sign in sign up
<<
Home , Seagrass meadow

University of South Carolina Scholar Commons

Faculty Publications Biological Sciences, Department of

11-3-1999

Nutrient Cycling in the Water Column of a Subtropical

Susan Ziegler University of Texas at Austin

Ronald Benner University of South Carolina - Columbia, [email protected]

Follow this and additional works at: https://scholarcommons.sc.edu/biol_facpub

Part of the Biology Commons

Publication Info Marine Ecology Progress Series, ed. Otto Kinne, Volume 188, 1999, pages 51-62. © Marine Ecology Progress Series 1999, Inter-Research.

This Article is brought to you by the Biological Sciences, Department of at Scholar Commons. It has been accepted for inclusion in Faculty Publications by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. MARINE ECOLOGY PROGRESS SERIES Published November 3 Mar Ecol Prog Ser

Nutrient cycling in the water column of a subtropical

Susan Ziegler*, Ronald Benner**

Marine Science Institute, University of Texas at Austin, 750 Channelview Drive, Port Aransas, Texas 78373, USA

ABSTRACT: The* cycling of was studied over a 16 mo period to determine how processes occurring between the water column and benthos influenced nutrient dynamics in a tes- tudinum dominated seagrass meadow. Nutrient concentrations were low and ranged from below detection to 0.59 pM ammonium (NJ&*), 0.04 to 0.29 pM plus nitnte (NO3- + NO<), and below detection to 0.22 phl soluble reactive phosphate (SRP).Water column and benthic fluxes of NO< +NO2- and SRP were usually below detection. The benthic fluxes of NH,' ranged from an uptake of -228 pm01 N m-' d-' to a release of 363 pm01 N m" d-l. Positive fluxes (i.e,directed out of the sediment) occurred primarily in light incubations and from seagrass-dominated sediments. Water column fluxes of NH,+ ranged from a net uptake of -145 pm01 N m d-' to a net regeneration of 643 pm01 N m-2 d-l. Thenet regeneration of NH,' in the water column usually exceeded the release of NH,' from the benthos. There was a significant correlation between the regeneration of NH4+m the water column and the light-mediated release of dissolved organic matter (DOM) from the benthos, indicating that benthic- derived DOM supported the regeneration of NH,' in the water column. Bacterioplankton growth effi- c~encieswere significantly and positively correlated to the regeneration of NH4+in the water column, possibly resulting from changes in the composition of DOM released from the benthos. The C:N ratios of the organic matter remineralized in the water column were variable and ranged from 14 to 81, with lowest values occurring in late summer and highest values in spring. The results of this study indicated that temporal variations in the source and composition of DOM significantly influenced the cycling of nutrients in the water column of this seagrass meadow.

KEYWORDS: Nutnent cycllng . . Benthic fluxes . Ammonium . Dissolved organic matter

INTRODUCTION nutrient enrichment has been reported throughout the (Orth & Moore 1983). Nutrient en- Seagrass-dominated are among some of the richment has been found to increase epiphytic growth most productive on earth, supporting and on seagrasses and possibly trigger algal blooms, both serving as nurseries for commercially important causing dramatic reductions in seagrass populations and feeding grounds for waterfowl (Hillman et al. (Kemp et al. 1983, Borum 1985, Flores-Verdugo et al. 1989). Worldwide degradation of seagrass is 1988, Dunton 1990). Increases in dissolved inorganic well documented (Dennison et al. 1993), with losses (DIN),possibly due to increases in agricultural attributed to increased human activities in coastal practices as well as a massive kill, may have been areas. For example, the coincidence of areas of greatest responsible for the onset of a brown tide in northern reduction in aquatic grasses with areas of greatest Laguna Madre, Texas, in June 1990 (Whitledge 1993). This brown tide has been implicated in about a 20% reduction of the seagrass Halodule wrightii between Present addresses: 1992 and 1995, and has dramatically increased the 'Carnegie Institution of Washington, Geophysical Labora- to seagrass production ratio in northern tory, 5251 Broad Branch Rd. NW, Washington. DC 20015- 1305, USA. E-mail: [email protected] Laguna Madre (Stockwell et al. 1993, Onuf 1996). "Department of Biological Sciences. University of South To help determine the influence of environmental Carolina, Columbia, South Carolina 29208, USA changes, such as increases in nutrient loading, on -

O Inter-Research 1999 Resale of full article not permitted 52 Mar Ecol Prog Ser 188: 51-62, 1999

grass ecosystems there is a need to understand the bio- soluble reactive phosphate (SRP) from June 1996 to geochemical cycling of nutrients within these systems. June 1997. These nutrient parameters were examined Nutrient dynamics in seagrass meadows and other in relation to measurements of phytoplankton and bac- shallow environments is typically dominated by ben- terioplankton production, benthic production, respira- thic processes. Because concentrations of nutrients are tion, and dissolved organic fluxes. low in the water column, seagrasses most likely obtain their nutrients from the sediments where concentra- tions are considerably higher (Iizumi et al. 1982, Short MATERIALS AND METHODS & McRoy 1984, Blackburn et al. 1994). Rapid recycling of nitrogen (N) in the sediments has been documented Site description. Laguna Madre is the southernmost in seagrass systems and appears to occur more rapidly located on the Texas , and it is separated in sediments dominated by seagrasses versus those from the Gulf of Mexico by Padre Island. Our study that are unvegetated (Boon et al. 1986). Iizumi et al. was conducted in a dominated (1982) and Short (1983) estimated that about 28 and seagrass meadow in the southern portion of Laguna 60%, respectively, of the ammonium (NH,') regener- Madre. Beds of and Halodule ated in the sediment of seagrass beds was released to wrightii as well as drift algal species such as Digenia the overlying waters. Remineralization of nutrients and simplex and Laurencia poitei also occur in southern nitrogen fixation in the benthos appear to be fueled by Laguna Madre (Humm & Hildebrand 1962, Onuf seagrass photosynthate released through the 1996).The study site was located east of the Gulf Intra- (Jerrgensen et al. 1981, Moriarty et al. 1986, Blackburn coastal Waterway at about 26" 10' N 97" 12' W. (For et al. 1994, Stapel & Hemminga 1997).The release and map see Herzka & Dunton 1996.) A total of ten 5 day remineralization of seagrass photosynthate in the ben- trips were made to southern Laguna Madre approxi- thos represents a possible source of inorganic N and P mately every 6 to 8 wk from February 1996 to June in the overlying water column in these ecosystems. 1997. Data from 2 of these 10 trips are not presented A significant fraction of the carbon fixed by sea- here due to brown tide blooms which interfered with grasses is released as dissolved organic matter (DOM) the collection of nutrient data. into the water column through leaching and exudation Water column parameters. Water temperature and and is subsequently utilized by heterotrophic bacterio- depth were recorded using a data logging system plankton (Moriarty & Pollard 1982, Moriarty et al. 1986, (YSI6000, YSI Inc.) secured at mid-depth (-0.5 m) for Chin-Leo & Benner 1991, Ziegler & Benner 1999).For the duration of each trip. Average NH4+,NO,- + NO,-, example, water column respiration accounts for about and SRP concentrations were determined from 12 to 24 60% of benthic net in a Thalassia samples collected over a 24 h period using an autosam- testudinum meadow of Laguna Madre (Ziegler & Ben- pler (see Ziegler & Benner 1998) Samples were col- ner 1999). The activity of benthic and macro- lected in 45 m1 glass bulbs (acid washed) at 60, 70 or phytes has been found to stimulate bacterioplankton 120 min intervals from about 0.6 m depth. Sample production and nutrient cycling in shallow estuaries bulbs were kept in the dark and collected from the where dissolved organic nitrogen appears to be the sampler every 12 h around sunrise and sunset. The major form of N released from the benthos (Moriarty et 45 m1 samples were transferred from the bulbs into al. 1986, Middelboe et al. 1998). 125 m1 high-density polyethylene bottles (acid washed) The influence of the benthic release of DOM on bac- and immediately placed on ice. Samples were frozen terioplankton activity has been observed, however, within 4 h of collection. All nutrient samples were ana- very little is known about its influence on nutrient lyzed within 4 wk of collection. cycling in seagrass ecosystems. How does the coupling Seagrass carbon and nitrogen. Whole live plants of processes between the water column and the ben- were collected using a 15 cm diameter corer at the thos influence the cycling of nutrients in these eco- study site in July, September, and November. The systems? Does the remineralization of DOM serve as a cores were sieved with water to remove live plants major source of inorganic nitrogen or ? If from the sediment. The plants were rinsed with seawa- so, is this source influenced by seasonal variations in ter and then kept on ice until they were brought back the source of DOM such as seagrass exudation and to the lab. The tissues were rinsed with distilled water leaching? The purpose of this study was to address and separated into above-(blades) and below-ground these questions by determining the and (roots and ) tissues. Once separated, the tis- cycling of nutrients in the water column of lower sues were placed in a drying oven at 45'C for 48 h. The Laguna Madre. Water column concentrations were dried tissues were milled, and carbon and nitrogen investigated, as well as benthic and water column content were determined on a Carlo Erba EA 1108 fluxes of NH,', nitrate plus nitrite (No3-+ NO2-),and elemental analyzer. Ziegler & Benner: Nutrient cycling in a seagrass meadow water column 53

Water column production, respiration, and nutrient triplicate live sample set to determine abiotic sorption fluxes. Water was collected around dawn and incu- of the labeled leucine. Incubations were terminated by bated in light and dark bottles (300 m1 biological oxy- filtration through a 0.2 pm pore-sized MF Nucleopore gen demand bottles or 90 m1 quartz bottles) for about membrane filter. Filters were immedately extracted 12 h during the day to estimate primary production, with ice cold 5% trichloroacetic acid for 5 min, fol- respiration, and net fluxes of nutrients in the water col- lowed by a 5 m1 rinse with ice-cold trichloroacetic acid, umn. Another water sample was collected around sun- stored in scintillation vials and refrigerated until mea- set and incubated overnight in dark bottles to estimate surement of radioactivity within 7 d of collection. respiration and nutrient fluxes at night. Dark bottles Prior to scintillation counting samples were extracted were wrapped in aluminum foil, and all bottles were at 50°C for 1 h using the tissue solubilizer Solvable incubated in situ on racks set at mid-depth (-0.5 m) in (Dupont, New England Nuclear Inc.) as described by the water column. To estimate primary production and Amon & Benner (1998). Rates of bacterial carbon pro- respiration, changes in dissolved oxygen were deter- duction were estimated from leucine incorporation mined in light and dark incubations using Winkler rates using conversion factors of 4.3 X 1016 cells mol-l, titrations with an automated titrator and potentiometric derived from experiments using Laguna Madre water endpoint detection (Biddanda et al. 1994). Rates of net (Chin-Leo & Benner 1991), and 20 fg C cell-' (Lee & O2 production or consumption were determined from Fuhrman 1987). the slope of the least-square linear regression (n 2 6) of Bacterial growth efficiencies were calculated as the the dissolved oxygen data versus time. Consumption of ratio of bacterial production to bacterial production dissolved oxygen in dark incubations during the day- plus water column respiration, except in March 1997. time were added to the value for light incubations to During March 1997 primary production was relativ- obtain estimates of gross primary production. Water ely high due to the presence of a brown tide. Respira- column gross primary production and respiration rates tion due to bacterioplankton, at this time, was as- based on O2 were converted to units of carbon using sumed to be 70% of the total plankton respiration. and respiration quotients of 1.2 and 1.0, This assumption was based on previous measure- respectively (Oviatt et al. 1986). ments of bacterial respiration ranging from 40 to 100% Rates of net nutrient production or consumption of plankton respiration for different marine environ- were determined from the slope of the least-square ments (Williams 1981, Chin-Leo & Benner 1992, Bid- linear regression analysis of nutrient concentration danda et al. 1994). versus time (n Z 6). Initial and final samples were dis- Benthic nutrient fluxes. Net fluxes of NH,', No3- + pensed into clean 125 m1 polyethylene bottles, imme- NO2- and SRP into and out of the benthos were deter- diately placed on ice and frozen within 4 h of collec- mined by measuring changes in nutrient concentra- tion. The daily net flux of nutrients occurring in the tions in light and dark chambers. Chambers were con- light was based on the number of daylight hours. The structed from 20 1 Nalgene polycarbonate carboys by number of hours of light for each day was based on the removing the bottoms and adding a sampling port at number of hours of saturating PAR light for seagrasses the shoulder. Caps of the chambers were fitted with a (-300 ~.IEm-' S-'; Herzka & Dunton 1996) determined current-driven stirring mechanism to mimic in situ from underwater continuous PAR llght data collected water movement (see Ziegler & Benner 1998). Covers close to the study site (K. Dunton, J. Kaldy & J. Kowal- for the chambers were constructed of dark-gray plastic ski unpubl. data). The daily net dark flux was calcu- and used for the dark incubations. Four chambers were lated from the number of hours below saturating light placed carefully about 8 cm into the sediment at 3 sea- levels. Daily net regeneration of NH,' in the water col- grass-dominated sites (primarily Thalassia testudinum) umn (dark incubations) was calculated from the sum of and 1 unvegetated area (void of all macrophytes) adja- the water column fluxes estimated for daytime and cent to the other chambers. An average of the 4 cham- nighttime. Positive fluxes indicated net regeneration bers was used to estimate net benthic nutrient fluxes in and negative fluxes indicated the net uptake of the lower Laguna Madre. The distribution of chambers nutrient. was based on estimates of the aerial coverage of sea- Bacterioplankton production. Bacterioplankton pro- grass-dominated (75 %) and unvegetated (25 %) sedi- duction was estimated from protein synthesis using ments in lower Laguna Madre (Quammen & Onuf 3H-leucine (Kirchman et al. 1985). Triplicate 20 m1 1993). Positive flux values indicated a net flux from the samples were incubated in the dark with 3H-leucine sediments to the water column, and negative fluxes (20 nM final concentration, specific activity of 50 Ci indicated a net flux from the water column into the mmol-', New England Nuclear) for 1 h in a circulating sediments. bath of Laguna Madre water. One formaldehyde (4% Benthic chambers were carefully deployed without final concentration) killed control was run with each caps the day prior to the benthic measurements. At 54 Mar Ecol Prog Ser 188: 51-62, 1999

dawn, water in the chambers was exchanged with Concentrations of NH,' were determined using the overlying water using a hand-held pump. Dark incu- phenol-hypochlorite method (Solorzano 1969). The de- bations were conducted early in the morning and late tection limit was 0.30 pM at a confidence level 289%. in the afternoon to avoid dramatic shifts in light levels. The precision of the NH,+ analysis at 0.3 pM was 15 % lncubations lasted for 1.5 to 4 h depending upon level (expressed as coefficient of variance). Concentrations of activity (i.e. incubations were the longest in winter of NO3- + NOz- were determined for 10 m1 samples and shortest in summer). Samples were collected for using the standard cadmium reduction method modi- nutrient analysis through the sampling port using a fied for small samples (Gardner et al. 1976). Imidazole 60 m1 syringe, dispensed into 125 m1 high-density poly- buffer was used in place of the NH,C1 in both the addi- ethylene bottles and treated as described for the die1 tions to the sample (0.05 M, pH = 7.5) and through the water column samples. After completion of all benthic column itself (0.1 M, pH = 7.5) in order to avoid NO3- incubations, volumes for all 4 chambers were deter- contamination often found in reagent grade NH4Cl. mined by injecting each with 15 m1 of 30 mM NaN03. Standards of KN03 and NaN03 were analyzed before Each chamber was stirred for 2 min before a sample and after every set of samples. The detection limit for was collected and later frozen. The concentration of NO3- + NO,- was 0.02 pM, and precision was 40 and NO3- in these samples was used to estimate the volume 4% at 0.02 and 1.00 PM, respectively. Samples col- of each chamber during every trip. Nutrient fluxes lected in June 1997 were analyzed for NO3- + NOz- measured in the water column were used to correct the using an Antek Model 745 Nitrate/Nitrite Reduction benthic nutrient fluxes for changes in nutrient concen- Assembly and Antek Model 7020 nitric oxide chemilu- trations due to processes occurring in the water con- minescence detector. The precision for l nM NO3- + tained in each chamber. The daily net benthic flux of NOz- was ~6%using chemiluminescence detection. nutrients occurring in the light was based on the num- SRP was determined using the standard colormetric ber of hours of saturating PAR light levels for sea- method (Strickland & Parsons 1972). The detection grasses, and the daily net fluxes occurring at night limit was 70 nM and the precision was l % at 70 nM. were calculated from the number of hours below satu- rating light levels. Nutrient analyses. All sanlples collected for nutrient RESULTS analyses were unfiltered to avoid additional steps that could have caused contamination in these san~ples Water column nutrient concentrations with low concentrations of nutrients. In March, how- ever, all samples were passed through a GF/F filter Nutrient concentrations in the water column of immediately after collection because of the unusually southern Laguna Madre were low throughout the high phytoplankton production caused by a brown tide year as compared with most estuarine systems, and bloom that occurred in the previous months. There was relative to most seagrass systems (Fig. 1) Consider- no difference between 0.2 pm (polycarbonate) filtered ably higher nutrient concentrations have been re- water and whole water collected in April, June, July, ported in other tropical and temperate seagrass September and November with respect to NH4', NO3- meadows (Pederson & Borum 1993, Johnson & John- + NOT concentrations (Table 1). There was a differ- stone 1995, Stapel et al. 1997). NH,' concentrations in ence in SRP concentration between the 0.2 pm filtered lower Laguna Madre ranged from ~0.30to 0.59 pM water and whole water collected in April; however, and h7ere highest in the late summer (Table 2, Fig. 1). SRP was usually below detection in both the filtered NO3- + NO2- concentrations ranged from 0.04 to and unfiltered water (Table 1). 0.18 pM, with the lowest concentrations in the sum-

Table 1. Average nutrient concentrations (pM) of whole and filtered (0.2 pm) water samples. Each value is reported as the mean (n 5 3) i 1 SD. SRP: soluble reactive phosphorus

Month NH,' NO3- + NOz- SRP Whole water 0.2 ].]m Whole water 0.2 pm Whole water 0.2 pm

APd 0.22 t 0.08 0.23 + 0.05 0.07 t 0.01 0.04 * 0.02 0.10+0 02 <0.07 June 0.21 * 0.03 0.26 t 0.03 0.12 + 0.03 0.14 i 0.02

IIIIIII ,IlIIII - c -Ammonium -Nitrate & Nitrite

Mar May Jul Sep Nov Jan Mar

Fig. 1. Mean ammonium, nitrate + nitrite, and soluble reactive Fig. 2. Die1 water column ammonium concentrations collected phosphate concentrations in the water column at the study at the study site on November 5-6, 1996 and March 8-9, 1997 site from February 1996 through June 1997. Error bars repre- sent il SD

Seagrass carbon and nitrogen content mer and fall. SRP was generally below detection limit and ranged from <0.07 to 0.17 PM. NO3- + NO2-, and The C and N content of above- and below-ground SRP concentrations peaked in March when phyto- Thalassiatestudinum tissues collected in July. Sep- plankton production was the highest due to a brown tember, and November indicated that there were tide. The concentration of SRP remained high in large differences in the N content of the seagrass tis- June 1997 when phytoplankton production remained sues between the summer and fall (Table 3). The C higher than normal. There was no apparent die1 pat- and N content decreased slightly from July to Sep- tern in water column nutrient concentrations except tember in both the above- and below-ground tissues. in March 1997, when NH,' concentrations increased Molar C:N ratios of the above- and below-ground tis- during the day and decreased at night (Fig. 2). Con- sues collected in July and September were similar. centrations of SRP in Florida Bay (annual average of The C:N ratios were much lower in November due to 0.03 PM) were similar to those in lower Laguna a higher N content in both the above- and below- Madre, whereas DIN concentrations (annual average ground tissues. The N content of the seagrasses col- of 2.7 PM) were generally higher (Fourqurean et al. lected in November were 28 and l l l % higher in the 1993). The ratios of DIN to SRP in lower Laguna above- and below-ground tissues, respectively, than Madre were below the of 16:l (Redfield those collected in September. The higher N content 1958) during the spring and winter (Table 2). From of the tissues in November suggests that N storage in April through November 1996, SRP was not detect- the fall could help support rapid growth in the spring able and thus only minimal values for the ratio of DIN (Pedersen & Borum 1993). to SRP could be calculated.

Table 2. Mean water temperature, amrnomum (NH,'), nitrate plus nitrite (NO3- Water column productioli + NO2-), and soluble reactive phosphorus (SRP) concentrations, and the ratio of and respiration dissolved inorganic nitrogen (DIN) to SRP. All errors are reported as i1 SD Water column gross primary produc- Month Temperature NH4+ NO3' + NO< SRP D1N:SRP tion was generally low, ranging from ("c) (PM) (PM) (PM) Ratio 0.4 to 9.2 mm01 C m-2 d-l, except dur- February 22.1k1.6 041i0.09 0.2920.04 0.17i0.01 4.11 ing January 1997, when it was as high April 22.7 * 1.4 <0.30 0.07 +0.05 <0.07 >5.29 as 35.5 mm01 C m-' d-' due to a brown June 30.2 ? 0.6 <0.30 0.13 +0.02 ~0.07 >4.86 tide bloom (Table 4).The water column July 30.7 + 0.1 0.43 + 0.12 0.04 i 0.03 <0.07 >6.71 was almost always net heterotrophic, September 30.5 + 0.1. 0.59 k 0.15 0.04 i 0.00 ~0.07 >9.00 with respiration ranging from 6.6 to November 23.9 + 1.7 0.31 i 0.08 0.06 i 0.01 <0.07 >5.29 March 20.7 * 1.5 0.37 20.16 0.18i0.15 0.13 i0.07 4.23 27.0 mm01 C m-2 d-' (Table 4). Bac- June 27.3 * 1.5 0.37 i 0.17 0.06 i 0.03 0.12 i 0.03 3.58 tenoplankton production in the water column ranged from 1.8 to 9.1 mm01 C 56 Mar Ecol Prog Ser 188: 51-62, 1999

Table 3. Carbon (C) and nitrogen (N) content of live seagrass when net regeneration of NH,' (76 pm01 N m-2d-' or blades and roots and rhizomes. The carbon to nitrogen ratios 0.10 pm01 N 1-' d-') was measured in the light (Table 3, (C:N)are reported as the molar ratio Total plant C:N is calcu- Fig. 3). Bacterioplankton N demand, calculated from lated assuming 85% of the total plant biomass 1s in below- ground tissues (Kaldy 1997) bacterioplankton production and C:N = 4.3, ranged from 450 to 21 10 pm01 N m-2 d-' and far exceeded the estimated NHdi regeneration, which may indicate that Month Tissue type C N C:N Total plant uptake of dissolved organic nitrogen was important. (W%) (wtuiu) C:N 1 Phytoplankton N demand, calculated from gross pri- Blades 34.97 2.08 19 62 39.06 mary production and C:N ratio of 6.6, ranged from < 10 Roots/rhizome 36.06 0.89 47.03 to 700 pm01 N m-2 d-l, indicating that the net uptake September Blades 32.45 1.89 20.03 42.06 Roots/ 33.30 0.75 51.87 (in the light) was usually an underestimate of total November Blades 33.57 2.41 16.27 22.63 uptake (Table 4). In the water column, net uptake of Roots/rhizome 32.89 1.58 24.35 NH4' in the light was generally much lower than net regeneration measured in the dark, and there was no seasonal pattern to the variability in fluxes measured in the light (Fig. 3). m-2 d-' and was similar to bacterial production rates The rates of water column NH,' regeneration, calcu- measured in the water column over other subtropical lated as the sum of NH,' release in dark incubations seagrass meadows, including upper Laguna Madre conducted during the day and night, ranged from 124 (Moriarty & Pollard 1982, Moriarty et al. 1990, Chin- to 652 pm01 N m-2 d-' or 0.10 to 0.59 pm01 N 1-' d-' Leo & Benner 1991). The estimated bacterioplankton (Table 4). On 25 June 1997, potential rates of NH,' growth efficiencies ranged from 21 to 38% and fell uptake and regeneration were estimated using the within the range of most aquatic ecosystems studied ''N isotopic dilution technique in incubations amended (del Giorgio et al. 1997). with 4 pM ISNH4C1 (Gardener et al. 1991, 1995). Po- tential regeneration rates were similar in both light (0.10 pM h-') and dark (0.11 pM h-') incubations Water column and benthic nutrient fluxes (Gardner unpubl. data), and were about 3 times the rate estimated from the net change in NH,' measured Net regeneration of NH,' in the water column in the water column during the same period (0.03 pM occurred primarily in the dark and ranged from 45 to h-'). Thls suggested that the absence of light and use of 643 pm01 N m-' d-' or 0.04 to 0.55 pm01 N 1-' d-' net changes in NH,' did not grossly underestimate the (Table 5). Water column regeneration of NH4+ mea- regeneration of NH,' in the water column. The pres- sured in the dark increased through the summer and ence of light has been reported to increase rates of decreased in the fall (Fig. 3). However, regeneration remineralization in systems with a greater proportion was greatest in March possibly due to the presence of of phytoplankton production (Gardner et al. 1996). a brown tide which may have provided additional dis- Water column fluxes of NO3- + NOz- were almost solved organic nitrogen for remineralization. NH,' was always undetectable, except in June 1997, when sam- primarily taken up in the water column in the light, ples were analyzed using chemiluminescence detec- and net fluxes ranged from -145 to -45 pm01 N m-2 d-' tion (Table 5).The more sensitive and precise chemilu- or -0.13 to -0.04 pm01 N 1-' d-l, except in June 1997, minescence method served to verify that the NO3- +

Table 4. Average water column estimates of respiration (R),gross primary production (GPP),bacterioplankton production (BP), bacterioplankton growth efficiency (BGE),phytoplankton nitrogen demand (Phyto. N demand) based on C:N = 6.7, and bacterio- plankton N demand (Bacterio. N demand) based on C N = 4.3, ammonium (NH,+)regeneration, and molar C:N radio of dissolved organlc matter (DOM) rem~neralized.All rates are in units of mm01 C m'' d.' or mm01 N m-2d-' Errors are reported as *l SD

Month R G PP BP BGE Phyto. Bacterio. NH,' C:N of DOM (X) N d.emand N demand regeneration remineralized

June 27 i 2 3 * 2 6 + 0.7 19 c0.01 1.20 0.33 + 0.06 81 * 0.2 July 14*6 0.4t3 6 * 0.4 3 0 ~0.01 1.08 0.55 + 0.17 26 0.7 September 9i1 3t1 6 + 1.9 38 0.10 1.35 0.65 + 0.10 14 t 0.2 November 7 i 2 1 t 1 2* 1.0 2 1 0.04 0.45 0.12 + 0.07 55 * 0.6 )larch 17i7 9*3 9 * 1.6 36 0.35 2.11 0.65 + 0.46 26 * 0.8 Junv 15 1 6*2 6 * 2.2 2 8 0.70 1.37 0.28 + 0.13 54 t 0.5 Ziegler & Benner: Nutrient cycling in a seagrass meadow water column 57

Table 5. Water column and benthic net nutrient fluxes (pm01 N or P m-' d-') measured in light and dark incubations. Daily rates for light and dark incubations were calculated from the number of hours of daylight or night, respectively. All rates are reported as mean (n 2 3) + 1 SD. -: not detectable, na: not available due to contamination

Date NH,' NO3- + NOz- SRP Llght Dark L~ght Dark Light Dark

Water column 12-13 Jun 1996 -99 + 382 182 * 35 - - - - 23-24Jun 1996 - 314 + 102 - - - - 10-11Sep1996 -145 + 56 421 * 60 - - - - 6-7 NOV1996 -45 i 40 45 * 32 - - - - 8-9 Mar 1997 -114 i 111 643 + 619 - - -116 i 0 101 i 155 24-25 Jun 1997 76 *41 277 -t 130 92 l0 5*4 - - Benthos 13 Jun 1996 253 i 136 na - - - 25 Jul 1996 187 + 0.2 -107 * 255 - - - 12 Sep 1996 310 * 327 28 -t 333 - - 76 r 77 5 Nov 1996 112 +233 296 -r 213 - - - 8 Mar 1997 363i293 -228 * 101 - - - 24 Jun 1997 123 + 220 306 i 295 -1 rt 9 6 _t 23 - -

NO2- fluxes were small and variable (9 + 5 pm01 N m-' column were -116 and 101 pm01 N m-* d-' in the light d-l). Water column fluxes of SRP were usually below and dark, respectively (Table 5). detection, except in March, when fluxes in the water Net benthic NH,' fluxes were always positive in the light and ranged from 112 to 363 pm01 N m-' d-' (Table 5, Fig. 3). Net fluxes in the dark were more vari- Water Column able and ranged from -228 to 306 pm01 N m-' d-l. The I large standard deviation was in part a result of the major differences in fluxes between vegetated and unvegetated sites. The net flux of NH,+ in the light was usually much lower in the unvegetated versus the veg- etated site (Fig. 4). The magnitude of the benthic release of NH,' was often lower than NH,' regenera- tion in the water column (Fig. 5).The NH,' released from the seagrass-dominated sediments on 25 June 1997 (a clear day) was more than double the flux mea- sured on the previous day when light levels were much lower (Table 6). The net fluxes of both No3- +NOz- and SRP were almost always undetectable. The samples

Jun Jul Sep Nov Mar Jun 1996 1997

Flg. 3. Average ammonium (NHdt)fluxes in the water column , U -1 00 1 I (n 2 3) and benthos (n = 4) measured in Light and dark incuba- Jun Jul Sep Nov Mar Jun tion~.Rates are reported as daily rates based on the number 1996 1997 of hours of saturating PAR light (for seagrasses) and the number of hours below saturation for light and dark rates. Fig. 4. Benthic ammonium (NH,') fluxes measured in the light respectively. Error bars represent r1 SD from vegetated and unvegetated sediments 58 Mar Ecol Prog Ser 188: 51-62, 1999

Table 6. Total underwater PAR irradiance, benthic nrt primary production when nutrients are in very low con- (NPP),d.issolved organic carbon (DOC) and ammonium (4H4+)fluxes measured centrations and carbon substrate is In the light for seagrass-dominated sedlments on 24 June and 25 June 1997 readily available (Rhee 1972), In Benthic rates are reported as the mean (n = 4) i1SD Laquna- Madre, water column net uptake of NH,+ in the light was always Date PAR" NPP~ DOC fluxh NH,' flux small relative to net regeneration of (E m-2 d-') (rnmol C m-' d-l) (mm01 C m-' d-l) (pmol C m-'d-l) NH4+ in the dark. Phytoplankton de- 24 Jun 1997 14 7.7 2 1.5 1142 + 219 2 * 9 mand for N represented only about 0.8 25 Jun 1997 36 15.6 * 2.4 2260 + 1298 21 526 to 34% of the total N demand in the dK.Dunton (unpubl. data); 'Ziegler & Benner (1999) water column, indicating that hetero- trophic uptake was probably responsi- ble for most of the utilization of NH4+ in the water column. Furthermore, collected in June and analyzed for NO3- + NO,- using potential NH,+ uptake rates, measured at the study chemiluminescence detection verlfied that, in fact, the site on 25 June 1997 using the "N isotope dilution ap- benthic fluxes of NO, + NO2- were extremely small proach, were essentially the same in light and dark and variable (Table 5). incubations (W. Gardner unpubl. data).

DISCUSSION Benthic source of water column nutrients

Nutrient limitation of phytoplankton production Release of nutrients from the benthos is often important for the growth of phytoplankton in the The close proximity of the seagrasses to high con- overlying water (Rowe et al. 1975). In various types of centrations of nutrients and bacterial activity in the estuaries, benthic nutrient fluxes have been reported benthos can provide them with an advantage over to range from -1 to 37 mm01 NH,' m-2 d-', -16 to phytoplankton where nutrients are in much lower con- 15 mm01 NO; + NOz- d-l, and -1 to 9 mm01 SRP centrations (Wetzel 1975, Sandjensen & Borum 1991). m-2 d-l (Cowan et al. 1996 and references cited In lower Laguna Madre, low inorganic nutrient con- within). In most shallow estuarine systems studied, centrations and the release of DOC from the benthos the availability of labile organic matter ultimately fueling heterotrophic activity (Ziegler & Benner 1999) regulates benthic nutrient regeneration rates (Nixon limit primary production in the water column. The 1981, Kemp et al. 1992). Net NH,+ regeneration rates degree to which phytoplankton are nutrient limited estimated for temperate seagrass beds ranged from may depend upon the magnitude of allochthonous as 7 to 106 mm01 N m-2 d-' for shallow and deeper sites well as the macrophytic input of DOC (del Giorgio & during different times of the year, suggesting that Peters 1994).Bacterioplankton have been found to out- remineralization in the benthos acts as a major source compete phytoplankton for ava~lable DIN at times of nutrients (Dennison et al. 1987). Fluxes of nutrients in shallow tropical environments, such as , suggest that efficient bacterial activity is responsible for tight recycling of the regenerated nutrients within the sediments (Alongi 1994). The tight recy- cling of nutrients within the sediments of shallow aquatic ecosystems may prevent benthic remineral- ization from providing nutrients to the overlying water colun~n. In subtropical, southern Laguna Madre fluxes of NH4', NO3- + NOz-, and SRP were all much smaller then those reported for many other estuaries, suggest- ing extremely efficient recycling of nutrients within the sediments in this . NH,+ was the major Jun Jul Sep Nov Mar Jun form of inorganic N exchanged between the water 1996 1997 column and the bcnthos. Rates of benthic NH4+ Flg. 5. A comparison of the rates of ammonium (NH,') regen- release were in the range of those calculated eration in the water column and total daily benthic flux of from measurements of NH,' regeneration and sea- ammonium grass assimilation for a marjna bed in Alaska Ziegler & Benner: Nutrient cycling in a seagrass meadow water column 59

(88 pm01 N m-2 d-'; Short 1983). Hoxvever they were Water column remineralization of DOM much Iower than fIuxes reported for Puttalam as a source of nutrients (-14 to 7 mm01 N m-' d-'), a more eutrophic tropical seagrass community in Sri Lanka that receives efflu- Remineralization of DOM was a major source of ent from local shrimp farming facilities (Johnson & water column NH4' in lower Laguna Madre. The mag- Johnstone 1995). The low nutrient fluxes and high nitude of NH,' regeneration in the water column was respiration in the benthos (Ziegler & Benner 1999) usually larger than the net release of NH4+ from the indicated rapid and tight recycling of N and P within benthos. NH,' regeneration in the water column was the sediments in Laguna Madre. NH4+regeneration in significantly correlated to the release of DOC from the the sediments of seagrass beds have been found to benthos (r2 = 0.91, p < 0.05, n = 4; Ziegler & Benner approximately equal utilization by seagrasses, sug- 1999), suggesting that benthic release of DOM was gesting that N is often tightly recycled within seagrass fueling the regeneration of NH4+in the water column. sediments (Boon et al. 1986) and not a major source of The correlation between NH4+ regeneration in the nutrients for the overlying water column. The rela- water column and water temperature was not signifi- tively high nutrient fluxes in Puttalam Lagoon may be cant, indicating that temperature was not as important indicative of a perturbed seagrass community (John- as the benthic supply of DOM in regulating NH4' son & Johnstone 1995). The large inputs of nutrients regeneration. and subsequent reduction in the productivity of the Phytoplankton N demand was always met by the seagrasses in Puttalam Lagoon may have destroyed sum of the NH4+regeneration in the water column and the tight recycling of nutrients evident and potentially the benthic release of NH,'. Bacterioplankton N de- important in southern Laguna Madre where alloch- mand generally exceeded the supply of NH,' from thonous input is very small. water column regeneration and the benthic fluxes. The Differences in benthic fluxes measured in Laguna sum of water column NH,' regeneration and benthic Madre on 24 and 25 June 1997 demonstrated that a release of NH4+ represented 37 to 73% of bacterio- link may exist between seagrass exudation of DOC plankton N demand, except in November, when it was and the benthic release of NH,+ (Table 6). This differ- about 117 % of bacterioplankton N demand. This indi- ence between 24 and 25 June was not observed in cated that dissolved organic nitrogen was probably a the unvegetated sediments, where fluxes of NH,' major source of N for bactenoplankton during most of were the same on the 2 consecutive days. Through- the year. out the year, most of the release of NH,' from the benthos occurred in the light and from the seagrass- dominated sediments, indicating that the source of Relationship between the source of DOM NH,+ was often related to seagrass exudation and its and nutrient cycling subsequent remineralization. The greatest light-medi- ated benthic release of NH,' occurred in March and Changes in the availability and composition of ben- coincided with the highest rate of benthic net primary thic-derived DOM in Laguna Madre significantly production (Ziegler & Benner 1999) and seagrass influenced the cycling of N in the water column. Bac- tissue N content (Kaldy 1997). In November, when terioplankton growth efficiencies were significantly benthic net primary production was lowest (Ziegler & correlated to water column NH4+ regeneration, with Benner 1999), the release of NH,' did not appear to the highest growth efficiencies occurring in March and be related to any light-mediated process. There is September (Fig. 6). The C:N ratio of organic matter some evidence that amino acids, released by the roots remineralized in the water column, calculated from of seagrasses during active photosynthesis, are de- water column respiration and NH4' regeneration, graded by thereby generating NH4+ (Jm- varied temporally and ranged from 14 to 81 (Table 4). gensen et al. 1981, Wood & Hayasaka 1981, Smith et Because there is no net accumulation of bacterio- al. 1984). Porewater NH,+ concentrations have been plankton biomass in Laguna Madre during the day observed to increase diurnally in the sediments of (Chin-Leo & Benner 1991) these C:N values are repre- some seagrass beds, suggesting that remineralization sentative of the DOM utilized by heterotrophic bac- of N in exudates could be a major source of NH4+ terioplankton. The temporal variability in the C:N of (Moriarty et al. 1986, Blackburn et al. 1994, Stapel et the bioreactive DOM and bacterioplankton growth al. 1997). Some proportion of this regenerated NH4+ efficiency indicated that the DOM released from the in the benthos may be released into the over- benthos and utilized in the water column was season- lying water column and could account for the small ally variable in composition. The C:N of the organic light-mediated release of NH,+ measured in Laguna matter remineralized in the water column was lowest Madre. in September and lghest in June, indicating that bio- Mar Ecol Prog Ser 188: 51-62, 1999

unpublished data relevant to this work. We also appreciate Mar* the assistance of Don Hockaday at the UT-PanAm Coastal , /&p Studies Laboratory. South Padre Island, Texas, who provided / needed laboratory space and facilities during field trips to lower Laguna Madre. This manuscript was improved by the review and comments of 3 anonymous reviewers. This research was supported by NSF grant DEBB 9420652. This is contribution number 1127 from the University of Texas Jun* Marine Science Institute.

r?= 0.74 . LITERATURE CITED Nov p c 0.03 Alongi DM (1994) The role oi bacteria in nutrient recycling in tropical and other coastal benthic ecosystems. Hydrobio1285:19-32 Bacterioplankton GE (%) Amon RWM, Benner R (1998) Seasonal patterns of bacterial abundance and production in the Mississippi River plume Fig. 6. Linear regression analysis of water column ammonium and their importance for the fate of enhanced primary pro- regeneration versus bacterioplankton growth efficiency (GE) duct~onMicrob Ecol 35.289-300 Biddanda B, Opsahl S, Benner R (1994) Plankton respiration and carbon flux through bacterioplankton on the Loui- reactive DOM was enriched in N during the late sum- siana shelf. Limnol Oceanogr 39(6):1259-1275 mer and relatively depleted in N during early summer. Billen G (1984) Heterotrophic utilization and regeneration of Processes responsible for the release of DOM, namely nitrogen. In: Hobbie JE, Williams PJ1 (eds) Heterotrophic actlvity in the sea. Plenum, New York, p 313-355 exudation and leaching, were probably quite different Blackburn TH, Nedwell DB, Wiebe WJ (1994) Active m~neral during these times of the year. During June, benthic cycling in a Jamaican seagrass sediment. Mar Ecol Prog net primary production was relatively high and sea- Ser 110:233-239 grass exudation would be expected to be greatest. Boon PI. Moriarty DJW, Saffigna PG (1986) Rates of ammo- nium turnover and role of amino-acld deamination in sea- Leaching, however, was probably more predominant grass (Zostera capncorni) beds of Moreton Bay. Aust Mar in September, when seagrasses were beginning to Biol 91 259-268 senesce and benthic net primary production was low Borum J (1985) Development of epiphytic communities on (Ziegler & Benner 1999). eelgrass () along a nutrient gradient in a Bacterial processes and N cycling have been found Danish estuary. Mar Biol 87:211-218 Chin-Leo G, Benner R (1991) Dynamics of bacterioplankton to be closely coupled to photosynthetic production of abundance and production in seagrass communities of a organic matter in other aquatic environments (Cole et hypersaline lagoon. Mar Ecol Prog Ser 73:219-230 al. 1982, 1988, Coffin et al. 1994). In Laguna Madre, Chin-Leo G, Benner R (1992) Enhanced bacterioplankton pro- release of DOM from the benthos was found to duction and respiration at intermediate in the Mississippi River plume. Mar Ecol Prog Ser 87:87-103 enhance regeneration of NH,' in the water column. Coffin RB. Connolly JP, Harris PS (1994) Availability of dis- Studies have indicated that efficient regeneration of solved organic carbon to bacterioplankton examined by nutrients by bacterioplankton may only occur when oxygen utilization. Mar Ecol Prog Ser 101:9-22 available substrates have a C:N < 10 (Fenchel & Black- Cole JJ, Likens GE, Strayer DL (1982) Photosynthetically pro- burn 1979, Billen 1984, Goldman et al. 1987). There- duced dissolved organic carbon: an important carbon source for planktonic bacteria. Limnol Oceanogr 27: fore, bacterial regeneration of NH,+ in Laguna Madre 1080-1090 may have been due to remineralization of a fraction of Cole JJ, Findlay S, Pace ML (1988) Bacterial production in DOM with a C:N < 10 or a combination of C-rich and fresh and salt-water ecosystems: a cross-system overview. N-rich compounds Previous studies in seagrass beds Mar Ecol. Prog Ser 43:l-10 Cowan JLW, Pennock JR, Boynton WR (1996) Seasonal and have demonstrated that DON constituted the majority interannual patterns of sediment-water nutrient and oxy- of interstitial N (Boon et al. 1986) and the dominant gen fluxes in Mobile Bay, Alabama (USA):regulating fac- form of N released from the sediments. Seasonal varia- tors and ecological significance. Mar Ecol Prog Ser 141: tions in bacterioplankton growth efficiencies and the 229-245 C N of bioreactive Doh? in Laguna Madre suggest that del G~orgioPA, Peters RH (1994) Patterns in planktonic P:R ratlos in lakes: Influence of lake trophy and dissolved the regeneration of NH4+ in the water column is organic carbon. Limnol Oceanogr 39(4):772-783 directly linked to the chemical composition of the del Giorgio PA, Cole JJ. Cimbleris A (1997) Respiration rates DOM released from the benthos. in bacteria exceed phytoplankton in unproductive aquatic systems. Nature 385:148-151 Dennison WC, Aller RC, Alberte RS (1987) Sediment ammo- Acknowledgements. The authors especially thank Beau Har- nium availability and eelgrass (Zosfera manna) growth. degree for extensive and essential field and laboratory assis- Mar Biol94:469-477 tance. We would like to thank Wayne Gardner for sharing Dennison WC, Orth RJ. Moore KA, Stevenson JC, Carter V, Ziegler & Benner: Nutnent cychg in a seagrass meadow water column 61

Kollar S, Bergstrom PW, Batuik RA (1993) Assessing water tion and physical exchange processes. Mar Ecol Prog Ser quality with submersed aquatic vegetation: habitat re- 85:137-152 quirements as barometers of Chesapeake Bay health. Bio- firchman DL, K'nees E, Hodson RE (1985) Leucine incorpora- science 43:86-94 tion and its potential as a measure of protein synthesis by Dunton KH (1990) Production ecology of Ruppia maritima L. bacteria in natural aquatic systems. Appl Environ micro- s.1. and Halodule wrightii Achers. in two subtropical estu- biol 49:599-607 aries J Exp Mar Biol Ecol 143:147-164 Lee S, Fuhrman JA (1987) Relationships between biovolume Fenchel T, Blackburn TH (1979) Bacteria and mineral cycling and biornass of naturally denved marine bacterioplank- Academic Press, New York ton. Appl Environ Microbiol 53: 1298-1303 Flores-Verdugo FJ, Day JW, Mee L, Briseno-Duenas R (1988) Middelboe M, Kroer N, Jorgensen NOG, Pakulski D (1998) Phytoplankton production and seasonal biomass variation Influence of sediment on pelagic carbon and nitrogen of seagrass, Ruppia maritima L., in a tropical Mexican turnover in a shallow Danish estuary. Aquat Microb Ecol lagoon with an ephemeral inlet. Estuaries 11(1):51-56 14:81-90 Fourqurean JW, Jones RD, Zieman JC (1993) Processes influ- Moriarty DJW, Pollard PC (1982) Die1 variation of bacterial encing water column nutrient characteristics and phos- productivity in seagrass (Zostera capncorni) beds mea- phorus limitation of phytoplankton biomass in Florida sured by rate of thym~dineincorporation Into DNA Mar Bay, FL, USA: inferences from spatial distributions. Estuar Biol 72:165-173 Coast Mar Sci 36:295-314 Moriarty DJW, lverson RL, Pollard PC (1986) Exudation of Gardner WS, Wynne DS, Dunstan WM (1976) Simplified pro- organic carbon by the seagrass Halodule wnqhtii Aschers. cedure for the manual analysis of nltrate in seawater. Mar and its effects on bacterial growth in the sediment Exp Chem 4:393-396 Mar Biol Ecol96:115-126 Gardner WS, Herche LR, St. John PA, Seitzinger SP (1991) Moriarty DJW, Roberts DG, Pollard PC (1990) Primary and High-performance Iiquid chromatographic determination bacterial productivity of tropical seagrass communities in of '5NH4:[14NH4+ I5NH,] ion ratios in seawater for isotope the Gulf of Carpentaria, Australia. Mar Ecol Prog Ser 61: dilution experiments. Anal Chem 63: 1838-1843 145-157 Gardner WS, Bootsma HA, Evans C, John PAS (1995) Nixon SW (1981) Remineralization and nutrient cycling in Improved chromatographic analysis of l5N:I4N ratios in coastal marine sediments. In: Neilson BJ, Cronin LE (eds) ammonium or nitrate for isotope addition experiments. Estuaries and nutrients. Humana, New York, p 11 1-138 Mar Chem 48:271-282 Onuf CP (1996) Biomass patterns in seagrass meadows of the Gardner WS, Benner R, Amon RMW, Cotner JB, Cavaletto JF, Laguna Madre, Texas. Bull Mar Sci 58:404-420 Johnson JR (1996) Effects of high-molecular-weight dis- Orth RJ, Moore KA (1983) Chesapeake Bay: an unprecedented solved organic matter on nltrogen dynamics in the Missis- decline in submerged aquatic vegetation. Science 222: sippi River plume. Mar Ecol Prog Ser 133:287-297 51-53 Goldman JC, Caron DA, Dennett MR (1987) Regulation of Oviatt CA, Rudnick DT, Keller AA, Sampou PA, Almquist GT gross growth efficiency and ammonium regeneration in (1986) A comparison of system (02 and COz) and C-14 bacteria by substrate C,N ratio Limnol Oceanogr 32: measurements of in estuarine mesocosms. 1239-1252 Mar Ecol Prog Ser 28:57-67 Herzka SZ, Dunton KH (1996) Seasonal photosynthetic pat- Pedersen MF, Borum J (1993) An annual nitrogen budget for terns of the seagrass Thalassia testudinrrm in the western a seagrass Zostera marina population. Mar Ecol Prog Ser Gulf of Mexico. Mar Ecol Prog Ser 152:103-117 101:169-177 Hillman K, Walker DI, Larkurn AWD, McComb AJ (1989) Pro- Quammen ML. Onuf CP (1993) Laguna Madre: seagrass ductivity and nutrient limitation. In: Larkum AWD, Mc- changes continue decades after reduction. Estu- Comb AJ, Shepard SA (eds) Biology of seagrasses. Else- aries 16(2):302-310 vier, Amsterdam, p 635-685 Redfield AC (1958) The biological control of chemical factors Humrn HJ. Hildebrand HH (1962) Marine algae from the gulf in the environment. Am Sci 46:205-221 coast of Texas and Mexico. Pub1 lnst Mar Sci 8:227-268 Rhee GY (1972) Competition between an alga and an aquatic Iizumi H, Hattori A, McRoy CP (1982) Ammonium regenera- bacterium for phosphate. Limnol Oceanogr 17(4):505-514 tion and assimilation In eelgrass (Zostera marina) beds. Rou7e GT, Clifford CH, Smith KL, Hamilton PL (1975) Benthic Mar Biol 66:59-65 nutrient regeneration and its coupling to primary produc- Johnson P, Johnstone R (1995) Productivity and nutrient dy- tivity in coastal waters. Nature 255:215-217 namics of tropical sea-grass communities in Puttalam Sandjensen K, Borum J (1991) Interactions among phytoplank- Lagoon, Sri Lanka. Amblo 24(7-8):411-417 ton, periphyton, and macrophytes in temperate freshwaters Jsrgensen NOG, Blackburn TH, Henr~ksenK. Bay D (1981) and estuaries. Aquat Bot 41(1-3):137-175 The importance of oceanica and Cymodocea Short FT (1983) The response of interstitial ammonium in eel- nodosa as contributors of free amino acids in water and grass (Zostera marina L.) beds to environmental perturba- sediment of seagrass beds. PSZN I: Mar Ecol 2:97-112 tions. J Exp Mar Biol Ecol 68:195-208 Kaldy JE (1997) Production dynamlc, reproduction ecology Short FT, McRoy CP (1984) Nitrogen uptake by and and demography of Thalassia testudinum (turtle grass) roots of the seagrass Zostera marina L. Bot Mar 27:547-555 from the Lower Laguna Madre, Texas. PhD dissertation, Smith GW, Hayasaka SS, Thayer GW (1984) Ammonification University of Texas at Austin of amino acids by the microflora of Zostera Kemp WM, Boynton WR, T~villeyRR, Stevenson JC, Means marina L. and Halodule wrjqhtli Aschers. Bot Mar 27: JC (1983) The decline of submerged vascular plants in 23-27 upper Chesapeake Bay: summary of results concerning Solorzano L (1969) The determination of ammonium in natural possible causes. Mar Technol Soc J 17:78-89 waters by phenol-hypochlonte method. Llmnol Oceanogr Kemp WM, Sampou P, Caffrey J, Mayer M, Henriksen K 14:799-801 (1992) Seasonal depletion of oxygen from bottom waters of Stapel J, Hemminga MA (1997) Nutrient resorption from sea- Chesapeake Bay: roles of benthic and planktonic respira- grass leaves. Mar Biol 128:197-206 62 Mar Ecol Prog Ser 188: 51-62, 1999

Stapel J, Manuntun R, Hemminga MA (1997) Biomass loss during a brown tide bloom. In: Smayda TJ, Shirnizu Y and nutrient redistribution in an Indonesian Thalassia (eds) Toxic phytoplankton blooms in the sea. Elsevier, hemprichii seagrass bed following seasonal low tide expo- Amsterdam, p 711-7 16 sure during daylight. Mar Ecol Prog Ser 148:251-262 Williams PJL (1981) Incorporation of microheterotrophic pro- Stockwell DA, Buskey EJ, Whitledge TE (1993) Studies on cesses into the class~calparadigm of the planktonic food conditions conducive to the development and mainte- web. Kiel Meeresforsch 5:l-28 nance of a persistent 'brown tide' in Laguna Madre, Wood DC, Hayasaka SS (1981) Chemotaxis of rhizoplane bac- Texas. In: Smayda TJ. Shirnizu Y (eds) Toxic phytoplank- tend to amino acids cornpnsing eelgrass (Zostera marina ton blooms in the sea. Elsevier, Amsterdam, p 693-698 L.) root exudate. J Exp Mar Biol Ecol50:153-161 Strickland JD, Parsons TR (1972) A practical handbook of sea- Ziegler S, Benner R (1998) Ecosystem metabolism of a sub- water analysis, 2nd edn. Bull F~shRes Bd Can 167 tropical seagrass-dominated lagoon. Mar Ecol Prog Ser Wetzel RG (1975) Limnology, 2nd edn. WB Saunders Com- 173:l-12 pany, Philadelphia Ziegler S, Benner R (1999) Dissolved organic carbon cycling Whitledge TE (1993) The nutrient and hydrographic con- in a subtropical seagrass-dominated lagoon. Mar Ecol ditions prevailing in Laguna Madre, Texas before and Prog Ser 180:149-160

Ed~torial respons~bdity:Otto ffinne (Editor), Submitted: September 3, 1998; Accepted: April 30, 1999 Oldendorf/Luhe, Germany Proofs received from authorls); October 25, 1999