Biological Consequences of El Nifno Richard T. Barber and Francisco P. Chavez event. During 1982 and 1983 triweekly observations were made at shore sta- tions in Paita and on the equator at the Galapagos Islands. In addition, ship- board observations were made quarterly along the five transects shown in Fig. 1. Conceptual Framework El Niiho is defined by the appearance and persistence, for 6 to 18 months, of anomalously warm water in the coastal and equatorial ocean off Peru and Ecua- dor. However, the anomaly in the east- ern tropical Pacific Ocean is only one facet of a large-scale phenomenon in- volving the global atmosphere and the entire tropical Pacific. In addition to major ecological and agricultural conse- quences in Asia and the Americas (1), the anomalous ocean conditions of El Niino are accompanied by large reduc- tions in plankton, fish, and seabirds in the normally rich waters of the eastern equatorial Pacific. To understand how the ocean changes affect the biology, it is necessary to describe the unique charac- ter of this region. The eastern equatorial Pacific over a wide band along the coast and equator normally is remarkably cool (2). In Au- gust 1982, for example, a swimmer on a beach at 5°S near Paita, Peru (Fig. 1), would have encountered water that was 170C despite continuous heating by the equatorial sun. This August 1982 condi- tion was close to normal [mean August temperature is 16.5°C (3)] and demon- strates that ocean processes must contin- uously export from the region a portion of the solar heat received. By October 1982, however, sea-surface temperature (SST) had risen 50C (Fig. 2), sea level had risen, the surface mixed layer had deepened, and the thermocline had been depressed by 50 m or more (4). The region's normal cool conditions result from a dynamic balance of heat transfer between ocean and atmosphere and between different parts of the ocean. El Ninlo is what results when the normal balance is upset. Wyrtki (5) suggested that El Nifio, in addition to transporting extra heat into the eastern equatorial Pacific, also interrupts the normal export of heat. Cool SST's are accompanied by three other correlates of reduced heat storage in the upper layer of the ocean: a low sea level, a shallow surface mixed layer, and a shallow thermocline (6). Of these four characteristics of the eastern boundary region of the Pacific, sea level does not 16 DECEMBER 1983 have direct biological sequelae, but the cool SST's, shallow surface mixed layer, and shallow thermocline result in high annual productivity at all trophic levels of the ecosystem (7). Upwelling ecosys- tems appear not to differ qualitatively from other marine ecosystems, but they differ quantitatively. It is this high pro- ductivity that El Niino disrupts. There is a clear theoretical basis for a decrease in biological productivity dur- ing El Ninio. The theory has two causal aspects, one dealing with inorganic plant nutrients such as nitrate, phosphate, or silicate and the other with the supply of light for photosynthesis. The major inor- ganic nutrient reservoir of the ocean is water below the thermocline (11); any process that depresses the thermocline away from the surface layer, where there Summary. Observations of the 1982-1983 El Nifio make it possible to relate the anomalous ocean conditions to specific biological responses. In October 1982 upwelling ecosystems in the eastern equatorial Pacific began a series of transitions from the normal highly productive condition to greatly reduced productivity. The highly productive condition had returned by July 1983. Nutrients, phytoplankton biomass, and primary productivity are clearly regulated by the physical changes of El Ninio. Evidence from 1982 and 1983 also suggests effects on higher organisms such as fish, seabirds, and marine mammals, but several more years of observation are required to accurately determine the magnitude of the consequences on these higher trophic levels. There is ample evidence for year-to- year variations in biological production along the Peru coast (7, 8), but the envi- ronmental changes responsible for these variations are not known. During the equatorial warming event of 1975 (9), anomalously warm water west of the Galapagos Islands at 95°W was nutrient- rich (10), but primary productivity was reduced about tenfold from values in the 1960's at the same location and season. At the onset of the 1976 El Ninlo a sudden bloom of the dinoflagellate phy- toplankter Gymnodinium splendens took place in the coastal waters along Peru when a sudden warming of the nutrient- rich surface layer increased the static stability. Interest in the causality of changes in biological productivity during thermal anomalies in the coastal and equatorial ocean led to a program in which the environment of plankton was studied in terms of both day-to-day and year-to-year changes. Anecdotal ac- counts from 1972 and 1976 suggested that biological consequences of El Nifio showed up almost simultaneously along the entire Peru coast, persisted for months, and then rapidly disappeared. The program, begun in early 1982, was designed to adequately evaluate such an is enough light for photosynthesis, will necessarily reduce productivity. Be- cause light decreases exponentially as a function of depth, the depth of the sur- face mixed layer in which phytoplankton are homogeneously distributed deter- mines the quantity of light that can be captured by the phytoplankton (12). If the mixed layer is deep, phytoplankton spend a greater proportion of time in the dark and water molecules, not phyto- plankton, absorb most of the light. Note that we have invoked two of the corre- lates of ocean thermal dynamics, ther- mocline depth and mixed layer depth, whose variability is an integral part of El Ninlo (6). Regions of the ocean, such as the eastern tropical Pacific, having a thermocline near the surface and a very shallow mixed layer are biologically rich- er than other regions of the ocean (13) because large-scale thermal structure provides the opportunity for enhanced simultaneous capture of light and nutri- ents. This enhanced primary production is reflected in all levels of the ecosytem Richard T. Barber is professor in the Department of Zoology and the Department of Botany and Francisco P. Chavez is a graduate student in the Department of Botany, Duke University, Beaufort, North Carolina 28516. 1203
Transcript
Biological Consequences of El NifnoRichard T. Barber and Francisco P. Chavez
event. During 1982 and 1983 triweeklyobservations were made at shore sta-tions in Paita and on the equator at theGalapagos Islands. In addition, ship-board observations were made quarterlyalong the five transects shown in Fig. 1.
Conceptual Framework
El Niiho is defined by the appearanceand persistence, for 6 to 18 months, ofanomalously warm water in the coastaland equatorial ocean off Peru and Ecua-dor. However, the anomaly in the east-ern tropical Pacific Ocean is only onefacet of a large-scale phenomenon in-volving the global atmosphere and theentire tropical Pacific. In addition tomajor ecological and agricultural conse-quences in Asia and the Americas (1),the anomalous ocean conditions of ElNiino are accompanied by large reduc-tions in plankton, fish, and seabirds inthe normally rich waters of the easternequatorial Pacific. To understand howthe ocean changes affect the biology, it isnecessary to describe the unique charac-ter of this region.The eastern equatorial Pacific over a
wide band along the coast and equatornormally is remarkably cool (2). In Au-gust 1982, for example, a swimmer on abeach at 5°S near Paita, Peru (Fig. 1),would have encountered water that was170C despite continuous heating by theequatorial sun. This August 1982 condi-tion was close to normal [mean Augusttemperature is 16.5°C (3)] and demon-strates that ocean processes must contin-uously export from the region a portionof the solar heat received. By October1982, however, sea-surface temperature(SST) had risen 50C (Fig. 2), sea levelhad risen, the surface mixed layer haddeepened, and the thermocline had beendepressed by 50 m or more (4).The region's normal cool conditions
result from a dynamic balance of heattransfer between ocean and atmosphereand between different parts of the ocean.El Ninlo is what results when the normalbalance is upset. Wyrtki (5) suggestedthat El Nifio, in addition to transportingextra heat into the eastern equatorialPacific, also interrupts the normal exportof heat.Cool SST's are accompanied by three
other correlates of reduced heat storagein the upper layer of the ocean: a low sealevel, a shallow surface mixed layer, anda shallow thermocline (6). Of these fourcharacteristics of the eastern boundaryregion of the Pacific, sea level does not
16 DECEMBER 1983
have direct biological sequelae, but thecool SST's, shallow surface mixed layer,and shallow thermocline result in highannual productivity at all trophic levelsof the ecosystem (7). Upwelling ecosys-tems appear not to differ qualitativelyfrom other marine ecosystems, but theydiffer quantitatively. It is this high pro-ductivity that El Niino disrupts.
There is a clear theoretical basis for adecrease in biological productivity dur-ing El Ninio. The theory has two causalaspects, one dealing with inorganic plantnutrients such as nitrate, phosphate, orsilicate and the other with the supply oflight for photosynthesis. The major inor-ganic nutrient reservoir of the ocean iswater below the thermocline (11); anyprocess that depresses the thermoclineaway from the surface layer, where there
Summary. Observations of the 1982-1983 El Nifio make it possible to relate theanomalous ocean conditions to specific biological responses. In October 1982upwelling ecosystems in the eastern equatorial Pacific began a series of transitionsfrom the normal highly productive condition to greatly reduced productivity. The highlyproductive condition had returned by July 1983. Nutrients, phytoplankton biomass,and primary productivity are clearly regulated by the physical changes of El Ninio.Evidence from 1982 and 1983 also suggests effects on higher organisms such as fish,seabirds, and marine mammals, but several more years of observation are required toaccurately determine the magnitude of the consequences on these higher trophiclevels.
There is ample evidence for year-to-year variations in biological productionalong the Peru coast (7, 8), but the envi-ronmental changes responsible for thesevariations are not known. During theequatorial warming event of 1975 (9),anomalously warm water west of theGalapagos Islands at 95°W was nutrient-rich (10), but primary productivity wasreduced about tenfold from values in the1960's at the same location and season.At the onset of the 1976 El Ninlo asudden bloom of the dinoflagellate phy-toplankter Gymnodinium splendens tookplace in the coastal waters along Peruwhen a sudden warming of the nutrient-rich surface layer increased the staticstability. Interest in the causality ofchanges in biological productivity duringthermal anomalies in the coastal andequatorial ocean led to a program inwhich the environment of plankton wasstudied in terms of both day-to-day andyear-to-year changes. Anecdotal ac-counts from 1972 and 1976 suggestedthat biological consequences of El Nifioshowed up almost simultaneously alongthe entire Peru coast, persisted formonths, and then rapidly disappeared.The program, begun in early 1982, wasdesigned to adequately evaluate such an
is enough light for photosynthesis, willnecessarily reduce productivity. Be-cause light decreases exponentially as afunction of depth, the depth of the sur-face mixed layer in which phytoplanktonare homogeneously distributed deter-mines the quantity of light that can becaptured by the phytoplankton (12). Ifthe mixed layer is deep, phytoplanktonspend a greater proportion of time in thedark and water molecules, not phyto-plankton, absorb most of the light. Notethat we have invoked two of the corre-lates of ocean thermal dynamics, ther-mocline depth and mixed layer depth,whose variability is an integral part of ElNinlo (6). Regions of the ocean, such asthe eastern tropical Pacific, having athermocline near the surface and a veryshallow mixed layer are biologically rich-er than other regions of the ocean (13)because large-scale thermal structureprovides the opportunity for enhancedsimultaneous capture of light and nutri-ents. This enhanced primary productionis reflected in all levels of the ecosytem
Richard T. Barber is professor in the Departmentof Zoology and the Department of Botany andFrancisco P. Chavez is a graduate student in theDepartment of Botany, Duke University, Beaufort,North Carolina 28516.
Large-scale trade winds blowingacross the Pacific from east to west setup the tilt in the thermocline that bringsits middle (the 20°C isotherm) to a favor-able, shallow depth of 40 m or less in theeastern boundary (1, 14). Mixed layerthinning in the eastern boundary currentis also a consequence of the large-scalezonal winds (6). Exploitation of thelarge-scale thermal structure involves aset of smaller scale physical processesthat act within the favorable, basin-widethermal structure of the Pacific but aresomewhat independent of it. These me-socale physical processes are set in mo-tion by meridional winds blowing towardthe equator along the coasts. The meridi-onal winds drive coastal upwelling; thisphenomenon, which takes place within50 km of the shore, transports waterfrom depths of 40 to 80 m to the surface(15). In a narrow band along the equator,wind-driven equatorial upwelling (16)provides the same final advective linkbetween the thermocline and the surfacelayer. Vertical mixing in the wake ofislands, shelf break upwelling, and geo-strophic upwelling resulting from currentshear all play ecological roles analogousto those of coastal and equatorial upwell-ing by providing local vertical transport
or mixing that can connect the nutrientpool with the light supply. But theselocal processes can enhance biologicalprocesses only if the large-scale thermalstructure is favorable in the sense thatnutrient-rich water is close to the sur-face.
In a 1974 description of El Ninlo,Wooster and Guillen (17) said that coast-al upwelling ceased or weakened duringthe events. This was a reasonable inter-pretation of the data at the time becausecool SST's were the major signature ofcoastal upwelling and this signature dis-appeared during El Ninlo. When coastalwind data became available it becameclear that in previous El Ninlo events (18)and the 1982-1983 event (19) the coastalwinds driving coastal upwelling did notweaken. In fact, coastal winds may in-tensify during El Ninlo because of in-creased thermal differences betweenland and sea (18). During El Ninlo itappears that coastal upwelling continuesbut that the water entrained is warmerand poorer in nutrients. As the thermno-cline is progressively depressed towardand below the depth of entrainment (40to 80 m), coastal upwelling (as well as theother mechanisms of local vertical fluxmentioned above) transports smaller andsmaller quantities of nutrients to the sur-face.
This conceptual model suggests thatEl Nifio affects the ecosystem by de-creasing the quantity of nutrients trans-ported to the surface, which in turncauses primary production of organicmaterial to decrease proportionally. Inaddition, the amount of light available toa phytoplankton population for the syn-thesis of organic material is decreased bya deepened mixed layer. In this mannerthe supply of both nutrients and light isreduced as El Nifio strengthens in inten-sity and the decrease in new primaryproduction available to the food chainafter some period of time causes propor-tional reductions in the growth and re-productive success of zooplankton, fish,birds, and marine mammals. Becausetemperature, nutrients, productivity,and food are tightly linked in upwellingecosystems, fish, seabirds, and marinemammals (20) have evolved behavioraladaptions that enable them to use tem-perature as an environmental cue to findareas of abundant food. Such behavior,of course, is disrupted by El Nifio, so theshort-term biological response of higherorganisms to the productivity anomaly ofEl Nihlo may be mediated through behav-ior responding to the thermal anomaly.Examples of both temperatures and foodresponses are evident in observations ofthe 1982-1983 El Nifio.
Fig. I (left). Chart of the eastern equatorial Pacific, showingthe location of the transects and observation sta-tions. Fig. 2 (right). (A) Triweekly SST measurementsmade 8 km offshore at the 100-m isobath at Paita (5004'S,81°15'W), temperature anomalies calculated by comparing c)the monthly averages of triweekly temperatures to the Talara(4035'S, 81017'W) 26-year monthly mean temperatures (1955to 1981), and monthly rainfall anomaly at Piura (5018'S,80035'W) calculated with the 26-year monthly mean rainfall(1955 to 1981). (B) Monthly catch of sardines, hake, jackmackerel, and shrimp along the northern coast of Peru(reported at Paita).
1204
16
12
01
00Z.I--8 ,
oc
4 coco
O
600EE
400 EE0
c0
200 cs.0
O
cI
-00 o
0Y0
_.0.5 co mE e)
0
0 C0E
Co-
I 40
3153 35 3. §3 5 3L 3- l;3 563. 5 3 .,5 K3
Sept. Oct. Nov. Dec. Jan. Feb. Mar. Apr. May June July Aug.
The anomaly arrived at Paita (Fig. 1)during the last week of September 1982as a 4°C SST increase in one 24-hourperiod (Fig. 2). The SST at the Paitaocean station, which is 8 km offshoreover the 100-m isobath, and at the pierwhere the historic Paita temperature rec-
ord is obtained (3), showed the samerapid change. About 400 km offshore(50S, 85°W), the mixed layer deepeneddramatically in the first week of October1982 without a very large change in SST(Fig. 3B). Comparison of October 1982with November 1981 shows the magni-tude of the mixed layer deepening thattook place in 1982 in the absence of a
significant change in coastal winds (19).The October 1982 profile shows that thelargest temperature anomaly associatedwith onset of El Niflo is 40 m below thesurface; this explains why the anomalyappears first at the coast (1), where localupwelling entrains the anomalouslywarm subsurface water and transports itto the surface.The mixed layer deepening seen in the
October 1982 profile could have de-creased primary productivity in the re-
gion, but nutrients remained favorablefor phytoplankton growth, with nitrateconcentrations over 4 FM. The relationbetween phytoplankton growth and am-
bient nutrient concentration is complex(21). Here it is sufficient to know thatwater with nitrate concentration of 4 FxMor more is nutrient-rich; that is, the up-
take versus concentration relation is sat-urated (21). Conversely, water with a
nitrate concentration of 0.1 ,uM or lessis nutrient-poor. Nitrate, silicate, andphosphate, the major inorganic nutrientanions (7), covary in the upper 100 m ofthe eastern equatorial Pacific, so nitrateis used as an index of the nutrient abun-dance of all the major nutrient anions.The difference in the nitrate concentra-tion in the water column between No-vember 1981 and October 1982 showsthat the processes normally keeping thesurface layer nutrient-rich were reducedin the first week of October 1982 but thatthe surface layer was not depleted ofnitrate.
It is tempting to speculate that theOctober 1982 to November 1982 progres-sion shows the interplay of remote andlocal processes. That is, a propagatedKelvin wave transiting from the equato-rial wave guide to the coastal wave guide(19) deepened the mixed layer by Octo-ber 1982; in the following month localheating increased the temperature of theupper 100 m and phytoplankton uptakestripped nutrients from the surface layer16 DECEMBER 1983
of the water column that was now isolat-ed from the nutrient-rich water of thethermocline. The trouble with this sce-nario is that it neglects large-scale hori-zontal water movements that were oc-curring around 5°S and 85°W in Octoberand November 1982. Drogued surfacebuoys (4) showed episodes of strong flowto the southeast in this region, and thesurface layer had salinities of less than 34parts per thousand, suggesting that thesurface water originated north of theequator and flowed into the region be-tween October and November 1982 (22).Further evidence of southward overflowof the Equatorial Front along 85°W canbe seen by comparing November 1981 toNovember 1982 in Fig. 3A. The Equato-rial Front was 2°S in November 1981,separating the cooler, nutrient-rich wa-ters of the Peru Current from the nutri-ent-depleted surface waters to the north.
51
10
E
cLcI*
50
100-
Temperature (OC)
O 'L ~ i .22\ 22 24
20
4 - 1~~18 16
'5
0 ('14
In November 1982 the front, as indicatedby the 24°C and 4 pM nitrate isopleths,was in the vicinity of 10°S, 800 to 900 kmsouth of the normal November location.On the basis of the October and Novem-ber 1982 observations we calculate thatthe Equatorial Front progressed south-ward at 16 km/day. The change fromOctober to November along 85°W proba-bly resulted from a combination of apropagating Kelvin wave (19), large-scale southward flow (4), and, to a lesserdegree, local heating and nutrient up-take. Figure 3 shows that, regardless ofthe processes involved, the ocean along85°W changed in October and November1982 from a nutrient-rich eastern bound-ary current condition to a nutrient-de-pleted condition typical of a centralocean gyre (23).Along the cross-equatorial transect at
Fig. 3. (A) Cross-equatorial profiles of temperature and nitrate along a transect at 85°W from2°N to 10°S. November 1981 shows normal conditions; October, onset of the anomaly. (B)Vertical temperature and nitrate profiles measured at 5°S, 85°W during November 1981,October 1982, and November 1982.
by the onset of El Nifio is clear; profilesfrom November 1979 and April 1982show presumably normal conditions dur-ing the cool season (October and No-vember) and the warm season (Marchand April) (24). During both seasons, andindeed throughout the annual cycle, theequatorial region at 95°W provides phy-toplankton with optimum nutrient andmixed layer conditions in that nitrateconcentrations are over 4 ,M and ther-mal stratification is strong in the upper50 m (that is, there is a very shallow ornonexistent mixed layer). The positionof the Equatorial Front at 95°W is shownby the 24°C isotherm in November andthe 26°C isotherm in April 1982. In No-vember 1982, after the onset of theanomaly, thermal stratification was ab-sent in the upper 75 m and the equatorialband from 1°N to 4°S was nutrient-rich(less than 8 FM but greater than 4 FMnitrate) and relatively warm (less than
26°C but greater than 24°C). The chloro-phyll proffies in Fig. 4 show that phyto-plankton biomass was low in November1982. A deepened mixed layer wouldreduce primary production, but the largedecrease in static stability of the layeralso may have enhanced diffusive andsinking losses of nonmotile phytoplank-ton such as diatoms. By March 1983 theanomaly was peaking; no surface signa-ture of equatorial upwelling was presentalong the 95W transect. The region at95°W showed nutrient depletion and re-ductions in specific photosynthetic activ-ity and the absolute quantity of primaryproductivity (Table 1). While the tran-sect along 95°W showed a fivefold reduc-tion in absolute productivity, a cross-equatorial transect along 92°W (Fig. 1)close to the Galapagos Islands showed a20-fold decrease (Fig. 5 and Table 1) inMarch 1983 at the peak of the 1982-1983El Nifio. Ocean waters around islands
Temperature (°C) Nitrate (aM) Chlorophyll (jLg/liter)
-2 5
26
'-5
'2 1 26' \2726 26\
24
16
26
~20>~-18
AY0.1
-~~-
12
8.
16
t87
20 8
*
02
. @.....
ICO.C04 ... .4. .
0
z
Fig. 4. Cross-equatorialC profiles of temperature,o nitrate, and chlorophyll a
along a transect at 95°W. from 2°N to 5°S. Novem-
ber 1979 and April 1982show normal conditionsfor those months. No-vember 1982 is during on-
set of the anomaly at95°W and March 1983 isduring the peak of theanomaly. In all profiles
cN temperature was mea-a sured as a continuous
vertical record; discreteo nitrate and chlorophyllZ observations were made
at the depths shown bythe dots.
00oco
Latitude on 950W
and coasts have inherently higher pro-ductivity than waters far removed fromland, so the proportional reduction by ElNifio was greater in the island (92°Wtransect) and coastal (5°S transect) wa-ters (Fig. 5 and Table 1). The phyto-plankton species (25) showed surprising-ly little taxonomic change between nor-mal (26) and El Ninlo conditions on the92°W transect, but there were fewer dia-toms and more microflagellates duringthe peak of the anomaly.The cross-equatorial transect at 95°W
shows onset of the event in November1982 (depressed thermocline, deepenedmixed layer, nutrient richness) and peakconditions (29°C SST, relatively strongstratification, nutrient depletion) inMarch 1983, but we do not know wheth-er the change from onset to peak wasrapid or gradual. Along the coast, devel-opment of the anomaly had a 5-monthmaturation phase between onset in Octo-ber 1982 and the peak in May 1983.Figure 2 illustrates development of theSST anomaly: a rapid rise in late Sep-tember, a slower but steady increasefrom October through March from 22°Cto 28°C, and a plateau at 29°C. Thesequence of transects along 5°S normalto the coastline (Fig. 6) shows the tempo-ral and spatial development at the off-shore and equatorward anomaly progres-sively moves inshore and poleward.
In November 1982 there was deepen-ing of the thermocline and the appear-ance of a layer of warm, low-salinity,low-nutrient, and low-chlorophyll waterat 40 m along the transect from 85°W to aposition at about 82°W, or 150 km off thecoast (Fig. 6). Coastal upwelling contin-ued to supply nutrients to the surfacelayer in a band next to the coast, but thenarrow 30- to 50-km band of enrichmentin November 1982 contrasts sharply tothe 400-km-wide region of high nutrientsand chlorophyll present in November1981. During November 1982 chloro-phyll concentrations all across the tran-sect were lower than in November 1981;however, stations within 30 km of thecoast during November 1982 measured 1to 10 p,g of chlorophyll per liter in abloom dominated by a diatom typical ofcoastal upwelling, Asterionella japonica(26).
Conditions in March 1983 were likethose in November 1982 except forwarmer SST's and a still narrower bandof coastal enrichment (Fig.. 6). Theseincreases, which are also evident in thePaita ocean station time series (Fig. 2),had begun in late September 1982. InMarch 1983 coastal upwelling still sup-plied a narrow inshore band with nutri-ents, and this band remained rich inphytoplankton, as shown by cross-shelf
profiles of chlorophyll. The coherence ofthe nitrate and chlorophyll profiles inFig. 6 shows that, to a first approxima-tion, the spatial distribution of phyto-plankton biomass in this ecosystem isdetermined by advective supply of newnutrients to the surface layer. A progres-sive decrease in the area of productiveinshore habitat started in November1982 and continued through March 1983,but as the size of the productive habitatdecreased the concentration of phyto-plankton biomass, as reflected by chlo-rophyll concentration, remained remark-ably high. Figure 6 shows that chloro-phyll in the extreme inshore area inMarch 1983 ranged from 1 to 6 ,ug/liter,concentrations that characterize thecoastal upwelling habitat during normalconditions (7, 8).The May 1983 profiles in Fig. 6 show a
50-m-deep layer of 29°C, nutrient-deplet-ed water against the coast and maximumexpression of the physical and biologicalanomalies of the 1982-1983 El Niino. Thethermocline is deep with the 20°C iso-therm at 150 m, isotherms tilt downtoward the coast, and a 50-m-deep mixedlayer has very low nutrient concentra-tions, low phytoplankton biomass, andlow productivity (Fig. 5 and Table 1).
Transects farther south along 10°30'Sindicated that the anomalous conditionswere delayed and somewhat reduced inintensity off central Peru compared withthe northern coast around 5°S. In No-vember 1982 water characteristic of theregion north of the Equatorial Front hadnot progressed southward to 10°30'S;surface layer concentrations of nitratewere between 4 and 8 p.M._ By March1983 the anomaly was present at 10'30'Sand reached to within 50 km of the coast.Comparison of these results with thosefrom 5SS establishes that the anomalywas less intense father from the equator.This supports the concept that in anoceanographic sense El Nifio is an equa-torial phenomenon that propagates pole-ward and progressively weakens alongthe coast (5, 14, 19, 27).Recovery from the 1982-1983 anomaly
started at the Paita ocean station early inJuly 1983 (Fig. 2). By 15 July 1983 thetemperature had decreased to 20°C; atransect along 5°S (Fig. 6) showed asimultaneous return to normal condi-tions in a 200-km-wide band next to thecoast. Figure 5 shows that the 200-kmband contained nitrate concentrations of4 to 16 ,uM and chlorophyll concentra-tions of I to 20 p.g/liter. The speed (Fig.2), spatial extent (Fig. 6), and intensity(Fig. 5) of the recovery in nutrient levels,phytoplankton biomass, and primaryproduction along the 5°S transect wasunexpected. During July and August
Table 1. Mean surface nitrate, chlorophyll, and primary productivity during normal and El Nintoconditions on transects at 95°W, 92°W, and 5°S (Fig. 1). Values were calculated by integratingthe values along the three transects (the space under the curves of Fig. 6) and dividing by thelength of the transect.
Nitrate Chloro- grams ofTransect Date (mmole/ phyll a of carbon
mi3) (mg/m3) carbon per milli-per gram of
cubic chloro-meter phyll
per day) per hour)
Equator at 95°W; April 1982 (normal) 5.3 0.22 15.6 7.02°N to 2°S March 1983 (El Ninlo) 0.1 0.16 3.0 1.8
Ratio (April 1982 to 53.0 1.4 5.2 3.9March 1983)
Equator at 92°W; April 1966 (normal) 7.8 0.57 77.8 13.62°N to 2°S March 1983 (El Nifno) 0.1 0.17 3.9 2.3
Ratio (April 1966 to 78.0 3.4 20.0 5.9March 1983)
Coast at 5°S; July 1983 (normal) 3.0 4.44 219.3 4.981°15'W to 85°W May 1983 (El Ninto) 0.1 0.21 10.3 4.9
Ratio (July 1983 to 30.0 20.9 21.3 1.0May 1983)
Latitude on 95°W20S 1 0 1 2°N
30 A-x. X-X/X-X-Mr-X
1 *- March 83
0
I-0
0)n)
25
20
30
April 82
25
20
is
a 10-3 l-
s- Anl8
March 83o- X-X-X-x-x-X
1I
0
z
a21
mVL
0.5-
Latitude on 92°W2°S 1 0 1 2°N
March 83
April 66
15-
10-April66
5-
March 830 -x-x x x
2
1
March 83
oJ100,
X-O0%E E
t0 -
E. 0
IL E
50-
0-
April 82
| L March 83
2°S 1 0 1 2°NLatitude on 95°W
00200-
100
0
30
25
Longitude on 50S85 84 83 82 81sW
I
- -
x-x May 83
il J
July 3
20'15- A
10- July 83
5-
May 83O- it X _x -X-X,'
201
10
O0
April 66
March 83X --. nx20S 1 0 1 2°N
Latitude on 92°W
1000
500
0
July83
May 83/
xx-x-
/July 83
May 83/
400 200 0
Kilometers offshore
Fig. 5. Comparison of SST, nitrate, chlorophyll a, and primary production along the 95°W,92°W, and 5°S transects during normal conditions and during the peak of the anomaly whenbiological productivity was affected most.
1983 the offshore water beyond 200 kmremained anomalously warm, nutrient-depleted, and low in productivity. Re-covery of the upwelling ecosystem start-ed next to the coast and progressedoffshore. Still, even the nutrient-rich andhighly productive water in the 200-kmband along the coast remained anoma-lously warm in July and August; satelliteobservations of the large-scale tempera-ture field off the coast of Peru and Ecua-dor for those months (28) gave no indica-tion of the large-scale recovery in pri-mary productivity that was taking place.
Effects on Higher Trophic Levels
Interannual variability in SST alongthe coasts of Ecuador and Peru and inthe Galapagos Islands is well knownbecause of the association of warmanomalies with reductions in fish (Fig.7). Most investigators of the effects of ElNinlo on fish, particularly Peruvian sci-entists most familiar with the phenome-non (8, 29), believe that reductions infish abundance are caused by decreasesin primary productivity that affect theentire food web. Evidence for a causal
Temperature (°C) Nitrate (.uM) Chlorophyll (jig/liter)
1
0,
50
1OO,
6
a
0-
so
100
26
$24
222
N20k\
29
26
22
0.~~~-05
12
20
01, 04,lo.0.2 o
i*.
2J
0.L
0
z
N
0
zFig. 6. Cross-shelfprofiles of tempera-ture, nitrate, andchlorophyll a along a5°S transect from85°W to Paita. No-vember 1981 shows
co the normal condi-tions; November 1982is during onset;March 1983 is duringmaturation; May 1983is during the peak ofthe anomaly; and July1983 shows recoveryof normal conditionsin the nearshore up-welling cell and per-sistence of the anom-
aly farther offshore.
CD)
0
Kilometers offshore at 5°S
1208
relation between primary productivityand changes in anchovy abundance hasbeen fragmentary because a completecycle of productivity changes during ElNinlo was not observed along with theconsequences of the perturbation athigher trophic levels. What was missingbefore 1982 were time series of nutrientsand productivity to compare with theexcellent time series of temperature, fishcatch, and seabird abundance obtainedin the past (29).There have been other views on how
El Ninlo affects fish. For example, adecade ago it was suggested that thePeruvian anchovy (Engraulis ringens)simply swam to deeper, cooler water andthat the missing fish would return in fullabundance after the warm anomalypassed (30). In the 1970's, however,multi-ship acoustic surveys of Peruviancoastal waters showed convincingly thatspatial redistribution does not explainhow El Ninlo affects the anchovy (31).For most species of the higher trophiclevels we believe that the major effect ofthe 1982-1983 El Nifio was an absolutedecrease in growth and reproductivesuccess caused by disruption of the nor-mal food web. The specific change was a5- to 20-fold reduction of primary pro-ductivity (Table 1) as this normally eu-trophic region changed to one with anoligotrophic character typical of a cen-tral ocean gyre (23).To our knowledge the first report of a
biological response to the 1982-1983 ElNinlo was the reproductive failure of sea-birds on Christmas Island (2°N, 157°W).In June the nesting of the blue-facedbooby (Sula dactylatra) was proceedingnormally with many adult birds, nest-lings, and eggs present on the island; byNovember 1982 Schreiber and Schreiber(32) found only three adult birds and oneunderweight nestling. Equally large re-productive failures were observed formost of the species of sea birds that neston Christmas Island. For example, thenumber of great frigate birds (Fregataminor) declined from 20,000 in June 1982to fewer than 100 in November 1982.Rains and flooding may have killed somenestlings, but Schreiber and Schreiberstated that disappearance of small fishand squid caused the adult birds to aban-don the island. Such abandonment ofnests and young at the onset of El Nihlowas observed frequently in the 1982-1983 event as well as in earlier events(33). The spatial extent of this behaviorin 1982-1983 was from the central Pacificto 15°S on the Peru coast. Nest abandon-ment and widespread dispersal from thenest sites is appropriate evolutionary be-havior for these relatively long-lived
birds because it allows the adults to useavailable food for survival rather thanreproduction. Adult seabirds along thecoast of Peru survived through March1983 without detectable mortality, but byApril there were reports of many deadadults on beaches throughout the region.It appears that, in addition to causingfailure of the class of 1982, the eventsignificantly reduced the adult popula-tion because of its intensity and dura-tion.
Seabirds responded rapidly to alteredconditions, but marine mammals (34)provide the clearest evidence for thefood stress that accompanies El Ninlo. In1982 and 1983 the Galapagos fur seal(Arctocephalus galapagoensis) and theCalifornia sea lion (Zalophus califor-nianus) were studied in the GalapagosIslands and the South American fur seal(Arctocephalus australis) and sea lion(Otaria byronia) were studied at PuntaSan Juan near 15°S on the southern coastof Peru. Limberger et al. (34) reportedthat all the Galapagos fur seal pups bornin 1982 had died by March 1983 and thatmany of the pups and some of the adultsof the other three species had died. Thepup mortality may have resulted fromlack of foraging success by the adultfemales (34); females stayed at sea for anaverage of 5 days rather than the typical1.5 days, and when they returned toshore they were unable to provideenough milk to prevent pups from starv-ing. Apparently the fish and squid thatthese marine mammals require were notavailable in adequate numbers betweenNovember 1982 and March 1983, eithernear the Galapagos or off Peru. In addi-tion to the loss of the individuals born in1982, it appears that the adult populationhas been reduced (34).
Figure 2B shows changes in the catchof several commercially important spe-cies in the northern Peru region. Catch isnot synonymous with abundance, but,according to the fisheries biologists whoprovided the data, it accurately reflectsthe presence or absence of these speciesbecause from November 1982 to August1983 the fishing fleet continuously trieda variety of techniques at locationsthroughout the region to catch whateverthey could. Changes in catch were close-ly related to the timing and intensity ofenvironmental changes in the differenthabitats of hake, shrimp, mackerel, andsardines. The first species to respondwas hake (Merluccius gayi). Apparently,these relatively large and motile bottom-dwelling fish moved down the continen-tal slope, staying with the cool water towhich they are adapted. In November,1982, when the 18°C isotherm moved16 DECEMBER 1983
down the shelf to below 100 m andsubsurface warming was observed in themoored current meter array (19), thehake catch slowly decreased to zero(Fig. 2); when 18°C water returned to theshelf break in July 1983 a few fish reap-peared in the catch.The increase in shrimp catches at Paita
shown in Fig. 2 is consistent with theidea that the altered currents of El Ninloredistributed shrimp throughout the re-gion (35). The catch of the three majorspecies of shrimp (Xiphopenaeus riueti,Penaeus occidentalis, and Trachypen-aeus byesi) was normal on the PacificCoast of Colombia from January 1982 toSeptember 1982, then decreased sharplyin the last quarter of 1982. The shrimpcatch along Ecuador was normal throughFebruary 1983, while the catch off north-ern Peru, which is usually very low,increased as the temperature increasedfrom November 1982 to February 1983(Fig. 2). The southward shift of shrimpabundance reflects the somewhat plank-tonic character of these organisms; theywere carried southward by the inshore,southward-flowing warm current charac-teristic of El Ninlo (6). Shrimp and hakehad no observed mortality in connectionwith the anomaly, but not all demersalspecies avoided mortality. Between De-cember 1982 and February 1983 largenumbers of dead corbina (Cynoscionxanthulus) were found floating off north-ern Peru (35). Cross-shelf temperatureprofiles (Fig. 6) show that by March 1983the habitat of this demersal species hadincreased in temperature from about 160to 24°C.The sardine (Sardinops sagax) is in the
ac00-
2m0
E-
x
0a
1980's an important resource for Ecua-dor, Peru, and Chile; Fig. 7 shows theincrease in this fishery in the past fewyears in Peru to an annual catch of over 1million metric tons. In January 1983 sar-dines disappeared from the Ecuadoriancoast, did not change in abundance alongthe Peru coast, and greatly increasedalong the coast of Chile. Maintenance ofthe sardine catch off northern Peruthrough March 1983 (Fig. 2) can be ac-counted for by the cross-shelf profilesfrom 50S (Fig. 6). From November 1982to March 1983 sardines remained in thedwindling band along the coast wherechlorophyll concentrations were over 1,ug/liter. Chlorophyll is an index of phy-toplankton abundance but not of zoo-plankton, the major food of sardines.However, because phytoplankton arethe major food of zooplankton, chloro-phyll concentration does delimit the hab-itat of sardines. As adult sardines con-centrated in the nearshore region theywere extremely vulnerable to fishermen,and their physiological condition wasdeteriorating, as shown by the decreaseof their oil content to less than 1 percentby weight (35). In mid-April 1983 (Fig. 2)sardines disappeared from the northernPeru catch; in confirmation, the May1983 profiles in Fig. 6 show that the richhabitat containing more than 1 ,ug ofchlorophyll per liter was virtually elimi-nated from the region.Jack mackerel (Trachurus symmetri-
cus) disappeared from the catch in De-cember 1982 (Fig. 2), between the timeswhen hake and sardines left the coastalregion. This timing is in accordance withthe alteration of the habitat of the three
0
0
U2z
0aCI.-00
1955 1960 1965 1970 1975 1980 1985Year
Fig. 7. Association of annual SST anomalies with annual catch of anchovies and sardines offPeru. The anomaly temperature scale is inverted: upwards indicates cooler temperatures;downwards indicates warmer temperatures-that is, El Ninlo. The annual temperature anomalyis relative to the 26-year mean temperature of 16.9°C at Chicama; it is calculated for the thermalyear from July to June spanning half oftwo calendar years, and the value is plotted in the middleof the thermal year at January. The fish catch for each calendar year is plotted in the middle ofthe calendar year at July.
species by El Nifio. The subsurfacewarming (19) and thermocline depres-sion (4) of the onset forced hake to movefirst. Jack mackerel occupy the offshoreboundary of upwelling circulation,where their euphausid food is most abun-dant (7). The warm, nutrient-poor, phy-toplankton-poor water seen offshore inthe November 1982 profiles is a habitatin which jack mackerel cannot survive,so this species was crowded into theinner 30-km band by the offshore-to-onshore progression of anomalous con-ditions. At this time predatory speciesthat eat jack mackerel-bonito (Sardachiliensis), dorado (Coryphaena hip-purus), and yellowfin tuna (Thunnus al-bacores)-became more abundant closeto the coast (35). Increasingly predationby these oceanic species, which are tol-erant of warm temperatures, on the con-centrated jack mackerel may have con-tributed to its disappearance in Decem-ber 1982.The most impressive biological conse-
quence of El Nidlo is the effect on thePeruvian anchovy (Engraulis ringens),once the basis of the world's largestfishery (29). Figure 7 shows the covaria-tion of thermal conditions and anchovyharvest. We believe that this relation iscausal and that the causality depends onincreased heat storage depressing thethermocline and increasing the depth ofthe mixed layer. As shown in Figs. 3, 4,and 6 and Table 1, thermocline depres-sion and mixed layer deepening are al-ways accompanied by reductions in theproductivity and biomass of phytoplank-ton. The 20-fold decrease in phytoplank-ton biomass and productivity that tookplace along the coastal transect in 1983will decrease the growth, survival, andparticularly reproductive fitness of theadult anchovy. Because larval survival isespecially dependent on the availabilityof phytoplankton (36), the reproductivesuccess of the species will be impaired.In combination these processes will re-duce the anchovy stock to a record lowlevel.
The coastal region around Paita wasclosed to anchovy fishing in 1982, sothere is no information on the catch ofthis species to relate to the local changesin physical conditions and productivityon the 5°S transect. Observations madeby the Instituto del Mar del Peru (35)confirm that the anomaly has had aneffect. During December 1982 to Febru-ary 1983 anchovy schools were concen-trated in small pockets next to the coastand numerous dead adult anchovieswere seen on the surface at several loca-tions. This observation is significant be-cause mortality of adults was not ob-served during the 1972 or 1976 El Nifioevents (35). It appears that for the ancho-vy, as for seabirds and marine mammals,the 1982-1983 El Ninlo profoundly affect-ed both reproduction and adult survival.
References and Notes
I. E. M. Rasmusson and J. M. Wallace, Science222, 1195 (1983).
2. S. Zuta and W. Urquizo, Bol. Inst. Mar Peru 8,461 (1972).
3. F. Vasquez and C. Raul Castillo, Estud. Reg.Fen6m. El Nifio Bol. 3. 9 (1983) (available fromComisi6n Permanente del Pacifico Sur, Quito,Ecuador).
4. D. Halpern, S. P. Hayes, A. Leetmaa, D. V.Hansen, S. G. H. Philander, Science 221, 1173(1983).
5. K. Wyrtki, J. Phys. Oceanogr. 5, 572 (1975);ibid. 9, 1223 (1979); Mar. Technol. Soc. J. 16, 3(1982). (The last of these is a nontechnicaldiscussion of El Nifio with figures showing theeffect of large-scale winds on sea level, thermo-cline tilt, and mixed layer thickness.)
6. M. A. Cane, Science 222, 1189 (1983).7. R. T. Barber and R. L. Smith, in Analysis of
Marine Ecosystems, A. R. Longhurst, Ed. (Aca-demic Press, New York, 1981), p. 31.
8. 0. Guillen, in Fertility ofthe Sea, J. D. Costlow,Ed. (Gordon & Breach, New York, 1971), p.187; and R. Z. Calienes, in ResourceManagement and Environmental Uncertainty:Lessons from Coastal Upwelling Fisheries, M.H. Glantz and J. D. Thompson, Eds. (Wiley,New York, 1981), p. 255.
9. K. Wyrtki, E. Stroup, W. Patzert, R. Williams,W. Quinn, Science 191, 343 (1976).
10. T. J. Cowles, R. T. Barber, 0. Guillen, ibid.195, 285 (1977).
11. A. C. Redfield, Am. Sci. 46, 205 (1958); R. C.Dugdale, Limnol. Oceanogr. 12, 685 (1967).
12. H. U. Sverdrup, J. Cons. Cons. Int. Explor.Mer. 18, 287 (1953).
13. J. H. Ryther, Science 166, 72 (1969); R. C.Dugdale, Geoforum 11, 42 (1972).
14. A. J. Busalacchi, K. Takenchi, J. J. O'Brien, inHydrodynamics of the Equatorial Ocean, J. C.J. Nihoul, Ed. (Elsevier, New York, 1983), p.155.
15. Figure 5 of R. T. Barber and R. L. Smith (7)shows the depth of entrainment in three coastal
upwelling ecosystems, as determined by densevertical arrays of current meters.
16. M. Vinogradov, in Analysis of Marine Ecosys-tems, A. R. Longhurst, Ed. (Academic Press,New York, 1981), p. 69.
17. W. S. Wooster and 0. Guillen, J. Mar. Res. 32,387 (1974).
18. D. B. Enfield, J. Geophys. Res. 86, 2005 (1981).19. R. L. Smith, Science 221, 1397 (1983).20. J. J. Walsh, ibid. 176, 969 (1972).21. J. J. McCarthy and J. C. Goldman [ibid. 203, 670
(1979)1; R. C. Dugdale, B. H. Jones, Jr., J. J.MacIsaac, and J. J. Goering [Can. Bull. Fish.Aquat. Sci. 211, 234 (1981)] and R. W. Eppley(ibid., p. 251) provide entree to the topic ofnutrient regulation of phytoplankton growth.
22. W. S. Wooster, Deep-Sea Res. 16, 407 (1969).23. M. Blackburn, in Analysis of Marine Ecosys-
tems, A. R. Longhurst, Ed. (Academic Press,New York, 1981), p. 1.
24. R. T. Barber, S. Zuta, J. Kogelschatz, F. Cha-vez, Trop. Ocean-Atmos. Newsl. No. 16 (1983),p. 15.
25. Phytoplankton were identified and counted byT. Arcos of Instituto Nacional de Pesca, Guaya-quil, Ecuador, for 1983 and by E. Hulburt ofWoods Hole Oceanographic Institution, WoodsHole, Mass., for 1966.
26. R. Jimenez [in Coastal Upwelling, F. A. Rich-ards, Ed. (American Geophysical Union, Wash-ington, D.C., 1981), p. 327] and B. Rojas deMendiola (ibid., p. 328) describe the phyto-plankton species normally found in Ecuadorianand Peruvian waters.
27. D. B. Enfield and J. S. Allen, J. Phys. Ocean-ogr. 10, 557 (1980).
28. R. A. Kerr, Science 221, 940 (1983).29. 0. Guillen, R. Z. Calienes, R. I. de Rondan, Bol.
Inst. Mar Peru 2. 49 (1969); G. J. E. Valdivia,Rapp. P. V. Reun. Cons. Int. Explor. Mer. 197,196 (1978); D. H. Cushing, in Upwelling Ecosys-tems, F. A. Richards, Ed. (American Geophysi-cal Union, Washington, D.C., 1981), p. 449; S.Zuta and 0. Guillen, Bol. Inst. Mar Peru 2, 157(1970); A. Bakun and R. H. Parrish, Calif.Coop. Oceanic Fish. Invest. Rep. 23, 99 (1982).
30. G. Murphy, Geoforum 11, 63 (1972).31. K. Johannesson and R. Vilchez, in Meeting on
Hydroacoustical Methods for the Estimation ofMarine Fish Populations, J. B. Suomala, Ed.(Charles Stark Draper Laboratory, Inc., Cam-bridge, Mass., 1981), p. 765.
32. R. W. Schreiber and E. A. Schreiber, Trop.Ocean-Atmos. Newsl. No. 16 (1983), p. 10.
33. P. D. Boersma, Science 200, 1481 (1978); R. C.Murphy, Oceanic Birds of South America(American Museum of Natural History, NewYork, 1936), p. 1210.
34. D. Limberger, F. Trillmich, G. L. Kooyman, P.Majluf, Trop. Ocean-Atmos. Newvsl. No. 21(1983), p. 16.
35. J. Valdivia, R. Jimenez, S. Avaria, 0. Mora,Informe de la Tercera Reunion del Comite6 Cien-t(fico del Estudia Regional del Fen6meno ElNiuo (Comisi6n Permanente del Pacifico Sur,Quito, Ecuador, 1983), p. 11.
36. R. Lasker, Rapp. P. V. Reun. Cons. Int. Explor.Mer. 174, 212 (1978).
37. Shiptime was provided by the Equatorial PacificOcean Climate Study of the National Oceano-graphic and Atmospheric Administration and bythe National Science Foundation. Research sup-port was provided by NSF grant OCE-8110702.The authors acknowledge the generous supportprovided by the governments of Peru and Ecua-dor for this work and the valuable assistance ofthe staff of the Paita laboratory of Instituto delMar del Peru. We thank J. E. Kogelschatz, V.Thayer, and J. C. Ramus for their contributions.
