ISSN 0141 0814. I11ternati01lalJo1lrnal of the Society for U1Iderwater Tech1l010gy, Val 26, No 1, pp 3.12, 2004underwatcrTECHNOLOGY

Lander techniques for deep-ocean biological research

PM BAGLEY, IG PRIEDE, AD ]AMIESON, DM BAlLEY, EJV BATTLE, C HENRIQUESandKM KEMPOeeanla4 Universiry cf Aberdeen, Nelllbllrgh, Aberdeen.shire, Seotland, UK

Abstract

Photographic landers have proved to be a useful tao Ifor deep-ocean biological research. This paper pres-ents a calculation to define the niche landers havewhen compared to wire gears used on research ves-sels, and then describes how landers have developedto enable experiments previously only possible in shal-low water or in the laboratory to be transported to thedeep ocean at depths to 6000m. A number of diversecase studies using the authors' landers are describedto illustrate these developments. These studies includefish tracking experiments using the AUDOS (AberdeenUniversity Deep Ocean Submersible) Lander to studydeep-ocean fish movements. Two experimentsdesigned to investigate deep-ocean fish physiologyare presented; the first investigated muscle perform-ance of fish in-situ using an electrical stimulator aboardthe 'Sprint' Lander; and the second describes a landercapable of trapping and measuring the in-situ oxygenconsumption of deep-ocean fish. A further case studyto investigate bioluminescence in both the water col-umn and on the sea floor to abyssal depths using theISIT (Intensified Silicon Intensified Tube) lander isdescribed. Finally, to investigate the effects of time sig-nals on deep-ocean animals a case study using theDOBO (Deep Ocean Benthic Observer) lander isdetailed.

1. Introduction

The oceanic and atmospheric sciences have been revo-lutionised in recent decades by the use of satellites inEarth orbit as instrument platforms that continuouslyscan large areas of the Earth surface. In particular,marine sciences have benefited greatly from the abilityto visualise oceanic fronts, gyres and areas of primaryproduction.1 For scientific studies beneath the surface,however, in-situ instruments are the only way to makeobservations. Whilst low frequency echosounders andsonars can map the sea floor at abyssal depths, high-res-olution studies require a towed body elose to the seafloor. For any other measurement or observation (Phys-ical, chemieal, biologieal) on the deep-ocean floor thepresence of a man-made instrument is required.

In deep-oeean biology, trawl sampling has greatlyenhanced our understanding.2 However, animalsbrought to the surface are almost inevitably killed due

to pressure and temperature change, henee our knowl-edge of deep-oeean life, until the advent of in-situ tech-niques, was restricted to the study of dead animals.Modern in-situ techniques such as manned submersiblesor RemoteIy Operated VehicIes (ROVs) have furthcrextended the boundaries of knowledge; however therequirement for the constant presence of a surface ves-seI precIudes them from long-term studies.3

Autonomous Underwater VehicIes (AUVs) showgreat promise for the future of in-situ marine rescarch.1However, sampling !imitations due to !imited dynamicpayload capability, and finite mission duration restrietstheir use for biological research. Particularly in deep-sea research, landers are perceived as an importantmeans of obtaining data that could not be achieved inany other way. They can be left to work autonomouslyfor periods of hours to years and, unlike manned sub-mersibles, ROVs, and other wire gears, thcy can oper-ate whilst the ship is back in port. Indeed, it is possiblefor a single ship to service a fleet of landers operatingover a \vide area of the sea floor.5 Landers also providca means for small research vesseIswithout winchcs bigenough to carry cable capable of rcaching thc abyssalsea floor to recover information from the deep-sea.

2. Lander technology

Lander definitionThe term 'lander' refers to a subsea mooring thatdescends to the sea floor, carries out aseries of tasksautonomously for periods of days to years storing dataon board, and then ascends to the surface at the end ofthe mission for recovery by a surface ship. A typicallan-der comprises a chassis, buoyancy module, recoverymodule and ballast with release mechanism. Dependingon the size of the lander one or more scientific payloadscan be attached for monitoring processes in the surfacelayers of the sediment and the bottom water.

The lander is usually ballasted to be negativclybuoyant so that it descends to the sea floor at a rate ofO.5-1.0mJs-1,which gives a relatively soft landing so asto reduce sediment disturbance. At the end of the mis-sion ballast weights are shed either by time release oracoustic command from the surface. The lander thenascends because of positive buoyancy to be recovercdby the surface ship. A recovery module is usuallyequipped with radio beacons, strobe lights and visualmarkers to aid location. Landers have evolved from

3

Bagley et al. Lander technigues

Fig 1: Comparison of bottom time (useful working time)per hour of ship time using landers and wire gear for anexperiment of 10h duration. Analysis assumes the shiphas other useful work to do during lander bottom time

wire is always less than the ship time. However, landersachieve about twice as much work as conventional wiregears. 1'his analysis assurnes that more than one landeris available 01'the ship has other useful work to do in thetime between deployment and recovery. It is notablethat although landers are associated with deep-searesearch, they are most advantageous in shallow waters.However, the increased probability in shelf areas of lan-ders being damaged by shipping 01'fishing activity hasto be offset against this apparent advantage.

1'0 determine at what point a lander becomes moreefIicient than a wire gear, this analysis can be extendedto compare landers and wire gears for experiments thathave different bottom time durations. If:

60002000 3000 4000 5000Depth (m)

1000o

o

0.5

t:::W

2.5~Q)

E 2:;::00.:c(f) 1.5QQ)

E:;::0

Ego.Deso~a:

where Tt = ship transit timeTd = gear descent timeTb = gear bottom timeTa = gear ascent time.

where, Tc = time rcquircd to cstablish acousticcantact and activate release

Tl' = lander recovery time after surfacing.

conventional oceanographic instrument moorings suchas current meter 01' sediment trap arrays but are eharac-terised by an instrument package standing or suspendedelose to the sea floor.

However, for lander work the ship does not have toremain on station after deployment (no bottom time,Tb)' but there is additional time required to locate thelander and time required to recover it from the surface.1'hcrcfore, for lander work:Shiptimerequired 1's = 1't+1'd +1'c+Ta +1'1'

Lander niche as a tool for marine researchIf experiments on the sea floor have durations ofmonths 01' years it is immediately obvious that anautonom aus platform should be plaeed on the sea floor.However, experiments of a few hours duration might becarried out either by suspending an instrument on a ver-tieal cable (eg, CTD, ROV) 01' by autonomous lander.1<01' conventional wire work the total ship time (Ts)required 1S:

Ts T t + Td + Tb + Ta

1'able I illustrates a simple example to compare wireand lander work for an experiment lasting lOh on thesca floor.

Fig I shows the results of this analysis giving theratio of bottom time over ship time for depths from100m to 6000m. Obviously, the bottom time using the

Ship time requiredfar wiregear (T,+Td+Tb+T) Wiregear

R=------=Ship time required (T,+Td+Tc+T,,+TJ Landerfor lander

WireWork landerWork

Transit time (T,) 4h Transit time (T,) 4hDescent rate (Td) 3600 meh" Descent rate (Td) 2400meh"8ottom time (T J 10h Bottom time (T J (10h)Ascent rate (T) 3600 meh" Search time (Tc) 0.5h

Ascent rate (Ta) 2400 meh"Recovery time (T,) 0.5h

Table 1: Parameters for a typical 10h experiment with40 nmiles steaming time between stations

Fig 2: Ratio of ship time achieved by landers comparedto wire gear for experiments of different durations but

otherwise characterised by data in Table 2. Valuesgreater than R=1 indicate that landers use ship time

more efficiently than wire gears

Fig 2 shows the results of this analysis, giving a set ofcurves for R for experiments of bottom time durationfrom 0.5h to lOh. At 2h duration the performance ofthe two systems are approximately equal (R= I). At 4hduration the lander system has advantages, although

a: 30

~ 2.5Q)

E 2:;::00.:.c(f)

G5 1.5uc~'ca:IQ)OJ 0.5[l?

~ 00 1000 2000 3000 4000 5000 6000

Depth (m)

4

und r at rTECHNOLOGY Vo126, No 1,2004

Acronym Name Institute Measurement

MAPI/II Module Autonome Pluridisciplinaire IFREMER,France Benthic boundary laver flowRAP 11 Respirometre IFREMER,France Benthic chamber sediment oxygen consumptionFFR Free- Fall Respirometer GEOMAR, Germany Benthic chamber sediment oxygen consumptionGoteborg Goteborg chamber and profile lander Goteborg, Sweden Benthic chamber sediment oxygen consumption,

solute fluxesIMBC Compact respirometer lander IMBC, Crete Benthic chamber sediment oxygen consumptionAUDOS Aberdeen University Aberdeen, Scotland 20 Acoustic tracking of fish ingesting

Oeep Ocean Submersible transponder tagsATIIS 1 Acoustic Telemetry and Transponder Aberdeen, Scotland 10 Acoustic tracking of fish ingesting

Interrogation System #1 transponder tagsATIIS2 Acoustic Telemetry and Transponder Aberdeen, Scotland 10 Acoustic tracking of fish ingesting

Interrogation System #2 transponder tagsATIIS3 Acoustic Telemetry and Transponder Aberdeen, Scotland 10 Acoustic tracking of fish ingesting

Interrogation System #3 transponder tagsATIIS4 Acoustic Telemetry and Transponder Aberdeen, Scotland 10 Acoustic tracking of fish ingesting

Interrogation System #4 transponder tagsLAFF 1/11 Large Abyssal Food Fall Aberdeen, Scotland Large carcass consumption rates on the sea floorISIT Intensified Silicon Intensified Aberdeen, Scotland Bioluminescence monitoring of the benthic

Tube Lander boundary layerFish Trap Baited cage Aberdeen, Scotland Capture of deep-sea fishesSPI Sediment Profile Imagery Galway, Ireland Sediment profiling imagery (wire gear)CTO Conductivity, Temperature, Oepth SOC, England Conductivity, Temperature, Oepth (wire gear)Multicore Multicore SOC, England Sediment cores (wire gear)

Table 2: Landers and other gears operated during the ALiPOR cruise D222 (leg A)

these have to be traded off against the requirement fora second lander (so the ship has other work to performwhile the first lande I' is operating on the sea floor), andthe possible risk of loss of the lander. For experimentshaving a bottom time lasting longer than 8h,autonomous landers emerge as the most efficient use ofship time (R~2).

Practical experience: The ALiPOR cruisesALIPOR (Autonomous Lander Instrument Packagesfor Oceanographic Research) was a EU MAST III(Marine Seienee and Teehnology) programme CT950010 coordinated by the University of Aberdeeninvolving nine partners in six EU countries. It is recog-nised that to carry out large-scale research in the deep-sea, the resources of single institutions and countriesare likely to be insufficient. A number of groupsthroughout Europe have developed deep-sea landertechnologies and it is important that these systemsshould be able to work together in joint programmesand also avoid mutual interference. The aim of theALIPOR programme was to create a compatibleEuropean fleet of lander vehicles and to gain experi-ence of operation of fleets of landers from one 01' moreships.

As part of the ALIPOR programme, research cruis-es were undertaken at locations around the PorcupineSeabight (40030'N; 15000'W) and Porcupine AbyssalPlain (48050'N; l6°30'W) areas of the north-east

Atlantic Gcean. During these cruises landers from eachinstitution were operated in cooperation to incrcasc thcknowledge of deep-sea ocean processes. Tablc 2 sllm-maries the type of landers and other gears uscd duringthe ALIPOR cruise D222 (leg A).6

These cruises were scheduled to maximise thcamount of bottom experimcntation time within thcavailable weather and logistics windows. This wasachieved by deploying multiple landers so that experi-mentation was in progress on thc sea floor even if thcship was hove to because of bad weather. The ship usedfor both emises was the RRS Discovery, a conventionalship with bow thruster, but no dynamic positioning.

The ship was in the working area for just over 26days (634.7h). She steamed betwcen three main work-ing areas (Porcupine Abyssal Plain 48°50'N; 16°30'W,OMEX site 49°II'N; 12°49'N, BIOTRANS 47°40'N;19°50'W). During this time 40 lander deployment andrecovery operations were carried out which is equiva-\ent to one lander operation every 7.9h. Fig 3 shows thedeployment time for each lander used on the cruise.There were also three wire gears (SPI, CTD, and mul-ticorer) that are shown at the bottom of the figure. Formost of the cruise at least four landers were deployedsimultaneously, and for periods, all 11 landers wereoperating on the sea floor simultaneously. The totalamount of subsea time achieved during the course ofthe cruise was 3804.6h, 01' almost six times (an efficien-cy of 599%) the ship time expended.

5

Bagley et al. Lander techniques

ALiPOR RRS Oiscovery Cruise 222 Leg A (1996)July August

27 2~ 2930 3) 01 02 03 O~Op06 07 08 O~ 1011 12 1~ 1~ 15161! 1~ 19202) 222324 2p 26 27

Ifremer MAP

Ifremer RAP

Geomar 1

Goteburg

Aberdeen LAFF

Aberdeen AUDOS

Aberdeen Fishtrap

Aberdeen AniS 4

Aberdeen AniS 3

Aberdeen AniS 2

Aberdeen AniS 1

SPI

CTD

Multicorer

Cruise duration 634.7h; Totallander deployment duration 3804,6h; Efficiency 599%

Fig 3: Time bar chart for the ALiPOR research cruise 222A (1996). Bar lines represent lander and wire geardeployment durations over the course of the cruise. Wire gears were SPI, CTD, and Multicore; all other gearswere landers

6

Onee the lander had surfaeed, the average timerequired to loeate it, man oeuvre the ship alongside andsueeessfully grapple it was 22 min. Bringing it onboardand securing it so that the ship eould get underway took14 min. Thus the average total recovery time was 36min onee the lander surfaced.

3. In-situ lander deep-ocean biologicalresearch: Case studies

A photographic lander is perhaps the simplest form ofdeep-ocean lander and eontinues to prove of greatvalue in studying the remote deep-sea environment.' Toshow how the lander technique ean be used to transportexperiments, previously only available to terrestrial andshallow-water biologists, to the deep ocean, a numberof case studies are described illustrating diverse experi-ments carried out by the authors.

The AUDOS fish tracking landerOne of the great challenges of oceanography is to

understand how agents such as current, waves andmarine organisms transport matter across the sea floor.The AUDOS (Aberdeen University Deep OceanSubmersible) lander was designed to address the small-scale transport of matter by marine organisms by track-ing the movements of deep-sea fish at abyssal depths.Fishes ean playa role in the transport of organie mat-

ter by virtue of consumption and dispersal of foodmaterial across the sea floor.

Miniature aeoustie and radiolocation tags have beenused for many years for the study of fish behaviour.Tags (pingers) attached internally or externally transmita short aeoustic pulse at a pre-defined repetition rate.Mitson & Storton-Wese enhanced this technique, traek-ing plaice (Pleuronectes platessa) using a 300kHz transpon-der tag. The tag, when interrogated by a shipbornesidescan sonar, appeared as a bright spot on a sonar dis-play. However, only one transponder eould be traekedat any time, and the whole resourees of the surfaee shipwere used to follow one fish.

Using the AUDOS fish traeking lander as the dcliv-ery system Baglei'10 developed a Code ActivatedTransponder (CAT) that required a unique interroga-tion code before a deployed CAT would respond. TheCAT was designed to operate at depths to 6000m froma sonar deployed on the AUDOS lander operating onthe sea floor.11 Potentially 32 fish eould be traekedsimultaneously.

Once on the sea floor fish were attracted to AUDOSby bait attached below the vehicle, in view of a camera.Some of the bait was used to make up small bait pareelseontaining the CAT, whieh the fish ingested. As the fish(with CAT tags in their stomaehs) moved away fromAUDOS the sonar aboard AUDOS traeked theirmovements to a nominal radius of 800m. The CATs

und r at rTECHNOLOGY• __ ••• • Vo126, No I, 2004

AUDOS interrogation JUl~ _CAT response

AUDO S hydrophone 1

AUDOS hydrophone 2 _

AUDOS hydrophone 3 _

Fig 4: The AUDOS fish trackinglander operation. Differences inperiods T1, T2, &T3 areexaggerated to show how the CATresponse is detected at slightlydifferent times at each of theAUDOS hydrophones

1151::'

CAT

AUDO~) t rro(p!.OI'

Transponder 440

800

Fig 5: Typical Code Activated Transponder track of ademersal fish (Coryphaenoides armatus) at 4800m

depth in the Porcupine Abyssal Plain, north-eastAtlantic Ocean. The AUDOS lander with onboard sonar

is located at the origin

400

Range (m)-200

EQ)OlCC1l([

-200The 'Sprint' landerIn-situ studies of whole fish and in vitro experiments withmetabolie enzymes suggest that the metabolic rates ofdeep-sea fish are lower than those of related shallow-water speeies at similar temperatures. Measuring fast-start performance provides a non-invasive but quantita-tive measure of ma.ximum aetivity eapaeity and mayindicate the relative importance of burst performanceto animals.'2 The 'Sprint' lander (Fig 6) was used todeliver an anima! stimulation experiment to the deep-sea floor. Anima!s were attraeted to bait sueh that theywere in view of a high-speed eamera, and between twoeleetrodes situated Im apart. At timed intervals a short

were miniature (67mm long, 16mm wide) acoustictransponders that only responded to a unique dual77kHz pulse position interrogation code generated bythe AUDOS sonar. If the time separation between thetwo pulses was the same as the time spacing pre-pro-grammed into the GAT during manufacture, the GATresponded with its own Sms duration 77kHz pulse. TheGAT response was detected back at AUDOS by threeequally spaced hydrophones. The arrival time of theGAT response at each of the hydrophones dependedupan the distance between the GAT and eachhydrophone. Therefore, by accurately timing the periodfrom interrogation to the detection of a response fromthe GAT at each hydrophone (TI, '1'2, & '1'3), the loca-tion (in two dimensions) of the GAT eould be ealculat-ed by triangulation (Fig 4).

Fig 5 shows a typical track of a deep-ocean demer-sal fish (Coryphaenoides armatus) at 4800m in thePoreupine Abyssa! Plain area of the north-east AtlantieGeean.

7

ßagley et al. Lander teclmiques

'"

(al

Baito

•• Luminescent Particle'" Impact Screen

Camera

The 181TlanderSunlight is an important faetor in the morphology

and behaviour of animals in the top 1000m of theoeean. Howevel~deeper than 1000m no sunlight pene-trates and yet the majority of animals including fishesstill have operational eyes and well developed visual cen-tres in the brain.17 The only source of light at suchdepths is of biological origin. Many marine animals arecapable of producing light 01' maintain a population ofluminescent bacteria.IB The eyes of deep-sea animalsare elearly advantageous in detecting such biologicallight sources either from the same 01' different species.Biolumineseence has been extensively investigated in thesurface layers of the oeean but litde is knowll about thefrequellcy and function of bioluminescence at depthsgreater than 1000m, either in mid-water (bathypclagie)01' on the deep-sea floor at bathyal and abyssal depths.

Experiments were designed to investigate thesourees of light likely to be seen by deep-sea fishes andother animals whose eyes are often adapted to detecttargets of very low intensity below the threshold ofhuman vision. An ISIT (Intensified Silicon IntensifiedTarget) camera (Konsberg-Simrad) provided a suitablcdetector with sensitivity corresponding approximatelyto that of many deep-sea animal eyes. :Forbatllypelagicstudies the ISIT camera was fixed downward facingand foeused on a mesh screen at 0.5m distanee (Fig 7a).

Fig 6: The 'Sprint' Lander, (1) Acoustic releases, (2)High speed camera, (3) Controller, (4) Electricstimulator, (5) Battery, (6) Electrode array

Fig 7: Modes of operation of the ISIT Lander; (a)Bathypelagic: ISIT lander free-falls through the water

column recording bioluminescent occurrences; (b) Seafloor: ISIT lander records spontaneous bioluminescent

events occurring at bait

pulse eleetrie field generated between the two eleetrodesstimulated animals at the bait to perform a fast start,escape response. Off-line analysis of the high-speedvideo on reeovery of the lander enabled museIe per-formance to be measured.

Although, electric fishing and clectrical stimulationof hsh museIe "vasweil founded in terrestrial studiesl3,ioI

litde was known of the levels of electrical stimulationreguired to stimulate deep-ocean animal fast startresponse. Although estimates from previous studiesl5,16

were made, flexibility in levels of elcetrical stimulationwas required. Therefore, an electric stimulator wasdcsigned that was capable of generating a finite rangeof voltages (16,21,26,31,36,40,47, and 57V), at aprogrammablc variable pulse width. Typically, an elee-trode array immersed in seawater can result in a resist-anee of approximately 0.1Q. Therefore, the electricalstimulator must be capablc of generating instantaneouscurrents in excess of 570A for the 57V amplitude set-ting. This was achieved using a bandgap referencedprogrammablc power supply trickle charging a capaci-tive array (0.198 F). Onee charged, under microproees-SOl' control the capacitors were discharged through theeleetrode array via a high power IGBT transistor stage.Pulse width was controlled under software by utilisingthe microprocessor's pulse width modulation output.

Suceessful fast start response was aehieved with the6sh Antimora l'Ostrala at 2500m in the Porcupine Seabightarea of the north-east Atlantic with a stimulation pulseof 40V for 2ms. A. rostrata with peak velocity of 0.7ms·1

and white museIe output power of 17.0VVkg-I has per-formance similar to shallow-water fish at the same tem-perature.12

8

und r at rTECHNOlOGY Vo1 26, No 1,2004

Fig 8: Examples of lumineseent reeording from the 181Tlander; (a) Bathypelagie lumineseent events generatedby impaeting the mesh sereen as the 181Tlander free-

falls through the water eolumn; (b) 8pontaneous seafloor lumineseent events observed moving aeross a bait

As the lander free-fell through the water eolumn anybiolumineseent partieles impaeting the sereen lumi-neseed in view of the eamera. This enabled a vertiealprofile of biolumineseent partieles to be measured fromthe surfaee to the abyssal sea floor.

:F'or sea floor experiments, the mesh sereen wasremoved and replaeed with bait situated at the foeallength of the downward looking ISIT eamera (Fig 7b).Onee on the sea floor, any spontaneous biolumineseentaetivity eaused by animal aetivity was observed.Reeording took plaee in time-Iapse segments spreadover the period of the deployment. Towards the end ofeach segment a LED and then a light were illuminatedfor calibration and identifieation purposes.

The ISIT lander "vas controlled by a 68000 basemieroproeessor (Onset Computer Corp) that tookinputs from an onboard real-time dock and eontrolprogram to sehedule events with respect to systempower-up time. Although the operating system softwarewas written in C, to enable ease of use and flexibility ofsystem operation a higher level ISIT lander instruetioncode (LIS) was ereated. The LIS program was in theform of a text file situated on removable flash eardmedia that the mieroproeessor would parse direetlyafter power up. Any illegal eommands would result inthe system not issuing an 'all deal" signal (10 flashes ofthe lamp) indieating to the user that the lander shouldnot be deployed until the problem had been resolved.Commands eould be seheduled to aresolution of Is.

Composite video from the ISIT eamera from time-lapse recording segments was fed to a miniature digitalvideo recorder (Sony Walkman GV-D300) eontainedwithin the controller housing aboard the lander. A eur-rent meter with onboard CTD instruments enabled

video reeordings to be taken against measurements ofthe loeal environmental eonditions.

Fig 8 shows typieal examples of lumineseent eventsoeeurring during free-fall (Fig 8a), and on the sea floorat bait (Fig 8b). The bathypelagie experiments havegiven insights into the vertieal proftle of biolumines-eenee in the water eolumn, and its ehanges betweenseasons. Biolumineseenee observed at bait on the seafloor has suggested the presenee of eommunieation notpreviously thought to exist on the deep-oeean floor.

The FRESP respirometry landerDeep-sea fishes are charaeterised by high water con-

tent,19,20low metabolie enzyme aetivitY'21small hearts/2

and other adaptations that are indicative of low meta-bolie rates. However, tllere is no eonsistent rdationshipbetween metabolism and depth for all animal groups(eg, 23,24,25). Conventionallaboratory studies of deep-seafish are not possible as these fish are typieally killed dur-ing recovery to the surfaee due to temperature ehangeand deeompression. The FRESP (Fish RESPirometry)lander was conceived to enable in-situ respirometry atdepths to 6000m.26

The FRESP lander eonsisted of a watertight trap inview of a downward looking digital video eamera.Initially, the trap was held above a base plate eontainingSkg of mackerel bait (Fig 9). The positioning of the trapallowed fish attraeted to the bait to be undisturbed byany struetures relating to the trap design. The behav-iour of deep-sea llsh was sufIieiently understood toallow the trap to be released by time release under theeommand of a eentral eontroller. Onee rcleased, thetrap fell under its own weight to seal to a base platetrapping any fish still at the bait.

9

10

Fig 9: The FRESP Respirometry Lander; (a) Internallyrecording digital video camera; (b) acoustic releases; (c)

Lights; (d) respirometry chamber in its raised position;(e) repirometry chamber base plate (bait located in the

centre of the plate); (n oxygen sensor and pump; (g)controller and data Jogging; (h) battery

During the trapping process the digital video cam-era was recording to assess the behaviour of the fish andtrapping system. Once the fish were trapped, animpeller eonstantly circulated the water within thechamber. At periodic intervals, the oxygen level of theehamber water was measured with a Beckman-typepolarographic oxygen electrode (Sea Bird ElectronicsInc SBE 23B). Water was drawn from the chamber atapproximately 50ml" by a pump (Sea Bird ElectronicsInc SBE 51') across the sensor and returned to thechamber. A customised, pressure compensated three-way valve alternated flow across the sensor between thechamber water for measurement and external ambientwater for sensor calibration. Sampies were taken at 40min intervals and consisted of 3 min of ambient watersampling to allow repolarisation of the electrode mem-brane, 3 min sampling from the chamber and a further3min of ambient water sampling. Typically, the experi-ment duration was three days.

A central controller, similar in design to the ISITlander system, controlled alllander sub-systems exceptthe acoustic releases. Each subsystem could be inde-pendently operated at specified times and durationsthrough a text-based program that was configured bytl1euser prior to deployment. Data from oxygen meas-urements were stored on-board the controller's flashcard memory, which was removed and downloaded onrecovery of the lander.

The background oxygen consumption due to thebait was measured during control experiments. TheFRESP lander was deployed with the respirometerchamber closed and ol\.")'genlevels recorded, ie, no fishin the chamber. The intent was to subtract these levelsfi'om any measurements witl1 fish in the chamber.However these levels were found to be very low andwithin the noise level of the measurements.

Data from trapping tl1e fish CoryPhaenoides armatus at4000m showed a low metabolie rate (3.6 ± 0.2 mLh·1 kg·l)in tl1isspecies when compared witl1other gadid species.26

During use in tl1e deep ocean otl1er species such asamphipods and bacteria were caught in the trap alongwith the target species. However, analysis showed tl1atoxygcn consumption of tl1e bi-catch species was witlllnthc range of tl1ebackground noise.

The lang-term 0080 landerThe deep sea has historically been perceived as a stableenvironment receiving a constant rain of organic mat-

Bagley et al. u11lder techniques

tel' from the surface layers above.27 The discovery of aseasonal pattern to this input has altered this percep-tion.28 Cyclical periodicity in reproductive strategies,and seasonal patterns of growth and activity among thcdeep-sea fauna further indicate that these animals areresponsive to, perhaps dependent on, time signals intheir environment/9

,30,31 yet the temporal and spatialvariability of these influences are still poorly under-stood. The DOBO lander (Deep Ocean BenthicObserver) was designed to monitor deep-ocean physicalparameters and observe the biological response.

The DOBO lander was designed to operate forperiods exceeding six months. In addition, it wasequipped with an ADCP (RDI, USA) and a currentmeter (FSI, USA) for background current measure-ments and a large capacity 35mm stills camera toobserve animals attracted to bait. Initially, the DOBOwas deployed with a large food fall (porpoise carcass).However, data are difficult to analyse against the back-ground of steadily deteriorating bait. Thercfore, toinvestigate changes in the frequency, behaviour, sizeand number of fish visitors to the bait over the winterand spring months, a periodic bait release system wasdeveloped. This consisted of an amino acid solutionthat was designed to mimic fish bait. A proportion ofthe solution was pumped into a sponge in-view of thccamcra and under the control of a very low powercontroller. The sponge allowed the amino acid solu-tion to slowly generate an odour plume rather than bewashed away in the ambient current. The very lowpower controller was a 68000 base microprocessor(Onset Computer Corp, USA) with controlling soft-ware that was able to shut down non-essential micro-pro cesSOl'systems and to slow down the system dockto conserve power during sleep periods between obser-vations. The low power control software was transpar-ent to the user, who used a text-based language similarto that described above to bias photography aroundbait releases.

und rwat r;TECHNOLOGV Val 26, No I, 2004

4.0 Conclusion

Landers provide a cost effective and efficient means oftransporting experiments to the deep-ocean floor. Forexperiments lasting 8h or longer landers can also beshown to provide a more efficient use of available shiptime than wire-deployed systems, providing the ship hasother useful work to do. However, there is always agreater risk of loss than with a wire gear. Fairly sophis-ticated experiments can be carried out using landers,but robust manufacture and design are of the utmostimportance, particularly with long-term landers whereinstrument failure or controller software crashes wouldnot be discovered until after recovery.

With the advent of sophisticated oceanographictools such as deepwater ROVs, manned submersibles,and AUVs, a lander retains a niche in sea floor experi-mentation where:• measurements 01' observations are required oververy long periods. In particular with the recent interestin cabled networks where sensor technology is still notmature enough for semi-permanent deployments of 10years 01' more/2 a lander system attached to a cablednetwork could provide a useful transport tool that couldbe connected to the sea floor network yet still be recov-ered to the surface for sensor servicing• investigating phenomena that occur infrequently orsporadically. Here it wou1dbe difficult to justify fundingfor long duration ROV or AUV support, but a landerused off a vessel with other useful work to do wouldprovide a cost effective solution. An example of thistype of experiment is the investigation of spontaneousbioluminescent events with tlle ISIT lander• use on ships of opportunity is possible where thoseships do not have suitable warp for deep-ocean deploy-ment or the DP and other systems necessary for ROVand submersiblc operations• rclativcly undistributed observation or sampling isrequired (ROV and manned submersible thrusters candisturb sea floor sediment and communities).

References

1. Priede, IG. (1983). Use qf satellites in marine biology.Experimental Biology at Sea. Eds A MacDonaldand IG Priede (New York: Academic) pp 3-50.

2. Merrett, NR and Haedrich. (1997). Deep-sea demersalfish andfishes. Chapman and Hall, London.

3. Gilchrist, I, MacDonald, AG and Priede, IG. (1983).Diver~ submersibles and unmanned vehicles. ExperimentalBiology at Sea, Eds AG Macdonald and JG Priede.Academic Press, London.

4. Doolittle, D. (2003). The payriff is in the payload: usingAUVs in scientific research. Undenvater Magazine,Doyle Publishing Co. March/April 2003.

5. Priede, IG. (1999). Autonomous lander instrument plat-

.Jorms.Jor oceanographic research (ALIPOR). Final report.EU Mast 3. Contract No MAS3-CT950010.

6. Priede, JG. (1996). RRS Discovery Cruise Report 222(leg A). 27 July-26 August 1996.

7. Bett, BJ. (2003). Time-lapse photography in the deep sea.Undenvater Technology 25 No 3, pp 12]-127.

8. Mitson, RB and Storeton-West. (1971). A transpond-ing acoustic fish tag. Radio & Electronic Engineer 41483-489.

9. Bagley, PM. (1992). A code activated transponder Jor indi-vidual identijication and tracking qf deep-seafislz. WildlifeTelemetry, Eds Priede and Swift. E]lis Honvard,Chichester, Eng]and.

10.Bagley, PM, Bradley, S, Priede, IG and Gray, P.(1999). Measurement qf fish movements at depths to 6000musing a deep-ocean lander incorporating. a short base-finesonar utilising miniature code-activated transponder teelmolo-g)!. Meas. Sci Technol 10 1214-1221.

II.Bagley, PM and Priede, IG. (1997). An autonomousJree-Jall acoustic tracking .rystem.Jor investigation qf flShbehaviour at abyssal depths. Aquat. Living. Resour. 1067-74.

12.Bailey, DM, Baglcy, PM,Jamicson, AJ, Collins, MAand Priede, IG. (2003). ln-situ investigation qf burstswimming and muscle perfOrmance in the deep-sea flShAntimora rostrata. (Gunther, 1878). Journal ofExperimental Marine Biology and Ecology 285-286, 295-31 ] .

13.Cowx, IG. (Ed) (1990). Developments in electricfishing.Fishing News Books, Blackwell ScientificPublications. Oxford. England.

14.Cowx, JG and Lamarque, P. (1990). Fishing with elec-tricity: Applications in Jreshwater fisheries management.Fishing News Books, Blackwell ScientificPublications. Oxford. England.

15.]ames, RS and Johnston, JA. (1998). Scaling qf muscleperfOrmance during escape responses in the fishMyoxocephalus scorpius L. Journal ofExperimental Biology 20 I 913-923.

16.]ohnson, TP andJohnston, IA. (1991). Power output qffish musclefibres perJorming oscillatory work: iffects qf acuteand seasonal temperature change. Journal ofExperimental Biology 157409-423.

17.Priede, IG, Williams, LM, Wagner, HJ, Thom, A,Brierley, I, Collins, MA, Collin, Sp' Merrett, NR andYau, C. (1999). lmplications qf the visual .rystem in regu-lation qf activity cycles in the absence qf solar light: 2_{25I]iodomelatonin binding sites and melatonin receptor geneexpression in the brains qf demersal deep-sea gadiform flShes.Proceedings of the Royal Society. Series B 266:2295-2302.

18.Herring, PJ. (1990). Bioluminescent communication in thesea. Light & Life in the Sea (Ed PJ Hcrring ct al), pp245-264. Cambridge.

19.Childress,]J and Nygaard, MH. (1973). The chemicalcomposition qf midwater fishes as a Junction qf depth qf

11

12

occurrence rdf southern CalijOrnia. Deep-Sea Research20, 1093-1109.

20. Childress, lJ, Price, MH, Favuzzi, J and Cowles, D.(1990). Chemical- composition r1 midwater fishes as ajUnc-tion r1 depth occurrence rdf the Hawaiian-islands - foodavailability as a selective ftctor. Marine Biology 105,235-246.

21. Childress, lJ and Somero, GN. (1979). Depth-relateden;;;yme activities in muscle, brain, and heart r1 deep-livingpelagic marine teleosts. Marine Biology 52, 272-283.

22. Greer- vValker, M, Santer, RM, Benjamin, M andNorman, D. (1985). Heart structure r1 some deep-seafish(Teleostei: Macrouridae). Journal of Zoology 205,75-89.

23. Childress, lJ and Thuesen, EV (1993). E.fJects qfhydrostatie pressure on metabolie rates qf six speeies qf deep-sea gelatinous zooplankton. Limnology & Oceanography38, 665-670.

24. Cowles, DL, Childress, lJ and Wells, ME. (1991).At/etabolie rates qf midwater erustaceans as a jUnctionqf depth oeeurrenee rdf the Hawaiian islands - foodavailability as a selective factor. Marine Biology 110,75-83.

ßagley et al. Lander techniques

25. Seibel, BA and Childress,1J. (2000). Metabolism qfbenthic octopods (Cephalopoda) as a function qf habitatdepth and o>rygenconeentration. Deep-Sea Research I 47,1247-1260.

26.Bailey, DM,Jamieson, AJ, Bagley, PM, Collins, MAand Priede, IG. (2002). Measurement qf in-situ o>rygenconsumption qf deep-sea fish using an autonomous landervehicle. Deep-Sea Research I, 868, I-li.

27. Sanders, HL. (1968). Marine benthie diversity: a eompar-ative study. Am. Nat 102, 243-282.

28. Billet, DSM, Lampitt, RS, Rice, AL and Mantoura,RFC. (1983). Seasonal sedimentation qf phytoplankton tothe deep-sea benthos. Nature 302, 520-522.

29. Palmer,JD. (2002). The living dock: The orchcstra-tor of biological rhythms. Oxford University Press.

30. George, RY and Menzies, RJ. (1967). Indieation qfcyclie reproductive activiry in abyssal organisms, Nature,215, p 878.

31. Tyler, PA. (1988). Seasonaliry in the deep sea. Oceanogr.Mar Bio\. Annu. Rev. 26 pp 227-258.

32. National Research Council (2000). Illuminating thehidden planet: the future of seafloor observatory sci-ence. National Academic Press, USA.