Estuaries Vol. 17, No.4, p. 829--838 December 1994
The Relationship Between Primary Production and the Vertical Export of Particulate Organic Matter in a River-Impacted Coastal Ecosystem
DONAL[) G. RI·:D:\I:fE 1 STEVEN E.. LOIIRENZ Cenla for Mmiue Scieuce University of Sou/hem MississijJpi Stennis Space Center, MississijJjJi 39529
GARY L. FAIINENSTIEL National Oceanic and Atmospheric Administration Great Lakes Environmental Research Laboratory Ann AdJm; Michigan 48105
ABSTRACf: As part of the National Oceanic and Atmospheric Administration's (NOAA) Nutrient Enhanced Coastal Ocean Productivity program, we have conducted four research cruises, July-August 1990, March 1991, September 1991, and May 1992, in the Mississippi River plume and adjacent shelf regions. Over this time period, photic-zone-integrated primary production varied significantly in both the river plume and shelf study regions, with greatest variability observed in the river plume region. In the river plume and the adjacent shelf, highest production occurred during July-August 1990 (8.17 g C m-2 d- 1 for the plume and 1.89-3.02 g C m-2 d-1 for the shelf) and the lowest during March 1991 (0.40- 0.69 g C m-• d-1 for the plume and 0.12-0.45 g C m-2 d- 1 for the shelf). The vertical export of POC from the euphotic zone, determined with free-floating MULTITRAP sediment trap systems, also varied temporally in both study regions, with highest values occurring in May 1992 (1.80 ::!: 0.04 g C m-2 d-1 for the plume and 0.40 ::!: 0.02 g C m-• d-• for the shelf) and the lowest values occurring during July-August 1990 (0.29 ± 0.02 g C m-• d-1 for the plume and 0.18 ± 0.01 g c·m-• d- 1 for the shelf). The fraction of production exported out of the photic zone was highly variable and was dependent, in part, on phytoplankton species composition and on the grazing activities of microzooplankton and me sozooplankton. The lowest ratio of export to production coincided with the time when production was greatest and the highest ratios occurred when production was the lowest.
Introduction nutrient concentrations in the river waters. One The Mississippi River drains more than 40% of consequence of river flow into the coastal Gulf of the continental United States. Over the past 35-40 Mexico is the development of stratification, with a yr there has been a twofold increase in the ob lower salinity layer overlying a more saline coastal seawater layer. This stratification and the develop served concentrations of N03- measured in river waters near the modern "birdsfoot" delta (Turner ment of salinity fronts has been related to obser and Rabalais 1991; Dinnel and Bratkovich 1993; vations of primary production with production en Bratkovich et al. 1994). Examination of monthly hanced along these salinity fronts (Lohrenz et al. records have led Dinnel and Bratkovich (1993) to 1990). It has been suggested that the trend of in creasing dissolved inorganic nutrient concentra conclude that seasonal variation in dissolved N03 - concentrations, which are superimposed on a gen tions could result in elevated levels of primary pro erally increasing trend, are linked with seasonal duction in the coastal regions of the Gulf of trends in river discharge. Higher nutrient concen Mexico (Sklar and Turner 19~1; Lohrenz et al. trations are associated with higher river discharge 1990). Further, it has been suggested that the in rates (e.g., in winter and spring). Conversely, low creased levels of primary production would give river discharge appears to be correlated with lower rise to increased sedimentation of particulate or ganic matter (POM) and possibly contribute to the frequently observed episodes of hypoxia on the in 1 Corresponding author. ner Gulf shelf (Turner et al. 1987). One of the
Cl 1994 Estuarine Research Federation 829 0160-83471941040829-10$01.5010 830 D. G. Redalje et al.
30
w 0 ;::J 10 f-o 29 ·--- ·~ f-o '"--- _____ .__...... ___./" - -40 ------./ ,_.-/ 28 l t--/ I 6"""""'------...... ---··- 92 91 90 89 88 LONGITUDE Fig. I. The NECOI> study has examined stations in the Mississippi River plume (filled circles) and in the a 1 diation (PAR), which was measured with a Bio density = 1.08 g kg- , to pre,·ent loss of collected sphet·ical Instruments PNF300 profiling nanu·al materials upon t·ecovery) containing 2% (v/v) for fluorometet- on the day pt·ior to that on which pro malin as a pt·esenrative. The base of each trap tube duction was to be meastu·ed. The three light was fiued with a 0.8-IJ.m Pot·etics filter and a drain depths chosen correspond to the light transmis port. After ,-ecovet·y, the preserved brine solution sion characteristics of the simulated in situ on-deck with the collected POM was drained thmugh this incubation system (Lohrenz et al. 1990, 1992h). filter. The POM collected on the filters was exam Samples were placed in 1-1 polycarbonate bottles, ined microscopically to remove any "swimmer" 14 inoculated with 250 1-1-Ci H C0:1- and incubated zooplankton, which would contaminate the sample ft·om sum·ise to sum·ise. Trace metal clean proce (Karl and Knauer 1989; Knauer et al. 1984; Lee et dures, as recommended by Fitzwater ct al. ( 1982), al. 1988). The collected POM was treated as de were employed in preparing incubation containers S TABLE L Emironmental characteristics of the surface waters patchy nature of the study region is further dem in the Mississippi River plume and adjacent shelf study regions. onstrated by the results of production experiments Ranges are given filr each parameter measured during the se ries of four research cruises conducted as part of the NECOP for the shelf region during Mat-ch 1991 and during program. May 1992 (Fig. 2b,d). On these occasions, IPP var ied by more than a factor of 3 from day 1 to day 2. July-Au~u:-.1 :'1.1.uch SqHt·mbc.-r Rq~ion 19~1(1 19YI 1991 The Yertical export of POC and rON deter Surface Nitrate Concentrations (!LM) mined during the sediment trap deployments in Plume 30-100 10-31 0.05-0.65 16-42 the two study regions exhibited a high degree of Shelf 1.7-23.6 0.4-13.2 <0.02 05-5.0 variability both between the two regions and ovet· Surface Phosphate Concentrations (!LM) time within each •·egion (Table 3). The vertical Plume 0.12-2.15 0.2-0.!1 ().~6-0.9 0.7-{1.1 flux rates of POC out of the photic zone were low Shelf o. 1 Surface Chlorophyll a Concentr·ation (!Lg 1- ) to a much lesser extent than POC export. As was Plume 12.H-14.4 1.1-1.7 1.9-2.1 4.8-10.7 the case for the vertical expon of POC, the lowest Shelf 10.1-11.8 0.3-{1.8 0.03-0.2 2.8-7.3 export of PON was observed during July-August Sm·face Suspended Particulate Matter· (mg 1" 1) 1990 and the highest (approximately twofold Plume 2.4-7.2 0.5-10.3 2.8-3.9 7.6-12.6 greater) during the May 1992 cruise. Shelf 0.9-5.7 0.3-1.8 1.7-4.1 2.2-5.7 Comparison of the vertical export of roc from Average K(PAR) Over Upper 5 m (m- 1) the photic zone in both the river plume and shelf Plume 0.3fHl.84 0.4!Hl.67 0.35-0.47 0.64 regions to measured primary production on each Shelf 0.52-0.64 0.26-0.68 0.06-0.28 0.13-0.41 day of the trap deployments shows that the fraction Surface Salinity of production exported from the surface varied Plume 25.3f>-27.15 12.55-IH.33 24.50-29.67 26.40-30.20 widely over the study period. During July-August Shelf 33.07-3:-1.12 20.64-25.:~6 29.88-32.93 32.18-34.36 1990, a relativley small portion of the Irr (2.9- 9.0%) was collected in the sediment traps, while a much larger portion (64-266%) of the IPP was col meters deeper for the shelf stations. In general, lected in the traps in March 1991. photic-zone-integrated primary production (IPP; It is instructive to examine the relationship be Table 2) rates were up to an order of magnitude tween the export of roc and PON from the pho greater in the river plume than in the adjacent tic zone and the various environmental parame shelf waters. Only during the March 1991 cruise ters, such as IPP, the near-surface salinities and were IPP values comparable in both study regions. concentrations (average values for the upper 1-2 In each study region, IPP varied over time during m) of N03-, P0 4 ~-. Si03-, and chi a, the mass of the field portion of the project. In the river plume, SPM, and ~;a, in the upper 5 m. It is important to IPP ranged from a low value of 0.40 g C m-2 d-1 determine the degree of interrelationship among during March 1991 to a value of8.17 g C m-2 d-1 these data in order to assess the impact of each during July-August 1990, over 20 times greater parameter on the observed rates of primary pro than the lowest observed value. Similarly, IPP val duction and the vertical export of POM to the shelf ues for the adjacent shelf region varied with time sediments. The Pearson Correlation was used to from a low value of0.12 g C m-2 d-1 in March 1991 relate the mean values of the data to the environ to a value 30 times higher, 3.02 g C m-2 d-1, during mental variables presented in Table I. The corre July-August 1990. The <8-jJ.m components of the lation matrix values (n = 8 and df = 6 for each phytoplankton community were responsible for variable) are presented in Table 4. All values of r 50% to 66% of the total IPP in the Mississippi River greater than 0. 707 are considered significant (p < plume and 40% to nearly 100% for the adjacent 0.05; Steel Torrie 1980). IPP is positively correlated shelf waters during the course of the study. The with the near-surface concentrations of N03-, F"~g. 2. Simulated in situ primary production experiments conducted during the research cruises in the Mississippi River plume and the adjacent shelf region. Postincubation size fractionation studies were conducted to determine the production of the <8-JLm components of the phytoplankton community. Symbols used are defined in each figure. Primary production profile$ are given for the July-August 1990 (a), March 1991 (b), September 1991 (c; squares represent data from day 1 and circles represent data from day 2), and May 1992 (d) NECOP cruises. -3 -1 -3 -1 3 1 3 1 Production. mg c m d Produclion. mg C m d Production, mgC m - d- Production, mgC m - d- 0 1 2 3 4 5 0 I 2 3 4 5 0 1000 2000 3000 0 250 500 7150 1000 0 0 0~--~----~--~~ 0 ,....----,...-----.---,---..., • Tolal • Day 2 5 2 2 .,. . ,. Total •b < B I'm E 4 - a a ..ci ~ - ..: c.. 10 :j 4 .... 4 c.. ? 2; 6 - r:l""" r:l""" 15 e e 8 10 20 River Plume Region Hypoxic Shelf Region Hiv<·r Plume Region llypox1c Sh<·lf Rt·~ion (a) USM CMS NECOP July/ August 1990 (b) USM CMS NECOf' 'v!arc II 1991 -3 -I -3 -1 -3 -I -3 -I Prod uclion, mg C m d Production, mg C m d Production, mg C m d Production, mg c m d 0 15 10 115 20 25 0.0 0.2 0.4 0.8 0.8 1.0 0 20 40 eo 80 0 4 8 12 0 0 0 0 /0 5 2· 2 5 10 6 4 6 6 4 6 • Day 1 Total .<::· .ci .... 15 .ci ..ci 10 .... 0 Day I <6 I'm 0...... 0.. Ill 0.. Ill Day 2 Total 0.. Q e Q Ill 8 Day I Total e Q 0 Day 2 <8 I'm Q Day I " e Day I Day I <8 I' 20 0• Day 2 0 Day 2 15 8 8 25 10 30 10 20 River Plume Region Hypoxic Shelf Region River Plume Region Hypoxic Shclr Hegion CIO (d) USM CMS NECOP M~y 1992 w (c) USM CMS NECOP September 1991 w 834 D. G. Redalje et al. TABLE 2. Photic-zone-integrated primary production (IPP; g TABLE 3. Vertical flux of particulate mganic carbon (POC) (g 2 1 C m- d- ) for the simulated in situ primary production expe•·• C m·~ d- 1) and particulate organic niu·ogen (PON) (g N m·~ imenlS conducted on each cruise in both the river plume and d·') out of the photic zone for the four NECOP cruises com adjacent shelf study regions. Values are given for those experi pleted thus far. The st.andard error and number of replicate menL~ associated \\ith the free-floating sediment u·ap array de samples are giwn in parentheses. ploymenlS. In addition IPP •·esuiL~ for the whole phytoplankton population and for the <8-1'-m size fr.Iction determined \\ith Crui-.c.· Plunw Rq;:iorl Shdf Rq.~ion postincubation size fractionation procedures are presented. n.d. July-August I 990 =no dat.a. POC flux 0.29 (0.02; n = 3) 0.18 (0.01; n = 4) lnt(_·gralt·d Priu1ary Producli~lll PON Flux 0.06 (0.003; n = 6) 0.03 (0.002; n = 8) rtume Rq~ion Shdf Rt·giHn :\larch 1991 Cruist· Si03 -, and chi a. The vertical export of POC and PON were significantly correlated. Additionally, both parameters were correlated (p < 0.05) with ave1·age 35-50% of IPP in our study region in the the near-surface concentrations of PO/- and SPM. northern Gulf of Mexico. Our results, however, do N03 - concentrations were correlated significantly not show any clear trends in the relationship be with chi a and with the concentrations of the other tween POC export and IPP. If one compares the 3 nutrients with the exception of P04 -. Using the data presented in Table 2 for IPP and Table 3 for model II geometric mean, least squares linear re the export of POC, it is clear that the ratio of ex gression (Laws and Archie 1981), we also exam port to POC varies from low values of 3-9% during ined the relationship between the export of POC July-August 1990 to high values during March and PON and the mass of SPM in the river plume 1991, when export exceeded measured IPP by a and adjacent shelf dataset. The linear regression factor of about 2. In contrast, there appears to be analysis indicated that both POC and PON export a strong relationship between the input of dis were significantly related to SPM concentrations solved N03 - and Si03 - from the outflow of the (r = 0.80, p < 0.05, Eq. 1 and r2 = 0.80, p < 0.05, Mississippi River and the observed rates of IPP as Eq. 2, respectively). is seen in the significant correlation of these pa rameters (Table 4). The increases in riverborne an POC.,xpon = -0.02 + 0.21[SPM] (1) thropogenically derived nutrients observed over PON.,xpon = -0.04 + 0.03[SPM] (2) the past several decades ( cf. Dinnel and Bratkovich 1993; Bratkovich et al. 1994) appear to have con Discussion tributed significantly to high rates of generation of It appears that the rates of primary production, POM and chi a. It is clear that greater temporal the vertical export of POM from the photic zone, coverage (e.g., more frequent observations) would and the relationship between IPP and the export be necessary before any conclusive relationship be of POM vary over time in both study regions in a tween rates of IPP and POC export can be derived. complex manner. Conventional wisdom suggests Our data are not sufficient to conclude that the that higher levels of IPP should be associated with higher rates of IPP we observed have resulted in a higher rates of vertical export of POM from the greater flux of POM to the shelf sediments· and photic zone (Eppley and Peterson 1979; Walsh et thus, contributed to the periodic episodes of hyp al. 1989). The observations discussed by Eppley oxia. This observation is reinforced by the results and Peterson (1979) were derived, in general, of correlation and regression analyses, which in from both coastal and open ocean environments dicated that the export of POC and PON were cor using short-term measurements of primary and related significantly with both the mass of SPM ~nd new production. The descriptive model presented the near-surface -concentrations of PO,.S- rather by Walsh et al. (1989) for the continental margins than with IPP (Table 4). Indeed, it would be pos in the Gulf of Mexico relies on the conceptual sible to relate the export of both POC and PON framework suggested by Eppley and Peterson to measurements of just SPM in the regions stud (1979). One would expect then, based on these ied (Eqs. 1 and 2). It is also interesting to note that treatments, that the vertical export of POC would near-surface concentrations of PO,.S- were highly ·'... Primary Production and Export Relationships 835 TABLE 4. The Pearson Correlation mau;x for mean values of selected rate processes (integrated primary production, IPP; vertical export of POC, Fro6 vertical export of PON, FroN; surface concenu-;nions of NO,-, ro;-, SiO,- and chi a; mass of suspended particulate matter, SPM; diffuse attenuation coefficient over photosynthetically available radiation. K,.AR; surface salinity, SAL; units are the same as in Table 3). Correlation coefficient values which exceed 0.707 (six degrees of freedom) are considered to be significant at the 5% level. trr FI' correlated to export while the other nutrients mea For this reason, we do not believe that resuspen sured during the study were highly correlated to sion of material presents any significant contribu IPP. Although we do not have sufficient data to tion to our export measurements. make a definitive statement, it does appear that Several researchers have examined the biases of PO/- cycles in the river plume and shelf regions sediment-trap collections in turbulent flow regimes are much different than those for either NO~- or (Butman 1986; Baker et al. 1988). From the results sio3-· of these studies, it has been suggested that flow The use of free-floating sediment traps in a rel velocities near the trap mouth, relative to those of atively shallow coastal system is often viewed with the array, on the order 12 em s-1 or less will not concern (Knauer and Asper 1989). It is possible adversely affect collection efficiency for cylindrical that due to hydrodynamic complications, materials traps such as those employed in this study. How collected in the traps may not accurately reflect ever, flow regimes greater than 30 em s-1 may se either the quality or quantity of the POM sinking riously impact sediment-trap collection efficien to the shelf seafloor. There are several pieces of cies. It is unclear what effect velocities in the 12- supporting evidence derived from other NECOP 30 em s-1 would have on collection efficiencies. investigators that suggest that we did obtain real The results from a related study conducted in the istic samples of sinking POM and can expect that Mississippi River plume and shelf regions may pro the fluxes we calculated reflect flux rates charac vide information on trap efficiencies that would teristic of the study region. An additional concern support the values we obtained in this study (Li et is that the material collected in the traps is more al. in preparation). Li et al. (in preparation) uti representative of resuspended sediments than the lized a ship-mounted acoustic Doppler current material sinking through the water column. The profiler (RD Instruments, 600 kHz) to obtain es stable isotope studies of Eadie et al. ( 1994) suggest, timates of current velocities near the surface and on the basis of ~)ISC studies, that the material col at the depths of the sediment traps. This was ac lected in the traps is marine in origin and is indis complished by conducting a series of transects tinguishable from that suspended in the water col both along and across the path of the trap array umn. It is equally clear that riverborne suspended during deployment. A strong north wind front POM of terrigenous origin does not make a sig passed through the sampling region during their nificant contribution to our trap-<:ollected materi trap deployments resulting in surface current ve al; sediment trap POC &1SC values reflect the char locities that averaged greater than 50 em s-1• Since acteristics of water-column POC in the region of the winds which occurred during the NECOP highest observed IPP. One other concern is that study never reached the velocities measured dur material which has been resuspended from the ing the Li et al. project, it is expected that relative sediments, with presumably similar &1SC character to the NECOP results, this would be a "worst case" istics as exhibited by particles suspended in the wa comparison. For the Li et al. study, the difference ter column, would be included in measurements in current velocities between currents at the depth of trap-<:ollected POC. Although we have not pre of the traps and that of the trap array averaged 20 sented the results here, the mass of POC collected em s-1• Thus the average flow regime at a trap in our deeper traps (see Materials and Methods) mouth during their study was well below the value was always less than that we measured in the traps of 30 em s-1, which is associated with inefficient placed at 15m, near the base of the euphotic zone. trap collections. A second study, conducted in the .' 836 D. G. Redalje et al. same region as our cruises, utilized a similar MUL the sediments: differences in phytoplankton com TITRAP system as was used here (Aspe1· personal munities present, differences in the activities of communication). In the Asper study, an electro both microzooplankton and macrozooplankton magnetic cunent meter (Woods Hole Instrument graze1·s, and horizontal advection of particulates Systems) was placed in-line in the trap array at the into or out of the study 1·egions. same depth as the individual sediment collectors. We do not have any information available on the Results of this deployment indicated that 1·elative advection of water through our study areas. How current flows at the trap depths were on the order ever, we can address the other two potential of 4 em s-I. Weather conditions were simila1· to sow·ces of variability. Olll- data suggest that those those encountered on our series of research cruis >8-J.Lm components of the phytoplankton com es (Asper personal communication). It is likely, munity, primarily diatoms, were of lesser impor theref(lre, that during the studies presented here, tance than smaller organisms in their contribution the current \"clocity at trap mouth depths relative to IPP for the river plume region. During both the to that of the array would be well within the range July-August 1990 and Ma1·ch 1991 cruises, >8-J.Lm whe1·e we would expect no adverse effects on sed size fraction was responsible for mo1·e than half of iment-u·ap collection efficiencies (cf. Knauer and the IPP for the shelf region. However, during the Aspe1· 1989). September 1991 cruise, the <8-J.Lm size f1·action It appears that our July-August 1990 cruise may was responsible for virtually all of the IPP. It is in have been at variance with long-term average con teresting to note that more diatoms, principally ditions for summer (cf. Dinnel and Bratkovich Skeletonenw costatum, were found in the sediment 1993) in that our cruise was preceded by a period traps deployed in the plume than in those de of higher than normal river discharge with con ployed in the shelf region during the July-August centrations of N03 - > 20 J.LM in the river plume. 1990 cruise (Dortch et al. 1992; Fahnenstiel et al. These conditions resulted in a highly stratified wa 1992). It is possible that the lower salinity and ter-column in both the plume and shelf regions, higher nutrient concentrations observed in the and gave rise to the highest values of IPP and the plume by Lohrenz et al. (1992a, 1994) may have lowest ratios of POM export to IPP observed in the contributed to diatoms being more important in study. Some of the environmental conditions mea the material collected in the plume sediment traps sured during the March 1991 cruise were similar than for those deployed in the shelf region. to those measured duringJuly-August 1990 (Table Zooplankton grazing activity is another factor 4). It should be noted that the water column dur that helps explain the observed differences in the ing March 1991 was stratified due to the high river proportion of the POM produced in the photic flow at that time and that incident PAR and surface zone which is exported to the sediments. One concentrations of both SiO,- and chi a were lower would expect that grazing by macrozooplankton than during the previous cruise. During March would be dominant in a coastal environment such 1991, we obtained the lowest values ofiPP and the· as that examined here (Ortner et al. 1989).0n highest ratios of POM export to IPP observed dur each of the cruises, dilution experiments were ing the study. It is clear that the parameters im used to examine the grazing activity of microzoa pacting IPP are different from those leading to the plankton (Landry and Hassett 1982; Dagg and Ort vertical export of POC and PON. ner 1992; Fahnenstiel e~ al. 1992). Results suggest Temporal variation in the irradiance field com that microzooplankton grazing, which resulted in bined with the absorption and scattering proper more tightly coupled production and regeneration ties of the dissolved and suspended materials in the of POM, was more intense during July-August plume and shelf areas may contribute to the ob 1990 than in March 1991. Our POM export results served variation in primary production (Lohrenz support this suggestion in that the fraction of IPP et al. 1990, 1992a, 1994). Yet Lohrenz et al. (1994) exported to the sediments was much lower when were able to show that variability in environmental intense microzooplankton grazing was observed. conditions could only explain part of the observed The results of Benner et al. (1992) and Gardner variation in photosynthetic-irradiance parameters et al. (1994) also support the greater regeneration (i.e., a 8 and P8 max>. The differences in the ratio of of organic matter during July-August 1990 than in the vertical POC flux to IPP cannot be explained March 1991. Their bacterial production studies in solely on the basis of differences in IPP resulting dicate that regeneration rates were an order of from heterogeneity in the physical and chemical magnitude greater in the summer than in the fol environment. It appears that at least three addi lowing winter. Conversely, when maerozooplank tional factors would likely contribute to temporal ton grazing activity was more intense (e.g., March differences in the fraction of the organic matter 1991), rates of bacterial regeneration were much produced in the photic zone which is exported to lower (Benner et al. 1992; Dagg and Ortner 1992). Primary Production and Export Relationships 837 It seems likely that decreased regeneration by bac ducti,;~-. Publication r-.:umber TAl'viU-SG-92-109, Sea Grant teria and lower rates of microzooplankton grazing Program, Texas A&l\1 Uni,·e•-sity, Gal\'eston, Texas. BRATKO\'tcll. A., S. P. Dl:-o:~EL, AND D. A. GOOl$BY. 1994. Vari allowed for a larger portion of the IPP to be ex ability and prediction of riYe•·ine forcing functions for the ported from the photic zone, as is seen in our sed Louisiana-Texas shelf system. Estuaries I 7:766-778. iment trap results. \Ve suspect that for the Septem BUTMAK, C. A. 1986. Sediment trap biases in turbulent flows: ber 1991 and May 1992 cruises, both bacterial Results from a laboratory flume study. jmnnal of A1arine Rr activity and microzooplankton grazing activities search 44:645-69:~. DAGG, M.j. AND P. B. 0RTKER. 1992. Mesozooplankton grazing were moderate because approximately half of the and the fate of cadxm in the nonhe•·n Gulf Mexico, p. 117- POC produced in the photic zone was exported to 121. In Nuuient Enhanced Coastal Ocean Producti,ity. Pub the sediments. This is consistent with the standard lication NumberT:\MU-SG-92-109. Sea Grant Program, Texas viewpoint of export production relative to primary A&M UniYersity. Galveston, Texas. production in coastal regions (Eppley and Peter DtNNEI., S. 1'. A:-o:ll A. HR.HKKOVICII. 199:t Water discharge, ni son 1979; Walsh et al. 1989). trate concentration and niu·ate flux in the lowe•· Mississippi RiYer. jourual of Marini' S)'slrm.< 4:31 f>-326. DoRTCII, Q.. D. ~III$TEn, N. N. RAIIAIAis, S. E. I.OIIRENZ, D. G. Conclusions REDALJE, M. .f. DA<:G, R. E. TuR."ER, AND T. E. Wllrn.ELx:E. Our data from the NECOP program studies in 1992. Role of silicate availability in phytoplankton species composition and the fate of carbon, p. 76--83./n Nutrient En the Mississippi River plume and the inner Gulf of hanced Coastal Ocean Pmductivity. Publication Number Mexico region suggest that during July-August TAMU-SG-92-109. Sea Grant Program, Texas A&M University, 1990, the production and regeneration of POM Galveston, Texas. within the photic zone were tightly coupled, giving EADIE, B.J., B. A. McKEE, M. B. LANSIN<:,j. A. ROUIIINS, S. METZ, rise to a low rate of POM export to the sediments. AND J. H. TREFRE\'. 1994. Records of nutrient enhanced coast al ocean productivity from the Louisiana continental shelf. Conversely, during March 1991, photic-zone pro E .·. 838 0. G. Aedalje et al. mcnt of oceanic particle flux-Arc .. s\\'immcrs" a pn>blem) S11o.w, W. T. AND B. \\'. LlL'M. 1976. Improved exLraction of Oaanogmphy lllagaziur 1:34-36. chlorophyll a and b from algae using dimethyl sulfoxide. l.im l..OHRENZ, S. E., l\1. .J. DAt:G, .\:"D T. E. WtiiTLEDGE. 1990. En •wlog:~· and OaauogmfJh_)· 21 :92(>-92!!. hanced primary production at the plume/oceanic interface SKL·\R, F. II. A:->ll R. E. TURNER. 1981_ Characteristics of phyto of the Mississippi River. Coutiurntal Shelf Rr.