landry, michael r., et al. iron and grazing constraints on … · 2015. 7. 29. · oceanography may...

LIMNOLOGY AND

OCEANOGRAPHY May 1997

Volume 42

Number 3

Limnol. Oceanogr., 42(J), 1997.405-41s 0 1997, by the American Society of Limnology and Oceanography, Inc

Iron and grazing constraints on primary production in the central equatorial Pacific: An EqPac synthesis Michael R. Lundry Department of Oceanography, University of Hawaii at Manoa, 1000 Pope Rd., Honolulu, Hawaii 96822

Richard T. Barber Duke University Marine Laboratory, Beaufort, North Carolina 285 16

Robert R. Bid&are Department of Oceanography, University of Hawaii at Manoa

Fei Chai Department of Oceanography, 5741 Libby Hall, University of Maine, Orono 04469-5741

Kenneth H. Coale Moss Landing Marine Laboratories, PO. Box 450, Moss Landing, California 95039-0450

Hans G. Dam Department of Marine Sciences, University of Connecticut, Groton 06340-6097

Marlon R. Lewis Department of Oceanography, Dalhousie University, Halifax, Nova Scotia B3H 451

Steven T. Lindley Carnegie Institution of Washington, Department of Plant Biology, Stanford, California 94305

James J. McCarthy Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138

Michael R. Roman and Diane K. Stoecker Horn Point Environmental Laboratory, University of Maryland, PO. Box 775, Cambridge 21613

Peter G. Verity Skidaway Institute of Oceanography, 10 Ocean Science Circle, Savannah, Georgia 31411

Jacques R. White College of Marine Studies, University of Delaware, Lewes 19958-1298

Abstract Recent studies in the central equatorial Pacific allow a comprehensive assessment of phytoplankton regulation in a

high-nutrient, low-chlorophyll (HNLC) ecosystem. Elemental iron enters the euphotic zone principally via upwelling and is present at concentrations (530 PM) well below the estimated half-saturation constant (120 PM) for the large cells that bloom with iron enrichment. In addition, the meridional trend in quantum yield of photosynthesis suggests that even the dominant small phytoplankton are held below their physiological potential by iron deficiency. Grazing by microzooplankton dominates phytoplankton losses, accounting for virtually all of the measured phytoplankton production during El Nina conditions and -66% during normal upwelling conditions, with mesozooplankton grazing and lateral advection closing the balance. Nitrate uptake is strongly correlated with the pigment biomass of diatoms, which increase in relative abundance during normal upwelling conditions. Nonetheless, the f-ratio remains low (0.07-0.12) under all conditions. Iron budgets are consistent with the notions that new production is determined by the rate of new iron input to the system while total production depends on efficient iron recycling by grazers. Although the limiting substrates differ, the interactions of resource limitation and grazing in HNLC regions are conceptually similar to the generally accepted view for oligotrophic subtropical regions. In both systems, small dominant phytoplankton grow at rapid, but usually less than physiologically maximal, rates; they are cropped to low stable abundances by microzooplankton; and their sustained high rates of growth depend on the remineralized by-products of grazing.

405

406 Landry et al.

Vast expanses of the global oceans, including most of the Grazing studies during the SUPER (Subarctic Pacific Eco- open ocean Pacific, are characterized by low and relatively system Research) Program revealed, however, that copepods stable standing stocks of phytoplankton, the dominance of exerted only a small grazing impact on phytoplankton (Frost small species, and the scarcity of diatoms that give the 1987, 1991., Landry and Lehner-Fournier 1988; Dagg 1993), plankton communities of nutrient-rich coastal ecosystems but grazing by small protistan “microzooplankton” fre- their classic seasonal blooms. In subtropical regions, ex- quently exceeded the growth rates of bulk phytoplankton tremely low levels of major nutrients and reduced seasonal chlorophyll as well as the rates for individual populations of variability in physical forcing (e.g. light and deep mixing) pica- and nanoplankton (Strom and Welschmeyer 199 1; provide sufficient explanation for such conditions. The ex- Welschmeyer et al. 1991; Landry et al. 1993b). The synthe- tent of the nutrient effect on phytoplankton growth rate and sis of resulls from the SUPER Program (Miller et al. 1991) the relative importance of nitrogen and phosphorus have emphasized that small ecologically dominant phytoplankton been issues of debate for these oligotrophic systems (e.g. in the subarctic Pacific grew at high rates with no obvious Goldman et al. 1976; Sharp et al. 1980; Karl et al. 1995), evidence of substrate limitation, that they were cropped ef- but the basic point of nutrient regulation has long been rec- fectively and maintained at low levels by the high grazing ognized. In contrast, after nearly four decades of speculation, rates and population growth potential of protistan consum- an entirely acceptable understanding is only beginning to ers, and that they depended for most of their growth on emerge for other open-ocean regions-including the subarc- remineralized ammonium, which suppressed the uptake of tic Pacific, the equatorial Pacific, and the Southern Ocean- nitrate (Wbteler and Kokkinakis 1990). Nonetheless, the rar- where phytoplankton stocks remain low despite high near- ity of large bloom-forming diatoms, which could escape the surface concentrations of major nutrients. Advances in food- combined grazing pressure of micro- and mesozooplankton web paradigms and measurement techniques have only re- (Landry et al. 1993a), was not adequately explained by the cently made possible a critical examination of the mecha- SUPER grazing hypothesis. Thus, some form of substrate nisms of phytoplankton control in these high-nutrient, low- (e.g. iron) limitation appeared to be needed to keep them in chlorophyll (HNLC) regions. check.

One of the goals of the U.S. JGOFS EqPac Program was to understand how standing stocks and production rates of phytoplankton were constrained to the low levels observed in the high-nutrient upwelling core of the central equatorial Pacific (Murray et al. 1992). Two main hypotheses-grazing control and iron limitation-provided the context for this effort. The origins of these hypotheses go back many years in the oceanographic literature, but both were revitalized in studies of the subarctic Pacific during the 198Os, emerging as competing explanations for the HNLC conditions in that area.

As originally developed by Russian and Canadian ocean- ographers, the grazing hypothesis attributed the lack of phytoplankton blooms in the subarctic Pacific to large endemic copepods of the genus Neocalanus (Beklemishev 1957; Mc- Allister et al. 1960; Heinrich 1962; Parsons et al. 1966).

Acknowledgments The success of the EqPac Program was due in large measure to

the efforts of J. Murray, who coordinated the shipboard scientific plan and subsequent data workshops, and to the superb logistical support of S. Kadar and the captain and crew of the RV Thomas G. Thompson. We also acknowledge the efforts of the many co- workers who contributed to the collection and analysis of the data presented. We note, in particular, J. Murray’s role in stimulating the discussion that led to this synthesis, and D. Foley for preparing Fig. 6. Lastly, the manuscript benefitted significantly from the construc- tive comments of K. Banse and J. Cullen.

Research support was provided by National Science Foundation grants OCE 90-22117,93-l 1246, and 93-15432 (M.R.L.), 90-22321 (R.R.B.), 90-24373 (R.T.B.), 88-13565 (K.H.C.), 90-22418 (H.G.D.), 90-22411 (J.J.M.), 90-24381 (M.R.R.), and 90-22318 (D.K.S. and P.G.V.), and by Office of Naval Research grant NO00 14-84-C-0619 (K.H.C.).

Contributions 4172 from the School of Ocean and Earth Science and Technology, University of Hawaii at Manoa and 2731 from CEES, University of Maryland. JGOFS contribution 236.

More than half a century after iron was first considered a possible limiting factor for ocean production (e.g. Gran 193 1; Hart 1934; Harvey 1938), the late John Martin and co-workers developed sensitive methods for the analyses of ambient oplen-ocean concentrations and proposed iron deficiency as an explanation for conditions in the subarctic Pa- cific and other HNLC regions (Martin and Fitzwater 1988; Martin and Gordon 1988; Martin 1990; Martin et al. 1990, 1991, 1994). The iron hypothesis was initially cast in terms of new production, with iron deficiency limiting the utilization of e:tcess surface nutrients and the drawdown of atmospheric CO, (Martin 1990). More recently, the hypothesis has been broadened to include the limitation of phytoplankton growth rate and biomass (Coale et al. 1996a; see also Cullen 199.‘:). At the time that the EqPac Program was being developed, the iron hypothesis was supported by two lines of evidence: extremely low (subnanomolar) concentrations of iron in the euphotic zones of HNLC regions relative to the requirements for phytoplankton growth and the availability of major nutrients (Martin and Gordon 1988; Martin et al. 1990:) and increases in phytoplankton biomass (chlorophyll and large species) and the rate of nitrate utilization in bottles with added iron (Martin and Fitzwater 1988; Coale 1991). Alt.hough these “grow-out” experiments clearly showed iron effects, the artificial circumstances of bottle incubations prompted concerns about “controls” that frequent- ly exhibited responses similar to iron treatments, about significant lag times before treatment effects could be observed, about the atypical nature of species responding to iron, and about possible effects of bottle treatments on the grazer community, including the exclusion of larger zooplankton (e.g. Banse 1990, 1991; de Baar et al. 1990; Dugdale and Wilk- erson 1990:l. Two mesoscale iron-enrichment experiments in the equatorial Pacific (IronEx I and II) have subsequently demonstrated responses of the phytoplankton community to nanomolar additions of iron that parallel the effects observed

Constraints on phytoplankton production 407

in bottles (Kolber et al. 1994; Martin et al. 1994; Coale et al. 1996a, b), though not necessarily the expected magnitudes of nutrient and CO, decreases (Watson et al. 1994; Cullen 1995).

Although the grazing and iron hypotheses were developed independently and argued, at least initially, as antagonistic propositions, neither has proven sufficient in itself to explain all of the characteristics of HNLC regions (Chavez et al. 1991; Frost 1991; Morel et al. 1991b; Price et al. 1991). Because phytoplankton species do not have identical iron requirements (Brand et al. 1983; Brand 1991; Hudson and Morel 1990), some large diatoms, for example, may be held below their physiological growth potential to a greater degree in HNLC regions than others (Miller et al. 1991; Morel et al. 199 la, b; Price et al. 1994). Further, the high rates of phytoplankton growth in the steady-state microbial communities of HNLC regions imply strong links to rapid rates of grazer removal and efficient mineral cycling (Frost and Franzen 1992; Cullen 1991; Cullen et al. 1992). In fact, despite enhanced rates of primary production, specific production and growth, the first mesoscale enrichment experiment provided clear evidence that grazing modulated, at least in part, the effects of iron addition on phytoplankton biomass and drawdown of inorganic carbon (Martin et al. 1994; Banse 1995; Cullen 1995).

This EqPac synthesis contributes further to the notion that iron regulation and grazing are complementary mechanisms, which together constrain phytoplankton production in the central equatorial Pacific. To build this case, we first develop the conceptual relationships between grazing and substrate regulation of phytoplankton production. Descriptive and experimental results from EqPac and related recent studies in the equatorial Pacific are then presented in separate sections dealing with specific effects of iron and grazing and their interactions via remineralization by grazers. We then bring these pieces of evidence together in a summary of iron flows through the equatorial plankton community. Lastly, we argue that the underlying principles and relationships from this synthesis apply equally well to the oligotrophic gyres.

Conceptual relationships

Figure 1 provides a simple conceptual context for orga- nizing our thinking about the factors that contribute to phytoplankton production and how they interact and relate to one another as potential control mechanisms. The measured rate of primary production is a function of two variables- the specific growth rate (p,) of phytoplankton and phytoplankton biomass (B); PP = k X B. Production is therefore constrained by all factors that directly or indirectly influence growth rate, standing stock, or both.

Iron availability could affect p and B in HNLC regions, similar to the effects of major nutrients in oligotrophic gyres. If the concentration of iron in near-surface waters is dispro- portionately low relative to the elemental composition of phytoplankton and the availability of other essential elements, the maximum daily rate of cell division at steady state may be constrained (i.e. regulated below maximal levels) by the daily rate of supply of biologically available iron. The

Trophic Transfer

t

N HJ+ Recycled

+ q , BO

b Advection

Fe, NOa- Sinking Flux

New Nutrients Upwelling & Aeolian

Fig. 1. Conceptual relationships between iron and grazing controls of phytoplankton production. Pools and fluxes are indicated by boxes and arrows, respectively.

effects of low iron concentration may be expressed to varying degrees, up to and including relatively minor depression of physiological growth potential, and these effects may vary contemporaneously among different populations of the community. In addition, even when relatively high rates of cell division are realized, the cells may show symptoms of iron deficiency, including reduced enzyme function, cell pigments, and cellular iron quota (Morel et al. 1991b; Fal- kowski et al. 1992; Geider and La Roche 1994; Sunda and Huntsman 1995). While such deficient cells may not show dramatic population growth responses to enhanced availability of iron, the improvement in their condition may be evident in certain diagnostic parameters, such as photosynthetic efficiency (e.g. Greene et al. 1994).

Iron becomes available to pelagic plankton community from photochemical transformation or biological remineralization of iron already in the euphotic zone (Bruland et al. 1991) or the introduction of “new” iron from deeper water (via upwelling or mixing) or the deposition of atmospheric dust. If iron is a limiting substrate, the total iron inventory in the euphotic zone, as opposed to some other essential element, must set an upper bound to the standing stock of the resident plankton community, including both auto- and heterotrophic components. If iron flux into the euphotic zone changes spatially or temporally, for example, with seasonal or interannual variability in upwelling or aeolian input, the total biomass of the plankton community should shift, within natural bounds, to a level that is higher or lower than the long-term mean value.

Grazers can contribute directly to the maintenance of phytoplankton steady state by cropping daily production (AZ?). However, other factors, such as cell sinking and advection, may also be significant removal terms in accounting for the fate of production. Therefore, the relative importance of

408

Dissolved iron (nmol Fe kg-l) Dissolved Iron (nmol Fe kg-l) 0.00 0.10 0.20 0.30 0.40

25

Landry et al.

125

E 5 175

e n 225

275

325

375 Fig. 2. Depth profile of iron concentration at the equator,

14O”W. Line is the best fit to all data (from Gordon et al. 1997).

grazing needs to be demonstrated by comparing the rate estimates for grazing losses against phytoplankton-specific growth (p,). Only if grazers have the potential to exceed the growth rates of their phytoplankton prey under ambient conditions and keep their populations suppressed when growth factors are available in excess can the grazers be said to control phytoplankton stocks. In addition to their direct impact on phytoplankton stock, grazers return to the phytoplankton remineralized elements, including iron (Hutchins et al. 1993), essential for sustaining a high p. The combination of direct cropping and remineralization processes provides a potentially powerful feedback mechanism by which grazers influence both the biomass and rate components of phytoplankton production (Stone and Weisburd 1992; Stone and Berman 1993). Moreover, because ammonium suppresses nitrate uptake (Wheeler and Kokkinakis 1990), the remineralization of nitrogen as ammonium by grazers is important in maintaining high levels of nitrate in HNLC systems.

Iron limitation

The evidence that ambient iron supply in the equatorial Pacific constrains phytoplankton processes consists of concentrations, physiological parameters, growth rates in bottle incubations, and natural community responses. We focus on three specific points. First, iron concentrations in the euphotic zone are at levels that regulate growth rates of at least some components of the phytoplankton community. Second,

250

Fig. 3. Depth distributions of iron at 14O”W from FeLine transect cruise in July 1990 (from Coale et al. 1996a).

symptoms of iron deficiency are exhibited by the ambient phytoplankton community, not just relatively rare species and groups. Third, phytoplankton new production is consistent with the flux rates of iron into the euphotic zone from external sources, as hypothesized by Martin (1990).

Near-surl’ace concentrations of iron are extremely low in the central equatorial Pacific, at or below the analytical de- tection limit of about 30 pM (Fig. 2). Immediately below the euphotic zone at the equator, the equatorial undercurrent is evident as a subsurface maximum of -300 pM Fe (Fig. 3). A local maximum in particulate aluminum also occurs in the core of the undercurrent, suggesting a terrigenous source for the iron found there. This probably occurs in the shelf area off Papua, New Guinea where the undercurrent origi- nates (Lindstrom et al. 1987).

Figure 4 is a compilation of results from equatorial grow- out experiments in which initial iron concentrations were measured and growth rates determined from initial PON and the incremental decrease in nitrate during the course of 5- 7-d incubations (Coale et al. 1996a). Each point represents a single bol:tle experiment, some of which were spiked with dissolved iron (iron chloride or sulfate), while others, probably with varying low levels of Fe contamination, were left as controls. Net community growth rates were determined from the ra,tes of increase in standing stock during several days of exponential growth after an initial lag of 2-3 d. According to these results, the most iron-responsive component of the plankton community, diatoms and other large cells (Martin et al. 1989; Chavez et al. 199 1, Coale 199 1; Price et al. 1994), had a maximum net growth rate of -0.7 d-l, and this rate was approached at concentrations of ap- proximately 1 nM (Fig. 4). Fitting a rectilinear hyperbola to these data gave a half-saturation constant (& = 120 PM) about four times the natural concentration of iron in equatorial surface waters. In comparison, Price et al. (1994) used tracer additions of 59Fe to determine uptake kinetics of the ambient phytoplankton community in equatorial waters and found a half-saturation constant of 34 pM Fe. Although cau- tion needs to be exercised in interpreting these results, they

Constraints on phytoplankton production

0

:+ . 0

0

01 I I I I I I I 0 2 4 6 8 10 12

Iron Concentration (nM)

Fig. 4. Net growth rate response of phytoplankton to total dissolved iron in grow-out experiments in the central equatorial Pacific (adapted from Coale et al. 1996a). The relationship between net growth rate (d-l) and iron concentration (nM) is described by non- linear fit of untransformed data to the Michaelis-Menten equation; P “,,, = (0.69 [Fe]) ([Fe] + 0.12)-l.

are consistent with the view that ambient concentrations of iron in the equatorial Pacific are sufficiently low to affect rate processes (uptake or growth) of at least some of the ecological dominants and particularly larger forms that are generally rare (Morel et al. 1991b).

IronEx fertilization experiments provide the most com- pelling evidence for iron limitation in HNLC equatorial waters (Martin et al. 1994; Coale et al. 1996b). The natural ecosystem not only responded significantly to low-level iron enrichment, which constitutes in itself a prima facie case for some form of iron regulation, but the responses were rapid and involved measurable enhancements in photochemical energy conversion efficiency (Greene et al. 1994; Kolber et al. 1994) and standing stocks of even small dominant mem- bers of the phytoplankton community (Martin et al. 1994). Symptoms of iron deficiency among the ambient equatorial phytoplankton were also revealed by Lindley et al. (1995) from latitudinal variations in the maximum -quantum yield of photosynthesis (@,J, an indicator of nutrient status (Cleveland and Perry 1987; Kolber et al. 1988). During normal upwelling conditions (August-September 1992), the pa- rameter Q,ff,, which is @‘, corrected for light absorption by photoprotective carotenoids, increased systematically from 0.015 on the southern end of the EqPac transect at 14O”W to 0.035 on the northern end (Fig. 5). Such a pattern is consistent with the greater level of aeolian input of iron to the central northern Pacific (Duce and Tindale 1991). Nonethe- less, some of the latitudinal variability in @‘ff, may reflect differences in community composition. For example, the high index on the northern end of the EqPac transect is in an-area dominated by Prochlorococcus spp. (Landry et al. 1996), cells that are unlikely to be nutrient stressed because of their small size. The local maximum of Q?ff, in the vicinity of the equator is presumably due to the input of additional new iron from upwelling, but it also coincides with a local minimum in diatom pigment biomass. In contrast, the de-

0.035 h r I

5 2 0.030

a a E 0.025

0 5

" 0.020

s-

0.015

0.010 1 I I I I I I

15"s IO 5 0 5 10 1 Latitude

5” N

Fig. 5. Latitudinal variation of @E,, the maximum quantum yield of photosynthesis corrected for light absorption by photoprotective carotenoid pigments. Vertical lines represent standard deviations of 4-12 replicate measurements (adapted from Lindley et al. 1995).

creases on either side of the equator correspond to local maxima in diatoms (see below).

We can conclude from the QE? index of physiological condition along the EqPac transect as well as the bulk community responses to the in situ iron fertilization experiment (IronEx) that the ambient community shows signs of iron deficiency. However, the extent to which this is reflected in the growth rates of small community dominants has not been entirely resolved. Based on cell cycle analysis, a nonincu- bation technique, Vaulot et al. (1995) found consistently high growth rates for Prochlorococcus spp. (0.7-0.9 d-l at 30 m and depth-integrated rates of 0.58 d-l to 150 m) during EqPac cruises and concluded that this component of the ambient community was not severely growth-rate limited at the equator. Taxa-specific results from dilution incubations were equivocal for Prochlorococcus spp., due to photobleaching at high near-surface light levels; nonetheless, other abundant taxa (e.g. Synechococcus spp., prymnesiophytes, and pela- gophytes) generally showed significant (40-300%) increases in growth rates during normal upwelling, relative to El Niiio conditions when the rate of iron supply to the euphotic zone was lower (Latasa et al. 1997).

Close to the equator, new iron comes predominately via the upwelling of water from the equatorial undercurrent. The time-averaged magnitude of this flux into the euphotic zone can be estimated from the composite iron profile overlying the undercurrent (Fig. 2) and the mean profile of upwelling velocity from Gargett (1991). At the base of the euphotic zone, the 0.1% light level at 120 m, the average iron concentration of 89 nmol me3 and a 1.23-m d-l upwelling velocity give an average upwelling flux of about 120 nmol Fe rnp2 d-l (Table 1; G or d on et al. 1997). Although upwelling velocity is higher at shallower depths within the euphotic zone, with a maximum at about 50 m (Gargett 1991; Chai

410 L.undry et al.

Table I. Estimates of iron fluxes to the euphotic zone of central equatorial Pacific (14O”W). Potential new production in carbon equivalents is computed from the iron fluxes assuming a C : Fe ratio of lo5 (adapted from Coale et al. 1996a).

Source Upwelling Atmosphere Eddy diffusion Total

Fe flux (nmol Fe m-* d-l)

120 5-25

0.2-14 125-I 59

Potential new production

(nmol C m-* d-l) 12

0.5-2.5 0.03-l .4

13-16

1995), the flux of dissolved iron decreases markedly at these depths as the supply from below is depleted by uptake into particulates. Overall, the input of “new” iron from upwelling at the equator is about an order of magnitude greater than the atmospheric contribution of soluble iron (5-25 nmol m-2 d - I ; Duce and Tindale 199 1); however, aeolian dust might be as, or more important, a source of new iron to the mixed layer. Diffusive flux, determined from the gradient in the composite iron profile (2.4 nmol m-4 at 120 m, Fig. 3) and a vertical eddy diffusivity coefficient (K,) of 0.086 m-* d-l (Peters et al. 1988), provides a negligible input of new iron (0.2 nmol m-2 d-l; Table 1) compared to either upwelling or atmospheric deposition. Carr et al.‘s (1995) higher value of Kp, -6 me2 d-l for the region from 1”N to 1’S gives an upper estimate for diffusive flux of 14 nmol Fe m-2 d- I, which is still an order of magnitude lower than that due to upwelling. Rates of new Fe supply to the euphotic zone should drop off rapidly within a degree or two of the equatorial divergence, where the contributions of upwelling and eddy diffusion become negligible (Carr et al. 1995) and where aeolian dust represents the dominant source.

The carbon-equivalent new production from these external sources of iron can be determined using a C : Fe ratio of lo”, which is in the midrange of measurements for equatorial plankton (Geider and La Roche 1994). The result, 13-16 mmol C me2 d-l, is in good agreement with nonE1 Nifio estimates of new production based on nitrogen uptake (Mc- Carthy et al. 1996). Thus, the rate of new phytoplankton production in the central equatorial Pacific appears to be constrained by the rate of supply of iron largely from below the euphotic zone, as has also been determined for HNLC regions of the Southern Ocean (de Baar et al. 1995). When more iron is upwelled into the euphotic zone, the magnitude of new production should increase. Such an effect was observed by McCarthy et al. (1996) in comparing new production rates between 2”N and 2’S during the March-Feb- ruary survey cruise, when upwelling of undercurrent waters was diminished by El Nifio conditions (mean new production = 4.8 mmol C m-2 d- l), to those in August-September 1992, when the system had returned to more normal upwelling conditions (mean new production = 18.5 mmol C m-2 d-l ).

In contrast to iron, the supply of nitrate to surface waters by upwelling does not determine the rate of new production in the central equatorial Pacific. Assuming Redfield C : N stoichiometry, the iron fluxes in Table 1 are sufficient to

account for the biological uptake of 2.0-2.4 mmol NO,- rnp2 d-l. This is substantially less than Carr et al.‘s (1995) estimate of the rate of NO,- supply to the euphotic zone between 1”N and 1’S, 15O”W (5.5 mmol m- * d I). Consequent- ly, the depletion of upwelled iron by new production, should still leave 4rO-60% of the nitrate unutilized, in good agreement with estimates of the meridional transport of nitrate away from the equatorial divergence (Carr et al. 1995).

Grazing bdance

In the theoretically simple scenario of steady-state phytoplankton abundances, the appropriate test of grazing balance is to compare estimates of phytoplankton growth rate to the combined estimates of grazing losses to zooplankton (Miller et al. 1991). Although the approach sounds straight- forward, the desired result has never been fully realized, even when intensively investigated in the subarctic Pacific (Strom and Welschmeyer 1991; Landry et al. 1993b). Other natural losses of phytoplankton production to lateral advection and sinking complicate the analysis. To allow for some losses to these other processes, our criteria for demonstrating balance in this analysis are that phytoplankton growth should be explained by the sum total of all measured losses and that grazing should be unequivocally the dominant removal process.

Phytoplankton growth rates (p) and grazing rates of micro- and :>200-p,rn mesozooplankton (g,i,,, and gmeso, respectively) were derived from experimental studies from 1”N to 1’S, 14O”W on four EqPac cruises. Stations in this area were occ@ed from 19 to 28 February and 23 March to 10 April 1992 during El Nifio conditions and from 24 August to 2 September and 2 to 20 October 1992 when hydrographic conditions were close to climatological mean values (i.e. normal upwelling). Rate estimates for individual parameters, generally from different teams of researchers, were similar for the two cruises in the same season. We, therefore, combined results when comparing the two phases of the system.

Estimates of phytoplankton p, came from two sources, depth-integrated values of 14C-primary production relative to chlorophyll standing stock (Barber et al. 1996) and the results of dilution incubations (Landry et al. 1995a,b; Verity et al. 1996). A C : Chl weight ratio of 58 (Eppley et al. 1992) was used to convert chlorophyll standing stock to phytoplankton carbon biomass for the former estimates, and the depth integration was taken to the 0.1% light level. Dilution incubations, were all conducted under simulated in situ conditions (temperature- and light-regulated deck incubators) for 24 h, but other details of the experimental protocols differed somewhat between cruises. For transect cruises (February and August-September 1992), three experiments were conducted simultaneously at each station, in conjunction with in situ 14C-uptake experiments, using water collected from the upper (30-50% of surface irradiance, I,), middle (7-14% I,), and lower euphotic zone (1% 1,) (Landry et al. 1995a). For time-series cruises (March-April and October 1992), one experiment was conducted on a given day, so depth-averaged estimates of F combined individual experiments run at different depths and at different times (as long as 5 d apart)


Table 2. Estimates of photoplankton growth and removal rates from the equator, 14O”W during El Nifio (February-April 1992) and normal upwelling (August-October 1992) conditions. g,,c,o and g,,,,, are grazing rates of microzooplankton and mesozooplankton, respectively. With the exception of advective transport, all estimates are mean specific rates (d-l) for the euphotic zone (2 = SD, n = number of depth profiles averaged). (Data not available-na.)

Measured variable

Growth rates

El NiHo Normal upwelling

Rate (d-l) n Rate (d-l) n

~.L--‘~C uptake F-dilution

Removal rates

0.46-t-0.09 15 0.57-c-0.08 14 0.43&O. 12 7 0.79*0.11 7

g,,,,-dilution 0.48&0.15 7 0.52kO.15 7 gmcso-14C uptake 0.03 20.02 11 0.11+0.08 13 g,,,,-gut fluorescence 0.03 20.02 11 0.03+0.02 13 Lateral transport-

drifter na 0.16?0.06 2

(Verity et al. 1996). All dilution experiments were analyzed according to the linear assumptions of Landry and Hassett (1982), with the underlying assumptions of the approach val- idated as described by Landry et al. (1995b). Community averaged estimates of F and g,nic,o were based on changes in chlorophyll a concentrations, determined fluorometrically.

Estimates of mesozooplankton grazing on phytoplankton were derived from gut pigment analyses and the uptake of radiolabeled food particles during short-term 14C-incubation experiments. The first approach uses chlorophyll and its fluo- rescent degradation products as a tracer for zooplankton grazing on phytoplankton biomass (Dam et al. 1995). The second measures the ingestion by planktonic animals of newly fixed primary production and is directly comparable to the assessment of phytoplankton production by 14C (Ro- man and Gauzens 1997). Because potentially variable pigment losses during gut passage (e.g. Conover et al. 1986; Lopez et al. 1988) were not quantified for the former meth- od, the latter is considered a more appropriate basis for the present comparison of phytoplankton growth and loss rates.

During El Nifio, the two estimates of depth-integrated phytoplankton growth were comparable (0.43 d-l for the dilution estimates vs. 0.46 d-l from the 14C-uptake experiments), and both were less than the combined grazing rates of micro- and mesozooplankton (0.51 d-l) (Table 2). Lateral advection was not measured during El Nifio, and vertical sinking, determined from the flux of chlorophyll into free- drifting sediment traps, was low (J. Newton pers. comm.). Even without these other removal terms, the results of this rate comparison are consistent with balanced growth and grazing processes, and microzooplankton are clearly the dominant grazers. These results are as expected, given the overwhelming dominance of small cells in the phytoplankton community during El Nifio conditions; 92% of chlorophyll passed through 2-km filters, and only 3% was retained on 14-pm filters (Bidigare and Ondrusek 1996).

Microzooplankton also dominated grazing during normal upwelling (Table 2). However, the relative contribution of mesozooplankton increased markedly over El Nifio due both

-6 ti 144 143 142 141 140

Longtitude (“W)

0.3

0 Tin’: (d)

4

Fig. 6. Advective transport of plankton away from the equator as measured with instrumented drifters released at 0” and l”N, near 14O”W. Left: drifter paths away from the equatorial divergence and to the west; open circles indicate drifter position at daily intervals. Right: local growth of phytoplankton is implied from increases in estimated concentrations of Chl a (mg m -“) at the drifters over the first few days of transport. Vertical bars show standard deviation of mean chlorophyll estimates.

to higher mesozooplankton biomass and to higher biomass- specific grazing rates. Presumably this reflected the higher abundances of large phytoplankton cells, particularly diatoms, on these cruises (8 1% of chlorophyll passed through 2-pm filters and 12% was retained by the 14-pm filters; Bidigare and Ondrusek 1996). Depth-integrated rates of phytoplankton growth were also higher, and there was a substantial difference between rate estimates from dilution and 14C experiments, which we attribute to an assumed constant ratio of C : Chl, rather than to an inherent problem with either measurement technique. Compared to dilution results, the combined micro- and mesozooplankton grazing rates leave an imbalance of about 0.16 d- I.

This imbalance is consistent with estimates of lateral advective transport from free drifters released near the equator (0” and 1”N; Fig. 6). The drifters were equipped with spectral radiometers that documented the increase in phytoplankton chlorophyll as water was advected away from our study site to the west and to higher latitude. Chlorophyll increased by 50 to > 100% during the first few days of transit time, giving estimates of mean net growth rate ranging from 0.12 to 0.2 d-l (Table 2). W e can conclude from this analysis that grazing accounted for most of the phytoplankton production during normal upwelling conditions at the equator, but some production was lost meridionally via lateral transport. Bi- modal peaks in phytoplankton abundance with a local minimum at the equator are consistent with this scenario. Bidi- gare and Ondrusek (1996) noted such peaks for normal upwelling, but they were not evident during El Nifio.

Grazing balance defines the observed steady-state condition in which phytoplankton cells are removed, on average, close to the rates at which they are produced. In this sense, it is more a demonstration of the internal consistency of the various rate measurements involved, rather than proof of

412 Landry et al.

0 0

0.1 0.2 0.3 0.4 Chlorophyll (mg ms3)

0.5

Fig. 7. Growth rates of phytoplankton as a function of initial chlorophyll. Data from dilution experiments conducted from 1”N to I’S, 14O”W on EqPac cruises in February-April and August-Oc- tober 1992. Line is the least-square linear regression for all data points (Y = 0.78 + 0.44X, Y = 0.16).

phytoplankton control by grazers. Balanced grazing removal is compatible, in fact, with regulation by a growth substrate such as iron, which sets the maximum level of community biomass that can be supported by the most scarce resource (Lindley et al. 1995). The remineralization feedback links growth and removal processes. In the absence of such a feedback, one would expect that growth of resource-limited population should be classically density dependent; that is, as biomass increases, growth rate should decline. Over a range of about four fold in initial phytoplankton chlorophyll, however, there was no appreciable decline in community estimates of p from dilution experiments (Fig. 7). The chemostat analogy of Frost and Franzen (1992) is appropriate here, with grazers functioning as both the outflow sink and the reservoir of nutrients to sustain further growth. Phytoplank- ton cannot long sustain population growth rates that deplete the micronutrient upon which further cell division depends. Thus, following perturbation, the system must naturally shift to a new level of biomass where growth is balanced by grazing. Phytoplankton in such a state may exhibit symptoms of nutrient deficiency, even while growing at rates that are nearly, but not quite, maximum (Laws et al. 1987, 1989).

Nutrient relationships

As Miller et al. (199 1) argued previously for the subarctic Pacific, efficient remineralization plays a key role in explain- ing two of the characteristic properties of HNLC regions: sustained high rates of phytoplankton cell division and consistently high concentrations of nitrate. The latter was ad- dressed in grow-out experiments conducted on the EqPac cruises with the addition of different inorganic nitrogenous nutrients (ammonium, nitrite) and with or without added iron. As might be expected from the preferential utilization of more reduced forms of nitrogen by phytoplankton (Wheeler and Kokkinakis 1990), ammonium was utilized first in these incubations, suppressing the uptake of nitrate

until ammonium concentration approached low ambient levels. This effect was also demonstrated for equatorial phytoplankton j n the experimental studies of Price et al. (1994), in which ammonium concentrations above 0.2 ~.LM substantially depressed uptake rates of nitrate. EqPac incubations also revealed that the various nutrient treatments (and controls) had no effect on the amount of nitrogen incorporated into any new growth of phytoplankton; observed increases in chlorophyll were always accompanied by reductions of inorganic nitrogen in a ratio close to 1.3 krnol N : kg Chl a. In the absence of other regulating factors, therefore, phytoplankton gr,owth rates of - 1 .O d-l should rapidly deplete any and all sources of inorganic nitrogen. These high rates of growth can only be maintained with equally high rates of ammonium (and iron) remineralization by grazers.

The dependency of the equatorial production system on recycled nutrients is clear in the nitrogen uptake studies conducted on EJqPac cruises (Fig. 8). In comparison to oligotrophic systems to the north and south, utilization rates of ammonium were significantly elevated in the vicinity of the equator during normal upwelling conditions and substantially less so during El Nina. The corresponding nitrogen-based estimates of the new-production ratio (fl Fig. 8) varied little with latitude during El Nifio, except for a slight depression at the equal:or. During normal upwelling, the f-ratio showed a bimodal distribution around the equator (1’S) overlying a strong trend increasing from south to north, which is very similar to the latitudinal pattern in the maximum quantum yield of photosynthesis (@z,) from the same cruise (Fig. 5). The greater relative rate of nitrate utilization in the northern portion of this transect is consistent with greater aeolian input of iron (Duce and Tindale 1991).

The bimodal pattern observed for new production, with local maxima on either side of the equator, is also found for other variables expected to vary positively with the uptake of N03--, such as mesozooplankton biomass (White et al. 1995; Zhang et al. 1995), organic carbon fluxes into sediment traps (Murray et al. 1996), and biogenic silica and organic carbon fluxes into deep moored traps (Honjo et al. 1995). In particular, nitrate uptake rates were strongly correlated with diatom chlorophyll biomass during EqPac cruises (r2 = 0.96, n = 26; Fig. 9), implying a tight connection between the low rates of nitrate utilization observed in the central equatorial Pacific and factors that hold diatoms to a small fraction of phytoplankton community biomass.

From the various physical, chemical, and biological measurements made during EqPac cruises (Table 2), we can rea- sonably quantify most of the fluxes in the conceptual model of Fig. 1. We use an average estimate of aeolian iron flux from Duce and Tindale (199 1) for both cruise periods because the winds were similar. However, modeled upwelling velocites al. 120 m differed by about a factor of three between Febr,uary-April and August-October 1992 (0.33 vs. 1.09 m d-l) (Buesseler et al. 1995; Chai 1995), and, using the Kp estimates of Cart- et al. (1995), diffusive fluxes were also reduced during El Nifio when the undercurrent was deeper. Combining the modeled upwelling velocities and measured iron concentrations at the base of the euphotic zone yields estimates of iron fluxes for the 1992 El Niiio and the subsequent normal period (48 vs. 134 nmol Fe m-2


g 30 A

----Q--m- El NiAo 9 E - Normal

0: ” ” ” ” ” ” ” ’

0.5

B -m-s Q -w-s El NiAo - - Normal

0.4

.g 0.3-

3 d 0.2-

) . --\

0.1 -

y 0.0 ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ’ ‘ ’ ’

12’S 9 6 3 0 3 6 9 12”N

Latitude Fig. 8. Distributions of nitrogen uptake parameters on EqPac

transect cruises during El NiHo (January-February 1992) and normal upwelling (August-September 1992) conditions: A. Ammoni- um uptake (mmol NH,+ m-2 d-l). B. The f-ratio = (NO,- uptake) : (uptake of NO, + NO,- + NH, ’ ). All estimates integrated to 1% Z, (adapted from McCarthy et al. 1996).

d- I; Table 3). Th e increase in iron flux during normal upwelling was evident in only a slight increase in mean phytoplankton stock size at the equator (1.36 compared to 1.6 nmol Fe equiv. m-* assuming a C : Fe ratio of 10”; Fig. lo), but biomass also accumulated to the north and south due to lateral advection. Iron-equivalent phytoplankton production was 50% higher for the normal upwelling period (1.2 vs. 0.8 p,mol Fe m-* d-l). Combined estimates of loss terms exceeded slightly the measured rates of primary productivity for both periods, but microzooplankton grazing was the dominant removal process during both periods, accounting for 93% of the combined measured removal rates during El Nifio and 62% during normal upwelling.

Recycling of iron is the unmeasured term in these budgets; the estimates of 0.74 p,mol Fe m -* d- I for El Nifio conditions and 1.08 for normal upwelling were computed from the difference between the iron needed to support measured rates

m 8 I ----Q---- NO3 (El Nifio) -m-B -0 -m-m Chl (El NiAo) - NO3 (Normal) - Chl (Normal)

6 3 0 3 Latitude

6 12”N

0 El NiFio l Normal

- Model 2

0 . Y = 0.46 + 1.06X

’ I I I I ' I I

0 1 2 3 4 5

[Ch~]diatorns (mg m-3) Fig. 9. The relationship between NO,- uptake and Chl biomass

in diatoms on EqPac transect cruises during El Nifio (January-Feb- ruary 1992) and normal upwelling (August-September 1992) conditions. A. Latitudinal distributions of nitrate uptake (mmol NO,- m-2 d-l; McCarthy et al. 1996) and chlorophyll biomass in diatoms (mg me2, Bidigare and Ondrusek 1996). All estimates integrated to 1% I,,. B. Relationship between NO,- uptake and diatom Chl biomass. The solid line corresponds to the fit obtained by model 2 linear regression analysis.

of production and the rate of supply of new iron. Compared to the rate estimates for combined micro- and mesozooplankton grazing, the computed rates of iron remineralization indicate that most of the iron in the phytoplankton consumed by grazers is recycled back to phytoplankton. Corresponding iron-based f-ratios for these two periods, 0.08 and 0.12, respectively, are consistent with, and quite close to, new production estimates from 15N studies, which give values of 0.07 and 0.12 (Table 3). In fact, if we assume Redfield C : N ratios, and thus an N : Fe ratio of 1.5 X 104, the implied Fe remineralization rates are in good agreement, in terms of nitrogen, with measured rates of ammonium uptake (11.2

414 hndry et al.

Table 3. Phytoplankton standing stocks and related rate estimates from the equator, 14O”W during El NiAo (February-April 1992) and normal upwelling (August-October 1992) conditions. All estimates are relevant to the euphotic zone between 1”N and 1”s. Grazing rates are calculated relative to measured rates of primary production using the ratios of g; : plJc in Table 2. A C : Chl ratio of 58 is assumed.

Variable El Nifio Normal Source

Phytoplankton biomass (mmol C m-*) Primary production (mmol C m-* d-l) Nitrate uptake (mmol N m-2 d-l) Nitrate uptake (mmol N m-2 d-l) Ammonium uptake (mmol N m-2 d-l) New production f-ratio* Microzooplankton grazing (mmol C m-2 d-l) Mesozooplankton grazing (mmol C m-2 d-l) Lateral transport (mmol C m-2 d-l) Iron upwelling (nmol Fe m-2 d-l) Diffusive flux (nmol Fe rnd2 d-l) Aeolian iron (nmol Fe m-2 d-l)

136 80 0.50 0.06 6.3 0.07

83 5 -

43 5

15

160 Barber et al. 1996 123 Barber et al. 1996

2.4 McCarthy et al. 1996 0.65 McCarthy et al. 1996

16.4 McCarthy et al. 1996 0.12 McCarthy et al. 1996

112 This paper-Table 2 24 This paper-Table 2 17 This paper-Table 2

120 Chai 1995; this paper 14 This paper-Table 1 15 Duce and tindale 1991

*f = (NO, - uptake): (uptake of NO3 + NO, + NH, I).

and 16.2 mmol N m--* d-l for the two cruises; in comparison to values in Table 3). Interestingly, the new production ratio during normal upwelling does not reflect the fact that a significant fraction of primary production is advected away from the equator in near-surface waters. This implies that another recycling process or source term may be unaccount- ed for, possibly the return flow of ammonium, dissolved organic material (DOM), and organically complexed iron from higher latitudes in the deeper euphotic zone. Complexed iron would not appear in the measured concentrations of dissolved, biologically available iron but may be released by photochemical processes as it is upwelled closer to the surface (Wells and Mayer 1991). To the extent that it is not balanced by export losses, atmospheric deposition during several days of poleward transport could also enhance the total iron inventory of the euphotic zone up to about double that of the upwelling source, thereby supporting an accu- mulation of plankton biomass on either side of the equator.

One of the conclusions that we can draw from these budgets is that iron is cycled with about the same efficiency as ammonium (as suggested by Bruland et al. 1991); otherwise, the f-ratios for iron would be substantially different than those for nitrogen. The turnover of iron is also relatively rapid compared to the concentration of free dissolved iron in the euphotic zone. For a 100-m euphotic zone characterized by iron concentrations on the order of 30 pM, for example, a recycling rate of 1 p,mol Fe m-* d-l represents a pool turnover rate of 3 d. The 59Fe-uptake rates (V,,, = 8.5 pmol Fe liter- I h- I) of Price et al. (1994) imply even faster turnover times of the dissolved iron pool (7.5 h) but may be exaggerated by the scavenging of tracer by nonliving particles. In either case, nitrogen turns over slowly compared to iron, with remineralization rates of ammonium accounting for only 2-3% of total inorganic nitrogen per day (based on nitrate concentrations of 3 and 6 FM, respectively, during the two cruise periods). Despite its small contribution to the total pool of dissolved inorganic nitrogen, however, ammonium cycling is sufficient to account for about 90% of the phytoplankton N requirement.

The EqPac synthesis

The results from EqPac and related programs in the equatorial Pacific suggest that iron regulation and grazing function jointly to constrain the standing stocks and production rates of phytoplankton. Iron affects the composition of the phytoplankton community by selecting for smaller dominant forms with high surface area : volume ratios for iron uptake (Morel et al. 1991b; Sunda and Huntsman 1995). However, iron regula:tion does more than account for the relative rarity of large phytoplankton. Even the small ambient taxa exhibit physiological symptoms of iron deficiency, and, judging from their population responses to mesoscale iron enrichments (Martin et al. 1994) as well as growth rates in dilution incubations1 (Latasa et al. 1997), the growth rates of these taxa are maintained somewhat below their physiological po- tentials by iron. Grazers, particularly microzooplankton, re- move most of the daily production of phytoplankton biomass and sustain relatively high rates of phytoplankton cell division with the return of remineralized nutrients. Recycling of nitrogen as ammonium contributes because preferential uptake of ammonium helps to suppress the uptake of nitrate. However, because ammonium can neither prevent for long the drawdown of other inorganic nitrogen constituents when iron is available in excess nor sustain phytoplankton growth if iron is not available, the role of grazers in iron cycling is a key process.

Grazing balance is an important part of this explanation, but not necessarily as the ultimate determinant of phytoplankton abundance. Although phytoplankton may be held by grazers to lower than the maximum concentration that would be attained if all iron was available only to them (e.g. Frost and FYanzen 1992), this only represents a redistribution of the iron inventory among several pools-phytoplankton, grazers, and dissolved. The addition of more iron by experimental manipulation or nature (upwelling vs. El Nifio) will allow phytoplankton to increase, despite grazing. Grazing balance defines more or less a steady-state condition, the approximate equivalency between phytoplankton growth and grazer-dominated removal processes, which is an inevitable

El Niiio Feb-Apr 7992

Mesozooplankton

t Microzooplankton

new Fe

Normal Upwelling Aug-Ott 1992

new Fe

Fig. 10. Comparison of iron budgets for rhr central equaror~al Pacific during the 1992 El NiIio and the subsequal normal upwelling period. Phyloplankton standing stock is expressed as p,mol Fe rmz assuming a C : Fe ratio of 10’. All rate eslimates are bmol Fe m-’ d ~I. Mesozooplankton grazing rates were computed relative 10 primary production rates using the ratios of g,,,,,, (“C uptake) to & (“C uptake) in Table 2. To minimize discrepancies in budget estimates due to uncertainlies in C: Chl, the computed microzooplankton grazing and advcctive losses were based on the ratios of these losses to p. estimates from dilution experiments (Table 2).

consequence of resource limitation coupled with rapid and efficient recycling. In a tightly coupled balance of growth, grazing, and remineralization processes (Aram et al. 1983), phytoplankton cannot grow faster than the rate of supply of iron will allow, and this iron comes predominately through the recycling of particulate iron consumed by grazers. Graz- ers, in turn, cannot consistently crop phytoplankton at rates exceeding daily primary production without substantially re- ducing their food resources and providing excess substrate to stimulate phytoplankton.

Pro&tan grazers have high maximum rates of population

growth, which may enable them w oven&e and control increases of some phytoplankton groups. It is not clear, however, that they achieve these high rates at food concentrations available to them in the open oceans. Although this is an open question, there is at least some evidence that protistan grazers can crop picophytoplankton and nanoplankton at rates as fast or faster than they can grow (e.g. Landry et al. 199%; Verity et al. 1996; Latasa et al. 1997). Selective grazing control probably explains why abundances of picoplankton populations were visually unchanged during El Niim and the nomml upwelling periods (Landry et al. 1996), while diatoms and other large forms increased substantially in abundance with normal upwelling (Bidigare and Ondrusek 1996). Near-surface concentrations of iron were low in both seasons, so smaller forms should have enjoyed a competitive advantage in either case. If grazers controlled the smaller phytoplankton within a given range in abundance, however, the higher iron inventory in the euphotic zone during normal upwelling conditions would have allowed for increased abundances of groups, like diatoms, capable of exploiting the excess resources.

Many potentially important aspects of the dynamics of HNLC systems may be overlooked in this simple explanation. For example, one might conclude from relative rarity of diatoms in the ambient community and their dramatic response to added iron that their growth rates must be held to much lower levels by iron relative to those of community dominants. However, tax-specific results from dilution experiments revealed just the opposite; diatom-specific growth rates were higher than other groups (generally exceeding I .O do ‘), and diatoms showed the least response to short-term or seasonal variations in Fe supply (Latasa et al. 1997; see also Strom and Welschmeyer 1991; Welschmeyer et al. 1991; Landry et al. 19936 for the subarctic Pacific). As discussed by Latasa et al. (1997). such results are not easily dismissed as the consequence of iron contamination because measured fluxes of biogenic silica in deep sediment traps (Honjo et al. 1995) and a carpet of diatoms on the seafloor (Smith et al. 1996) are consistent with high diatom growth rates. Cell sinking (aggregation and buoyancy regulation) may therefore be important for understanding diatom dynamics in HNLC systems. It is notable, in this regard, that one of the diatoms identified from the seafloor and dominating a dense aggregation of phytoplankton al a convergent front at 2”N (Murray et al. 1994; Yoder et al. 1994) was a large buoyant diatom (Rhizosolenia sp.). Presumably diatoms like this are capable of exploiting the very near-surface environment, including the surface film, where aeolian dust enters the ocean and where diatoms might have a competitive advantage over light-sensitive taxa like Prochlorococcus spp. Higher iron concentrations in this layer may allow luxury uptake by diatoms for future growth (Sunda and Huntsman 1995). The striking correlation between estimates of nitrate uptake and diatom pigment biomass (Fig. 9) could also indicate a par- titioning of the phytoplankton community between ammonium and nitrate specialists, as proposed by Morel et al. (1991b). The latter would include diatoms and possibly other taxa correlated with them. If this is the case, the relative constancy of small ammonium-dependent species, now as- cribed solely to protistan grazers, might ultimately involve

416 Lmdry et al.

both N-limitation and grazing constraints. These remain during l.he JGOFS EqPac program. Deep-Sea Res. Part 2 42: open issues for future research. 77-804

The present synthesis derives many of its essential fea- tures from previous explanations of the processes that main- tain HNLC ecosystems (e.g. Miller et al. 1991; Morel et al. 1991b; Frost and Franzen 1992; Price et al. 1994), but there are also some significant differences-notably the roles as- cribed to chronic resource deficiency, the linkage of iron cycling with new and recycled production, and our distinc- tion between grazing balance and control. Our conclusions are not very different in principle from the generally accepted explanation for ecosystem processes in the oligotrophic subtropical gyres. By most accounts, the phytoplankton in such ecosystems are nutrient limited; they are dominated by small forms that grow at rapid but usually less than physiologically maximal rates; they are cropped to low stable abundances by microzooplankton grazers; and their sustained growth is dependent on the remineralized by-products of grazing. One may be limited by iron, the other by nitrogen or phosphorus (Karl et al. 1995), but we emphasize the fun- damental similarities in structure, function, and mechanistic constraints between HNLC and oligotrophic systems. We conclude that some form of resource limitation is the com- mon feature of most of the open Pacific Ocean. The low and relatively stable standing stocks of phytoplankton in such resource-limited, grazing-balanced ecosystems must depend on their being sufficiently isolated from major perturbations in the rate of supply of the limiting resource. Otherwise, they would bloom.

CARR, M.-E., M. R. LEWIS, D. KEILEY, AND B. JONES. 1995. A physical estimate of new production in the equatorial Pacific along I5O”W. Limnol. Oceanogr. 40: 138-147.

CHAI, E 1995. Origin and maintenance of high nitrate condition in the equatorial Pacific: A biological-physical model study. Ph.D. Thesis, Duke Univ. 170 p.

CHAVEZ, E I?, AND OTHERS. 1991. Growth rates, grazing, sinking and iron limitation of equatorial Pacific phytoplankton. Limnol. Oceanogr. 36: 1816-1833.

CLXVELAND, J. S., AND M. J. PERRY. 1987. Quantum yield, relative specific absorption and fluorescence in nitrogen-limited Chae- toceros gracilis. Mar. Biol. 94: 489-497.

COALE, K. II. 199 1. Effects of iron, manganese, copper, and zinc enrichments on productivity and biomass in the subarctic Pa- cific. Limnol. Oceanogr. 36: 185 1-1864.

-, S. 13. FITZWATER, R. M. GORDON, K. S. JOHNSON, AND R. T. BARI~ER. 1996n. Control of community growth and export production by upwelled iron in the equatorial Pacific Ocean. Nature 379: 621-624.

ANII OTHERS. 1996b. A massive phytoplankton bloom in- duced by an ecosystem-scale iron fertilization in the equatorial Pacific. Nature 383: 495-50 1.

CONOVER, El. J., R. DURVASULA, S. ROY, AND R. WANG. 1986. Probable loss of chlorophyll-derived pigments during passage through the gut of zooplankton, and some of the consequences. Limnol. Oceanogr. 31: 878-887.

References

AZAM, E, AND OTHERS. 1983. The ecological role of water-column microbes in the sea. Mar. Ecol. Prog. Ser. 10: 257-263.

BANSB, K. 1990. Does iron really limit phytoplankton production in the offshore subarctic Pacific? Limnol. Oceanogr. 35: 772- 775.

- 199 1. Iron availability, nitrate uptake, and exportable new production in the subarctic Pacific. J. Geophys. Res. 96: 741- 748.

-. 1995. Community response to IRONEX. Nature 375: 112. BARBER, R. T., AND OTHERS. 1996. Primary production and its reg-

ulation in the equatorial Pacific during and following the 1991- 92 El NiHo. Deep-Sea Res. Part 2 43: 933-969.

BEKLEMISHEV, K. V. 1957. The spatial relationships of marine zoo- and phytoplankton. Tr. Inst. Okeanol. Akad. Nauk SSSR 20: 253-378.

CULLEN, J. J, 1991. Hypotheses to explain high-nutrient conditions in the open sea. Limnol. Oceanogr. 36: 1578-1599.

- 1995. Status of the iron hypothesis after the Open-Ocean Enrichment Experiment. Limnol. Oceanogr. 40: 1336-l 343.

-, M. R. LEWIS, C. 0. DAVIS, AND R. T. BARBER. 1992. Photosynthetic characteristics and estimated growth rates indicate grazing is the proximate control of primary production in the equatorial Pacific, J. Geophys. Res. 97: 630-654.

DAGG, M. J. 1993. Grazing by the copepod community does not control phytoplankton production in the subarctic Pacific Ocean. Prog. Oceanogr. 32: 163-l 83.

DAM, H. G., X. ZHANG, M. BUTL~ER, AND M. R. ROMAN. 1995. Mesozooplankton grazing and metabolism at the equator in the central Pacific: Implications for carbon and nitrogen fluxes. Deep-Sea Res. Part 2 42: 735-756.

DE BAAR, H. J. W., AND OTIIERS. 1990. On iron limitation of the Southern Ocean: Experimental observations in the Weddell and Scotia Seas. Mar. Ecol. Prog. Ser. 65: 105-122.

AND OTMERS. 1995. Importance of iron for phytoplankton blooms and carbon dioxide drawdown in the Southern Ocean. Nature 373: 412-415.

BIDIGARE, R. R., AND M. E. ONDRUSEK. 1996. Spatial and temporal variability of phytoplankton pigment distributions in the central equatorial Pacific Ocean. Deep-Sea Res. Part 2 43: 809-833.

BRAND, L. E. 1991. Minimum iron requirements of marine phytoplankton and the implications for the biogeochemical control of new production. Limnol. Oceanogr. 36: 1756-1771.

-, W. G. SUNDA, AND R. R. L. GUILLARD. 1983. Limitation of marine phytoplankton reproductive rates by zinc, manganese, and iron, Limnol. Oceanogr. 28: 1182-l 198.

BRULANII, K. W., J. R. DONAT, ANL) D. A. HUTCHINS. 1991. Inter- active influences of bioactive trace metals on biological production in oceanic waters. Limnol. Oceanogr. 36: 1555-1577.

BUESSEIZR, K. O., J. A. ANDREWS, M. C. HARTMAN, R. BELAS- TOCK, AND E CHAI. 1995. Regional estimates of the export flux of particulate organic carbon derived from thorium-234

DUCE, R. A., AND N. W. TINDALE. 1991. Atmospheric transport of iron and its deposition in the ocean. Limnol. Oceanogr. 36: 1715-1726.

DUGDALE, IR. C., AND E P WILKERS~N. 1990. Iron addition experiments in the Antarctic: A reanalysis. Global Biogeochem. Cycles 4: 13-19.

EPPLBY, R. W., F. I? CHAVEZ, AND R. T BARBER. 1992. Standing stocks of particulate carbon and nitrogen in the equatorial Pa- cific at 15O”W. J. Gcophys. Res. 97: 655-661.

FALKOWSKI, F! G., R. M. GREENE, AND R. J. GRIDER. 1992. Phys- iological limitations on phytoplankton productivity in the ocean. Oceanography 5: 84-9 1.

FROST, B. W. 1987. Grazing control of phytoplankton stock in the open Pacific Ocean: A model assessing the role of mesozooplankton, particularly the large Calanoid copepods Neocalanus spp. Mar. Ecol. Prog. Ser. 39: 49-68.


- 1991. The role of grazing in nutrient-rich areas of the open sea. Limnol. Oceanogr. 36: 16 16-1630.

AND N. C. FRANZEN. 1992. Grazing vs. iron limitation in the’ control of phytoplankton stock and nutrient concentration: A chemostat analogue of the Pacific equatorial upwelling zone. Mar. Ecol. Prog. Ser. 83: 291-303.

GARGETT, A. E. 199 1. Physical processes and the maintenance of nutrient rich euphotic zones. Limnol. Oceanogr. 36: 1527- 1545.

GEIDER, R. J., AND J. LA ROCHE. 1994. The role of iron in phytoplankton photosynthesis, and the potential for iron-limitation of primary productivity in the sea. Photosynth. Res. 39: 275- 301.

GOLDMAN, J. C., J. J. MCCARTHY, AND D. G. PEAVEY. 1976. Growth rate influence on the chemical composition of phytoplankton in oceanic waters. Nature 279: 210-2 15.

GORDON, M. R., K. H. COALE, AND K. JOHNSON. 1997. Iron distributions in the equatorial Pacific: Implications for new production. Limnol. Oceanogr. 42: 419-43 1.

GRAN, H. H. 1931. On the conditions for the production of phytoplankton in the sea. Rapp. P-V. Cons. Int. Explor. Mer 75: 37-46.

GREENE, R. M., Z. S. KOLBER, D. G. SWIFT, N. W. TINDALE, AND I? G. FALKOWSKI. 1994. Physiological limitation of phytoplankton photosynthesis in the eastern equatorial Pacific determined from variability in the quantum yield of fluorescence. Limnol. Oceanogr. 39: 1061-1074.

HART, T. J. 1934. On the phytoplankton of the southwest Atlantic and Bellingshausen Sea, 1929-31. Discovery Rep. 8: l-268.

HARVEY, H. W. 1938. The supply of iron to diatoms. J. Mar. Biol. Assoc. U.K. 22: 205-219.

HEINRICH, A. K. 1962. The life histories of plankton animals and seasonal cycles of plankton communities in the oceans. J. Cons. Cons. Int. Explor. Mer 24: 455-464.

HONJO, S., J. DYMOND, R. COLLIER, AND S. J. MANGANINI. 1995. Export production of particles to the interior of the equatorial Pacific Ocean during the 1992 EqPac experiments. Deep-Sea Res. Part 2 42: 831-870.

HUDSON, R. J. M., AND E M. M. MOREL. 1990. Iron transport in marine phytoplankton: Kinetics of cellular and medium coor- dination reactions. Limnol. Oceanogr. 35: 1002-1020.

HUTCHINS, D. A., G. R. DITULLIO, AND K. W. BRULAND. 1993. Iron and regenerated production: Evidence for biological iron recycling in two marine environments. Limnol. Oceanogr. 38: 1242-1255.

KARL, D. M., AND OTHERS. 1995. Ecosystem changes in the North Pacific subtropical gyre attributed to the 1991-92 El NiHo. Na- ture 373: 230-234.

KOLBER, Z. S., AND OTHERS. 1994. Iron imitation of phytoplankton photosynthesis in the equatorial Pacific Ocean. Nature 371: 145-148.

- J. ZEHR, AND P FALKOWSKI. 1988. Effects of growth irradiance and nitrogen limitation on photosynthetic energy conversion in photosystem II. Plant Physiol. 88: 923-929.

LANDRY, M. R., J. CONSTANTINOU, AND J. KIRSI~TEIN. 1995a. Mi- crozooplankton grazing in the central equatorial Pacific during February and August, 1992. Deep-Sea Res. Part 2 42: 657- 671.

-, D. J. GIFFORD, D. L. KIRCHMAN, I? A. WHEELER, AND B. C. MONGER. 1993a. Direct and indirect effects of grazing by Neocalunus plumchrus on plankton community dynamics in the subarctic Pacific. Prog. Oceanogr. 32: 239-258.

- AND R. P HASSETT. 1982. Estimating the grazing impact of ‘marine microzooplankton. Mar. Biol. 67: 283-288.

-, J. KIRSI-ITEIN, AND J. CONSTANTINOU. 1995b. A refined dilution technique for measuring the community grazing im-

pact of microzooplankton, with experimental tests in the central equatorial Pacific. Mar, Ecol. Prog. Ser. 120: 53-63.

- -, AND -. 1996. Abundances and distributions of ‘picoplankton populations across the central equatorial Pa- cific from 12”N to 12”S, 14O”W. Deep-Sea Res. Part 2 43: 871- 890.

- AND J. M. LEHNER-FOURNIER. 1988. Grazing rates and behavior of the copepod Neoculunus plumchrus: implications for phytoplankton control in the subarctic Pacific. Hydrobiolo- gia 167: 9-19.

- B. C. MONGER, AND K. E. SELPII. 1993b. Time-depen- de;cy of microzooplankton grazing and phytoplankton growth in the subarctic Pacific. Prog. Oceanogr. 32: 205-222.

LATASA, M., M. R. LANDRY, L. SCHLUTER, AND R. R. BIDIGARE. 1997. Pigment-specific growth and grazing rates of phytoplankton in the central equatorial Pacific. Limnol. Oceanogr. 42: 289-298.

LAWS, E. A., G. R. DITULLIO, F! R. BETZER, D. M. KARL, AND K. L. CARDER. 1989. Autotrophic production and elemental fluxes at 26”N, 155”W in the North Pacific subtropical gyre. Dccp- Sea Res. 36: 103-120.

- - AND D. G. REDALJE. 1987. High phytoplankton grdwth and’ production rates in the North Pacific subtropical gyre. Limnol. Oceanogr. 32: 905-918.

LINDLEY, S. T., R. R. BIDIGARE, AND R. T. BARBER. 1995. Phyto- plankton photosynthesis parameters along 14O”W in the equatorial Pacific. Deep-Sea Res. Part 2 42: 441-463.

LINDSTROM, E. R., AND OTHERS. 1987. The Western Equatorial Pacific Ocean Study. Nature 330: 533-537.

LOPEZ, M. D. G., M. E. HUNTLEY, AND I? E SYKES. 1988. Pigment destruction by C&anus pacificus: Impact on the estimation of water column fluxes. J. Plankton Res. 10: 715-734.

MCALLISTER, C. D., T. R. PARSONS, AND J. D. H. STRICKLAND. 1960. Primary productivity and fertility at station “P” in the north-east Pacific Ocean. J. Cons. Cons. Int. Explor. Mer 25: 240-259.

MCCARTHY, J. J., C. GARSIDE, J. L. NEVINS AND R. T. BARBER. 1996. “New” production along 14O”W in the equatorial Pa- cific during and following the 1992 El Nifio event. Deep-Sea Res. Part 2 43: 1065-1093.

MARTIN, J. H. 1990. Glacial-interglacial CO, change: The iron hypothesis. Paleoceanography 5: 1-13.

- K. H. COALE, AND OTHERS. 1994. Testing the iron hy- poihesis in ecosystems of the equatorial Pacific. Nature 371: 123-129.

- AND E. E. FITZWATER. 1988. Iron limits phytoplankton grdwth in the northeast Pacific subarctic. Nature 331: 341-343.

-, AND R. M. GORDON. 1988. Northest Pacific iron distributions in relation to phytoplankton productivity. Deep-Sea Res. 35: 177-196.

-- AND S. FITZWATER. 1990. Iron in Antarctic water;. Nature’345: 156-158.

-, AND - Ockanogr. 36: 1793-1802:

1991. The case for iron. Limnol.

-- -, AND W. W. BROENKOW. 1989. VERTEX: Ph;toplank;on/iron studies in the Gulf of Alaska. Deep-Sea Res. 36: 649-680.

MILLER, C. B., AND OTHERS. 1991. Ecological dynamics in the subarctic Pacific, a possibly iron-limited ecosystem. Limnol. Oceanogr. 36: 1600-1615.

MOREL, E M. M., R. J. M. HUDSON, AND N. M. PRICE. 1991u. Limitation of productivity by trace metals in the sea. Limnol. Oceanogr. 36: 1742-l 755.

-, J. G. RUETER, AND N. M. PRICE. 1991b. Iron nutrition of phytoplankton and its possible importance in the ecology of

418 Landry et al.

open ocean regions with high nutrient and low biomass. Ocean- ography 4: 56-61.

MURRAY, J. W., R. T. BARBER, M. R. ROMAN, M. I? BACON, AND R. A. FEELEY. 1994. Physical and biological controls on carbon cycling in the equatorial Pacific. Science 266: 58-65.

-, M. W. LEINEN, R. A. FEELY, J. R. TOGGWEILER, AND R. WANNINKHOF. 1992. EqPac: A process study in the central equatorial Pacific. Oceanography 5: 134-l 42.

AND OTHERS. 1996. Export flux of particulate organic carbon from the central equatorial Pacific determined using a combined drifting trap-*““Th approach. Deep-Sea Res. Part 2 43: 1095-l 132.

PARSONS, T. R., L. E GIOVANDO, AND R. J. LEBRASSEUR. 1966. The advent of the spring bloom in the eastern subarctic Pacific Ocean. J. Fish. Res. Bd. Can. 23: 539-546.

PE~‘ERS, H., M. C. GREGG, AND J. M. TOOLE. 1988. On the param- eterization of equatorial turbulence. J. Geophys. Res. 93: 1199-1218.

PRICE, N. M., B. A. AHNER, AND E M. M. MOREL. 1994. The equatorial Pacific Ocean: Grazer-controlled phytoplankton populations in an iron-limited ecosystem. Limnol. Oceanogr. 39: 520-534.

---, L. E ANDERSEN, AND E M. M. MOREL. 1991. Iron and nitrogen nutrition of equatorial Pacific plankton. Deep-Sea Res. 38: 1361-1378.

ROMAN, M. R., AND A. L. GAUZENS. 1997. Copepod grazing in the equatorial Pacific. Limnol. Oceanogr. 42. In press.

SHARP, J. H., M. J. PERRY, E. H. RBNGER, AND R. W. EPPLEY. 1980. Phytoplankton rate processes in the oligotrophic waters of the central North Pacific. J. Plankton Res. 2: 335-353.

SMITH, C. R., AND OTHERS. 1996. Phytodetritus at the abyssal seafloor across 10” of latitude in the central equatorial Pacific. Deep-Sea Res. Part 2 43: 1309-1338.

STONE, L., AND T. BERMAN. 1993. Positive feedback in aquatic ecosystems: The case of the microbial loop. Bull. Math. Biol. 55: 919-936.

AND R. S. J. WEISBURD. 1992. Positive feedback in aquatic ecosystems. Trends Ecol. Evol. 7: 263-267.

STROM, S. IL., AND N. A. WELSCHMEYER. 1991. Pigment-specific rates of phytoplankton growth and microzooplankton grazing in the open subarctic Pacific Ocean. Limnol. Oceanogr. 36: 50- 63.

SUNDA, W. G., AND S. A. HUNTSMAN. 1995. Iron uptake and growth limitation in oceanic and coastal phytoplankton. Mar. Chem. 50: 189-206.

VAULOT, D , D. MARIE, R. J. OLSON, AND S. W. CHISHOLM. 1995. Growth of Prochlorococcus, a photosynthetic prokaryote, in the equatorial Pacific Ocean. Science 268: 1480-1482.

VERITY, F? G., D. K. STOECKER, M. E. SIERACKI, AND J. R. NELSON. 1996. Microzooplankton grazing of primary production in the equatorial Pacific. Deep-Sea Res. Part 2 43: 1227-1255.

WATSON, A. J., AND OTHERS. 1994. Minimal effect of iron fertilization on sea-surface carbon dioxide concentrations. Nature 371: 143-145.

WELSCHME‘(ER, N., R. GOERICKE, S. STROM, AND W. PB~ERSON. 1991. Phytoplankton growth and herbivory in the subarctic Pacific : A chemotaxonomic analysis. Limnol. Oceanogr. 36: 1631-1649.

WELLS, M. L., AND L. M. MAYER. 1991. The photoconversion of colloidal iron oxyhydroxides in seawater. Deep-Sea Res. 38: 1379-l 395.

‘WHEELER, I? A., AND S. A. KOKKINAKIS. 1990. Ammonium recycling limits nitrate use in the oceanic subarctic Pacific. Limnol. Oceanogr. 35: 1267-1278.

WHITE, J. I;:., X. ZHANG, L. A. WELLING, M. R. ROMAN, AND H. G. DAIV~. 1995. Latitudinal gradients in zooplankton biomass in the tropical Pacific at 14O”W during the JGOFS EqPac study: Effects of El NiHo. Deep-Sea Res. Part 2 42: 715-733.

ZHANG, X., H. G. DAM, J. R. WHITE, AND M. R. ROMAN. 1995. Latitudinal variations in mesozooplankton grazing and metabolism in the central tropical Pacific during the U.S. JGOFS EqPac study. Deep-Sea Res. Part 2 42: 695-714.

YODER, J. A., S. G. ACKEI.SON, R. T BARBER, I? FLAMENT, AND W. M. BALCH. 1994. A line in the sea. Nature 371: 689-692.

Received: I6 November 1995 Accepted: 22 September 1996

Amended: 14 January 1997

landry, michael r., et al. iron and grazing constraints on … · 2015. 7. 29. · oceanography may...

Documents