Grazing and Zooplankton Production As Key Controls of Phytoplankton Production in the Open Ocean
Total Page:16
File Type:pdf, Size:1020Kb
FEATURE GRAZING AND ZOOPLANKTON PRODUCTION AS KEY CONTROLS OF PHYTOPLANKTON PRODUCTION IN THE OPEN OCEAN By Karl Banse THANKS TO NASA's Coastal Zone Color Scan- Phytoplankton Concentrations and Seasonality ner (CZCS, 1978-1986), nearly ocean-wide cov- Regions Without Marked Phytoplankton erage of the distribution of phytoplankton pig- Seasonality ment in the upper part of the euphotic zone and, In much of the open ocean, the seasonal change in effect, in the mixed layer of the open ocean, of phytoplankton pigment and, by implication, of is now available. This coverage includes season- biomass is small. Figure 1 juxtaposes monthly me- ality and interannual variability. The CZCS ob- dians of CZCS-derived pigment of a cool-temper- servations confirm that the physics of the ocean ate area (Fig. la; see Frost, 1991, for regional provide the backbone of the geographic and background), an equatorial area (Fig. lb; see Bar- temporal patterns of pigment distribution and, ber and Chavez, 1991), and a subtropical area by inference, of primary production rate. For ex- (Fig. lc; see Venrick et al., 1987). The first two ample, phytoplankton concentrations and pro- areas are persistently nutrient-rich, the last is per- duction are enhanced where upwelling or sea- sistently nutrient-depleted. Patterns similar to sonal overturn of the water column replenishes those in Fig. la prevail in the subantarctic water nutrient concentrations in the mixed layer, ring (Banse, unpub, observations) that encom- whereas the timing of this enhancement may be passes about N0 of the ocean area; patterns similar controlled by upwelling, seasonal overturn, or to those in Fig. lc prevail in all subtropical central the mixed layer becoming shallower than the gyres that I estimate comprise close to '/3 of the critical depth. ocean area (cf. Banse and English, 1994). I propose that the next task is to understand Figure l a shows that a marked seasonal cycle the cause of the phytoplankton concentrations of incident light in the year-round presence of nu- I propose that the and to be able to predict them and the temporal trients, as observed at -50°N, does not necessarily next task is to rate of change of concentrations on the scale of lead to clear seasonality of phytoplankton, e.g., to several days to seasons. I will show that this spring and fall blooms of several mg m 3 of pig- understand the cause major challenge for biological oceanography ment that supposedly are typical for temperate lat- of the phytoplankton cannot be addressed without a vastly improved itudes. (The 2 values > 1 mg m -3 occur during a understanding of the zooplankton. Understanding season when inaccurate values were recorded by concentrations and to the animals is of intrinsic interest, which is also the CZCS in 1980, Banse and English, 1994). At be able to predict true for phytoplankton. However, the task of pre- the subtropical site (Fig. lc), the trend to seasonal- dicting phytoplankton concentrations and their ity is caused by enhanced vertical mixing and nu- them and the rates of change also may be considered as a sub- trient supply during the cool season. Further, com- set of the challenge of anticipating the effect of parison of Figure 1, b and c indicates that temporal rate of climate change on, e.g., the geographic and tem- drastically different nutrient (N, P) concentrations change of poral (seasonal) distribution of the sign of the have little effect on pigment levels; the same small CO2 gradient between atmosphere and sea via effect holds for the rate of nutrient supply from concentrations on the the connection between phytoplankton concen- below the thermocline at the two sites. The low tration and photosynthesis, or on plankton com- scale of several days pigment levels in situations rich in N and P as in munity composition and its feedback to the at- Figure 1, a and b are not likely to be caused by to seasons. mosphere from changes in the distributions of lack of iron. In or near these areas, phytoplankton DMS producers. enclosed in the control flasks (no iron added) of the recent iron bioassays (mostly by Martin et al., 1991) doubled to quadrupled in biomass over a K. Banse, School of Oceanography, WB-10, University of few days (Martin et al., 1991 for the eastern tropi- Washington, Seattle, WA 98195, USA. cal Pacific; see Banse, 1991a, for the southern OCEANOGRAPHY'VoI.7, NO. l'1994 13 (3 b 1- 0.5 • 1" 1" 0.8- 0.4t i E c~ 0.6 0.3- E o 4> ^ • E]~ E (1.) E 0.4 0.2. C~ {3. • o [] [] 0.2 0.1. • ]t A - <> ~ Vt ~ ~,& 8 1'o 1'2 monlh of yeor month of year C 0.5 d 0.4 • 1978 [] 1982 0 1985 'E tm 0.3 E E • 1979 0 1983 ,O, 1986 (D [] E 0.2 13_ [] • <> <> [] o • • 1988 /k 1984 -- median 0.1 • 1981 o o 2 ~, 6 8 1'o 1'2 month of yeQr Fig. 1: Monthly medians of CZCS-derived pigment m three oceanic domains for 9 years. Medians for each month are connected by lines• The areas are approximately parallelograms with the coordinates of the southwestern, southeastern, and northeastern corners indicated. (a) Subarctic Pacific south of Gulf of Alaska, 45°N, 155°W; 45°N, 138°W; 50°N, 144°W (the last coordinates approximately those of the for- mer Ocean Weather Station PAPA; note reduced scale of ordinate and omitted values [arrows at top margin], 1.16 mg m 3for Sept. 1983 and 1.71 for Oct. 1981). (b) Eastern equatorial Pacific, 6°S, 124°W; 6°S, 97°W; 4°N, 97°W. (c) Central North Pacific around CLIMAX area, 25°N, 160°; 25°N, 144°W; 30°N, 148°W. (d) Symbols for years. (see Banse and English, 1994, for methods). • . the rate of Gulf of Alaska and Antarctic sites), despite more, horizontal advection will average out the ef- change of demonstrated adsorption of iron to the container fect of patchiness of pigment. Thus the horizontal walls. terms in the equation in Figure 2 can be neglected. phytoplankton As to the understanding of the rate of temporal Further, Figure 1 shows that the percentage daily change of pigment due to seasonality, if any, is concentration is change of pigment, only in the last few years have accurate growth rates of bulk phytoplankton be- very small so that d[chl]/dt in Figure 2 can be set principally come available, which permit us to draw up a bal- to zero. Then, the measured rates of cell division k are balanced by the three vertical terms and mor- understandable as ance sheet (Fig. 2; see Banse, 1992, for sources supporting the following discussion). In the geo- tality, g. Since each of the vertical terms is known the balance between graphic domains identified (Fig. 2, bottom part), to be equivalent to a few percent per day, the rate horizontal gradients of pigment are slight, espe- of change of phytoplankton concentration is prin- the rates of cell cially in the zonal direction. Also, considering the cipally understandable as the balance between the division and mortality. time change at a fixed station over a few days or rates of cell division and mortality. 14 OCEANOGRAPHY°VoI. 7, No. 1o1994 mixing, advection _ [Chl]u .I ~ upper ( = euphoric) t adv. mix. sink. ~ lower layers [Chl]l = 0 ~l<,,temp' d [Chl]u dt - [Chl]u (k - a - m - s - g + x) ~, 0 vertical "be k a m s g SUBANTA RCTIC 0.25 0.04 0.05? -<0.02 0.t57 ( Summer) (0.5)-~ SUBARCTIC 0.5 <0.0t ~0.03 0.0t ~0.45 ("Papa", Summer) (0.8)"~ SUBTROPICAL GYRE 1.2 <0.01 <0.01 _<0.01 ~1.2 {t.7)* EQUATORIAL 1.0 0.04 0.02 0.01 ~0.9 UPWELLING (t. 4)* ~max. instant,growth rate Fig. 2: The temporal balance of rates of algal growth and loss in the mixed layer at 4 representative sites of the open sea. The equation, based on the model above, compares the instantaneous rates of growth (k) and loss (a, m, s, g, x, all as d ~) for bulk phytoplankton in the euphotic zone (= mixed layer), which is as- sumed to overlie phytoplankton-free water. The maximal instantaneous rates for the respective tempera- tures also are given, mostly representing diatoms at replete nutrients and light, with varying daylengths taken into account. To obtain division rates (d-~), divide by 0.69; the doubling times (d) are the inverse of the division rates (from Banse, 1992, with permission of Plenum Press). In normal, near-steady state situations with atom cells when counting phytoplankton under the • . cell mortality high algal division rates, as in the domains of Fig- microscope. Even these cells might have died ure 2, cell mortality mainly is caused by grazing. from grazing by heterotrophic dinoflagellates with mainly is caused by Most of the grazing is due to small, unicellular an- digestion outside the cell body, i.e., without break- grazing. imals (protozoans) with division rates that are po- age of the prey (cf. Jacobson and Anderson, tentially similar (given sufficient food) to those of 1986)• the phytoplankton or even higher, so that the ani- Note regarding the grazing term in Figure 2, mals quickly can adjust the grazing pressure to an that g (d ') equals the fraction of water (liter/liter) enhanced food supply and the underlying phyto- that is being cleared of particles per day. With di- plankton production rate. The only other source of visions approaching 2 d ' (3rd line of Fig.