ec h in o d e rm (three asteroids an d a cidaroid
u rch in) occu rred on sedim ents at d ep th s o f less th a n
8 50 m . Tw o species o f d em ersal fishes (a n o to th e n iid
a n d a zoarcid) o cc u rre d on sedim ents at dep th s o f m ore
th a n 950 m , a n d only th e n o to th e n iid species occu rred
Proc. R. Soc. B (2012)
8
7
05 ^
« 6
5
§ 4
_o 3
u
a, ?
xra z
1
0
50
100
150
no. individuals
200
250
Figure 4. Rarefaction diversity curves for the sedim entdwelling m egafauna in Palm er D eep at depths of m ore than
950 m (where N. yaldwyni and its traces were abundant;
dashed line) and less th an 850 m (where N. yaldwyni and
its traces were absent; solid line).
at d ep th s o f less th a n 850 m . F ish densities w ere h igher
at m o re th a n 950 m th a n at less th a n 850 m d ep th s
(3700 versus 1900 k m -2 ), b u t sm all sam ple sizes (« =
20 an d 1, respectively) p rec lu d ed statistical significance
u sing lin e-tran sect analysis.
Species richness an d species density o f epifauna on
d ro p stones w as substantially higher, an d th e p ercen tag e
o f clean d ro p stones substantially low er, at d ep th s o f
less th a n 850 m (i.e. above th e d ep th zone o f Aí yaldwyni)
th a n w ith in th e lith o d id d ep th zone at d ep th s o f m o re
th a n 950 m in P alm er D ee p (figure 3, table 2). In p a rti­
cular, m e an species d ensity (i.e. n u m b e r o f species
p e r 500 cm 2) was fourfold low er (tw o-sam ple t-test,
p = 0 .0 0 6 ), an d th e p ercen tag e o f d ro p stones devoid o f
epifauna 3 .5-fold h ig h er (F ish er’s exact test, p < 0 .0 0 3 ),
in th e p resen ce o f N . yaldw yni. E ch in o d erm s w ere
w holly ab sen t from d ro p stones at d ep th s o f m o re th a n
950 m , w hile fo u r species (a b risin g id asteroid, tw o
species o f o p h iu ro id s an d an echinoid) o cc u rred on
d ro p stones at d ep th s o f less th a n 850 m.
(e) P a lm e r D eep h y d ro g ra p h y
D u rin g th e RO V dive, in situ seafloor tem p eratu res
declined from 1.53°C at 1419 m d e p th to 1.45°C at
600 m (figure 5). Since th e lith o d id an d its traces d isap ­
p ea red at d ep th s o f less th a n 850 m , N. yaldw yni
ap p eared to b e restricted to w ater te m p eratu re s m o re
th a n o r eq u al to 1.48°C w ith in P alm er D eep.
B o tto m te m p eratu re s in P alm er D eep have exhibited a
w arm in g tre n d from 1982 to 2 0 1 0 , w ith th e te m p eratu re
m in im u m increasin g from 1.20°C to 1.47°C below 800 m
d ep th s (i.e. increasing at a rate o f approxim ately
0.01 C y r '; figure 5). In 1998, w h en crab traces w ere
p resen t, th e P alm er D eep b o tto m tem p eratu re was
approxim ately 1.4°C . D u rin g th e tim e interval 1 9 9 0 2 005, co m p o site su m m er a n d w in ter h y d ro g rap h ic sec­
tions across th e WAP shelf in to P alm er D eep suggest
th a t b en th ic o r larval stages o f lithodids dispersing n ear
th e seafloor across th e sill d ep th o f approxim ately 450 m
in to P alm er D ee p w o u ld p ro b ab ly have ex perienced te m ­
p eratu res no co ld er th a n approxim ately 1.0°C (figure 5;
see electronic su p p le m e n tary m aterial, figure S2).
M in im u m te m p eratu re s w ith in th e d e p th range o f
4 0 0 - 6 0 0 m in P alm er D eep (i.e. at th e m e an d e p th of
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
K in g crabs crossing the Antarctic S h elf
(b)
(a)
C . R. S m ith et al.
400
,2010
600
2005
64.6° S
800
65.0° S
1000
1023
2000
1995
1990
65.4° S
1200
66°W
65° W
64° W
63°W
1985
'
1400
1.1
(c)
1.2
1.3
1.4
temperature ( °C )
1.5
1980
( d )
r=0.0101 xyear-18.7157 0 -0 .9 2 )
1.4
^
ü
1.4
O
3
r—
03
<U
B
S
1980 1985 1990 1995 2000 2005 2010
year (station date)
1 .0---------------------------------------------1980 1985 1990 1995 2000 2005 2010
year (station date)
Figure 5. Tem perature-profile data from Palm er D eep from 1982 to 2010. (a) Locations of hydrographic profiles indicated by
blue dots. (b) Profiles of in situ water tem perature versus depth in Palm er D eep. Colours (see scale bar) indicate tim e of data
collection, (c) In situ w ater tem perature over the 8 0 0 -1 4 0 0 m depth interval versus tim e in Palm er Deep. T he linear regression
equation at the top indicates a w arm ing trend of approxim ately 0 .0 1°C yr- 1 . (d ) In situ w ater tem perature over the 4 0 0 -6 0 0 m
depth interval in Palm er D eep versus time.
Table 2. M ean size and epifaunal species diversity for drop stones 2 0 0 -1 0 0 0 cm 2 in plan area at depths > 9 5 0 m (N..
yaldwyni and crab traces abundant) and < 8 5 0 m (N.. yaldwyni and crab traces absent) in Palm er Deep. ‘C lean’ drop stones
had no identifiable epifauna. s.e., standard error.
depth zone (m)
> 950 (9 5 0 -1 4 1 9 )
< 8 5 0 (6 0 0 -8 5 0 )
no. of
stones
33
17
m ean stone plan
area (cm -2 )
478
493
s.e.
total no. of
species in zone
35 8
40 14
th e co n tin en tal shelf in the region; figure 1) exhibit su b ­
stantial seasonal an d in te r-a n n u a l variability, b u t ap p ear
to have risen from approxim ately 1.04 to 1.36°C since
1992 (i.e. at a rate o f approxim ately 0 .0 1 -0 .0 2 ° C y r-1 ;
figure 5).
4. D ISCU SSIO N
O u r R O V survey indicates th a t large, ad u lt lithodids,
m o st p ro b ab ly N . yaldw yni, w ere a b u n d a n t over large
areas o f th e P alm er D ee p seafloor at d ep th s o f 9 5 0 1419 m in F e b ru a ry 2010. T h e p o p u la tio n density o f N.
Proc. R. Soc. B (2012)
m ean species density
(per 500 cm -2 )
0.488
2.09
s.e.
proportion of
clean drop stones
0.142 0.64
0.505 0.18
yaldw yni estim ated for P alm er D eep (10 600 k m -2 , 95%
C L o f 8 0 3 0 - 1 3 900 k m -2 ) is g reater th a n th e m e an d en ­
sity (5100 k m - 2 ) o f th e com m ercially im p o rta n t red king
crab Paralithodes camtschaticus in an A laskan fjord, an d
sim ilar to th e ab u n d a n ce o f th e lith o d id Paralomis formosa
b ein g co n sid ered for co m m ercial exp lo itatio n at d ep th s o f
7 1 9 - 1 5 1 8 m on th e S o u th G eorgia slope (m ean =
8313 k m - 2 ) [21]. T h e b io m ass o f N . ya ld w yn i in P alm er
D eep is also po ten tially large; if we apply th e lith o d id
b ody-w eight m o d el o f C ollins et al. [21] to ad u lt N . ya ld ­
wyni, each N. ya ld w yn i has a w eight o f approxim ately
500 g, yielding an estim ated crab b iom ass in P alm er
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
1024
C . R. S m ith et al.
K ing crabs crossing the Antarctic S h elf
D ee p o f approxim ately 5000 kg k m - 2 . O u r R O V tran se ct
only crossed a p o rtio n o f th e P alm er D ee p seafloor d eep er
th a n 950 m , b u t king crabs an d crab traces w ere bro ad ly
d istrib u ted along this approxim ately 2 km tran sect; if we
assum e th a t th e estim ated N . yaldw yni ab u n d a n ce was
representative o f th e entire P alm er D ee p m ore th a n
9 50 m in d ep th , the p o p u la tio n size o f N . yaldw yni in
P alm er D ee p exceeded 1.55 x IO6 individuals an d the
b io m ass w as substantial (approx. 800 to n n es).
Som e lithodids are know n to aggregate d u rin g certain
life stages [21,22]. Is it conceivable th a t th e high densities
o f N . yaldw yni rec o rd e d along the R O V tran se ct reflect a
rec en t lith o d id aggregation in th e process o f dispersing?
T h e b ro a d d istrib u tio n o f lithodids an d a b u n d a n t crab
traces th ro u g h o u t o u r approxim ately 2 km RO V tran se ct
below 950 m , as w ell as th e ab u n d a n ce o f traces in
w idely spaced b o tto m p h o to g rap h s from 1998, are n o t
co n sisten t w ith an aggregation in the process o f dispersing.
T h u s, we th in k it very p ro b ab le th a t lithodids are bro ad ly
d isp ersed across the floor o f P alm er D eep below 950 m.
T h ere is evidence th a t N. yaldw yni exerts substantial
im pacts o n bioturbation, ichnofabric, and the diversity
an d taxonom ic structure o f m egabenthos in P alm er D eep.
W alking an d feeding activities o f lithodids in P alm er
D eep created a high abundance o f traces and a gashed/
lum py sedim ent texture covering the seafloor at depths of
m ore th a n 950 m (figures 2 a an d 3; see electronic sup­
p lem entary m aterial, videos S I and S2). T h is gashed/
lum py texture closely resem bles th e sedim ent texture of
‘digging pits an d tracks’ created by invasive red king crabs
{Paralithodes camtschaticus) in the B arents Sea in the
Arctic [23]. We cannot directly calculate a b io turb atio n
rate from the activities o f N. yaldw yni in P alm er
D eep, b u t th e sedim entation rate of approxim ately
0 .1 - 0 .4 cm y r-1 [12] indicates th a t the centim etre-scale
lithodid traces should b e obscured by sedim ent accum u­
lation (i.e. b uried b en e ath 0 .5 - 1 .0 cm o f sedim ent) in
less th an 10 years. In addition, the fresh appearance an d
overlap o f m any lithodid traces (figures 2a and 3; see elec­
tronic supplem entary m aterial, videos S I and S2) indicate
th a t crab retracking is causing trace turnover at m u ch
sh o rter tim escales th a n sedim ent accum ulation [24].
T h u s, the heavy cover o f lithodid traces at depths o f m ore
th a n 950 m (figure 3) suggests th a t the particle-displacem e n t an d p redatory activities o f king crabs are influencing
m u c h o f the sedim ent to depths o f 1 - 4 cm over tim escales
m u c h less th a n 10 years. O th er studies have show n th a t the
walking, resting an d foraging activities o f epifaunal crabs
(including lithodids) can m ix surface sedim ents to depths
o f several centim etres [25], disru p t infaunal tubes an d b u r­
rows (i.e. alter ichnofabric), and change rates an d pattern s
o f bio-irrigation by rem oving infauna perform ing im p o rta n t
ecosystem functions [23,26]. T h u s, N. yaldw yni is likely
to be acting as a m ajor ecosystem engineer [26,27] by
substantially m odifying the biotu rb atio n regim e, sedim ent
stratigraphy and sedim ent geochem istry (e.g. the location
o f labile organic m a tte r decom position an d th e d ep th of
th e redox potential discontinuity o f P alm er D eep sedi­
m en ts), an d by rem oving habitat-form ing echinoderm s
(see below). T hese sedim entological alterations could
prove useful for reconstructing the history o f king crabs in
P alm er D eep from sedim ent stratigraphy.
T h e diversity o f epibenthic m egafauna, b o th on sedi­
m e n ts an d on drop stones, was substantially low er in
Proc. R. Soc. B (2012)
th e lith o d id d ep th zone (m ore th a n 950 m ) th a n at
d ep th s lacking lith o d id s in P alm er D ee p (less th a n
850 m ). In p articu la r, ech in o d erm s, w h ich are co m m o n
p rey item s for lith o d id s [1,26], w ere ab sen t on soft sedi­
m en ts a n d d ro p stones w ith in th e lith o d id d e p th zone,
yielding su b stan tial declines in biodiversity. M eg afau n a
on sed im en ts an d o n d ro p stones w ere d o m in a te d by tu b icolous ce rian th id an em o n es a n d sponges, respectively;
b o th th ese taxa ap p e ar to b e lightly p reyed u p o n by lith o ­
dids in o th e r regions [26]. T h e diversity a n d taxonom ic
stru ctu re o f m eg ab en th o s in th e P alm er D ee p lith o d id
zone w as also m arkedly different from sim ilar d ep th s on
th e n earb y o p en slope lacking crabs; in ph o to su rv ey s at
8 0 0 - 1 0 0 0 m d ep th s off K in g G eorge Islan d , P iep en b u rg
et al. [28] fo u n d n o lithodids, relatively h ig h m egafaunal
diversity ( 1 7 - 2 4 species versus 8 in o u r P alm er D eep
data) an d a b u n d a n t ech in o d erm s ( 3 .8 - 6 .0 m -2 ). T h ese
d ata are highly co n sisten t w ith th e h ypothesis th a t p re d a ­
tio n b y th e d u rap h ag o u s p re d a to r N . yaldw yni red u ces th e
diversity a n d alters th e co m m u n ity stru ctu re o f A n tarctic
m eg ab en th o s below d ep th s o f 950 m in P alm er D eep , as
p red ic ted for lith o d id invasions o f A n tarctic co m m u n ities
[1,2,4]. Sim ilar su b stan tial red u ctio n s in epifaunal diver­
sity have b ee n associated w ith king crab invasions in
A rctic fjords [23,26]. N o n eth eless, b ecau se o u r results
on th e im p acts o f N . ya ld w yn i on b en th ic diversity in
P alm er D eep are correlative, th ey n ee d to b e confirm ed
by ad d itio n al studies (e.g. g u t-c o n te n t an d b io m ark er
analyses, an d crab-exclusion experim ents).
B ased on o u r RO V survey, N . yaldw yni an d its traces
were ab sen t from d ep th s o f less th a n 850 m in P alm er
D eep. Extensive p h o to g rap h ic an d traw l surveys from
2 0 0 8 - 2 0 1 0 at n in eteen stations at dep th s o f 4 3 0 - 7 2 5 m
along th e W AP shelf a n d in fjord basins from 63° to 68° S
also failed to d etect lithodids o r th eir traces (figure 1; see
electronic su p p lem en tary m aterial, table S 2). T h e total
area surveyed an d traw led (approx. 238 000 m 2) at the
n in eteen stations is approxim ately 40-fold g reater th a n
th e area surveyed by th e R O V in P alm er D eep, providing
strong evidence th a t N. yaldw yni or o th er lithodids did
n o t o ccu r at d ep th s shallow er th a n 725 m o n m u d d y h ab i­
tats on the W AP shelf from 2008 to 2010. Similarly, T h atje
et al. [4] failed to find lithodids on th e co n tin en tal shelf at
dep th s o f 5 0 0 - 6 0 0 m in M arg u erite Bay in 2007, an d
extensive video an d traw l surveys o f th e WAP shelf w est
o f P alm er D ee p at dep th s o f 5 0 0 - 6 0 0 m failed to d etect
lithodids betw een 1999 an d 2001 [11,20]. We conclude
th a t lith o d id crabs, in clu d in g N. yaldw yni, were ab sen t or
extrem ely rare o n th e W AP shelf at d ep th s o f less th a n
725 m at least u p to 2010. I t is n o tew o rth y th a t m egab en th ic ab u n d an ce an d diversity in these lithodid-free,
soft-sedim ent W AP co m m u n ities at 4 3 0 - 7 2 5 m d ep th s
were d o m in ated by echinoderm s, w hich are ab sen t from
th e lith o d id zone in P alm er D eep ([20]; C . R. S. S m ith,
L. J. G range & C. C lark 2 0 0 8 -2 0 1 0 , u n p u b lish ed data).
T h e a b u n d a n ce , large b o d y sizes, ap p a ren t isolation
an d extensive b io tu rb a tio n effects o f N . yaldw yni in
P alm er D eep suggest th a t this p o p u la tio n h a d b ee n estab ­
lished for years. T h e ab u n d a n ce o f a p p a re n t crab traces
at th e P alm er D eep floor in 1998, 12 years p rio r to o u r
RO V survey, is co n sisten t w ith this conclusion. It seem s
unlikely th a t N . yaldw yni in P alm er D ee p is m a in tain ed
by co n tin u o u s m ig ratio n o f w alking life stages across th e
shelf b ecau se o f th e ap p a ren t absence o f b en th ic life
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
K in g crabs crossing the Antarctic S h elf
stages from th e shallow er d ep th s su rro u n d in g P alm er
D eep over the p e rio d 1 9 9 9 -2 0 1 0 (see electronic su p ­
p le m en tary m aterial, table S2) [20]. Similarly, th e p o o r
dispersal abilities o f dem ersal, lecithotrophic lith o d id
larvae [29,30] suggest th a t co n tin u o u s tra n sp o rt of
N . yaldw yni larvae over distances m o re th a n 120 km
across th e shelf is unlikely. T h e recovery o f an ovigerous
N . yaldw yni fem ale from P alm er D eep indicates th a t this
lith o d id p o p u la tio n is capable o f rep ro d u ctio n ; th u s, th e
p o p u la tio n m ay be self-sustaining. A lternatively, th e
hig h ab u n d a n ce o f king crabs in P alm er D eep co uld
have resu lted from one o r a few large re c ru itm e n t events
m e d iated by b en th ic im m igration, or, p erh ap s m o re p ro b ­
ably, by p u lse d larval tra n sp o rt from th e co n tin en tal slope
across th e shelf in U C D W [14]. D em o g rap h ic, rep ro d u c­
tio n an d genetic studies o f th e P alm er D eep N . yaldw yni
p o p u la tio n are n e e d ed to resolve its o cc u p atio n history.
T h e ab u n d a n ce o f N . yaldw yni in P alm er D eep below
950 m , a n d its ap p a ren t absence at d ep th s o f less th a n
850 m an d w ater te m p eratu re s below 1.4°C , has in te rest­
ing im plications for th e invasion o f WAP shelf w aters by
this species. I f th e d istrib u tio n o f lithodids in A n tarctic
w aters is lim ited by a low er te m p eratu re thresh o ld , as
argued by a n u m b e r o f au th o rs [ 1 - 4 ] , o u r results suggest
th a t the P alm er D ee p p o p u la tio n o f N . yaldw yni ca n n o t
disperse an d /o r rep ro d u ce in te m p eratu re s co ld er th a n
approxim ately 1.4°C . In 1982, P alm er D eep b o tto m
te m p eratu re s w ere significantly colder th a n 1.4°C , at
approxim ately 1.25°C (figure 5), an d w ere likely to have
b ee n w arm in g for som e decades [5]; th u s a low er te m p ­
erature lim it o f approxim ately 1.4°C w ould suggest th a t
N . yaldw yni invaded P alm er D ee p w ithin th e last app ro x i­
m ately 50 years. A low er lim it o f 1.4°C is significantly
w arm e r th a n th e 0.5 °C w aters from w hich individual
ad u lt N . yaldw yni have b ee n re p o rte d from th e Ross Sea
[2,19], so different p o p u latio n s o f N . yaldw yni m ay have
different te m p eratu re tolerances. A lternatively, o th e r
u n k n o w n environm ental factors m ig h t exclude N . ya ld ­
w yni from d ep th s o f less th a n 850 m in P alm er D eep;
how ever, fish p re d a tio n on crab larvae an d juveniles
[26] is unlikely to be im p o rta n t since dem ersal fishes
are, if anything, m ore a b u n d a n t in th e lith o d id zone of
P alm er D eep th a n ju st above. I f te m p eratu re s o f less
th a n 1.4°C are in d e ed lim iting for th e P alm er D eep p o p u ­
lation, N . yaldw yni m ay n o t long b e excluded from th e
W AP shelf. C o n sid erin g th e rate at w hich P alm er D eep
w aters are w arm in g (approx. O .O L C y r-1 ) an d th e fact
th a t W AP shelf w aters (i.e. at 4 0 0 - 6 0 0 m depths)
aro u n d P alm er D eep now have te m p eratu re m inim a
aro u n d 1.35°C (figure 5; see electronic su p p le m e n tary
m aterial, figure S 2), th e P alm er D eep N . yaldw yni co uld
m ove u p o n to th e W AP shelf w ithin 1 - 2 decades. T h u s,
th e p o p u la tio n size, d e p th d istrib u tio n an d ecological
im pacts o f the P alm er D ee p N . yaldw yni p o p u la tio n
m e rit careful study in th e n ea r futu re to elucidate th e
rates an d co nsequences o f WAP shelf invasion by d u rap h agous lith o d id p red ato rs. T h e P alm er D eep lith o d id
p o p u la tio n is likely to serve as an im p o rta n t m o d el for
th e p o te n tia l invasive im pacts o f crush in g p red ato rs,
especially king crabs, on A ntarctic shelf ecosystem s du e
to clim ate w arm ing.
We are very grateful to the extraordinary support at sea
provided by R PSC , the captain and crew of the R V IB
Proc. R. Soc. B (2012)
C . R. S m ith et al.
1025
N. B. Palmer, and our colleagues in the LARISSA Project
during cruise N B P10-01. D ata collection occurred as parts
of the LARISSA and FO O D B A N C S2 Projects, and was
generously supported by N S F -O P P grant
num bers
0732711 and 0636806 to C.R.S, 0732467
to E.D .,
0732450 to C. L. Van Dover and 074735 to B.H. T he
ROV was funded by Im pulsfinanciering of the Special
Research F u n d of G hent University, and participation of
the ROV team (Dries Boone, K atrien H eirm an and Lieven
N audts) was co-financed by the BELSPO project
H O LA N T. We thank the referees A. Clarke, H . Ducklow
and J. Pearse for substantially improving the m anuscript.
This is publication num ber 8479 from SOEST, University
of Hawaii at M anoa and num ber 7492 from L am ontD oherty E arth Observatory.
REFERENCES
1 Aronson, R. B., T hatje, S., Clarke, A., Peck, L. S., Blake,
D. B., Wilga, C. D. & Seibel, B. A. 2007 Clim ate change
and invasibility of the A ntarctic benthos. A nnu. Rev. Eco!.
Evei. Syst. 38 , 1 2 9 -1 5 4 . (doi:10.1146/annurev.ecolsys.
38.091206.095525)
2 Haii, S. & T hatje, S. 2010 Tem perature-driven biogeo­
graphy of the deep-sea family Lithodidae (Crustacea:
Decapoda: Anom ura) in the Southern Ocean. Polar
Biol. 34 , 3 6 3 -3 7 0 . (doi:10.1007/s00300-010-0890-0)
3 T hatje, S. & Lörz, N . 2005 First record of lithodid crabs
from A ntarctic waters off the Balleny Islands. Polar Biol.
28 , 3 3 4 -3 3 7 . (doi:10.1007/s00300-004-0686-l)
4 T hatje, S., H aii, S., H auton, C ., H eld, C. & Tyler, P.
2008 E ncounter of lithodid crab Paralomis birsteini on
the continental slope off Antarctica, sam pled by ROV.
Polar Biol. 31 , 1 1 4 3 -1 1 4 8 . (doi:10.1007/s00300-0080457-5)
5 M eredith, M . P. & King, J. C. 2005 Rapid climate change
in the ocean west of the A ntarctic Peninsula during the
second h alf of the 20th century. Geophys. Res. Lett. 32,
L19604. (doi: 1 0 .1029/2005G L024042)
6 Clarke, A., M urphy, E. J., M eredith, M . P., King, J. C.,
Peck, L. S., Barnes, D. K. A. & Sm ith, R. C. 2007
Climate change and the m arine ecosystem of the w estern
A ntarctic Peninsula. Phil. Trans. R. Soc. B 362 , 1 4 9 -1 6 6 .
(doi: 10.1098/rstb.2006.1958)
7 Ducklow, H . W., Baker, K ., M artinson, D. G ., Quetin,
L. B., Ross, R. M ., Sm ith, R. C., Stam m erjohn, S. E.,
Vernet, M . & Fraser, W. 2007 M arine pelagic ecosystems:
the West Antarctic Peninsula. Phil. Trans. R. Soc. B 362 ,
6 7 -9 4 . (doi: 1 0 .1098/rstb.2006.1955)
8 T hatje, S. & A rntz, W. 2004 A ntarctic reptant decapods:
more than a myth? Polar Biol. 27, 1 9 5 -2 0 1 . (doi:10.
1007/s00300-003-0583-z)
9 García Raso, J. E., M anjón-C abeza, M . E., Ram os, A. &
Olaso, I. 2005 N ew record of Lithodidae (Crustacea:
Decapoda: Anom ura) from the Antarctic (Bellingshausen
Sea). Polar Biol. 28 , 6 4 2 -6 4 6 . (doi:10.1007/s00300-0050722-9)
10 Domack, E., Amblas, D., Gilbert, R., Brachfeld, S.,
Camerlenghi, A., Rebesco, M ., Canals, M . & Urgeles, R.
2006 Subglacial morphology and glacial evolution of the
Palmer D eep outlet system, Antarctic Peninsula.
Geomorphology 75, 125-142.
11 Sm ith, C. R., M incks, S. & D eM aster, D. J. 2008 T he
FO O D B A N C S project: introduction and sinking fluxes
of organic carbon, chlorophyll-a and phytodetritus on
the w estern Antarctic Peninsula continental shelf. Deep
Sea Res. (77 Top. Stud. Oceanogr.) 55, 2 4 0 4 -2 4 1 4 .
(doi:10.1016/j.dsr2.2008.06.001)
12 Kirby, M . E ., Dom ack, E. W. & M cC lennen, C. E. 1998
M agnetic stratigraphy and sedimentology of H olocene
Downloaded from rspb.royalsocietypublishing.org on May 26, 2014
1026
13
14
15
16
17
18
19
20
C . R. S m ith et al.
K ing crabs crossing the Antarctic S h elf
glacial m arine deposits in the Palm er D eep, Bellingshau­
sen Sea, Antarctica: implications for climate change?
Mar. Geo!. 152, 2 4 7 -2 5 9 . (doi:10.1016/S0025-3227
(98)80043-1)
Sm ith, D. A., H ofm ann, E. E., Klinck, J. M . & Lascara,
C. M . 1999 H ydrography and circulation of the West
A ntarctic Peninsula C ontinental Shelf. Deep Sea Res.
(7 Oceanogr. Res. Pap.) 46 , 9 2 5 -9 4 9 . (doi:10.1016/
S0967-0637 (98)00103-4)
Klinck, J. M ., H ofm ann, E. E., Beardsley, R. C.,
Salihoglu, B. & How ard, S. 2004 W ater-mass properties
and circulation on the west A ntarctic Peninsula C onti­
nental Shelf in Austral fall and w inter 2001. Deep Sea
Res. (77 Top. Stud. Oceanogr.) 51, 1925-1 9 4 6 . (doi:10.
1016/j.dsr2.2004.08.001)
Shevenell, A. E. & K ennett, J. P. 2002 A ntarctic H olocene climate change: a benthic foraminiferal stable
isotope record from Palm er D eep. Paleoceanography 17,
8000. (doi: 10.1029/2000PA000596)
Buckland, S. T., A nderson, D. R., B urnham , K. P.,
Laake, J. L ., Borchers, D. L. & T hom as, L. 2001 Intro­
duction to distance sampling: estimating abundance of
biological populations. Oxford, UK: Oxford University
Press.
T hom as, L., Buckland, S. T., Rexstad, E. A., Laake,
J. L., Strindberg, S., Hedley, S. L., Bishop, J. R. B.,
M arques, T. A. & B urnham , K. P 2010 D i s t a n c e soft­
ware: design and analysis of distance sampling surveys
for estimating population size. J. Appl. Eco!. 47 , 5 -1 4 .
(doi:10.1111/j.l365-2664.2009.01737.x)
Ahyong, S. T. & Daw son, E. W. 2006 Lithodidae from
the Ross Sea, A ntarctica, with descriptions of two new
species (Crustacea: D ecapoda: A nom ura). Zootaxa
1303 , 4 5 -6 8 .
Ahyong, S. T. 2010 T he m arine fauna of N ew Zealand:
king crabs of N ew Zealand, Australia and the Ross Sea
(Crustacea: D ecapoda: Lithodidae). NIW A Biodiv.
Memoir 123, 3 -1 9 4 .
Sum ida, P. Y. G ., B ernardino, A. E , Stedall, V. P.,
Glover, A. G. & Sm ith, C. R. 2008 Tem poral changes
in benthic megafaunal abundance and com position
across the West Antarctic Peninsula shelf: results from
video surveys. Deep Sea Res. (77 Top. Stud. Oceanogr.)
55, 2 4 6 5 -2 4 7 7 . (doi:10.1016/j.dsr2.2008.06.006)
Proc. R. Soc. B (2012)
21 Collins, M . A., Yau, C ., Guilfoyle, E , Bagley, P.,
Everson, I., Priede, I. G. & Agnew, D. 2002 Assessment
of stone crab (Lithodidae) density on the South Georgia
slope using baited video cameras. IC E S J. Mar. Sei. 59 ,
3 7 0 -3 7 9 . (doi: 10.1006/jm sc.2001.1167)
22 Powell, G. C. & N ickerson, R. B. 1965 Reproduction of
king crabs, Paralithodes camtschaticus (Tilesius). J. Fish.
Res. Board Can. 22 , 1 0 1 -1 1 1 . (doi:10.1139Æ65-009)
23 Oug, E ., Cochrane, S., Sundet, J., N orling, K. & Nilsson,
H . 2010 Effects of the invasive red king crab Paralithodes
camtschaticus on soft-bottom fauna in Varangerfj orden,
n o rth ern Norway. Marine Biodiversity 41 , 4 6 7 -4 7 9 .
(doi: 10.1007/s 12526-010-0068-6)
24 W heatcroft, R. A., Sm ith, C. R. & Jum ars, P. A. 1989
Dynamics of surficial trace assemblages in the deep sea.
Deep Sea Res. (7 Oceanogr. Res. Pap.) 36 , 7 1 -9 1 .
25 Solan, M ., W igham, B. D., H udson, I. R., K ennedy, R.,
Coulon, C. H ., N orling, K ., N ilsson, H . C. & Rosenberg,
R. 2004 In situ quantification of bioturbation using time
lapse fluorescent sedim ent profile imaging (f SPI), lum i­
nophore tracers and m odel simulation. Mar. Eco!. Prog.
Ser. 271 , 1 -1 2 . (doi:10.3354/m eps271001)
26 Falk-Petersen, J., Renaud, P. & Anisimova, N . 2011
Establishm ent and ecosystem effects of the alien invasive
red king crab (Paralithodes camtschaticus) in the Barents
Sea— a review. IC E S J. M a r Sei. 68, 4 7 9 -4 8 8 . (doi:10.
1093/icesjms/fsql92)
27 Jones, C. G ., Law ton, J. H . & Shachak, M . 1994 O rgan­
isms as ecosystem engineers. Oikos 69 , 3 7 3 -3 8 6 . (doi:10.
2307/3545850)
28 Piepenburg, D., Schm id, M . & G erdes, D. 2002 T he
benthos off K ing George Island (South Shetland Islands,
A ntarctica): further evidence for a lack of a latitudinal
biomass cline in the Southern Ocean. Polar Biol. 25,
1 4 6 -1 5 8 .
29 Thatje, S., Calcagno, J. A., Lovrich, G. A., Sartoris, F J. &
Anger, K. 2003 Extended hatching periods in the subantarctic lithodid crabs Lithodes santolla and Paralomis
granulosa (Crustacea: Decapoda: Lithodidae). Helgol.
Mar. Res. SI, 110-113. (doi:10.1007/sl0152-003-0145-3)
30 T hatje, S., Anger, K ., Calcagno, J. A., Lovrich, G. A.,
Pörtner, H .-O . & Arntz, W. E. 2005 Challenging
the cold: crabs reconquer the Antarctic. Ecology 86 ,
6 1 9 -6 2 5 . (doi: 10.1890/04-0620)