Deep-Sea Research, 1976, Vol. 23, pp. 729 to 754. Pergamon Press. Printed in Great Britain.

Interactions between zooplankton and phytoplankton profiles in the eastern tropicalPacific Ocean

ALAN R. LONGHURST*

(Received 4 August 1975; ill revised form 29 January 1976; accepted 5 February 1976)

Abstract-Data from 86 detailed zooplankton profiles taken during the EASTROPAC cruises of1967 to 1968 have enabled a first-order description to be made of zooplankton distribution in theupper 1000 m of thewater column in relation to density, light, oxygen, and phytoplankton. A layerofabundant epiplankton contains a subsurface maximum that tends to coincide with the bottom ofthe mixed layer and with the depth of maximum carbon fixation, but lies above thechlorophyll-amaximum. Zooplankton abundance declines sharply downwards across the pycnocline, forming adiscontinuity between the epiplankton and the low-biomass plankton below. Coincident with thedeep sonic scattering layers are diurnal layers of migrant interzonal species which rise at night intothe epiplankton. Major regional differences are caused by the shoaling of the pycnocline at theequatorial divergence, and the very deep pycnocline of the southern gyral region. An hypothesis,based upon theEASTROPAC data, isstated: that theform of phytoplankton profiles, in periods orregions of stable production, may be primarily determined by a depth-differential in herbivoregrazing pressure, ratherthan bydifferential cell-sinking rates as is suggested in classical productionmodels.

INTRODUCTION plankton profiles been obtained of comparableZOOPLANKTON abundance in low latitudes appears resolution to profiles of phytoplankton and itsto be greatest in a relatively shallow layer which related variables, which can be obtained by pumpsis separated from water containing much less with much less expenditure of ship-time; com­zooplankton at some depth between 100 and monly, only about five levels are sampled down300 m. VINOGRADOV (1968) reports many zoo- to about 500 m with opening-closing nets (e.g.plankton profiles for low Pacific latitudes in VINOGRADOV, 1968). When higher resolution haswhich plankton abundance decreases by about an been obtained, as with the new series of RMTlorder of magnitude between the surface and and RMT8 nets (BAKER, CLARKE and HARRIS,1000 m to a very low biomass, which is itself 1973), several hours are required to obtain a singlereduced by another order of magnitude only profile, during which period vertical migrationbetween 3000 and 5000 m. It is also evident from may to some extent occur even if twilight isthese profiles that biomass in the upper kilometre avoided, and it is usually impossible to work manyof the water column contains a discontinuity at such stations during one cruise or survey.some depth above 500 m, at which most of the For these reasons, the number of stations fordecrease occurs. Such a discontinuity may occur which we have zooplankton profiles of highelsewhere in the oceans, though its depth varies resolution is relatively smalI, and although theconsiderably. (e.g. FOXTON, 1956; GRICE and details of plant and animal profiles must, in fact,HULSEMANN, 1965). be causally matched by the processes of grazing

It is predictable that zooplankton profiles and regeneration of nutrient salts, we have notshould contain a degree of complexity to match advanced much farther in our interpretation ofwhat is well known to occur in pumped phyto- these relations than the models of RILEY, STOM~IELplankton profiles; however, because of the restric- and BUMPUS (1949) and of STEELE and YENTSCHtions imposed by opening-closing nets for zoo- *Institute for Marine Environmental Research, 67plankton collection, only very rarely have 200- Citadel Road. Plymouth PLI 3DH, England.

729

730 ALAN R. LoNGHURST

(1960) in which grazing was considered to be auniform factor throughout the euphotic zone­which it clearly is not.

The EASTROPAC expeditions in 1967 to1968were designed to monitor seasonal change ina large part of the eastern tropical Pacific Ocean,between about 200 N and 200S; cruise and stationnumbers used in the present paper are the sameas presented in the EASTROPAC Atlases (LOVE,1971). Zooplankton hauls with multiple serialplankton samples (LONGHURST, REITH, SEIBERTand BOWER, 1966), or Longhurst-Hardy PlanktonRecorders (LHPR's), as they have come to becalled, were included in standard EASTROPACstation procedures together with profiles of plantpigments, carbon fixation, inorganic nutrients,oxygen, transparency and physical variables; allstation and data processing procedures aredescribed by LOVE (1971). These data enable us toexamine, over a very large area of the ocean andover all seasons, the relationships between profilesof zooplankton and other variables in the upperkilometre of the water column.

For the purposes of this investigation it wasassumed that net-caught zooplankton were pre­dominantly herbivores (p. 749); it was furtherassumed, except in some special cases discussedseparately, that carnivores would aggregate, ingeneral, at similar depths jo their herbivore prey.

DATA ACQUISITIONThe zooplankton profiles were taken with

0.5 m2 LHPR equipment, fitted with 0.3 mmnylon filtering gauze. Successful profiles wereobtained at 86 stations (Fig. 1) during four cruisesof the David Starr Jordan, as follows:

EASTROPAC cruise12-February-March 1967-32 stations3Q-June-July 1967-20 stations6Q-December-January 1967-68-21 stations76-March 1968-13 stations.

All stations were worked within about onehour of midday and midnight in order to sampleonly when diel migrants were at their most stablein respect of depth levels, and to avoid the duskand dawn periods of relatively very rapid verticalmovement. Even though, as ROE (1974) has

recently confirmed, some vertical shifts occurduring the periods of full daylight and darkness,midday and midnight levels effectively representthe depth at which diel migrants spend most oftheir time.

The standard LHPR haul comprised a singleoblique profile to the surface; the gear was sunkas rapidly as possible, usually in little more than5 min, until it was judged that it had reached500 m, at which depth it was held for 2 min tostabilize itself. It was then hauled slowly andobliquely to the surface with about 50 samplingperiods, each of 30-s duration, giving an averagesampling depth interval of 9.3 m and a totalhauling time of about 30 min. There were 73 suchstandard stations on cruises 12, 30 and 60, whichwere found subsequently to have varied from 200to 700 m in sampling depth, though half fellwithin the range 450 to 500 m. In addition, oncruise 76 there were 10 special shallow haulsdesigned to give greater resolution within themixed layer; these had an average sampling depthinterval of only 2.5 rn between the surface anddepths of 150to 300 m. Finally, on the same cruisethree special deep hauls from 1000 m wereworked successfully.

The total number of individual planktonsamples from these 86 stations was 3886, eachrepresenting about 10 m3 filtered; the actualamount filtered and the sampling depth intervalswere recorded for each sample by the instrument,and used to normalize the data. The LHPRequipment had no bridles in front of the netaperture and was coloured black: these measuresappear to have so reduced net-avoidance by zoo­plankters that no consistent day/night ratio couldbe detected in numbers of organisms per cubicmetre, and the problems discussed by BRINTON(1967) were apparently not important.

The plankton nets and recorder were carefullyexamined on recovery; whenever serious hang-upof organisms on the nets was found to haveoccurred, the haul was rejected, as were hauls inwhich subsequent sample editing showed thatentrainment of specimens in the recorder hadoccurred. Rejected hauls were about 10% of allhauls taken, and are not listed here.

Interactions between zooplankton and phytoplankton profiles 731

•10

../I

o V

~o _

o•

o..- CI

o v-._~~-- ~o ___

o -- _

.'~Q

o

Jl

o

o

o

- --- - _.~-;;;-----.__. _-'

--' I /1

o90

0 - y

, -CRUISE 30 t> 50

1..-- ....,.-- - - --.-,- - - - -,.- - ---,.-'1 0·

FEBR ARY - APA:l

v O f 'CRU;SE 12 t. 76, I

1()(1° 9::/

Fig. 1. Location of LHPR profilesand allocation of stations to oceanographic regions I to VII; open circles- middaystations, closed circles-midnight stations.

The samples were subsequently sorted into18 major categories (all coelenterates, all cope­pods, chaetognaths, molluscs, cladocera, etc.)and 18 minor categories including some genera(e.g. Euchaeta, Eucalanus, Candacia), some arti­ficial groups (e.g. large and small species ofPleuromammdy and some single species (e.g.Rhincalanus nasutus, R. cornutus, Euphausiadiomediae, E. gibboides). Twenty-one profilesfrom cruise 12 were sorted for six species ofClausocalanus as part of a different investigation:these data are included in those considered here.Biomass, expressed as dry weight in mg m-3, wascomputed from volumetric measurements made oneach sub-sample using factors experimentallyderived for 10 categories of zooplankton com­position as sampled by the auxiliary net fitted to

the LHPR (LONGHURST, REITH, SEIBERT andBOWER, 1966); each sub-sample was allocated tothe zooplankton composition category it mostclosely resembled, to determine the appropriateconversion factor.

THE EASTROPAC AREAThe EASTROPAC area comprises the eastern

tropical Pacific Ocean from 200 N to 20cS andfrom the American continent out to 120oW. Thestation lines on which LHPR profiles wereobtained run north-south between 98 and118cW, so as to form transects normal to thezonal current system (Fig. I). Rather than groupthe profiles within rectangles based on latitudeand longitude in the same way as BLACKBURN,LAURS, OWEN and ZEITSCIlEL (1970), in their

732 ALAN R. LONGHURST

studies of the multiple correlation of biotic andabiotic variables in the horizontal plane, I havechosen instead to identify a series ofoceanographicregions within the EASTROPAC area, each withconsistent and definable characteristics, and thezooplankton profiles have been pooled accord­ingly.

Examination of all available station data,mostly as presented in the EASTROPAC atlases,indicated that there were seven such regimescharacterized by the vertical variability ofproperties known to influence the vertical distribu­tion of zooplankton. The areal .extent of theseregimes and the allocation of zooplankton profilesto them is indicated in Fig. I ; they have the follow­ing characteristics:

Region I. A triangular area bounded by theMexican coast and by looN and l lO''W, ofvariablecurrents, the surface water originating in the northfrom January to May after which, with theseasonal development of the north equatorialcounter-current in Region III, the Costa RicaCurrent brings equatorial surface water insideCalifornia Current water which it entirely replacesin the area until August to October when asoutherly component from the California Currentagain invades the area. Surface temperatures arehigher here than in the rest of the EASTROPACarea, frequently exceeding 29°C from April toSeptember; mixed layer depths are alwaysmoderate so that the bottom of the isothermallayer, of Dl (WYRTKI, 1964), lies between 20 and50 m, the thermocline having a complex topo­graphy with much seasonal and areal variation.In the second half of the EASTROPAC year Dlwas shoaler (20 m) and mixed layer temperatureswarmer (28 to 29°C) than in February-March(50 m, 26 to 28°C), and northward coastal flowafter June was associated with replacement of athermocline ridge 300 to 400 miles off the coast bya trough, from which Dl shoaled towards thecoast. This region overlies the greatest oxygendeficiencies in the North Pacific, and values below0.25 ml 1-1 occur within the oxygen mininallayer.

Region II. To the west and southwest of thefirst area the extension of the California Current

as the north equatorial current dominates surfaceflow throughout the year; temperatures arelower, and vary more during the year than closerto the coast, from 21 to 26°C seasonally, with theeffects of the northern hemisphere winter and ofcoastal upwelling at Baja California. The oxygenminimum tends to have slightly higher minimalvalues (0.25 to 0.50 ml l") than in Region I,and the mixed layer tends to be deeper, so that Dlis typically greater than 50 m.

Region III. The north equatorial counter­current traverses the northern part of the studyarea between 5 and lOON except between Februaryand May, when it does not extend east of aboutl200W at the surface, though it is detectable in thethermocline topography farther east at this time.The occurrence of this flow from the west isindicated in the slope of the thermocline down­wards towards the south, from a zonal ridge at10 to l2°N along the northern boundary of thecurrent to a zonal trough centered at about 4°N,which indicates its southern boundary. Thus,Dl varies from 90 m in the south to only 20 m inthe north, and there is considerably greatervariation in its depth during anyone cruise thanthere is between seasonal means. Profiles wereobtained from the north equatorial counter­current, however, only in the second half of theyear and these showed D I to be much shoalerthan in Region I north of the thermocline ridge at10 to l2°N. The oxygen minimal layer includesvalues no lower than 0.5 ml l'<, and rangesfrom 300 to 500 m in thickness.

Region IV. About 2° north and south of theequator the water column is very stronglyinfluenced by seasonal equatorial Ekman diver­gence and by the shallow, eastward-flowingequatorial under-current; the latter lies close tothe surface, its core being scarcely deeper than50 m, and its latitudinal extent and relativevelocity indicated by the equatorial troughing ofisotherms in the lower part of the thermocline(12 to 20°C, Fig. 2) and by the location there of asalinity core (34.9 to 35.2%0)'

Equatorial divergence causes the thermoclineto approach the surface so closely that the mixedlayer may be extremely shoal, commonly less than

Interactions between zooplankton and phytoplankton profiles 733

E

".c-Q.C,l

o

SOUiH1J'

I

<O'~O@D------::..----==<

200 '. •

•!-IT ZOOPLA."JKTO '

~...:-:.~:..,. . . . . . .. . .

..::::::::::::::9"

50

1eo

200

Fig. 2. Sections along EASTROPAC transects at 1l0'W; the symbols along the 'baselines indicate station positions.Chlorophyll-a is contoured at seven levels, from> 0.4 to < 0.05 rng m-3 ; zooplankton at six levels, from> 2000 to

< 100 g (100 rn3) - l ,

734 ALAN R. LONGHURST

10 m deep. From May until about November thisprocess appears to surface the equatorial under­current (CROMWELL, MONTGOMERY and STROUP,1954; JONES, 1969) so that the permanent tropicalmixed layer is obliterated, probably assisted byturbulent diffusion (WILLIAMS and GIBSON, 1974),resulting in a deeper homogeneous layer (<2 40 m),though at considerably lower temperatures(<2 20°C) than earlier in the year. This situationwas established at some stations in June and July,1967 (cruise 30) and remained so throughout therest of the year; at all Region IV stations inDecember (cruise 60) the primary thermocline hadbeen eliminated and a cool mixed layer (22°C) ofrather variable depth was established. By Feb­ruary 1968 (cruise 76) the tropical mixed layerwas re-established, though it was everywhere veryshoal.

Consequent upon the general shoaling of theisotherms at the equator, the O2 minimum layer,which has minimum values greater than0.50 ml l-', lies very shoal.

Region V. Throughout the year, equatorialsurface water is transported westwards by thesouth equatorial current broadly over theEASTROPAC area south of about 9°N and it iswithin this westward zonal drift that Region IVoutlined above is embedded as a very narrowequatorial ribbon.

Region V is recognized as the region ofstrongest zonal flow of. the south equatorialcurrent, its northern extent limited by the southernboundary of the north equatorial counter-currentat the 4°N thermocline trough or by a broad zoneof convergence at 7 to 9°N when the northequatorial counter-current is absent. South of theequator, the permanent thermocline deepenssouthwards throughout the EASTROPAC region.An interruption in this slope by a region in whichsouthward shoaling of the summer thermoclineoccurs is associated with a weak, subsurface southequatorial counter-current (TSUCHIYA, 1974); thesouthern edge of this current, where the mixedlayer deepens again southwards, 'is taken as thelimit of Region V.

This region has a mixed layer with relativelystrong small-scale thermal layering at many

stations both north and south of the equatorialdivergence, this being especially true of those witha deeper D1. Variations in thermocline depthappear to contain no seasonal component and tobe due to the relative instability of the upper100 m here compared, in particular, with Region r.

Region VI. From about 7 to 90S, the velocityof the south equatorial current slackens, thesurface water mass changes its characteristics,and a summer thermocline occurs; this is a regimeof gentle westward drift, increasing mixed layerdepths and increasingly high values in the O2

minimum towards the south. A number of distri­butions of planktonic animals and bathypeIagicfish (AHLSTROM, 1971; MCGOWAN, 1971) appearto coincide with the boundary between RegionsV and VI.

Region VIT. In the extreme south of the area,from about 15 to 16°S, the south Pacific centralwater mass is evident in the water column, beingprincipally indicated by a salinity maximum whichprogressively strengths and deepens towards thesouth, though the separation from Region VI isnot very well defined. 01 increases progressivelysouthwards, as does the summer mixed layerabove it, and reaches 250 m at 20oS. Thermoclinegradients are slight and only exceptionally isthere no mixed layer temperature stratification,even above the summer thermocline. There arevery low standing crops ofall biota, and deepeningof features of the phytoplankton profile, especiallysouth of 14°S where the chlorophyll-a maximumoccurs deeper than 200 m at some stations;oxygen values lower than 1.0 ml I-I do notnormally occur in the water column.

GENERAL FEATURES OF THEZOOPLANKTON PROFILES

Zooplankton profiles for biomass, totalnumbers of organisms and for all sorted cate­gories were computer-generated for each station;examination of the general profiles for biomassand numbers immediately indicated a high degreeof commonality among their general features,especially when the profiles were pooled region­.ally. Copepods, being more numerous in allprofiles by factors varying between 7 and 10,

Interactions between zooplankton and phytoplankton profiles 735

together with chaetognaths and euphausiids,whose biomass is also very important, dominatedthe form of the profile for biomass and number oforganisms.

Three main features (Figs. 3 and 4) can beseen in most profiles and are defined below: anepiplankton, a planktostad, and discrete layerswithin the planktostad.

From the surface down to depths ranging from

50 to 150 m in all except nine of the 86 profiles ofbiomass and total numbers there is a layer ofabundant zooplankton having dry weight valuesranging from 1 to 1000 mg m-3, and withnumbers of all organisms frequently as high as900 m-3 and rarely below 100 m-3• It is convenientto apply the term epiplankton to this layer, follow­ing FOWLER (1898), VINOORADOV (1968) andothers.

TOC N X103

10 15 20 25 15 30

76120 ... DZNN

.. DZX E

C

... DPS100

'4 I;>SL

200

Ep

~

.c+-c.Q)

°30024 48

DZN76152 DZX

c E

.. DPS

100

P

Fig. 3. Two exemplary shallow daytime profiles (total organisms 100 m-S) to demonstrate their principal features;N = neuston, E = epiplankton, C = planktocline, P = planktostad, Remainder of notation as in text, p. 737.

~wO'l

J A AD RHIN L S EECOEL CHAE COPEP MYS AMPH EUPH EUPH FISH EUCA CORN PLEU PLEU EUCH CAND OIOM GI BB

Temperat ure Biomass Ind iv iduals5 10 15 20 25

1001 ./ I ... I I r r I I I I I I rlOO

9676 478 2293 8034 31 224 313

20 0 200

300 300

400 400

l l t'">

f'''jI E ~(J

.c SIi zo600 600 OJ

~C

ooj

I700 700•800

800

900900

10001000

Fig. 4. Special deep daytime profile (76202) at 14"N to show layering of taxa; in this computer-derived plot the horizontal scales are arbitrary ; against eachhistogram are indicated the maximum numbers per 100 m3 or maximal biomass in g (100 01 3) - 1 reached in that profile. Abbreviations are of major and minor

sorted categories, defined in text, p, 731.

Interactions between zooplankton and phytoplankton profiles 737

The lower limit of the epiplankton is usuallywell defined at the bottom of a depth zone acrosswhich plankton abundance usually decreases by100 to 1000 organisms m-3 per meter of depthincrease, or about 2 to 10% of the epiplanktonaverage value per meter and which can con­veniently be described as a planktoclinei" in dayprofiles a zooplankton minimum frequently liesat 10 to 25 m separating a layer of more abundantnear-surface zooplankton from the main body ofthe epiplankton below. A major faunistic bound­ary occurs at the bottom of the epiplankton, whichforms the lower limit for many species and generaof copepods, for all c1adocera, for most poly­chaetes, and for some decapods.

Below the planktocline, plankton biomass andnumbers are from one to several orders ofmagnitude lower (0.1 to 1.0 mg m? and ca. 10individuals m-3) and change very little withincreasing depth to at least 1000 m. This cantherefore be described as a planktostad,* whichcontinues downwards for several thousandmeters (e.g. VINOGRADOV, 1968: pp. 188 to 189),and within which rates of change are so small asto be unquantifiable in the present data.

Single interzonal species or groups of speciesoccur both within the epiplankton by night andthe planktostad by day, but are more prominentwithin the latter, where they form importantdiscrete layers; these may, or may not, coincidewith deep scattering layers (DSL), they extendthroughout the sampled depth range of theplanktostad, and they usually represent daytimepopulations of diel vertical migrants, or upperinterzonal species in the sense of VINOGRADOV(1968, p. 51).

The biomass of such depth-discrete layers mayapproach 5 mg m-3, and may be greater than theplanktostad background by at least an order ofmagnitude, and also they may be extremely thin(5 to 10 m is not unusual) in the vertical sense.Such layers have previously been reported inLHPR profiles in the California Current(LONGHURST, 1967), the North Sea (KINZER,1970), the North Pacific (BARRACLOUGH,LEBRASSEUR and KENNEDY, 1969) and the NorthAtlantic (Williams, personal communication)

and, from our general knowledge of the oceanicdistributions of DSLs, are probably a commonfeature in many parts of the ocean.

From examinations of the profiles of biomassand of total individuals, depths were assigned tothe major features of the profiles; in almostevery case the two profiles agreed but, when theydid not, the data derived from the biomass profilewere discarded:

DPS -depth of the bottom of the epiplankton,defined by the planktocline-planktostadboundary.

DZX-depth of the subsurface zooplanktonmaximum that may occur within theepiplankton.

DZN-depth of the subsurface zooplanktonminimum that frequently separates themain body of the epiplankton from anear-surface layer, having its ownmaximum abundance within 10 to15 m of the surface.

In addition, from the temperature record madeby LHPR for each profile and from associatedEASTROPAC station data, the depths of anumber of related environmental parameters weredetermined:

D 1 -the depth of the base of the mixedlayer, from STD casts;

D2 -the depth of the base of the thermo-cline (WYRTKI, 1964), also from STDcasts;

DE -the depth of the bottom of theeuphotic zone, at the 1% incidentlight level, determined from Secchidisc extinction depths by appropriateconversions (LOVE, 1971);

DCLX-the depth at which the subsurfacechlorophyll-a maximum occurs(yENTSCH, 1965), determined fromLexan bottle casts (LOVE, 1971);

DCX -the depth at which the maximal rateof uptake of 14-C was measured in

*The general suffixes-e-cline and stad-used in oceano­graphy to denote the presence or the absence, respectively,of a gradient (as in thermo-, halo-, pycno-, OX)'-, etc.) arehere used for the same purpose, as a matter of convenience,in relation to plankton.

738 ALAN R. LoNGHt1RST

simulated in situ primary productionexperiments (LOVE, 1971);

DPHX-the depth of the subsurface maximumof phaeopigment concentration, fromthe same bottle samples as used forchlorophyll-a determination.

Finally, though not used subsequently in theformal correlation programme, information wasextracted on:

DCLN-the depth of the bottom of the majorchlorophyll-a maximum where valuesfell to below a value of 0.05 mg m-3 ;

DON, DNOX-respectively, the depths of theoxygen minimal layer, and of theshallow N02 maximum, determinedfrom reversing bottle casts.

The 19 sets of profiles of specially sorted taxawere then examined and depths derived for majorlayers (or for maximum abundance in unlayeredtaxa) both within the planktostad and theepiplankton (Table 1). In the case of diel migranttaxa, separation was made between depths of themain migratory parts of the population and thedepths of the residual, non-migrating layers whichare a characteristic feature of such taxa (e.g.LONGHURST, 1967).

The data derived from the general profiles werestratified by time (day/night), season (cruise) andregionally, and an analysis of variance betweenthe strata was performed, as was a linear regressionin all unstratified permutations of two of the eightvariables (Table 2). A stratified table ofdifferences

Table 1. Mean depths of variables, stratified by region; XD-N is the mean difference between dayand night depths.

Zooplar.kto'l Ph"toplankton Er:.:Jil'On::':ent

R~Jio~1 Drs DZX DCX DCLX DeLli DPBX 01 D2 DE D~02

98.:: 39.0 25.0 49.5 107.0 83.9 37.8 119.7 65.6 70.4

II 82.5 41.3 30.0 70.0 105.0 95.0 47.5 120.0 84.3

111 89.4 31.1 29.3 49.4 112.0 75.0 30.9 93.6 86.1 64.2

IV 17.6 11.9 22.9 38.4 108.0 66.6 25.2 98.9 56.8 51.6

V 89.0 25.7 31.5 52.8 122.0 16.8 36.1 113.3 64.3 58.9

VI ao.o 60.7 22.1 58.6 130.0 96.4 51.6 195.0 56.3 111.4

VII 35.0 39.2 120.8 150.0 146.7 68.3 251.7 77.2

X 88.5 30.8 27.6 55.3 119.0 84.0 37.9 126.6 66.7 71.3('2.3) (0.2) ('1.3) (H.7l (±Z.8) (0.0) (t1.6) (t4.1) (tl.6) (tl.1)

X D-~ .7.7 +10.4 '3.5 -6.0 'Il.I .4.7

Table 2. Correlation coefficients (r) of linear regressions, all permutations of two of eightselected variables.

Dl 02 DE DCLX DCX DPHX Drs

01

D2 +.626***

DE •• 161 NS '.178 NS

DCLX +.555U1r .... 666**'" +.413u lr

DCX +.209· '.218' +.444··· +.299 U

DPHX '.115 NS +.314"'* ••057 NS +.262* -.0.5 NS

Drs +.'l38Ht ,.053 NS +.204 'NS -e, 342** +.2&6* -.130 NS..DZX +..396u * 1'.413U I\' •• 065 NS +.294u +.148 NS +.066 NS 1'.482 i11 U

NS-not significant (P > 0.05). ·_just significant (P O.Ol-t>.05), "-significant (P O.OOI-t>.OI).···-very significant (P < 0.(01).

Interactions between zooplankton and phytoplankton profiles 739

Drs - Dl +58 *** +47 .** +.56 *** +57 *it. +".9 ± 2.S ***- D2 -uu +20 .* -24 ** -29 .. -25.8 f. 4.0 *..- DE +26 ..* +21 Hi +16 ." +19 * +22.6 ± 2.5 .....

- DCU +47 u* +38 ••• +41 *.. +37"'· +40.0.!. 2.4 ...

- Del +69 .." +56 flU +63 *** +5,5 ....: ...61 ..1! 2.4 ...

- D1RX +18 .* +11 ." +7 ~S +7 ~S +11.1 ± 4.0·.

DZI - Dl +3 ES -20.** -6 ss -3 ~S -5.0 ± 3..3 ss

- D2 ·70 ••• ..88 **. -89 *** ..86 *** -88.1 t 4.5 *..- DE -29 *** -40 *.* -47 *.. -43·" -35.4 t 3.7 ....

- DCU ·9 * ·28 ••• -21 *.. ...26 *•• ..20.' t 3.5 ...

- llCX +18 ** -11 .. +2 ss ·7 ss +3.4 ± 3.1 ~S

-DI'lll -38 *.. -49 *** -55 *... -51·" -49.1 t 4.4 ***

Table 3. Depth differences (m) stratified by timeand region, between DZX, DPS and six environ­mental variables. Abbreviations as in text and Table2. Stated significance is that between depths of

indicated pairs of variables.

(Table 3) between DPS and DZX and the othersix variables was constructed and each differencesubjected to analysis of variance and ofsignificance.

These processed data were then used as thebasis of an examination of the relationshipbetween the variables and of possible reasons forthe observed form of the zooplankton profiles.

VERTICAL EXTENT OF EPIPLANKTON

Preliminary examination of the general pro­files showed that while the general featuresassociated with the epiplankton could berecognized at nearly every station, there wasgreat variability in detail, principally concerningthe depth at which features occurred; it was alsoevident that the variability was rather stronglylinked with the regional pooling of the stations.It is to be supposed that the planktocline, thedepth at which the vertical extent of the abundantepiplankton is limited, would be the result of thepresence of a constraint in the physical environ­ment or by the insufficiency of the food supplybelow such depth; these possibilities are nowexplored.

Density and temperatureIn tropical seas, temperature is a practical

indicator of density of sufficient precision forpresent purposes and so we may define thepycnocline by Dl and D2, or as coincident withthe thermocline.

Because of this relatively direct relationshipbetween density and temperature, the lower partof the pycnocline coincided with the same groupsof isotherms over the whole region; the datasuggest that DPS and the 14°C isotherm coinciderather frequently, though this is most unlikelyto be in any sense a depth-limiting factor in thedistribution of the epiplankton, and the possibilitywas not investigated formally.

Analysis of the depth differences DPS-Dl andDPS-D2 showed that DPS usually occurred in thelower part of the pycnocline, and there was nocase in which it occurred shoaler than Dl, orwithin the mixed layer; in only 16 of the 75 pro­files at which DPS could be identified did it occurdeeper than D2. The mean depth differences werevery significant in each case:

DPS = Dl + 54 ± 2.5 m (P < 0.(01)DPS = D2 - 26 ± 4.0 m (P < 0.(01).

Significance at this level was maintained forall strata (time, season and region), and the linearregression DPSjDl (Dl being the more satis­factorilydetermined parameter of the thermo­cline) showed significant covariance (r = 0.338;P 0,001 to 0.01) for all data strata.

An examination of the 16 exceptions showedthat in 11 stations at which the epiplanktonappeared to extend below the base of the thermo­cline (i.e, where DPS > D2), the difference wasby a margin of only 2 to 17 m, possibly to beaccounted for by the imprecision ofthe quantifica­tion of D2 from the temperature profiles.

As for the remaining exceptions, two stations(30215 and 60257) lay within the equatorialdivergence cell in Region IV to the west of theGalapagos Islands where the thermocline hadsurfaced and D2 was relatively very shoal, atonly 38 m in the case of 30215; two other stations(30246 and 60284) lay on the southern slope ofthe Costa Rica dome, having mixed layer depthsless than 20 m and rather shoal values for D2 of:< 60 m (regional means were 92 to 117 m),

I-VII

DAr + ,'fIG1:i":NIGET

I-III IV J V

DJ.l'

I-lIt tv I v

740 ALAN R. LOI'GHURST

At the final station 76120, at which DPSapparently lay 101 m below D2, there was a deepmigrating layer lying unusually shoal, and in­separable from the epiplankton, so DPS musthave been calculated in error.

The eleven stations at which DPS could notbe identified were, with a single exception, allat the extreme south of the EASTROPAC areawhere Dl and D2, but especially the latter, lievery deep (p. 734 and Table 1). In Regions VIand VII, DPS could be identified only at nightstations in Region VI; at day stations in VI, andboth day and night stations in VII, it was notpossible to identify a planktocline below anabundant layer of epiplankton. These profilessuggest that in these regions of very deep pycno­clines it is only at night, as a result of thecrepuscular rise of diel migrants that a plankto­cline is formed, and that even this does not occursouth of about 15°S. Unfortunately, there weretoo few stations in Regions VI and VII to analysestatistically the relationship between DPS anddensity or temperature.

The single station (30233) in low latitudes atwhich DPS could not be identified was situatedin an area of dynamic upwelling to the west ofthe Galapagos Islands, and no explanation forthis anomaly is evident.

LightEASTROPAC data on light penetration are

derived from 30-cm Secchi disc extinction depthsand therefore are available only for day stations.Reflecting the fact that the only important sourceof turbidity in this area is pelagic biologicalproduction, light penetration is closely correlatedwith the depth of the layer of abundant plant cells(Table 2). The mean value of DPS is deeper by23 ± 2.5 m than DE (P < 0.001), and althoughthe range of variability in this difference is notgreat, the attempt to demonstrate linear cor­relation fails (r = 0.204, P > 0.05), and it seemsimprobable that DPS is primarily determinedby the light regime; even though DPS shows a veryminor diel depth fluctuation, which has the correctsign to be caused by normal vertical migrationpatterns, the day-night difference (+ 7.7 m,

Table I) is not significantly different from zero(P > 0.10).

OxygenThe eastern Pacific oxygen-deficient layer is

shoal and very strongly developed in Region I,to the southwest of Baja California and theMexican coast (BRANDHORST, 1959; \VYRTKI,1962). Earlier studies in this layer (LONGHURST,1967) have shown that oxygen levels only below0.2 ml 1-1 may constrain the distribution ofinterzonal migrant zooplankton. This oxygenisopleth was as shoal as 75 m at some stations inRegion I and at four of these (12006, 12030,30250 and 30258) DPS was as much as 30 mdeeper, and well into water with very high oxygendeficiency. Such observations suggest that oxygenplays a negligible role in determining the depthdistribution of the epiplankton. Moreover, inmost of the EASTROPAC profiles, the 0.2 ml l?

isopleth lay much deeper than DPS.

PhytoplanktonOf the 86 LHPR stations, there were 77 at

which profiles of chlorophyll-a and phaeopigmentswere also obtained and 48 daytime stations atwhich profiles of 14-C uptake rate were taken.These three parameters of the phytoplankton hada rather constant relationship: the subsurfacechlorophyll maximum was underlain by a deeperphaeophytin maximum, while the depth of maxi­mum gross production rate lay shoaler. At IIsoutherly stations the phytoplankton samplingwas not carried sufficiently far down to reach thechlorophyll maximum which lay very deep indeed(> 125 m) at some stations in Regions VI andVII.

Since the simplest hypothesis relating thedepths of abundance of animals and plants is thatzooplankton do not exist in abundance below thedepth at which a significant standing stock ofphytoplankton occurs the relations of DCLNwith DPS were examined first.

DCLN lay between 105 and 1I5 m right acrossRegions I to V, showing remarkably littlevariance (Table I): it did not appear to react toshoaling of D2 in very low latitudes in Region IVand deepened less strongly than D2 in the

Interactions between zooplankton and phytoplankton profiles 741

southernmost Regions VI and VII. Because of itsrelation with D2, therefore, DPS always layshoaler than DCLN by 25 to 50 m in RegionsII to VII; only in Region I, with its very well­defined and relatively shoal mixed layer, wasthere a closer correspondence between D2, DCLNand DPS. The lack of better correspondencebetween DPS and DCLN might indicate that anincorrect chlorophyll value was selected to definethe latter; however, a value which would give abetter match in Regions I to V (say the 0.1 mg m-3

.isopleth) gives a rather poor match in thesouthernmost regimes, as would a value basedupon a percentage of the maxmimal chlorophyllvalue.

As expected, the formal analysis of differenceand correlation confirmed that DPS laysignificantly deeper than DCLX and DCX:

DPS = DCLX + 40 m ± 2.4 m (P < 0.001)DPS '= DCX + 61 m ± 2.4 m (P < 0.001),

and although the linear regressions indicate somecovariance, this is not very significant (for DPSjDCLX, r = 0.342; P 0.001 to om; for DPSjDCX, r = 0.286; P om to 0.05).

Taken together, the evidence on density, light,oxygen and phytoplankton indicates that thelower limit of the epiplankton is usually deter­.mined by the lower part of the thermocline,though not by its deepest isotherms and isopycnals,and that it occurs in a rather loose associationwith the lower limit of chlorophyll-rich water; itappears to be unrelated to light penetration or tooxygen content.

DEPTH CONTROL OF EPIPLANKTON

MAXIMA

Within the epiplankton, DZX could beidentified in 80 of the 86 profiles of biomass, oftotal numbers of organisms, of total copepodsand of total chaetognaths. There is a greater dielvariation in DZX than in DPS, and the regionalmean day-night differences range from 6.8(Region IV) to 45.0 m (Region VI), and becauseDZX is much shoaler than DPS, its day-nightdifferences are a much higher percentage of itsday depth, regional means ranging from 30 to58%.

Diel variation in the depth of maximum plank­ton aggregation reflects two different processes:the nocturnal rise (or sinking) of the aggregatedepiplankters forming the diurnal DZX, and therise of interzonal species from the planktostad totheir nocturnal layer depths within the epi­plankton. A nocturnal deepening of DZX doesnot, therefore, necessarily imply nocturnal sinkingof any individual plankters, for it may be duesimply to the layering of interzonal species atnight deeper than daytime DZX.

The preferential aggregation of zooplanktersat any particular depth may be a response to afavourable feeding situation at that depth, or totheir crowding against a barrier or constraintimposed by the physical or chemical environment.Though the first possibility seems a priori to bethe more likely in the present profiles, both arenow examined.

Constraint is only likely to be caused bydensity, because it has already been suggestedthat other possible contraints (light, oxygen) areinoperative even in the case of DPS which liesdeeper, and so more under their possible influencethan is DZX.

DensityThe zooplankton maximum occurs very close

to Dl in all strata, the overall mean value ofCZXbeing 5 ± 3.3 m shoaler than Dl; the differenceDZX-Dl is not significantly different from zero(P < 0.05) for the whole data pool and for allseparate strata with the single exception of day­time profiles in very low latitudes in Regions IVand V; some anomalously deep values for D 1 inthis area, especially where mixed layers arerelatively difficult to interpret in the equatorialdivergence cell; caused the mean difference forthis stratum to be DZX = Dl - 20 ± 5.6 rn.The linear regression DZXjDl shows verysignificant covariance (P < 0.001). Thus, thelayer of maximum zooplankton abundance withinthe epiplankton generally occurs some 50 mshoaler than DPS, and is very closely associatedwith the bottom of the mixed layer.

The aggregation of zooplankton at densitydiscontinuities has .recently been reported by

742 ALAN R. LoNGHURST

BOYD (1973) using an electronic particle counter,although the cause of the phenomenon remainsobscure. Two possible explanations suggestthemselves: firstly, that it may be a simple physicaleffect of the changed sinking rates to which anysinking particle will be subject at DI or, secondly,it may be the result of an active food-seekingprocess. That the zooplankton profiles are notconsistently skewed downwards towards DPSsuggests that the physical process alone is not thecause.

PhytoplanktonThe data show that zooplankton aggregation

frequently occurs rather close to the depth ofmaximum 14-C uptake (DCX) and usuallysignificantly shoaler than the depth of maximumchlorophyll-a standing stock (DCLX). In fact,the overall mean depths of zooplankton aggrega­tion and maximum 14-C uptake are so similar(DZX = DCX + 3 ± 3.1 m) that the differencebetween them is not significant (P < 0.001);correspondence is closer at night ·than by day,when marginally significant differences do exist.Surprisingly, the linear regression DZX/DCX isnot significant (Table 2); this may be because themean depth difference is so small, or becausenight data for DCX were derived from the nearestday station (p. 740). Because DZX and DCXcoincide so closely, DZX lies above DCLX byabout the same distance as does DCX: the meandifference is DZX = DCLX - 20.5 ± 3.5 m(P < 0.001) and significant difference is main­tained for all strata of the data (Table 3); further­more, in this case the two variables show a highdegree of positive correlation (r = 0.294,P 0.01 to 0.001).

This suggestion that zooplankters tend toaggregate in depths of actively dividing (high14-C uptake) phytoplankton in preference todepths where less actively dividing (or low 14-Cuptake) but more abundant phytoplankton occurshas important implications for the: interpretationof the form of phytoplankton profiles to whichwe shall return later (p. 749); previous detailedcorrelations between animal and plant plankton

profiles have suggested a rather different coinci­dence. ANDERSON, FROST and PETERsoN (1972)and MULLIN and BROOKS (1972) have shown acorrelation at three stations off Oregon and sixoff California, respectively, between maximumnumbers of copepods and chlorophyll-a maximain the mixed layer. HOBSON and LoRENZEN (1972)at about 70 stations in the Gulf of Mexico, theBenguela Current and in the North Atlantic' at35 to 42°N, found a coincidence betweenchlorophyll-a maxima, microzooplankton (Lugolsiodine samples) and pycnoclines. Unfortunately,neither of these sets of observations included14-C uptake data, in the absence of which it isimpossible to be sure that the conclusions reachedwould not have been different had all three sets ofdata been available.

In all EASTROPAC regions, the phaeopig­ment maximum (DPHX) was deeper thanDCLX, as expected (e.g. YENTSCH, 1965;LoRENZEN, 1967), by about 20 to 30 m so that itsdepth is much greater than that of ·DZX(DZX = DPHX - 49 ± 4.4 m), the area ofpresumed highest rate of production of thesepigments during the grazing process. It seemsprobable that the closer coincidence of DPHXwith DPS, which is only slightly, though signi­ficantly, deeper is not functional and that bothvariables are likely to be independently related tothe lower part of the pycnocline where sinkingrates of both plant cells and faecal pellets willdecrease.

Also occurring in close proximity in allregions to DPHX is the shallow NO! maximum(DNOX), an indicator of the depth of maximumremineralization oforganic material by herbivoresand heterotrophs (WADA and HATTORI, 1971);DNOX usually lay well below DZX (Table 1) andthough within the depth range of the epiplanktonis clearly unrelated to the depth of maximumzooplankton abundance but rather to the depthof aggregation of faecal material and senescentplant cells where microbial remineralizationprocesses are most active.

These observations on the relations betweenDZX, density features and phytoplankton in thegeneral zooplankton profiles can be, to a limited

Interactions between zooplankton and phytoplankton profiles 743

extent, extended by the data from the specialsorting categories.

Of the six such categories which alwaysoccurred in their maximum abundance withinthe epiplankton (the genera Euchaeta, Candacia,Acartia, Eucalanus, Lucifer and Stylocheironjoniy the last taxon showed important die! migra­tion (mean 38 m). The nocturnal layer depth ofStylocheiron and the 24-h layer depths of the othertaxa occurred very close to DZX and, thus, toDCX (Tables 1 and 4).

Despite this general coincidence with DCXthere are individual deviations: the discretelayers of these six taxa lie with rather specificinter-relationships (Table 4). Thus, Lucifer lies ina relatively narrow layer (mean vertical rangeonly 25 m by day), and shoaler (Table 4) than thecopepods Candacia, Euchaeta and Acartia whichlie closer to DZX; Eucalanus lies rather deeper(day mean 40 m) and Stylocheiron deeper yet

(day mean 71 m). Since each of these taxa, withthe probable exception of Lucifer, compriseseveral species, it has to be assumed that theepiplankton is composed of very many over­lapping specific distributions, information uponwhich was not obtained from the samples thoughit was contained there: 21 profiles fromEASTROPAC cruise 12 were analysed by BruceFrost for six species of Clausocalanus (p. 731) andshow how several co-occur at stations frequentlyand abundantly, but have specific layer depthsand migration patterns. Three species (C. para­pergens, C. jobei and C. arcuicornis occur in day­time in the planktostad and lower part of theplanktocline (mean daily depths of maximumabundance 170, 79 and 85 m, respectively) butrise at night to different levels within theepiplankton (49,54 and 40 m). On the other hand,C. furcatus, C. minor and C. mastigophorus occurin their greatest abundance very shoal in the day-

Table 4. Mean depths (m) of the abundance maxima ofthe main layers ofspecially-sorted taxa, of thedaytime vertical extent of these layers, of their range of vertical migration and of their residual

layers for all regions combined.

FISIDUAL

L41ER rEFTb.';; LArER ttrtss

Dx.! Night E.rtent F.::::n.gs Da-J la'ight

lPlI'LA/lITON TAXA.

Eucata=. .pp. 40 •• 78 "E""hast<> .pp. 28 11 50 -11

candoaia .pp. 23 30 13 ••Acartia .pp. 23 10 84 -13

Luci,.1' .pp. 19 11 42 -2

StyZocMi1'on Ipp. 11 33 56 -38

X 34 25 63 -9

IllItlZOllAr. TAXA.

lthi>:Ol%w.u. ccrr.utu3 229 19 11a -205 30 166

R. na.e:;.(tuI" 300 ]0 14 -270 41 267

PlMlrOl'r'4!tP'.:% gra.ci-tu 206 20 62 -186 40 261

P. abaoninaU./J:iphicu 30a 36 63 -272 30 100

Euphauaia diGfr:edias 322 16 59 -.30.') ]0 100

E. B:im.a 302 11 102 -285 22 100

E. gibbci#. 305 29 so -276 70 120

Ne~to6cBZ£6 grdlis 283 51 22 -226 15

E""",~brccJ,ion f1..J:ip.. 3H 10 26 -275 11 110

EucaZ=u8 bungU 391 405 56 -14

X 289 33 11 -255 33 156

744 ALAN R. Lo:>GHURST

BiC'm.l:;~ 20 II " I" 11.1

fctal ,"('pero.fs 1) II 8. J

.Joven i l e euphaus i id s JZ H zz.oAdult eupbau s i Ld s 71 7. J> 4J sr.:rh.,JctoGn.lth~ 10 1

Atiuit fisl: ~o 38 -J Z7 30.0

/i.;'i'.cal.(:):u~ ....!'rn:..1'-;~.= 4 24 14.3

R. "la~:.lt:.~ J1 30 ,~ .1

Pl€t<r~.l (La r ge species) 11 48 62 60.2

Pl.curo-narna (S!!I.:ltl spedes) 24 18 54 3!. 7

time (12, 21 and 10 m) apparently sinking atnight, the mean day-night difference being about28 m.

Obviously, the nocturnal migration of inter­zonal species upwards into the epiplankton mustcause a major shift in the vertical balance ofbiomass and numbers of individuals. To test this,progressive integrations were performed from thesurface to 500 m, or to the bottom of the profilesif shoaler, for a number of the most importantsorted categories: biomass, numbers of copepods,euphausiids, chaetognaths, pelagic fish, Rhin­calanus cornutus, R. nasutus, large and smallPleuromamma species (mainly P. xiphias, P.abdoniinalis and P. borealis, P. gracilis, respec­tively). From these step-wise integrations it waspossible to calculate the percentages of planktonabove and below any given level, and hence therelative amount in the epiplankton. Table 5shows that about 18%of the total biomass above500 m was transferred from the planktostad tothe epiplankton at night by migration of inter­zonal species.

Surprisingly, the copepod data show that lessthan to %ofall individuals down to 500 m partici­pated in this translocation and added relativelysmall numbers to the very abundant smallepiplanktonic species which, in all regions, com­prised from 60 to 80 % of all copepods above500 m. The largest part of the nocturnal increasein epiplankton biomass is derived from migranteuphausiids and small mesopelagic fish, thoughthese are much fewer in number.

Table 5. Percentages of certain categories ofplankton above 500 m which more into the epi-

plankton from the planktostad at night.

111 IV 1-\'11

Both juvenile euphausiids, in relation toadults, and small species of Pleuromantnui inrelation to large, show reduced contribution to theepiplankton at night, in addition to lying shoalerin their daytime residence layers within theplanktostad. Of the groups examined in this way,chaetognaths show the least consistent pattern ofnocturnal reinforcement of the epiplankton, thedata indicating a more scattered dispersal.

Presence ofnear-surfacemaximumIn 38 of the 86 profiles, it was possible to

identify a ncar-surface zooplankton maximumseparated by a minimum from the main body ofthe epiplankton below (p. 737). The depth of thisminimum (DZN) usually occurred between toand 20 m with a day-night difference « 2 m)insignificantly different from zero.

There was variability in the frequency ofoccurrence ofDZN both spatially and temporally.Most importantly, it was more frequent (50%)in daytime than at night (36 %); this is evidentlythe effect of nocturnal shoaling of DZX and theconsequent difficulty of separation of surface andsubsurface maxima within the epiplankton.

In Regions r to lIT, where Dl is moderate,DZN was identified in the highest proportion ofprofiles, occurring in 22 (= 63 %). Ratherfrequently in Regions IV and V, and especiallyat those stations most influenced by the equatorialdivergence, the only maximum occurring withinthe epiplankton was very shoal and lacked theminimum above it: DZN was identified here inonly 13 profiles (34 %). This can best be inter­preted as the result of the surfaceward displace­ment of DZX by the shoaling of the densityparameters of the water column. In Regions VIand VII only three profiles (23 %), none of whichwas in the gyral conditions of Region VIr,included a subsurface minimum, presumably inresponse to the very great depths of the mixedlayer and of DPS, and the very low planktonabundances, so that any aggregation within theepiplankton was hard to detect.

The chlorophylI-a profiles rather frequently(in 51 cases of the 86 examined) showed evidenceofa secondary pigment maximum very close under

Interactions between zooplankton and phytoplankton profiles 745

the surface, generally no deeper than 10 m. InRegions IV and V, within the influence of theequatorial divergence and where a very highpercentage of zooplankton profiles showed near­surface maxima, less than half of the profilesincluded such pigment maxima, perhaps indicatingthe grazing effects of 'surfaced' zooplankters.

LAYERING OF INTERZONAL SPECIESThe analysis of interzonal species was con­

cerned with three components of the plankton:the deep, non-migrant taxa, the shoaler daytimelayers of dicl migrants, and the nocturnal layerswithin the epiplankton of diel migrants from theplanktostad.

Layering within the planktostadThe layers of relatively abundant interzonal

plankton that occur within the low-biomass waterof the planktostad may form narrow strata onlyseveral tens of meters deep, or as rather thicker,more diffuse strata. The thin layers are distributedthroughout sampled depths in the EASTROPACmaterial, while the thicker layers occur 'at ratherconstant depths and apparently in a ratherconstant relationship with sonic scattering layers.The upper and lower limits of the main layer atthe 39 stations at which it was identified werefrom 226 to 330 ill at daytime stations, and 227to 349 m at a very few night-time, stations: thus,in general, it extends for rather more than 100 mbelow a very consistent depth of about 225 m.

Both this main, thick layer ofinterzonal speciesand the bottom of the epiplankton can be cor­related in about two-thirds of the daytime stationswith the records from a 30-kHz echo-sounderwhich was operated continuously on high gainbetween stations. At each LHPR station theechogram was examined in detail, using the scale­expansion capabilities of the instrument to in­vestigate the extent to which DPS, and the top(0 ') and bottom (0") of a layer of interzonalorganisms could be recognized on the expandedechograms. This showed that there is a 70%probability that an echogram will predict DPSwith a precision of ± 15 m, a 64% probabilityfor D' and a 52% probability for 0"; that the

probability decreases with depth of the featurepredicted does not necessarily imply increasinglayer-diffusion with depth, and is probably dueto the effects of signal attenuation over longersound-paths. It is reasonable to suppose thatDPS and (at least) D' were monitored in anapproximate manner by the 30-kHz underwayechograms taken throughout EASTROPAC andFig. 5 shows echograms at 112 and 115°W inFebruary to March 1967 which, compared withFig. 4, considerably expand the description of themain features of the migrating layers within theplanktostad.

Figure 5 shows that, while there is a good dealof small-scale variability in the depth of featureswithin regions, there is remarkably little consistentvariability between regions except at the southern­most end of the lines in Regions VI and VII,where the consistent deepening confirms the datashown in Fig. 2. The EASTROPAC data do notpermit us to examine light levels at D', which isconsiderably deeper (mean factor, X 3.3) thanthe depth of the euphotic zone though there is nosimple correlation between these variables: manydifferent species form the DSLs (see below) soperhaps a simple correlation is not to be expected.

However, DSL depths are well-known to bephoto-regulated (e.g. BODEN and KAMPA, 1967)and light penetration in these oceanic areas whereall turbidity is of biological origin (p. 740) isprobably more uniform than most other variablesin the water column which determine layer-depth.

Of the minor 'sorted categories of organisms,nine taxa (including seven single species), areinterzonal diel migrants (Table 4); together withbathypelagic fish (principally the myctophidlanternfish Diogenichthys spp. and the gono­stomatiid lightfish Vinciguerria spp.) these ninetaxa comprise the bulk of the diel migrants in thesamples.

As the table shows, most of the interzonal taxahad daytime layer depths, coincident with DSLdepths, of between 250 and 300 m, and mostlayers were less than 100 m thick. The anomalouslayer thickness shown for Euphausia eximia islargely influenced by a single station (76158) inthe south equatorial current at which this species

...a"..0\

~

1000

NIGH

r!O

i~

750

1000

WI CQIlSr

'0\'. COAST

100

N

FigS. Sections of sonic scattering layers at IIZoWand along a line from IzoN IIZoW to the Mexican coast on the same scale. Contoured at 3 arbit rary levelsof signal strength, in addition to 'no signal return'.

Interactions between zooplankton and phytoplankton profiles 747

occurred in large numbers down to about 300 mwithout any layering.

As with the relatively non-migratoryepiplankters, so the layer depths of the interzonalspecies do not coincide: Euphausia eximia andNematoscelis gracilis regularly lie shoaler in thedaytime than do the other species of euphausiids.E. diomediae, which generally lies deeper than anyother euphausiid in daytime, also performs thegreatest migrations and lies shoalest at night.PONOMAREYA (1963) confirms that this lasteuphausiid occurs very deep in the daytime inlow precision opening-elosing net hauls; shefound maximal numbers at 200 to 500 m, butstragglers down to 1500 m.

The large species of Pleuromamma (P. xiphias,P. abdominalis) regularly lie deeper than thesmaller species (P. gracilis, P. borealis), the latteroccurring generally shoaler than 200 m; at night,however, the layer depths of the two groups ofspecies are not so clearly separated. The largespecies of Pleuromamma were the most activelymigrating copepods of the taxa examined in thisstudy, occurring at night 250 to 300 m above theirdaytime levels.

Again, Rhincalanus nasutus and R. cornutushave mean daytime residence depths separated byabout 75 rn; compared with Pleuromamma, theseare evidently much less consistent migrants,especially in the case of R. cornu/us. Much morecommonly than some other interzonal plankters,these species leave residual layers of individualswhich fail to migrate both in the epiplankton indaytime and in the planktostad at night. ROE(1972a), indeed, found negligible diel migrationin Atlantic forms of these two species off theCanary Islands in the SOND data, as didVINOGRADOY (1968) in all regions except in thetropical divergence that occurs at the boundarybetween north equatorial and equatorial counter­currents, where he found a distinct nocturnalshoaling of the population. Thus, the occurrenceof residual layers of non-migrant individuals morecommonly in these species than is' the case forPleuromamma occasions no surprise.

Two main regional phenomena can be detectedin the data for layer depths within the planktostad

and these are connected (I) with the thermoclineshoaling at the equatorial divergence and (2)with the greater thermocline depth in the regionof the southern summer thermocline.

Examination of individual station data showsthat in the equatorial divergence in Region IV, thedaytime layer depths of euphausiids remainedgenerally deeper than 200 mat 8 stations, althoughat 10 others the daytime layers were shoaler than50 m, even occurring in the upper 10 m; thisoccurrence of adult euphausiids at the surfacewas confirmed at several stations by actualobservation (from the underwater ports of thebow-chamber) of red, adult euphausiids in largeswarms. This was not an entirely passive reactionto water dynamics, for at each station not alleuphausiids were surfaced, or normally deep; at12100,for instance, E. gibboides remained centeredat 320 m, while E. diomediae, E. eximia, Nemato­scelis gracilis and Nematobrachion flexipes cen­tered at 30 to 35 m. Similarly, among the cope­pods, at three of six stations in Region IV, bothspecies of Rhlncalanus and of Pleuromamma hadlayer depths at less than 30 m, mostly at less than10 m, although at equatorial stations, part of thePleuromamma population remained in daytime atdepths exceeding 200 m,

In Regions VI and VII, to the south of themain stream of the south equatorial current,daytime layer depths of interzonal species as ofthe epiplankton, were greater than elsewhere inthe survey. In almost every case, layer depths inVII were greater than in VI; at night, also, thetabulated data indicate that layer depths weredeeper in these regions than elsewhere. These arethe expected consequences of greater mixed layerdepths and greater depths of features in thephytoplankton profile that have been discussedabove, and are apparently unrelated to differentiallight penetration, which is remarkably slight(Table 1).

One interzonal taxon, Eucalanus bungli,which occurs deeper than aU other sorted taxa,at depths in which black bathypelagic augaptalidcopepods and scarlet hoplophorid carideans werebeginning to occur in the sub-samples, does notappear to be a diel migrant; this species occurs

748 ALAN R. LoXGHURST

extremely deep in only a few samples and formsdiscrete layers at depths ranging down to 860 m.The most prominent layers of this species occurin the special deep samples (76197, 76202, 76158)in which strong layers occurred at 600 to 860 m,each being itself less than 100 m thick. Thisspecies shows indication neither of verticalmigration, nor of equatorial shoaling, and clearlyplays an unimportant role in the tropical pelagicecosystem, yet E. bungii is a very important andabundant interzonal plankter in the northernNorth Pacific where it lies generally below 200mand shows little diel migration (VINOGRADOV,1968); in the California Current (LONGHURST,1967) it is widespread below 300 m and evidentlyit is even further submerged in the EASTROPACarea through which it becomes progressively lessfrequent southwards (station occurrences: RegionsI to III = 34 %. IV and V = 27 %. VI and VII =15%).

Nocturnal layers within the epiplanktonWe have seen earlier (p. 741) that the layer of

general plankton aggregation (DZX) lies veryclose· to DCX and 20 to 30 m shoaler than thedepth of the chlorophyll maximum; the data inTable 4 show that the depths to which individualtaxa of interzonal migrants rise at night alsocorrespond more closely to DCX than to DCLX.Thus, although the numbers of animals whichparticipate in the diel translocation between theplanktostad and the epiplankton are not relativelygreat (p. 743), they do contribute to the generalobserved aggregation around DZX.

The migrant interzonal taxa for which datawere acquired have mean nocturnal layer depthswhich range from 16 to 70 m (Table 4); only twospecies, Nematoscelis gracilis (mean 57 m) andNematobrachion flexipes (mean 70 m) differ morethan 12 m from mean DCX, and five species hadtheir maximum abundances with 10 m of DCXat night.

As within the planktostad by day, so there isspecific variation between layer depths of theinterzonal taxa within the epiplankton at night;among the four copepod taxa, those which laydeepest in the p1anktostad during the daytime

iRhincalanus nasutus, Pleuromamma abdominalis­xiphias) lay deepest in the epiplankton at night.The five species of euphausiids did not followsuch a simple pattern, however, and the shoal estin the daytime (Nematoscelis gracilis) lay deepestat night. The most abundant euphausiid in theEASTROPAC area, Euphausia diomediae, regu­larly lay very shoal at night, having maximumabundances at less than 25 m in all regions.

ROE (1972b) showed quite clearly from theSOND data for six species of Pleuromamma thatthe larger species lay deeper than the smallerduring the daytime and his Fig. 8 shows someindication that the small species also lie relativelyshoaler at night, but the EASTROPAC datareviewed here suggest that lack of vertical resolu­tion in the SOND samples (only five levels to250 m) may have obscured the fine detail thatevidently occurs within the upper 50 m at night.

CONCLUSIONSThe principal conclusions to be drawn from

this study can be stated quite simply: zooplanktonabundance is highest above the lower pycnoclineand high biomass values rarely extend deeper thanthe layer of abundant chlorophyll; within thislayer, both by day and by night, zooplanktontends to aggregate closer to the depth of maximumcarbon fixation than to the depth of maximumchlorophyll standing stock. Finally, and perhapsless importantly, the presence of a near-surfaceenrichment of both animals and plants appearsto have two possible causes: (i) extreme shoalingof the pycnocline which is associated with theshifting of biological processes close towards thesurface, and (ii) an extremely stable mixed layerof moderate depth allowing the developmentwithin its upper few meters of a specializedneuston based on plant production processeswhich are probably to an extent isolated fromthose below.

Although several theoretical models whichaccount for the general form of phytoplanktonprofiles have been published, these withoutexception include zooplankton grazing pressureas a constant, usually throughout the watercolumn modelled. This could hardly be otherwise,

Interactions between zooplankton and phytoplankton profiles 749

for only very few and very recent studies haveinvestigated zooplankton-phytoplankton verticalrelations in detail.

The model of STEELE and YENTSCH (1960) toaccount for the vertical distribution ofchlorophyll,based on those of RILEY, STmtMEL and BUMPUS(1949), explains the observed chlorophyll distribu­tion by the sinking of plant cells; accumulationof cells is assumed to occur either in response todensity discontinuities or to the decreased sinkingrates of plant cells below the euphotic zone dueto their changing physiology in the absence oflight.

These models recognize two conditions: thevery shallow chlorophyIl maximum which occurswheri a phytoplankton population is increasingexponentiaIly, and a deeper chlorophyll maximumwhen the phytoplankton population reaches anapproximately steady state, later in the summer inhigh latitudes. In the former case, the chlorophylland gross photosynthesis maxima coincide; in thelatter, the chlorophyIl maximum occurs deeperthan the photosynthesis maximum.

As BLACKBURN, LAURS, OWEN and ZEITSCHEL(1970) showed, EASTROPAC standing stocks ofphytoplankton, zooplankton, micronekton andfish varied seasonally by a factor of less than two,a very small variability relative to high latitudes;moreover, except in the eastern EASTROPACarea, where no LHPR hauls were taken, and theseasonal cycle is strongest, there was no detectablelag between the seasonal maxima of phyto­plankton and zooplankton. Blackburn's study,which is based on the most comprehensive time­series ofdata available for the low-latitude pelagicecosystem, thus confirms the previous predictionof HEINRICH (1962) that producers and con-·sumers of plant material should co-vary season­ally in the open tropical ocean. It further suggeststhat here one cannot expect to find the detailedcoincidence in the vertical sense between maximaof chlorophyll-a and carbon fixation rate whichcan exist ephemerally in higher latitudes withgreater instability in the seasonal production cycle.Thus the EASTROPAC profiles might generallybe expected to approximate the summer chloro­phyll profiles of much higher latitudes, and the

sections presented in the EASTROPAC atlases(LOVE, 1971) confirm that this is the case.

What is new in the LHPR profiles is thedemonstration that zooplankton aggregation, andhence grazing pressure on plant ceIls, is strikinglynon-uniform with. respect to depth and con­centrates closer to the depth of maximal plantproduction than to maximal plant material. It ishard to avoid the suggestion, therefore, that acausal relationship exists between the form of theprofiles ofherbivore and producers. That, in short,the form of the chlorophyll profile is influenced,or even determined, by a depth-differential in thegrazing pressure of herbivores, in addition to theeffects of differential sinking rates invoked in theearlier models.

Of course, not all the zooplankters profiled inthis study were, in fact, obligate herbivores; nordid they comprise all the pelagic herbivorespresent in the water column. However, thenumerical profiles are dominated by smallcopepods which are at most only one trophiclevel above the primary herbivores whose biomassis likely to be dominated by very small organisms,and hence it is reasonable to consider the planktonprofiles studied here as dominated by herbivorousorganisms. There is evidence that levels of maxi­mum abundance of the microzooplankton, whichin this region must be extremely important con­sumers of primary production, coincides atEASTROPAC stations with DZX as determinedin this study (e.g. BEERS and STEWART, 1971).It seems unlikely that the larger organisms of themicro nekton (BLACKBURN, .LAURS, OWEN andZEITSCHEL,1970) contribute importantly to herbi­vore grazing pressure, so it seems reasonable toequate LHPR-plankton in this region withherbivore distribution.

An hypothesis similar to that put forwardhere has already been proposed by LORENZEN(1967) on the basis of a smaIl number of LHPRstations, at which phytoplankton profiles werealso obtained, off Baja California. Further.VENRICK, MCGOWAN and MANTYLA (1973) havesuggested that differential zooplankton grazingfrom animals concentrated above the chlorophyllmaximum layer in low latitudes may help to

750 ALAN R. LoNGHURST

maintain a sharp gradient above it, but proposethat the deep chlorophyll maximum is ultimatelydetermined by the nutrient regime. Such sugges­tions are an advance on the earlier models dis­cussed above which depend on decreased sinkingrates of plant cells, but the hypothesis has yet tobe tested directly to validate the relationship whichcould be demonstrated in the EASTROPAC dataonly in a statistical sense; direct validation dependson simultaneous and matched high precisionprofiles not only of plant pigments and zoo­plankton but also of plant production rates. Sofar as I am aware, production 'profiles of therequired precision have not yet been made.

The grazing hypothesis also raises the questionof why additional grazing pressure does notdevelop deeper than DZX, where the chlorophyllprofile suggests that food availability for grazingorganisms is higher than at DCX. Why do theinterzonal herbivores traverse the chlorophyllmaximum in their nocturnal rise (p. 742), to seekdepths where phytoplankton biomass is apparentlyalways lower?

Two possible answers suggest' themselves:firstly, that the phytoplankton population at thechlorophyll-a maximum is physiologically andbiochemically less attractive as a food sourcethan the plant cells at the depths of maximumgross photosynthesis or, secondly, that the otherfactors presumed to determine the rate andextent of diel migration, and of residence depth ofnon-migrants-light and temperature-over-ridethe selection ofthe depth ofmaximum chlorophyll.

Certainly, the plant cells in the photosynthesismaximum (DCX) differ biochemically from thosedeeper (e.g. JEFFREY, 1974) and actively dividingcell populations will contain a higher proportionoflarger pre-division cells which have been shownto induce increased feeding rates in a calanoidcopepod (RICHMAN and ROGERS, 1969); thedeeper cells are senescent, may not be activelydividing, and may be less attractive to herbivorousplankton. On the other hand, one must beimpressed by the rigidity of migrant plankters'response to light intensity (e.g. BODEN andKAMPA, 1967; BACKUS, CLARKE and WING, 1965)which simply may not permit sufficient flexibility

of response for zooplankton grazing pressure todevelop on plant cell concentrations at whateverdepth these may be encountered.

The general hypothesis that there is a relation­ship between the nature of zooplankton profilesand the vertical distribution of density andphytoplankton production is now examined inthe light of regional and seasonal differenceswithin the EASTROPAC area.

It is in Region I that plankton profiles (Fig. 6)most frequently show all major characteristicsof the epiplankton, and this is clearly related tothe isothermal mixed layer defined by an extremelysharp thermocline. In general, Dl is heresufficiently deep for the development within themixed layer of a sub-surface zooplankton maxi­mum separated from a surface maximum andthis is associated with the development of charac­teristic phytoplankton profiles.

As expected, DZX shoaled in the second halfof the year to about half its mean depth forFebruary to March but, despite this crowdingsurfaceward by the shoaling of Dl, the zoo­plankton profiles retained their typical form; veryhigh zooplankton volumes occurred near lOOoWin cruises 30 and 60 in relation to doming offCosta Rica. Even though tunicates and coelen­terates were unusually abundant in these profiles(especially 30254) their essential profile-formremained typical of the region.

Only four stations were occupied in the northequatorial current in Region II, one in Februaryto March 1967 and three in the December toJanuary following: these are insufficient tocharacterize the plankton profiles in this region,but give no evidence ofbeing significantly differentfrom those characteristics of Region I.

Because of its seasonal occurrence in thestudy area (p. 732) profiles were obtained from thenorth equatorial counter-current (Region Ill)only in the second half of .the year and theseshowed Dl to be much shoaler than in Region Ito the north of the thermocline ridge marking thenorthern boundary of the north equatorialcounter-current, Related to this, the zooplanktonprofiles frequently showed no separation betweensurface and subsurface maxima within the

S'S~o'~

i

Ic~ji;"

~8'::lro::lc,

~8''E.ro

~e'0

a[

~1348

\ ' 11

7U112 " II

.-WJ...==..·..;;;:;.;..:-·:·:·::;:«··,.

~jllitifffit{

VII

14!>19

\'I

v

..

litVIIV

7W43

12221 1V

..,y .. .«-»7h«<::::-::-:-~~:~:~,<;??:~~;i~~"xf-:~~::::::~:::-:::'W~··:S:~·'*~?::~:~=*~~~$%'~;1':4::=':@~

...IV

1200 IV

2450 ._'....,===~fW ':-:1

@.~Mt#.t!i111

i{{-M'?Wfi-t.~'t'iiitX

DAY

12003 I

12002 I

-'=~"~'"

0»

2Qt)-

(]

E

~

Fig. 6. Two exemplary profiles for each of four regions (I, IV, VI and VII) by day and by night, to show main characters of typical profiles of these regions; thethermocline, as indicated by depth interval between Dl and D2, is stippled. Maximal numbers of organisms (100 rn8)-1 are indicated above each profile.

....:lU1

752 ALAN R. LmmUURST

epiplankton, and mean DPS was shoaler than inReg ion I.

In the second half of the year in Region IVthe very strong vertical eddying as the equatorialunder-current surfaced resulted in a relativelycold and deep mixed layer, which was reflectedin the distribution of the main features of theepiplankton; however, early in the year when DIis less than 10 m, the epiplankton extendsrelatively deeper into the thermocline than inregions with moderate depths of D I (Fig . 6). Theratio DPS/DI for Regions I, II, III and V rangedfrom 1.7 to 3.8 for all seasons; in Region IV,however, it was 3.5 in June to July when thethermocline was deepest, but reached 9.6 inDecember to January and 45.8 in February toMarch when the thermocline was shoalest. Thisconfirms that in extreme conditions of shoalthermoclines, the apparent lower limiting depthof the epiplankton responds principally to somefactor other than density, for which there is nodiscontinuity at D2; the difference in depthbetween DZX and DCX remains relatively slightat all seasons in this region despite the"importantvariations in DI, suggesting that under these veryunusual conditions the link between plants andanimals is greater than that between either anddensity.

Values for DI in Region VI arc intermediatebetween those in Regions I to III and V to thenorth and the greater depths in VII to the south,and the decreased mixed layer stability in VIimplied by the step-wise temperature profilesabove D I is apparently reflected in the zoo­plankton profiles which here mostly do not includea subsurface minimum (Fig. 6); the existence ofa depth zone of relatively sparse plankton in theupper part of the epiplankton is evidently acharacteristic of regions with a therrnostad aboveD I. Most profiles in Region VI have a simplesubsurface maximum within the epiplankton orelse a relati vely complex system of maxima whichare usually rather obviously correlated withdensity features (Fig. 6). The profiles of chloro­phyll -a and carbon fixation are nat sufficientlydetailed, however, to demonstrate similar cor­relations with density.

To the south of 14°S, in the gyre proper wheresouth Pacific central water characteristics occur,stations were worked in Region VII only inFebruary to March, in the southern summer.These zooplankton profiles (Fig. 6) arc mostlyunlike those at any other stations in theEASTROPAC area; daytime profiles from 8 to14°S have a very shoal surface maximum andseveral discrete . subsurface ma xima, commonlyat depths exceeding 100 m, while in the southerngyral water the profiles were composed of almostuniformly low zooplankton densities from thesurface to the greatest depths sampled. At nightfrom 8 to 14°S, nocturnal migration upwardsresulted at several stations in relatively heavyplankton densities from the surface to a plankto­cline at less than 100 m depth.

Because neither DZX nor DCX could be aswell defined as in more northerly regions, andbecause DCLX was apparently often locateddeeper than chlorophyll sampling was carried,the relationship between plant and animal profilesappeared to be relati vely loose here. This maywell be a result which is more apparent than real,however, and to some extent is certain to be areflection of the inadequacy of the standardEASTROPAC sampling protocol under theseconditions.

This analysis of the EASTROPAC datasuggests a generalization which may be extra­polated to other areas of the ocean: in fact, towherever the phytoplankton profile includes achlorophyll maximum which underlies a primaryproduction maximum. VENRICK, McGOWAN andMANTYLA (1973) review deep « 100 m) chloro­phyll maxima in the Indian Ocean, the SargassoSea, the Gulf of Mexico and the northwestPacific, and their own data show that one extendsfrom 80'W to 1800E and from 45°N to 25°S in thePacific Ocean. The EASTROPAC data suggestthat the chlorophyll maximum is usually overlainby the depth of the maximum rate of primaryproduction, itself probably corresponding to thedepth of maximum photosynthetic rate per unitof chlorophyll (ICHIMURA, 1956).

In higher latitudes this pattern occurs only fora few weeks or months each summer, as STEELE

Interactions between zooplankton and phytoplankton profiles 753

and YENTSCH (1960) have suggested and asANDERSON, FROST and PETERSON (1969) haveshown off Oregon at about 45°N. In areas ofstrong divergence or coastal upwelling it cannotbe expected to occur at all, as is seen in theEASTROPAC profiles otT South American(LOVE, 1971), and in the Somali Currentprofiles published by YENTSCH (I965). Not onlydoes the instability of the water column in suchplaces preclude the establishment of the relation­ship postulated here, but also strong seasonalvertical migration occurs in the zooplankton.As in high latitudes, where the zooplankton profilein winter is quite different from those observedin the present study due to deep over-winteringpopulations of sub-adult copepods (e.g.VINOGRADOV, 1968), in intermittent low-latitudeupwelling situations the same may occur in theinterstadial (LONGHURST, 1967). Under suchcircumstances the vertical distribution of grazingpressure must be quite different from that instable water columns in the EASTROPAC area.

Acknowledgements-This study was begun as part of theScripps Tuna Oceanography Program, under Bureau ofCommercial Fisheries Contracts 14-7·ooo7-n,and continuedas part of the programme of the National Marine FisheriesService at the Fishery-Oceanography Center at La Jolla,California. It was completed at the Institute for MarineEnvironmental Research of the Natural EnvironmentResearch Council, at Plymouth, England. I am gratefulto my EASTROPAC colleagues for supplying me withmuch data prior to publication, and especially to DoxSEIBERT, Scripps Institution for Oceanography, who wasentirely responsible for data acquisition from the originalLHPR samples. BRUCE FROST (University of Washington)generously provided me with his data on Clausocalanusprior to its use in his doctoral thesis (University of Cali­fornia, San Diego). JIM ZWEIFEL (La Jolla), and CHARLESFAY, LINDA WOODWARD and PUluP RADFORD (Plymouth)gave me essential help with computer processing andstatistical treatment of the data.

REFERENCESAHLSTROM E. (1971) Kinds and abundance of fish

larvae in the eastern tropical Pacific, based oncollections made on EASTROPAC I. FisheryBulletin, U.S. Department of Commerce, 69, 3-78.

ANDERSON G. C. (1969) Subsurface chlorophyllmaximum in the northeast Pacific Ocean. Limno­logy and Oceanography, 14, 386-391.

ANDERSON G. C., B. W. FROST and W. K. PETERSON(1972) On the vertical distribution of zooplanktonin relation to chlorophyll concentration. In:

Biological oceanography of the northern NortltPacific Ocean, A. Y. TAKEr-:OlITI, editor, IdemitsuShoten, pp. 339-345.

BACKUS R. B ., G. L. CLARKE and A. WING (1965)Behaviour of certain organisms during the solareclipse of July 20, 1963. Nature, 205, 989-991.

BAKER, A. DE C., M. R. CLARKE and M. J. HARRIS(1973) The N.J.O. combination net (RMTl + 8)and further development of rectangular midwatertrawls. Journal ofthe Marine Biological Associationofthe United Kingdom, 53,167-184.

BARRACLOUGH W. E., R. J. LEBRASSEUR and O. D.KENNEDY (1969) Shallow scattering layer in thesub-Arctic Pacific Ocean: detection by high­frequency echo-sounder. Science, 166, 611-613.

BEERS J. R. and G. L. STEWART (1971) Micro­plankters in the plankton communities of theupper waters of the eastern tropical Pacific. Deep­Sea Research, 18, 861-883.

BLACKBURN M., R. M. LAURS, R. W. OWEN and B.ZEITSCIIEL (1970) Seasonal and areal changes instanding stocks of phytoplankton, zooplanktonand micronekton in the eastern tropical Pacific.Marine Biology, 7, 14-31.

BODEN B. P. and E. M. KAMPA (1967) The influence ofnatural light on the vertical migrations of an animalcommunity in the sea. In: Aspects of marinezoology, N . B. MARSHALL, editor, Symposia ofthe Zoological Society ofLondon, 19, 15-26.

BOYD C. M. (1973) Small-scale spatial patterns ofmarine zooplankton examined by an electronic insitu zooplankton detecting device. NetherlandJournal of Sea Research, 7, 103-111.

BRANDHORST W. (1959) Nitrification and denitrifica­tion in the eastern tropical Pacific Ocean. Journaldu Conseil. Conseil permanent international pourl'exploratlon de la mer, 25, 3-20.

BRINTON E. (1967) Vertical migration and avoidancecapability of euphausiids in the California Current.Limnology and Oceanography, 12,451-483.

CROMWELL T., R. B. MONTGmtERY and E. D. STROUP(1954) Equatorial undercurrent in the PacificOcean revealed by new methods. Science, 119,648-649.

FOWLER G. H. (1898) Contribution to our knowledgeof the plankton of the Faeroe Channel. Pro­ceedings ofthe Zoological Society ofLondon, 4, 545.

FOXTON P. (1956) The distribution of the standingstock of zooplankton in the Southern Ocean.Report. "Discovery" Expedition, 28,191-236.

GRICE G. D . and K. HULSEMANN (1965) Abundance,vertical distribution and taxonomy of calanoidat selected stations in the northeast Atlantic.Journal ofZoology, 146, 213-262.

HEINRICH A. K. (1962) The life histories of planktonanimals and seasonal cycles of plankton com­munities in the oceans. Journal du Conseil. Conseilpermanent international pour l'exploration de lamer, 27, 15-24.

HOBSON L. A. and C. J. LORENZEN (1972) Relation­ships of chlorophyll maxima to density structure

754 ALAN R. LoXGlIURST

in the Atlantic Ocean and Gulf of Mexico. Deep­Sea Research, 19, 297-306.

ICHIMURA S. (1956) On the ecological meaning oftransparency in the production of matter inphytoplankton community of lake . BotanicalMagazine, Tokyo, 69, 219-226.

JEFFREY S. W. (1974) Profiles of photosynthetic .pig­ments in the open ocean using thin-layer chromato­graphy. Marine Biology, 26, 101-110.

JONES J. H. (1969) Surfacing of the Pacific equatorialundercurrent: direct observations . Science, 163,1449-1450.

KINZER J. (1970) On the contributions of euphausiidsand other planktonic organisms to deep scatteringlayers in the eastern North Atlantic. In: Pro­ceedingsofan international symposium on biologicalsound scattering ill the ocean, O. B. FARQUHAR,editor, U.S. Navy Me Report 005, pp. 476-489.

LoNGHURST A. R. (1967) Vertical distribution of zoo­plankton in relation to the eastern Pacific oxygenminimum. Deep-Sea Research, 14, 51-63.

LONGHURST A. R., A. D. REITH, D. L. R. SEIBERT andB. BoWER (1966) A new system for the collectionof multiple serial plankton samples. Deep-SeaResearch, 13, 213-222.

LoRENZEN C. J. (1967) Vertical distribution of chloro­phyll and phaeo-pigments: Baja California. Deep­Sea Research, 14, 735-746.

LoVE C. (1971) EASTROPAC Atlas. NationalOceanic and Atmospheric Administration, FisheriesCircular, 330, pp. 1-12 + figs.

MCGOWAN J. A. (1971) Oceanic biogeography of thePacific. In: The micropalaeantology of the oceans,B. M. FURNELL and W. R. RIEDEL, editors,Cambridge University Press, 3-74.

MULLIN M. M. and E. R. BROOKS (1972) The verticaldistribution ofjuvenile Calanusand phytoplanktonwithin the upper 50 m of the water off La Jolla,California. In: The biological oceanography of thenorthern North Pacific Ocean, A. Y. TAKENOUTI,editor, Idemitsu Shoten, pp. 347-354.

PONOMAREVA L. A. (1963) Euphausiids of the NorthPacific, Nauka Publications, pp. 1-154.

RICHMAN S. and J. N. ROGERS (1969) The feeding ofCalanus helgolandicus on synchronously growingpopulations of the marine diatom Ditylum bright­.....ellii. Limnology and Oceanography, 14,701-709.

RILEY G . A., H . STOMMEL and D. F. BUMPUS (1949)Quantitative ecology of plankton of the westernNorth Atlantic. Bulletin of the Bingham Oceano­graphic Collection, 12, 1-169.

ROE H. S. J. (1972a) The vertical distribution andvertical migrations of calanoid copepods collectedon the SOND cruise, 1965 (II). Journal of theMarine Biological Association of the UnitedKingdom, 52, 315-343.

ROE H. S. J. (1972b) The vertical distribution andvertical migrations of calanoid copepods coll ectedon the SOND cruise, 1965 (III). Journal of theMarine Biological Association of the United King­dom, 52, 525-552.

ROE H. S. J. (1974) Observations on the diurnalvertical migrations of an oceanic animal com­munity. Marine Biology, 28, 99-114.

STEELE J. H. and C. S. YENTSCH (1960) The verticaldistribution of the chlorophyll. Journal of theMarine Biological Association of the UnitedKingdom, 39, 217-226.

TSUCHIVA M. (1974) Variation in the surface geo­strophic flow in the eastern intertropical PacificOcean. Fishery Bulletin. U.S. Department ofCommerce, 72, 1075-1086.

VENRICK E. L., J. A. MCGOWAN and A. W. MANTYLA(1973) Deep maxima of photosynthetic chloro­phyll in the Pacific Ocean. Fishery Bulletin, U.S.Department of Commerce, 71, 41-52.

VINOGRADOV M. E. (1968) Vertical distribution of theoceanic zooplankton, Nauka Publications, pp,1-320.

WADA E. and A. HAlTORI (1971) Nitrite metabolismin the euphotic layer of the central Pacific Ocean.Limnology and Oceanography, 16, 134-139.

WILLIAMS R. B. and C. H. GIBSON (1974) Directmeasurements of turbulence in the Pacificequatorial undercurrent. Journal of PhysicalOceanography, 4, 104-108.

WVRTKI K. (1962) The oxygen minima in relation toocean circulation. Deep-Sea Research, 9, 11-24.

WYRTKI K. (1964) The thermal structure of theeastern Pacific Ocean. Deutsche hydrographischeZeitschrift, Erganzungsheft Reihe A, 6, 1-84.

YENTSCH C. S. (1965) Distribution of chlorophyll andphaeophytin in the open ocean. Deep-Sea Research,12, 653-666.