This article was downloaded by: [Florida International University]On: 20 August 2013, At: 02:10Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number:1072954 Registered office: Mortimer House, 37-41 Mortimer Street,London W1T 3JH, UK

International Journal ofRemote SensingPublication details, including instructions forauthors and subscription information:http://www.tandfonline.com/loi/tres20

Dynamics of chromophoricdissolved organic matterand dissolved organiccarbon in experimentalmesocosmsE. J. Rochelle-NewallPublished online: 25 Nov 2010.

To cite this article: E. J. Rochelle-Newall (1999) Dynamics of chromophoricdissolved organic matter and dissolved organic carbon in experimentalmesocosms, International Journal of Remote Sensing, 20:3, 627-641, DOI:10.1080/014311699213389

To link to this article: http://dx.doi.org/10.1080/014311699213389

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of allthe information (the “Content”) contained in the publications on ourplatform. However, Taylor & Francis, our agents, and our licensorsmake no representations or warranties whatsoever as to the accuracy,completeness, or suitability for any purpose of the Content. Anyopinions and views expressed in this publication are the opinions andviews of the authors, and are not the views of or endorsed by Taylor& Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information.Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly inconnection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private studypurposes. Any substantial or systematic reproduction, redistribution,reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of accessand use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

int. j. remote sensing, 1999, vol. 20, no. 3, 627 ± 641

Dynamics of chromophoric dissolved organic matter and dissolved

organic carbon in experimental mesocosms

E. J. ROCHELLE-NEWALL, T. R. FISHER, C. FAN andP. M. GLIBERTUniversity of Maryland Center for Environmental Science,Horn Point Laboratory, PO Box 775, Cambridge MD. 21613, USA;e-mail: rochelle@hpl.umces.edu

Abstract. Since CDOM can signi® cantly in¯ uence the optical characteristics ofCase II waters, we investigated potential sources of CDOM and DOC during amonth-long mesocosm experiment. Nutrient additions resulted in a large phyto-plankton bloom, followed by a rapid decline. Despite this, DOC and CDOMincreased slowly and linearly and were only weakly correlated with Chla. However,there were stronger correlations between CDOM and bacteria, and sedimentaryCDOM ¯ uxes varied from 0 to a value large enough to explain the increase ofCDOM in the tanks. These results suggest that bacteria and sediments may bemore important in CDOM production than phytoplankton on this time scale.

1. Introduction

Chromophoric or coloured dissolved organic matter (CDOM) is the fraction ofthe dissolved organic matter that absorbs light in both the ultraviolet and visibleranges. The presence of CDOM in natural waters has been known since the late1940s, when Kalle (1949 ) ® rst noticed that seawater ¯ uoresced when illuminated byultraviolet light. Since then many studies have been conducted to determine theoptical characteristics (e.g. Coble et al. 1990, Green and Blough 1994) and spatialdistribution of CDOM (e.g. Hoge et al. 1993, Chen and Bada 1992). Also known asgilvin’, gelbsto� ’, or yellow stu� ’, CDOM is ubiquitous to all water bodies. Highconcentrations are found in marshes, lakes, and rivers with lower concentrations inestuaries and coastal regions (Kirk 1994). The lowest concentrations of CDOM arefound in open ocean regions, particularly mid-oceanic gyres such as the SargassoSea (Vodacek et al. 1995).

Because CDOM absorbs both in the visible and ultraviolet ranges, it can representa signi® cant fraction of the total absorption of light in the water column. Forexample, in nearshore waters in August, DeGrandpre et al. (1996) showed theabsorption of CDOM at 442 nm easily exceeded that of particulates, which includedabsorption due to chlorophyll a (Chla). This is in contrast to the situation theyobserved in November, when the particulate fraction was dominant. At their furthesto� shore station they observed a similar season reversal in dominance; however, theabsorption by both components was much lower. This variability means that algo-rithms used to estimate Chla concentration from ocean colour images have to takeinto account not only the presence of CDOM but must also consider the spatial

International Journal of Remote SensingISSN 0143-1161 print/ISSN 1366-5901 online Ñ 1999 Taylor & Francis Ltd

http://www.tandf.co.uk/JNLS/res.htmhttp://www.taylorandfrancis.com/JNLS/res.htm

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.628

and temporal variability in CDOM concentration. In coastal and estuarine systems,where CDOM is higher than in o� shore regions, it is therefore particularly importantto identify the sources and sinks of CDOM in order to assess the processes a� ectingthe variability observed in CDOM concentrations.

Chromophoric dissolved organic matter is a part of the bulk pool of dissolvedorganic matter, usually measured as dissolved organic carbon (DOC). Until recently,DOC was thought to be a large pool with a long turnover time. However, recentdevelopments in methodology have permitted the study of smaller scale dynamicsof DOC cycling (Sugimura and Suzuki 1988, Sharp et al. 1993, 1995 ), and it hasbeen shown that signi® cant fractions of the DOC pool are consumed and producedat relatively short time scales (Kirchman et al. 1991, Carlson et al. 1994).Consequently, there has been a greater emphasis placed on the study of theimportance of various sources and sinks of DOC.

Dissolved organic carbon is produced both autochthonously and allochthonouslyin natural waters. Within the water column, DOC is produced by cellular releaseduring photosynthesis in the form of exudates (Sharp 1977), release of dissolvedcellular contents as a consequence of viral lysis (Fuhrman and Suttle 1993) and fromzooplankton sloppy feeding (Lampert 1978), release from zooplankton faecal matterand marine snow dissolution (Smith et al. 1992), and release during cell death anddecay (Fukami et al. 1985). External sources of DOC include riverine in¯ ows,wetland discharge, release from sediments (Burdige et al. 1992), and airborne inputssuch as in precipitation and dry deposition (Velinsky et al. 1986). However, therehas been very little research linking the in situ dynamics of DOC with those ofCDOM.

There have been several studies of the external sources of CDOM. Concentrationsare usually high in river waters, especially those with high humic and fulvic acidconcentrations derived from soils. In black water rivers, very high CDOM concentra-tions from podzolic soils or peats are responsible for the tea-like colour of the water(Kirk 1994). Other identi® ed sources of CDOM in estuaries includes sedimentaryrelease (Skoog et al. 1996a) and inputs from wet deposition (Chen and Bada 1992).

In contrast, no direct biological sources of CDOM have been clearly identi® ed.In recent work at the Bermuda Atlantic Time Series Station, Nelson et al. (1998)proposed that bacteria are a potential source of CDOM. They hypothesized thatDOC from the spring phytoplankton bloom is processed into CDOM later insummer, resulting in a peak in ¯ uorescence that is decoupled from the peak DOCconcentration (Nelson et al. 1998). Moreover, Carder et al. (1989) also postulatedthat the increased CDOM concentrations that they observed in the Gulf of Mexicomay have been due to the decline of a phytoplankton bloom that had been presentin that region 11 days previously.

In contrast to our lack of information on biological sources of CDOM, a knownremoval mechanism for CDOM is photolysis (Skoog et al. 1996b). ChromophoricDOC is degraded by light into DOC compounds that do not absorb light in theultraviolet and visible range (Keiber et al. 1989, Keiber et al. 1990; see also reviewby Moran and Zepp 1997). Photolysis of CDOM has also been shown as a sourceof inorganic nitrogen to coastal systems (Bushaw et al. 1996), and inorganicphosphate to fresh waters (Franko and Heath 1982).

The aim of this study was to investigate the dynamics of CDOM and DOC in acontrolled mesocosm study. Production of DOC has been successfully studied inmesocosms, usually by observing the changes of DOC concentration following a

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 629

phytoplankton bloom (e.g. Norrman et al. 1995, Smith et al. 1995). Here we extendthis approach to examine the mechanisms by which CDOM may be biologicallyproduced in such closed systems during an intense algal bloom initiated by experi-mental fertilization. This approach allowed us to estimate the time scales upon whichCDOM may be produced in natural systems, which may have important implicationsfor remote sensing of ocean colour. Mesocosms provide a mechanism by whichstudies of environmental ecosystems can be examined on smaller scales in controlledsituations. They provide the investigator with the ability to control and even removevarious factors from the experiment and therefore give the investigator the abilityto isolate parts of the system. While useful for identifying potentially importantmechanisms or processes, mesocosms may also provide information that must becarefully applied to natural systems that are much more complex.

2. Materials and methods

The data presented in this paper were collected during a 28-day experiment inOctober 1996. The mesocosms are part of the Multiscale Experimental EcosystemResearch Center (MEERC) at Horn Point Laboratory, which is funded by US EPA.The goals of MEERC are to understand the principles of scaling in the ecology ofaquatic systems.

Three tanks of identical shape and size were used. The tanks are made of aninert plastic material, with dimensions of 1.13 m diameter and 1 m water depth. Theexterior of each tank is surrounded with a silver foil covered bubble wrap insulationwhich e� ectively stops external radiation from entering the tanks from the side. Thetanks are lit from above with ¯ uorescent lights (Phillips Fluorescent LightF72T12/D/VHO) providing 200 mE mÕ

2 sÕ 1 (or 8E mÕ2 dÕ

1 ) of PAR (photosyn-thetically active radiation, 400± 700 nm); the lights operate on a 12-hour light: 12-hourdark cycle, commencing at 0700. The daily irradiance totals are within the rangeexpected for October in Maryland (2± 41EmÕ

2 dÕ1 ) but are lower than the average

monthly irradiance (26 Ô 9 EmÕ2 dÕ

1 ; Fisher et al. in press). Mixing within thetanks was obtained by mechanically controlled paddles operating at 3 rpm on a 4hour on: 2 hour o� cycle to achieve near-continuous vertical mixing and continuouslyoxic sediment surfaces (Sanford 1997).

Prior to the start of the experiment, a 10 cm deep layer of homogenized sedimentwas placed in the bottom of each tank and allowed to settle for a week. The sedimenthad been initially collected in May 1995 from an adjacent cover in the ChoptankRiver and had been stored in a holding tank except when it was in use in otherexperiments. The older mud was chosen instead of fresh mud for the experimentbecause of its low activity, which results in a small nutrient ¯ ux from the sediment(Cornwell, personal communication).

The tanks were ® lled with un® ltered estuarine water from the Choptank River,a sub-estuary of Chesapeake Bay. On Day 0 of the experiment 1.0m3 of water witha salinity of 9.3 PSU, was carefully added to each of the three tanks to preventsediment disturbance. The water was distributed to the three experimental tankssimultaneously to ensure uniform water characteristics between the three tanks. Forthis experiment, river water had been initially collected and stored in a 10m3 holdingtank for a month prior to ® lling the smaller (experimental) tanks. The water wasstored in order to encourage the natural phytoplankton assemblages to deplete thenutrients present in the water, and the initial concentrations of dissolved nitrogenin each of the tanks were low (NO3= 0.89 Ô 0.13 mM N, NO2= 1.03 Ô 0.07 mM N,

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.630

NH4= 1.23 Ô 0.15 mM N, urea= 0.09 Ô 0.03 mM N). During this time, the holdingtank was subject to both the same mixing and light regimes as the experiment inthe smaller tanks.

The experiment reported here was part of another investigation, the goals ofwhich were to examine the fate of nitrogen from di� erent sources into various sizeranges of the particulate fraction. To address this objective, three tanks and threenitrogen sources (nitrate, ammonium and urea) were used. This also provided theopportunity to investigate the e� ect of nitrogen source on the production of CDOMand DOC in these systems. Nitrate, ammonium, and urea can all be used byphytoplankton as their primary nitrogen source (McCarthy et al. 1977), and onlyone nitrogen source was added to each tank at the beginning of the experiment. Ondays 1, 2, and 3 at 0900, additions were made to each tank to increase concentrationsby 5mM N, and on day 4 a large pulse of nitrogen (25 mM N) was added to eachtank in the respective nitrogen forms. On day 4, a pulse of phosphate (1mM PO4 )was also added to relieve any phosphate limitation that might be occurring in thetanks with the large additions of nitrogen.

Water column samples were removed daily (0800 ) for the measurement of dis-solved organic carbon (DOC), chromophoric dissolved organic matter (CDOM),chlorophyll a, nutrients, and bacterial number. On several days a more frequentsampling interval was adopted and samples were collected every four hours for aminimum of twenty four hours.

The water samples for all of the DOC/CDOM measurements were ® lteredthrough GF/F glass ® bre ® lters in an all-glass, precleaned, ® ltration ¯ ask. Pre-cleaning of all glassware consisted of acid washing (10% HCL and copious rinseswith deionized water), followed by combustion at 450ß C for 1 hour; plastic wasavoided whenever possible. Samples were stored in glass pre-cleaned bottles, sealedwith Te¯ on lined caps, and frozen. The samples were kept frozen until the DOC andCDOM analyses were performed.

Prior to the DOC and CDOM measurements, the samples were defrosted atroom temperature. Dissolved organic carbon concentrations were measured usingan adaptation of the persulphate method (Sharp 1973, Sharp et al. 1995). For theanalysis, 10ml of sample was pipetted into an ampule, acidi® ed with 100 ml 10%H2SO4 and gas stripped with oxygen for 10 mins. After initial CO2 removal, 100 mgpotassium persulphate were added as the oxidizing agent, and the ampules were gasstripped for a further 30 secs to remove any CO2 introduced to the sample with theoxidizing agent. The ampules were then sealed under continuous ¯ ow of oxygen gasand immediately autoclaved for 55 minutes to convert organic carbon into CO2 gas.During later analysis to measure the CO2 derived from DOC oxidation, the ampuleswere broken in air and 7 ml of liquid was carefully removed into a 20ml syringe.Then 0.25 ml of 10% H2SO4 was carefully added to the syringe without introducingair, followed by 7 ml of He gas. The syringe was shaken for 1 min to equilibrate theCO2 between the gas and liquid phases, and the headspace gas was then injectedinto a gas chromatograph (Hach Carle Series 100 AGC) connected to a Hewlett-Packard 3393A Integrator. The CO2 gas produced by DOC oxidation was calibratedagainst sucrose standards of known concentration ranging over 0± 600 mM DOC.The sucrose standards were prepared in deionized water that had been previouslyautoclaved with potassium persulphate (10 g in 1 l DI water) and ampulatedas outlined above for samples. The autoclaved DI water was also used as theexperimental blank and was usually equivalent to 20± 30 mM DOC.

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 631

The ¯ uorescence of the chromophoric fraction was measured as soon as possibleafter the DOC analysis and always within 24 hrs. Prior to the analysis, the sampleswere stored in the dark in a refrigerator, and immediately prior to the opticalmeasurements the samples were allowed to slowly warm to room temperature (20ß C).The ¯ uorescence of the samples was measured on an Aminco-Bowman Series 2Luminescence Spectro¯ uorometer using a 1cm quartz cell. An excitation wavelengthof 355 nm was used, resulting in a water Raman response at 404 nm and a CDOMemission centered around 450 nm. Milli-Q water was used as a blank for the ¯ uores-cence measurements. All sample CDOM ¯ uorescence intensities were normalized tothe Raman and to an external standard, 10 mg lÕ 1 quinine sulphate in 10% H2SO4

(Green and Blough 1994), which is equivalent to ten normalized ¯ uorescence units(NFIU, Hoge et al. 1993).

Concentration of Chla of the samples was measured using the ¯ uorometricmethod of Parsons et al. (1984) using excitation and emission wavelengths of 450 nmand 670 nm respectively. Brie¯ y, duplicate volumes of water were ® ltered onto GF/Fglass ® bre ® lters, and the ® lters were frozen until analysis. The ® lter was ground andChla extracted in 90% acetone and the ¯ uorescence was measured using a ¯ uoro-meter (Turner Designs Fluorometer 10), calibrated with commercially available Chlastandard (Sigma Chemical Co.).

Water samples for the ambient nutrient concentrations were ® ltered throughWhatman GF/F glass ® bre ® lters. The ® ltrate was frozen for later determination ofnitrate (NO3 ), nitrite (NO2 ), ammonium (NH4 ), urea and total dissolved nitrogen(DON). Prior to analysis samples were defrosted at room temperature. Nitrate,nitrite and urea were measured colorimetrically using the methods of Parsons et al.(1984), ammonium was measured using the protocol of Solorzano (1969), and totaldissolved nitrogen was measured on an Antek DON Analyzer.

The acridine orange direct count (AODC) procedure was used for bacteriacounting. Water samples (5 mls) were preserved with glutaraldehyde, dyed with 80%acridine orange, then counted using epi¯ uorescence microscopy (Hobbie et al. 1977).

Sediment cores were taken from the tanks at the end of the experiment to measurethe ¯ uxes of both DOC and CDOM from the sediments. Two cores were collectedfrom each tank and were incubated in benthic ¯ ux chambers (Cowan and Boynton1996). Each chamber had an internal diameter of 10 cm and contained a 10 cm deepsediment core overlaid with 20cm of ® ltered Choptank River water. The benthic¯ ux chambers were sealed and incubated for 48 hours. The chambers were subjectedto a 12-hour light: 12-hour dark cycle for 48 hours and were constantly stirred bymagnetic stirrers within the chambers. Water samples were collected for DOC andCDOM every 12 hours. A blank core of ® ltered river water was also run in parallelwith the sediment cores.

We used both correlation and regression statistical tests to evaluate the statisticalrelationships in this research. We used both Pearsons’ Product Moment correlationfor the variables that were normally distributed and Spearmans Rank OrderCorrelation for the variables that failed a normality test (Sigma Stat, Jandel Scienti® cSoftware). Linear regressions were used to evaluate the time course data and thesediment ¯ ux data. All statistical relationships are reported as not signi® cant (NS,p> 0.05), signi® cant (*, 0.05 > p> 0.01), or highly signi® cant (**, 0.01 > p).

3. Results

A large phytoplankton bloom occurred in all three tanks following the nutrientadditions (® gure 1). The stepwise nutrient additions increased the original back-

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.632

Figure 1. Time course of Chla (panel a), concentration of added nitrogen nutrient (panel b),and bacterial number (panel c). Bacterial number increased exponentially with time(r2= 0.60**). Filled circles= NO3 enriched tank, open circles= NH4 enriched tank,® lled triangles= urea enriched tank.

ground concentration of 1 mM N NO3 , NH4 , or urea and resulted in a peak nutrientconcentration of about 30± 40 mM N in the tanks on days 3± 5 (® gure 1 (b)).Concentrations declined rapidly between days 5± 10 and were relatively low there-after. The highest phytoplankton biomass occurred on day 5 (70± 110 mg Chla lÕ 1 ;® gure 1 (a)) and then fell quite rapidly through day 10, apparently in response to theremoval of nitrogen from the water (® gure 1 (b)). Thereafter, Chla remained between5± 20 mg Chla lÕ 1 for the duration of the experiment. Bacterial numbers (® gure 1)increased exponentially from 0.67 to 3.57 Ö 106 mlÕ 1 towards the end of the experi-ment (r2= 0.60**). These data are similar to bacterial numbers obtained from similarexperiments done previously in these tanks (e.g. Bryant 1995).

No signi® cant di� erences were noted between the three tanks in the positiveexponential rise to the peak in algal biomass, in the exponential fall to low biomasslevels, nor in the relatively constant biomass during days 10± 28 (ANOVA). Becausethere was no signi® cant di� erence between the three tanks, we have treated the threetanks as replicate mesocosms with equal N enrichments in subsequent data analyses.

Both CDOM and DOC increased linearly over the course of the experiment

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 633

(r2= 0.86** and r2= 0.27**, respectively, ® gure 2 (a,b)). As in ® gure 1, there were nosigni® cant di� erences in the CDOM or DOC production rate between the threedi� erent nitrogen sources, and all data were pooled to obtain single net rates ofproduction for both CDOM and DOC. CDOM, measured as NormalizedFluorescence Units (NFIU), increased steadily over the course of the experiment ata rate of 0.30 NFIUdÕ

1 . There was no peak in CDOM production that correspondedwith the peak and crash of the phytoplankton bloom (® gure 1). DOC concentrationalso increased steadily at a rate of 0.174mmol C lÕ 1 dÕ

1 , also with no clear relationto the phytoplankton bloom and crash.

There were weak but signi® cant relations between CDOM and Chla (r2= 0.09**,® gure 3 (a)) and DOC and Chla (r2= 0.18**, ® gure 3(b)). However, it is likely thatthese weak statistical results are caused by the large number of data points at lowerChla concentrations and re¯ ects no functionally signi® cant relations. Contrasting® gures 1 and 2, there was no obvious connection between the bloom dynamics(® gure 1(a)) and the steady increase in CDOM and DOC (® gure 2).

Bacterial abundance was more strongly related to CDOM. There was a signi® c-ant, positive correlation between CDOM and bacterial number (r2= 0.40**,

Figure 2. Changes in CDOM (A) and DOC (B) throughout the experiment. Legend as in® gure 1.

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.634

Figure 3. Relation between CDOM, DOC and Chla. CDOM and DOC are weakly inverselycorrelated with Chla concentrations (r2= 0.09**, r2= 0.18**, respectively). Legend as in ® gure 1.

® gure 4(a)), again with no signi® cant di� erence between N sources. However, therewas no relation between DOC and bacterial number (r2= 0.01, NS, ® gure 4(b)), andthere were no signi® cant di� erences in the mean DOC values between N sources.Overall, DOC concentrations were quite high (mean= 362 mM C), increased slowlyduring the experiment (® gure 2(b)) were unrelated to bacterial numbers (® gure 4(b)),and were inversely related (possibly spuriously) to phytoplankton abundance(® gure 3(b)).

In all but one sediment core there were no signi® cant ¯ uxes of CDOM or DOCeither into or out of the sediment (® gures 5 and 6). Blank cores (® ltered water)showed no signi® cant changes in DOC or CDOM over the course of the incubationand are omitted for clarity. One core from the nitrate tank exhibited a signi® cant¯ ux of CDOM (r2= 0.90*) out of the sediment at a rate that increased concentrationin the overlying water by 1.14 Ô 0.22 (s.e.) NFIUdÕ

1 .

4. Discussion

There is a paucity of information available on studies of the biological productionof CDOM and the time scales over which this production occurs. Some studies haveobserved correlations between nutrient pro® les and CDOM concentration in oceanicand coastal environments (Hayase et al. 1988, Momziko� et al. 1992). For example,

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 635

Figure 4. Relation between CDOM, DOC, and bacterial population size. Bacterial numberand CDOM are positively correlated (r2= 0.40**, panel a), but DOC and bacterialnumber are not (r2= 0.01NS, panel b). The mean DOC concentration is 362 mM.(c) Legend as in ® gure 1.

there are strong correlations between apparent oxygen utilization (AOU), nutrientpro® les, and CDOM ¯ uorescence pro® les in the deep waters of the Central Paci® c(Hayase and Shinozuka 1995). The higher concentrations of CDOM in the deeperwaters probably arise from the bacterially mediated regeneration of particulatematter in deeper waters combined with photolysis of CDOM in surface waters,resulting in a pro® le characteristic of nutrients such as phosphate or nitrate (e.g. lowconcentrations at the surface and higher concentrations in the aphotic zone).Therefore, it seems probable that deep water production of CDOM is at leastpartially bacterially mediated. In contrast to this, Momziko� et al. (1992), workingin the Ligurian Sea region of the Mediterranean, have hypothesized that the increasesin CDOM they observed were of both phytoplanktonic and zooplanktonic origin.They observed strong correlations between the position of the peaks in Chla andCDOM in their pro® les. The correlation was strongest in the region farthest fromshore, and was less well de® ned in the stations nearer shore. The position of theCDOM peaks and the position of a scattering layer caused by zooplankton, whichwere coincidental in many of their pro® les, may indicate a zooplanktonic origin ofCDOM (Momziko� et al. 1992).

In this study, the relation between algal biomass and CDOM concentration was

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.636

Figure 5. Sediment core incubations for measurements of DOC ¯ uxes. No signi® cant ¯ uxeswere observed for any of the sediment cores, and there were no signi® cant di� erencesbetween the blank cores (® ltered river water, data not shown) and the duplicatesediment cores from any of the tanks. Filled circles= sediment core a, open circles=sediment core b.

very weak. This is perhaps surprising as there was a large phytoplankton bloom(> 70 mg Chla lÕ 1 ) followed by a rapid crash. Despite these large changes in phyto-plankton abundance there was no clear, functional relation between algal biomassand CDOM concentration. While there were weak negative correlations betweenChla, CDOM, and DOC which were statistically signi® cant, these relations areprobably not functionally signi® cant, and in any event suggest an inverse relationbetween these three parameters. This seems to indicate that phytoplankton, at leastin this type of experiment, are not a direct source of CDOM during the relativelyshort time scale of the experiment (28 days). Similarly, there was no clear relationbetween the crash of the phytoplankton bloom and the appearance of DOC. Thisis in contrast to the results of Norrman et al. (1995) who observed an increase inDOC in their mesocosm experiment following the crash of the phytoplankton bloom.In a culture experiment, Chen and Wangersky (1993) also observed increases inDOC concentration over the twelve days of their experiment. Using the data ofNorrman et al. (1995 ), we calculated a DOC increase of 3.16 mM C per unit changein Chla (mg lÕ 1 ) during their experiment. If an equivalent release had occurred in

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 637

Figure 6. Sediment core incubations for measurements of CDOM ¯ uxes. Only one coreexhibited a signi® cant ¯ ux of CDOM out of the sediment (r2= 0.90*), in all othercores including the blanks (® ltered river water, data not shown) no signi® cant ¯ uxwas observed in either direction. Legend as ® gure 5.

our experimental tanks, we would have expected an increase in DOC of 245 mM Cin each of the tanks. However, during the entire period, the increase in DOC overbackground that we observed in the tanks was 108 mM C. It is not clear why therewere di� erences between the data presented here (no DOC increase directly relatedto the crash of the bloom) and those reported by others. There was considerableday to day variability in DOC concentration in the tanks, which, combined with ahigh background DOC concentration, may partially explain our lack of an observableto observe increase in DOC directly related to the crash of the bloom. However, ourdata do not support the concept of a DOC pulse resulting from the crash of aphytoplankton bloom under our conditions.

There was also an apparent lack of ¯ uxes of both CDOM and DOC either intoor out of the sediments, except in one core from the nitrate enriched tank. The mudsurface and overlying water were oxic at all times throughout the experiment, aswas also the case during the ¯ ux chamber experiments. This may be importantbecause Skoog et al. (1996a) have reported that they failed to see ¯ uxes of CDOMout of sediment cores from the Baltic until oxygen de® cient conditions were present

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.638

in the overlying water. This may mean that the sediment CDOM/DOC ¯ uxes weobserved may represent a lower ¯ ux than what would be observed in a naturalsystem with seasonal anoxia.

However, for the one core that did have a positive ¯ ux of CDOM out of thesediment, we have calculated the impact of this on the increases of CDOM observedin the tanks. Remembering that 1 NFIU is equivalent to the ¯ uorescence of 1mg mÕ

3

quinine sulphate (in order to manipulate units), then we can estimate the ¯ ux ofCDOM out of the core (F, mg quinine sulphate unitsmÕ

2 dÕ1 ), using

F = (d [CDOM]/dt )z (1)

where t= time in days and z= water depth in metres above the sediment (Fisheret al. 1982). The volumetric change in CDOM (1.14 NFIUdÕ

1 ) in the one core witha signi® cant ¯ ux into the overlying water (z= 20cm) was used to compute a ¯ ux of0.23 mg quinine sulphate units mÕ

2 dÕ1 . We can similarly use the observed concentra-

tion change in CDOM in the mesocosm (0.30 NFIUdÕ1 ) to compute a ¯ ux, assuming

all CDOM originates from the sediments in the mesocosm. Given the average waterdepth of 0.75 m in the mesocosms, this rate of change of CDOM is equivalent to a¯ ux of 0.22 mg quinine sulphate unitsmÕ

2 dÕ1 , essentially equivalent to the one

observed ¯ ux from the nitrate mesocosm. Thus, unfortunately our sediment ¯ uxexperiments gave results that ranged from 0 (5 of 6 cores) to a rate that can explainthe observed rate of change of CDOM in the water column of the mesocosm (1 of6 cores). It is likely that our core incubations did not have enough data pointsand/or sensitivity to resolve the CDOM ¯ uxes from the sediments, and the role ofCDOM ¯ uxes from sediments remains unresolved by our data.

It is also interesting to note that there was no concomitant ¯ ux of DOC out ofthe sediment for any core, including the one with the signi® cant ¯ ux of CDOM. Itis likely that the CDOM measurements were able to detect smaller changes thanthe DOC measurements, as the precision of the ¯ uorescence measurement is muchbetter than that for the DOC measurements. The depth of water column overlyingthe sediment within the cores was also quite large, which may explain why we failedto see any other signi® cant ¯ uxes of DOC or CDOM from any of the other cores.Furthermore, the high background DOC and CDOM precludes highly sensitivemeasurements of CDOM and DOC ¯ uxes.

The one statistically signi® cant ¯ ux represents an upper limit of what could haveoccurred in the tank during the experiment. Spatial heterogeneity of sediments is tobe expected, and it is possible that the one measurable ¯ ux did not exist all over thesurface of the mud in the tanks. However, using the data presented here, we cannotrule out the possibility that the sediments have the potential to be a signi® cant, andperhaps the dominant source of CDOM to the overlying water.

The bacterial biomass data is similar to that from previous experiments in thesetanks (Bryant 1995). The increase in bacterial number towards a stable populationsize may have been a consequence of either resource or grazing limitation, as alsonoted by Bryant (1995 ), who observed an increase in heterotrophic nano¯ agellatesthat are known to graze on bacteria. The relationship between the bacterial biomassand the increase in CDOM concentration is signi® cant and positive (® gure 4). Themechanisms by which CDOM is associated with bacteria are unclear from thisexperiment, but it is possible that the bacteria are a direct source of CDOM, eitherby direct release of CDOM, or as a by-product of some other process. It has beenpreviously shown that bacteria may be integral in the production of CDOM (Nelson

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 639

et al. 1998). Such evidence as the increased concentrations of CDOM in deeperwaters mirroring the nutrient pro® les in the deep ocean supports this view.

The relative importance of sediments and bacteria as sources of CDOM is di� cultto separate from this experiment. The shallow depth of the water column undoubtedlyin¯ uences the outcome and increases the importance of sediment-derived inputs.However, in natural systems with a deeper water column, bacterially derived produc-tion of CDOM is likely to be more important. Moreover, if the relationship betweenbacterial biomass and CDOM concentration proves to be a functionally importantone in natural systems, this will be evidence to support the importance of bacterialprocesses in the production of CDOM. It is clear from these data, however, thatphytoplankton blooms do not appear to be a large source of CDOM on the timescale of approximately one month.

Acknowledgments

We wish to thank Dr N. V. Blough (Department of Chemistry and Biochemistry,University of Maryland), for the use of his ¯ uorometer. We would also like to thankMichael Lomas, Teresa Coley and Jennifer Zelenke for their assistance. This studywas funded by grants from the EPA Multiscale Experimental Ecosystem ResearchCenter (R819640), Maryland Sea Grant (SC3527606N) to PMG, and NASA Missionto Planet Earth (LCLUC-0024) to TRF. This is contribution number 3058 from theUniversity of Maryland Center for Environmental Science.

References

Bryant, A. L., 1995, E� ects of mesocosm dimension on the factors regulating heterotrophicbacterioplankton dynamics. MS. Thesis, University of Maryland, USA.

Burdige, D . J., Alperin, M . J., Homstead, J., and Martens, C. S., 1992, The role of benthic¯ uxes of dissolved organic carbon in oceanic and sedimentary carbon cycling.Geophysical Research L etters, 19, 1851± 1854.

Bushaw, K . L., Zepp, R. G ., Tarr, M . A., Shulz-Jander, D ., Bourbonniere, R. A., Hodson,R. E., M iller, W . L., Bronk, D . A., and Moran, M . A., 1996, Photochemical releaseof biologically available nitrogen from aquatic dissolved organic matter. Nature ,381, 404± 407.

Carder, K . L., Steward, R. G ., Harvey, G . R., and Ortner, P. B., 1989, Marine humicand fulvic acids: Their e� ects on remote sensing of ocean chlorophyll. L imnology andOceanography, 34, 68± 81.

Carlson, C. A., Ducklow, H . W ., and M ichaels, A. F., 1994, Annual Flux of dissolvedorganic carbon from the euphotic zone in the Northwestern Sargasso Sea. Nature ,371, 405± 408.

Chen, R. F., and Bada, J. L., 1992, The ¯ uorescence of dissolved organic matter in seawater.Marine Chemistry, 37, 191± 221.

Chen, W ., and Wangersky, P. J., 1993, High-temperature combustion analysis of dissolvedorganic carbon produced in phytoplankton cultures. Marine Chemistry, 41, 167± 171.

Coble, P. G ., Green, S. A., Blough, N . V., and Gagosian, R. B., 1990, Characterization ofdissolved organic matter in the Black Sea by ¯ uorescence spectroscopy. Nature , 348,

432± 435.Cowan, J. L., and Boynton, W . R., 1996, Sediment-water oxygen and nutrient exchanges

along the longitudinal axis of Chesapeake Bay: Seasonal patterns, controlling factors,and ecological signi® cance. Estuaries, 19, 562± 580.

DeGrandpre, M . D ., Vodacek, A., Nelson, R. K ., Bruce, and Blough, N . V., 1996,Seasonal seawater optical properties of the U.S. Middle Atlantic Bight. Journal ofGeophysical Research, 101, 22727± 22736.

Fisher, T. R., Carlson, P. R., and Barber, R. T., 1982, Sediment nutrient regeneration inthree North Carolina estuaries. Estuarine and Coastal Shelf Science, 14, 101± 116.

Fisher, T. R., Gustafson, A. B., Sellner, K ., Lacouture, R., Haas, L. W ., Wetzel, R. L.,

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

E. J. Rochelle-Newall et al.640

Magnien, R., Everitt, D ., M ichaels, B., and Karrh, R., in press. Spatial andtemporal variation of resource limitation in Chesapeake Bay, Marine Biology.

Franko, D . A., and Heath, R. T., 1982, UV-sensitive complex phosphorus: Association withdissolved humic material and iron in a bog lake. L imnology and Oceanography, 27,

564± 569.Fuhrman, J. A., and Suttle, C. A., 1993, Viruses in marine planktonic systems. Oceanography,

6, 51± 63.Fukami, K ., Simidu, U ., and Taga, N ., 1985, Microbial decomposition of phyto- and

zooplankton in seawater: 1. Changes in organic matter. Marine Ecology ProgressSeries, 21, 1± 5.

Green, S. A., and Blough, N . V., 1994, Optical absorption and ¯ uorescence propertiesof chromophoric dissolved organic matter in natural waters. L imnology andOceanography, 39, 1903± 1916.

Hayase, K ., and Shinozuka, N ., 1995, Vertical distribution of ¯ uorescent organic matteralong with AOU and nutrients in the equatorial central Paci® c. Marine Chemistry,48, 283± 290.

Hayase, K ., Tsubota, H ., Sunada, I., Goda, S., and Yamazaki, H ., 1988, Vertical distributionof ¯ uorescent organic matter in the North Paci® c. Marine Chemistry, 25, 373± 381.

Hobbie, J. E., Daley, R. J., and Jasper, S., 1977, Use of nucleopore ® lters for countingbacteria by ¯ uorescence microscopy. Applied and Environmental Microbiology, 33,

1225± 1228.Hoge, F. E., Vodacek, A., and Blough, N . V., 1993, Inherent optical properties of the ocean:

Retrieval of the absorption coe� cient of chromophoric dissolved organic matter from¯ uorescence measurements. L imnology and Oceanography, 38, 1394± 1402.

Kalle, K ., 1949, In T he ¯ uorescence of dissolved organic matter in the sea, by E. K. Duursma1974. In Optical Aspects of Oceanography, edited by N. G. Jerlov and E. SteemannNielsen. (New York: Academic Press), pp. 237± 256.

Keiber, D . J., McDaniel, J., and Mopper, K ., 1989, Photochemical source of biologicalsubstrates in sea water: implications for carbon cycling. Nature , 341, 637± 639.

Keiber, R. J., Zhou, X., and Mopper, K ., 1990, Formation of carbonyl compounds from u.v.induced photodegradation of humic substances in natural waters: Fate of riverinecarbon in the sea. L imnology and Oceanography, 35, 1503± 1515.

K irk, J. T. O., 1994, L ight and Photosynthesis in Aquatic Ecosystems, 2nd edn. (Cambridge:Cambridge University Press).

K irchman, D . L., Suzuki, Y., Garside, C., and Ducklow, H . W ., 1991, High turnover ratesof dissolved organic carbon during a spring phytoplankton bloom. Nature , 352,

612± 614.Lampert, W ., 1978, Release of dissolved organic carbon by grazing zooplankton. L imnology

and Oceanography, 24, 831± 834.McCarthy, J. J., Taylor, W . R., and Taft, J. L., 1997, Nitrogenous nutrition of the

phytoplankton in the Chesapeake Bay. 1. Nutrient availability and phytoplanktonpreferences. L imnology and Oceanography, 22, 996± 1011.

Momzikoff, A., Dallot, S., and Pizay, M-D., 1992, Blue and yellow ¯ uorescence of ® lteredseawater in a frontal zone (Ligurian Sea, northwest Mediterranean Sea). Deep SeaResearch, 39, 1481± 1498.

Moran, M . A., and Zepp, R. G ., 1997, Role of photoreactions in the formation of biologicallylabile compounds from dissolved organic matter. L imnology and Oceanography, 42,

1307± 1316.Nelson, N . B., Siegel, D . A., and M ichaels, A. F., 1998, Seasonal dynamics of colored

dissolved material in the Sargasso Sea. Deep Sea Research, 45, 931± 957.Norrman, B., Zweifel, U . L., Hopkinson, C. S., and Fry, B., 1995, Production and utilization

of dissolved organic carbon during an experimental diatom bloom. L imnology andOceanography, 40, 898± 907.

Parsons, T. R., Maita, Y., and Lalli, C. M ., 1984, A Manual of Chemical and BiologicalMethods for Seawater Analysis, 1st edition (New York: Pergamon Press).

Sanford, L. P., 1997, Turbulent mixing in experimental ecosystem studies. Marine EcologyProgress Series, 161, 265± 293.

Sharp, J. H ., 1973, Total organic carbon in seawaterÐ Comparison of measurements usingpersulphate oxidation and temperature combustion. Marine Chemistry, 1, 211± 229.

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013

Estuarine and coastal processes 641

Sharp, J. H ., 1977, Excretion of organic matter by marine phytoplankton: Do healthy cellsdo it? L imnology and Oceanography, 22, 381± 399.

Sharp, J. H ., Suzuki, Y., and Munday, W . L., 1993, A comparison of dissolved organiccarbon in North Atlantic Ocean near shore waters by high temperature combustionand wet chemical oxidation. Marine Chemistry, 41, 253± 259.

Sharp, J. H ., Benner, R., Bennet, L., Carlson, C. A., Fitzwater, S. E., Peltzer, E. T.,and Tupas, L. M ., 1995, Analyses of dissolved organic carbon in seawater: the JGOFSEqPac methods comparison. Marine Chemistry, 48, 91± 108.

Skoog, A., Hall, P. O. J., Hulth, S., Paxeus, N ., Rutgers van der Loeff, M ., andWesterlund, S., 1996a, Early digenetic production and sediment-water exchange of¯ uorescent dissolved organic matter in coastal environment. Geochimica etCosmochimica Acta, 60, 3619± 3629.

Skoog, A., Wedborg, M ., and Fogelqvist, E., 1996b, Photobleaching of ¯ uorescence andthe organic carbon concentration in a coastal environment. Marine Chemistry, 55,

333± 345.Smith, D . C., Simon, M ., Alldredge, A. L., and Azam, F., 1992, Intense hydrolytic enzyme

activity on marine aggregates and implications for rapid particle dissolution. Nature ,359, 139± 142.

Smith, D . C., Steward, G . F., Long, R. A., and Azam, F., 1995, Bacterial mediation ofcarbon ¯ uxes during a diatom bloom in a mesocosm. Deep Sea Research, 42, 75± 97.

Solorzano, L., 1969, Determination of ammonia in natural waters by the phenolhypochloritemethod. L imnology and Oceanography, 14, 16± 23.

Sugimura, Y., and Suzuki, Y., 1988, A high temperature catalytic oxidation method for thedetermination of non-volatile dissolved organic carbon in seawater by direct injectionof a liquid sample. Marine Chemistry, 24, 105± 131.

Velinsky, D . J., Wade, T. L., and Wong, G . T. F., 1986, Atmospheric deposition of organiccarbon to Chesapeake Bay. Atmospheric Environment, 20, 941± 947.

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 airborne ¯ uorescence measurements to determine UVabsorption coe� cients and DOC concentrations in surface waters. L imnology andOceanography, 40, 411± 415.

Dow

nloa

ded

by [

Flor

ida

Inte

rnat

iona

l Uni

vers

ity]

at 0

2:10

20

Aug

ust 2

013