Carbon cycling in the upper waters of the Sargasso Sea

Carbon cycling in the upper waters of the Sargasso Sea

Material Information

Carbon cycling in the upper waters of the Sargasso Sea
Bissett, W. Paul
Place of Publication:
Tampa, Florida
University of South Florida
Publication Date:
Physical Description:
ix, 203 leaves : ill. ; 29 cm.


Subjects / Keywords:
Carbon cycling (Biogeochemistry) -- Computer simulation -- Sargasso Sea ( lcsh )
Optical oceanography ( lcsh )
Dissertations, Academic -- Marine Science -- Doctoral -- USF ( FTS )


General Note:
Includes vita. Thesis (Ph. D.)--University of South Florida, 1997. Includes bibliographical references (leaves 186-203).

Record Information

Source Institution:
University of South Florida
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
023911174 ( ALEPH )
37642128 ( OCLC )
F51-00200 ( USFLDC DOI )
f51.200 ( USFLDC Handle )

Postcard Information



This item is only available as the following downloads:

Full Text


CARBON CYCLING IN THE UPPER WATERS OF THE SARGASSO SEA by / W. PAUL BISSETT A dis se rtation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Marine Science University of South Florida April1997 Major Professor: John J Walsh, Ph.D.


Graduate Schoo l University of South Florida Tampa, Florida Ph.D Dissertation This is to certify that the Ph.D. Dissertation of W. PAUL BISSETT with a major in Marine Science has been approved by the Examining Committee on December 4, 1996 as satisfactory for the dissertation requirement for the Doctor of Philosophy degree Examining Committee: Major Professor: John J.Wa l sh, Ph.D Member: Kendall L. Carder Ph D Member: Denni s A. Hansell, Ph.D. Member: Frank E. Mtiller Karger, Ph.D. Member: Gabriel A. Vargo Ph.D.


Copyright by W. Paul Bissett All rights reserved


Acknowledgements Financial support for this dissertation wa s provided by : National Aeronautic s and Space Administration Office of Naval Research National Science Foundation Rotary International Bermuda Biological Station f or Re s earch John Lake Foundation University of South Florida William and Pamela Bi ss ett This dissertation is th e culmination of s everal years of research. While it has only one author, it could not have been accomplished without help from many people This help included hours of discussions with my committee members, Drs. J. J. Walsh, K. L. Carder, D A. Hansell, F E Miiller-Karger, and G A. Vargo, as well as D A Dieterle Any errors or omis s ion s are solely the responsibility of the author.


Table of Contents List of Table s . . . . . . . . . . . . . . . . . . . . . . . . 111 List of Figures iv Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . viii 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Overview and broad objectives . . . . . . . . . . . . . . 1 1.2 Tools of research and specific objectives . . . . . . . . . . . 5 2. Biological Model . . . . . . . . . . . . . . . . . . . . . . 9 2.1 Biological Method s . . . . . . . . . . . . . . . . . . 9 2.1.1 Biological Methods Overview . . . . . . . . . . . 9 2.1.2 Phytoplankton State Equations . . . . . . . . . . . 10 light-limted growth rate . . . . . . . . . . 13 nutri e nt -limi ted growth rate . . . . . . . . . 18 biomass los s terms . . . . . . . . . . . . 23 2 .1. 3 Bacterioplankt on State Equations . . . . . . . . . . 26 2 1.3 1 carbon-limited growth rate . . . . . . . . . 27 nitrogen-limited growth rate . . . . . . . . . 28 total carbon uptake . . . . . . . . . . . . 30 2. 1.3 .4 bacterial loss terms . . . . . . . . . . . . 32 2.1.4 Dissolved Organic Matter State Equations . . . . . . . 32 DOC1 32 relict DOC 2 . . . . . . . . . . . . . . 36 2.1.5 Fecal Pellet State Equations . . . . . . . . . . . . 37 2.1.6 Nitrogen State Equations . . . . . . . . . . . . . 40 2.1.7 Dissolv ed Inorganic Carbon State Equation . . . . . . . 41 2.1.8 Physical Overview . . . . . . . . . . . . . . . 44 vertical mixing . . . . . . . . . . . . . 44 2 .1. 9 Initialization . . . . . . . . . . . . . . . . . 45 2.1.10 Boundary Conditions and Numerical Considerations . . . 48 2.2 Biological Results . . . . . . . . . . . . . . . . . . . 49 2.2.1 The Strong Mixing, Simple Biological Case Case 1 . . . . 50 organic . . . . . . . . . . . . . . . . 50 inorganic . . . . . . . . . . . . . . . 65 2.2.2 The Weak Mixing, Simple Biological Case-Case 2 . . . . 66 organic . . . . . . . . . . . . . . . . 66 inorganic . . . . . . . . . . . . . . . 80 2.2.3 The Weak Mixing, Complex Biological Case-Case 3 . . . 80 2.2.3 1 organic . . . . . . . . . . . . . . . . 80 inorganic . . . . . . . . . . . . . . . 98 2.3 Biological Discu ss ion . . . . . . . . . . . . . . . . . 102


2.3.1 Model Fidelity ............................... 2.3.2 The Complex Biological Model .................... di sso lved organic matter .................. nitrification ........................... particulate carbon and nitrogen regeneration .... nitrogen fixation ................ ........ dissolved inorganic carbon ........... ...... 2.4 Summary 3. Bio-Optical Model ......................................... 3.1 Bio-Optical Methods ............................... ... 3 .1.1 Bio-Optical Overview ........................... 3.1.2 Light Energy at the Se a Surface .................... 3.1.3 Inherent Optical Properti es ....................... absorption-water ....................... absorption-particulate ..... ............ . absorption-CDM (C DOC ) ................. backscattering ......................... 3.1.4 Apparent Optical Properties ....................... 3.1.5 CDOC (CDM) Photolysis ............. . ......... 3.2 Bio-Optical Results ................................... 3.2.1 Light Energy at the Sea Surface .................... 3.2.2 Absorption .................................. 3.2.3 Spectral Diffu se Attenuation Coefficient .............. 3.3 Bio-Optical Discussion ................................ 3.3.1 Seasonal and Vertical Cycles of th e Diffuse Attenuation Coefficients . . . . . . . . . . . . . . . . 3.3.2 Accessory Pigment Effect on . .... . . 3.3.3 Potential CDM Interference of Remotely-Sen se d Chlorophyll a .............................. . 3.4 Summary . ....................... ......... ....... 4. The Big Picture ..................... ....... . ............. 4.1 Overview ......................................... 4.2 Conceptual framework .. .............. ........ ....... 4.2.1 EcoSim ..................... .. .. .. ......... 4.3 Where does DIC accumulate? ..... ...... ................ 4 4 Summary ...... ................................. 5. Conclusions 102 110 110 113 115 118 119 120 122 122 122 1 23 124 124 125 129 136 138 140 142 142 142 148 152 152 156 164 166 167 167 168 173 176 183 184 Reference s . . . . . . . . . . . . . . . . . . . . . . . . 186 Vita . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 11


List of Tables Table 201 0 Maximum Functional Group Growth Rates 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 13 Table 2020 Ph ytop lankton Parameters for the Comp l ex Model 0 0 0 0 0 0 0 0 0 0 0 0 0 0 47 Table 2 03ao Ph ytop lankton Parameter s for the Simple Model 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 50 Table 2.3bo Conceptual differences betwe en the Simp l e and Complex Models 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 50 T able 2.4ao Integrated Concentrations 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 99 Table 2.4bo Average Daily Carbon Fluxe s 100 Table 2 .4c o Average Daily Nitrogen Fluxes 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 100 T a bl e 2050 WIC versus L1DIN beneath t h e Euphotic Zone 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 116 Table 3 olo Parameter s for Li g ht Limited Carbon to Chlorophy 11 a Ratios 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 126 Table 3 020 Parameter s for Nutrient-Limited Carbon to Chloroph yll a Rati os 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 127 Tabl e 3030 Accessory Pigment s as a Function of Chlo r ophyll a 0 0 0 0 0 0 0 0 0 0 0 0 128 Tabl e 4010 C02 Fluxe s and Inte gra t ed Changes in DIC Concentrations 0 0 0 0 0 0 0 1 77 ill


List of Figures Figure 1.1. Typical dissolved inorganic profiles for each of the world's oceans . . . . . . . . . . . . . . . . . . . . . 4 Figure 1.2. Bermuda Atlantic Timeseries Study site . . . . . . . . . . 7 Figure 2.1. Interpolated NODC (a) temperatures and (b) salinities used to drive the model . . . . . . . . . . . . . . . . . 46 Figure 2.2. Case 1 integrated primary productivity for each functional group and the total integrated primary productivity . . . . . . . 51 Figure 2.3. Case 1 (a) downwellin g irradiance at noon; (b) vertical eddy diffusivity ; (c)-(f) realized growth rates for functional groups 1 4 respectively . . . . . . . . . . . . . . . 52 Figure 2.4. Case 1 (a) ( d ) carbon biomass for functional groups 1 4, respectively; (e) total phytoplankton carbon biomass; (f) fecal pe ll et carbon stocks . . . . . . . . . . . . . . . 54 Figure 2.5. Case 1 (a)-(d) carbon t o nitrogen ratios for functional groups 1 4, respectively ; (e) total phytoplankton carbon to nitrogen ratios; (f) fecal pellet carbon to nitrogen ratios . . . . . 57 Figure 2.6. Case 1 (a) ( d ) chlorophyll a concentrations for functional groups 1 -4, respectively; (e) total chlorophyll a stocks . . . . . 59 Figure 2.7. Ca s e 1 accessory pigment concentrations(a) and (b ) c hl orophyll b concentrations for functional groups 1 and 2, re s pectively ; (c) functional group 3 phycoerythrin concentrations; (d) and (e) functional group 4 chlorophyll c and photosynthetic carotenoid concentra t io n s . . . . . . . . 61 Figure 2 8. Case 1 (a) dissolved inorganic carbon stocks; (b) n itrate stocks; and (c) ammonium stocks . . . . . . . . . . . . 63 Figure 2 9. Case 2 (a) downwelling irradiance at noon; (b) vertical eddy diffusivity; (c)-(f) realized growth rates for functional groups 1 4 re spectively . . . . . . . . . . . . . . . 67 Figur e 2.1 0. Case 2 integrated primary productivity for each functional gro up and the total integrated primary productivity . . . . . . . 69 IV


Figure 2.11. Case 2 (a)-(d) carbon biomass for functional groups 1 4, respectively; (e) total phytoplankton carbon biomass ; (f) fecal pellet carbon stocks . . . . . . . . . . . . . . . 70 Figure 2.12. Case 2 (a)-(d) chlorophyll a concentrations for functional groups 1 -4, respectively; (e) total chlorophyll a stocks . . . . . 72 Figure 2.13. Case 2 accessory pigment concentrations(a) and (b) chlorophyll b concentrations for functional groups 1 and 2 respectively; (c) functional group 3 phycoerythrin concentrations; (d) and (e) functional group 4 chlorophyll c and photosynthetic carotenoid concentrations . . . . . . . . 7 4 Figure 2.14. Case 2 (a)-(d) carbon to nitrogen ratios for functional groups 1 -4 respectively; (e) total phytoplankton carbon to nitrogen ratios; (f) fecal pellet carbon to nitrogen ratios . . . . . 77 Figure 2.15. Case 2 (a) dissolved inorganic carbon stocks; (b) nitrate stocks; and (c) ammonium stocks . . . . . . . . . . . . 79 Figure 2.16. Case 3 (a) integrated primary productivity for each functional group and the total integrated primary productivity; and (b) daily integrated nitrogen-fixation rates . . . . . . . . . . . 81 Figure 2.17. Case 3 (a) downwelling irradiance at noon; (b) vertical eddy diffusivity; (c)-(f) realized growth rates for functional groups 1 4, respectively . . . . . . . . . . . . . . . 83 Figure 2.18 Case 3 (a)-(d) carbon biomass for functional groups 1 -4, respectively; (e) total phytoplankton carbon biomass; (f) fecal pellet carbon stocks . . . . . . . . . . . . . . . 85 Figure 2.19 Case 3 (a)-(d) chlorophyll a concentrations for functional groups 1 4 respectively ; (e) total chlorophyll a stocks . . . . . 87 Figure 2.20 Case 3 accessory pigment concentrations(a) and (b) chlorophyll b concentrations for functional groups 1 and 2, respectively; (c) functional group 3 phycoerythrin concentrations; (d) and (e) functional group 4 chlorophyll c and photosynthetic carotenoid concentrations . . . . . . . . 89 Figure 2.21. Case 3 (a)-(d) carbon to nitrogen ratios for functional groups 1 4, respectively; (e) total phytoplankton carbon to nitrogen ratios; (f) fecal pellet carbon to nitrogen ratios . . . . . 91 Figure 2.22 Case 3 (a) dissolved inorganic carbon stocks; (b) nitrate stocks; and (c) ammonium stocks . . . . . . . . . . . . 94 F ig ure 2 23. Case 3 (a) labile dissolved organic carbon; and (b) relict dissolved organic carbon; carbon to nitrogen ratio for (c) labile dissolved organic matter; and (d) relict dissolved organic matter . . . . . . . . . . . . . . . . . . . 96 Figure 2.24. Case 3 daily integrated nitrification rates 101 v


Figure 2 25. Time series of integrated primary production (circles) and 150m trap fluxes (squares) at the BATS site . . . . . . . . 103 Figure 2.26. Time series of phytoplankton pigments at the BATS site from December 1989 to June 1990: (a) chlorophyll a; (b) chlorophyll b; (c) zeaxanthin; (d) 19'butanoyloxyfucoxanthin; (e) fucoxanthin; (f) 19'hexanoyloxyfucoxanthin; (g) chlorophyll c1+2; (h) chlorophyll c3 . . . . . . . . . . . . . . . . . . 104 Figure 2.27 Time series of nitrate stocks at the BATS site . . . . . . . . 107 Figure 2.28. Six year average of CZCS estimated chlorophyll a concentrations at the BATS site compared against Case 1 and 3 surface chlorophyll a values . . . . . . . . . . . . . 109 Figure 2.29. Seasonal dissolved organic carbon stocks at the BATS site . . . 112 Figure 2 30. Nitrite profiles from the BATS study site . . . . . . . . . 114 Figure 2 31. Dissolved inorganic carbon on Julian day 0 (solid line) and 365 (dashed line) from Case 3 of the simulation . . . . . . . 121 Figure 3 .1. Pigment-specific absorption curves used in the model 130 Figure 3 2. Chlorophyll a-specific absorption curves (a)-(d) for functional group 1 4, respectively . . . . . . . . . . . 131 Figure 3.3. (a) The simulated clear sky irradiance at the sea surface (solid line) and the sea surface irradiance (dashed line) estimated from the Comprehensive Ocean Atmosphere Data Set that is assumed to integrated the effects of clouds; and (b) the simulated downwelling irradiance at 1.25 m . . . . . . 143 Figure 3.4. Simulated (a) particulate absorption and (b) colored detrital material at 442 nm . . . . . . . . . . . . . . . . . 144 Figure 3 5. Photoprotective carotenoids (a)-(d) for functional groups 1 4, respectively . . . . . . . . . . . . . . . . . . 146 Figure 3.6. Diffuse downwelling attenuation coefficients at (a) 412 nm and (b) 442 nm . . . . . . . . . . . . . . . . . . 149 Figure 3 7. Diffuse downwelling attenuation coefficients at (a) 467 nm and (b) 487 nm . . . . . . . . . . . . . . . . . . 150 Figure 3.8. Diffuse downwelling attenuation coefficients at (a) 522 nm and (b) 567 nm . . . . . . . . . . . . . . . . . . 151 Figure 3 9. The ratio of to which is indicative of CDM interference The dashed line is the simulated depth of the mixed layer . . . . . . . . . . . . . . . . . . . 153 Vl


Figure 3.10. S i mulated (a) particulate absorption and (b) colored detrital material at 412 nm . . . . . . . . . . . . . . . . . 157 Figure 3.11. Simulated (a) particulate ab s orption and (b) colored detrital material at 487 nm . . . . . . . . . . . . . . . . . 158 Figure 3.12 The chlorophyll a -s pecific ab so rption during the (a) sp rin g peak and (b) summer peak in su bsurface chlorophyll a concentrations . . . . . . . . . . . . . . . . . . 159 Figure 3.13. Comparison of the chlorophyll a-specific absorption s from the surface and subsurface chlorophyll a maximum. Demonstrate s the shift in functional groups over the course of the year . . . . . . . . . . . . . . . . . . . 160 Figure 3.14. S easonal ratio of divinylc hloroph y ll a to total chlorophyll a concentration s around Bermuda . . . . . . . . . . . . 162 Figure 3.15. (a) Ratio of simulated CDM absorption to particulate absorption at 442 nm ; (b) s imulated s urface chlorophyll a concentrations (so lid line ) and s ix year average of CZCS e s timated chlorophyll a concentration at the BATS s ite (dashed line); and (c) ratio of CZCS estimated chlorophyll a to s imu l ated chlorophyll a (solid line). . . . . . . . . . . 165 Figure 4 .1. Grap hi c depiction of differe n ces betwee n re l at i ve C02 flu x during (a) cold deep winter mixing, and (b) warm, s hallow s ummer mixin g . . . . . . . . . . . . . . . . . . 169 Figure 4.2. Seasonal pr o file s of DIC. Da she d line s s how anthropogenic C02 increm e nt . . . . . . . . . . . . . . . . . . 170 Figure 4.3. TC02 concentrations from BATS . . . . . . . . . . . . 174 Figure 4.4. Residual differences between predi c ted and observed DIC concentration s in th e South Atlantic . . . . . . . . . . . 175 Figure 4 .5. Case A (a) DIC profi l es on Julian day 0 (so lid line ) and Julian day 365 ( dashed line ); and ( b) change in DIC concentration with depth . . . . . . . . . . . . . . . 178 Figure 4.6. Cas e B (a) DIC profile s on Ju l ian day 0 (s o li d l ine) and Julian day 365 (das hed line) ; and ( b ) change in DIC co ncentration with depth . . . . . . . . . . . . . . . 179 Figure 4. 7. Case C (a) DIC profile s on Julian day 0 (solid line ) and Julian day 365 ( dashed line) ; and ( b) change in DIC concentration with depth . . . . . . . . . . . . . . . 1 8 1 Fi g ure 4 .8 C aseD (a) DIC profiles on Julian day 0 (solid line) and Julian day 365 ( dashed line); and (b) change in DIC concentration with depth . . . . . . . . . . . . . . . 1 82 vii


CARBON CYCLING IN THE UPPER WATERS OF THE SARGASSO SEA by W. PAUL BISSETT An Abstract Of a di ssertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philos op hy D e partment of Marine Science University of South Florida April1997 Major Profes sor: John J. Walsh, Ph D. Vlll


A complex ecosystem model is developed to test the hypothesis that anthropogenic C02 sequestration occurs in the upper waters of the Sargasso Sea. The model is physically driven by seasonal changes in spectral light, temperature, and water column mixing. Autotrophic growth is represented by four functional groups of phytoplankton. The groups have lightand nutrient-utilization characteristics that reflect those of Prochlorococcus, Synecho cocc us and Chromophycota species. The model includes differential carbon and nitrogen cycling, nitrification and nitrogen-fixation to effect a realistic allochthonous and autochthonous nutrient source to the euphotic zone. This simulation yields realistic seasonal and vertical succession of phy t oplankton functional groups, as well as estimates of bioma ss, productivity, air -sea C02 exchange, settling losses, and diffuse attenuation coefficients The results also suggest that the Coastal Zone Color Scanner (CZCS) estimates of s urface chlorophyll a stocks may have been contaminated by co l ored detrital material in the Sargasso Sea during the late spring and early fall of e ach year. The s imulation predict s yearly C02 influxes of 0.59 mol m-2. at an atmospheric C02 concentration of 355 ppm Of thi s 0.24-0.30 mol m-2 (41 52%) accumulates between 150500 m. This suggests a large fraction of the oceanic C02 uptake in oligotrophic gyres is being sequestered at relatively shallow depths. The return of thi s sequ es tered C02 to the surface mixed layer is capped at the base of the euphotic zo ne by biological processes that preferentially recycle nitrogen relativ e to carbon. Abstract Approval: M ajor Profe sso r : John J. Walsh, Ph D. Professor, Department of Marine Science Date Approved: ___ _____ lX


1. Introduction 1.1 Overview and broad objectives Since the beginning of the industrial revolution, atmospheric concentrations of C02 have increased at almost an exponential rate, from approximately -280 parts per million by volume (pprnv) in the late 1800's to a 1994 estimate of 358 pprnv This amounts to an addition of -165 Pg C (Pg C = 1015 gram s carbon as C02) by humans to the atmospheric inventory (Moore and Braswell, 1994; IPCC, 1996). The earth is heated by the energy it receives from the sun (in the form of short-wave radiation). This heat is radiated back to space at a lower energy level (long-wave radiation). As a "greenhouse gas, C02 is fairly transparent to the short-wave energy arriving from the sun, but traps the long-wave energy leaving the earth, providing a thermal blanket for the earth. One of the net results of this anthropogenic increment of C02 is an increase in the Earth's trapping of long-wave radiation, resulting in a warmer atmosphere The warming response to this increased atmospheric concentration of C02 along with other greenhouse gases and aerosols, is an estimated 0 9 to 3.5 C change in global temperatures by 2100 (IPCC, 1996). The predicted global changes wrought by such atmospheric increases include rising sea levels and altered weather patterns that have tremendous socioeconomic implications. There are many difficulties to overcome when trying to project the magnitude of increased temperatures (IPCC, 1996). A lar ge part of the potential inaccuracies stern from the inability to estimate the future levels of C02 in the atmosphere This is due, in part, to difficulties in balancing the global C02 budget. For years, researchers added up the amount of C02 produced by the burning of fossil fuels and deforestation, subtracted the increment within the atmosphere, limited the oceanic sink to the amount of new production


( c.f. Eppley and Peterson, 1979), and determined that there was a "missing" sink, since this budget did not balance (i.e. less accumulated in the atmosphere than was expected) So this missing sink has led to a scientific race to determine where this carbon is going, and what possible feedback effects exist for this sink (Sarmiento 1993). This intensive effort to describe the global carbon cycle has yielded many observations that have tightened the budget. Measurements of the isotopic C02 distribution in the oceans (Quay et al., 1992) and 02 evolution in the atmosphere (Keeling and Shertz, 1992) appear to have narrowed the range of estimates for the recent oceanic carbon sink. With a firm estimate for the oceanic sink the terrestrial sink can be solved for by reconstruction (simply assuming that all the missing carbon is somewhere in the terrestrial biotaPost et al., 1990). A recent attempt at balancing the current C02 budget (Siegenthaler and Sarmiento, 1993; IPCC 1996) suggests that 2.0 Pg C of the 5 5 Pg C released annually from fossil fuel burning is sequestered in the oceans, while the remaining 3.5 Pg Cremains in the atmosphere. The net deforestation of tropical forests is approximately balanced by the net reforestation of temperate forests and increased carbon fixation by C3 plants (Mooney and Koch, 1994). The oceans would appear to be a major sink of anthropogenic C02, accounting for -42% of the annual release of C02 from the burning of fossil fuels. The remaining -58% accumulates in the atmosphere each year. Most of this atmospheric increment is merely temporary, as the oceans could buffer all of the anthropogenically released C02 quite easily. In fact, the oceans could buffer the burning of the entire -4000 Pg C of easily recoverable fossil fuel C02 mostly through the dissolution of CaC03 (Broecker and Peng, 1982), as this recoverable fossil fuel represents only -10% of the total oceanic carbon pool of -40,000 Pg C (However, at present there is little evidence that net dissolution of marine CaC03 has begun (Broecker et al., 1979) ) It is the mismatch of the timing between the annual release of C02 and its eventual sequestration that makes it difficult to 2


project future atmospheric C02 concentrations over the time scale of socioeconomic concerns. As a result, projections of increases in global heating are difficult to make. While the present annual C02 budget appears to be balanced there is considerable uncertainty surrounding the future oceanic uptake of C02. From a strictly chemical standpoint, uptake by the ocean surface waters should slow due to a decrease in buffering capacity as C02 increases This effect has been demonstrated in a simple box diffusion model of C02 cycling (Kwon and Schnoor, 1994). Yet, such simple models have been unable to describe the change in atmospheric C02 during the last deglaciation (Sundquist, 1993) There appears to be more to the oceanic C02 cycle than can be explained by simple parameterization of the complex biological, chemical, and physical interactions in a box model. Many of these more complex processes are now being identified The input of new nutrient s to the ocean v ia nitrogen fixation (Karl et al., 1992; Karl et al., 1995 ; Walsh 1996) the uncoupling of "Redfield fixation of inorganic carbon and nutrients (Sambrotto et al 1993 ; Banse, 1994), and the effects of differential carbon and nutrient rernineralization on carbon sequestr a tion (Shaffer, 1993 ; Denman et al ., 1996) are a few of the processe s that must be quantified before a predictive understanding of the oceanic carbon cycle can be accomplished (Denman et al 1996) The transport of C02 into the deep seas is mainly effected by the settling fluxes of organic and calcium carbonate materials from the surface waters to the deep sea, as the abiotic flux due to sinking polar water remove s only a small fraction -0.2 Pg C yr-1 ( Sundquist, 1993) of the anthropogenic C02 signal It would therefore appear to be important to understand the processes that are effecting the movement of biotic carbon While i t may be true that the total biological flux will not change unless there i s increased fertilization of the necessary limiting nutrient in the oligotrophic oceans (Sundquist, 1993) the vertical gradient of I. C02 (Figure 1 .1) is directly attributable to biological proce s ses e s peciall y within the surface waters in contact with the atmosphere The magnitude of the C02 exchange of the surface waters with the atmosphere is a direct function of this 3


0 1 ,.--..2 ] ...._, :I:3 t 04 5 6 2000 2100 TC02 (J!Mikg) 2200 2300 2400 Figure 1.1. Typical dissolved inorganic profiles for each of the world's oceans. Data are averages for the following ocean regions: NA North Atlantic; SA South Atlantic; AA, Antarctic region south of latitude 45s; SI-South Indian; NI-North Indian; SPSouth Pacific; and NP-North Pacific (redrawn from Sundquist, 1985) 4


gradient ; hence an understanding of this gradient's controlling mechanism are necessary before any prediction of future fluxes can be made The importance of understanding the changes in this vertical gradient are highlighted by the fact that. prior to the industrial revolution, the oceans :were outgassing about 0.6 Pg C yr-1, presumably to offset the riverine loading of carbon into the ocean (Broecker and Peng 1992; Siegenthaler and Sarmiento, 1993). At today's influx of 2 0 Pg C yr-1, there has been a net change in the oceanic flux of2.6 Pg C yr-1. Since there is only -0.4 Pg C yr-1 accumulating in the surface ocean (Siegenthaler and Sarmiento 1993), there must have been a 2.2 Pg C yr-1 change in the net flux between the deep ocean and the surface waters. Given that most of the biogenic organic carbon is remineralized within the upper 1 km of the oceans (Martinet al., 1987), one would expect that most of the accumulation of this C02 has occurred in this depth region below the surface. The net effect would be a build up of C02 at this depth interval and an increase in the vertical gradient of C02 from here to the surface Theoretically, this would force more C02 into the surface waters, and could potentially be a positive feedback to increasingly higher atmospheric C02 concentration, allowing less storage in the surface ocean This scenario leads to an important question that needs to be addressed if one is to try to predict future changes in the ocean carbon cycle does C02 of biogenic origin accumulate in the upper ocean beneath the euphotic zone, with increasing atmospheric C02? This question is the underlying thrust of this dissertation. The focus of this work is to synthesize the current understanding of the biological-physical-chemical interactions of the upper ocean into a predictive computer simulation. 1.2 Tools of research and specific objectives The complexity of the interactions between biological, chemical and physical processes make it difficult to design experimental programs which can simultaneously 5


measure all the parameters deemed significant in a particular ecosystem. Experimental procedures typically measure one set of parameters over the course of time, holding all else constant, or attempt to measure many things instantaneously. Neither of these methods provides a framework for studying the interactions of many variables over long time scales. Simulations, however, allow the exploration of hypotheses on the interactions themselves, as well as a way of studying system-wide dynamics. Simulations focus attention on areas of weak understanding, and as a result, provide a necessary feedback mechanism into the design of experimental procedures To address the question stated in Section 1 .1, a complex one-dimensional computer simulation of the nonlinear biological, chemical and physical interactions effecting the seasonal cycles of organic and inorganic carbon will be used to describe the upper 1.0 km of the ocean at the Joint Global Ocean Flux Study (JGOFS) Bermuda Atlantic Time-series Study (BATS) station (Figure 1.2) This simulation analysis involves the seasonal cycles of primary productivity dissolved organic matter, four functional group s of phytoplankton, two forms of dissolved inorganic nitrogen, dissolved inorganic carbon, and the associated processes of nitrogen-fixation, grazing, and nutrient regeneration throughout the upper water column. This site was chosen for a number of reasons, the most important being the existence of a long term JGOFS time-series station that has been occupied since 1988. The location is also important since the greatest influxes of C02 are thought to occur in the central gyre regions (Tans et al., 1990) In addition, this region has been intermittently studied since the 1950's providing an historic data set rich in biological chemical and physical detail. Validation data for the simulation thus comes from the BATS (Knap et al. 1991; Knap et al., 1992 ; Knap et al., 1993; Knap et al., 1994; Knap et al., 1995) and other data sets produced from the region (e.g. Menzel and Ryther, 1960; Jenkins, 1988; Spitzer and Jenkins, 1989; Lohrenz et al., 1992; Malone et al., 1993; Roman et al., 1993; Carlson et al., 1994; Michaels et al., 1994b; Waters and Smith 1994; Siegel et al., 1995; Bacastow et al., 1996 ; Bates et al 1996; Michaels and Knap, 1996; Siegel and Michaels, 1996) 6


BATS DEPLOYMENT AREA efore July 1994 (31 45'N, 64 10"W) After July 1994 (31 40'N, 64 10"W) Bennuda HYDROSTATION S (32 lO'N, 64 30"W) OFF SITE (31 o SO'N, 64 lO"W) Figure 1.2. Bermuda Atlantic Time-serie s Study site. (redrawn from Michaels and Knap, 1996) 7


A number of numerical experiments have been performed on this ecosystem becau se of the rich data set available from thi s site and the surrounding Sargasso Sea (e.g. Wroblewski et al 1988; Fasham et al., 1990; Taylor et al., 1991; Fasham et al ., 1993 ; Bissett et al., 1994; Doney et al., 1996; Hurtt and Armstrong, 1996). The differences between this simulation analysis and past works are inclusion of-1) differential carbon and nitrogen cycling; i.e. non-Redfield growth and rernineralization; 2) incorporation of four functional groups of phytoplankton; 3) nitrogen fixation ; 4) dissolved inorganic carbon cycling; 5 ) phytoplankton growth based upon the s pectral quality of light; and 6) two pool s of dissolved organic matter. The incorporation of these additional factors increases the size of the validation data set that constrain the simulation, and works to improve the veracity of the predictions 8


2. Biological Model 2.1 Biological Methods 2.1.1 Biological Methods Overview The model used in thi s s tud y is a one-dimensional simulation of the seasonal phytoplankton dynamics in the upper 1000 meters of a subtropical gyre. Autotrophic biomass is divided into four functional groups of phytoplankton. Growth of these functional groups is effected by spectral light absorption and differential nutrient uptake. Loss of biomass is effected by grazing sinking, and excretion. The vertical mixing of biomass and nutrients is accomplished with daily specification of turbul ence driven by water column stability and surface wind s tr ess. The state equations in this s imulation are for phytoplankton ( Pi), bacteria biomass (B), nitrate (N 03, where the charge is ignored in future usage ), di sso lved organic carbon (DOCi), dissolved inorganic carbon ( DIC ), and zooplankton fecal pellets ( F). The s ub script i on the phytoplankton symbol has a value of 1 through 4 and repre se nts the algal functional groups throughout all the state equations. The subscription the dis so lved organic carbon sy mbol has a value of 1 through 2. The currency of these state equations is carbon and everything is converted into C liter-I for the calcu lation s. There are analogous independent s tate equations for the nitrogen pools of phytoplankton (PNi), bact e ria (BN), di sso l ved organic nitrogen (DONi), and f ecal pellet nitrogen (FN). The nitrogen state equations are linked to the carbon state equations by C:N 9


ratios, which in some cases bound both the nitrogen and carbon state equation solutions. There are also independent equations for color dissolved organic carbon (CDOCi). All the equations are solved in finite difference form for each depth interval (Bissen et al., 1994). The depth interval,&, is equal to 2.5 meters. The total number of depth intervals, NZ; is equal to 400 (1000 meters I 2 5 meters). The time interval, is equal to 150 seconds. The following description includes the parameters used in the basic version of the simulation The results section will include descriptions of the different parameter choices used during the testing phase. 2.1 2 Phytoplankton State Equations The phytoplankton functional groups in this simulation represent the dominant autotrophic biomass in the Sargasso Sea as determined from pigment and size fractionation studies (Gieskes and Kraay, 1986; Glover et al., 1986; Prezelin et al., 1986; Prezelin and Glover, 1991; Goericke and Repeta, 1993; Goericke and Welschmeyer, 1993; Neveux and Lantoine 1993; Michaels et al., 1994b). These functional groups also define the number of identifiable groups that can be created with the weight-specific pigment absorption curves from Bidigare et al. ( 1990). The pigments include chlorophyll a, chlorophyll b, chlorophyll c, photosynthetic carotenoids (PSC), photoprotective carotenoids (PPC), as well as phycoerythrin (PE) from Synechococcus stains WH7803 and WH8103, containing the phycoerythrobilin (PEB) and phycourobilin (PUB) chromophores, respectively A prior simulation analysis utilized two functional groups of phytoplankton to effect the seasonal cycle of autotrophic biomass (Bissett et al., 1994). This simulation analysis seeks to use the additional data sets of accessory pigments (e.g Michaels et al., 1994b) and spectral diffuse attenuation coefficients (Siegel et al., 1995; Siegel and Michaels 1996) as validation data. Hence, the need for more than two functional groups of phytoplankton 10


results from the desire for additional validation data The four functional groups with their pigment suites are defined as follows: -2 Prochlorococcus-like groups size= -0. 6 J.Lm FG 1 chl a, chl b (low chl alehl b ratio), PPC FG2-chl a, chl b (high chl alchl b ratio), PPC -1 Synechococcus-like group, size = -1.0 J.Lm FG3-chl a, PE (high PUBIPEB ratio cum WH8013), PPC -1 Chromophycota-like group, size= -2.5 J.Lm FG4-chl a, chl c, PSC, PPC FG4 encompasses the Chrysophyta algae, i.e., diatoms pryrnnesiophytes, and chrysophytes The exact range of pigment concentrations as a function of phytoplankton carbon concentrations can be found in the Bio-Optical Section (Section 3). There are two groups of Prochlorococcus species because Prochlorococcus represents a significant fraction of the summer-time biomass and productivity in stratified oligotrophic waters (Olson et al., 1990; Goericke and Welschmeyer 1993; Campbell et al., 1994; Vaulot et al., 1995; Bucket al., 1996). However, there are distinct differences between the species that grow in the high-light surface waters, and the low-light subsurface chlorophyll a maximum (Moore et al 1995) In particular the low-light adapted species are extremely light sensitive, achieving a maximal growth rate at 20-40 J.Lmol quanta m-2 s-1. This rate is reduce to zero by -150 J.Lmol quanta m-2 s-1 (Moore et al. 1995). In order to represent the ecosystem under stratified conditions, it is necessary to include two species of this organism Phytoplankton biomass is carried as rnicromoles carbon per liter. All rates of growth and loss are converted into carbon-specific terms before being incorporated into the simulation The generic state equation for phytoplankton is: 11


where the first term is the growth of phytoplankton biomass The realized net growth rate, J..L, is the minimum of the light-limited (J..Lu), and nutrient-limited (J..Lru) growth rates. In this form, growth is not limited by. a multiplicative effect of light and nutrient limitation (Walsh and Dieterle, 1994). It is instead limited by the scarcest available resource of light or nutrients (Blackman, 1905; Walsh, 1975). The phytoplankton particulate nitrogen state equation is: ( a aP;) (N) + az K7. az c P, 2 2 where PNo3 i and PNH4_i are the terms for nitrogen uptake. N/C is the nitrogen to carbon ratio of the fluxing material from the donating grid box. The determination of J..Lu and J..l.In begins with a function that describes the effects of temperature on balanced carbon specific growth (Eppley, 1972 ): II II 0 .0 633 (f -27) , ,....maxexp where llmax is the maximum growth rate normalized to a 24 hour period at 2 T C. The growth rates for each functional group are given in Table 2.1. 2.3 The temperature dependent maximum growth rate,> is calculated each day, at each depth interval. The daily resolution is an operational definition because the temperature data comes from monthly means (NOAA, 1994) interpolated to daily values. The equations for calculating J..Lu and llru utilize the temperature-dependent maximum growth,> a s the maximum growth rate for that day. However, J..Lu and are 12


calculated at each .1t since the availability of light and nutrients will change dramatically over the course of a day Table 2.1. Maximum Functional Group Growth Rates 24 hour, Function group 27 c growth rate, d 1 FG1 I 1.28 FG2 I 1.45 FG3 I 2.00 FG4 I 3.70 1. (Moore et al., 1995 ) 2 ( Partensky et al., 1993) 3.-(Cuhel and Waterbury, 1984) 4. -(Kana and Glibert 1987) 5.-(Brand and G u illar d 1981) 6. (Geider and Osborne, 1987) 7. (Goldman and McCarthy, 1978) 2.1.2.l li g ht-limted growth rate reference 1 2 I 1 2 I 1, 3, 4 I 5 6, 7 The light-limited carbon growth rate is calculated at each depth and time interval when the quantity of light at that depth exceeds 1.0 J.lmol quanta m 2 s 1 When the light level i s below 1.0 J.lmol quanta m2 s -1, the rate i s set to zero. The light-limited growth rat e can be initially described by a hyperbolic tan ge nt function of growth versus irradiance (J ass by and P l att, 197 6): [ a ( E0(z) -E0(comp) ) l J.111 = tanh J.1 m J.lm where Eo(z) i s the scalar irradiance at depth z in Jlmol quanta m-2 Eo(comp) i s the compensation light flux at which net growth i s equal to zero fo r the individual functional 13 2.4


group in j.lmol quanta m-2. E o (comp) is 1.0 for FG 1 (Moore et al., 1995), 6.0 for FG2 and FG3 (Moore et al., 1995), and 10.0 for FG4 (Richardson et al., 1983; Sakshaug et al ., 1987). Alpha, a is the photosynthetic efficiency (j.lmol quanta m-2)-1. It varies as a function of ambient light conditions and intracellular nutrient status since the absorption of light and utilization efficiency varies under changing environmental conditions (Laws and Bannister, 1980; Geider, 1987; Sakshaug and Andresen, 1989). Photosynthetic efficiency can be described by (Kirk, 1994): 2.5 where max_ i is the maximum quantum yield of photosynthesis (mol C (mol quanta) -I) and a* ph_p_ i is the spectrally weighted specific absorption coefficient of the photosynthetically active pigments for each phytoplankton group (meters-!). This specific absorption coefficient is defined as: 700 I aph_ p_i (A.,z) E0 (A.,z) dA. 4 00 a ; h p i (z) = ---=7=-=-oo=-------2.6 I E0(A.,z ) dA. 4 0 0 where aph_p_i(A.,z) reflects the spectral absorption calculated for the functional group i at a wavelength A. and a depth of z in units m-1 (while a ph_p_i is defined as an integral, it is calculated as a sum at 5 nrn resolution) aph_p_ i (A,z) is defined as: a p h p _i(A.,z) = packaging effect L (pigments [mg m -3 ] a;ig [m2 mg -1]) 2.7 14


where packaging effect is the reduction in weight specific pigment absorption because of self-shading (Kirk, 1994); pigments refer to the concentration of each photosynthetically active pigment within the functional group i at that depth; and a* pig is the weight-specific absorption of the particular pigment at that wavelength The packaging effect is assumed to be negligible for the Prochlorococcus-like FG1 and FG2 (Morel et al. 1993; Moore et al., 1995) as well as for the Synechococcus-like FG3 (Bidigare et al., 1989; Moore et al. 1995); their values for packaging effect are thus set equal to 1.0 For FG4, packaging is instead a function of the carbon:chlorophyll a ratio. At the minimum carbon to chlorophyll a ratio (which occurs at the lowest photon flux density of growth for FG4 of 10 quanta m-2 s-1 ), the packaging effect reduces aph_p_ 4 to 50% of its unpackaged value (Geider and Osborne, 1987; Bricaud et al., 1988). The reduction for FG4 is minimized (package effect is equal to 1.0) when the carbon to chlorophyll a ratio approaches its maximum light-limited value, or 60:1. The value is maintained at 1.0 under nutrient-limited growth, where the carbon to : chlorophyll a ratio can reach a maximum value of 150 : 1 (a more complete description of the functional groups pigment suites and the conditions under which they vary can be found in the Bio-Optical Section). The following is an example of the calculation of aph_p_1 for FG 1 at 100 meters Assuming the concentration of pigments are 0.50 mg m-3 chlorophyll a and 0.50 mg m-3 chlorophyll b (the absorption of light resulting from photoprotective carotenoids are not included in the photosynthetic efficiency calculation). The a* pig for chlorophyll a at 442 nm is 0.025 m2 mg-1 and 0 0126 for chlorophyll b (Bidigare et al. 1990b). Since packaging effect for FG 1 is equal to 1.0, the value for aph_1 ( 442,1 00) then equals (0.50 0.025) + (0.50 0 0126) = 0.0188 m-1. This value for aph_p_1 is then used in Equation (2 6) to derive the spectrally weighted specific absorption coefficient. The necessity for weighting the spectral absorption of the phytoplankton biomass by the radiant light field becomes obvious, when one recalls that the photon flux changes its spectral dependency as it penetrates the water 15


column. Equation (2 6) thus gives a competitive advantage to the functional group whose pigment suite most efficiently absorbs the ambient spectral light field at depth. While cultures of both Prochlorococcus and Synechococcus appear to have relatively unpackaged absorption spectra (Morel et al., 1993; Moore et al. 1995), their absorption efficiencies are quite different et al., 1993). The decrease of size for Prochlorococcus, relative to Synechococcus ( -0.6 11m vs. -1.0 11m), appears to be matched by an increase of internal pigments relative to diameter. The increase of relative pigmentation does not result in a reduction in the weight-specific absorption (packaging effect). As a result, the probability of a photon impinging on the geometric cross-section of the Prochlorococcus being absorbed is about twice as high as the probability for Synechococcus Or, put differently, their optical cross-sections are about equal. If one normalizes this efficiency by the biovolume, or carbon content, the relative optical efficiency of absorption is about 4 times higher for Prochlorococcus (Morel et al ., 1993) Equation (2.5) is modified to incorporate this enhancement: 0: ; = enhance max_ i a;h_i 2.8 where enhance is equal to 1.5 for FG 1 and FG2. It is equal to 1.0 for FG3 and FG4. While a value of 4.0 could be justified, the work by Morel et al. ( 1993) used a higher growth irradiance for the Synechococcus experiments than in the Prochlorococcus experiments This could have led to some of the decreased relative efficiency, as the pigmentation of the Synechococcus culture was lower than other culture experiments (Kana and Glibert, 1987a; Kana and Glibert, 1987b ). Therefore, a more conservative value of 1.5 is chosen The minimum mole quanta needed to fix one mole carbon Cmax_i-1) is 8 (Kirk 1994 ). However, as a practical limit this value is probably not reached in nature because of leakage in the energy transduction system of the photosystems (Geider et al 1986a; Kirk, 16


1994 ) Cultured diatoms grown in low light have been shown to utilize a minimum of 11 mol quanta to fix 1.0 mol C (Sakshaug et al., 1991). A similar value of 10 has been found at low light for Prochlorococcus (Partensky et al 1993), but these values are the exception rather than the norm A more conservative value of 12 or max_i = 0.083 in Equation (2 8), for all functio nal groups is used in this simulation (Bissett et al., 1994) With max_ i and a* ph_p i, a i can be calculated fro m Equation (2.8), and can be calculated with Equation (2.4). Light inhibition is only considered for FG 1 and FG2 (Moore et al., 1995), as it is not considered a factor for cultured species, other than Prochlorococ c us that are acclimated to high-light growth regime s (A nderson and Roels 1981; Falkowski et al. 1985 ; Post et al., 1985; Kana and Glibert, 1987a; Kana and Glibert 1987b; Sakshaug and Andresen, 1989; Iriarte and Purdie, 1993; Moore et al ., 1995 ). From the growth versus irradiance curves of Moore et al. ( 1995), light begins to inhibit FG1 at 40 quanta m-2 s-1. A zero gro wth rate is reached at 110 quanta m 2 s-1. Inhibition is assumed to be an exponential reduction of the maximum growth rate which is effectively zero when the exponential reduction factor reaches 0 001 or e xp( 7) The exponential decay rate is solved as a linear function of light energy: decay rate = 7 0 = 0.1 quanta]-1 110 40 so that for Eo(z) greater than 40 quanta m-2 s-1, Equation (2.4) for FG 1 becomes: = tanh 2.9 [ a1 ( E0(z) E0(comp)) l 2.10 exp [ 0.1 ( E0(z)-40.0 ) ] 17


The decay rate for FG2 is solved in a similar fashion, where the inhibiting light level is 105 J..Lmol quanta m-2 s-1 and a zero growth rate is achieved at 10,300 J.lmol quanta m-2 s-1. Note, the culture work of Moore et al. ( 1995) did not measure growth at this extremely high light level. The zero growth light intensity was determined by continuing the log-linear plot of growth versus irradiance to its zero intercept at light intensities greater than 105 J..Lmol quanta m-2 s -1. What this effectively means is that the growth FG2 is only slightly negatively affected by high-light intensities The exponential decay rate is solved in a similar manner a s Equation (2.9) and its value is 0 001 [J..Lmol quanta m 2 s-1]-1. At light intensities greater than 105 J..Lmol quanta m-2 s-1 Equation (2.10) (substituting the proper terms for the FG2 exponential decay) is used to calculate J..lll_2 2 1.2.2 nutrient-limited growth rate In the oligotrophic Sargasso Sea, nitrogen and phosphorous both limit primary productivity (Menzel and Ryther, 1960) However this simulation analysis will focus on nitrogen limitation, as phosphorous is considered less limiting than nitrogen in the open ocean on times scales relevant to biological productivity (Codispoti, 1989). While there is some evidence that carbon dioxide might limit the growth of those diatoms unable to utilize bicarbonate (Riebesell et al 1993), DIC is not considered a rate limiting nutrient for the ecological system as a whole (Walsh and Dieterle 1994) Hence, it is not included in the nutrient-limited growth calculation. It has been widely noted that the uptake of dissolved inorganic nitrogen (V, units of time-I) and the growth rate of organic particulate nitrogen (J..L, units of time-I) can vary independently, depending upon the nutritional status (Q) of the phytoplankton cell ( Goldman, 1980 ; Goldman and Glibert, 1982 ; Goldman and Glibert 1983) The utilization of a singular Michaelis-Menten function to describe both nutrient uptake and growth is only valid when J..l is equal to J..lmax (Goldman and Glibert, 1983). It may be true 18


that as a general rule may be close to one in the open ocean (Goldman, 1980) ; and if so, a singular relationship between nutrient uptake and growth might suffice. However, a J.l:J.l.max of one does not describe the actual realized value of Jl, which varies significantly over temporal and spatial scales The reason for this variation in growth rate is the changing composition of the phytoplankton assemblage over time and space. The dominant phytoplankton species in a particular assemblage may be growing at, or near, its maximal rate The nutrient uptake and growth for this species could be described by a Michaelis-Menten function. However, a single function would not describe the myriad of species within a seasonally shifting phytoplankton assemblage that have different Jlmax values A method is needed to allow the dominant functional group to shift over time as a result of their competitive advantages The method used here is a combination of uptake and growth equations The larger the phytoplankton species the greater the ability for the individual phytoplankter to take up excess nutrients for use at a later time. This luxury consumption gives larger organisms a competitive advantage over smaller organisms that can not assimilate the nutrients at a rate greater than they can incorporate the nutrients into macromolecules Diatoms have the greatest potential for luxury consumption of nitrogen, followed by flagellates, and then other smaller phytoplankton species (Wheeler, 1983) However, smaller cells have an intrinsic competitive uptake advantage re s ulting from their larger surface area to volume ratio (the larger this ratio the greater the diffusive supply of nutrients to the cell per unit volume). The ability for luxury uptake by larger cells partially offsets thi s diffusive uptake advantage of smaller cells In this s imulation the maximum nitrogen uptake rate, v' m _j, at depth z and time t, is as follows: y = Q,.a_i m _i 19 2.11


where Qi is the nutritional status of the functional group, defined as the nitrogen to carbon ratio. The exponent, ai, describes the curvature of the maximum uptake to maximum growth rate relationship. For FG 1, FG2, and FG3, the maximum uptake rate is assumed to be equal to the maximum growth rate and ai is equal to zero. This does not mean that they cannot store nitrogen above their cellular quota for balanced growth. It means that the uptake rate is equal to the nutrient-limited growth rate (Glibert and Ray, 1990) For FG4, uptake can exceed incorporation, and V' m 4 is set to equal twice the value of f...Lmt 4 when -Q4 is minimized The value is determined a s a log-linear function of the cellular nitrogen quota. Dividing both sides of Equation (2 11) by f...Lmt_ 4 taking the log of both sides, and rearranging the terms results in the following: 2 12 The minimum nitrogen to carbon ratio for Q 4 (also denoted in this simulation as KQ_4 ) allowed for FG4 is 1:14 (Goldman and McCarthy, 1978 ; Laws and Bannister, 1980; Sakshaug and Andresen, 1989; Flynn et al., 1994), and the ratio of uptake to growth in the first log term is set to 2. The maximum allowed N:C for this equation is 1:6 625 (denoted a s Qm), as this is considered the ratio of balanced growth. Uptake should equal growth at balanced nutrient incorporation, so the value of the ratio of uptake to growth is set to 1.0. This provides the end points for a straight line versus Q 4 with a slope of 3 3032 and an intercept of -0.4986. Substitution into Equation (2.12) of Q4 and at each time step and depth interval describes a family of curves whose maximum value for nutrient uptake is found where the nutritional status of the cell is at a minimum level of nitrogen, Ko_ 4 The actual amount of nitrogen uptake for each functional group is defined as a function of the ambient concentration of nitrate, N03, and ammonium, NJ4, and the total 20


particulate nitrogen in each functional group, PNi. The absolute transport flux, p (units of ma ss volume-! time-1), becomes: P NH i = V 1 max_i PN i .2.13 2.14 where the maximum absolute transport flux is equal to V' max_i (PNi) The total nitrogen transport for functional group i is equal to PN03_i + PNH4_i. The parenthetic equations reflect the total amount of nitrogen available and provides for preferential uptake (or the repression of N03 uptake b y the presence of NH4, Flynn 1991; Smith et al., 1992) by phytoplankton (Wroblewski, 1977; Bi sset t et al. 1994). The half-saturation con s tants for nitrate, KsN03_i, and ammonium KsNH4_i, are assumed to be 0 100 and 0 050, respectively, for FGl and FG2. These are perhaps the lowe st half-saturation constants found i n the literature (Eppley et al 1977) prior to the advent of the technique s to measure nanomolar concentrations of nitrogen (Harrison et al 1996). However, the hal f-saturation constants assumed for this study are conservative estimates of those measured in the o li gotrophic North Atlantic that were found by using the tracer te chni que s in nutrient kinetic experiments ( Harri so n et al., 1996 ). The K s values for FG3 and FG4 are assumed to be greater than tho se of FG 1 and FG2 given surface area to volume considerations of the cellular diffusive flux. The Ks values are proportionally increased by their ratio of the volume to surface area (as opposed to normalizing by the ratio of s urface area to volume, since this would give FG3 and FG4 smaller Ks values). Using a mean diameter of 0 .6 J..Lm for FG 1 and FG2, and 1 0 J..Lm for FG3, KsN03_3 and KsNH4_ 3 are equal to 0 167 and 0.083, respectively. Using a mean diameter of2.5J..Lm for FG4, KsN03_4 and KsN H4 _4 are equal to 0.417 and 0.208, resp ec tively. These values 21


provide a slight competitive advantage in nutrient uptake to smaller functional groups under very low nutrient conditions. When Equation (2.13) was first developed, the half-saturation constant was set to 1 0 J..l.m for N03 and NJ4, with the 'I' value set to 1.462 (Wroblewski, 1977). However, when utilizing lower half-saturation constants, 'I' must be increased to keep the total parenthetic value below 1.0 (since Pi should never be greater than Pma.x_D. The values for 'I' are now 10.0 10.0 6.5 and 2.6 for FG1 FG2 FG3 and FG4, respectively : Nitrogen uptake is calculated separately from carbon growth and all nitrogen taken up by the phytoplankton is added to the particulate nitrogen pool of each functional group. During light limitation the nutrient uptake rate can greatly exceed that of theJight-lirnited carbon growth rate. As a result particulate nitrogen may accumulate faster than particulate carbon, and the total particulate nitrogen: particulate carbon ratio may exceed that of Redfield ratio of 1.0:6.625 Such differential carbon and nitrogen uptake at both greater than and less than the Redfield ratio, has been found in culture experiments under light and nutrient limitation (Laws and Bannister, 1980 ; Falkowski et al., 1985; Kana and Glibert, 1987a; Sakshaug and Andresen 1989 ; Sakshaug et al., 1991; Smith et al 1992). The C:N ratio is allowed to drop to a minimum of 5.5 during the uptake of nitrogen (remember also that carbon uptake has not been calculated yet, and so the carbon content is from time step t 1). If the C:N ratio is lower than this minimum 5.5, then nitrogen is released back to the water column. Nitrate is preferentially released followed by ammonium until this minimum ratio is achieved. Nitrogen uptake is allowed to continue 24 hours a day. If the amount of light at the current depth interval exceeds that of Eo_i( comp ) then a new Q value is calculated. The nutrient-limited carbon growth rate is then detennined from the Droop equation (Droop 1973), as it appears to be a better mathematical representation of growth under nutrient limiting conditions (Marra et al. 1990; Haney and Jackson, 1996) It is described by: 2 2


2 .15 where J.l. is defined as the unattainable growth rate at infinite Q . In cases where the limiting nutrient makes up only a very small percentage of the cell, i.e Qro >> KQ, J.l. approaches J.l.. In the case of nitrogen, which makes up approximately 15% of the cell at a C:N of 6:1 J.l. is about 85% J.l. (Goldman and Glibert, 1983). The true maximum growth, J.l.mto will be reached when the nitrogen to carbon ratio is maximized at Qm. Since changes daily with temperature, J.l. i must also change daily. It is calculated by rearranging Equation (2.15) : Q i [ K l -1 J.l.i =J.l.i 1Q 2.16 and setting J.l.i equal to J.l.mU KQ_i is set to 1.0/6 625 for FG1, FG2, and FG3 It is set to 1.0114.0 for FG4. Qi is set to equal Qm_i (1.0 + Pmax_i)/6.625 This collapses to (1.0 + for all functional group s s ince V' max_ i is equal to J.l.mt_i at Qm_ i (see equation 2.11). J.l.i is then used at each !:ll and Equation (2.15) to calculate the nutrient-limited growth rate, J.l.nU Note, the maximum attainable growth rate is J.l.mU It is reached at a C:N ratio of 6.625 for all functional groups If Qi is less than 6.625 in Equation (2.16), then the realized nutrient lirnited growth rate would exceed J.l.mt_ i Since Q is allowed to drop to 5.5, this may occur If it does, J.l.ni_ i is reduced to equal J.l.mU The minimum of the two growth rates, J.l.n_i and J.l.nl_i is the realized growth rate, J..lr_ i. and is u sed in Equation (2.1) to calculate the change in functional group carbon as a result of phytoplankton growth 2 1.2.3 biomass loss terms After net photosynthesis, the loss of phytoplankton carbon in Equation (2.1) is accomplished through the excretion of labile dissolved organic carbon (ei) grazing (gi ) 23


sinking (ws_i), and movement from the depth interval by other physical processes (diffusion). Labile dissolved organic carbon excretion is held to be a constant 5% fraction of biomass This reflects the permeability of low molecular weight, monomeric, hydrocarbons through the cell wall 1988). Grazing constitutes the majority of the loss of the biomass. The role of zooplankton as a controlling factor on phytoplankton production has long been recognized (for a review see Banse, 1992), as well as its role in the downward transport of organic material below the euphotic zone (Longhurst and Harrison, 1988). Unfortunately accurate spatial and temporal estimates of zooplankton biomass and grazing rates are not as routinely available as are measurements of phytoplankton biomass and productivity (Peinert et al. 1989; Banse 1991; Longhurst, 1991; Banse, 1992) The loss of phytoplankton carbon from pathogenic infections (Suttle et al., 1990; Bratbak. et al., 1993) is also considered part of the grazing function. These losses are even less understood and quantified than the zooplankton losses. As a result of these uncertainties in the exact loss rates of phytoplankton carbon, grazing stress represents the closure term for this simulation (Steele and Henderson, 1992). A prior simulation of this region, featuring two functional groups of phytoplankton utilized a Michaelis-Menten function for the large functional group and a linear function for the small functional group (Bissett et al., 1994). In this dissertation, a Michaelis-Menten function based on the functional group's biomass is used for all functional groups. 2.17 For FG 1 and FG2, the maximum grazing rate is set to equal the maximum growth rate as these are essentially the same-sized organism. In this equation Pi is replaced with the sum FG 1 and FG2 (P1 + P2). Total grazing stress for FG 1 and FG2 is then proportionally weighted by the difference in biomass between the two functional groups 24


K s g i is equal to 1.0 JlM C for FGl, FG2, and FG3 K5g__4 is set to 2.0 JlM C for FG4 to allow a greater lag between biomass growth and grazing pressure for this larger functional group Another way to view Equation (2.17) is that the growth rate of the herbivores is similar to the growth rate of the phytoplankton, and Pi is similar to the herbivore biomass Such similarities are found in oligotrophic regions (Bucket al., 1996). A refuge carbon level is established for each functional group equal to 0.02 JlM C from the surface to 250 meters. When the functional group biomass drops below this value grazing is set to zero. This can be converted into cells per liter by fust using the diameters of 0 6 Jlm for FG 1 and FG2 (Morel et al. 1993 ; Moore et al 1995), 1.0 Jlm for FG3 (Kana and Glibert 1987a ; Moore et al., 1995), and 2 .5Jlm for FG4, along with a conversion factor of 470 fg C Jlm-3 (Verity et al., 1992; Campbell et al., 1994) to yield carbon per cell values of 60 fg C cell-1 for FG 1 and FG2 (Li et al., 1992; Campbell et al., 1994), 250 fg C cell-1 for FG3 (Kana and Glibert, 1987a) and 3850 fg C cell-1 for FG4. Converting these values into Jlmol C cell-I and dividing into the 0.02JlM C refuge level yields refuge populations of 4 0 x 106 cells per liter for FG 1 and FG2, 9.6 x 1 os cells per liter for FG3, and 6.2 x 104 cells per liter. These values are 2 to 4 orders of magnitude less than observed peak cell counts in the Sargasso Sea for each functional group (Li and Wood, 1988 ; Olson et al., 1990; Li et al. 1992). Grazing also constitutes the major pathway of labile DOC 1 production through e x cretion and sloppy feeding by zooplankton (Roy et al., 1989) and other heterotrophs The fraction of gi that is released into DOC 1 is described with the DOCi state equation Sinking losses in this simulation are small. The sinking velocity, w5_i is held to 0 0 m d 1 for FG1, FG2 and FG3, and 1.0 m d 1 for FG4 (Bienfang, 1980 ; Bienfang 1981; Bienfang et al 1982; Takahashi and Bienfang, 1983). The sinking velocity is multiplied by the change in phytoplankton concentration with depth, since the change in concentration at a point in time and depth is a function of the divergence/convergence of the 25


flux (if there is a uniform profile of biomass, and all the biomass is sinking at a constant speed there is no change in biomass at depth z as a result of sinking) A s particulate nitrogen and particulate carbon are both being accounted for independently (with the exception of diffusion), the loss terms reflect both carbon and nitrogen loss at the respective C:N ratio of the functional group. The last term of the phytoplankton state equation is the change in concentration from vertical diffusion The calculation for the vertical eddy diffusivity, Kz, is explained in the physics section (Section 2.1 8). Diffusion is calculated as a function of the carbon gradient of the functional groups Particulate carbon transferred to a different depth interval as a result of diffusion carries particulate nitrogen to this depth interval at the C : N ratio from whence it came. Functional group pigment concentrations are carried in the same manner as particulate nitrogen. Total pigments are, therefore conserved as a function of diffusion during the night. Phytoplankton diffusion is set to zero at levels less than 1 x 10-10 C liter-1 as this carbon level is much smaller than the cellular carbon of the smallest functional group. 2.1 3 Bacterioplank.ton State Equations The state equation for bacterial carbon (B) and nitrogen (BN) biomass are given as : aB = net ( U ooc_r ) GGEC + A toN -gaB az 7 0 2.18 aBN = net ( U DON + U NH 4 ) GGE n + ( AtoN ) ( ) az 7.0 C B -gaB. t 2 19 26


where net refers to the net uptake of DOC, DON + NI-4 after excretion ; GGEc and GGEn are the gross growth efficiencies for carbon and nitrogen, respectively; and gb is the grazing rate on bacteria carbon-limited growth rate The growth of bacteria can be limited by the availability of both dissolved organic carbon and the dissolved forms of organic and inorganic nitrogen (Wheeler and Kirchman 1986; Fuhrman et al., 1989 ; Kirchman et al. 1990; Pomeroy et al. 1995). Therefore, growth of bacterial biomass is assumed to be limited by both carbon and nitrogen and their growth rates equal to the minimum of the carbon-or nitrogen-limited growth rates The organic carbon for simulated bacterial growth is supplied by total labile dissolved organic carbon; i.e. the sum ofDOCt and CDOC1 (the sources of which are described in Section 2.1.4 ), and fl c is the calculated growth rate of bacteria when they are carbon limited. This carbon-limited growth rate is described by: [ DOC! + CDOCI l ll c = llmt_b K + DOC + CDOC S_DOCI I I 2 .20 where flmt_b is the temperature-dependent maximum bacterial growth rate ; and Ks_DOCI is the half-saturation constant for the Michaelis-Menten function flmt_b is found with the temperature dependent equation from Walsh and Dieterle ( 1994): 0 092 (T -27) llmt_b = llmax_bexp where the maximum specific growth rate llm_b is assumed to be 2 0 day-!. 2.21 Ks DOC! in Equation (2 20) is estimated from the growth of bacteria computed by 3H-Thymidine incorporation from the BATS 43 cruise at 40 meters (Knap et al ., 1994). The average 3H-Thymidine incorporation rate was 0.876 pmole liter-! hour-I. Utilizing 27


conversion factors of 4 x 1018 cell mole -I (Fuhrman et al., 1989) and 20 fg carbon cell-I (Ducklow et al., 1993; Campbell et al., 1994) results in a bacterial carbon growth of 0.1402 j.!M C day-1. Bacterial biomass was given in the same report as 5.87 108 cells kg1 Converting this value to j.lmol carbon kg-1 yields a particulate carbon biomass of 0.9783 j.lmol C kg-1 (1.004 j.lmol C.liter-l), that in turn yields a malized daily carbon specific growth rate of0.1396 d-1. Assuming that the total labile DOC (DOCt and CDOC1) concentration was 20 j.lmol (Carlson et al., 1994) and the in situ temperature was 19.936 C, and utilizing Equation (2.21) yields a resultanrvalue for 1-lmt_b of 0 522. Therefore Ks_ooct is approximately 130 j..tM C. nitrogen-limited growth rate Nitrogen for the simulated bacterial growth is supplied by both labile dissolved organic nitrogen, DON 1 and ammonium, The equations for determining the source of the nitrogen utilized for growth are based on those of Fasham et al. ( 1990), with some additional considerations. The equations and parameters are briefly discussed here. Bacterial growth is assumed to occur at a C:N ratio of 5 The amount of ammonium uptake (Goldman et al., 1987) will then depend upon the gross growth efficiency for nitrogen (GGEn) and carbon (GGE c ), and the C:N ratios of the dissolved organic matter (DOM) substrate and bacteria. The bacterial increment in biomass as a result of ammonium and DON 1 uptake i s defined as e and d, respectively. This increment (in nitrogen mass units) h, is then: 2.22 If the C : N ratios of DOM and bacteria are and Rb, respectively, then the carbon-based bacterial production, H, will be: 28


H = Rb h = GGE R d c d Dividing Equation (2.22) by (2 23) and rearranging : e d 2.23 2.24 By assuming balanced C:N growth (growth of bacteria occurs only at a C:N of 5.0), the ratio of NH4 uptake to DON 1 uptake is fixed for given C:N ratios of the bacteria and DOM. This concept i s formalized by first defining a nitrogenous s ubstrate S for bacteria as: 2.25 with ll from Equation (2 24) The bacterial uptake rate of nitrogen on sources of DON 1 and is calculated as Michaelis-Men ten functions: 2 .26 U NH4 = llmt_b B ( ) [ K NH l B S _NB 4 2 .27 where Kss is the half-saturation constant for nitrogen uptake. Kss is s et to be 115 of Ks DOCI (bacterial C : N equals 5). The GGEn is set to be 100% (Fasham et al., 1990); thus, the DON 1 nitrogen uptake rate ( UooN) is equal to the temperature-dependent maximum growth rate multiplied by the bacterial nitrogen biomass multiplied by the Michaelis-Menten function. uptake is equal to UNH4 Note that the division of B by 5.0 in Equations (2.26) and (2.27) reflects the bacterial nitrogen biomass (bacterial C : N equals 5) 29

PAGE 43 t o tal carbon uptake Total carbon uptake by bacteria i s greater than their carbon-based growth since the gross carbon g rowth efficiency is a ss umed to be less than 100%. Their carbon gro ss growth efficiency (GGEc) i s set to 30% ( Banse 1990 ; Amon and Benner 1994 ) and assumes that 70% of the carbon taken up by the bacteria is respired as DIC. Total realized carbon uptake, Uooc_r. is th en: 2.28 where Uooc c = (GGEcr1 ll c B 2.29 Uooc_ N = (GGEcr1 L ( UooN + u N H.) 2.30 where Jlc i s the carbon-limited growth rate of bacteria. Whe n Uooc_c is minimal the uptake of nitro g en would exceed the balanced nitrogen requirement s In thi s case, the total nitro ge n uptake, (UooN + UNH4), is reduc e d so th at total nitrogen uptake matches that set by carbon limitation. The excess nitrogen is released from th e bacterial particulate nitrogen pool as NJ-4, first by releasing the nitrog e n taken up as then by releasing the nitrogen taken up as DON 1 int o form. Thi s provides a mechanism for NJ-4 remineralization when the C:N ratio of DOM 1 is le ss than 16.67 (C:N of bacteria divided by GGEc). The total nitrogen released i s removed from UNH4 UNH4 may therefore becom e negative if the total released Nl-4 include s nitrogen from the DON 1 uptake At this point it is referred to as net( UNH4 ) and is seen in the ammonium state equation. Total carbon uptake is split between the total l abile concentration s of DOC : U DOC = U DOC_r (DOC I l 30 2.31


[ CDOC1 l u C D OC = u DOC _r DOC I + CDOC I 2 .32 Excretion of relict or unusable dissolved organic matter by bacteria (Brophy and Carlson, 1989) provides a source of DOC2 and DON2 in this simulation. DOC2 excretion ebc is set to 4.0% of the DOC1 uptake Uooc (Brophy and Carlson, -1989). DON2 excretion, ebn is set to 0.5% of DON 1 uptake This value of 0.5% is calculated by multiplying ebc by bacterial C : N ratio and dividing by the assumed relict DOC2:DON2 ratio of 40: 1 (taken from TOC:TON measurements near 300 meters in upwelled waters from the North Pacific, Hansell and Waterhouse, 1996) Net carbon uptake, net(Uooc_r). is again tested against net nitrogen uptake, net( UooN + UNH4), to make sure there is sufficient carbon and nitrogen to grow at the GGEc. If the carbon and nitrogen are out of balance (GGEc *net carbon uptake I net nitrogen uptake is not equal to 5 0), then the excess carbon or nitrogen required to balance growth is released as DOC1 (excess labile DOC is released as colorles s DOC,) and Nf4, d e creasing net(Uooc_r) and net(UNH4), respectively By computing the excretion of relict DOM2 after the uptake DOC 1 and DON 1 is calculated the realized gross growth efficiencies for carbon and nitrogen are actually a bit lower than those stated for GGEc and GGE0 The realized values are closer to 29% and 99% respectively The next term, AtoN, refer s to nitrifying bacteria. Nitrification occurs very rapidly at high rates in oligotrophic water, as much as 40 nmol day! (Eppley et al 1990). The numerical description of this process is given in the nutrient state equation section Nitrifying bacteria are chemolithotrophs, and as such, fix C02 into organic carbon. The ratio of N14 oxidized to C02 fixed is held to be 7.0 (Kaplan 1983). The quantity of chemolithic bacterial particulate nitrogen, BN, is calculated by dividing the chemolithic carbon fixed, AtoN /7.0, by the bacterial C:N ratio of 5 :1. 31

PAGE 45 bacteria/loss terms The seasonal amplitude about the mean concentration in heterotrophic bacterial biomass is far le ss than that of phytoplankton, -30% versus -150%, at the BATS site (Carlson and Ducklow, 1996). Bacterial biomass is thus held at a constant carbon and nitrogen .concentration that decreases .with depth (see initial conditions). At each time step all of the bacterial carbon and nitrogen growth is assumed to be grazed. As with phytoplankton functional groups FG 1 -3, bacterial sinking, w5b, is assumed to be 0.0 meters day -!. 2.1.4 Dissolved Organic Matter State Equations There are two forms of DOMin this simulation. Like the state equations for phytoplankton DOMi diffusion is calculated along the dissolved organic carbon gradient. The DONi diffusive flux is based upon the DOCi flux divided by the C:N ratio of the diffusing DOCi In addition, there are separate state equations for CDOCi. The two forms ofDOM are: 1) A labile form that is available for biological and photo-degradation, DOM1; and 2) A relict form that is available for photo-degradation, DOM2. The forms of CDOC are analogous to the form of DOC. The spectral characteristics of the colored forms and their equations are described in the Bio Optical Method s Section. DOC1 The state equations for the labile disso l ved organic carbon and labile colored dissolved organic carbon are: aooc, at = eiP i + FecDOCi (1.0-colorFR,) g i Pi+ BacDOC (1.0 -colorFR1 ) gb B net ( U ooc,) + UVDOC 1 2 33 a aooc, + UVDOC2 + NFIX 5.5 + OZ K z OZ 32


where eiPi is the excretion from each functional group; FecDOCi is th e fraction of the grazing stress from functional group i that is lost as DOCt ; colorFRt is the fraction of total DOC released a s CDOCt (described in Section 3); and giPi is the grazing stress on functional group i; Uooc is the bacterial uptake ofDOC1 ; BacDOC is the fraction of the bacterial grazing stress gbB, that is lost as DOC 1 ; UVDOC 1 is the release .of DOC 1 from the photolysis of relict CDOC1 ; UVDOC 2 is the release ofDOC1 from .the photolysis of CDOC2 ( UVDOCi are de s cribed in Section 3); NFix i s the amount of DON 1 fixed from N 2 multiplied by a DOC1 conver s ion factor ; and the last term is the vertical diffusive flux The state equation for CDOC1 is: acDOC1 at = FecDOC; colorFR1 g; P ; + BacDOC colorFR1 gb B-U cooc UVDOC1 -UVDIC1 I 2.34 a acDOCt + az Kz az where U cdoc is the bacterial uptake of CDOCt; and UVDICt is the release of DIC from the photoly s is of CDOCt. The state equation for labile dissolved organic nitrogen is : aDON1 at = FecDOC ; gi PN; + BacDOC gb BNUDON ( a aDOC 1 ) ( N ) NFIX + az Kz az C oo c 2.35 where (N/C)ooc 1 i s the nitrogen to carbon ratio of the DOC 1 It is ass umed that all of th e color of DOMi is represented by CDOCi; thu s, there are no equations for colored DONi. The fraction of the phytoplankton grazing stress released into the DOC 1 pool, FecDOCi is based on the number of trophic levels through which grazed material mu s t pais before incorporation into rapidly sinking fecal pellets. For example, assume that for 33


the largest phytoplankton group, FG4, fecal pellets of herbivores and omnivores are formed in the first trophic step, which sink at a velocity of 100 meter day-!. Assume further that 25% of the ingested material is metabolized as respired C02 25% is released as fecal pellets, 25% is added to grazer biomass, and 25% is released as DOC1 from either sloppy feeding or direct excretion (Jumars et al., 1989; Roy. e t al.,.1989 ; Banse, 1990) If grazer biomass is assumed to be in steady state, then the 25% that is assimilated into the grazer biomass must be equally distributed between the metabolism, the fecal pellets, and the release of DOC 1 of higher trophic levels. Hence the values for FecCYC4, FecPEL4 and FecDOC4 are 33.33% in steady state (FecCYCi and FecPELi are the percentages of phytoplankton grazing released as C02 and fecal pellets, respectively) For the smaller phytoplankton species, FG 1 FG3, the fecal pellets produced by the microheterotrophs are assumed to have a neutral buoyancy. These fecal pellets must be ingested by an omnivore at a second trophic level before being incorporated into a rapidly sinking fecal pellet : Likewise, the growth of the microheterotrophs pass through the second trophic level as well. Again by assuming steady state after this second trophic level, fecal pellet production is now 25% of the initial 25% of the fecal fraction, plus 25% of the second trophic step growth, and 113 of the remaining grazer biomass growth, or 16 .67% (25% 25% + 25% 25% + 1/3 12.5%) Steady state metabolism (FecCYC) and DOCt production (FecDOC) are equal to their initial 25%, plus 25% of each of the ingested first trophic level fecal pellets and assimilated grazer biomass, plus 113 of the remaining growth; i.e ., equals 41.66% (25% + 25% 25% + 25% 25% + 1/3 *12.5%) The modeled fate of the size fractionated biomass is similar to previous estimates of oceanic food webs (Michaels and Silver, 1988). The percentage of bacterial grazing stress that is release as DOC 1, BacDOC 1 is estimated in the same fashion as that of FecDOC 1, with the exception that one more trophic level is added before the ingested organic matter is incorporated into rapidly sinking fecal pellets. In this case, the DOC1 release (BacDOC) and grazer metabolism (BacCYC) are 34


equal to 45 83% (25% + 25% 25% + 25% 25 % + 25% 12.5 % + 25 % 12 5 % + 113 6 .2 5 % ) The percentage of bacterial grazing that is incorporated into fecal pellets i s thus 8.33 % (25% *12. 5 % + 25% 12 5 % + 113 6 25 % ) The photolysi s of labile CDOC 1 and relict CDOC 2 provides a source of simple, labile organic carbon, e.g pyruvate and carbon monoxide, that could be utilized by the heterotrophic microbial population (Kieber et al., 1989; Mopper and Zhou, 1990 ; Mopper et al., 1991). Photolysis of total CDOC al so releases dissolved inorganic carbon ( Miller and Zepp, 1995 ; Wetzel et al ., 1995 ). DON1 not released from the photolysis ofDOM2. DON 1 is as s umed to remain inert. The effect of this process is to increase the C:N ratio of the labile DOC 1 in surface waters during photolysis. The rates of photolysis are described in the Bio-Optical Section. Nitrogen fixation NFix, i s included in this model as a release of DOC 1 /DON 1 from the nitrogen-fixing cyanobacterium Trichodesmium spp Since this species is not included as an explicit functional group of phytoplankton, it s net effect is to increase the nitrogen available to other phytoplankton and bacteria in this model. The nitrogen is fixed into DON 1 without consideration of the depletion of di-nitrogen, N2. While the supply of N2 is obviously not infinite, it does dominate the dissolved pool of inorganic nitrogen in the s urface waters of oligotrophic gyres by 4 orders of magnitude ( 400-500 f..Lmol N2 liter-1 Karl et al ., 1992) Therefore, N2 is not carried as a state variable in this simulation. The quantity of nitrogen fixed in this model is a function of the estimated number of trichomes liter-! and the rate of nitrogen fixation per trichome per time The number of trichomes liter-! of surface water is, in turn, estimated from the temperature versus Tri c hodesmium colony relationship found in the Caribbean and Western Sargasso Sea s ( Carpenter and Romans, 1991). The maximum temperature in these simulations is -2T C. Such a temperature supports maximum colony concentration of 1 colony liter !, or 250 trichomes liter-! (Carpenter and Capone, 1992). The minimum temperature in thi s model is 19 C, implying a zero trichomes liter-! ( Carpenter 1983 ). The daily surface concentration 35


of nitrogen-fixers is determined from a linear interpolation of their stocks versus the daily temperature ; i.e., trichomes liter = 250 trichomes 2TC19 c ( daily temperature -19 C ) The concentration is held constant from the surface to a depth of 15 meters and then reduced to 10% of the surface concentration at 50 meters (Carpenter, 1983 ). 2 36 The maximum rate of nitrogen fixation is taken to be 750 pmol N colony-1 at 1500 quanta m-2 s-1 (Capone et al., 1994) or 3 pmol N trichome-! This rate is reduced by the ratio of Ect(z)/1500 to account for the reduction of irradiance with depth The amount of DOC1 production is set to be 5.5 times the rate of DON 1 production (Carpenter and Romans, 1991). This model assumes that all of the nitrogen fixed from N 2 is released as DON 1 Grazing losses by large herbivores on nitrogenfixers appears to be slight, perhaps due to the toxicity of the cyanophytes (O'Neil and Roman, 1992; Sellner, 1992). Sedimentation losses of Trichodesmium are minimal as a result of internal gas vacuoles, thus nearly all of the nitrogen fixed is projected to remain in a tightly coupled bacterial loop within the surface waters (Sellner, 1992). Since Trichodesmium spp. are not explicitly included as state variables, this model ignores the assimilation of N 2 into algal biomass thus the nitrogen fixed in this model is released solely as DON 1 relict DOC2 The state equations for the relict dissolved organic carbon and colored dissolved organic carbon are : 2.37 36


dCDOC2 dt = colorFR2 ebc U ooc_r UVDOC2 -UVDIC2 a acDOC2 + dz K z dz 2.38 where colorFRz (described in Section 3) is the fraction oftotal relict DOC released as CDOCz and ebcUnoc r is the excretion of total relict DOC by bacteria during the uptake and conversion of total labile DOC to bacterial biomass UVDOC2 and UVDIC 2 repre s ent the release of colorles s DOC1 and DIC from the photoly s i s of CDOC2 res pectively. 2 .39 where ebnUnoN_I is the excretion of relict DON2 by bacteria during the uptake and conversion of DON 1 to bacterial biomass One last comment about DOM in this simulation I t is not strictly defined here as the matter that would pass through a 0 2 filter. It is more generally defined as that organic matter that is not growing ( or does not have the capacity to grow ) and is also not part of the rapidly sinking fecal pellet pool. 2.1 5 Fecal Pellet State Equations The state equations for the rapidly sinking fecal pellet carbon and nitrogen pools are : dF at = FecPEL i giPi + BacPEL g8B RegenC(z ) F dF + l_K dF -WF dz dz 2 dz 2 .40 37


where FecPELi is detennined in the same fashion as FecDOCi and FecCYCi Note that the sum of + FecDOCi + FecCYCi is equal to 1 .0. BacPEL is also detennined in the same fashion as BacDOC and BacCYC. The sum of these three terms are also 1.0. RegenC(z) is the remineralization rate of the fecal pellet carbon at depth z. Wf is the sinking speed of the fecal pellets, and the last term represents vertical diffusion. The equivalent nitrogen equation is: dFN ---at = FecPEL i giPNi + BacPEL g8BN RegenN(z) FN (w dF + dF F dz C F dz z dz C F where (N/C)F is the nitrogen to carbon ratio of the sinking and fluxing material. 2.41 The regeneration rates, RegenC(z) and RegenN(z), are calculated as a function of depth below 100 meters from the equations of Martin et al. ( 1987) Their sinking flux of carbon or nitrogen at depth for oligotrophic oceanic waters is given by: Flux(z) = Flux!OO( r 2.42 where Fluxwo refers to the amount of material fluxing through 100 meters, z is the depth of interest, and exponent b is the log-log slope of the relationship between depth and measured flux. The exponent b is different for carbon and nitrogen, -0 .858 and -0.988, respectively. What is of interest in simulating the remineralization of sinking particles is the percent reduction of the particulate flux between depth intervals, !::a.. This can be calculated with Equation (2.42) below 100 meters between any two depths as: 38


% reduction = 1.0 -X 100 2.43 where the z1 and z2 refers to the upper and lower depth intervals respectively and the exponent b is the same for each Flux10o divides out of the equation and the solution is the reduction in the mass of fluxing matter between z 1 and z 2 By assuming a s inking speed wp, for fecal pellet s of 100m d-1 (Deuser 1986 ; Wals h et al., 1991 ; Bissett et al ., 1994 ; Walsh and Dieterle, 1994 ), the percentage value from Equation (2.43) has units of s-1 from division of the depth interval, !:J.z, by the sinking rate and then dividing this result into the % reduction For example, between 100 m and 102 5 m the %reduction in carbon mass is equal to 2 108%. Dividing the depth interval of 2.5 m by the sinking speed of 0 0016 m s-1 (100m d-1) returns a value of 2160 s, which means the fecal carbon mas s remains in the depth interval for 2160 s Therefore, divid i ng 2 108 % by 2160 s returns the amount of carbon remineralized in this depth interval per second (0.000976% s-1). By solving Equation (2.43) for each depth interval below 100 m, an array of % reduction per second values is found that is greatest near the surface, decreasing exponentially with depth. The % reduction in depth intervals above 100 m are initially set to equal the % reduction at 100 m. This value is then increased by the temper a ture-dependent function of the bacterial growth rate (equation 2.21 ) since bacteria are a s sumed to be partly responsible for the particulate regeneration process Fecal nitrogen, FN, is treated in a similar manner except for a different exponent b -0 988 (Martinet al 1987) for Equation (2.43) Simulated in this fashion nitrogen is remineralized faster than carbon 39


2.1.6 Nitrogen State Equations The state equations for nitrate and ammonium are: 2.45 dNH4 dt = -I. PNH4_i + FecCYCi gyNi + BacCYC g8BN -AtoN [ 1.0 + ( )J -net ( U NH4 ) 2.46 () dNH4 + RegenN (z) FN + dz Kz dz where the fust term in each equation represents the net phytoplankton uptake of each nutrient. AtoN is amount of ammonium that is nitrified. It is modeled as a Michaelis-Menten function: ( NH4 ) AtoN = 0.04 K NH SNot+ 4 2.47 where the maximum nitrification rate is set to 0.04 J.Lmol N d-1 (Ward et al., 1982; Eppley et al., 1990). The half-saturation, Ks_Nit> constant is set to 0.10 J.Lmol NI4 (this is set to be a conservative estimate of Ks_Nit as the maximum nitrification rates were measured when N02 concentrations were -0.10 J.Lmol). The modification of the AtoN term in Equation (2.46), compared to Equation (2.4 7), is the result of chemolithic bacterial carbon growth that uses energy derived from ammonium oxidation. C02 is fixed into bacterial biomass at a 1 :7 ratio of Nl4 oxidized (see above). If C02 is incorporated into bacterial carbon biomass, it must be balanced by an increase in bacterial nitrogen biomass. It is ass umed that the nitrogen source is ammonium and that the nitrogen is fixed at a C:N of 5 : 1 Net loss of ammonium, as a 40


result of nitrification is therefore slightly more than that oxidized to nitrate. Lastl y, nitrification appears to be inhibited by light (Ward et al., 1982 ) ; to account for photo-inhibition, nitrification is set to zero when light energy at the depth interval Zi 1 j.lmol quanta m-2 s-1. FecCYCi in the ammonium state equation refers to the amount of the grazed phytoplankton nitrogen that is metabolized to NH4 (metabolized carbon and nitrogen are released at the C:N ratios of the grazed phytoplankton). It is modeled in the same fashion as FecDOCi (see above) BacCYC i s the analogou s release of from the grazing of bacteria. Net( UNH4) refers to the net uptake of NH4 by bacteria. Remember that can be released during catabolism of DOC1 when the C : N ratio of DOC1 is less than bacterial C : N divided by the GGEc RegenN(z) is the percentage of the fecal pellet nitrogen FN, that is remineralized into from Equation s (2.42) and (2.43). Finally, the last term in both of the nutrient equations refers to the passive flux of the nutrients over the water column 2.1.7 Dissolved Inorganic Carbon State Equation The change in total inorganic carbon within the water column is effected by the utilization of inorganic carbon by phytoplankton and bacteria during photosynthesi s and nitrification, respectively; the release of C02 during the metabolism of particulate and dissolved organic carbon ; photolysis of CDOC; and diffusive fluxes. These processes are described by: C1DIC dt = -J.l,_iP i + FecCYCi g i P i + net ( U00c_, ) (1.0-GGEc) + Bac CYC g B -AtoN + RegenC ( z) F B 7 0 + UVDIC 1 + UVDIC2 + l_ K C1DIC dZ z dZ 41 2.48


where the first term represents the net production of phytoplankton. FecCYCi is re s piration by grazers. The third term represents the amount of labile DOC metabolized by heterotrophic bacteria during its gross growth. The fourth term represents the respiration of bacterial predation of the microbial web. The fifth term is net carbon fixation by nitrifying bacteria. The sixth term is the amount of fecal carbon remineralized at depth UVDIC1 and UVDIC2 represent the photolysis of DOC1 and DOC 2 respectively, into DIC. The las t term is the diffusive flux. With the exception of photolysis and diffusion, all of the terms have been previously described. The DIC state equation has an open boundary at the air-sea interface The flux of C02 across this boundary is described by: 2.49 where k is the gas transfer velocity, Lis the solubility of C02 expressed in units of concentration/pressure and pC02 is the partial pressure of C02 in air and seawater (W anninkhof, 1992) The more correct form of this equation is written in terms of fugacity (which incorporates the changes of chemical potential as a function of pressure and temperature), rather than partial pressure. However, fugacity is approximately equal to partial pre ss ure at sea level, and the partial pressure of C02 is more commonly referenced in the literature. Thus, partial pressure is used for the remainder of this work The gas transfer velocity is found with Equation (8) from Wanninkhof(l992), whose formulation include s the chemical enhancement of C02 gas exchange that is especially evident at low wind speeds: k = [2. 5 ( 0.5246 + 1.6256 10-2t + 4.9946 10 -4t2 ) + 0.3u2 ] ( Sc I 660 ro 5 42 2.50


where t is the sea surface skin temperature, u is the wind speed, and Sc is the Schmidt number of C02 in seawater (660 is the Schmidt number of C02 in sea water at 20 C) The Schmidt number is found with a least squares polynomial fit versus temperature (Wilke and Chang, 1955; Wanninkhof, 1992). The gas solubility constant, L in Equation (2.49) is found with a least squares polynomial fit versus temperature and salinity (Weiss, 1974; Peng et al., 1987; Wanninkhof, 1992). The atmospheric pC02 value is assumed to be constant over the entire course of the simulation at 355 ppm (Keeling et al., 1995). The partial pressure of C02 in seawater can be calculated using estimates of total alkalinity, temperature, and DIC along with apparent dissociation constants of carbonic acid, boric acid, the solubility of C02, and the activity of the hydrogen ion (Peng et al. 1987). In using the methods of P e ng et al. (1987) the silicate, phosphate, and water alkalinities were ignored since they contribute less than 0.3% to the total alkalinity (Walsh and Dieterle, 1994). The exception to the methods of Peng et al. ( 1987) was finding the activity of the hydrogen ion. Rather than solving interactively for this activity, the root solution of a cubic equation for the hydrogen ion activity ( Gieskes, 1974) was instead found. Nitrate alkalinity was also ignored in this model (Walsh and Dieterle, 1994). Total alkalinity is a linear function of salinity and was set to 2390 kg-! at a salinity of 36.6 psu (Bates et al., 1996). This alkalinity was multiplied by a factor of S/36.6, where S is the daily surface salinity (see below). The temperature used in the above equations is an estimate of the sea surface skin temperature. The sea surface skin temperature is on average 0.2 C colder than the bulk sea surface temperature (Schluessel et al ., 1990) and its effects on gas exchange can be dramatic (Sarmiento and Sundquist, 1992 ) Hence, the water temperature of the surface layer was reduced by 0.2 C in these calculations. The DIC cycle is affected by calcium carbonate formation that decreases total alkalinity and increases pC02. It is excluded in this model, as the fixation of organic 43


carbon dominates the biological flXati on of DIC into particulate carbon ( -80% of total DIC fixation Broecker and Peng, 1982 ; Denman et al., 1996) 2.1.8 Physical Overview The physical inputs of the model are temperature (T), salinity (S), wind speed (WND), and spectral light. The light formulation is covered in the Bio-Optical Section. Daily temperature, salinity, and wind speed are interpolated from monthly means and are used to generate a daily one-dimensional mixing regime. While higher frequency physical forcing may change the ecosystem structure (Bissett et al., 1994), higher frequency in situ data is unavailable and a three-dimensional physical simulation is beyond the scope of this work. 2.1. 8.1 vertical mixing The time and depth dependent mixing regime is parameterized by the vertical eddy diffusivity, Kz. It is found daily as the solution of a turbulence closure scheme (Mellor and Yamada, 1982; Chen et al 1988) coded by Walsh and Dieterle (1994) This scheme depends upon the vertical structure of density and the horizontal velocity shear. Density is computed from daily values ofT and S. The velocity shear is defined as a function of the surface stress induced by wind This daily stress is derived using a formulation of the wind induced drag coefficient (Hellerman and Rosenstein, 1983). Scalar wind is interpolated daily from monthly averages at the appropriate latitude and longitude (Wright, 1988) of the BATS study site. T and S values are interpolated from monthly values at the appropriate latitude and longitude (NOAA, 1994) These monthly values are given at fixed depth intervals. The monthly values are first interpolated to a 2.5 meter depth resolution. These twelve profiles ofT and S at a 2.5 m resolution to 1000 m are then interpolated to daily values (Figure 44


2 .1). These daily values are used in all of the above calculations, i.e. mixing growth, alkalinity etc., that utilize temperature and salinity data. 2.1.9 Initialization Table 2.2 summarizes the model's parameter values The initial conditions for the simulation are drawn from a number of sources. Phytoplankton carbon is set to 0.12 JlM C from the surface to a depth of 250 meters, 0.00 JlM C from 252.5 to 1000 meters The carbon levels correspond to -0.13 Jlg Chl a literI in the upper 250 meters if total phytoplankton carbon is divided by a C:Chlorophyll a ratio of 45. Phytoplankton nitrogen is initialized at a C:N of 6 625 for all functional groups. Bacterial carbon is set to 0 85 J..lM C in the surface waters decreases with a hyperbolic tangent function (HT AN (30,000 I (z)2) 0 85) to 0 10 J..lM C at 500 meters and is constant from there to the bottom of the model. Assuming a conversion ratio of 20 fg C cellI (Ducklow et al., 1993 ; Campbell et al., 1994) this equates to 5 1 x 108 at the surface, and 0 .61 x 108 cells liter-! between 500 and 100 meters Bacterial nitrogen is initialized at a C : N ratio of 5.0. Total relict dissolved organic carbon DOC2 is set to 55. 0 J..lmol in the surface waters, decreases with a hyperbolic tangent function, (HT AN ( 1000 I (z)2} 0.85), to 45 Jlmol at 500 meters, and is constant to the bottom of the model. Total labile DOC 1 is set to 8.0, 3 0 and 0 001 Jlmol between depths 0-200, 200250, 2501000 meters, respectively. These carbon concentrations are converted into DONi at C:N ratio s of 6.625 and 40.0 for DOC1 and DOCz, respectively 45



Table 2.2. Phytoplankton Parameters for the Complex Model FG1 FG2 FG3 FG4 Eo(comp) 1.0 6.0 6.0 10.0 quanta m-2 s-I Maximum Package none none none reduce to 50 % Effect Optical Cross1.5 1.5 1.0 1.0 Section enhancement max_r1 12 12 12 12 mol quanta [mol C]-I Light inhibition 0.1 0.001 0.0 0.0 decay rate Onset of light inhibition 40 105 0 0 quanta m-2 s-I Absolute Maximum Uptake 2.0 X d-1 Slope and intercept 0.0 0.0 0.0 3.3032 ofai 0.0 0.0 0.0 -0.4986 MaximumC : N I 6.625 I 6.625 I 6.625 I 14.000 MinimumC:N 5 .500 5.500 5.500 5.500 K sN03 I 0.100 I 0.100 I 0.167 0.417 K sNH4 I 0.050 I 0 .0 50 I 0.083 I 0.208 Nitrate uptake 10.0 10.0 6.5 2.6 repression, 'I' carbon excretion,ei I 5.00 % I 5.00% I 5.00 % I 5 .00% K sg I 1.0 I 1.0 I 1.0 I 2.0 Fecal Fraction of 16.67% 16.67 % 16.67% 33.34% Grazing, FecPEL DOC 1 Fraction of 41.67 % 41.67 % 41.67 % 33.33 % Grazing, FecDOC Recycled Fraction of 41.66 % 41.66 % 41.66 % 33.33 % Grazing, FecCYC Refuge Population 0 .02 0 .02 0.02 0.02 Sinking Velocity, w5 0.0 0 .0 0.0 1.0 md1 47


Nitrate profiles are taken from BATS cruise 52 (January, 1993, Knap et al., 1995) Ammonium is assumed to be constant at 5 nmol at all depths. Fecal pellet carbon is set to 0 .002 1-1M C. Fecal nitrogen is converted from fecal carbon at a C:N of 8.0 Dissolved inorganic carbon concentrations from surface to 250 meter are taken from Knap et al. ( 1995) during BATS cruise 52;. The remaining DIC concentrations to 1000 meters are from GEOSECS data in the North Atlantic (Sundquist, 1985). 2 .1.1 0 Boundary Conditions and Numerical Considerations Fluxes at the sea surface are defined to be zero, with the exception of C02. The bottom boundary is assumed to be permeable Concentrations forB, DOC3, N03, and DIC state variables (as well as their nitrogen counterparts) at 1002.5 meters (or NZ plus 1) are set to equal their initial values at 1000 meters. Pi, DOC1 DOC2, Nf4, and F state variables are set to equal 0.0 at NZ plus 1. Since these state variables are primarily driven by production in the surface waters and are non-conservative, setting these values to zero at NZ plus 1 allows for one-way flux from the simulation All the partial differential equations are solved explicitly with a forward-in-time centered in-space method (0' Brien, 1986). The von Neumann stability condition for the diffusion terms in the above equations is 2AtKzf.6.z2 1.0. The maximum Kz is ca. 200 x 10 -4m2 s-1 and requires a At of 156 seconds to satisfy the stability criteria. It is set to 150 seconds and satisfies the biological considerations of nutrient utilization as well (Bissett et al., 1994). The particulate sinking terms are solved with an upwind difference technique. The artificial viscosity from this method is assumed to have a negligible impact on the numerical solution (Walsh and Dieterle, 1994) 48


2.2 Biological Results After 3 years of numerical transients from the initial conditions, the simulated seasonal cycle of the state variables are similar from year to year under the same repeated forcing functions. The results of three cases are thus shown from the fifth year of the simulation; i.e., Julian day 1460 to 1825. The first case (Case 1) is forced by strong mixing below the surface mixed layer (8.00 cm2 s-1) and utilizes a simplified biological parameter set. A prior simulation analysis (Bissett et al., 1994 ) found that a vertical eddy diffusivity of 8.00 cm2 s-1 induced a nitrogen flux at 400 m that approached that estimated by radiotracer measurements (Jenkins 1988) In this simplified biological version there are no forms of DOM; i .e., DOC exudation and DOM release by grazing are set to zero (Table 2.3a); there are no bacteria; there is no nutrient regeneration from fecal pellets ; i.e the fecal pellets remain intact, and are removed by sinking through the bottom boundary; there is no nitrification of ammonium ; and there i s no nitrogen-fixation The prior simulation analysis employed a s imilar simplified biological model. The differences between the simplified model and the complex model described in Section 2 are summarized in Table 2.3b. The second case (Case 2) is forced by weak mixing below the surface mixed layer (0.37 cm2 s-1) and utilizes the simplified biological parameter set. The vertical eddy diffusivity of 0.37 cm2 s-1 was estimated from microscale gradients in temperature and velocity in the Sargasso Sea (Lewis et al ., 1986). The third case (Case 3) utilizes the weak mixing from case 2 and the complete biological parameter set described in Section 2. 49


Table 2.3a. Phytoplankton Parameters for the Simple Model FGl FG2 FG3 FG4 carbon excretion,ei I 0 .00% I 0.00 % I 0.00% I 0.00 % Fecal Fraction of 20.00% 20 .00% 20.00 % 50.00 % Grazing, FecPEL DOC 1 Fraction of 0.00% 0.00% 0.00 % 0.00% Grazing, FecDOC Recycled Fraction of 80 .00% 80.00% 80.00 % 50 .00% Grazing, FecCYC Table 2.3b. Conceptual differences between the Simple and Complex Models Case 1 Case2 Case 3 DOC cycling I no I no I yes Heterotrophic no no yes Bacteria Fecal Pellet Regeneration no no yes Nitrification I no no I yes Nitrogen-Fixation I no I no I yes Minimum Vertical Eddy Diffusivity, Kz 8.00 0.37 0.37 cm2 s-1 2 2 1 The Strong Mixing, Simple Biological Case Case 1 organic The seasonal peak in daily integrated primary productivity, -850 mg C m-2 d-1, occurs on Julian day 78 (Figure 2.2), resulting from increased so lar radiation (Figure 2.3a; see also Figure 3.3) and water column stability (as evidenced by the decrease in vertical mixin g; Figure 2 3b). The seasonal cycles of the individual functional group carbon concentration can be seen in Figure (2.4). FG4 dominates during the spring when the increa se in light energy (Figure 2.3a) creates a favorable environment for the rapid uptake 50


Daily Primary Productivity 800 600 "'C::l ':'a u 00 400 a 2 0 0 / / /o I ,/ \ /.. .. All FG 1 FG2 FG3 FG4 / ::;.: ------0 100 200 julian d a y 300 Figure 2.2 Cas e 1 integrated prim ary productiv i ty for each functional group and the total integrated primary productivity. 51


A B Downwelling lrradiance at noon E. ? STin; .-.-.-.-.-.-.-.-::::::: ::: : ... JO" ------, ................ 100 200 llfinm's' V e rti c al Eddy DiffusivityK. c:m's' Realized G r owth Rate Group I ----- . "'----. 2--------too 200 carbon s pecifi c g rowth d 300 300 Figure 2.3 Case 1 (a) downwelling irradiance at noon; (b) vertical eddy diffu s ivity; (c) -(f) realized growth rates for functional groups 1 4, respectivel y (co ntinued on n ext page) 5 2


D E F 0 100 200 300 carbon specific growth d Realized Grow th Rate Group 3 0 100 200 300 carbon specific growth d Realiz e d Growth Rate Group 4 o--------====;;o _-:..=====--------1 00 200 carbon specir,c growth d 300 Figure 2.3. (continued) 53


A B -250 0 c 0 C::S-"""0.)00 o.lOO o=----------------------. o oo---------:- -.-.-----.-. --. --. 007$" -.. -.---.-. "0050" -.---. - -.. -.... -0.025"------......... . 0.075" -. ----------- -. 0 02J" .. - 1 00 200 300 J.lmol C liter' Ph y toplank to n Grou p 3 0. 100 ---------------o=--------- .-.-- - -. . --.-----. "00)(1 . ... ---.... : ............. "0025' ............... ... .. 200 f.lmOlCliter 300 Figure 2.4. Case 1 (a)-(d) carbon biomas s for functional groups 1 4, respectively; (e) total phytoplankton carbon biomass; (f) fecal pellet carbon stocks ( continued on next page) 54


D -E F Figure 2 .4 (continued) --------0 .20()--------. ......... 0 .073" -.---.-.. . . 0 .050"-- -.-. 0 .025 200 JllllO ICliocr Ph y t op lankt o n G r oup All .... 1.00 0 .10 0 .60 ----------o .co---------tOO 0..05" .---- . 2 00 F ec al P elle ts 300 . .. . ... ... .... .. . . . . 0 0 55


of nutrients, which were brought upward from the deep waters during the winter mixing. The realized daily carbon specific growth rate of 1.4 d-1 is then nearly double the next fastest growing functional group, FG3 (Figure 2.3c 2.3f). The spring peak in total phytoplankton carbon (Figure 2.4e) is matched by a peak in the fecal pellet carbon (Figure 2.4f). Phytoplankton C:N ratios are at a minimum in the surface waters for all groups at -5.5 (Figure 2 5) during this peak in production. This demonstrates that light is the limiting factor for growth throughout the euphotic zone, as the accumulation of excess nitrogen (C:N ratio less than 6.625) only occurs during light limitation The spring peak of total chlorophyll a concentrations (Figure 2.6e ), -0.35 mg m-3 occurs during the spring at about 40 m, it is dominated first by FG 1 (Figure 2 6d ) followed later by FG3 (figure 2.6c). This shift within the phytoplankton assemblage can also be seen in the spring cycle of chlorophyll c and photosynthetic carotenoids (FG4 accessory pigments), and phycoerythrin (FG3 accessory pigments; Figure 2.7). The effects of winter mixing, which rea c hes its maximum at the beginning of February ( Figure 2.3a), can be seen by the deep penetration of chlorophyll a (Figure 2.6e ) and phytoplankton carbon stocks (Figure 2.4e) on Julian 44. FG4 and FG3 are able to increase their chlorophyll a and carbon stocks, despite deep mixing as a result of their higher growth rates (Figure 2.3e and 2.3f) These functional groups are poised to greatly increase their stocks at the onset of stratification, resulting in the springtime maximum of organic carbon and chlorophyll a. Vertical mixing continues to decrease (Figure 2.3b) as the simulation moves into summer. The incorporation of nitrogen into particulate form, and its sub s equent packaging into sinking fecal pellets by grazers, drives surface nutrient stocks to their lowest levels ( Figure 2 8b and 2.8c). During this period FG1 and FG2 growth rates are the g reate s t ( >0.4 d-1; Figure 2.3c and d), and their contribution to integrated production i s maximal (Figure 2.2 ) By late summer, the diffusive flux of nutrients from below th e maximum of winter mixing influence (200 m) has replenished the base of the euphotic zone, as 56


A 0 100 E ISO SO 0 B 0 S O t: e ISO SO 0 c 0 t: e ISO 0 C:N Phy t o plankt o n Group-I 100 200 C:N Phywplankton Group 2 . ............................ : : .... --.: "'' 100 200 C:N Phy t oplankton Group-3 . ----::: ... S.!" ---------------------100 200 300 300 300 Fig u re 2 5 Case 1 (a) (d) car b o n to n i tr ogen ra t ios for f u nc ti o na l g r o up s 1 4, --------r espective ly; (e) t ota l p h y t o pl ank t o n carbon to nit roge n r a t ios; (f) fecal pe ll et car b o n to n itrogen r at i os (conti n ued on next page) 57


D 0 : -250 0 E 0 -50 f-100 f--150 f-200 1-250 0 F 0 1--200 -400 ---: v. .1i E 600 -800 -C : N P h ytop l ankto n G r o up -4 100 200 300 C : N Phytoplank to n Grou p All .. . --. .: -: ..... --. .......... ---........ -.---.-.... ... : ... . . :::::: ::::::::::::::::::::::::::::::: ::w---100 .. .. --200 300 C:N F ec al P e ll ets -= 6.5' ----:::::::: .s-. -_-... .. _.:::: .-:::::: :: : __::: .. ::: _..... ---. -----. ... .. : __ :_: Figure 2.5 ( continued ) 58


A B -----. - -.-. "O.()j.Q".-.- -- --.--. . --------.--. 0 .0"15 -. -----.( ___ c __ ___________ _,o:> ... . -.. ---.. --.-----. -----0.075"-------100 O --.-.--.------.---.--. --0015" -----.-------.-.--.-------. 0..02.5".-. ------200 118 Chi a liter' ..,.. -300 300 Figure 2.6. Case 1 (a)-(d) chlorophyll a concentrations for functional groups 1 4, respectively; (e) total chlorophyll a stocks (continued on next page) 59


D .'. ._ :: :." ; :::::: .: . 0 .07J".-.... -.. -.---.-.-. ------------------"0. 0$0-------------- " O OlS" "". -----E .. o. o,,---------QOj()------------o .ou-.-------Figure 2.6. (co ntinued) 60


A -25 0 0 B 0 -50 f--100 9 ISO f--200 25 0 0 c . . . . . . . -oo6 .. o.oa ... . .... ......... "' '-_ ---------------0.20 . ..... . ... ---100 ----- -.-----0J)4"-.--.-----------------om ---200 1'8 C hi blitu Chi b Ph y toplankton Group 2 "001---. -- ---------. ------100 100 ----------------------200 1'8 Chi b l iter HPUB Phy toplank ton Group 3 I S oa ----.. ------. 0 6 "-----.-------. --.---.... --o--.-.------.-... 0 2 . 200 )l g HPUB liter 300 . oOl 300 300 Figure 2 7. Case 1 accessory pigment concentrations(a) and (b) chlorophyll b concentrations for functional groups 1 and 2, respectively; (c) functional group 3 phycoerythrin concentrations ; (d) and (e) functional group 4 chlorophyll c and photosynthetic carotenoid concentrations. (continued on next page) 61 ---


F ( 100 E Chi c Phytoplankton Group 4 O.ozao---------------- 0 .00>() o oon ......... --200 Ch i c liter 300 PSC Ph y t op lankt on Group 4 Q : __ :._:;._: __ :_: :_::._:_:_; __ :_:_;: . _._,., ... .... . Figure 2. 7. (continued) o .gi1 0 - ........... . .. ..................... 'OJ).&'. o m --.-200 PSC liter 62 . -. 0.02 . 300


A B !S e c !S e 0 v 1120 -:: Inorganic Carbo n f.-2 130 mo-:: -200 21.., 2140= mo mo -2160 2160--= -400 I--2170 2110-2110 2110-: 2190 ,. .. ,.,. 2200--600 2210 wG= 2UO 1!30 mo--800 2240 v..:O: 22>0 2:!.10-"60 2.160--1000 0 100 200 C liler' 300 -200 -400 -600 -800 -1000 _-_ -_-_-_-___-:-_.-;;:::::::::::;;:;::::::: :_:_:_:_:_:_:_:_ :_:_:_:_:_:_ :_:_:_ ------:_:_:_:_ 2 00 4.00 4 00--:: 6 .ooa .oo a .oo-10. 00 10. 00----o:: 11. 00 12.0014.00 ... oo-= 16. oo 1100 11.oo-20. 00 2D. oo 0 100 200 300 Ammonium -0.200 0.200--: -200 Fo200 -400 t-600f-o-'o.o,----------------------------------.-------o .ow----------------------------------------o .oso---------------------________ _____________ ___ '0.0$0".-.------.... ---.---.-... ----.-.. --------. -8001-o. 02J . .. --. --- -.----.. - -- --. . -.-. -.. .. -.-100 200 300 Figure 2 8. Case 1 (a) dissolved inorganic carbon stocks; (b) nitrate stocks; and (c) ammonium stocks 63


evidenced by the 0.5 jl.mol N03 isopleth crossing 100 meters (Figure 2.8b). This allows phytoplankton carbon levels to again approach their summer peak in concentration (Figure 2.4e). The growth rates ofFG3 and FG4 reach their summertime maximum of -0.6 d 1 (Figure 2.3e and 2.3f), however their biomass levels are lower than FG 1 and FG2 Thus, the summer peak in phytoplankton carbon which is similar to the carbon peak during the spring, does not have a commensurate increase in fecal carbon levels, as the summer phytoplankton biomass is now dominated by the smaller functional groups While the summer growth rates for each of the functional groups are similar (Figures 2.3c through 2.3f), the depth of each functional group's maximum growth rate is different. While FG 1 and FG2 have the greatest affin i ty for nitrogen the growth of FG 1 is inhibited by strong sunlight. Hence, FG2 is the dominant Prochlorococcus-like functional group in the bright sunlight surface layers (Figures 2.4b, 2.6b, and 2 7b) FG 1 has a greater ability to harvest light and its peak in growth, maximum carbon concentration, and pigment stocks can be found at the deepest levels of the euphotic zone (Figures 2.3c, 2.4a, 2.6a, and 2.7a). FG3 and FG4 compete for the region between the lowest light levels and the lowest nutrient concentrations. As FG3 is a little more efficient at nutrient uptake and light utilization, it establishes a growth rate peak slightly deeper than the peak in FG4. (Figures 2.3e and f). This can also be seen in their respective carbon and pigment stocks (Figures 2.4 2 6 and 2.7) As summer progresses and nutrients become limiting for growth in the surface layers the C:N ratios for FG2, FG3, and FG4 approach their maximum values (Figure 2.5). FG 1 is inhibited by light at the surface, hence it does not exhibit any carbon based growth. However its nutrient uptake is unaffected until it reaches its minimum C:N ratio The onset of fall leads to increases of mixing and lower sub-surface light levels (Figures 2.3b and 2.3a), forcing a shoaling of the chlorophyll a maximum (Figure 2 6e). This shoaling is followed by a subsequent dilution of biomass and pigment concentrations throughout the mixed layer as mixing increases (Figures 2.4 2.6, and 2.7). Surface 64


phytoplankton carbon concentrations begin to increase while the deep concentrations decrease, resulting from increased mixing of nutrients towards the surface, lower light levels (Figure 2.3a and 3 .3 a), and a shift in the phytoplankton assemblage towards the larger functional groups. The increases of growth rate and carbon biomass are matched by a greater flux of fecal carbon, which results from both the increased total phytoplankton carbon production (Figure 2 2) and the greater fraction of FG4 carbon (Figure 2.4 ). The vertical structure of C : N ratios is homogenized (figure 2.6) after the onset of winter mixing resulting from dilution and increases in the nutrient concentrations in the euphotic zone (Figures 2 .8b and 2 8c) 2.2.1 2 inorganic While there is relatively little seasonal change in the DIC concentrations below 200 rn, the surface concentrations follow the seasonal cycle of mixing (Figure 2 8a). Maximum surface concentrations of DIC of 2124 !lrnolliter-1 (2070 !!mol kg-1) are found on Julian day 121 (Figure 2 8a) approximately 80 days after the deepest winter mixing (Julian day 44; Figure 2.3a). This surface maximum results from the upward mixing of DIC-rich waters and in-gassing when the water temperatures are cold (Figure 2.1), which overwhelms the reduction in surface DIC during the peak in primary production on Julian day 78 (Figure 2.2). The minimum stocks of DIC of 2113 !lrnolliter-1 (2060 !!mol kg-1) are simulated during the summer on Julian day 292. This minimum in DIC concentrations resulting from the out-gassing of C02 and the incorporation of DIC into particulate organic carbon. The deep water stocks of N03 also show little change over the course of the year (Figure 2.8b) below a depth of 200 rn. Since there is no exchange of N03 with the atmosphere the surface concentrations are maximal during the peak in winter mixing on Julian day 48 at 0.37 !!mol N liter-! (Figure 2.2b). The minimal surface concentration of 65


N03 (Figure 2.8b) ofO.OO 11mol N liter-! is reached during the late summer peak of stratification on Julian day 165 (Figure 2.2b) The concentration of Nl4 displays a subsurface maximum at depth 100 200 m (Figure 2.8c) as a result of phytoplankton utilization in the euphotic zone and zooplankton excretion throughout the water column. Thus, the gradient for Nl4 is positive towards the euphotic zone and deeper waters. This vertical s tructure sets up a diffusive removal of nitrogen from the surface waters, as the N03 incorporated as phytoplankton biomass is released to the Nf4 pool via grazing, which then fluxes to depth. Like the nitrate pool, surface concentrations of Nf4 are greatest during winter mixing and lowest during summer stratification (Figure 2.2b). 2.2 2 The Weak Mixing, Simple Biological Case-Case 2 2 2.2.1 organic This case considers a 20-fold weaker minimum mixing in the regions beneath the surface mixed layer (below the bottom contour of Figure 2.9b). As a result, the seasonal cycle of phytoplankton primary productivity and biomass is much the same as in the prior, strong mixing case, but at much lower levels. The seasonal maximum of daily productivity occurring on Julian day 52 is now only -103 mg C m-2 d-1 (Figure 2.10). Phytoplankton carbon is dominated by FG 1 (Figure 2.11 ), as opposed to FG4 in the strong mixing case and is followed by FG2, FG3 and then FG4. The relative shift in species dominance is also seen in the simulated spring chlorophyll a and accessory pigments (Figures 2.12 and 2.13) The spring growth rates for all functional groups are -0.2 d-1 (Figures 2.9c-2.9f) A 19and 27-fold smaller fecal pellet concentration and flux respectively were computed after the meager spring peak production in this weak mixing case (Figure 2.9), compared to the strong mixing case Chlorophyll a concentration reach a seasonal maximum of -0.15 66


A B c 100 Downwelling lrradiance at noon E. 400 .;: / yy?t< !:T ')() ........... . ---- - 10 -- -! Vertical Eddy Diffusivity K. a.o .. ------. -.a .-. -: : .. :.::.:. -=o. 0? 200 carbon specific growth d .. . . . ::: :.o. 0 .-'-::::.. ---:..... -.-.. 300 Figure 2 9 Case 2 (a) downwelling irradiance at noon; (b) vertical eddy diffusivity ; (c) -(f) realized growth rates for functional groups 1-4, respectively ( continued on next page) 67


D 200 carbon specific &rowth -d E Realized Growth Rate Group 3 01 -250 0 tOO 200 300 carbon specific growth -d F Realized Gr owth Rate Gro u p 4 0 -50 r-l -tOO -t: e t50 '--200 f2 S O 0 tOO 200 300 carbon specific growth -d Figure 2.9 (continued) 68


-: >. ro '"0 <:'E u 0!) E Daily Primary Productivity 150 100 50 . : . I .... ... ;----..... All FG 1 FG2 FG3 FG4 0 .---... /. ;;...= ... =-=-:--...:-_ :=--::--=-:0 100 200 julian day 300 F i gure 2.10 Case 2 integrated primary productivity for each functional group and the total integrated primary productivity 69


B -250 0 c 0 "' -50 100 9 --ISO f-200 f--250 0 100 .. .. .. .. 100 200 Phytoplankton Group 3 300 "0.02j --.-.. .... -.... .. -.. -...... ... "00"..5 :: ---: : :. :. :. :_ :. : :. :. :. :. :: o. ow----_ _ _____ .. .----200 300 ---Figure 2.11. Case 2 (a) (d) carbon biomass for functional groups 1 4, re s pectively ; (e) total phytoplankton carbon biomass; (f) fecal pellet carbon stocks (continue d on next page) 70


D 0 -50 ,...--100 ,...__ 9 -150 -200 -2 5 0 0 E 0 -250 0 F 0 2 00 1--400 r-9 E 600 1--8001-g t: 0 ;; i \ ' -1 00 100 Figure 2.11. (continued) Ph y toplankton Group -4 -. .. "0.02.1 .. 200 )1mo1Cliter' Phyt o plankton Group All -3 00 --0 0!5.---. -.--.-. --.-----------.--. -.-.. -.-2 00 )lmo l C liter' F e cal Pellets --.. ... 0 .006]"----- -.--.-. -.. ----71 300 --


A B :: s e c t! s e --. . .. _. -: 002$' :: :::."_:: .. ---. -250 0 0 -50 -100 t--150 t-2 00 1--250 0 0 -50 t--100 1--150 ,__ -200 ,__ -250 0 . . . . 100 - ----.......... . -oo!IO -------'002$'----------200 Chi a tiLer' Chi a Phytoplankton Group 2 300 ---------------. -.---------"00:!5 -. -.------.-......... -... --..... -. -.. -. .-: _.::: _:: :. : o .050 . . ----------100 100 "0.02$. -.-.-.- --.--- -200 C hi a l ilt r Chl a Ph yto plankton Group 3 _-: _-: __: _.... .-.-........ 200 a tiLer' 300 300 --Figure 2.12. Case 2 (a)-( d ) chlorophyll a concentrations for functional groups 1 4, respectively ; (e) total chlorophyll a stocks. (conti nued on next page) 72


D 1:? g e E 0 1-1 00 1-I S O 1--200 -250 0 0 :. ... ' : 8 . ;; ----. Chi a Phytoplankton Group 4 -: ::. -.. 1 00 200 300 Chlaliltr Chi a Phytoplankt o n Group All ...... .. o:o:O. . .: . . . -. -_ -_ ---. . . . . -. . . -. .: . -. . .. ... ... ...... o .ru ... ----. ... ----.0 1 00 200 300 Chi a l iltr Figure 2.12. (continued) 73


A B 200 Chi b liter' 300 Chi b Phyto p lank ton Grou p 2 -200-. o.oao--. Oo10 200 )18 Chl b li ter' HPUB Ph yto plankton Group 3 300 -o. z -.. ... .... ..... ....... __ .......... -.:-:-:-:-::: ... o .. --------200 HPUB liter' 300 Figure 2.13. Case 2 accessory pigment concentrations(a) and (b) chlorophyll b concentrations for functional groups 1 and 2 respectively; (c) functional group 3 phycoerythrin concentrations; (d) and (e) functional group 4 chlorophyll c and photosynthetic carotenoid concentrations. (continued on next page) 74 ----


F 0 Chi c Phy toplankton Group 4 -50 f.-100 f.e ... ISO f.... ... -.. --250 0 100 200 300 IISChlcliu:r' E PSC Phytoplankton Group 4 0 -50 f.-100 f..-. q'l>J. 150 f-200 1---250 0 100 200 300 118 P S Cliu:r Figure 2.13. (continued) 75


mg m-3 during this peak period of production (Figure 2.12e ), approximately half of the chlorophyll a peak in case 1. A larger growth rate of -0.4 d-1 is simulated for FG1 during the summer period. In spite of this larger growth rate, fecal concentrations are 8.5-fold smaller, following the reduced production by FG4 during the summer (Figures 2.9f and 2 11f) The depth-integrated primary productivity is also lower than the seasonal maximum in the spring (62 versus 103 mg C m-2 d-1, Figures 2.2 and 2.10). The secondary peak of summer production occurs during the peak in light penetration (Figure 2.9a). This light penetration to the base of the nutricline at -175 m is spectrally weighted towards the blue end of the spectrum. FG 1, with the highest photosynthetic efficiency is best able to utilize this dim blue light. This functional group achieves a carbon growth rate of -0.43 d-1 on Julian day 178 (Figure 2 9c). The peak in phytoplankton carbon concentration also occurs during this period 0.60 C liter-! just above the nutricline (Figure 2 11e) The summer period is marked by the simulated peak in annual chlorophyll a concentration of -0.24 mg m-3 (Figure 2.12e). Again the dominance ofFG1 is noted by the peak in chlorophyll b (Figure 2 13a), at the depth of the peak in phytoplankton carbon (Figure 2.11e). The reduction in insolation (Figures 2.9a; see also 3.3) in the fall is marked by a decrease in productivity, biomass, and growth (Figures 2 10 2.11 and 2.9). The peak in biomass begins to shoal, following the depth of the euphotic zone (Figures 2.11e and 2.9a). The simulated fall biomass peak maintains its structure and position (Figure 2.11) until it is destroyed by the winter mixing cycle. The low minimum mixing of case 2 reduces the upward flux of nutrients and effects an increase in the functional group carbon to nitrogen ratios (Figure 2.14). These ratios are at their maximum levels for all three cases over the annual cycle, reflecting the low levels of nutrients in the euphotic zone (Figure 2.15). 76


A 50 --200 t--B -sot.. -!SOt-200 t-250 0 100 200 300 c 0 C : N Ph y toplankton Group 3 -SO -ISO-200--Figure 2 14. Case 2 (a)-(d) carbon to nitrogen ratios for functional groups 1-4, respectively ; (e) total phytoplankton carbon to nitrogen ratios; (f) fecal pellet carbon to nitrogen ratios. (continued on next page) 77


E : . -250 0 100 200 300 F C:N Fecal PelleiS 0 :6._1...:. ... r-. ....... ................ ......::::::::..::::::::.. -400 -B e --0 100 200 300 Figure 2.14. (co n tin u ed) 78


A B c 0 -200 I--600 -800 r--1000 0 1120 1 00 Inor ganic Carbo n :!1:00 mo 2 1<0 2UO 2160 2 1 70 2110 1200 221 0 mo lliO ""' 2250 2260 200 11mol C U ter 211D -m/ 2120 mo= 21<0 -,.,._ 2160----= 2190 noo-2210-;::;; m r 22 .. mo 1260 -300 N i trate l .OO 2 .00-4.00 .oo-" 6.00 6 .00 -400 a.oo a.oo-= 10.00 1o. oo -1 2 00 14.00 l<. oo-= -600 16.00 16.00--= 11. 00 11. 00 -20.00 :!0.00--= 100 200 1'1110 1 Ntiter 300 Ammoni um ,,, -,,,, ,_, ,, ,,, ,, , , -" "'-'.--.<...,.,., I 2001-- "0. 015" . ----.. --- ---. 0 .07$""". ---- --- -.. -. -- ---o .o$0 --------o .oso-------------------------------------------"0.0"..1"----------------------'-------------------4001-6001--8001-1 00 200 IJmoiNtiter' 300 Figure 2. 15. Case 2 (a) dissolved inorganic carbon stocks; (b ) nitrate stocks; and (c) ammonium stocks 79


2.2.2 2 inorganic The seasonal cycle of DIC in this second case is similar to that of the strong mixing case. At the surface, DIC stocks peak at 2124 )lmolliter-1 (2070 )lmol kg-1) on Julian day 135 (Figure 2 15a), and are a minimal2106 )lmolliter-1 (2053 )lmol kg-1) on day 292 (Figure 2.15a) The maximum surface DIC stocks in this weak mixing case are -11 )lmol liter-! (-10 )lmol kg-1) greater than the surface DIC stocks of case 1. This is the results of both the lower levels of spring primary production (Figure 2 9) and the dominance of phytoplankton assemblage by FG 1 (Figures 2.11), whose grazers have a higher recycling efficiency (Table 2 2). The minimum DIC stocks for case 2 are -7 )lmolliter-1 ( -7 Jlmol kg-!) lower than the surface DIC stocks of case 1. Since the mixed layer depth is very shallow during the summer period ( -15 m; Figure 2.1 Ob ), these lower minimum surface DIC stocks are simulated by the smaller diffusive fluxes of DIC from depth, as the vertical eddy diffusivity below the surface mixed layer is 20-fold smaller in this weak mixing case. At depths< 200m, seasonal changes N03 and NH4 are now slight (Figures 2.15b and 2.15c) This reflects the slow upward diffusive flux across the permanent thermocline. At deeper depths, the subsurface maximum is half that of case 1, and penetrates only to 300 m compared to 800 m in case 1. 2.2.3 The Weak Mixing, Complex Biological Case Case 3 2 2.3.1 organic The seasonal cycle of primary production in case 3 is again similar to the previous simulations The daily production (Figure 2 16a) of the more complex biological model falls between the values of weak and strong mixing scenarios of the simpler biological model. The maximum production occurs during the spring on Julian day 78 at a rate of 544 mg C m-2 d-1 (Figure 2.16a). A secondary summer peak in production occurs on Julian day 236 (303 mg C m-2 d-1), resulting mainly from nutrient supplies from nitrogen 80


A B Daily Primary Productivity -: :>.. ro "'0 "' E u b.() E -: :>.. ro "'0 "' E z 0 E E 800 All FG 1 FG2 FG3 FG4 600 400 200 / / 0 :; ; :; ; \ :::.;:_; ."::::" ,:; "'-; 0,;:. -:..::c. ::.c.:?.' .... 0 1.0 0.8 0 6 0.4 0.2 0 100 200 julian day Integrated Daily Nitrogen Fixation 100 200 julian day 300 300 Figure 2.16. Case 3 (a) integrated primary productivity for each functional group and the total integrated primary productivity ; and (b) daily integrated nitrogen-fixation rates 81


fixation (Figure 2 16b ) This compares with a production on this day o f 398 and 49 mg C m-2 d-1 for case 1 and case 2, respectively Through the late winter-spring period the primary production i s once again dominated by FG4, like case 1 followed by FG3 FG2 and FG 1 (Figure 2 16a). During the period of nutrient depletion after the spring bloom, FG 1 briefly dominates before giving way to FG2 During the summer period, FG4 rebounds to second place in response to nitrogen fixation within the surfa c e waters FG3 competes marginally until the fall overturn when it is the co-dominant functional group with FG4 ( Figure 2 16a). The maximum growth rate for each functional group occurs at different depth s and day s, effecting competition and s easonal succes s ion (Figures 2 .17c-2 17f) For FG 1 the maximum realized growth rate is 0 34 d-1 at 87.5 monday 194; FG2 has its maximum of 0.40 d 1 at 7 5 monday 227 ; the maximum for FG3 is 0 55 d-1 at 5 monday 79 ; and the maximum growth rate for FG4 is 0.94 d 1 at 2 5 monday 67. The seasonal contributions of each functional groups to particulate carbon (Figure 2.18) chlorophyll a (Fi gure 2.19), and accessory pigments (Figure 2.20) f ollow their seasonal growth rate (Figure 2 17) The s i mulated transitions from FG4 to FG3, and thence to FG1, are similar to the observed succession of golden-brown algae in February March to a shallow maximum of S y ne c ho c occus in April to a deep maximum of Prochlorococcus in May June around Bermuda (Michael s et al. 1994b) Like the previous cases 1 and 2, their depth-dependent populations depend upon the interplay of growth as driven by spectral lightand d i fferential nutrient-utilization, and grazing losses. Cellular carbon to nitrogen ratios reach their maximum values for the groups during the nutrient-deleted surface conditions following the spring bloom except for FGl who s e carbon growth is inhibited at strong sunlight (Figure 2.21) The C : N ra tio s stay near their maximum values during summer until winter mixing brings nutrients into the euphotic zone Carbon to nitrogen ratios also decrease slightly during the peak period of nitrogen fixation (Figures 2.21 and 2 16b ) The total C : N ratio for all groups fluctuates between 82


A IOOf-p ISOf--200fB 0 c o.z--. Downwelling Irradiance at noon -E. --. 10' - -.-------------------Vertical Eddy Diffusivil y K, a..o ... -. .. --.,,--..-_ : : .. -o .. ----------1 00 200 cm's 100 200 carbon s pecific zrowth d' ---.. -..... . 300 O l 300 Fig ure 2 .17 Case 3 (a) downwelling irradiance at noon; (b) vertical eddy diffusivity; ( c) -(f) realized growth rates for functional groups 1 4, respectively. (co ntinued on next page) 83


D E 0 F Figure 2 17. (continued) 100 200 carbon specific growthd R ealize d Growth Rate Group 3 200 carbon specific growth -d Realiz e d Growth Rat e Group-4 r----------o ,---------100 200 carbon specific growth -d 84 0 < 300


A B c -250 II 0 2 5 0 () 0 Phytopl ankton G roup I _;..;__ _ -_ -_-_ _ _ --_-__ _-__.:...:_-.:... ___! <' : -_--: : -:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-:-: -:-:-:-_-_-_-_-_-_-__ . 100 2IXI 300 J.lmoiC iitcr Phy toplankton Group 2 0.300 .. .-_:..:::: : .. --.. . 1 00 2!XI 300 J.lmoiCiitc r Phytoplankton G r o up 3 ./ . .. . .. . . : ........................................... .,. .. 0 HXl 2!XI 300 J..lmoi C iitcr' Figure 2.18. Case 3 (a)-(d) carbon biomass f o r fun ctional gro u ps 1 4, respectively; (e) total phytoplankton carbon biomass ; (f) fecal pellet carbon stocks. (continued on next page) 8 5


D F Phytoplankton Group -4 ---------------------woo .... -_ -_ -_ -_ -_ -_ -_ -_ -_ _ _ -_ -_ -_ -_ -_ -_ -_ -_ -_ -_ -_ _ -_ -_ -_ -_ -__ tOO .. "0.025"-----. ----. 2 00 llffiO I C l i ter o .os -o .as---- - --- tOO 200 llffiOICiiter Fecal Pellets -. "0,(06)'. 0 Figure 2.18 (c ontinued ) 86

PAGE 100

A B c .. . o.o2l --....... .... --. --..... -..... ..... o.010 0 1 00 200 300 0 ""' ISO I--2001-.-. 1 00 (J . .. . . :.!. .. .. . . .. 11g Chi a liter' Chi a Ph y t o plankt o n Gr o up 2 200 11g Chi a l iter' Chi a Phyto p lankt o n Group 3 ----o .nso ::... -. .oo10 o:
PAGE 101

D E jg o .. . -. 100 200 118 Chla Q02j :: : C hi a Phytoplankton Group -All 0.300 O.lOO . . : : :: : : : : : : : : ::: .-::: .:: :: _-. -- ......... ... o .ctt..s. Figure 2.19 (continued) 88 300

PAGE 102

A . . . . . . . . . .. . . . . . . _.,., g ... : : : : : : : -_ : : : : : : : : : : :: : : : : : : : : : : : :": : : :" : : :' : ..... 0 .10 !If>' .. > 0 100 B 0 0,0;! 100 t! 3 ISO 200 0 100 c 0 .. 100 200 11g Chi b lill!r C hl b Phytoplankton G ro u p 2 ..... __ . . _ _ . __ _. ... o.ol 200 11g Chi blill!r-' HPUB Phytoplankton Group 3 200 11gHPUB lill!r' oOl 300 300 300 .. .. Figure 2 .20. Ca s e 3 acce ss ory pigment concentrat ions-(a ) and ( b) chlorophyll b concentrations for functional groups 1 and 2, respectively; (c) functional group 3 phycoerythrin concentrat ions ; (d) and (e) functional group 4 chlorophyll c and photosynthetic carotenoid concentrations (contin ued on next page) 89 --

PAGE 103

1 00 Chi c Phytoplankton Group 4 "O.OQl.S' .......... -200 J.l8 Chic 300 E PSC Phytoplankton Group 4 .... . ... ... .. :" .. ;: .... ... -:::.: .... Figure 2.20. (continued) -oor -----------90 :. .. g . . .... o.Ol . >:

PAGE 104

A 0 -50 -100 -!: e 150 -200 --250 0 B 0 50 r-100 r-!: e 1 50 r-200 r--25 0 0 c 0 -50 r-100 -!: e 150 --200 -250 0 C : N Ph y t o plank to n Grou p I .'::. .-.-.. _-___ _-.-__-_-___-_-_-_. : _-: :::: : : ::: : ::: : :5_5: : ::: :::-. -_ -. -.-.-.-.:: : _._ -.. . .. .. ':.-.-----..:: "::::,,, _. _. _. _. _. _. _. _. _._. _. _. _. _. _-_-.-. : :.--.-. o s s::;:: _.: ; : : . .. .... .. .. . 100 200 300 C : N Phyto pl ank t o n G r o up 2 -:._--::::::::::: _. -_-_._._._ -_ _ ____ _ _ _ _ _ _ _ _ _._ ---.: ;_;::. :: .: .::: : : : : _..-.-::: ... 100 200 300 C : N Ph y t o plankton Gro up -3 ... : _. .------. . -.::;:: ::: ::: ;:::: :--: :: :---s; :----... :._ --:::: ::: ;: .-.-: ... : -. 100 200 300 . -. ... -------Figure 2.21. Case 3 (a) -( d) carbon to nitrogen ratios for fun ctio nal gro u ps 1 4, respective l y ; (e) total phytoplankto n carbon to nitrogen ratios ; (f) fecal pe llet carbon to n itroge n ratios. ( c o nt i n ued o n n ext page) 91

PAGE 105

E .,. "" ...... ,'.',(>. '.', V C : N Phytop l ankt o n Group 4 ..... : ... -..--------. . --.-. .. ... -,:.-;;::. ... _,_ ,_, _,_:,::: ;: :::;; ; : : : : :_: : ; ;_:_:;; : : ;:::: C : N F ecal Pellets . .. ----------Figure 2 21. (continued) 92 :.:---. .. : .. -

PAGE 106

5 5 and 7.1 (Figure 2.21e) reflecting the C:N ratio of the dominant functional group in term of biomass. Fecal pellet concentration (Figure 2.18f) is greatest during the spring at the time of maximum productivity; a secondary peak follows in the summer. While the C : N ratios of the total phytoplankton biomass remain below 7.1 (Figure 2 21e ), the C:N ratios of the fecal pellets in the euphotic zone are as high as 8 8 just after the spring bloom and 8 6 in the summer (Figure 2.21f). This pellet ratio is higher than that of the total phytoplankton biomass ratio pellet, since pellet formation is strongly dependent on the amount of FG4 biomass. The C:N ratio ofFG4 is maximal during these periods of the year (Figure 2.21f) because DIN stocks are at their lowest (Figures 2.22b and 2.22c) Following the simulated spring bloom, the stocks of FG4 are still a significant fraction of total phytoplankton biomass (Figures 2 16a and 2.18). However, the low stocks of DIN reduce FG4's competitive advantage, and the growth rate slows (Figure 2.17f). Since this functional group has the capacity for growth at C : N ratios> 6 625, the C : N ratio of the group climbs, even as the total stocks decrease from grazing losses. Thus, the fecal pellet C : N increases (Figure 2.21f) During the summer nitrogen fixation adds nutrients to the surface waters that is quickly utilized by FG2 and FG4 (Figure 2 16b and 2.18) The nutrient additions are not enough to cause a bloom of these functional groups and grazing losses keep these populations in check. However, the relative increase in FG4 is strongly reflected in the fecal pellet concentrations (Figure 2.18f) and C:N ratio (Figure 2 21f). Note that the fecal C:N ratio in the deeper waters also depends on bacterial production. Since the C : N of bacteria is 5.0, bacteria production acts to lower the C:N ratio of the sinking phytoplankton fecal material. This is particularly evident in the decrease of fecal C:N ratios with depth between 100 and 150m (Figure 2 21f) during the summer despite the preferential rernineralization of nitrogen that acts to increase the C:N ratio of the sinking pellet. As a result of rernineralization fecal pellet stocks decrease with depth 93

PAGE 107

A B E c !,! E 0 K_ -200 -600 -800 ---1000 0 12 10 100 Inorganic Carbon 21l0 mo 2140 21>0 211i'M= mo-21 ..... 21>0-21ro-2170 mo-2190--:: mo-,,_ ,, .. ,,.-= ""'-300 Nitrate -200 -400 -600 800 1000 ..... .. _: :. :_ :_ :_ :_ :_ '-'-' ' ' ----' .. '{';:" -------.-.-.---.-----------.-.-.----------------.-.----------.;.-:--1- 00 .oo_::::::: 10.00 12. 00 1 6 .00 1!. 00 10.00-12. 00 -14. 00 -!Om-= r-----------------------------------,.,oo,-------------------------------------------,.,m-1-0 100 200 llffi O I N li!tr' 300 Ammoniu m 0 :-: .. -200'--400 --600 ---1000 0 100 200 300 llrnOINliltr' Figure 2.22. Case 3 (a) dissolved inorganic carbon stocks ; (b) nitrate stocks; and (c) ammonium stocks. 94

PAGE 108

(Figure 2.18f). This is in opposition to the results from cases 1 and 2 (Figures 2.4f and 2.11 f), where there is no remineralization of fecal material. The fecal carbon and nitrogen fluxes for case 3 are 0 .16 and 0.018 mrnol m2 y-1, respectively; this compares well with the 0.21 and 0.022 mrnol m2 y-1 from open ocean sediment trap data at 1000 m (Martinet al., 1987) Labile DOC exhibits a seasonal cycle (Figure 2 23a) with the maximum surface stocks occurring after the spring bloom on day 98 at 5.5 DOC liter-!. We shall find in the Bio-Optical Section that the CDOC component of total DOC stocks leads to a contamination of the satellite-sensed chlorophyll a biomass after Julian day 80 Most of the labile DOC is found within the upper 250 m ; within this depth interval the maximum labile DOC stock on day 93 is 0.71 mol DOC m-2, compared to 0.37 mol DOC m-2 on day 351 (Figure 2.23a). There is a smaller seasonal signal in the relict DOC pool (Figure 2.23b), with a slow increase in the surface waters. The simulated total DOC pools within the upper 250m ranges from 14.8 to 15. 2 mol DOC m-2, compared to observations of 14.9 to 15. 9 mol DOC m-2 (Carlson et al., 1994). The C : N ratio of the labile DOC pool fluctuates between a maximum of 5 7 following the spring bloom to a low value of 4 7 during the winter mixing (Figure 2.23c) The low C : N ratio ofDOC1 results from carbon-limited bacteria which are not impacted b) ammonium availability The heterotrophic bacteria are never nitrogen limited in this simulation and are always a source of NH4 from the utilization of DOM. The relict DOC C:N ratio only varies from 40.0 to 42. 6, increasing in the surface waters as a result of photolysis. The simulated seasonal cycle of nitrogen fixation (Figure 2 16a) provides a maximum influx of 1.84 mmol DOC/0 33 mrnol DON m-2 d-1 on Julian day 215 The model's minimum of zero release occurs on day 71. 95

PAGE 109

A Dissol ve d O rganic C ar b o n L a bil e 0 s 50 1 00 4 3 (J'J 2 .....
PAGE 110

c "-l .... (!.) ..... (!.) a D "-l .... (!.) ..... (!.) a C : N Labile Dissolved Organic Matter -501-100 1--150 r-200 -250 0 0 -42 -50 . . . .. 100 I 200 I 300 C:N Relict Dissolved Organic Matter I I ---41 -100 --150 --2000 100 200 300 Figure 2.23. (continued) 97

PAGE 111 inorganic The seasonal cycle of DIC over the upper 200 m is strongly dependent on the physical environment. Deep and rapid winter mixing destroys the prior vertical DIC structure (Figure 2.22a), with a downward displacement ofthe 2120-2130 J..lmolliter-1 isopleths Stratification in the spring and summer effects the establishment of a vertical gradient of DIC with depth, which results from both out-gassing at the surface and net removal of DIC from the euphotic zone by sinking organic Winter in-gassing and upwards mixing of DIC rich water still lead to maximal surface DIC values on Julian day 135 of2120 J..lmolliter 1 (2066 J..lmol kg-1), slightly lower than the maximum value of 2124 J..lmolliter-1 (2070 IJ.mol kgI) in the first case of stronger mixing. The minimum surface concentration of DIC is found on day 292 at 2098 J..lmol liter (2045 J..lmol kg-1). Net influx of C02 from Julian day 0 to 365 is equal to 0 .59 mol C m-2 y-1 (Table 2.4), in the middle of the estimate of 0 .22-0 83 mol C m-2 y-1 calculated with observed pC02 values at the BATS site (Bates et al., 1996). The seasonal change in nitrogen is again similar to DIC and there is little change below 200m. Above 200m, the 0.25 IJ.molliter-1 isopleth reaches the surface in the winter (Figure 2.22b), unlike the weak mixing scenario, case 2 (Figure 2.15b), more closely resembling the strong mixing scenario (Figure 2.8b). However, the source of nitrate at the bottom of the euphotic zone is now nitrification, rather than strong cross isopycnal thermocline mixing in case 1. The seasonal cycle of ammonium is very different from cases 1 and 2 (Figure 2.22c). With the introduction of explicit recycling via DOM pools and nitrification the NH4 budget is more dependent upon the imbalances between its release and remineralization, rather than its downward diffusive fluxes Unlike the fust two cases, there is no downward diffusive flux of NI4 below 200 m The Nl4 stocks are less than half as much as both case 1 and 2 beneath the sunlit waters. Since the sub-euphotic zone stocks in case 3 versus case 1 are much lower there is less to mix upwards in the winter; 98

PAGE 112

thus the winter euphotic zone stocks are much less, as we ll The grazing of p h ytoplankton is remineralized s h ortly after it is prod u ced beneath the sunlit waters. As a consequence of the sea s onal cycles i n daily nitrification rates from 0 1000 m are 2.52 mmol N m-2 d-1 on day 73 and 0.82 mmol N m-2 d-1 on day 180 (Figure 2.24) T ab l e 2.4a Integra t ed Concentratio n s Julian day 365 (mmol m-2) State Variable I Case 1 Case2 Case3 FG1 I 25.09 9.81 24.26 FG2 I 26 78 I 4.69 25.71 FG3 I 26.80 4 .27 I 21.11 FG4 I 32 .01 I 2 38 22.65 All I 110 68 I 21.15 I 93.73 Fecal P e ll ets I 105 .7 0 I 1.90 7.53 DOC1 I N/A I NIA I 373.36 CDOC1 I N/A I NIA I 14.88 DOC2 I NIA NIA I 49,576.58 CDOC2 I NIA NIA 300.40 N03 I 10, 1 54.66 I 11, 4 1 3.60 I 12, 1 42.57 NH4 I 104.61 23.47 4 6 DIC ( 1 03) I 2 1 89 87 2195.87 2 1 99 23 99

PAGE 113

Table 2.4b. Average Daily Carbon Fluxes (mmol C m-2 d-1) I Case 1 I Case2 I Case 3 Growth-FG1 I 7.14 I 2.56 I 4.55 Growth-FG2 I 8.88 I 0.63 I 6.09 Growth-FG3 I 8.44 I 0.32 I 4.83 Growth-FG4 I 12 .17 I 0.15 I 7.64 Growth-All I 36.63 I 3.66 I 23.11 UptakeBacteria I NIA I NIA I -13.29 DIC surface flux I -0.66 I 0.45 I 1.61 DIC bottom flux I 9.90 I 0.13 I 0.09 Fecal bottom flux I -10 .98 I -0 .78 I -0.43 DOC bottom flux I NIA I NIA I 0.00 Table 2.4c. Average Daily Nitrogen Fluxes (mmol N m-2 d-1) I Case 1 I Case2 I Case 3 FG Uptake N03 I -1.82 I -0 .13 I -1.13 FG Uptake I -4.34 I -0.49 I -2.12 N03 bottom flux I 1.37 I -0 .01 I -0 .01 bottom flux I -0.01 I 0 00 I 0.00 Fecal bottom flux I -1.79 I -0 .13 I -0.05 DON bottom flux I NIA I NIA I 0.00 Nitrification (N03) I NIA I NIA I 1.22 Nitrogen Fixation I NIA I NIA I 0.14 100

PAGE 114

4 3 '"C) ':'8 2 z 1 0 Integrated Daily Nitrification 100 200 julian day Figure 2.24. Case 3 daily integrated nitrifi c ation rates. 101 300

PAGE 115

2.3 Biological Discussion 2.3.1 Model Fidelity Which is the most accurate simulation?. It is not likely to be the second case of weak mixing and a simple biological model. Case 2 simulates a low productivity system, who se annual primary production is 16 g C m-2 y-1, an order of magnitude lower than 14C estimates ( Lohrenz et al. 1992; Malone et al., 1993; Michael s et al., 1994b ; Michaels and Knap 1996 ) Thi s is also evidenced by th e case 2 seaso nal cycle of productivity ( Figure 2.9), whose range is 15 to 103 mg C m2 d-1, also an order of magnitude less than the observed seasonal cycle ( Figure 2.25). Phytoplankton biomass is extremely low throughout the year in case 2, and i s dominated by Prochloroco cc u s-like FG 1 and FG2 Thi s is contrary to the observed chlorophyll band other accessory pigment d a ta, that suggest this species is unimportant during the fall-spring seasons (Figure 2.26) The s ubsurface chlorophyll a maximum (SCM) is down around 150m, much deeper than observed subsurface maxima ( Figur e 2.26, see also Siegel et al. 1995; Si ege l and Michaels, 1996). While i t appears obvious th a t thi s case is not representativ e of the ecosystem at the BATS site, th e results are displayed as transition between th e unrealistic stro n g mixin g of case 1 and more reali stic weak mixin g of case 3. The determination of accuracy between the s imple ecosystem of case 1 and the complex ecosystem of case 3 i s not so clear. Both cases have annual productivities i n th e range of 14C observations (160 and 101 g C m-2 y 1 for c ase 1 and case 3, respectively). Both cases simulate observations of spring influx es of nitrate ( Fi g ure s 2.7b, 2.20b, and 2.27), and the resultant blooms of primary prod uct io n ( Figures 2.2, 2.16a, and 2.25), and chlorop hyll a stocks (Figures 2.6e, 2.19e, and 2.26). Each case also simulates the observed seaso nal s ucce ss ion of the phytoplankton asse mbla ge as seen in the accessory pigment data ( Fi g ure s 2. 7, 2.20 and 2.25). 102

PAGE 116

1200 200 _.., 180 1000 "'""'" Ul 8 160 0 -a u 800 140 n bL) "'1 8 "-" 120 g c 600 0 -100 ..... <.) = 400 80 "0 0 1-o 60 c 200 40 8 -0 1-o 20 0 Figure 2 25. Time series of integrated primary production (circles) and 150m trap fluxes (squares) at the BATS site (redrawn from Michaels and Knap 1996) 103 ::s ::!2 c: :>< _.., a OQ n -a N 0. "-"

PAGE 117

A 0 50 ,--.._ E 100 '-' ..c ...... 0.. 150 0 200 250 M s 1990 B 0 Chlorophyll b (ng/kg) 50 ,--.._ 100 E '-' ..c ...... 150 0.. 0 200 250 s c 0 Zeaxanthin (ng/kg) 50 ,--.._ 100 E '-' ..c ...... 0.. 150 0 200 250 s Figure 2.26 Time series of phytoplankton pigments at the BATS site from December 1989 to June 1990 : (a) chlorophyll a; (b) chlorophyll b; (c) zeaxanthin; (d) 19'butanoyloxyfucoxanthin; (e) fucoxanthin; (f) 19 hexanoyloxyfucoxanthin; (g) chlorophyll c1+2 ; (h) chlorophyll c3 (redrawn from Michaels et al. 1994) (continued on next page) 104

PAGE 118

D 0 19 Butanoyloxyfucoxanthin (ng/kg) 50 ......... E 100 '-' ..c 0.. 150 d) 0 200 250 s E 0 Fucoxanthin (ng/kg) 50 ......... E 100 '-' ..c _. 150 0.. d) 0 200 250 s F 0 19-Hexanoyloxufucoxanthin (ng/kg) 50 ......... E 100 '-' ..c 0.. d) 150 0 200 250 s Figure 2.26. (continued; continued on next page) 105

PAGE 119

G ..c 0. 150 11) 0 200 s H 0 Chlorophyll c3 (ng/kg) 50 ,-._ 5 100 '-' ..c 0. 150 11) 0 200 250 s Figure 2 26 (continued) 106

PAGE 120

,...-._ so a .._, 100 ..c ..... 0.. 0 150 200 Nitrate (!lmoVkg) Figure 2 27. Time series of nitrate stocks at the BATS site ; dashed contours are 0.1 f.!mollkg (redrawn from Michaels and Knap 1996). 107

PAGE 121

The summer SCM is shallower in case 1 than the observations -75 m versus -100 m. Yet, given that maximum summertime irradiance is only -1130 Jlmol quanta m-2 s-1 in response to cloud effects (see the Bio-Optical Methods), this is not unreasonable. The inclusion of variable carbon to chlorophyll a ratios in both cases results in simulated surface chlorophyll a time series that approximate the CZCS data (Figure 2.28). The peaks in carbon biomass and pigment distributions are not as defined in case 1, compared to case 3; but again, they are not unreasonable given the vertical and temporal sampling scheme at BATS. Were it not for the lack of a physical basis for a vertical eddy diffusivity of 8.00 cm2 s-1 in the stratified thermocline, the solutions of case 1 would appear reasonable particularly, if the only validation data for this simulation are the productivity and chlorophyll a observations (Figures 2 25 and 2 26). Such was the conclusion of an earlier study of this ecosystem (Bissett et al 1994) Yet, upon a more complete examination of the system the simulated results of case 1 begin to appear untenable The settling fluxes of particulate carbon and nitrogen in case 1 are five times larger than measured fluxes at 150 m: a simulated flux of 48 g C m-2 y-1 versus a measured flux of 9.2 g C m-2 yr-1, and a simulated flux of 9.1 g N m-2 y-1 compared to a loss of 1.5 g N m-2 y-1 (Michaels et al., 1994b). The particulate fluxes of case 1 are 20 times larger than those measured fluxes a depth of 1000 m within other oligotrophic waters (Martinet al., 1987). Closer examination of the simulated physical environment reveals other subtle problems with case 1. While the springtime peak of primary productivity (800 mg C m-2 d-1) is within the range of measurements, it is at the high end of the inter-annual variance (Michaels and Knap, 1996). Such large daily primary productivities are usually, but not exclusively, associated with periods of deep convective mixing (Lohrenz et al., 1992; Michaels et al., 1994b; Michaels and Knap, 1996). The maximum mixed layer depth of 150m in this simulation is at the shallow end of the observation (Michaels and Knap 108

PAGE 122

E .......... ro :a u .......... OJj E Simulated Chlorophyll a versus CZCS Chlorophyll a 0.25 0.20 / ,, I \I I 0.15 0.10 0.05 0 100 Case 1 Case 3 czcs T . { 7\ \ I J ..._ \. f \ II I \ 200 julian day 300 Figure 2.28. Six year average of CZCS estimated chlorophyll a concentrations at the BATS site (data from Bissett et al., 1994) compared against Case 1 and 3 surface chlorophyll a values. 109

PAGE 123

1996). The total annual production of 160 g C m-2 y-1 is also on the high side, which is most often associated with years of maximum thermocline ventilation (depths> 200m). Another problem with case 1 of the simulation is its fidelity in mimicking the observed vertical gradients of DIC and nitrate (Figure 2 27), not to mention the large accumulation of ammonium. The large Kz removes these gradients below the euphotic zone Case 1 is also a net heterotrophic ecosystem, releasing 0.2 mol C02 m-2 y-1 to the atmosphere, in contrast to the estimated net autotrophic system from pC02 measurements (Bates et al., 1996). If the goal of a simulation effort is to predict carbon fluxes in the face of rising anthropogenic releases, this discrepancy is cause for great concern. What once appeared to be a reasonable representation of an oligotrophic ecological system is not nearly as convincing when additional parameters and validation data are included 2.3 2 The Complex Biological Model Sparse measurements of productivity and pigment distribution do not completely define an ecosystem. Case 3 simulates the observed particle fluxes, vertical gradients, and gas exchange, more accurately than case 1 by the inclusion of additional state variables and processes (bacteria differential carbon and nitrogen cycling, DIC, DOM, nitrification, nitrogen fixation and fecal pellet remineralization) in a more realistic physical habitat of 0.37 cm2 s-1 mixing through the thermocline. dissolved organic matter The cycling of labile dissolved organic matter is very rapid in case 3, without accumulation of any labile DOC over the year. Such rapid bacterioplankton utilization of a small fraction of the total DOC pool is suggested by in situ measurements (Coffin et al. 1993; Carlson et al., 1994; Ducklow et al., 1995; Hansell et al., 1995; Carlson and Ducklow, 1996; Carlson et al. 1996). The depth-integrated labile DOC over 0-100 m 110

PAGE 124

varies from 6-9% of the total DOC pool. Over the upper 250 m, the total DOC stocks only vary between 14.9 and 15.2 mol C m-2, similar to observations of 14.9 to 15.9 mol C m-2 (Figu re 2.29). Export of labile DOC below a depth of 100 m occurs in the simulation during deep winter mixing (Figure 2.17b). Calculated by the method of Carlson et al. ( 1994), the depth-integral of labile DOC between 100-250 amounts to a 0.16 mol C m-2 y-1 export of DOC from the euphotic zone. There is also an export of relict DOC of 0.06 mol C m-2 y-1, for a total calculated DOC export of0.22 mol C m-2 y-1. Such a simulated export is much lower than the export calculated by Carlson et al. (1994) of 1.02 mol C m-2 y-1. The measured export is only temporary, since the exported DOC is utilized, and co nverted to DIC, by heterotrophic bacteria during the following summer and late-fall periods. The simulated DOC export is also converted to DIC during these periods. The difference between the simulated and the measured export of DOC may lie in the definition of lability of the DOC; i.e. all of the model's labile DOC is instantaneously labile. In the "real" world, the various components of the measured labile DOC pool may not be. There may instead be different turnover times for different fractions of the DOC pool, and bacterial utilization efficiencies. In addition, the half-saturation constant for labile DOC of the model was calculated from growth rates of bacteria, determined by 3 H -t hymidine incorporation. There are some indications that the 3H-thymidine incorporation method overestimates the net growth rate of bacteria (Carlson et al 1996); a rate that would result in a higher half -sa turation constant for DOC utilization This would yield a longer period of labile DOC accumulation, which, in turn, would allow for a g reater export beneath a depth of 100 m One particular noteworthy feature of thi s simulation i s the result that bacteria are never nitrogen limited but instead are always carbon limited. Such a paradigm has been hypothesized from culture experiments (Go ldman et al., 1987 ; Kirchman et al., 1989 ; Kirchman et al., 1990; Carlson and Ducklow, 1996), and budgets of bac t erial growth Ill

PAGE 125

16 0 1 15.8 0-250 m 15. 6 0 08 N 8 15.4 u 1 5 2 I 0 0 6 "0 N I -1 5 0 E 1 4.8 E 0 .04 u 0 1 4 6 0 .02 E 14.4 1 4 2 0 7 2 7 0-100 m 6 8 N 6.6 I E u 6.4 -6 .2 0 E 6 5 8 5.6 5.4 9 6 9.4 100-250m 9 2 N I E 9 u 8.8 0 E 8 6 8.4 8.2 8 ...... ...... C'l C'l C'l C'l 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 0'1 ...... ...... ...... ...... ...... ...... ...... ....... 0.. u d 0.. u Q) Q) ::s Q) Q) (/.) a ...... (/.) a d 0.. u ::s Q) Q) ........ (/.) a Fig u re 2 29. S e aso nal d isso l ve d orga n ic car b o n s t ocks at the B ATS site (redrawn from Carl so n et a l., 1994). 112

PAGE 126

(Banse, 1995). The excretion of labile DOC by phytoplankton and zooplankton is small, about 12 g C m-2 y-1, yielding a labile DOM pool with a C:N ratio of 5 5 approximating that of the living POC pool (Figure 2.23c). With a GGEc equal to 30% and a bacterial biomass C : N ratio of 5.0, the necessary C:N substrate ratio would need to be> 16.7 for nitrogen to be the limiting nutrient. Thus, the bacteria are net producers of NJ4, derived from the excess nitrogen within the DOM source This effect would not be resolved in a model involving solely carbon or nitrogen as the "currency" state variable. Hence the cycling of the nutrients and carbon would be inaccurately simulated. Almost 60 % of the yearly production in case 3 passes through the bacterial pool. The inability to accurately simulate the fate of DOC/DON poses great difficulties in building a predictive carbon cycle simulation nitrification Ammonium concentrations reach a maximum of 0.15 J.Lmol Nf4liter-1 after the spring bloom in case 3 (Figure 2.22c) Only 0 05 J.Lmol NJ4liter-1 were measured after the 1994 spring bloom (Figure 2.30), using fluorescence techniques that have a 2 nmol precision (Lipschultz et al., 1996) However the intermediate oxidation product of N02 is not a state variable of case 3. Nitrite concentrations, determined by chemiluminescent techniques with a 1 nmol precision, are> 0.20 J.Lmol N02liter-1 at 150m, forming the primary nitrite maximum (PNM) at the BATS and Hydrostation S sites (Lipschultz et al., 1996) The seasonal cycle of "ammonium" in case 3, instead matches the observed seasonal cycle of the sum ofNI4 + N02 The model's average nitrification rate of 1.2 mmol N02 d-1 replicates the 1.0 mmol N02 d-1 estimated from geochemical arguments over depths of 150 to 1000 m (Zafiriou et al., 1992). 113

PAGE 127

Nitrite (nM) 0 5 10 15 20 25 1000 Figure 2.30 Nitrite profiles from the BATS study site taken during the spring bloom of 1994 (redrawn from Lipschult z et al., 1996) 114

PAGE 128 particulate carbon and nitro gen regeneration The d i fferential rernineralization of the particulate carbon and nitrogen provides a mechanism to selectively move more carbon to greater depths than nitrogen Such a hypothesis resulted from sediment trap data, used in the formulation of case 3 (Martin et al., 1987), as well as from the results of other sediment traps studies in the North Atlantic and North Pacific (Knauer et al., 1979; Knauer and Martin, 1981; Karl et al., 1988; Knap et al., 1991; Taylor and Karl, 1991; Knap et al., 1992; Knap et al. 1993; Knap et al., 1994; Knap et al., 1995) While there are some questions about the relative percentages of sinking material caught by the traps (Buesseler 1991 ; Michaels et al 1994a ; Michaels and Knap 1996), there is no rea so n to assume that the traps se lectively catch nitrogen-poor particles at greater depths If enhanced rernineralization of nitrogen relative to carbon occurs, its consequence should be seen as changing M)IC/M)IN ratio beneath the euphotic zone as a function of depth during stratified conditions These products of remineralization would not be dependent on the efficiency of sediment traps Care must be taken in making the calculation of this ratio ; the depths of sampling must be removed from autotrophic influences at the base of the euphotic zone The growth of phytoplankton at the base of the euphotic zone in oligotrophic regions is dependent on the upward flux of nutrients, and their growth quickly drives nutrient stocks to nanomolar concentrations. Thus the M)IC/M)IN gradient should approximate the classic Redfield carbon and nitrogen incorporation ratio of 6 .6 across this base of the euphotic zone Beneath the euphotic zone of these s tratified waters away from the influence of phytoplankton growth, the dominant process effecting the gradients of carbon and nitrogen is instead rernineralization. If there is a selective return of nitrogen the M)IC/M)IN ratio should be smaller that than the Redfield ratio Table 2.5 lists the BATS cruises between Julian day 129 (May 9th) and 295 (October 17th) for the years 1989 through 1994, when concurrent DIC and DIN (N03 + N02 ) measurements were made over depths 150-250 m 115

PAGE 129

Table 2 5 M)JC ver s us miN beneath the Euphotic Zone BATS First Second miC/ Crui s e Depth Depth DICt DIC 2 DINt DIN 2 M)JN and Date 8-5/89 I 175.8 248 6 1 2 010.1 2078.2 I 3 .11 I 4 80 I 4 79 9-7/8 9 I 148.9 251.3 1 2069.4 2073.5 I 2.02 3 .55 I 2 68 10-11s9 1 148 5 1 97 9 1 2064 o 2072.0 I 2 06 2.63 I 14.04 20 5/9o 1 182.5 249.9 1 2os1.1 2080.3 I 3 .31 4.41 I -0 73 21 6190 160.1 251.0 1 2067.7 2066 5 I 2.18 2 35 I -7.06 32-5/91 I 161.9 251 1 1 2063 5 2074 8 I 1.67 3 66 I 5 68 33-6 /91 I 157 7 253.4 1 2o12.4 2073 5 I 1.55 2.95 I 0.79 34-7/91 I 170 3 203.0 1 2o73.3 207 3. 7 I 2 .35 2 .33 1 -2o oo I 203 0 250.7 1 2073 7 20 7 2 5 I 2.33 2.31 I 60 00 35s191 1 159.7 200 8 1 2o12 5 2075 6 1.81 2 59 I 3 97 44 5/92 1 160.8 198.7 2075 0 2075.0 1.19 1.29 I 0 00 I 198.7 253 8 2075 0 2075.0 1.29 1.31 I 0 00 45-6/92 161.2 201. 7 2078.0 2082 0 1.86 2 .56 5 .71 201 7 247 0 2082 0 2086.0 2.56 3.11 7 27 46 -7/92 1 199.7 I 246.8 2091.0 2086.0 3.22 3 05 29.41 47s192 1 159.7 I 200 7 2084 0 2085 0 1.16 1.85 1.45 I 200.7 I 250 1 2085 0 2091.0 1.85 2 70 7 06 48-9/ 92 1 159 8 I 199.6 2082 0 1 2os3 o 2.08 2.71 1.59 4910/92 1 159 8 I 200.2 2083 0 1 2os3 o 2.56 2.74 0 .00 I 200 2 I 249 7 2083.0 I 2086 0 2 74 3 .17 6 98 56 5/93 1 148 0 I 193.7 2075.0 I 20 7 5 0 1.91 2.07 0 00 193 7 I 245.7 2075 0 2077 0 2 07 2 .16 22 22 57-6/93 1 161.8 I 201.8 2076 0 2077 0 1.98 2 .10 8 .33 I 201 8 I 255.0 2077 0 1 2os2.o 2 .10 2.46 13. 89 58 -7/93 1 161.4 I 2 01.4 2091.0 1 2os9 o 2.45 3.16 -2.82 I 201.4 I 250.0 2089 0 1 2os9 o 3.16 3 26 0 00 59 s/93 1 160 0 I 200 5 2093 0 1 2o9 2. o 1.44 2 88 I -0 69 200 5 251.9 2092 0 1 2089 o 2 .88 3 .19 I -9. 68 (continued on next page ) 116

PAGE 130

Table 2 5 (continued) BATS First Second I DIC LIDIC/ Cruise Depth Depth I 1 DIC2 DIN 1 DIN2 LID IN and Date 60-9/93 I 159.0 I 201.1 1 2085.o 2083.0 I 2 03 I 2.25 I -9 09 I 201.1 I 252.2 I 2083.0 I 2082.0 I 2.25 I 2.59 -2.94 61 wt93 1 160 .2 I 202.3 I 2080.0 I 2080 0 I 2.48 I 2 .51 I 0 00 I 202.3 I 250.7 I 2080 0 I 2080.0 I 2 .51 2.54 I 0 00 68-5/94 I 160.4 I 200.4 I 2081.0 2086.0 I 1.66 I 2.59 I 5.38 I 200.4 I 251.0 1 2o86 o I 2089.0 I 2.59 I 3.34 I 4.00 69-6/94 I 139.5 201.2 1 2o16.o 2076 0 I 1.48 1.87 I 0 00 201.2 I 250.6 I 2076.0 I 2081.0 I 1.87 I 2.29 I 11.90 71 8/94 I 161.1 I 201.4 I 2078.0 I 2081.0 I 2.05 I 2.41 I 8.33 I 201.4 I 252 5 I 2081.0 I 2086.0 I 2.41 I 3.15 I 6 76 12-9/94 1 160.4 I 251 .9 I 2078.0 2081.0 I 1.21 I 2.20 I 3.03 7310/94 1 160.4 I 200.9 I 2088.0 I 2087.0 I 1.92 I 2.43 I -1.96 200.9 I 251.2 I 2087.0 2090.0 I 2.43 I 3.04 4.92 80 5/95 I 161.0 I 199.5 I 2075.0 2075.0 I 1.75 1.59 I 0 00 I 199.5 I 252.1 I 2075.0 I 2081.0 I 1.59 I 1.40 1 -31.58 Of the 43 observations 5 actually show DIN decreasing with depth, suggesting a source of DIN from the s urface 10 of the observations have LIDIC/LIDIN ratio s >6.6. The remaining 28 samples show LIDIC/LIDIN ratios <6.625, suggesting nitrogen is being returned to this depth interval at a faster rate than carbon from sinking particles. Since the precision of DIC measurements are approximately 1 DIC kg 1 (Bate s et al 1996) and the reported DIC values after September, 1991 did not include significant digits, proving a hypothesis based on the impacts of a more rapid nitrogen cycle strictly from the data may be difficult. Future increases in the precision of the DIC measurements should help to address this hypothesis 117

PAGE 131 nitrogenfixation The simulated yearly rate of nitrogen fixation of 53 mmol N m-2 y-1 is similar to those estimated from the tropical Atlantic during non-bloom conditions (9-47 mmol m-2 N y-1) by Goering et al. ( 1966), from non-bloom densities of 104 trichomes m-3 (39 mmol N m-2 y-1) by Carpenter ( 1983), and estimated in the North Pacific (49 mmol m-2 y-1) by Gundersen et al. (1976). However, the simulated N 2 fixation is on the low side of yearly estimates based on bloom conditions for 1 day (80 to 100 mmol N m-2 y-1) by Karl et al. ( 1992) in the North Pacific. The maximum daily rate of 0.33 mmol N m-2 d-1 in case 3 is also much lower than those estimated by Carpenter and Romans ( 1991) in the tropical North Atlantic (0.71 to 3.57 mmol N m-2 d-1). Growth of cyanophytes, resulting even from low levels of nitrogen-fixation, requires phosphorous at some elemental ratio. Since the BATS site does not appear to have measurable quantities of dissolved inorganic phosphate (DIP) in the upper euphotic zone during the summer, some source of phosphorous needs to be invoked to balance the stoichiometric growth ratios inferred in case 3 of the model. One method to supply the necessary phosphorous may be the Trichodesmium PTransport model (Karl et al., 1992). Trichodesmium also displays alkaline phosphatase activity (Yentsch et al., 1972) and has a high affinity for phosphomonoesters (McCarthy and Carpenter, 1979). A labile dissolved organic phosphorous (DOP) pool would then provide a potential source of phosphorous to the nitrogen-fixing organisms. This DOP may be supplied from upward fluxing of low density organic material (Karl et al., 1992) or be left in the euphotic zone by the previous spring bloom (Walsh, 1996) Trichodesmium also appear to have a flexible particulate nitrogen to particulate phosphorous ratio cell quota (-50 to 125:1, Karl et al., 1992; Letelier and Karl, 1996) at higher than typical Redfield ratios This flexibility would allow it to thrive in the face of what would typically be considered limiting phosphorous conditions. Lastly, phosphorous is considered to be regenerated during decomposition more rapidly than nitrogen (Redfield, 1958). This would allow greater 118

PAGE 132

fixation of nitrogen per unit phosphorous than if nitrogen and phosphorous were regenerated at the same rate (similar to the selective regeneration of nitrogen versus carbon above). 2.3 2.5 dissolved inorganic carbon The seasonal cycle of DIC in the Sargasso Sea is well represented by case 3 (Figure 2.22a) The modeled seasonal DIC surface water maximum lags by a couple of months the minimum surface temperature of 19.00 Con Julian day 72 (Figure 2 .1a ); and maximum surface DIC concentration of 2120 J.lmolliter-1 occurs on Julian day 135 (Figure 2 22a). This lag is similar to that mea s ured at Bermuda (Bates et al., 1996) The minimum DIC concentration al s o lags the maximum in temper a ture (26.93. Con Julian day 237, minimum surface DIC 2098 J.lmolliter-1 on Julian day 292). However there are some differences between the field data and model results that need to be discussed The simulation's minimum DIC of 2073 J.lmolliter-1 is -25 Jlmol greater than the lowest concentration found at the BATS site. There is also a difference of approximately 21J.1mol DIC liter-! between the simulated and measured values as deep as 50 meters during the late summer This depth-integrated difference of 1.03 mol DIC m2 is equal to the supposed carbon -c ycle imbalance at the BATS site (Michaels et al ., 1994a) Since the simulated ecosystem is balanced and the proftles of biomass DOC, productivity particulate flux inorganic nutrients, nitrogen-fixation nutrient regeneration and phytoplankton pigments appear to be well represented the problem would appear to lie in either the model's calcu lation of gas flux during low wind speed conditions or perhaps three-dimensional effects that are unresolved in this one-dimensional model ; i e, downwelling The resolution of this simulation's summer DIC discrepancy will require the completion of a high-resolution three dimensional ecosystem model. The simulated maximum DIC stocks are similar to the observations ; and over the annual period, Julian day 0 and 365 the 150m stocks are relatively unchanged (Figure 119

PAGE 133

2.31). Thus, the -25 C liter-I discrepancy in the simulated summer concentrations, compared to the measurements, results only in a reduced seasonal amplitude of C02 flux across the air-sea interface. This reduced amplitude should not affect the case 3 result of a net autotrophic system. 2.4 Summary A simple biological model, with unrealistic physical mixing (case 1), mimics the seasonal cycle of productivity and chlorophyll a stocks at the BATS site. However it does not mimic the cycle of DIC and NJ-4, nor does it reproduce the vertical gradients of DIC, NJ-4, and N03. A more complex biological model that includes heterotrophic bacteria, DOM, nitrification, particulate remineralization and nitrogen-fixation, and uses a realistic vertical mixing (case 3), does replicate the seasonal cycle of productivity, chlorophyll a, phytoplankton species succession, DIC, DOC, NH4, nitrification, and nitrogen-fixation In particular the addition of local nitrification, remineralization, and nitrogen-fixation removes the need for an unrealistically high upward vertical flux of nitrate to mimic the productivity and chlorophyll a stocks. Case 3 of the model suggests that bacteria are always limited by DOC, rather than by ammonium or DON. This result suggests that other simulations based solely on nitrogen dynamics will not accurately simulate the carbon cycle. In this respect, data are needed on the lability of DOM and a possible set of different DOM pools in the sea. Nitrification and differential particle remineralization of carbon and nitrogen as a function of depth will also affect the calculations of carbon removal from the surface waters of the ocean. A more rapid return of nutrients would yield a greater sequestration of carbon per unit nutrient. Again, more observations are required to test this hypothesis 120

PAGE 134

Dissolved Inorganic Carbon (DIC) 0 2100 2110 2120 2130 2140 Jlmol C liter"1 Figure 2.31. Dissolved inorganic carbon on Julian day 0 (solid line) and 365 (dashed line) from Case 3 of the simulation. 121

PAGE 135

3. Bio-Optical Model 3.1 Bio-Optical Methods 3.1.1 Bio-Optical Overview The ecosystem model, hereafter referred to EcoSim, utilizes the spectral distribution of light energy, in conjunction with temperature and nutrients, to effect autotrophic growth. The downwelling energy distribution just below the sea surface-atmosphere interface, Ed(A.,O), is derived from a previously developed atmospheric radiative transfer model designed for oceanographic applications (Gregg and Carder, 1 990). The attenuation of light energy (apparent optical property) with increasing water depth is a function of the magnitude and spectral distribution of light energy, as well as the spectral absorption and scattering (inherent optical properties) of the water. The inherent and apparent optical properties (lOPs and AOPs respectively) are determined at each time step of the simulation, when radiant energy strikes the sea surface (i.e. they are not calculated at night). There are three broad classes of materials that effect the attenuatio n of light-water, living organic material, and nonliving o rgani c material; i.e., dissolved organic material and particulate detritus. The t emporal changes of the living and nonliving organic materials are described by the equations in the Biological Section. All the values for lOPs and AOPs are calculated at the middle of each vertical grid box (Llz x 0.5), unless otherwise noted. 122

PAGE 136

3.1.2 Light Energy at the Sea Surface The irradiance model of Gregg and Carder ( 1990) hereafter referred to as IRRSim, provides a spectral distribution of downwelling direct and diffuse irradiance, Ed_dir(A.,O-) and Ed_dit{A.,O), respectively, just beneath the sea surface for a cloudless maritime atmosphere. The computer code was provided by Dr. K. L. Carder. The output of this model are the irradiance values over wavelengths between 400 700 nm, at a 1.0 nm resolution The inputs of IRRSim are latitude, longitude, atmospheric pres s ure air mass type (1 10), relative humidity precipitable water, wind speed, visibility, and ozone I chose coordinates of 33 degree N, 60 degree W where the latitude is the same as the COADS wind data. Atmospheric pressure is set at the default 29 .92 inches of mercury The air mass type is set for at 1 for a maritime atmosphere (10 is a continental atmosphere ). Precipitable water is set at the default 1.5 em. Wind speeds come from the monthly means of the COADS data (Wright, 1988) and interpolated to daily values as described in the Biological Section Visibility is assumed to be 25 km. Total ozone is calculated by the model from climatological values. The sky over the Sargasso Sea is not cloud free at all times. The monthly climatological values of shortwave solar radiation (Oberhuber 1988) do integrate the effect of cloud s on downwelling irradiance These climatological data, in W m-2, are multiplied then by 0 5 (Morel and Smith, 1974) to convert them to monthly photo sy nthetically active radiation (PAR), with interpolation to daily values These PAR estimates are next divided by the daily energy IRRSim output values, summed for every second of the day at a 1.0 nm resolution in W m-2 to give Ed(O+), in order to calculate the daily fractional energy coefficient, FRACLi : FRACLi = Daily Climatological Energy IRRSirn Ed ( o+) 123 3 1

PAGE 137

FRACLi is then used in a second iteration of IRRSim, which returns direct and diffuse irradiance values, W m-2, multiplied by FRACLi. This procedure reduces the spectral downwelling energy, and assumes that the clouds are spectrally neutral. Spectral downwelling energy is then finally converted to Jlmol quanta m-2 s-1. The spectral resolution of the light forcing is reduced for EcoSim to sampling the IRRSim output at 5 0 nm and 15 minute resolution The Julian date, time of day, in-water solar zenith angle (solar zenith angle after refraction calculated by Snell's Law and a refractive index for sea water of 1.34), the sum of direct and diffuse irradiance Ed( A 0+), and the percentage of diffuse irradiance, %Ed_dit\A, 0+) at each wavelength are used as input file s for EcoSim. 3 1 3 Inherent Optical Properties 3 1.3.1 absorption-water The spectral absorption of light is defined by three classes of material water, living phytoplankton, and other material, which is broadly classified as colored detrital material (CDM). In this model CDOC and CDM represent the same class of material and the terms are interchangable. The total light absorption by the complete water medium
PAGE 138

3 1.3.2 absorption-particulate The absorption of light from living biomass aph_1(A) is a function of the four functional groups of phytoplankton : 4 a ph_r (A) = L aph_i (A) i=l where 3 3 aph_u(A) =packaging effect I, (pigments [mg m-3 ] a:ig [m2 mg-1]) 3.4 Packaging effect in Equation (3.4 ) is the reduction in weight-specific pigment absorption because of self-shading (Kirk, 1994) pigments refer to the concentration of each pigment within functional group i at that depth a pig is the weight-specific absorption of the particular pigment at that wavelength Only FG4 has a reduction in aph_ 4(A) as a result of packaging effect (described in Biological Section) Equation (3.4) differs from Equation (2. 7) in that all of the pigments are utilized in the calculation of particulate absorption, whereas only the photosynthetically-active pigments are utilized in Equation (2 7). A s noted in the Biological Section the C : chlorophyll a ratio is not fixed. Instead it varie s as a function of light or nutrient limitation. Since this ratio 8, is allowed to vary between some maximum and minimum number the determination of the ratio at e ac h time s tep depends on an optimal C:chlorophyll a ratio, S opt_r_i and a rate of change from the previous C :c hlorophyll a ratio. In the case of light-limited growth, the optimal carbon to chlorophyll a ratio Sopt_ll_i. versus irradiance is described as a linear function of irradiance (Geider et al., 1986b; Geider, 1987) : eop!_ll_i = eO_ll_i + slopee_ll_i. Eo(Z) 3.5 125

PAGE 139

The intercepts, So_i, slopes, slopee_ll_i> and ranges of C:chlorophyll a for each functional group are given in Table 3.1. Table 3.1. Parameters for Light-Limited Carbon to Chlorophyll a Ratios FG C :Chla So slope reference 30-100 1 (from 0 to 100 J..Lmol 30 .70 1, 2 quanta m-2 s-1) :j: 20-150 2 (from 0 to 450 J..Lmol 20 .29 1 2 quanta m-2 s-1) :j: 30-160 3 (from 0 to 1330 J..Lmol 30 .10 2, 3,4 quanta m 2 s-1) :j: 25-60 4 (from 0 to 300 J..Lmol 25 12 5, 6, 7 quanta m-2 s-1) :j: :j: -maximum of range is maximum value of calculation regardless of irradiance 1. -(Li et al 1992) 2. -(Moore et al., 1995) 3. -(Kana and Glibert, 1987) 4. (Kana et al., 1992) 5.-(Laws and Bannister 1980) 6. -(Geider et al. 1986) 7. (Geider and Osborne, 1987) A similar equation is used for the nutrient-limited 8, Sopt_nl_i> since this relationship is also approximately linear (Sakshaug and Andresen, 1989) In this case, the independent variable is the carbon to nitrogen ratio, rather than irradiance. 3 6 The slopes and intercepts of each functional group for the nutrient-limited C : chlorophyll a are shown in Table 3.2. The maximum of the lightand nutrient-limited, Sopt_ll_ i and 126

PAGE 140

8opt_nl_j, is the C:chlorophyll a ratio to which the functional group strives to achieve, 8 opt_r_i. I assume that phytoplankton will devote the minimum amo u nt of resources necessary to achieve their realizable growth rate. Table 3.2. Parameters for Nutrient-Limited Carbon to Chlorophyll a Ratios FG C:chl a 1 no change 2 no change no change 3 (arrested growth) 60-150 4 (from C:N 6.625 -14) 1.-(Moore et al. 1995) 2 -(Kana et al., 1992) 3.-(Glibert and Ray, 1990) 4.-(Laws and Bannister, 1980) eo slope reference N/A N/A 1 N/A N/A 1 N/A N/A 2, 3 60 12.2 4 The change in the actual e towards the optimal 8 occurs at the growth rate, Jlr, calculated for the current time step (Falkowski and Wirick, 1981 ): 3.7 where R is the chlorophyll a to carbon ratio, l/8i; Rw is the optimal Chlorophyll a : C ratio, 118opt r i; and, is the realized growth rate (see Biological Section 2.0). The changes in accessory pigments are defined by their relationship to chlorophyll a (Table 3.3). They occur instantaneous l y with tho se of chlorophyll a. The accessory pigments are not conserved during changes in their concentration as a result of chlorophyll 127

PAGE 141

a optimization Accessory pigment concentration is, however, conserved during physical mixing of the functional groups (Biological Section) Table 3.3. Accessory Pigments as a Function of Chlorophyll a FG chi b/chl a chi c/chl a 1 2.40.8 N .A. s lope = -0.023 2 0 .2-0 06 N.A. s lope= -0.00 1 3 N.A. N.A. .34 .17 4 N.A. slope= 0.005 1. (Moore et al., 1995) 2.-(Kana and Glibert, 1987) 3. -(Jeffrey, 1976) 4.(Perry et al., 1981) 5. (Haxo, 1985) 6 -(Cleveland and Perry, 1987 ) 7. (Schofield et al., 1990) PSC/chl a N.A. N A N.A. 2 0 0 5 slope= 0.012 over C/CHL range 25 150 8. (Hoepffner and Sathyendranath 1992) 9.-( Barl ow and Alberte, 1985 ) PPC/ch l a 0.3 1.4 s lope = 0.016 0.3 -2 5 s l ope = 0.017 0.3 1.5 s lope = 0.009 WH8013 0.10 constant PE/chl a refer-ence N.A. I N.A. l 203 slope= 0.13 1 2, 9 WH7803 3, 4, 5, N.A. 6 7 8 Since J.lr_i has to be greater than zero t o effect a change in the chlorophyll a:C ratio, pigment changes only occur when there is sufficient light or nutrients to grow. Also note in Equ ation (3.5) that th e change in chlorophyll a is only a function of total light intensity, without any spectral dependence. Therefore, changes in accessory pigments are also a function of total photon flux, without any spectral relationships as s uggested by culture experiments ( Jeffrey and Vesk, 1977; Glover et al., 1987 ; Hooks et al., 1988; Olson et al., 1988; Moore et al., 1995). In this model, the major source of chromatic adaptation is a 1 28

PAGE 142

change of group dominance within the phytoplankton assemblage, rather than by enhanced production of one accessory pigment versus another within an individual functional group (Bidigare et al 1990a). The specific absorption values, a* pig(A), of the unpackaged pigments are taken from Bidigare et al. (1990). These curves (Figure 3.1) are assumed to represent the in vivo pigment specific absorptions of the individual functional groups without regard to solvation effects (Bi s sett et al. 1996). Their curves are first interpolated to a 1 nm resolution and then subsampled at 5 nm intervals. The maximal and minimal chlorophyll a-specific absorption curves for each functional group are shown in Figure 3 2. absorption-CDM (CDOC) The absorption of light by colored detrital matter includes the absorption of dissolved organic matter and particulate detritus (Roesler et al., 1989; Carder et al., 1991) The particulate detritus, in tum, includes the non-extractable cellular material dead organic matter larger than 0.2 J.Lm in size and heterotrophic bacteria. Terrigenous solids, such as clay minerals are ignored in this model of oceanic waters. Total particulate organic carbon (POC) represents a small fraction of the total organic carbon in the water, approximately 2 3 % (Carlson et al., 1994 ; Knap et al., 1995). Hence, the spectral signal of the CDM is assumed to be controlled by the quantity and type of color dissolved organic carbon (CDOCi). The absorption of light by CDM is therefore: 3 acoM (A.) = L acoo c i 3.8 i = l where a cooc i (A) = acoo c i ( 41 0) exp [ S ( 410 A ) ] 3 9 129

PAGE 143

Pigment-Specific Absorption "'\ I \ I I '/'\ I : I ,_ I ' I .';(\ .. ;, I /: \ \ /. : ) \ \ 1, \ \ 450 500 550 wavelength (nm) chl a chl b chl c PSC PPC HPUB 600 650 700 Figure 3 .1. Pigment-specific absorption curves used in the model (data taken from Bidigare et al. 1990) 130

PAGE 144

A B 0.15 .,. 0.10 0.0 8 N 8 i ro 0 .05 0 .00 400 0 .15 'o.o 0 .10 8 0 .05 / / Chlorophyll a-Specific AbsorptionFG 1 / I \ highest carbon:chl a I \ lowest carbon : chl a I I I \/ \ I I I ..... I \ I / ------450 500 550 600 650 wavelength (run) Chlorophyll a-Specific Absorption-FG 2 / "\ / \ \ highest carbon:chl a lowest carbon:chl a 700 0 .00 400 450 500 550 600 650 700 wavelength (run) Figure 3.2. Chlorophyll a specific absorption curves (a) (d) for functional group 1 4, respectively Acce s sory pigment and pigment packaging are a function of chlorophyll a concentration, which in turn is a function of the available light and nutrients (continued on next page ) 131

PAGE 145

c -; b.O E N E i D 'o.o E N E "8. 0.15 0 .10 0.05 I / Chlorophyll a-Specific Absorption -FG 3 .... / / I highest carbon :c hl a lowest carbon:chl a 0 00 400 0.15 0 .10 0.05 / / / / / 0.00 400 450 500 550 wavelength (nm) 600 650 Chlorophyll a-Specific Absorption -FG 4 ---450 500 550 wavelength (nm) highest carbon:chl a lowest carbon:chl a 600 650 700 700 Figure 3.2. (continued) The availability of light and nutrients is reflected in the carbon to chlorophyll a ratio and these curves represent the chlorophyll a-specific absorption at the highest and lowest carbon to chlorophyll a ratios for each functional group. 132

PAGE 146

describes of the spectral absorption of CDOC (e.g. Carder et al., 1989; Roesler et al., 1989; Carder et al., 1991; Kirk, 1994). The exponential decay of CDOC absorption with increasing wavelength is denoted by the variableS. The absorption of light at 410 nm, acooc_j(410), is defined as: acooc ;(410) = ;(410) CDOC; --3.10 where a*cooc_i is the weight-specific absorption of CDOCi, m2 g-1 and CDOCi is the concentration of CDOCi, g m-3. CDOCi is a varying fraction of DOCi The exact proportion is a function of the production/destruction of the light absorbing chemical bonds by microbial and photochemical processes (Hayase and Tsubota, 1983; Hayase and Tsubota, 1985; Carder et al., 1989; Hochman et al., 1995; Lindell et al. 1995 ; Miller and Zepp, 1995; Wetzel et al., 1995; Bushaw et al. 1996) The fraction of DOC that is CDOC and its weight-specific absorption coefficients have yet to be determined at the same time. Hence, estimates are derived from many disparate reports on DOC concentration, CDOC absorption and fluorescence, and photolysis of CDOC. CDOC fluorescence has been linearly related to absorption across a wide range of salinities (Hoge et al 1993; Vodacek et al., 1995). This suggests that the molecular bonds absorbing light may be similar within terrestrial and marine DOC. The specific absorptions of highly concentrated humic and fulvic acids from commercial, fresh water, and soil sources, presumably unexposed to sunlight, are 2.63 and 1.40 m2 g-1 DOC, respectively, at 410 nm (Zepp and Schlotzhauer, 1981 ) All humus (assumed to be humic+fulvic acids), however, is not colored The actual amount of carbon of the humic and fulvic acids involved in the absorption of light may be only 30%, as humic acid concentrations are reduced by this percentage when they are exposed to intense UV-B radiation (Wetzel et al. 1995). Dividing the above humic and 133

PAGE 147

fulvic specific absorptions coefficients by 30 % suggests that the specific absorptions of the colored fraction of each are 8.77 and 4 67 m2 g-1, respectively The humus fraction of total DOC found in oligotrophic region s ( -40-45 J..Lmol DOC) is 31% (Harvey et al 1983) ; and of this humus, 10% is humic acid and 90% i s fulvic acid. The colored fraction of unexposed refractory oceanic DOC is thus assumed to be 9.3% (30% colored humus x 31 % of total DOC). The weighted average specific absorption of this CDOC is then 5 08 m2 CDOC (10 % humic* 8 77 + 90% fulvic x 4 67 ). Accordingl y, the unexposed (non-photolyzed) total relict DOC specific absorption would be 0.4724 m2 g-1 DOC (5.08 x 9.3 % ) This v a lue is somewhat less than the value of 0.6132 m2 g-1 DOC found of freshwater systems (Morris et al 1995), but these have higher concentrations of humus. The absorption of light by DOC at 3000 min the Gulf of Mexico is 0.0253 m-1 at 410 nm (Green and Blough, 1994). The concentration of DOC is approximately 45 J..Lmol in oceanic waters of 1000 m depth within both the Gulf of Mexico and the North Atlantic (Guo et al., 1995). Therefore, the total DOC spec i fic absorption of the DOC a t 3000 m in the Gulf of Mexico equal to 0.0469 m2 g 1 (0 0253 m-1/ 0 54 g DOC m-3). This value i s 1% of the above unexposed DOC specific absorption, and it suggests that the refractory DOC of deep water is 89% bleached, relative to unexposed DOC (1.0 % /9.3% ) The CDOC2 fraction is, thus, initialized at 1% of the deep water DOC2 The labile CDOC 1 fraction of DOC 1 is then established by starting with a measured downwelling attenuat i on coefficient, of0. 060 m-1 at 75 meters in the Sargasso Sea during July 1992 (Siegel et al., 1995). This attenuation coefficient can be converted to total absorption plus backscattering, at(410) + bbt( 410) by multiplication with an average cosine value of the downwelling photons (Sathyendranath and Platt 1988 ) : 3 .11 134

PAGE 148

The value of the assumed to be 0.8 (Kirk, 1994 ), since the optical depth, (Ka{ 410) x z), is approximately 3 (Siegel et al., 1995) and the briar ratio is approximately 2.3 (calculated from data and total scattering equations below). The sum of absorption and backscattering at 410 nm is then equal to 0 .0 48 m-1. Next, the water fraction of this sum, aw(410) + bbw(410), of0.0196 m-1 (Smith and Baker, 1981 ), is subtracted to yield an attenuation for the remaining constituents of 0.0284 m-1. Chlorophyll a concentrations were -0.15 mg m-3 in July (Siegel et al., 1995); multiplying this stock by a specific absorption coefficient for chlorophyll a of 0.04 m2 mg-1 at 410 nm, which includes the absorption of accessory pigments (Morel, 1988 ; Siegel et al., 1995), yields a parti c ulate absorption aph_r. of 0.006 m-1. I assume that particulate backscattering is included in the chlorophyll a specific absorption. Subtracting this value from 0 0284 m-1 yields an attenuation of 0.0224 m-1, which must result from CDOC. Note this CDOC attenuation can not be used to calculated the total DOC concentration, since surface DOC stocks have been exposed to sunlight and contain labile DOC and CDOC, which will have different specific absorptions. Backscattering is assumed to be zero for CDOC (Gordon et al., 1988), hence the attenuation is strictly the result of absorption of light by CDOC. Fluorescence data from the Sargasso Sea (Mopper et al. 1991) suggests that the absorption of deep-water, relic CDOC decreases by approximately 112 near the euphotic zone (assuming fluorescence is linearly related to absorption). Therefore, the absorption resulting from relic DOC in the euphotic zone can be removed by using the deep water specific absorption value of 0.0253 m-1 multiplied by 112 to yield 0 0127 m-I. The absorption of light by labile DOC1 must then be 0.0224 m-I minus 0.0127 m-1, or 0.0097 m-I. Concurrent measurements of DOC equaled -65 July 1992 (Carlson et al., 1994) The labile fraction of this DOC would be 20 Jlmol (65 total DOC45 relic DOC2). Dividing the absorption of 0 0097 m-I by the quantity of labile DOC (0.12 g m-3) yields a specific absorption for total labile DOC 1 equal to 0.0808 m2 g-I. 135

PAGE 149

The same process was used for April 1992 at 75 m. Using the following values for Kci( 410) Jld, chlorophyll a, and labile DOC 1 of 0 07 m-1, 0 75, 0 .20 mg m-3, 13 )lmol respectively, yields a labile DOC1 specific absorption of 0 0782 m2 g-1. The average of the April and July, 1992 values is 0 0795 m2 g-1. There is really no way to determine the actual fraction of labile DOC that is colored from the available data. Obviously, the entire quantity is not colored, since bacterial production still occurs in the surface waters during the summer when there is no absorption resulting from DOC (Knap et al., 1995 ; Siegel et al 1995; Siegel and Michaels, 1996) There are two reasonable assumptions one could take to guess at the fraction The first is that color is conserved and bacterial action does not affect the molecular bonds that absorb light. This would yield an estimated colored fraction of 1.6 % (0 0795 m2 g 1 I 5.08 m2 g-1). The other is to assume that the color fraction is constant between the two forms at 9.3% The actual value probably depends on many factors, e g type of labile DOC, relative degradation, bacterial modification, etc. but may lie somewhere between 1.6% and 9 .3%. Therefore, the colored fraction of unexposed labile DOC 1 is set to 5.5% This yields a CDOC1 specific absorption of 1.4587 m 2 g-1 (0.0795 I 0.055) The spectral slope for CDOC2 S2, is taken from measurements of deep Sargasso Sea water to be 0 025 nm-1 (Green and Blough 1994). The spectral slope of CDOC1, S 1 is set to 0 .014 nm-1. This is a much lower value than CDOC2 as it includes phytodetritus that typically has a very shallow spectral slope (Kishino et al. 1986 ; lturriaga and Siegel, 1988; Morrow et al., 1989). 3 1.3.4 backscattering Like at(A), the total backscattering component of attenuation bbt(A), can be separated into its component parts (Morel 1988 ; Sathyendranath and Platt, 1988): bbt(A) = bbw. bw(A) + bbp(A) bp(A) 136 3 12

PAGE 150

where bw and bp are the total scattering coefficient s for water and particulates, respectively The relative proportion of photons that are backscattered by water and particulates are bow and bop, respectively. The spectral scattering for water is taken from Smith and Baker (1981), and the backscattering proportion is set to a spectrally invariant 1/2 bw (Smith and Baker, 1981; Morel, 1988 ; Sathyendranath and Platt 1988). The total particulate scattering bp(A), is set to be a function of the total chlorophyll a concentration (Morel 1991): b,{A.) = 0.30 [Chi a]062 ( 3.13 This power function increa ses particulate scattering as chlorophyll a concentration increases, however the chlorophyll a-specific scattering decreases, since the exponent is less than one. Particulate backscattering is also spectrally dependent and is estimated by (Morel 1988 ) : 6 0 62 [ ( I I ) ( 550 ll b p bp(A) = 0.30[Chla] 0.002 + 0.02 24log[Chla] --:;: 3 .14 A s noted a bove backscattering i s assumed to be zero for CDOC (Gordon et al., 1988 ). Thi s raise s a s mall inconsistency, since the absorption of the colored detrital material (CDM) i s accounted for in the CDOC term of the total absorption equati o n but the detrital backsc a ttering i s accounted for in the total particulate backscattering Unfortunately there i s a lack of available data on the abs orption scattering and composition, of naturally occurrin g colored particulate and dissolved detrital material on a siz e and weight specific basis. This description of the underwater lOP i s a semi-empirical attempt to account for all of the spectral constituents in Cas e I-ill (Jerlov, 1976) ocean waters 1 3 7

PAGE 151

3 .1.4 Apparent Optical Properties The downwelling attenuation coefficient, Kar (where the wavelength dependence is dropped for convenience), depends upon the lOPs and the geometric structure of the downwelling light field (cf. equation 3 .11 and 3.16) This geometric structure is functionalized by the average cosine value of the downwelling photons, Jld Just beneath the surface, the average cosine for downwelling irradiance is computed as (eq 14 Morel 1991): 3.15 where J.ldirect is the cosine of the in-water solar zenith angle and J.ldiffus e is the average cosine of the diffuse component of skylight-. While J.ldiffus e is solar altitude dependent it is approximated by a constant value equal to 0.86 (Morel, 1991) Using Jlo as the value for the average cosine, J.ld, and rearranging Equation ( 3.11) : 3 16 where at and bbt are the lOPs just beneath the surface. This equation gives the value for the downwelling attenuation coefficient just beneath the surface. The average cosine is depth-dependent because the deeper a photon travels into the water the more likely it will be scattered. In addition, the more oblique the angle of penetration, the more likely the photon will be absorbed within a given depth interval, as the pathlength for the photon has increased (Kirk, 1994) Eventually, the angular distribution of photons reaches an asymptotic value J.ldinfinity' at the base of the euphotic zone, which is dependent on the interaction between absorption and scattering 138

PAGE 152

This value of lldinfinity is described by a linear function of Jldinfinity versus the spectral single-scattering albedo scattering, COQ, (Fig. 9.4 Mobley, 1994): 3.17 where OlQ is equal to the scattering coefficient divided by the beam attenuation coefficient, b1 I (a1 + b1). The average cosine at (zJ..ld(z-62), refers to the average cosine at the middle of the grid box above the current grid box (at the surface this value is equal to Jlo from equation 3.15). The constant, 0 5, is the Jldinfinity when roo equals 1.0. The value for the average cosine at the current grid depth then becomes a linear function of optical depth, where the optical depth of Jldinfinity is assumed to be 7.0, or at the 0 1% isolume. In order to derive the slope between J..ld(z-62) and lldinfinity an estimate of the optical depth at the vertical grid point is needed. This is calculated with Equation (3 .16) and using a1 and bbt from the present grid point and J..ld(z62) The slope between !ld(z-62) and lldinfinity is then calculated by: Jl,. lld(z-6.z) slope_Jld = 7.0 est. Kd z (note, the minimum value allowed for the denominator in this equation is 1.0.) The average cosine for the present grid point, Jld( z) becomes: lld = lld ( z -6.z> + slope_Jld (est. Kd ) With this value of Jld(z) Equation (3.16) is used to set the Kd at the current grid depth. The spectral downwelling light Ed, at z is then equal to: 139 3.18 3.19

PAGE 153

3.20 This is operationally defined as: 3.21 where grid is the current depth interval. The light energy at each depth is calculated at the center of the interval, hence the 1/2 l1z term The downwelling light energy is calculated for each grid, until total Ect (the sum across 'A, 400 700 nm) at depth z is le ss than 0.1 % of the surface irr a diance This grid point defines the bottom of the euphotic zone at each time step during the day Phytoplankton growth i s dependent on the total light available, not just the downwelling light field Since EofEd is equivalent to (Morel 1991), Ect is converted to Eo by multiplying Ect by 3.1.5 CDOC (CDM) Photolysis The color of DOC is reduced upon exposure to natural sunlight, mostly as a result of UVB radiation (Kieber et al 1989 ; Hochman et al. 1995 ; Lindell et al., 1995 ; Miller and Zepp, 1995 ; Morris et al., 1995 ; Wetzel et al 1995 ; Bushaw et al ., 1996 ; Siegel and Michaels, 1996) This reduction in color is matched by a release of DIC (Miller and Zepp 1995; Wetzel et al., 1995) and labile DOC (Kieber et al., 1989; Mopper and Zhou, 1990; Mopper et al., 1991; Miller and Zepp 1995 ; Wetzel et al ., 1995). The maximum absorption-specific production rate of DIC is set to equal the value of acidified Gulf of Mexico water at 0 21 ).!mol m /-1 hr-1 (Miller and Zepp 1995) This rate is calculated from an absorption at 350 nm [ acooc(350)]. 140

PAGE 154

acooc is estimated at 410 run, using a spectral slope of 0.018 n.rn-1 (typical slope for extracted CDOC, Green and Blough, 1994), and equals 0.62 m l-1 hr-1. The maximum rate of labile DOC production from CDOC is set to the deep-water Sargasso Sea production rate of 0 0267 !J.mol hr-1 (Mopper et al., 1991), divided by the deep water absorption at 410 run of 0.0469 m-1 (see above) to yield an absorption-specific production rate of 0 57 j.lmol m l-1 hr-1. These rates are given seasonal and temporal dependence by multiplying the maximum value by the ratio of total Ed(z) I 1500 !J.mol quanta m-2 s-1, where 1500 !J.mol quanta m-2 s-1 i s as s umed to be the maximum noon time irradiance at depth 0 during the summer under a cloudless sky at the BATS site These maximum rates of photolysis mu s t be attenuated with increasing depth in the water colunm A decay rate with depth is detennined from the diffuse attenuation coefficient ofUV-B for oceanic water. for the UV-B region of solar irradiance (280 -320 nm) ranges from 0 306 to 0 094 m-1. However, the sunlight-normalized action spectra of CDOC photolysis is maximal around 300 run (Kieber et al ., 1990), since the solar irradiance between 280-320 is dominated by energy between 300-320 run (Bird and Hulstrom 1982). The diffuse attenuation coefficient around 300 run is 0 2 m-1 (Smith and Baker, 1981 ). Thus, the decay rate of the surface, absorption-specific rate of photolysis is set to 0.2 m -1. Photolysis is assumed to be zero below the depth where the rate is 1.0 % of the surface value The total absorption of both forms of CDOC is calculated at 410 run with Equation (3.10). This absorption is multiplied by the absorption-specific production rates of DIC and labile DOC to yield the absolute amount of carbon moved into each pool. The relative contribution from CDOC 1 and CDOC 2 is derived from their respective proportion of total absorption. Labile DOC1 created from the photolysis of CDOC1 and CDOC2 is assumed to be colorless, hence the ratio CDOC1 /DOC 1 falls with photolysis Note the production of DOC1 from CDOC1 does not change the quantity of carbon in the DOC1 pool, it just 141

PAGE 155

reduces the CDOC1/DOC1 ratio. All the of photolyzed CDOC2 is removed from DOC2 pool, and placed in the respective DIC or colorless DOC 1 pool. 3.2 Bio-Optical Results This section presents the optical results from the weak mixing, complex ecological model case 3. 3.2 1 Light Energy at the Sea Surface The clear sky simulated and CO ADS-estimated PAR values are displayed in Figure 3.3 As expected, the COADS observations are lower, reflecting the integrated effects of clouds and changing atmospheric conditions on the downwelling irradiance, Ect(O+). The reduction in Ect(O+) from the cloud parameterization is greatest during the summer and lowest during the winter (however, the percentage reduction in clear sky energy is greatest in the winter). The simulated downwelling irradiance in the surface level at 1.25 m of the model, Ect(l.25m), mirrors the 'COADS' curve. There is a slight oscillation with a period of 5 days around the mean trend of the curve. This oscillation in Ect is the result of an oscillation in the climatological ozone calculation The amplitude of the effect is -5 JliilOl quanta. 3 2 2 Absorption The simulated particulate absorption coefficient at 442 nm (aph_t(442) ; Figure 3.4) mirrors the pigment profiles from the Biological Section (Figures 2.19 and 2 20) Particulate absorption at 442 nm is greatest, 0.0216 m-1, after the peak of spring production on Julian day 120 at 85 m The vertical range of this peak narrows and deepens 142

PAGE 156

A Simulated and "COADS" Photosynthetically Active Radiation B 150 100 50 0 1500 'U'l "! E 1000 ro ..... 1:: ro ;:1 o< 0 E 500 0 / / / / / / 100 200 julian day ' ' ' ' 300 Simulated Downwelling IrradianceEd(total,1.25m) 100 200 300 julian day Figure 3.3. (a) The simulated clear sky irradiance at the sea surface (solid line) and the sea surface irradiance (dashed line) estimated from the Comprehensive Ocean Atmosphere Data Set (from Oberhuber 1988) that is assumed to integrated the effects of clouds; and (b) the simulated downwelling irradiance at 1 .25 m 143

PAGE 157

A 0 -20 -40 60 Cll 1-< (\) (\) E -80 -100 -120 -140 0 B 0 -20 -40 -60 Cll 1-< (\) ...... (\) E -80 100 -120 -140 0 () Particulate absorption a,h ( 442) m 0 002 I 100 0.014 0 010 200 julian day CDM absorption 442) m 100 200 julian day 300 300 Figure 3.4 Simulated (a) particulate absorption and (b) colored detrital material at 442 nm. 144

PAGE 158

as the season changes from spring to fall, following the cycle of isolumes (Figure 2 17) While chlorophyll a remains fairly constant at -0.3 J..Lg liter-I throughout the summer, particulate absorption continues to decline. This results from the change of dominance among the functional groups, with FG4 being replaced by FG 1. This shift is reflected by reduction of the accessory pigments (Figure 2.20) of chlorophyll c and photosynthetic carotenoids (PSC) of FG4, and an increase of chlorophyll b pigments of FG 1 and FG2 (Figure 2.20). Since the specific absorption at 442 run of chlorophyll b is -32% of chlorophyll c, and -72% of PSC, the particulate absorption at 442 run decreases when the rapid-growing Chromophycota-species are succeeded by the slow-growing Prochlorococcus. Vertical changes in the particulate absorption coefficient at 442 run also follow the vertical changes in pigment concentrations. Within the euphotic zone, above the subsurface biomass maximum, chlorophyll a stocks are at their lowest at the surface. The low pigment results from the low nutrient availability and commensurate low phytoplankton biomass, as well as the high levels of solar irradiance that reduce chlorophyll a to carbon ratios (equation 3 5). As the photosynthetic accessory pigment concentrations of each functional group are direct functions of their individual chlorophyll a concentration (Table 3.3), the photosynthetic accessory pigment concentrations are also at their lowest at the surface The photoprotective carotenoid (PPC) concentrations (Figure 3.5) are an inverse function of chlorophyll a. The increasing PPC will offset the effect on aph_t(442) of decreasing photosynthetic pigments. However, PPC stocks are also related to total biomass concentrations. Hence, the PPC concentrations are relatively highest, with respect to chlorophyll a, in the surface waters of the upper euphotic zone, but their total concentration is maximal towards the bottom of the euphotic zone In contrast to aph_tC 442), the simulated colored detrital material (CDM) absorption coefficient, arom(442) (Figure 3.4), reflects the seasonal and vertical cycle of the CDM. Photolysis by UV-B reduces the concentration ofCDM, with decreasing effectiveness at 145

PAGE 159

A B Photoprotecti ve Carotenoids -FG 1 -50--100.:: -150 -200--250 0 0 -50 -100 .:.:. .. ........... -150--200 -.. --0 o 'C,-: 0 .oso .. .. _. --------------------------.--:. ... ... o.oso .................... 0 025 ..... 100 200 Jlg PPC liter' 300 Photoprotective Carotenoids -FG2 .. ... . . . . . . .. : ---..... ..:.. .............. ........ . ............... o o s o :: .. o .otS .. . . & -----------c:::; -0 100 200 300 Jlg PPC liter'' Figure 3 5 Photoprotective carotenoids (a)-(d) for functional groups 1 4, respectively. (continued on next page) 146

PAGE 160

c D Photoprotective CarotenoidsFG3 1001--150 f--200 .. . .. -0 100 200 300 J..Lg PPC liter"1 Photoprotective Carotenoids All 0 O .IOQ .. .:.:::::: :.: .. :: : : ::.:: :::::. : : :.. :. : . :.: ... 0 025 . ...... 100 200 J..Lg PPC liter"1 300 Figure 3.5. (continued) 147

PAGE 161

depth as the high-energy photons are attenuated. During the s pring peaks of phytoplankton production and biomass (Figures 2 .16 and 2.18), the water column is still mixing rapidly (Figure 2.17) The concentrations of CDM are, thus diluted even while the s tocks of DOMt and DOM2 are increasing A summer maximum of CDM accumulates at the bottom of the euphotic zone during periods of high stratification Fall mixing and near-surface photolysis reduce CDM to its uniform winter stocks The absorption of light by CDM is, therefore, highest at depth just before the fall; and it is lowest in the surface waters during the peak in summer stratification 3 2 3 Spectral Diffuse Attenuation Coefficient The diffuse attenuation coefficient KJ(A.,Z,t), at wavelengths 412, 442,467, 487, 522, and 567 nm, display the interactions between the changing lOPs and the zenith angle of solar irradiance (Figures 3 .6-3 8) The seasonal profiles of KJ(A.) at these wavelengths generally follow those of the chlorophyll a profiles. There are some minor differences as a result of accessory pigments and CDM. For example the seasonal profiles of KJ(467) corresponds to those of chlorophyll b found in groups FG1 and FG2 (Figure 2.19) Whereas KJ( 487) mimics the seasonal profiles of phycoerythrins of group FG3 (Figure 2.19) At the longer wavelengths of 522 and 567 nm, the effects of particulate absorption are smaller as pigment specific absorption approach their minimum values. In addition, water absorption is much higher and therefore, dominates the absorption of light as it through the water column. As the water absorption coefficient, aw(A), does not change with time or depth, the KJ(522) and KJ(567) generally increase with depth as a result of a decreasing average cosine function The seasonal changes in KJ(522) and KJ(567) are, for the most part, effected by seasonal changes in the solar zenith angle, which is reflected in the surface average cosine. For the blue-green wavelengths scattering 148

PAGE 162

A Diffuse Downwelling Attenuation Coef.m l 0 d'"t:::l -20 <:;:) 0 -40 0 .1>-0 -60 Cll 1-< o.oso v ....... v -80 8 -100 b --1 -120 0 -140 0 100 200 300 julian day B Diffuse Downwelling Attenuation Coef. 442) m 1 0 -20 \:; -40 -60 Cll O.os0 1-< v j ....... 8 -80 -100 1 o ro 0% fl ao -120 -140 0 100 200 300 julian day Figure 3.6. Diffuse downwelling attenuation coefficients at (a) 412 nm and (b) 442 nrn. 149

PAGE 163

A Diffuse Downwelling Attenuation Coef. 467) m1 0 -20 0 rf 'i c:. 0 -40 -60 "-l '""' Q) ...... 11) -80 E -100 o r 0 QOoo -120 -140 0\ 0 100 200 300 julian day B Diffuse Downwelling Attenuation Coef. 487) m 1 0 -20 \:! -40 -60 "-l '""' 11) ...... Q) -80 E } -100 -120 -140 0 100 200 300 julian day Figure 3 7. Diffuse downwelling attenuation coefficients at (a) 467 nm and (b) 487 nm. 150

PAGE 164

Figure 3.8. Diffuse downwelling attenuation coefficients at (a) 522 nm and (b) 567 nm. 151

PAGE 165

is also a greater fraction of beam attenuation Hence, the average cosine is lower (equation 3.17), and the are commensurately higher decreases slightly below 100m. This results from decreasing particulate backscattering, and a reduction in the accessory pigments of chlorophyll band phycoerythrins. There are some small seasonal variations in all the simulated diffuse attenuation coefficients that result from the seasonal and vertical changes of scattering. For example, particulate scattering and backscattering are a function of total chlorophyll a concentration (equation 3 14). Changes in the chlorophyll a concentration will therefore lead to slight changes in the diffuse attenuation coefficients directly through Equations (3. 11) and (3.16) and indirectly through Equation (3.17) These variations are most evident in the that are least effected by pariculate absorption (yellow wavelengths). 3 3 Bio-Optical Discussion 3.3 .1 Seasonal and Vertical Cycles of the Diffuse Attenuation Coefficients The bio-optical results of EcoSirn compare quite favorably with the published data of the Bermuda Bio-Optical Program (BBOP) operating in conjunction with the JGOFS BATS program (Siegel et al., 1995; Siegel and Michaels, 1996). The ratio of is a measure of the relative amount of CDM attenuation (Siegel et al 1995). The seasonal variations of the simulated (Figure 3.9) compare favorably with the observations from BBOP. The lowest values of this ratio are found in the surface waters, during peak insolation and water column stratification The highest values are found in the lower euphotic zone after the peak stratification, much like that found during BBOP 152

PAGE 166

Diffuse Downwelling Attenuation Coef 0 -20 -40 -60
PAGE 167

The simul a ted Kct( 412): Kct( 487) ratios do not reach the magnitude of those observed suggesting a difference between the simulated system and the BATS system. One difference is evident from the simulated aph_t (442) and acdm(442) values The simulated aph_t(442) are generally greater than those of By using a spectral regression analysis of the data and representative spectral curves for the attenuation coefficients for chlorophyll a and CDM, Kch1(440 z,t) and Kcdm(440,z t) respectively Siegel and Michaels ( 1996) suggest that these values should be close to equal throughout much of the surface waters for most of the year The two major exceptions to this hypothesis are 1) at the surface during high insolation where photolysis reduces the concentration of CDM. Here, the attenuation of light due to phytoplankton was found to be much greater than that of CDM; and 2) at the bottom of the euphotic zone during the period of high stratification. Here the attenuation values for CDM can be twice as much as that for phytoplankton. This hypothesized seasonal variation of !
PAGE 168

Another discrepancy between the peaks in aph_t(442) and Kchi(440) is their correlation to chlorophyll a The peak values in aph_tC 442) are found during peaks in simulated chlorophyll a which is be to expected However, the peaks in Kch1(442) do not necessarily correspond to the peaks in observed chlorophyll a. The difference between the magnitude of Clcdm( 442) and KromC 440) at its peak is much the same as the aph_t(442) and Kchi(440) The maximum values for ClcctmC442) are -0.005 m-1 compared to a Kcctm(440) of -0. 03 m-1. The larger difference between ClcctmC442) and Kcctm(440) versus the difference in aph_tC442) and Kch1(440) suggests that the problem with the simulated Kct(412):Kct(487) ratio may lie in the absorption by CDM, rather than the particulate absorption In Section 2, the amplitude of the simulated DOM cycle was thought to be low, compared to observed values of DOC. This would also be reflected in the CDM cyc1e, since bacterial utilization of total labile DOM1 also includes CDOC1 (cf. equation 2.31, 2 .32, and 2.34) The reasons for the low amplitude DOM were previously discussed (Section 2). One of them was related to the lability of the DOM 1 If the lability of the DOM 1 was assumed to be lower in EcoSim; e.g., a larger half-saturation constant for DOC1 utilization, then the CDM1 would have accumulated to greater concentrations prior to the winter overturn. This could also be accomplished by having more than two pools ofDOM (and therefore CDM), each with a different time period of lability That such may exist is found in the exponential slope of the simulated Clcctm(412) to ClcctmC487) (Figure 3.10 and 3.11). The slope, during the peak absorption of 0 .01 for ClcctmC412) and 0.002 for ClcctmC487), in the fall is 0.0215 nm-1. This value means that the greatest relative fraction of the simulated absorption results from CDM2 (CDOC2), rather than CDM1 (CDOC1) as the slope for CDM2 is 0 025 nm-1 compared to the slope of CDM1 of 0.014 nm-1. The derived slope of Kcctm(A) from the observations is closer to 0.014 nm-1, suggesting a large relative fraction of particulate detritus and/or more labile DOM in the observed diffuse attenuation coefficients An additional simulated pool of DOM with the 155

PAGE 169

same spectral slope as DOM1 but less labile than DOM1 would result in a better replication of observed DOC concentration and a CDM spectral signal This result again suggests the need for further data collection on the lability of DOM and CDM. Another potential source of error in this work, and those of inverse modeling techniques may result from the use of the spectral water absorption coefficients of Smith and Baker ( 1981 ). A more recent study suggests these absorption coefficients are too high in the blue (Pope, 1993) If a lower water-specific absorption at 410 nm is used in Section then the derived CDOC1-specific absorption would have been higher This would have yielded a greater Kct( 412):Kct(487) per unit CDOC 1 It is unclear how this would have influenced the comparison between the simulate and the derived from the observations, since Siegel and Mi c haels ( 1996) utilized the absorption coefficients from Smith and Baker ( 1981) to derive their Kcdm(A). 3.3.2 Accessory Pigment Effect on Kct(412):Kct(487) As the CDM pool appears to be low in the simulation the ratio of aph_t(412):aph t(487) will have a greater relative effect on the simulated Kct(412): Kct(487) ratio What is evident in this ratio is that aph_t(412) is about 60% of the aph_t(487) for much of the year at all depths (Figures 3.10 and 3 11) It is also evident in the chlorophyll a-specific absorption curves (Figures 3 12 and 3.13). There is conflicting evidence as to the veracity of this simulated result. The works of Morel ( 1988) and Bricaud et al. ( 1995) suggest the aph_t(412):aph_t(487) ratio should be equal to/or greater than 1. Open ocean work in the Northwest Atlantic suggest this ratio could be less than 1 in the lower euphotic zone (Hoepffner and Sathyendranath 1992). The simulated results are a function of the high concentrations of the accessory pigments chlorophyll b, phycoerythrin and PPC. As there are no measurements of phycoerythrins at the BATS site, there is no way to ascertain whether their simulated concentrations are high The conversion of FG3 biomass during 156

PAGE 170

A -100 -140 0 B 0 -20 -40 -60 en ..... Q) ....... Q) E -80 -100 -120 -140 0 Particulate absorptionm -1 100 200 julian day CDM absorptionm-1 100 200 julian day 300 300 Figure 3.10 Simulated (a) particulate absorption and (b) colored detrital material at 412 run. 157

PAGE 171

A 0 -20 -40 -60 Cl) ..... Q) _. Q) 8 -80 -100 -120 -140 0 B 0 0 Particulate absorption 487) m 100 200 julian day CDM absorptionClrom(487) m 100 200 julian day 300 300 c 0 0 Figure 3 .11 Simulated (a) particulate absorption and (b) colored detrital material at 487 nm. 158

PAGE 172

A B Chlorophyll a-Specific Absorption Day 93 0.1 0 .............. ---.-....,::..._r--r---.--.--..--. I 0.08 s:: g e. 0 Vl 0 .06 () f.;: -() 11) 0.. 0.04 t::$ ->. ..s:: 0.. 0.02 :a () surface SCM I \ I \. I I 0. 00 ._..__,__,__...___,___,___.____..___,_---.J...._--'---J'---'---'--.l......J.--'--'--..o..___.L---'---'--..___,____.l___._,___,_____.__::j s:: .g e. 0 Vl () f.;: -() 11) 0.. Vl I t::$ --::>--..s:: 0.. 0 1-< 0 :a () 400 450 500 550 wavelength (nm) 600 650 700 .Chlorophyll a-Specific AbsorptionDay 180 surface SCM '\ I \ I \ J 0. 00 L...J....___,___._.____L____,__--'---''---'-_...L_ ............. _.__,__L.....o..___,___._I..--L.__.__..J.......J__,__.L_...L.-.l.---'-___.__:.j 400 450 500 550 wavelength (nm) 600 650 700 Figure 3 12 The chlorophyll a specific absorption during the (a) spring peak and (b) summer peak in subsurface chlorophyll a concentrations. 159

PAGE 173

A Chlor o p hy ll aS pec i fic A b sorp t ion-S urface 93 v. 1 80 0.10 i Day 93 ro I 0.08 I Day 1 80 c:: f '-/ 0 I -..... fr 0 , ..c: 0.. 0 ..... 0 0.02 :a <.) ,.... 0.00 4 00 450 500 550 600 650 700 wave l e n gt h (nm) B Chl oroph yll a-S p ec i fic A b sorpt i on SCM 93 v. 1 80 0.10 i Day 93 ro I 0.08 Day 1 80 c:: .s ..... fr 0
PAGE 174

the summer period to cell counts (using the conversions factor in Section 2 ) yields -5000 cells milliliter-1, similar to that measured near the BATS study site (Olson et al ., 1990) However there are measurements of total chlorophyll b and zeaxanthin (Knap et al., 1991 ; Knap et al 1992; Goericke and Repeta 1993 ; Goericke and Welschmeyer, 1993; Knap et al. 1993; Knap et al 1994 ; Michaels et al.; 1994b; Knap et al. 1995 ), as well a s measurements of divinyl chlorophyll a and b which indicate the presence of Pro c hloro c o c cu s species. The simulated chlorophyll b levels appear to be high (Figure 2 26), suggesting that the relative concentrations of FG 1 and FG2 are too high This is further evidenced by the relative fraction of chlorophyll a represented by FG 1 and FG2. In the simulation, the FG 1 plus FG2 fraction of total chlorophyll a approaches 85 % during the summer period. Observations of divinyl chlorophyll a suggest this ratio should be closer to 40 to 60% ( Figure 3.14, Goericke and Welschmeyer, 1993), although higher fractions of biomas s have been reported (Campbell and Vaulot, 1993; Goericke and Welschmeyer 1993 ; Campbell et al., 1994). The overestimation ofFG1 and FG2 is also suggested in the PPC concentrations (Figures 2 .26 and 3 5). The maximum total concentrations are approximately double that of the measured zeaxanthin/lutein (Knap et al., 1991 ; Knap et al. 1992 ; Knap et al ., 1993 ; Knap et al 1994 ; Michaels et al 1994b ; Knap et al., 1995). As the simulated total concentration s primarily result from FG 1 and FG2, this would again suggest an overestimation of the biomass of FG 1 plus FG2 The conversion of the biomass of FG 1 plus FG2 into cells rnilliliter-1 (us i ng the constants from Section 2) yield a maximum of -150,100 cells rnl-1. This is close to, but slightly greater than observations near the BATS site of -100,000 cell rnl-1 (Olson et al 1990) The overestimation of accessory pigment concentration may re s ult from an inaccurate description of chlorophyll a/accessory pigment ratio (Table 3 3), or perhaps, simul a ted conditions that favor Pro c hloroc o ccus-like FG 1 and FG2 over that of Synecho co cc us-like FG3. The exclu s ion of a functional group to represent eukaryotic 161

PAGE 175

A: Temperature (C) 1/85 3/85 6/85 8/85 10/85 12/85 1/86 3/86 5/86 7/86 9186 11/86 2187 i2 -60 2b -100 / 14 0 ..... 20 ............... -180 -220 B : Chlll:l (ng L-1 ) 3/85 6/85 8/85 10/85 12/85 1 /86 3/86 5186 7/86 9/86 11/86 2187 -20 -60 -100 140 -180 220 C : Ratio Chlll:l /total Chi a 3/85 6/85 8/85 10/85 12/85 1 /86 3/86 5/86 7/86 9186 11/86 2187 -20 60 -100 -140 -180 -220 Figure 3 .14. Sea sonal ratio of divinyl-chl oro phyll a to total chlorophyll a concentrations around Bermuda Diviny l-chlorophyll a is an indicator pigment for Prochlorococcus species (redrawnfrom Goericke and Welshmeyer, 1993). 162 -20 -60 -100 -140 -180 -220 -20 -60 -100 -140 180 -220 -20 60 -100 -140 -180 -220

PAGE 176

picoplank:ton, e.g. prasinophytes, to compete with FG 1 and FG2 may also contribute to potential overestimation of these groups. Lastly, a higher grazing stress on FG 1 and FG2 would have reduced their biomass and pigment stocks. In any case an overestimation of chlorophyll b and PPC will directly impact the ratio, since the aph_t(412):aph_t(487) dominates the signal for waters that have little CDM absorption. This has implications beyond just the current simulation results. In a case where chlorophyll band/or PPC are high, the ratio of aph_t(412): aph_t( 487) may be less than one, as seen in the observed data as well as the simulated data. In oligotrophic surface waters where the ratio is relatively unaffected by CDM; i.e., during summer stratification and insolation, the aph_t(412):aph_t(487) may dominate the signal An aph_t(412):aph_t(487) greater than 1 would cause great problems for inverse methods that calculate chlorophyll a concentration (eq. Carder et al., 1996; Siegel and Michaels, 1996) from observed K!'s and/or sea surface reflectance using preconditioned spectral absorption curves for aph_t and CDM Siegel and Michaels ( 1996) work around this problem with their summer samples by setting the derived absorption of light by phytoplankton and CDM to zero when the observed dropped below a certain value. Since the collection of remote data is biased towards periods of high insolation and calm weather condition, it may be common to experience aph_t(412):aph_t(487) less than 1 in oligotrophic waters Such observations of low should be explained, not eliminated. The reasons for these low should be quantified and incorporated into remote sensing algorithms, in order to lower variance between observed chlorophyll a and those estimated from remote sensing techniques 163

PAGE 177

3.3 3 Potential CDM Interference of Remotely-Sensed Chlorophyll a The optical paradigm for the Case I waters of the Sargasso Sea has been that "nothing besides algae with their associated retinue, can influence the optical properties" (Morel 1991) of these waters. The CZCS chlorophyll a algorithm was primarily developed in these waters types, with the overriding assumption that any interference by CDM would co-vary with chlorophyll a (Gordon and Morel, 1983; Gordon et al 1988). So even if CDM were a constituent of the ocean color signal, a co-varying fraction of CDM would allow the retrieval of chlorophyll a stocks by empirical techniques. The ratio of simulated particulate absorption to CDM absorption at 442 nm at the surface (Figure 3 15a) suggests that chlorophyll a and CDM do not co-vary. The implications are that even in Case I waters, seasonal interference of remote sensing reflectance by non co-varying quantities of CDM will cause errors in the retrieval of chlorophyll a estimates. Thus, explicit incorporation of non-algal colored material should be included in remote sensing reflectance algorithms that estimate chlorophyll a concentrations, even in the clearest oceanic waters (Carder et al., 1986; Carder et al., 1989 ; Carder et al ., 1991; Walsh et al., 1992; Siegel and Michaels 1996). Several assumptions are made when comparing the simulated surface chlorophyll a results with those of the CZCS time-series from Bissett et al. ( 1994) The first is that the average timing and strength of the water column mixing during the period that the CZCS data were collected, is approximated by the simulated mixing. Thus, the timing of the simulated chlorophyll a signal would approximate that of the CZCS signal. The second assumption is that the aliasing of the CZCS data towards clear weather conditions will not render the comparison invalid Lastly, this comparison assumes that this one-dimensional model accurately reflects the sources and sinks of light attenuating material in the three dimensional system of the Sargasso Sea. 164

PAGE 178

100 200 j ulian da y 300 Figure 3 .15. (a) Ratio of simulated CDM absorption to particulate absorption at 442 nm; (b) simulated surface chlorophyll a concentrations (solid line) and six year average of CZCS estimated chlorophyll a concentration at the BATS site (dashed line); and (c) ratio of CZCS estimated chlorophyll a to simulated chlorophyll a. 165

PAGE 179

While acknowledging the above assumption may be imperfect, Figure 3 .15 a also suggests that non co-varying CDM interference may be evident in the 6 year record of CZCS images at the BATS location. Such interference seems to appear in a plot of simulated surface chlorophyll a stocks and CZCS chlorophyll a ( Fi g ure 3 .15 b). The peak s of sim ulated aph(442):acnM(442) are in phase with the peaks of the ratio of CZCS:ECOsim chlorophyll a concentrations (Figure 3.15c), suggesting the CZCS estimation of chlorophyll a may have been contaminated by CDM. This contamination appears t o be greatest just after the sp rin g bloom during the maximum in DOM/CDM concentration (Figure 2.23). It also appears to be evident after the onset of fall mixing, when the seasonal maximum of CDM at depth i s mixed to the surface. The seasonal CDM contamination mimics that of coastal waters that undergo seasonal stratification ( Hochman et al., 19 95). 3.4 Summary The simulated seasonal and vertical cycle of the diffuse downwellin g attenuation coefficients compare well with the observed cycles of spectral downwellin g attenuation. The discrepancies between the s imulated and observed Ko's appear to re s ult fro m an underestimation of li g ht absorption by color detrital material, and a pos s ible overesti mation of accessory pigments. As noted in Section 2, additional data on the proc esses invol ved with the DOM/CDM cycle are needed. The simulated absorption of li g ht by CDM at the surface does not co-vary with chlorophyll a concentrations Thi s seasonal interference by CDM appears to be evident in the CZCS estimation of ch l orophyll a concentrations around the BATS stud y s ite. This simulation s uggests that the absorption of light by colored detrital material needs to explicitly incorporated into ocean color algorithms. 166

PAGE 180

4. The Big Picture 4.1 Overview Since the beginning of the industrial revolution, the atmospheric carbon dioxide concentration has increa s ed by -160 Pg C or -:-80 ppm; via anthropogenic sources (Schle s inger, 1991; Siegenthaler and Sarmiento, 1993; Moore and Braswell, 1994). However, this atmospheric increase doe s not represent the total release of carbon dioxide from the burning of fo s sil fuels and deforestation as the ocean and terrestrial systems have buffered a portion of the total release. During the 1980's anthropogenic releases by fossil fuel emissions and deforestation were estimated to be -5. 4 Pg C y-1 and -1.6 Pg C y -1, respectively (Siegenthaler and Sarmiento, 1993). During this period, the atmosphere inventory only increased by -3.4 Pg C y 1 or about 60% of the C02 released from the burning of fossil fuel. The oceanic uptake has been estimated (with some controversy) to be -2 Pg C y-1 ( Keeling and Shertz, 1992 ; Quay et al 1992; Siegenthaler and Sarmiento, 1993), leaving the -1.6 Pg C y-1 released by deforestation to be taken up by a terrestrial sink. The estimation of these past sinks of oceanic carbon dioxide has been controversial ( eg. Tans et al. 1993) However predictions of the future rate of oceanic uptake has been even more so. That the oceans could buffer the entire -4000 Pg C of easily recoverable fossil fuel C02 is not in question (Broecker and Peng, 1982). What is questioned is the timing, and possible feedbacks, of the oceanic uptake Of the 2 Pg C y-1 uptake, -0.4 Pg C y-1 has been estimated to be accumulating in the oceanic surface waters, with the remaining -1. 6 Pg C y-1 accumulating in the intermediate and deep waters (Siegenthaler and Sarmiento, 1993). As the ventilation time of these waters ranges from 1 to 1000 year s 167

PAGE 181

the sequestration period of carbon dioxide in the oceans is not a constant. The time period of sequestration depends on the depth of the carbon dioxide accumulation, as well as the ventilation time of the waters at that depth (or along a particular isopycnal surface). Hence, part of the problem with estimating oceanic carbon sinks and feedbacks is found in the estimation of the sequestration period for the anthropogenic C02 The oceanic removal of C02 by deep water formation in the polar regions has been estimated to be -0.4 Pg C y-1 (Tans et al., 1990). This would leave a net 1.6 Pg C y-1, of the total2.0 Pg C y-1 that entered the ocean, to influx between the latitudes of so Nand so S. Yet, the equatorial regions (between 15 Nand 1s S) are a net source of C02 to the atmosphere, on order of -1.6 Pg C y-1 (Tans et al 1990). Hence, a large oceanic sink, some 3.2 Pg C y-1, must be influxing into the central gyre regions of the worlds oceans. This work attempts to simulate the downward flux ofDIC from the atmosphere into the ocean at a central point in the Sargasso Sea. With the use of an ecological system simulation (EcoSim), the mechanisms by which C02 is transported through the water column are quantitatively explored. In addition, feedbacks to the flux of C02 are identified and the location of the maximum DIC build up in the region of the Bermuda Atlantic Time-Series (BATS) site is hypothesized. 4.2 Conceptual framework The direction of the C02 flux across the air-sea interface in the Sargasso Sea at the BATS site is related to the seasonal change of the physical, chemical, and biological interactions. During the fall/winter period, the mixed layer deepens and the mixing rate increases as a function of wind stress and cooling temperatures (Figure 4.1). The deep mixed layer brings up dissolved inorganic carbon (DIC)-rich waters from depth and mixes them with DIC-poor waters at the surface (Figure 4 2). Such an increase of surface DIC concentration, that would typically decrease the solubility of C02 gas, is overwhelmed by 168

PAGE 182

Winter WindG Stress -40 ppm I Flux Physical Accumulation of COz Summer Wind G Stress . +30 ppm Mixed Layer jFl!l,X Biological Flux Figure 4.1. Graphic depiction of differences between relative C02 flux during (a) cold, deep winter mixing, and (b) warm, shallow summer mixing. 169

PAGE 183

=Summe I -----------Euphotic Zone Permanent Thermocline Figure 4 .2 increment. Seasonal profiles of DIC. Dashed l i nes show anthropogenic C02 170

PAGE 184

the increase in solubility resulting from the cooling temperature. As a result a solubility gradient draws C02 from the atmosphere into the sea surface (negative of -40 ppm Bates et al. 1996) where it is equally distributed within the mixed layer. During the spring/summer, the mixed layer shoals and the mixing rate decreases as insolation increases and wind stress is minimized (figure 4.1). This causes a maximum in stratification within the seasonal thermocline and limits the exchange of water across density surfaces within the euphotic zone. The rising temperatures reduce the solubility of C02 gas and the so lubility gradient swi tches direCtions driving C02 from the sea surface to the atmosphere (positive of -30 ppm Bates et al., 1996 ) The shallow mixed layer, lower wind stress, and strong stratification of the seasonal thermocline all act in concert to limit the flux of C02 back to the atmosphere Another limiting mechanism of the inorganic efflux to the atmosphere is the biological fixation of DIC into organic carbon. This DIC, fixed into organic carbon by autotrophic production, moves through the food web until it is stored in sinking particles, or transported by migrating zooplankton beneath the euphotic zone. Remineralization processes act rapidly on this flux of organic matter, returning both the inorganic carbon and nutrients back to the water column (7 5 % of the carbon, 91% of the nitrogen within the upper 500 meters, Martinet al ., 1987). Since mixing is very low in the thermocline, the recycled amounts of carbon and nutrient increase, until the concentration differences between the euphotic and aphotic zone are large enough to allow diffusive transport of DIC and nutrients back into the base of the euphotic zone. The carbon and nutrients are fixed here again into particles and transported back beneath the euphotic zone without ever moving into the summer surface mixed layer. This biological cap of subsurface phytoplankton populations limits the flux of inorganic carbon into the surface of a stratified water column. Furthermore nitrogen is more rapidly removed from the sinking particles (Martin et al 1987 ; K.nap et al., 1991; K.nap et al., 1992; Knap et al ., 1993; Knap et al ., 1994 ; Knap 171

PAGE 185

et al., 1995). Thus, it is recycled at shallower depths in the water column, and can be returned to the euphotic zone more quickly than DIC. This process results in a net inorganic carbon to nitrogen loss from the euphotic zone that is higher than the original Redfield uptake ratios of the phytoplankton (eg. Sambrotto et al., 1993). The nitrogen cycle is also enhanced by nitrogen fixation (Carpenter, 1983; Carpenter and Romans, 1991 ; Carpenter and Capone, 1992; Karl et al., 1992; Letelier and Karl, 1996 ; Walsh, 1996), further accelerating the removal of DIC, relative to upward-fluxing nitrogen, to the deep waters. The net result of these processes is that each year during the fall/winter, atmospheric C02 fluxes into the homogenous deep, cold mixed layer. This increment of DIC will not return to the atmosphere during the next spring/summer period because of a combination of the initial photosynthesis the subsequent biological capping of the recycled materials at the base of the euphotic zone, and the concurrent low rates of air-sea gas exchange. As long as the net flux of C02 into the surface waters was greater during the cooler periods than the out-gassing during the warmer periods, DIC would accumulate within and below the permanent thermocline. What happens in a scenario with an incremental increase in atmospheric C02 from anthropogenic sources over the course of the year? Assuming there in no climatic change that affects the oceanic mixing cycles the physical and biological processes will operate at the same rate. However, during each fall/winter period an increasing amount of C02 fluxes into the mixed layer. Commensurately, during the spring/summer period there is less efflux of C02 Hence, there will be an incremental increase of DIC in the surface waters. This increase in surface DIC will reduce the upward flux of DIC from depth, as the vertical gradient between the surface and deep waters is now smaller. Yet, th e upward flux of nutrients is still the same; therefore, the relative fraction of anthropogenic DIC incorporated into biogenic particles will be enhanced. The differential recycling of nitrogen 172

PAGE 186

and carbon will further accelerate this process of anthropogenic DIC removal to depth. Thus the biological pump is enhanced without an increase in nutrient supply. Recently observed increases in DIC stock s in the 18. mode waters at the BATS site (Figure 4.3, Bates et al. 1996) and between 150-500 min the South Atlantic (Figure 4.4, Wallace et al., 1996) suggest that the seasonal cycle of mixing, the biological pump, and the biological cap are acting in concert to remove surface anthropogenic influxes to the intermediate waters between -150 500 m. The present work is a small step in the long process to test this hypothesis. It sought to answer the following questions -1 ) does anthropogenic DIC accumulate in the intermediate waters of 150-500 m ; and 2 ) what are the driving mechanisms that effect such an accumulation? 4.2.1 EcoSim A predictive ecosystem model has been developed for the central Sargasso Sea around the Bermuda Atlantic Time-Series site. This model is different from past numerical studies (e.g Wroblewski et al., 1988; Fasham et al., 1990; Sarmiento et al., 1992; Fasham et al. 1993; Bissett et al 1994 ; Doney et al 1996; Hurtt and Armstrong 1996; Lawson et al., 1996) in a number of ways. Constant ratios are no longer invoked in the seasonal cycling of carbon and nitrogen, allowing the calculations of carbon uptake and remineralization to occur at different rates than those of nitrogen. Autotrophic growth is also modeled as a function of the quantity of spectral light at depth This allows depth varying and seasonal succession of phytoplankton, resulting from the interaction of light, nutrient, and temperature regulation of the growth processes The diverse phytoplankton community of oligotrophic waters was represented by are four functional groups of phytoplankton: two types of Prochloroco c cus, Syne c h oco ccus and Chromophycota. 173

PAGE 187

bn 2100 -.------------------. 150-250 m cr.J Q.) 0 2080 a "-" 2060 u 2040 ro Q.) 1989 1990 1991 1992 1993 b ,-. 2070-cr.J Q.) 0 a "-" N 0 u 2060-2050-. 2040- .. mode water ... .. .... . 2030 ---'--1,.-1-98_9___,1_1_9-90-r1-1-99-1--.-1-19_9_2-,.-1-19_9_3 _l ,-. bJ) 2090 0 2080 "-" N 0 2070 u 2060 0-250 m 1993 Figure 4.3 TC02 concentrations from BATS (redrawn from Bates et al., 1996) 174

PAGE 188

0 1000 -5 2000 0 3000 4000 -20 e WOCE 1993 -D-SAVE 1989 _._ GSECS 197 4 -15 -10 I A I -5 1 1A:1111e1 01 A I I AO Bl I -'CJI I I I l--iOt+---A---; 0 Residuals (J..L mol kg-1) 5 Figure 4.4. Residual differences between predicted and observed DIC concentrations in the South Atlantic (redrawn from Wallace et al., 1996) Residual differences in the surface waters result from the accumulation of anthropogenic C02. 175

PAGE 189

These groups led to the establishment of different food web structures as a function of depth and season, yielding major changes of carbon and nitrogen fluxes. To allow a realistic description of weak mixing across the thermocline, local nitrification and nitrogen fixation are included as autochthonous and allochthonous sources of nitrogen. Finally, C02 gas exchange at the sea surface, incorporating chemical enhancement at low wind speeds, and labile and relict pools of DOC and CDOC, provide a realistic assessment of the dissolved forms of inorganic and organic carbon The inclusion of the additional state variables and processes greatly expand s the validation data set against which EcoSim can be evaluated. The model results are validated not only by chlorophyll a and productivity observations, but also by sediment trap fluxes profiles of C02, N03, and NH4+N02 POC and PON biomass, pigment profiles, spectral diffuse attenuation coefficients, and dissolved organic carbon profiles In addition, the functional relationships in the state equations were derived or taken directly from culture/laboratory experiments. Therefore, the time series data is used strictly for validation, as opposed to being used to derive input parameters. 4.3 Where does DIC accumulate? The seasonal cycle of C02 is relatively well resolved by the simulation. The summer observations are overestimated by -20 J.Lmol kg-1; but, as the wintertime DIC concentrations and C02 partial pressures of the model and field data are the same the net result is just a reduction in the seasonal amplitude of the total gas exchange across the air -sea surface. The model predicts an annual invasion of 588 mmol C02 m-2 y-1 of atmospheric C02 compared to 220-830 mmol C02 m-2 y-1 for the BATS site (Bates et al., 1996), at an atmospheric concentration of 355 ppm. At a constant atmospheric C02 level, the surface waters between 0 150 m store only 3.6 % of the 588 mmol C02 m-2 y-t influx (Case A; Table 4.1) Total water column 176

PAGE 190

storage of the atmospheric flux from 0-1000 m is 262 mmol C02 m 2 y-1 (0-1000 m increase minus the bot t om flux of DIC) s u ggest ing t hat 241 mol C02 m-2 y-1 (or 41%) of the atmospheric flux is accumu l ating in the 150500 m depth interval. The yearly increase of DIC as a fu n ction of depth (Figure 4.5) s uggest s that there is a peak in DIC accumula tio n around 320 m. Table 4.1. C02 Fluxe s and Inte g r a ted Changes in DIC Concentration s Kz iAtm Surfa ce Bottom L1 0-L1 150L1500-L1 0Case (10-5 C02 Flux Flux 150m 500m 1000 m lOOOm m 2 s-1) (p pm) ( mmol (mmo l ( mmol ( mmol ( mmo l ( mmol m-2y-1) m-2 y-1) m-2 y-1) m-2 y-1) m-2y-1) m-2y-1) A I 3.7 I 0.0 I 588 I 33 I 2 1 I 423 I -149 I 295 B I 3.7 I 2.0 I 624 I 33 I 56 I 423 I -148 I 331 c I 1.0 I 0.0 I 59 1 I 6 I 6 I 305 I 16 I 327 D I 1.0 I 2.0 I 626 I 6 I 41 I 305 I 16 I 362 Assuming a scenario of increa si n g atmospheric C02 (C as e B; Table 4 1 ) of 2 ppm y 1 ( Moore and Braswell 1994 ), th e s urface fluxe s are now 624 mmol C02 m-2 y-1, of which 8.9 % accumu l ates in the surface waters. The integrated accumulation from 0-1000 m of atmospheric C02 is now 298 mmol C02 m-2 y-1 ( total accumulation bottom flux) with 242 mmol C02 m-2 y 1 accumulating in the 150500 m inte rval. The incr ease d DIC accumulation over case A is mo s tly in the surface waters However as a r es ult of the increas e in 0 500 m interval, the upward flux from 500 1000 m interval i s s lightly r e duced. The peak accumulation i s again around 320 m (Figure 4.6). There is a decr ease in integrated DIC betw ee n 500 1000 m, which is smal l and hardly noticeable in the DIC profile s ( Fi g ure s 4.5 and 4.6 ) However, it does re s ult in an upward flux of DIC that is trapped in the 150500 m interval. I t may suggest that thi s interval is out of balance with the downward D I C flux in this one-dimensional model and I 177

PAGE 192

A B Dissolved Inorganic Carbon (DIC) -800 2050 2100 -200 -400 -600 -800 2150 J..Lmol C kg DIC Accumulation 2200 -1.0 -0.5 0 0 0 5 J..Lmol C kg 1.0 1.5 2.0 Figure 4.6. Case B (a) DIC profiles on Julian day 0 (solid line) and Julian day 365 (dashed line); and (b) change in DIC concentration with depth 179

PAGE 193

do not know if this particular fmding approximates the actual processes within the 500 1000 m depth range As there is a large DIC gradient change between 600 and 1000 m, these results could be driven by an overestimation of the vertical mixing within this depth interval The vertical eddy diffusivity, Kz, of 0.37 cm2 s-1 was chosen from actual microscale measurements in the North Atlantic oligotrophic gyre (Lewis et al 1986 ) This value for Kz appears reasonable by other accounts (Gargett 1984 ) ; however a lower Kz would reduce the upward flux ofDIC, particularly within the 500-100m interval Cases C and D (Table 1; Figures 4.7 and 4.8) of this simulation represent the effect of a lower minimum Kz of 0.1 cm2 s 1 on the DIC solution (this lower K z also effects a small reduction in the simulated primary productivity of 17 g C m2 y-1). The constant atmospheric C02 concentration of 355 ppm of Case C yields a slightly greater net influx of C02 of 3 mmol m 2 y 1 over Case A. However, the net 0150m increment is now only 1.0 % of total influx (Table 4.1). There is a net increase over the whole domain from 0-1000 m of 321 mmol DIC m-2 y 1 from atmospheric sources of which 311 mmol m-2 y 1 is stored in the 0 500 m interval. There is a net 15 mmol increase in the 500-1000 m interval, of which 6 mmol m 2 y 1 fluxes upward from the bottom boundary The 305 mmol m-2 y 1 increase in the 150500 m interval is lower than the total increment within the same depth interval of Case A. However, 149 mmol m-2 y 1 of the increment in Case A comes from the net upward DIC flux from the 500 1000 m interval. Hence, there is a 74 mmol m-2 y-1 greater atmospheric C02 storage within the 150 500 m interval of Case C, compared to Case A. The lower K z effects a lower upward flux of DIC and results in a greater storage of atmospheric C02 within the intermediate waters of 150500 m. The peak DIC accumulation in Case Cis at 280m, -40 m shallower than Case A (Figure 4.7) A 2 ppm atmospheric C02 increase over the year results in an increase in the total influx of C02 to 626 mmol m 2 y-1, 2 mmol m 2 y-1 greater than that of Case B. The 180

PAGE 194

A ell """ 0 ..... 0 E B Dissolved Inorganic Carbon (DIC) -200 \ :--. -400 -600 -800 2050 2100 -400 -600 -800 2150 J.Lmol C kg DIC Accumulation 2200 -0.5 0 0 0 5 1.0 1.5 2.0 J.Lmol C kg Figure 4 7. Case C (a) DIC profiles on Julian day 0 (solid line) and Julian day 365 ( dashed line) ; and (b) change in DIC concentration with depth. 181

PAGE 195

PAGE 196

surface storage is now 41 rnmol m-2 y-1, representing 6 5 % of the total influx Like the other cases, the greatest storage of atmospheric C02 is within the 150-500 m depth interval (Figure 4 8). While there are some differences between in the model results when driven by different K z's, the main conclusion of a peak DIC accumulation within the 150-500 m depth interval is unaffected 4.4 Summary The re s ults from this complex model suggest that C02 of anthropogenic origin is accumulating within the intermediate waters of the central gyres, for the most part between depths 150-500 m This depth interval of DIC sequestration, and the magnitude of its accumulation, match the results at the BATS study site (Bates et al. 1996) those of the South Atlantic Ventilation Experiment (SAVE : 1988-1989) and the World Ocean Circulation Experiment Hydrographic Programme (WHP: 1991-1995) (Wallace et al., 1996) The model's results also mirror the accumulation and penetration depth of isotopically light carbon in the Pacific Ocean (Q ua y et al., 1992). The sequestration of excess DIC beneath 150m is capped by the biological processes at the base of the euphotic, which are enhanced by the more rapid remineralization of nutrients relative to carbon In essence, the biology acts not only as a forcing mechanism for the downward carbon flux, but also as a capping mechanism to keep the accumulated DIC from reaching the sea surface. The combination of these biological proces s es along with the seasonal mixing cycle, increases the buffering capacity of the oceanic surface waters for anthropogenic C02. Numerical simulations of the oceanic C02 cycle should include these biological processes to better estimate the effects of future anthropogenic C02 releases on the global carbon cycle 183

PAGE 197

5. Conclusions The goal of this work was to test a hypothesis on the accumulation of anthropogenic C02 in the Sargasso Sea. To achieve this goal, a complex numerical model was built that simulates the seasonal change s in the upper water ecological sy s tem of this region. This simulation analysis yielded the following conclusions : 1) A complex numerical description of the autotrophic community simulated the seasonal profiles of phytoplankton species succession that match those collected by the BATS program This succession was effected by chromatic adaptation to a spectral light field and differential nutrient utilization 2) Numerical descriptions of differential carbon and nitrogen cycling, local nitrification, and nitrogen-fixation, in place of unrealistic influxes of "new" nitrate across the thermocline, provided realistic estimates of biomass productivity, air-sea C02 exchange, settling losses, and diffuse attenuation coefficients that matched those data collected by the BATS and BBOP programs. 3) The explicit numerical description of carbon and nitrogen utilization by heterotrophic bacteria simulated a population that was not nitrogen-limited in these waters Numerical descriptions of ecosystem s based solely on nitrogen dynamics, or fixed carbon to nitrogen ratios, may yield an inaccurate prediction of carbon and nitrogen fluxes. 4) The simulated absorption of light by colored detrital material did not co-vary with particulate absorption, matching the data collected by BBOP. Except for two seasonal periods, the surface chlorophyll a concentrations appeared to mimic those estimated by a long-term average of the CZCS data Major differences occurred during the month following the spring bloom, and at the onset of fall mixing The differences between the simulated chlorophyll a and those estimated from the CZCS were in phase with the peaks in 184

PAGE 198

relative light absorption by CDM, suggesting that the CZCS signal was periodically contaminated in these clear oceanic waters. 5) Anthropogenic C02 was sequestered at shallow depths, 150 500 m, within this simulation The return of the sequestered anthropogenic C02 to the surface waters was capped by the biological processes that preferentially recycled nutrients relative to carbon. This sequestration matched WOCE, SAVE, and BATS observations 185

PAGE 199

References Amon, R.M.W., and Benner, R. 1994. Rapid cycling of high-molecular weight dissolved organic matter in the ocean. Nature. 369: 549-552 Anderson, S.M., and Roels O A. 1981. Effects of light intensity on nitrate and nitrite uptake and excretion by Chaetoceros curvisetus Marine Biology 62 : 257-261. Bacastow R.B., Keeling, C.D. Lueker, T.J., Wahlen, M ., and Mook, W G. 1996. The 13C Suess effect in the world surface oceans and its implications for oceanic uptake of C02 : Analysis of observations at Bermuda. Global Biogeochemical Cycles. 10: 335-346. Banse, K. 1990. New views on the degradation and disposition of organic particles as collected by sediment traps in the open sea Deep-Sea Research. 37: 1177-1195. Banse, K. 1991. False Advertising in the Greenhouse ? Global Biogeochemical Cycles. 5 : 305-307. Banse K. 1992. Grazing, temporal changes of phytoplankton, and the microbial loop in the open sea p. 409-440. In Falkowski, P G ., and Woodhead, A.D. [eds ], Primary productivity and biogeochemical cycles in the sea. Plenum Press Banse, K. 1994. Uptake of inorganic carbon and nitrate by marine plankton and the Redfield ratio Global Biogeochemical Cycles 8: 81-84 Banse, K. 1995. Zooplankton: Pivotal Role in the control of ocean production. ICES Journal of Marine Science 52 : 265-277. Barlow, R.G., and Alberte R.S. 1985. Photosynthetic characteristics of phycoerythrin containing marine Synechococcus spp Marine Biology. 86: 63-74. Bates, N.R., Michaels, A.F., and Knap, A.H. 1996 Seasonal and interannual variability of oceanic carbon dioxide species at the U S JGOFS Bermuda Atlantic Time Series Study (BATS) site. Deep-Sea Research II. 43: 347-384 Bidigare, R.R., Marra, J., Dickey, T D lturriaga, R Baker, K.S., Smith, R.C., and Pale, H. 1990a. Evidence for phytoplankton succession and chromatic adaptation in the Sargasso Sea during spring 1985. Marine Ecology Progress Series 60: 113122 Bidigare, R R., Ondrusek, M.E., Morrow, J H ., and Kiefer, D A 1990b In vivo absorption properties of algal pigments p. 290-302. In Spinrad, R.W. [ed ], Ocean Optics X, Proc. SPIE 1302 SPIE The International Society for Optical Engineering. 186

PAGE 200

Bidigare R.R., Schofield, 0., and Prezelin, B B 1989 Influence of zeaxanthin on quantum yield of photosynthesis of Synechococcus clone WH7803 (DC2). Marine Ecology Progress Series. 56 : 177-188 Bienfang, P .K. 1980 Phytoplankton sinking rates in oligotrophic waters off Hawaii USA. Marine Biology. 61: 69-77. Bienfang, P.K. 1981. Sinking rates of heterogeneous, temperate phytoplankton populations. Journal of Plankton Research 3: 235. . Bienfang P.K., Harrison, P.J. and Quarmby, L.M 1982 Sinking rate response to depletion of nitrate, phosphate and silicate in four marine diatoms. Marine Biology. 67: 295-302. Bird, R.E., and Hulstrom R.L. 1982. Terrestrial Solar Spectral Data Sets. National Technical Information Service, U. S. Department of Commerce Bissett, W P Meyers, M B., Walsh, J J. and MullerKarger, F.E 1994 The effects of temporal variability of mixed layer depth on primary productivity around Bermuda. Journal of Geophysical Research. 99 : 7539-7553. Bissett, W.P. Patch, J.S. Carder, K.L., and Lee, Z. 1996 Pigment packaging and chlorophyll a-specific absorption in high light oceanic waters Limnology and Oceanography. (in press). Bjjljrnsen, P.K. 1988 Phytoplankton exudation of organic matter : Why do healthy cells do it? Limnology and Oceanography. 33 : 151-154. Blackman, F .F. 1905 Optima and limiting factors. Ann. Bot. 19: 281. Brand, L.E., and Guillard, R.R.L. 1981 The effects of continuous light and light intensity on the reproduction rates of twenty-two species of marine phytoplankton Journal of Experimental Marine Biological Ecology 50: 119-132 Bratbak, G Egge, J K., and Heldal, M. 1993 Viral mortality of the marine alga Emiliania huxleyi (Haptophyceae) and termination of algal blooms Marine Ecology Progress Series 93: 39-48 Bricaud, A., Babin, M., Morel, A., and Claustre, H. 1995. Variability in the chlorohyll specific absorption coefficients of natural phytoplankton: Analysis and parameterization. Journal of Geophysical Research. 100: 13,321-13 331. Bricaud, A., Bedhomme A., and Morel, A. 1988 Optical properties of diverse phytoplanktonic species : experimental results and theoretical interpretation Journal of Plankton Research. 10 : 851-873 Broecker, W S ., and Peng, T. 1992 Interhemispheric transport of carbon dioxide by ocean circulation. Nature. 356 : 587-589 Broecker, W.S., and Peng, T .H. 1982. Tracers in the Sea Lamont-Doherty Geological Observatory. Columbia University. Broecker, W.S., Takahashi, T Simpson H.J ., and Peng T.-H 1979 Fate of fossil fuel carbon dioxide and the global carbon budget. Science 206 : 408-418. 187

PAGE 201

Brophy, J E., and Carlson, D.J 1989. Production of biologically refractory dissolved organic carbon by natural seawater microbial populations. Deep-Sea Research. 36: 497-507. Buck, K.R., Chavez, F.P., and Campbell L. 1996. Basin-wide distributions of living carbon components and the inverted trophic pyramid of the central gyre of the North Atlantic Coean, summer 1993. Aquatic Microbial Ecology. 10 : 283-298 Buesseler, K.O. 1991. Do upper-ocean sediment traps provide an accurate record of particle flux? Nature. 353 : 420-423. Bushaw, K.L. Zepp, R.G., Tarr, M.A., Schulz-Jander, D ., Bourbonniere, R.A., Hodson, R.E., Miller, W.L., Bronk, D.A ., and Moran, M.A. 1996. Photochemical release of biologically available nitrogen from aquatic dissovled organic matter. Nature 381:404-407. Campbell, L., Nolla, H.A. and Vaulot, D. 1994. The importance of Prochlorococcus to community structure in the central North Pacific Ocean. Limnology and Oceanography. 39 : 954-961 Campbell, L., and Vaulot, D 1993. Photosynthetic picoplankton community structure in the subtropical North Pacific Ocean near Hawaii (station ALOHA). Deep-Sea Research I. 40: 2043-2060 Capone, D.G., Ferrier, M.D and Carpenter, E.J 1994. Amino acid cycling in colonies of the planktonic marine cyanobacterium Trichodesmium thiebautii. Applied and Environmental Microbiology. 60: 3989-3995. Carder, K.L., Hawes, S K., and Lee Z P 1996. SeaWiFS Algorithm for chlorophyll a and colored dissolved organic matter in a subtropical environment. Journal of Geophysical Research. (submitted). Carder, K.L., Hawkes, S K., Baker, K.A. Smith, R.C., Steward R.G., and Mitchell, B.G. 1991. Reflectance model for quantifying chlorophyll a in the presence of productivity degradation products. Journal of Geophysical Research. 96: 20,599-20, 611. Carder, K.L., Steward, R.G Harvey, G.R., and Ortner P.B. 1989 Marine humic and fulvic acids: their effects on remote sensing of ocean chlorophyll Limnology and Oceanography 34: 68-81 Carder, K L., Steward, R.G., Paul, J H., and Vargo G .A. 1986. Relationships between chlorophyll and ocean color constituents as they affect remote-sensing reflectance models Limnology and Oceanography. 31: 403-413. Carlson, C.A., and Ducklow, H W 1996. Growth ofbacterioplankton and consumption of dissolved organic carbon in the oligotrophic Sargasso Sea. Aquatic Microbial Ecology. 10 : 69-85 Carlson, C.A., Ducklow, H.W and Michaels A.F. 1994 Annual flux of dissolved organic carbon from the euphotic zone in the northwestern Sargasso Sea. Nature. 371:405-408 188

PAGE 202

Carlson, C.A., Ducklow H W., and Sleeter, T.D. 1996. Stocks and dynamics of bacterioplankton in the northwestern Sargasso Sea. Deep-Sea Research IT. 43: 491516. Carpenter E.J. 1983 Nitrogen fixation by Marine Oscillatoria (Trichodesmium) in the world's oceans. p 65-103 In Carpenter, E.J., and Capone, D.G [eds.] Nitrogen in the Marine Environment. Academic Press. Carpenter, E J., and Capone, D.G. 1992. Nitrogen fixation in Trichodesmium blooms p. 211-217. In Carpenter, E.J. Capone, D G. and Rueter, J.G. [eds ], Marine Pelagic Cyanobacteria: Trichodesmium and other Diazotrophs. NATO ASI series ; 362. Kluwer Academic Publishers. Carpenter, E.J., and Romans, K. 1991. Major role of the cyanobacterium Trichodesmium in nutrient cycling in the North Atlantic Ocean Science 254 : 1356-1358. Chen, D. Horrigan, S G ., and Wang, D.P. 1988. The late summer vertical nutrient mixing in Long Island Sound. Journal of Marine Research 46:753-770. Cleveland, J S., and Perry, M.J 1987. Quantum yield, relative specific absorption and fluorescence in nitrogen-limited Chaetoceros gracilis. Marine Biology. 94: 489-497. Codispoti, L.A. 1989. Phosphorus vs. nitrogen limitation of new and export production. p. 377-394. In Berger, W H Smetacek V.S. and Wefer, G [eds.] Productivity of the Ocean: Present and Past. Wiley Coffin, R.B., Connolly, J P and Harris P.S. 1993. Availability of dissolved organic carbon to bacterioplankton examined by oxygen utilization. Marine Ecology Progress Series 101: 9-22 Cuhel, R.L. and Waterbury, J.B 1984. Biochemical composition and short term nutrient incorporation patterns in a unicellular marine cyanobacterium, Synechoccus (WH7803) Limnology and Oceanography. 29: 370-374. Denman, K., Hofmann, E., Marchant, H., Abbott, M.R., Bates, T S Calvert, S.E., Fasham, M.J., Jahnke, R., Kempe, S., Lara, R.J., Law, C.S., Liss, P S., Michaels, A.F., Pedersen, T.F., Pefia, M A., Platt, T ., Sharp, J. Thomas D.N., Scoy, K.A.V., Walsh, J J., and Watson, A.J. 1996 Marine biotic responses to environmental change and feedbacks to climate. p 485-516 In Houghton, J T., Filho, L.G. M., Callander, B.A., Harris, N ., Kattenberg, A., and Maskell, K. [eds ], Climate Change 1995: The Science of Climate Change University Press Deuser, W.G. 1986. Seasonal and interannual variations in deep-water particle fluxes in the Sargasso Sea and their relation to surface hydrography. Deep-Sea Research 33 : 225-246 Doney, S.C., Glover, D.M., and Najjar, R.G. 1996. A new coupled, one-dimensional biological-physical model for the upper ocean: Applications to the JGOFS Bermuda Atlantic Time Series (BATS) site. Deep-Sea Research IT. 43 : 591-624. Droop, M.R. 1973. Some thoughts on nutrient limitation in algae. Journal of Phycology. 9: 264-272 189

PAGE 203

Ducklow, H W., Carlson, C.A., Bates, N.R., Knap, A.H., and Michaels, A F. 1995. Dissolved organic carbon as a component of the biological pump in the North Atlantic Ocean. Philosophical Transactions of the Royal Society, series B. 348: 161-167. Ducklow, H W., Kirchman, D.L., Quinby, H.L., Carlson, C.A., and Darn, H.G. 1993 Stocks and dynamics of bacterioplankton carbon during the spring bloom in the eastern North Atlantic Ocean. Deep-Sea Research IT. 40: 245-263 Eppley, R.W. 1972. Temperature and phtyoplankton growth in the sea. Fishery Bulletin. 70: 1063-1085. Eppley R.W., Garside, C., Renger, E.H., and Orellana, E. 1990. Variability of nitrate concentration in nitrogen-depleted subtropical surface waters. Marine Biology. 107: 53-60. Eppley, R.W., and Peterson, B.J. 1979. Particulate organic matter flux and planktonic new production in the deep ocean. Nature. 282: 677-680. Eppley, R.W. Sharp, J.H., Renger, E.H., Perry, M.J and Harrison, W.G. 1977. Nitrogen assimilation by phytoplankton and other microorganisms in the surface waters of the central North Pacific Ocean. Marine Biology. 39 : 111-120. Falkowski, P G., Dubinsky, Z and Wyman, K. 1985. Growth-irradiance relationships in phytoplankton Limnology and Oceanography. 30: 311-321. Falkowski, P.G. and Wirick, C.D .1981. A simulation model of the effects of vertical mixing on primary productivity. Marine Biology 65 : 69-75 Fasham, M.J.R. Ducklow, G.W., and McKelvie, S.M. 1990. A nitrogen-based model of plankton dynamics in the oceanic mixed layer Journal of Marine Research 48: 591-639. Fasham, M.J.R., Sarniento, J.L., Slater, RD., Ducklow, H.W. and Williams, R. 1993. Ecosystem behavior at Bermuda Station "S" and Ocean Weather Station "India": A general circulation model and observational analysis. Global Biogeochemical Cycles 7 : 379-415. Flynn, K.J. 1991. Algal carbon-nitrogen metabolism : a biochemical basis for modeling the interactions between nitrate and ammonium uptake. Journal of Plankton Research 13: 373-387. Flynn, K.J., Davidson K., and Leftley, J.W. 1994. Carbon-nitrogen relations at whole cell and free-amino-acid levels during batch growth of Isochrysis galbana (Prymnesiophyceae) under conditions of alternating light and dark. Marine Biology. 118: 229-237 Fuhrman, J.A., Sleeter, T D ., Carlson, C.A., and Proctor, L.M. 1989. Dominance of bacterial biomass in the Sargasso Sea and its ecological implications Marine Ecology Progress Series. 57: 207-217. Gargett, A.E. 1984. Vertical eddy diffusivity in the ocean interior. Journal of Marine Research 42: 359-393 190

PAGE 204

Geider, R.J. 1987. Light and temperature dependence of the carbon to chlorophyll a ratio in rnicroalgae and cyanobacteria: implications for physiology and growth of phytoplankton New Phytology. 106 : 1-34. Geider R.J and Osborne, B.A. 1987 Light absorption by a marine diatom: experimental observations and theoretical calculations of the package effect in a small Thalassiosira species. Marine Biology. 96: 299 -308 Geider, R.J Osborne, B.A., and Raven, J.A. 1986a Growth, photosynthesis and maintenance metabolic cost in the diatom Phaeodactylum tricomutum at very low light levels. Journal of Phycology. 22: 39-48. Geider, R.J Platt T., and Raven, J.A. 1986b Size dependence of growth and photosynthesis in diatoms: a synthesis. Marine Ecology Progress Series. 30: 93104. Gieskes, J.M. 1974. The alkalinity-total carbon dioxide system in seawater. p 123-151. In Goldberg, E.D. [ed ], The Sea-Marine Chemistry Volume 5 John Wiley & Sons. Gieskes, W W., and Kraay, G.W. 1986. Analysis of phytoplankton pigments by HPLC before, during and after mass occurrence of the rnicroflagellate Corymbellus aureus during the spring bloom in the open northern North Sea in 1983. Marine Biology 92: 45-52 Glibert, P M., and Ray, R.T. 1990. Different patterns of growth and nitrogen uptake in two clones of marine Synechoco cc us spp. Marine Biology. 107: 273-280 Glover H.E., Campbell, L., and Prezelin B B. 1986. Contribution of Synechococcus spp. to size-fractioned primary productivity in three water masses in the Northwest Atlantic Ocean. Marine Biology 91: 193-203. Glover, H.E., Keller M D., and Spinrad, R.W. 1987. The effects of light quality and intensity on photosynthesis and growth of marine eukaryotic and prokaryotic phytoplankton clones Journal of Experimental Marine Bioligical Ecology 105: 137-159. Goericke, R. and Repeta, D J. 1993 Chlorophylls a and band divinyl chlorophylls a and bin the open subtropical North Atlantic Ocean. Marine Ecology Progress Series 101: 307-313. Goericke, R., and Welschmeyer, N.A. 1993. The marine prochlorophyte Prochlorococcus contributes significantly to phytoplankton biomass and primary production in the Sargasso Sea. Deep-Sea Research 40: 2283-2294. Goering, J.J., Dugdale, R.C., and Menzel, D.W. 1966. Estimates of in situ rates of nitrogen uptake by Trichodesmium sp. in the tropical Atlantic Ocean. Limnology and Oceanography. 11: 614-620. Goldman, J.C. 1980. Physiological processes, nutrient availability, and the concept of relative growth rate in marine phytoplankton ecology. p. 179-194. In Falkowski P.G. [ed.], Primary Productivity in the Sea Plenum Press. 191

PAGE 205

Goldman J C., Caron, D.A., and Dennett M.R. 1987 Regulation of gross growth efficiency and ammonium regeneration in bacteria by subtrate C:N ratio. Limnology and Oceanography 32: 1239-1252 Goldman, J.C., and Glibert, P.M. 1982 Comparative rapid ammonium uptake by four marine phytoplankton species. Limnology and Oceanography. 27 : 814-827. Goldman, J.C., and Glibert, P.M. 1983. Inorganic nitrogen uptake by phytoplankton. p. 233-274. In Carpenter, E.J. and Carpone, D.G. [eds.], Nitrogen in the Marine Environment. Academic Press. Goldman J.C., and McCarthy, J.J. 1978. Steady state growth and ammonium uptake of a fast-growing marine diatom Limnology and Oceanography. 23: 695-703 Gordon, H.R. Brown, O.B., Evans, R H., Brown, J.W., Smith, R.C Baker K.S., and Clark D.K. 1988. A semianalytic radiance model of ocean color Journal of Geophysical Research. 93: 10,909-10,924 Gordon, H.R., and Morel, A. 1983. Remote assessment of ocean color for interpretation of satellite visible imagery, A review. SpringerVerlag Green S.A., and Blough, N.V. 1994. Optical absorption and fluorescence properties of chromophoric dissolved organic matter in natural waters. Limnology and Oceanography. 39: 1903-1916. Gregg, W.W., and Carder, K.L. 1990 A simple spectral solar irradiance model for cloudless maritime atmospheres Limnology and Oceanography. 35: 1657-1675. Gundersen, K.R., Corbin J.S. Hanson, C.L., Hanson, M .L., Hanson, R.B. Russell D.J., Stollar, A., and Yamada, 0. 1976. Structure and biological dynamics of the oligotrophic ocean photic zone off the Hawaiian Islands Pacific Science. 30: 45 68. Guo L., Santschi, P.H., and Warnken, K.W 1995. Dynamics of dissolved organic carbon (DOC) in oceanic environments Limnology and Oceano gra phy. 40: 13921403. Haney, J.D., and Jackson G.A. 1996. Modelin g phytoplankton growth rates. Journal of Plankton Research. 18: 63-85. Hansell, D.A., Bates, N .R., and Gundersen K 1995. Mineralization of dissolved organic carbon in the Sargasso Sea. Marine Chemistry. 51: 201-212. Hansell, D A., and Waterhouse, T Y 1996 Controls on the distribution of organic carbon and nitrogen in the eastern Pacific Ocean. Deep Sea Res e arch (in press). Harri so n, W.G. Harris, L.R. and Irwin, B.D 1996 The kinetics of nitro ge n utilization in the oceanic mixed layer : Nitrate and ammonium interactions at nanomolar concentrations. Limnology and Oc e anography 41: 16-32 Harvey, G R., Boran, D.A ., Che sal, L.A. and Tokar, J.M 1983. The structure of marine fulvic and humic acids. Marine Chemistry. 12: 119-132. 192

PAGE 206

Haxo, F. T. 1985. Photosynthetic action spectrum of the coccolithophorid, Emiliania huxleyi (haptophyceae): 19' hexanoyloxyfucoxanthin as antenna pigment. Journal of Phycology. 21: 282-287. Hayase, K., and Tsubota, H. 1983. Sedimentary humic acid and fulvic acid as surface active substances. Geochimica et Cosmochimica Acta 47: 947-952 Hayase K., and Tsubota, H 1985. Sedimentary humic acid and fulvic acid as fluorescent organic substances Geochimica et Cosmochimica Acta. 49 : 159-163 Hellerman, S., and Rosenstein, M. 1983. Normal monthly wind stress over the world ocean with error estimates Journal of Physical Oceanography 13: 1093-1104. Hochman, H.T., Walsh, J J ., Carder K.L., Sournia, A., and Mtiller-Karger F.E 1995 Analysis of ocean color components within stratified and well-mixed waters of the western English Channel. Journal of Geophysical Research. 100 : 10,777-10,787. Hoepffner, N., and Sathyendranath, S. 1992. Bio-optical characteristics of coastal waters: absorption spectra of phytoplankton and pigment distribution in the western North Atlantic. Limnology and Oceanography 37: 1660-1679 Hoge, F E Vodacek, A., and Blough, N.V. 1993 Inherent optical properties of the ocean: retrieval of the absorption coefficient of chromophoric dissolved organic matter from fluorescence measurements. Limnology and Oceanography. 38: 13941402. Hooks C.E., Bidigare R.R., Keller, M.D ., and Guillard, R.R.L. 1988 Coccoid eukaryotic marine ultraplankters with four different HPLC pigment signature. Journal ofPhycology. 24: 571-580 Hurtt, G C., and Armstrong, R.A. 1996 A pelagic ecosystem model calibrated with BATS data. Deep Sea Research ll. 43: 653-684. IPCC. 1996 Technical Summary of Working Group I Report. p 9-49 In Houghton, J.T., Filho, L.G.M ., Callander, B.A., Harris, N ., Kattenberg, A., and Maskell, K. [eds.], Climate Change 1995: The Science of Climate Change. University Press. Iriarte A., and Purdie, D.A. 1993. Photosynthesis and growth response of the oceanic picoplankter Pycnococcus provasolii Guillard (clone Q48 23) (Chlorophyta) to variations in irradiance photoperiod and temperature. Journal of Experimental Marine Biological Ecology. 168: 239-257 Iturriaga, R., and Siegel, D.A. 1988. Discrimination of the absorption properties of marine particulates using a microphotometric technique. p 277-287 In Blizard, M A. [ed.], Ocean Optics IX, Proc SPIE 925. SPIEThe International Society for Optical Engineering. Jassby, A.D., and Platt, T. 1976 Mathematical formulation of the relationship between photosynthesis and light for phytoplankton. Limnology and Oceanography. 21: 540-547 Jeffrey S W. 1976 The occurrence of chlorophyll Ct and C2 in algae. Journal of Phycology 12 : 349-354 193

PAGE 207

Jeffrey, S.W. and Vesk, M. 1977. Effect of blue-green light on photosynthetic pigments and chloroplast structure in the marine diatom Stephanopyxis turris. Journal of Phycology. 13: 271-279. Jenkins, W.J. 1988. Nitrate flux into the euphotic zone near Bermuda. Nature. 331 : 521-523. Jerlov, N G 1976. Marine Optics. Elsevier Scientific Pub. Co : Jumars, P.A., Penry, D.L., Baross, J.A., Perry, M.J and Frost, B.W. 1989. Closing the microbial loop: dissolved carbon pathway to heterotrophic bacteria from incomplete ingestion, digestion and absorption in animals Deep-Sea Research. 36: 483-495 Kana, T M., Feiwel, N.L., and Flynn, L.C 1992 Nitrogen starvation in marine Synechococcus strains: clonal differences in phycobiliprotein breakdown and energy coupling. Marine Ecology Progress Series. 88: 75-82. Kana, T.M., and Glibert P M 1987a Effect of irradiances up to 2000 J.lE m sI on marine Synechococcus WH7803--I. Growth, pigmentation, and cell composition. Deep-Sea Research. 34: 479-495 Kana, T M., and Glibert P.M. 1987b. Effect of irradiances up to 2000 J.lE m sI on marine Synechococcus WH7803--II. Photosynthetic responses and mechanisms. Deep-Sea Research. 34: 497-516 Kaplan, W.A. 1983. Nitrification. p. 139-190 In Carpenter, E J and Carpone, D.G. [eds ], Nitrogen in the Marine Environment. Academic Press Karl, D.M., Knauer, G.A., and Martin J.H 1988. Downward flux of particulate organic matter in the ocean : a particle decomposition paradox. Nature. 332 : 438-441 Karl, D.M ., Letelier, R., Hebel D., Tupas, L. Dore J. Christian, J and Winn, C 1995 Ecosystem changes in North Pacific subtropical gyre attributed to the 199192 El Nifio Nature 373 : 230-234. Karl, D.M., Letelier, R., Hebel, D V., Bird, D F., and Winn, C.D. 1992. Trichodesmium blooms and new nitrogen in the North Pacific Gyre. p. 219-238 In Carpenter E J ., Capone, D G and Rueter, J.G [eds ], Marine Pelagic Cyanobacteria: Trichodesmium and other Diazotrophs, NATO ASI series; 362. Kluwer Academic Publishers. Keeling, C.D Whorf, T P. Wahlen M. and Plicht, J.w.d. 1995. Interannual extremes in the rate of rise of atmospheric carbon dioxide since 1980 Nature 375: 666-670. Keeling R.F., and Shertz, S .R. 1992. Seasonal and interannual variations in atmospheric oxygen and implications for the global carbon cycle. Nature. 358 : 723-711. Kieber, D.J ., McDaniel, J., and Mopper, K. 1989 Photochemical source of biological substrates in sea water: implications for carbon cycling Nature 341: 637-639. Kieber, R.J ., Zhou X ., and Mopper, K. 1990. Formation of carbonyl compounds from UV -induced photodegradation of humic substances in natural waters: Fate of reverine carbon in the sea. Limnology and Oceanography 35 : 1503-1515. 194

PAGE 208

Kirchman D.L. Keil, R.G., and Wheeler P.A. 1989 The effect of amino acids on ammonium utilization and regeneration b y heterotrophic bacteria in the subarctic Pacific Deep-Sea Research. 36: 1763-1776. Kirchman, D.L., Keil, R.G., and Wheeler, P.A. 1990 Carbon limitation of ammonium uptake by heterotrophic bacteria in the subarctic Pacific Limnology and Oceanography. 35: 1258-1266. Kirk, J T O 1994 Light and Photosynthesis in Aquatic Ecosystems, 2nd ed. University Press . . . Kishino M Okami N Takahashi M and lchimura, S 1986 Light utilization efficiency and quantum yield of phytoplankton i n a thermally stratified sea. Limnology and Oceanography. 31: 557-566. Knap, A.H. Michaels, A F Dow, R.L. Johnson R.J., Gundersen, K ., Knauer, G .A., Lohrenz S E., Asper V .A., Tuel, M Ducklow, H Quinby, H and Brewer, P 1991 U S. JGOFS Bermuda Atlantic Time Series Study Data Report B-1, BATS 1-12, October 1988-September 1989, U.S. JGOFS Planning Coordination Office Knap, A H ., Michaels A F. Dow, R.L., Johnson R.J Gundersen, K., Sorensen, J.C., Close, A.R., Bates, N ., Howse, F Hammer, M ., Best, M Doyle, A.P., Michaels, C., Hansell D Waterhouse, T Kelly, R., Caporelli E ., Bahr, F a nd Little, R. 1995. U S. JGOFS Bermuda Atlantic Time-Series Study Data Report B5, BATS 46-60, October 1992-September 1993, U.S JGOFS Planning Coordination Office. Knap, A H ., Michaels, A.F Dow, R.L. Johnson, R J Gundersen K., Sorensen, J C Close, A.R., Hammer, M., Knauer, G .A., Lohrenz, S.E Asper, V.A Tuel, M ., Ducklow H Quinby, H., Brewer P., and Bidigare R. 1992. U S. JGOFS Bermuda Atlantic Time-Series Study Data Report B-2, BATS 13-24, October 1989-September 1990, U.S JGOFS Planning Coordination Office Knap, A.H., Michaels A F., Dow R.L., Johnson, R.J. Gundersen, K., Sorensen, J.C., Close, A.R., Howse, F., Bates, N Best M Hammer, M and Doyle A.P. 1994 U.S. JGOFS Bermuda Atlantic Time-Series Study Data Report B-4, BATS 37-48, October 1991-September 1992, U.S JGOFS Planning Coordination Office Knap A.H., Michaels A.F., Dow, R.L., Johnson, R.J ., Gundersen K., Sorensen, J C ., Close, A.R., Howse, F. Hammer, M., Bates, N Knauer, G A Lohrenz, S E Asper, V.A., Tuel, M., Ducklow, H and Quinby, H 1993 U.S. JGOFS Bermuda Atlantic Time-Series Study Data Report B-3 BATS 25-36 October 1990September 1991, U S. JGOFS Planning Coordination Office. Knauer G A., and Martin, J .H. 1981. Primary production and carbon-nitrogen fluxes in the upper 1500 M of the Northeast Pacific Limnology and Oceanography 26: 181186 Knauer, G.A ., Martin, J.H., and Bruland, K.W 1979 Fluxes of particulate carbon, nitrogen, and phosphorus in the upper water column of the northeast Pacific. Deep Sea Research. 29: 97-108. 195

PAGE 209

Kwon 0., and Schnoor, J.L. 1994 Simple global carbon model: the atmosphere terrestrial biosphere-ocean interaction. Global Biogeochemical Cycles. 8: 295-305. Laws, E.A., and Bannister, T.T. 1980. Nutrientand light-limited growth of Thalassiosira fluviatilis in continuous culture, with implications for phytoplankton growth in the ocean. Limnology and Oceanography. 25: 457 473. Lawson, L.M., Hofmann, E.E., and Spitz, Y.H 1996. Time series sampling and data assimilation in a simple marine ecosystem model. Deep Sea Research IT. 43 : 625652. Letelier, R.M., and Karl, D.M 1996. Role of Trichodesmium spp. in the productivity of the subtropical North Pacific Ocean Marine Ecology Progress Series. 133: 263273. Lewi s, M R. Harrison, W G ., Oakey, N.S., Herbert D. and Platt T. 1986. Vertical nitrate fluxes in the oligotrophic ocean. Science. 234: 870-873. Li, W K.W., Dickie, P.M. Irwin B D. and Wood, A.M. 1992. Biomass of bacteria, cyanobacteria, prochlorophyte s and photosynthetic eukaryotes in the Sargasso Sea Deep-Sea Research. 39: 501-519. Li, W.K.W., and Wood, A.M 1988 Vertical distribution of North Atlantic ultraphytoplankton: analysi s by flow cytometry and epifluorescence microscopy. Deep-Sea Research. 35: 1615-1638. Lindell, M.J., Graneli W., and Tranvik, L.J. 1995. Enhanced bacterial growth in response to photochemical tran sfo rmation of dissolved organic matter. Limnology and Oceanography. 40: 195-199. Lipschultz F., Zafuiou, O.C., and Ball, L.A. 1996. Seasonal fluctuations of nitrite concentrations in the deep oligotrophic ocean. Deep-Sea Research IT. 43: 403-420. Lohrenz S.E., Knauer, G .A., Asper, V L. Tuel, M Michaels, A.F., and Knap, A H. 1992. Seasonal variability in primary production and particle flux in the northwestern Sargasso Sea: U.S. JGOFS Bermuda Atlantic Time-series Study. Deep-Sea Research. 39: 1373 1391. Longhurst, A.R. 1991. Role of the marine biosphere in the global carbon cycle. Limnology and Oceanography. 36: 1507 1526. Longhurst, A.R. and Harrison W.G 1988. Vertical nitrogen flux from the oceanic photic zone by diel migrant zooplankton and nekton. Deep-Sea Research. 35: 881-889. Malone, T.C., Pike, S.E., and Conley, D.J. 1993. Transient variations in phytoplankton productivity at the JGOFS Bermuda time series station. Deep-Sea Research 40 : 903-924. Marra, J. Bidigare R.R., and Dickkey T.D. 1990. Nutrients and mixing, chlorophyll and phytoplankton g rowth. Deep Sea Research. 37: 127-143. Martin, J H ., Knauer, G.A., Karl, D.M., and Broenkow, W .W. 1987. Vertex: carbon cycling in the northeast Pacific. Deep Sea Research. 34: 267-285. 196

PAGE 210

McCarthy, J.J., and Carpenter E.J 1979 Oscillat o ri a ( Tri c h o desmium) thi e bautii ( Cyanophyta ) in the central North Atlantic Ocean. Journal ofPhycology. 15: 75-82. Mellor, G L. and Yamada, T 1982 Development of a turbulen c e clo s ure model for geophysical fluid problems Review of ge o physical Space Physics 20: 851-87 5 Menzel, D W., and Ryther, J.H 1960. The annual cycle of primary production in the Sargasso Sea off Bermuda Deep-Sea Research 6: 351-367. Michaels A F., Bates, N R., Buesseler, K.O., Carlson, C.A., and Knap A.H. 1994a. Carbon-cycle imbalances in the Sargasso Sea Nature 372 : 537-540. Michaels, A.F ., and Knap, A.H 1996. Overview of the U.S. JGOFS Bermuda Atlantic Time-series Study and the Hydro s tation S Program Deep Sea Research IT. 43: 157-198 M i chaels A.F., Knap A H., Dow, R.L., Gunder s en, K., John s on R.J., Sorensen, J ., Clo s e A ., Knauer G A. Lohrenz, S E. A s per V.A., Tuel M., and Bidigare, R 1994b. Seasonal patterns of ccean biogeochemistry at the U S JGOFS Bermuda Atlantic Times eries Study site Deep-Sea Research. 41: 1013-1038 Michaels, A.F. and Silver, M W. 1988 Primary production, sinking fluxes and the microbial food web. Deep-Sea Research 35 : 473-490. Miller W L., and Zepp, R.G 1995 Photochemical production of dissolved inorganic carbon from terrestrial organic matter : Significance to the oceanic organic carbon cycle Geophysical Research Letters 22: 417-420 Mobley, C.D 1994 Light and Water. Academic Press. Mooney, H A. and Koch, G W. 1994 The impact of rising C02 concentrations on the terrestrial biosphere AMBIO. 23 Moore, B., and Braswell, B.H. 1994 Planetary metabolism : understanding the carbon cycle. AMBIO. 23: 4-14 Moore, L.R., Goericke R. and Chisholm, S W 1995 Comparative physiology of Synechococcus and Prochloroccus: influence of light and temperature on growth pigments fluorescence and absorptive properties Marine Ecology Progres s Series 116 : 259-275 Mopper, K ., and Zhou, X. 1990 Hydroxyl radical photoproduction in the se a and it s potential impact on mar i ne proce ss es Science. 250 : 661-664 Mopper, K., Zhou, X Kieber R.J. Kieber, D.J ., Sikorski R.J., and Jones R D. 1991 Photochemical degradation of dissolved organic carbon and it s impact on the oceanic carbon cycle Nature 353 : 60-62 Morel, A. 1988 Optical modeling of the upper ocean in relation to its biogenou s m a tter content (Case I waters). Journal of Geophysical Research. 93 : 10,749-10,768 Morel A. 1991 Light and marine photosynthesis : a spectral model with geochemi c al and cl i matological implications Progress in Oceanography 26: 263 -306. 197

PAGE 211

Morel, A., Ahn, Y H., Partensky, F., Vaulot, D., and Claustre, H. 1993. Prochlorococcus and Synechococcus: A comparative study of their optical properties in relation to their size and pigmentation Journal of Marine Research 51: 617-649 Morel, A., and Smith, R. 1974. Relation between total quanta and total energy for aquatic photosynthesis Limnology and Oceanography. 19: 591-600. Morris, D.P., Zagarese, H., Williamson, C.E ., Balseiro, E.G., Hargreaves, B.R., Modenutti, B.M Moeller, R. and Queimalinos, C. 1995. The attenuation of solar UV radiation in lakes and the role of dissolved organic carbon Limnology and Oceanography. 40: 1381-1391. Morrow, J.H., Chamberlin, W.S ., and Kiefer, D.A. 1989. A two-component description of spectral absorption by marine particles. Limnology and Oceanography. 34: 1500-1509. Neveux, J., and Lantoine, F 1993. Spectrofluorometric assay of chlorophylls and phaeopigments using the least squares approximation technique. Deep-Sea Research. 40 : 1747-1765. NOAA. 1994. World Ocean Atlas (CD-ROM) U S. Department of Commerce. 0' Brien J.J 1986. The diffusive problem. p. 127-144. In Brien, J.J.O. [ed ], Advanced Physical Oceanographic Numerical Modelling. D. Reidel Publishing Co O'Neil, J.M., and Roman, M.R. 1992. Grazers and associated organisms of Trichodesmium. p 61-73 In Carpenter, E J. Capone, D.G and Rueter, J.G. [eds ], Marine Pelagic Cyanobacteria: Trichodesmium and other Diazotrophs, NATO ASI series; 362. Kluwer Academic Publishers. Oberhuber, J.M. 1988 An atlas based on the 'CO ADS' data set: The budgets of head, buoyancy and turbulent kinetic energy at the surface of the global ocean Max Planck Inst. for Meteorol. Olson, R.J., Chisholm, S.W Zettler, E.R., Altabet, M A., and Dusenberry, J.A. 1990. Spatial and temporal distributions of prochlorophyte picoplankton in the North Atlantic Ocean Deep-Sea Research. 37: 1033-1051. Olson, R.J., Chisholm S W., Zettler, E R., and Armbrust, E.V. 1988 Analysis of Synechococcus pigment types in the sea using single and dual beam flow cytometry. Deep-Sea Research. 35: 425-440. Partensky, F., Hoepffner, N., Li, W.K.W Ulloa, 0., and Vaulot, D. 1993. Photoacclimation of Prochlorococcus sp. (Prochlorophyta) strains isolated from the North Atlantic and the Mediterranean Sea Plant Physiology. 101: 285-296 Peinert, R., Bodungen, B.v., and Smetacek, V.S 1989. Food web structure and loss rate. Wiley Peng, T., Takahashi, T., Broecker, W.S., and Olafsson, J. 1987. Seasonal variability of carbon dioxide, nutrients and oxygen in the northern north Atlantic surface water: observations and a model. Tellus 39B : 439-458 198

PAGE 212

Perry, M.J., Talbot, M.C., and Alberte, R.S. 1981. Photoadaption in marine phytoplankton: response of the photosynthetic unit. Marine Biology. 62: 91-101. Pomeroy L.R., Sheldon, J.E. W M Sheldon, J., and Peters, F. 1995. Limits to growth and respiration of bacterioplankton in the Gulf of Mexico. Marine Ecology Progress Series 117: 259-268. Pope, R .M. 1993. Optical absorption of pure water and sea water using the integrating cavity absorption meter. Ph.D. dissertation. Texas A&M University. College Station, TX. Post, A.F., Dubinsky, Z., Wyman, K., and Falkowski, P.G. 1985. Physiological responses of a marine planktonic diatom to transitions in growth irradiance Marine Ecology Progress Series. 25: 141-149 Post, W.M., Peng, T., Emanuel, W R King, A.W. Dale, V.H., and DeAngelis, D.L. 1990 The Global Carbon Cycle. American Scientist. 78: 310-326 Prezelin, B B., and Glover, H.E. 1991. Variability in time/space estimates of phytoplankton, biomass and productivity in the Sargasso Sea. Journal of Plankton Research. 13: 45-67 Prezelin, B.B., Putt M., and Glover, H.E. 1986. Diurnal patterns in photosynthetic capacity and depth-dependent photosynthesis-irradiance relationships in Synechococcus spp. and larger phytoplankton in three water masses in the Northwest Atlantic Ocean Marine Biology. 91: 205-217. Quay, P.D., Tilbrook, B., and Wong, C.S. 1992. Oceanic uptake of fossil fuel C02 : carbon-13 evidence Science. 256: 74-79. Redfield, A. C 1958 The biological control of chemical factors in the environment. American Scientist. Autumn: 205-221. Richardson, K., Beardall, J., and Raven, J.A. 1983. Adaptation of unicellular algae to Irradiance: an analysis of strategies New Phytologist. 93: 157-191. Riebesell, U., Wolf-Gladrow D A., and Smetacek, V 1993 Carbon dioxide limitation of marine phytoplankton growth rates. Nature 361: 249-251. Roesler, C.S., Perry, M J ., and Carder, K.L. 1989. Modeling in situ phytoplankton absorption from total absorption spectra in productive marine waters. Limnology and Oceanography 34: 1510-1523 Roman, M.R., Dam, H.G., Gauzens, A.L., and Napp, J .M. 1993. Zooplankton biomass and grazing at the JGOFS Sargasso Sea time series station. Deep-Sea Research. 40: 883-901. Roy, S., Harris, R.P., and Poulet, S.A. 1989. Inefficient feeding by Calanus helgolandicus and Temora longicomis on Coscinodiscus wailesii : quantitative estimation using chlorophyll-type pigments and effects on dissolved free amino acids. Marine Ecology Progress Series. 52: 145-153 199

PAGE 213

Sakshaug, E and Andresen, K. 1989 A steady state description of growth and light absorption in the marine planktonic diatom Skeletonema costatum. Limnology and Oceanography. 34: 198-205. Sakshaug, E., Demers, S., and Yentsch, C.M. 1987 Thalassiosira oceanica and T. pseudonana: two different photoadaptational responses. Marine Ecology Progress Series. 41: 275-282. Sakshaug, E., Johnsen G., Andresen, K., and Vernet, M. 1991. Modeling of light dependent algal photosynthesis and growth: experiments with the Barents Sea Diatoms Thalassiosira nordenskioeldii and Chaetoceros furcellatus. Deep-Sea Research. 38: 415-430. Sambrotto, R.N. Savidge, G Robinson, C., Boyd, P., Takahashi, T Karl, D M., Langdon, C., Chipman, D ., Marra, J., and Codispoti L. 1993. Elevated consumption of carbon relative to nitrogen in the surface ocean Nature. 363: 248250. Sarmiento, J L. 1993. Ocean carbon cycle. Chemical and Engineering News special report: 30-43. Sarmiento, J.L., Slater, R.D., Fasham, M.J.R., Ducklow H.W., Toggweiler J.R. and Evans, G.T. 1992. A seasonal three-dimensional ecosystem model of nitrogen cycling in the North Atlantic euphotic zone. Journal of Geophysical Research. : 1-86. Sarmiento, J.L., and Sundquist, E.T. 1992. Revised budget for the oceanic uptake of anthropogenic carbon dioxide. Nature 356 : 589-593 Sathyendranath, S., and Platt, T. 1988. The spectral irradiance field at the surface and in the interior of the ocean : A model for applications in oceanography and remote sensing. Journal of Geophysical Research. 93: 9270-9280. Schlesinger W.H. 1991. BiogeochemistryAn analysis of global change Academic Press. Schluessel P., Emery W J. Grassl, H., and Mammen, T 1990 On the bulk-skin temperature difference and its impact on satellite remote sensing of sea surface temperature. Journal of Geophysical Research 95: 13,341-13,356. Schofield 0., Bidigare R.R., and Prezelin, B B 1990. Spectral photosynthesis quantum yield and blue-green light engancement of productivity rates in the diatom Chaetoceros gracile and the prymnesiophyte Emiliania huxleyi. Marine Ecology Progress Series 64: 175-186 Sellner, K.G. 1992. Trophodynamics of marine cyanobacteria blooms p 75-94 In Carpenter, E J ., Capone, D.G. and Rueter, J.G [eds.], Marine Pelagic Cyanobacteria: Trichodesmium and other Diazotrophs, NATO ASI series; 362. Kluwer Academic Publishers. Shaffer, G 1993. Effects of the marine biota on global carbon cycling. p. 431-455. In Heimann, M. [ed.], The Global Carbon Cycle. Springer-Verlag. 200

PAGE 214

Siegel, D.A., and Michaels, A.F 1996. Quantification of non-algal light attenuation in the Sargasso Sea: Implications for biogeochemistry and remote sensing. Deep Sea Research IT. 43 : 321-345. Siegel D.A., Michaels, A.F Sorensen, J.C., Brien M C.O and Hammer, M.A. 1995. Seasonal variability of light availability and utilization in the Sargasso Sea Journal of Geophysical Research. 100 : 8695-8713. Siegenthaler, U., and Sarmiento, J.L. 1993. Atmospheric carbon dioxide and the ocean. Nature. 365: 119-125. Smith, G J Zimmerman, R.C., and Alberte, R.S 1992 Molecular and physiological responses of diatoms to variable levels of irradiance and nitrogen availability: Growth of Skeletonoma costatum in simulated upwelling conditions. Limnology and Oceanography. 37: 989-1007. Smith, R C ., and Baker, K.S. 1981. Optical properties of the clearest natural waters (200800nm). Applied Optics. 20: 177-184. Spitzer, W.S and Jenkins, W.J. 1989. Rates of vertical mixing gas exchange and new production: estimates from seasonal gas cycles in the upper ocean near Bermuda. Journal of Marine Research 47 : 169 196. Steele, J H and Henderson, E W 1992. The role of predation in plankton models. Journal of Plankton Research 14: 157-172. Sundquist, E.T 1985. Geological perspectives on carbon dioxide and the carbon cycle. p. 5-59. In Sundquist E.T., and Broecker, W.S. [eds.], Carbon Cycle and Atmospheric C02 : Natural Variations Archean to Present, Geophysical Monograph 32. America! Geophysical Union. Sundquist, E T. 1993. The global carbon dioxide budget. Science. 259: 934-941. Suttle, C.A., Chan, A.M ., and Cottrell, M T 1990 Infection of phytoplankton by viruses and reduction of primary productivity Nature 347: 467-469. Takahashi, M., and Bienfang, P .K. 1983 Size structure of phytoplankton biomass and photosynthesis in subtropical Hawaiian waters. Marine Biology. 76 : 203 -2 11. Tans, P.P., Berry J.A., and Keeling, R.F. 1993. Oceanic I3C/I2C observations: A new window on ocean C02 uptake. Global Biogeochemical Cycles. 7: 353-368. Tans, P P Fung, I.Y ., and Takahashi T 1990. Observational constraints on the global atmospheric C02 budget. Science 247: 1431-1438. Taylor A H ., Watson, A.J., Ainsworth, M., Robertson, J.E ., and Turner D.R 1991. A modelling investigation of the role of phytoplankton in the balance of carbon at the surface of the North Atlantic Global Biogeochemical Cycles 5 : 151-171. Taylor G T ., and Karl D.M 1991. Vertical fluxes of biogenic particles and associated biota in the eastern north Pacific : implications for biogeochemical cycling and productivity. Global Biogeochemical Cycles 5 : 289-303. 201

PAGE 215

Vaulot, D. Marie, D. Olson, R.J. and Chisholm, S.W 1995. Growth of Prochlorococcus, a photosynthetic prokaryote, in the equatorial Pacific Ocean Science. 268: 1480-1482. Verity P.O., Robertson, C.Y. Tronzo, C.R., Andrews M.G., Nelson, J.R. and Sieracki M.E. 1992 Relationships between cell volume and the carbon and nitrogen content of marine photosynthetic nanoplankton Liminology and Oceanography . 37 : 1434-1446. Vodacek A., Hoge F.E. Swift, R.N ., Yungel, J.K., Peltzer E T and Blough, N V 1995 The use of in situ and airbourne fluorescence measurements to determine UV absorption coefficients and DOC concentration s in surface waters Limnology and Oceanography. 40: 411-415. Wallace, D.W R., Johnson, K.M., Holfort J. and Schneider B. 1996 Detection of the changing C02 inventory in the ocean. p 31-33 In 1996 U S WOCE Report U S. WOCE Implementation Report, Number 8. U S WOCE Office Walsh J J 197 5. A spatial simulati o n model of the Peru upwelling ecosystem Deep-Sea Research 22 : 201-236 Walsh J J 1996. Nitrogen-fixation within a tropical upwelling ecosystem : E v idence for a Redfield budget of carbon/nitrogen cycling by the total phytoplankton community Journal of Geophysical Research 101: 20,607-20,616 Wal s h J.J. Carder, K.L., and Miiller-Karger F.E 1992. Meridional fluxes of dissolved organic matter in the North Atlantic Ocean. Journal of Geophysical Research 97 : 15 625-15 637. Walsh, J .J., and Dieterle D.A. 1994 C02 cycling in the coastal ocean I-A numerical analysis of the southeastern Bering Sea with a pplications to the Chukchi Sea and the northern Gulf of Mexico Progress in Oceanography 34: 335-392. Walsh, J J., Dieterle, D A., and Pribble, J.R. 1991. Organic debri s on the continental margins : a simulation analys i s of source and fate Deep-Sea Research 38 : 805828. Wanninkhof, R. 1992 Relationship between wind speed and gas exchange over the ocean. Journal of Geophysical Research. 97: 7373-7382 Ward B.B. Olson, R.J and Perry M J 1982 Microbial nitrification rates in the primary nitrite maximum off southern California. Deep-Sea Research. 29: 247-255 Waters K.J. and Smith R.C 1994 Phytoplankton production in the Sargasso Sea a s determined u s ing optical mooring data Journal of Geophy s ical Re s earch 99: 18 385-18,402. Wei ss R.F 1974 Carbon dioxide in water and seawater : The solubility of a nonideal ga s Marine Chemistry 2 : 203-215 Wetzel, R.G., Hatcher P O. and Bianchi T S 1995. Natural photolysis by ultraviolet irradiance or recalcitrant dissolved organic matter to simple substrates for rapid bacterial metabolism Limnology and Oceano g raphy. 40 : 1369-1380. 20 2

PAGE 216

Wheeler P.A. 1983. Phytoplankton nitrogen metabolism. p 309-346. In Carpenter E.J. and Carpone, D G [eds ] Nitrogen in the Marine Environment. Academic Press Wheeler P.A., and Kirchman, D L. 1986 Utilization of inorganic and organic nitrogen by bacteria in marine systems Limnology and Oceanogr a phy. 31: 998-1109. Wilke C.R. and Chang, P. 1955. Correlation of diffusion coefficients in dilute solutions AIChE J 1: 264 270. Wright P B 1988. An atlas based on the 'COADS data set: Fields of mean wind, cloudiness and humidity at the surface global ocean. Max Planck Inst. for Meteorol. Wroblew s ki, J.S 1977. A model of phytoplankton plume formation during variable Oregon upwelling Journal of Marine Research 35 : 357-394 Wroblew s ki, J.S ., Sarmiento J .L., and Flierl G .R. 1988. An ocean basin scale model of plankton dynamic s in the North Atlantic Global Biogeochemical Cycles 2 : 199 218. Yentsch, C.M., Yentsch C.S ., and Perras, J P 1972. Alkaline phophatase activity in the tropical marine blue-green alga Osc illatoria erythaea ('Trichodesmium') Limnology and Oceanography 17: 772-774. Zafiriou, O.C. Ball, L.A., and Hanley, Q. 1992. Trace nitrite in oxic waters Deep-Sea Research. 39 : 1329-1347. Zepp, R.G and Schlotzhauer P F. 1981. Compari s on of photochemical behavior of various humic substances in water: ill. Spectroscopic properties of humic substances Chemosphere 10 : 479-486. 203

PAGE 217

Vita Paul Bissett received a Bachelor s Degree in Finance and Economics from the University of Florida in 1986 He worked in New York City for three years in fixed income portfolio management before enrolling as a graduate student in Marine Science at the University of South Florida in 1991. Paul received a Master's Degree in Marine Science in 1992 In 1993, Paul was a Rotary Ambassadorial Scholar to Australia and attended classes at James Cook University He returned to USF in 1994 to complete his Ph. D He is also senior author on two publ i cations completed during his tenure as a graduate student.


Download Options

No images are available for this item.
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.