Pelagic processes in extensive shrimp po

Aquaculture 175 Ž1999 121–141 Pelagic processes in extensive shrimp ponds of the Mekong delta, Vietnam Daniel M Alongi a,) , Paul Dixon a , Danielle J Johnston Doan Van Tien b, Tran Thanh Xuan b a,1 , a b Australian Institute of Marine Science, PMB No 3, TownsÕille M.C., Queensland 4810, Australia Research Institute for Aquaculture No 2, 116 Nguyen Dinh Chieu St., District 1, Ho Chi Minh City, Viet Nam Accepted 19 February 1999 Abstract Rates of pelagic primary production, respiration, and bacterioplankton dynamics were measured in relation to water quality parameters in three extensive shrimp ponds in the Mekong delta, Vietnam There were few consistent differences in pelagic characteristics among different locations within these ponds, among the three ponds, or between the ponds and adjacent river water Rates of primary production Ž14C uptake ranged from - to 94 mg C my3 dy1 Rates of gross primary production Žlight–dark bottle technique ranged from y184 to 2697 mg C my3 dy1 Rates of pelagic respiration ranged from 60 to 3783 mg C my3 dy1 Primary production rates measured by oxygen flux were greater than those measured via 14C uptake, with PrR ratios varying widely Žy0.45 to 3.4., but the mean PrR at each site was - with a grand mean among ponds of 0.7 Bacterioplankton numbers Žmean range: 0.2 to 27.2 = 10 cells mly1 and productivity Žrange: 2.5 to 297.3 mg C my3 dy1 did not vary consistently among ponds with season Mean bacterioplankton growth rates were fastest Ž m s 0.29 dy1 in the pond with highest shrimp production and slowest in the poorest yielding pond Ž m s 0.08 dy1 The primary production and respiration rates, and bacterioplankton dynamics, indicate that these pond waters are net heterotrophic Plankton measurements rarely correlated with changes in physicochemistry ŽpH, dissolved O , salinity, temperature or nutrient concentrations as within-site and temporal variability were large for most parameters The diurnal cycles of physicochemical characteristics were similar to those measured in other unfertilized, low-alkalinity ponds, but these cycles were dampened by intense rainfall Our data indicate that low net primary production, high rates of ) Corresponding author Tel.: q61-7-47-534313; Fax: q61-7-47725852; E-mail: d.alongi@aims.gov.au Present address: School of Aquaculture, University of Tasmania, P.O Box 1214, Launceston, Tasmania 7250, Australia 0044-8486r99r$ - see front matter q 1999 Elsevier Science B.V All rights reserved PII: S 0 4 - 8 Ž 9 0 - 122 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 respiration, moderate rates of bacterioplankton production, high suspended solid and nutrient concentrations, low and very variable pH and dissolved O concentrations, and variations in salinity due to intense rainfall episodes limit shrimp production in these extensive ponds q 1999 Elsevier Science B.V All rights reserved Keywords: Primary production; Bacterial production; Water quality; Nutrients; Oxygen production; Oxygen consumption; Bacteria; Extensive shrimp culture Introduction Shrimp aquaculture expanded rapidly in Southeast Asia up until the mid-1990s to the extent that land clearing and construction of shrimp ponds in the region is the leading cause of losses of mangrove forests and other forms of coastal deterioration World shrimp production has leveled off in recent years, as many aquaculture farms have either collapsed or experienced declining yields ŽLucien-Brun, 1997 Lack of sustainability is due to many factors, such as poor management, and low quality and overexploitation of seedstock Knowledge of ecological factors controlling pond production is also crucial for more effective management and sustainability of aquaculture enterprises An ability to estimate the production capacity of shrimp ponds relies upon knowledge not only of the reproductive and physiological capabilities and tolerance limits of shrimps, but also of system-level processes Processes such as rates of natural and supplemented food inputs, internal recycling, and outputs Žharvesting, losses via respiration and bacterial production ultimately regulate productivity The latter information is crucial in constructing ecosystem-level budgets that can aid in estimating maximum sustainable yields and in minimizing losses from ponds Sustainability problems and coastal degradation due to intensification of coastal shrimp aquaculture have plagued most Southeast Asian countries, such as the Philippines, Thailand, Taiwan and China ŽLucien-Brun, 1997 In Vietnam, there has been intensification of shrimp culture with concomitant loss and damage to coastal habitats This degradation has been especially severe for mangrove forests in the Ca Mau province in the Mekong delta of southern Vietnam This situation has resulted in the establishment within Ca Mau province of 22 mixed shrimp farming–mangrove forestry enterprises, where both shrimp and mangrove wood are harvested by individual farmers Shrimp farming in Vietnam is mostly extensive with low average yields Vietnam accounts for only 4% of world production Ž; 30,000 mt., but there are ; 200,000 in production, making Vietnam in area one of the largest shrimp farming nations in Southeast Asia ŽRosenberry, 1996 Although there is some indication that the rate of coastal degradation has slowed in the Mekong delta, current evidence indicates that yields of both shrimp and mangrove wood are declining As in other countries experiencing declining yields, poor shrimp yields in the Mekong delta have been attributed to several factors, such as pollution, poor management and infrastructure, acid sulfate soil, low quality and quantity of seedstock, uncontrolled coastal development, and overexploitation of wild stocks The aim of this study was to determine if poor water quality and low phytoplankton productivity were responsible for the poor shrimp production in these extensive ponds D.M Alongi et al.r Aquaculture 175 (1999) 121–141 123 In this paper, we compare and contrast rates of pelagic primary productivity, bacterioplankton production and growth rates, and bacterioplankton abundance and respiration in relation to water quality parameters and shrimp yield in separate and mixed, extensive shrimp farm–mangrove forestry enterprises in the Mekong delta Methods and materials 2.1 Study sites The study was conducted in shrimp ponds within one enterprise, Tam Giang III Žlatitude 8.88N, longitude 105.28E., located within the Ca Mau province of the Mekong delta of southern Vietnam ŽFig Tam Giang III is an enterprise consisting of extensive farms utilizing two different types of farming systems: Ž1 separate pond Ž20–30% of farm area and mangrove forest Ž70–80% areas, and Ž2 mixed shrimp pond–forest systems Ponds in both systems consist of a series of long Ž250–800 m narrow Ž3–4 m wide canals or channels dug either through Žmixed or adjacent Žseparate to the forest, and separated by levees In the mixed systems, the levees are vegetated by mangroves, whereas in the separate systems, the levees are mostly bare of vegetation Two ponds ŽPonds 12 and 23 were sampled in October 1996, May 1997 and November 1997; a third pond ŽPond 22 was sampled in November 1997 Sampling in November 1997 occurred week after Typhoon Linda passed over the region All sampling was conducted during the day, except for some hydrological measurements in May and November 1997 Žsee below The first system, Pond 12, was a mixed shrimp pond–mangrove farm ŽFig with a pond area of 2.4 of a total farm area of 13 The second system, Pond 23, was a separate farm with a pond area of 0.7 of a total farm area of 1.8 The third farming system, Pond 22, was a separate system with a pond area of 1.1 of a total farm size of 10 Water depth in the ponds ranged from 0.4–0.7 m, and averaged 0.5 m ŽClough and Johnston, 1997 Ponds 22 and 23 are connected directly to Song Dam Chim, a tributary of the lower Mekong Pond 12 is connected to a canal leading to Song Dam Chim Each pond is connected to the river by a single sluice gate located on the main waterway of each farm The gates, made either of wood or cement, are composed of a series of horizontal boards which are raised to allow water to flow into and out of the pond Water flow and shrimp recruitment is tidal The ponds not receive supplementary food, aeration, water flow, liming or fertilizer treatment The ponds are usually drained to excavate sludge on the bottom once per year After excavation, ponds are allowed to fill naturally on the first spring tide and a 15-day grow-out cycle is initiated, during which recruitment and harvesting occurs on consecutive flooding and ebbing tides over 3–5-day intervals during spring tides Harvesting is accomplished by placing nets at the front of the opened sluice gate, while the ponds are drained to ; 20 cm depth During the next neap tide, the sluice gate is closed for 10–12-day grow out There is little water exchange during this period, although some farmers allow ; 20 cm exchange of water 124 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 125 This recruitment-harvesting cycle is repeated for the rest of the year The dominant shrimp species cultured are Metapenaeus ensis Ž48.3% of total harvest., M lysianassa Ž32.2% and Penaeus indicus Ž9.7% Average annual shrimp yields are low, ranging from 380 kg hay1 yry1 and 390 kg hay1 yry1 at Ponds 23 and 12, respectively, to 1166 kg hay1 yry1 at Pond 22 ŽClough and Johnston, 1997 2.2 Water quality sampling Water samples for dissolved and particulate nutrients in each pond were collected ; 5–10 cm below the surface using sterile plastic syringes: one set of triplicate samples y y 3y for dissolved inorganic ŽNHq and organic ŽDOC, DON, DOP , NO q NO , PO4 nutrients, and another set of triplicates for suspended solids and particulate Žtotal carbon and nitrogen, total organic carbon nutrients Samples were filtered Ž0.45 mm cellulose acetate filters for dissolved nutrients; 0.4 mm Nuclepore filters for DOC into acidwashed test tubes using disposable syringes Samples were maintained on ice until frozen on return to Australia Dissolved inorganic nutrient concentrations were determined using standard automated techniques ŽRyle et al., 1981; Ryle and Wellington, 1982 DON and DOP concentrations were determined on subsamples allowed to digest overnight in a LaJolla UV photooxidizer DOC samples Ž10 ml were acidified with 100 ml HCl to liberate inorganic carbon and kept cool on ice in a container DOC concentrations were determined on a Shimadzu TOC-5000 analyzer Blanks were run concurrently on double-distilled water Samples for particulates were obtained by filtering onto pre-combusted and weighed GFC filters, which were individually air-dried and wrapped in foil after filtering 100–300 ml of pond water were filtered, depending on the suspended load in each pond Before analysis, the filters were dried Ž808C for days., weighed and crushed to powder for carbon and nitrogen analysis on an Antek Model 707C nitrogen analyzer in tandem with a Beckman carbon analyzer TOC concentrations were obtained by reacting filter subsamples with hydrochloric acid followed by analysis on a solid-sample Shimadzu analyzer Accuracy and precision were determined by analysis of certified reference materials digested with a blank filter; accuracy was within certified limits and precision was within 3–5% In October 1996 and November 1997, nutrient samples were taken at the front, mid and rear of each pond and in the river adjacent to Ponds 12 and 23 In May 1997, samples were taken at the same pond locations, but no samples were taken in adjacent river water Samples from the ponds were taken when the sluice gates were closed or when water exchange was nearly complete In May 1997, particulate samples were taken from the three ponds near the sluice gates during both harvesting and subsequent recruitment stages Temperature, salinity, dissolved oxygen and pH in pond waters were measured using Hydrolabq DataSonde dataloggers The dataloggers were calibrated as per factory Fig Map of the location of Ponds 12, 22 and 23 within Enterprise Tam Giang III and the enterprise position in the Mekong delta of southern Vietnam 126 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 instructions In October 1996, vertical profiles were measured in the ponds and adjacent river water and a datalogger was deployed for one diel cycle in each pond and in the river adjacent to Pond 23 Deployment was similar in November 1997, but including Pond 22 In May 1997, dataloggers were deployed for days in Ponds 12 and 23 only during grow-out periods In each location, dataloggers were secured in situ Žat mid-depth by means of a rope tied onto a pole placed perpendicular to the pond channel 2.3 Primary production and oxygen flux measurements Primary production was estimated by 14 C uptake following isotope preparation and labware cleaning outlined by Fitzwater et al Ž1982 Samples were taken of surface water in each pond by allowing the incubation bottles to fill slowly 100 ml water samples were spiked with mCi of 14 C bicarbonate Žspecific activity: 57 mCi mmoly1 and incubated in situ by placing the bottles in clear acrylic tubes which were weighed and deployed such that they floated just below the surface Three light and three dark polycarbonate bottles were run for each location Samples were incubated for h, between 1000 and 1400 h, and incubations terminated by filtration onto GFF filters After filtration, filters were placed in sterile plastic scintillation vials and acidified with 100 ml HCl until counting on return to the laboratory Internal standards were used for quench correction Primary production rates were calculated following Parsons et al Ž1984 Production on a daily areal basis Žmg C my2 dy1 was estimated by dividing by the water depth and multiplying by assuming that one-half the total daily irradiance occurs during 1000–1400 h Primary production measurements were made at the front, mid and rear areas of Ponds 12 and 23 and adjacent river locations in May 1997 and at all three ponds and adjacent to Pond 22 in November 1997 Gross primary production and respiration were estimated from changes in oxygen concentrations in pond and river water incubated in triplicate light and dark bottles ŽRobertson et al., 1993 At each location, replicate light and dark bottles were gently filled with surface water and a magnetic stirrer bar placed into each bottle An oxygen probe was sealed into the top of each bottle containing a specially made cap, then placed into a magnetic stirrer within a large plastic bin containing ambient water The stirrers were run using propellers powered by a battery-operated water pump The water in each bottle was stirred continuously Each oxygen probe was connected to a TPSq Model WP82 oxygen meter The electrodes were calibrated as per factory instructions The experiments were usually run from 0900 to 1500 h The dark bottles were usually allowed to incubate longer Žusually to 1800 h depending upon rate of oxygen consumption The incubations were conducted in full sunlight Bin water was replenished as much as possible to mimic in situ temperatures Light readings were taken at 10–20 intervals using a hand-held light meter with quantum sensor Rates of gross primary production and respiration were calculated following the formulae in Parsons et al Ž1984 The data were converted to carbon equivalents assuming an RQ and a PQ of Oxygen flux experiments were conducted only in the front area of each pond and in river water adjacent to Ponds 12 and 23 during the three expeditions D.M Alongi et al.r Aquaculture 175 (1999) 121–141 127 2.4 Bacterioplankton measurements Quadruplicate 20 ml samples for bacterioplankton abundance were taken immediately below the surface at each location using a 50-ml sterile plastic syringe Each sample was gently decanted into a sterile glass scintillation vial and fixed with 500–1000 ml filtered, buffered formalin Samples were kept cool and dark until return to the laboratory Bacteria were stained and counted using the DAPI method ŽPorter and Feig, 1980; Velji and Albright, 1993 Bacterial abundance were converted to biomass assuming 25 fg C celly1 ŽBell, 1993 Bacterioplankton production was estimated via incorporation of tritiated thymidine ŽBell, 1993 Briefly, 15 ml water samples Ž n s collected in the same manner as for cell counts were spiked with 18.44 ml of undiluted H-thymidine Ž41.0 Ci mmoly1 specific activity to give a final concentration of ; 30 nM Tdr The samples were incubated for 25 in clear acrylic tubes in the same manner as for the 14 C Fig Vertical profiles of temperature, salinity, dissolved oxygen and pH in shrimp Ponds 12 and 23 ŽP12 and P23 Žmiddle and front sections of each pond., and in Song Dam Chim river ŽR12 and R23 adjacent to both ponds, October 1996 See Fig for details of locality 128 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 measurements Triplicate blanks containing 1.1 ml formaldehyde were run concurrently Initial isotope dilution and time course experiments indicated that a final Tdr concentration of 30 nM and incubation time of 25 were optimal for the productivity measurements After incubation, samples were processed to estimate isotope incorporation into bacterial DNA as described in Bell Ž1993 Recovery efficiency was assumed to be 100%; internal standards were used to correct for quenching.The rate of thymidine incorporation was converted to carbon production assuming a thymidine conversion factor of = 10 18 cells moly1 and 25 fg C celly1 ŽBell, 1993 Specific daily growth rate Ž m was calculated by dividing the mean production estimate by the mean standing crop Bacterial measurements were made in May and November 1997 at the front, mid Fig Ža Diurnal cycles of water temperature and salinity in shrimp Pond 12 over days of the grow-out cycle in May 1997 Curves are means integrated by h intervals of datalogger readings taken every 20 See Fig for details of locality Žb Diurnal cycles of dissolved oxygen and pH in Pond 12 over days of the grow-out cycle in May 1997 Curves are means integrated by h intervals of datalogger readings taken every 20 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 129 and rear areas of Ponds 12 and 23 and adjacent river water; Pond 22 and adjacent river water was sampled only in November 1997 2.5 Data analysis Comparisons among sites and sampling times were made using either one- or twofactor ANOVA techniques ŽSokal and Rohlf, 1995 Any significant site or season effects were further examined using the Student–Newman–Keuls test Data were log-transformed if Fmax tests indicated heteroscedasticity Linear regression was used to calculate rates of oxygen production and consumption Correlations between variables were calculated using Pearson product–moment correlation tests on nontransformed data Level of significance was accepted at P s 0.05 Results 3.1 Physicochemical characteristics Vertical profiles of physicochemical characteristics indicated stratification of pond waters ŽFig Salinity did not vary significantly, but temperature and dissolved oxygen decreased significantly with increasing water depth Vertical change was not as abrupt in Song Dam Chim reflecting tidal intrusions of warmer, more saline South China Sea water The vertical profiles of dissolved oxygen suggest that respiration exceeded photosynthesis below the upper 5–10 cm Some of the pH profiles Že.g., R23, P23 Front Table Diurnal range of temperature, pH, salinity and dissolved oxygen Žmg ly1 in shrimp pond waters in October 1996, May 1997 ŽPond 23 only and November 1997, Enterprise Tam Giang III, Mekong delta, Vietnam Variable Pond 12 Pond 22 Pond 23 October November October November October May November 8C Day Night 29.1–30.1 NA 27.5–31.0 NA 28.5–32.3 29.6–32.1 28.8–33.5 NA 28.2–31.6 30.7–31.6 28.0–32.7 27.3–30.2 28.2–31.7 NA pH Day Night 5.0–5.5 NA 6.4–6.6 NA 4.7–5.3 4.5–5.0 5.5–6.2 NA 5.0–5.6 5.0–5.8 6.9–7.4 6.8–7.7 6.8–7.1 NA Salinity Day Night 16.0–16.4 NA 19.4–20.1 NA 13.8–17.2 15.5–17.1 16.0 NA 16.1–17.0 16.7–17.0 30.5–34.1 30.4–34.0 16.5–16.6 NA O2 Day Night 3.8–5.4 NA 0.9–3.2 NA 3.2–7.5 2.1–5.4 1.6–2.8 NA 0.2–0.5 0.3–0.7 0.8–9.9 0.6–10.9 2.3–4.7 NA Data for Pond 12 in May 1997 is in Fig D.M Alongi et al.r Aquaculture 175 (1999) 121–141 130 and P12 Mid contradict the DO profiles, but pH was within a comparatively narrow range ŽFig Pond and adjacent river waters were acidic, with pH ranging from 4.5–7.6 with most measurements ranging from 4–6 Highest pH levels were measured in Ponds 12 ŽFig 3b and 23 ŽTable in May 1997 and in November 1997, when most values were ) 6.5 and as high as 7.7 The 9-day monitoring in May 1997 of physicochemical characteristics in Ponds 12 and 23 showed diel cycles of temperature, dissolved oxygen and pH dampened on Day by intense rainfall ŽFig 3; Pond 12 only as reflected in lower temperatures, DO concentrations and salinity There were significant seasonal differences in salinity, dissolved oxygen and pH Pond and river water temperatures showed greater diel than seasonal change; temperatures at all sites ranged from 27–338C Salinity was greatest at all sites in May 1997, ranging from 26.5–34.4 ŽFig 3a, Table Salinity was lower in Table Variations in particulate nutrients, mean molar C:N ratios and suspended solids among sites in shrimp ponds and adjacent river, ŽA October 1996, ŽB May 1997 and ŽC November 1997, Enterprise Tam Giang III, Mekong delta, Vietnam Pond and site Nutrients TC TOC TN ŽA October 1996 P23F P23M P23R 1.64"0.52 1.07"0.34 1.78"0.38 1.05"0.12 0.84"0.13 1.74"0.36 0.41"0.09 0.20"0.09 0.28"0.04 3.0 4.9 7.4 65.7"6.3 88.7"7.3 115.5"23.5 P12F P12M P12R 0.36"0.21 0.43"0.01 0.66"0.21 0.33"0.10 0.40"0.0 0.48"0.02 0.05"0.03 0.06"0.01 0.09"0.02 7.7 7.8 6.2 68.7"12.0 65.1"4.5 74.4"13.9 P22F P22M P22R 0.51"0.10 0.98"0.0 1.26"0.10 0.51"0.05 0.57"0.10 1.05"0.08 0.07"0.01 0.14"0.02 0.21"0.02 8.5 6.3 5.8 81.9"3.0 92.4"11.3 95.4"3.0 ŽB May 1997 P23 a P12 a P22 a 1.80"0.4 1.30"0.10 1.90"0.5 1.50"0.2 1.20"0.2 1.60"0.5 0.25"0.02 0.15"0.01 0.18"0.05 7.0 9.3 10.3 164.0"7.7 184.4"12.5 177.7"10.8 ŽC November 1997 P23 a P12 a P22F P22M P22R Ri12 Ri23 1.60"0.10 0.65"0.50 1.51"0.02 0.70"0.05 0.65"0.05 3.40"0.90 1.66"0.10 1.58"0.05 0.56"0.02 1.28"0.02 0.50"0.02 0.56"0.02 1.81"0.02 1.51"0.10 0.17"0.04 0.05"0.01 0.12"0.03 0.05"0.03 0.06"0.02 0.13"0.01 0.15"0.02 10.8 13.1 12.8 11.7 12.9 16.2 11.7 74.9"4.5 175.8"3.2 115.5"2.1 83.3"2.1 78.7"5.1 286.3"8.3 310.3"3.7 a CN SS Values from front, middle and rear locations combined due to lack of significant differences Values are means Ž"1 SE Particulate nutrients are mg ly1 and suspended solids are mg DW ly1 Abbreviations: TC s total carbon; TOC s total organic carbon; TNs total nitrogen; SSssuspended solids; P s pond; F s front; M s mid; R s rear; Ri s river adjacent to pond no D.M Alongi et al.r Aquaculture 175 (1999) 121–141 131 October 1996 and in November 1997 ŽTable 1., ranging from 13.8–20.1 Dissolved oxygen concentrations were normally low in both ponds and the adjacent river water, with most measurements - mg ly1 Values - 0.7 mg ly1 were recorded in Pond 23 in October 1996 ŽTable 1.; DO levels in Pond 23 were significantly greater in May 1997 and in November 1997 ŽTable In contrast, highest daytime DO levels in Pond 12 were measured in October 1996 ŽTable 1.; DO concentrations in this pond oscillated between 0.6 and 3.8 mg ly1 during the other two sampling periods DO levels in Pond 22 were significantly lower in November 1997, ranging from 1.6–2.7 mg ly1 ; values oscillated between 2.1–7.5 mg ly1 during the other sampling periods 3.2 Nutrients Dissolved and particulate nutrient concentrations varied significantly among sites and seasons, but highly significant location= time interactions in the statistical tests suggest that differences among locations were inconsistent over time Within-site variability was large ŽTables and For example, there were significant differences in POC concentrations among within-pond locations in the three ponds in October 1996, but such was not the case for all three ponds in May 1997 ŽTable There were no clear, Table Variations in dissolved nutrients ŽmM among sites in shrimp ponds and adjacent river, ŽA October 1996, ŽB May 1997 and ŽC November 1997, Enterprise Tam Giang III, Mekong delta, Vietnam Pond and site Nutrients PO43y NHq y NOy qNO ŽA October 1996 P23 0.13"0.14 P12 0.07"0.02 P22 0.10"0.03 Ri12 0.10"0.02 Ri23 0.13"0.06 1.8"2.2 0.3"0.2 1.8"0.03 0.5"0.1 0.7"0.1 5.0"4.1 2.2"0.2 14.2"10.3 8.5"0.73 13.6"6.63 ŽB May 1997 P23 P12 P22 1.4"2.3 0.7"0.5 2.9"4.1 0.7"0.2 0.3"0.1 0.3"0.1 0.3"0.1 0.3"0.1 0.5"0.1 0.6"0.04 0.27"0.19 0.41"0.17 0.16"0.17 ŽC November 1997 P23 0.05"0.60 P12 0.06"0.03 P22F 0.06"0.03 P22M 0.02"0.01 P22R 0.04"0.01 Ri12 0.13"0.10 Ri23 0.07"0.02 DOC DON DOP 483"25 317"17 358"17 317"8 375"17 17.6"8.3 7.1"2.5 8.5"1.3 11.3"2.0 10.1"0.8 0.69"0.31 0.37"0.06 0.59"0.07 0.53"0.05 0.58"0.05 28.4"3.7 19.6"5.9 17.7"5.7 617"16 1025"25 867"83 16.4"4.1 17.8"12.8 24.2"15.5 0.72"0.14 0.52"0.20 0.58"0.10 0.2"0.04 0.2"0.16 0.2"0.04 0.2"0.0 0.0 0.2"0.04 14.3"2.1 433"50 542"42 517"58 542"33 591"8 600"25 433"42 14.2"1.3 15.8"3.1 12.6"2.9 10.4"0.8 9.0"1.6 20.1"5.1 14.3"2.1 1.1"0.11 1.2"0.2 1.0"0.3 1.0"0.4 0.8"0.1 1.8"0.2 1.0"0.2 Values are summed means Ž"1 SE of replicates from front, middle and rear areas of each pond due to lack of significant location differences Abbreviations as in Table 132 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 Fig Rates of primary production Ž14 C uptake in surface waters in various shrimp pond and adjacent river locations in May and November 1997 Abbreviations: P23F s Pond 23 Front; P23Ms Pond 23 Mid; P23R s Pond 23 Rear; R23s river adjacent to Pond 23, etc Note that P23R and P12F were not sampled in May and November, respectively Values are means and vertical bars depictq1 SE Asterisks denote significant differences between seasons within location See Fig for details of locality consistent patterns for nutrients between ponds and adjacent river water over time, but some patterns emerged: Ž1 atomic C:N ratios were highest among ponds in November 1997 with the ratio ranging from 10.8–16.2; in the other seasons, the ratio ranged from 3.0–10.3; Ž2 suspended sediment loads were generally greatest in May 1997, ranging from 164–184 mg DW ly1 ; SS loads and POC concentrations in the ponds were lower than in adjacent river water ŽNovember 1997.; and Ž3 DOP and DON concentrations were significantly greater than dissolved inorganic N and P 3.3 Primary production and respiration Rates of primary production measured by 14 C uptake ranged from - to 94 mg C m dy1 ŽFig In May, primary production rates in Ponds 12 and 23 were greater y3 Fig Rates of gross primary production Žoxygen flux and Pr R ratio Žbottom graph in various shrimp pond and river locations, October 1996, May and November 1997 Values are means and vertical bars depict"1 SE Numbers above each set of Pr R ratios for each location are means of seasonal ratios See Fig for details of locality D.M Alongi et al.r Aquaculture 175 (1999) 121–141 133 134 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 than in adjacent river water, but in November 1997, this was true only for Pond 23; rates of primary production in some, but not all, sites within Ponds 12 and 22 were greater than in Song Dam Chim ŽFig Rates of primary production were not significantly different between Ponds 23 and 12 in different seasons; but in November 1997, primary production was greater in Pond 22 than in the other two ponds, except in comparing Pond 22 Front with P23 Rear ŽFig Seasonally, rates of primary production were significantly greater in May than in November in Ponds 23 and 12 Differences in river water were significantly different only adjacent to Pond 23 ŽFig Rates of primary production correlated only with salinity Ž r s q0.78 Rates of primary production measured by oxygen flux were significantly greater than measured by 14 C uptake Rates of gross primary production ranged among locations from y184 to 2697 mg C my3 dy1 ŽFig 5, top There were no consistent differences within sites among seasons Rates of oxygen consumption ŽFig ranged from 60 to 3783 mg C my3 dy1 There were significant differences among sites and within sites by season In October 1996, there were no significant differences among sites, but in May 1997, respiration rates were greater in Ponds 22 and 23 than in Pond 12 In November 1997, oxygen consumption rates were greatest in river water and greater in Pond 22 than in Pond 23, Fig Rates of pelagic respiration Žoxygen consumption in various shrimp pond and river locations, October 1996, May and November 1997 Values are means and vertical bars depictq1 SE Asterisks denote significant differences between seasons with location See Fig for details of locality D.M Alongi et al.r Aquaculture 175 (1999) 121–141 135 but equivalent to rates in Pond 12 Seasonal differences within sites were not consistent ŽFig By difference, rates of net primary production ranged from to 1131 mg C my3 dy1 , but were usually zero There were no clear, consistent differences among sites or seasons in rates of net primary production, owing to large variability of oxygen fluxes among replicates The range of PrR ratios varied widely ŽFig 5, bottom., ranging from y0.45 to 3.4 The overall mean PrR of each site was - 1, ranging from 0.2 to 0.9, with a grand mean of 0.7 Rates of gross primary production and respiration did not correlate significantly with any other measured variables, but the PrR ratios correlated inversely with suspended solids Ž r s y0.83 and total organic carbon Ž r s y0.85 3.4 Bacterioplankton dynamics Bacterioplankton numbers were significantly greater in May than in November in Ponds 23 and 12, and greater in Pond 22 than at all other locations in November 1997 ŽFig There were no consistent differences between Ponds 12 and 23 between Fig Bacterioplankton densities at each shrimp pond and river location with season Values are meanq1 SE Asterisks denote significant differences between seasons within location Abbreviations as in Fig and H s high tide; L s low tide See Fig for details of locality 136 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 seasons Bacterial abundance was lower in river water than in Ponds 12 and 23 ŽMay only In November, bacterial numbers were higher in Pond 22 than in Song Dam Chim, but only P23 Front was significantly higher than densities in river water in May 1997 Rates of bacterioplankton production ranged from 2.5 to 297.3 mg C my3 dy1 ŽFig Productivity was significantly greater in Pond 22 than at the other sites in November 1997 There was only one difference in production rates between Ponds 12 and 23 in November, with slowest rates at Pond 23 Front In May, rates of bacterioplankton production were greater within Pond 23 compared to their within-pond counterparts in Pond 12 Production rates were usually greater in ponds than in adjacent river water ŽFig Specific growth rates of bacterioplankton ŽFig were greater in November than in May 1997, except at Pond 23 Front and Rear Growth rates ranged from 0.005 to 0.072 in May 1997, indicating turnover rates of 14 to 200 days In November, growth rates ranged from 0.003 to 0.66, equating to turnover times of 1.5 to 333 days Averaging seasons and within-pond locations, grand mean growth rates were fastest in Pond 22 Ž m s 0.29 followed by Pond 12 Ž m s 0.23 and Pond 23 with the slowest average rates of bacterial growth Ž m s 0.08 These grand means equate to turnover times of 3.4, 4.3 Fig Bacterioplankton production at each shrimp pond and river location with season Values are meanq1 SE Asterisks denote significant differences between seasons within location Abbreviations as in Fig and H s high tide; L s low tide See Fig for details of locality D.M Alongi et al.r Aquaculture 175 (1999) 121–141 137 Fig Specific growth rates of bacterioplankton at each shrimp pond and river location with season Values are means Abbreviations as in Fig and H s high tide; L s low tide See Fig for details of locality and 13 days, respectively Bacterial abundance, production and growth rates did not correlate with any of the other measured variables Discussion The dynamics of plankton communities and water quality characteristics in aquaculture ponds have been measured almost exclusively from semi-intensive and intensive ponds ŽEgna and Boyd, 1997 The sparse data available from extensive farms in the tropics indicate low primary and faunal productivity ŽOlah et al., 1986; see reviews of Fast and Lannan, 1992 and Diana et al., 1997 Our data support the notion that extensive pond systems are characterized by low primary productivity—shrimp yields in Ponds 12, 22 and 23 average only 390, 1166 and 380 kg hay1 yry1 , respectively ŽClough and Johnston, 1997 It is probable, however, that in addition to low food inputs Žlow primary productivity, no fertilizers or supplementary foods and inadequate management practices, several interacting physical, chemical and biological characteristics play a crucial role in maintaining poor shrimp yields in these ponds An earlier and more extensive survey of 138 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 shrimp ponds in the Mekong delta ŽBinh et al., 1997 found that bottom soil pH, alkalinity, gate width ratio, pond age, flooding levels and phosphate significantly correlated with shrimp yield Our data from Ponds 12, 22 and 23 indicate several common characteristics that likely play an important role in limiting shrimp production: Ž1 low net primary productivity due to light limitation; Ž2 moderate rates of bacterioplankton production; Ž3 rates of pelagic respiration4 pelagic primary production; Ž4 high suspended solid loads; Ž5 low and very variable pH and dissolved oxygen levels; and Ž6 very variable and high dissolved inorganic and organic nutrient concentrations These conditions are endemic to the region as there are no consistent differences in water quality and pelagic carbon production and consumption between ponds and adjacent river water Many of these pelagic characteristics are typical of estuarine waters in Southeast Asia ŽHinrichsen, 1998 The impact of Typhoon Linda appeared to be minimal with no clear inhibition or stimulation of plankton activity In this study, there were few significant correlations between variables The lack of clear relationships is likely to be due more to different sampling techniques, small-scale heterogeneity, and the lack of clear or consistent differences in most measured variables among sampling locations, than to a lack of real relationships The rates of primary productivity correlated with salinity, indicating either a negative impact of freshwater inputs into the ponds, or that alkalinity was a factor limiting primary production Boyd Ž1990 and Diana et al Ž1997 showed that in unfertilized fish ponds, phytoplankton productivity is limited by total alkalinities - 120 mg ly1 We measured concentrations of ÝCO in near-bottom water of the ponds in November ŽAlongi et al., 1999.; values ranged from 21.3–23.1 mg ly1 suggesting that alkalinity may have been limiting Assuming that the well-known salinity–alkalinity relationship ŽParsons et al., 1984 is valid for these waters, alkalinities were likely to have been greatest in May; higher rates of primary production were measured in May than in November This is reasonable considering the low pH of these waters: at pH 4, most inorganic carbon is in the form of CO , rather than bicarbonate Alkalinity alone may not be limiting ŽDiana et al., 1997 as other dissolved constituents may limit carbon fixation Rates of primary productivity did not correlate with the high suspended solid loads or with nutrient concentrations It is unlikely that primary production was nutrient-limited in these ponds, considering the high concentrations compared to other tropical coastal waters ŽFurnas, 1992 It is more likely that rates of carbon fixation were light-limited, given the high suspended particle loads Plankton studies in tropical estuarine waters, including aquaculture ponds, have found a negative impact of high suspended loads on phytoplankton production ŽTeichert-Coddington et al., 1992; Ezenwa et al., 1994 The high suspended loads and dissolved and particulate nutrient concentrations are characteristic of coastal Asian waters and are due to a variety of factors: high rates of river runoff, shoreline erosion, resuspension, heavy boat traffic, agricultural and forest runoff, and dumping of organic waste The diurnal cycles of temperature, dissolved oxygen and pH were somewhat typical of those measured in waters of other unfertilized, low alkalinity ponds ŽFast and Lannan, 1992 The cycles were, however, influenced by intense rainfall with noticeable declines in temperature, dissolved oxygen and salinity ŽFig The pH and absolute concentra- D.M Alongi et al.r Aquaculture 175 (1999) 121–141 139 tions of dissolved oxygen were within sublethal or lethal levels to shrimp ŽSeidman and Lawrence, 1985; Noor-Hamid et al., 1994 Dissolved oxygen concentrations of - mg ly1 and rates of DO change ) 1.5 mg ly1 hy1 and pH change of unit hy1 were measured in Pond 23 in October and May ŽTable and in Pond 12 in November ŽTable and May ŽFig 3b The exceedingly low levels in Pond 23 in October 1996 coincided with a disease outbreak suggesting that the shrimp were stressed, increasing their susceptibility to disease The low pH of these waters is probably due to several factors, including the naturally acidic nature of many Asian waterways, leaching from pond soil and adjacent acid sulfate soils, lateral inputs from interstitial mangrove water, inputs of acid rain, and excavation and subsequent oxidation of pond bottom deposits Low dissolved oxygen concentrations may reflect poor water circulation, and the dominance of heterotrophic over autotrophic processes Ži.e., net heterotrophy The vertical stratification of temperature, oxygen and pH ŽFig indicates poor circulation and stagnation, slow diffusive exchange with the atmosphere, and the influence of benthic andror near-bottom heterotrophic processes The impact of these processes is likely to be enhanced by the low pond water:sediment ratios that reflect the narrow and shallow nature of these ponds The abundance, productivity and growth rates of pond bacterioplankton were within the range of values measured in other coastal waters Žsee review of Ducklow and Shiah, 1993., but less than those measured in semi-intensive ponds ŽMoriarty, 1986, 1997; Costa-Pierce and Craven, 1987; Jana and De, 1993; Allan et al., 1995; Guo et al., 1997 Rates of bacterial carbon production and pelagic respiration Žpresumably mostly microbial exceeded rates of primary productivity The rates of primary production as measured by 14 C uptake and by the light–dark bottle method did not agree well at specific locations and sampling dates, but both sets of measurements agree in that net production was frequently low or zero This finding supports earlier measurements of low Ž0.0–0.28 mg ly1 chlorophyll a concentrations ŽClough and Johnston, 1997 The high gross primary production rates as measured by oxygen flux reflected, in most instances, the rapid rates of respiration Although the PrR ratio exceeded unity in some locations, averaging over seasons indicates that the PrR ratio of each pond and river position was - 1, with a grand mean of 0.7 ŽFig This finding supports the 14 C primary production and bacterioplankton production data indicating that these waters are net heterotrophic The ratio of bacterioplankton production to primary production Ž14 C uptake varies among pond and river locations, ranging from 47–238% in May 1997 and from 59–980% in November 1997 The methods used to measure bacterioplankton and primary production are subject to considerable uncertainties, such as conversion efficiencies and carbon content, so these rates are not absolute Even ignoring the efficiency at which pelagic bacteria convert organic matter into biomass Žgrowth efficiencies of pelagic bacteria vary widely., the ratios imply that sources of organic matter other than pelagic autotrophs are required to support bacterial growth and production Recycling within the pelagic food web may meet some of the bacterial demand andror some proportion of the bacterioplankton community may originate from the benthos via resuspension, but some portion of bacterioplankton production must be fueled by 140 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 allochthonous carbon These sources may include benthic production, litter or subsurface inputs from adjacent mangrove forests, tidal exchange, or some combination of these sources In any case, rates of bacterioplankton production far exceed rates of primary production in these extensive ponds: an ecological characteristic of unproductive aquatic systems Ždel Giorgio et al., 1997 Acknowledgements This project was funded by a grant ŽPN 9412 from the Australian Centre for International Agricultural Research, and by the Australian Institute of Marine Science We thank the Ministry of Fisheries, Vietnam, for their support of this project, and the Director and staff of the Research Institute for Aquaculture No and Tam Giang III for their help Mrs B.G Huyen of RIA No is thanked for her unstinting help and care We thank Mr Hoang, Mr Hung and Mrs Anh for their hospitality and permission to sample on their farms C Payn, F Tirendi, L Trott and J WuWon provided lab and field assistance, and B Smith and B Clough are thanked for their encouragement and support D McKinnon, M Kenway and F Tirendi critically reviewed the manuscript Contribution No 942 from the Australian Institute of Marine Science References Allan, G.L., Moriarty, D.J.W., Maguire, G.B., 1995 Effects of pond preparation and feeding rate on production of Penaeus monodon Fabricius, water quality, bacteria and benthos in model farming ponds Aquaculture 130, 329–349 Alongi, D.M., Tirendi, F., Trott, L.T., 1999 Rates and pathways of benthic mineralization in extensive shrimp ponds of the Mekong delta, Vietnam Aquaculture, in press Bell, R.T., 1993 Estimating production of heterotrophic bacterioplankton via incorporation of tritiated thymidine In: Kemp, P.F., Sherr, B.F., Sherr, E.B., Cole, J.J ŽEds , Handbook of Methods in Aquatic Microbial Ecology CRC Press, Boca Raton, pp 495–503 Binh, C.T., Phillips, M.J., Demaine, H., 1997 Integrated shrimp–mangrove farming systems in the Mekong Delta of Vietnam Aquacult Res 28, 599–610 Boyd, C.E., 1990 Water Quality in Ponds for Aquaculture Agriculture Experiment Station, Auburn University, Auburn, 482 pp Clough, B.F., Johnston, D.J., 1997 Project PN9412: Mixed Shrimp Farming–Mangrove Forestry Models in the Mekong Delta Annual Report to the Australian Centre for International Agricultural Research Australian Institute of Marine Science, Townsville, 123 pp Costa-Pierce, B.A., Craven, D.B., 1987 Estimating microbial production and growth in aquaculture ponds using rates of RNA and DNA synthesis Aquaculture 66, 69–78 del Giorgio, P.A., Cole, J.J., Cimbleris, A., 1997 Respiration rates in bacteria exceed phytoplankton production in unproductive aquatic systems Nature 385, 148–151 Diana, J.S., Szyper, J.P., Batterson, T.R., Boyd, C.E., Piedrahita, R.H., 1997 Water quality in ponds In: Egna, H.S., Boyd, C.E ŽEds , Dynamics of Pond Aquaculture CRC Press, Boca Raton, pp 53–72 Ducklow, H.W., Shiah, F.-K., 1993 Bacterial production in estuaries In: Ford, T.E ŽEd , Aquatic Microbiology Blackwell, Oxford, pp 261–287 Egna, H.S., Boyd, C.E ŽEds , 1997 Dynamics of Pond Aquaculture CRC Press, Boca Raton, 437 pp Ezenwa, B., Uzukwu, P., Anyanwu, P., 1994 Studies on primary productivity in a tidal fish farm at Ikoyi, Lagos J Aqua Trop 9, 1–7 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 141 Fast, A.W., Lannan, J.E., 1992 Pond dynamic processes In: Fast, A.W., Lannan, J.E ŽEds , Marine Shrimp Culture: Principles and Practices Elsevier, Amsterdam, pp 431–456 Fitzwater, S.E., Knauer, G.A., Martin, J.H., 1982 Metal contamination and its effect on primary production measurements Limnol Oceanogr 27, 544–551 Furnas, M.J., 1992 The behavior of nutrients in tropical aquatic ecosystems In: Connell, D.W., Hawker, D.W ŽEds , Pollution in Tropical Aquatic Systems CRC Press, Boca Raton, pp 29–65 Guo, X.Z., Fang, X.Z., Xie, J., Yu, T.B., Zhang, W.Y., 1997 Bacterial productivity in an intensely cultured high-yield fish pond using w3 Hx Tdr tracing analysis techniques In: Mathias, J.A., Charles, A.T., Baotong, H ŽEds , Integrated Fish Farming CRC Press, Boca Raton, pp 139–145 Hinrichsen, D., 1998 Coastal Water of the World: Trends, Threats, and Strategies Island Press, Washington D.C., 275 pp Jana, B.B., De, U.K., 1993 Management-induced variability of the bacterioplankton in fish farming ponds J Aqua Trop 8, 131–140 Lucien-Brun, H., 1997 Evolution of world shrimp production: fisheries and aquaculture World Aquacult 28, 21–35 Moriarty, D.J.W., 1986 Bacterial productivity in ponds used for culture of penaeid prawns Microb Ecol 12, 259–269 Moriarty, D.J.W., 1997 The role of microorganisms in aquaculture ponds Aquaculture 151, 333–349 Noor-Hamid, S., Fortes, R.D., Parado-Estepa, F., 1994 Effect of pH and ammonia on survival and growth of the early larval stages of Penaeus monodon Fabricius Aquaculture 125, 67–72 Olah, J., Sinha, V.R.P., Ayyappan, S., Purushothaman, C.S., Radheyshyam, S., 1986 Primary production and fish yields in fish ponds under different management practices Aquaculture 58, 111–122 Parsons, T.R., Maita, Y., Lalli, C.M., 1984 A Manual of Chemical and Biological Methods for Seawater Analysis Pergamon, Oxford, 173 pp Porter, K.G., Feig, Y.S., 1980 The use of DAPI for identifying and counting aquatic microflora Limnol Oceanogr 25, 943–948 Robertson, A.I., Daniel, P.A., Dixon, P., Alongi, D.M., 1993 Pelagic biological processes along a salinity gradient in the Fly River estuary, Papua New Guinea Contin Shelf Res 13, 205–224 Rosenberry, B., 1996 World Shrimp Farming Shrimp News International, San Diego, 167 pp Ryle, V.D., Wellington, J., 1982 Reduction column for automated determination of nitrates Analytical Laboratory Note No 19 Australian Institute of Marine Science, pp Ryle, V.D., Mueller, H.R., Gentien, P., 1981 Automated analysis of nutrients in tropical seawater Australian Institute of Marine Science Monograph Series OS-81-4, 24 pp Seidman, E.R., Lawrence, A.W., 1985 Growth, feed digestibility and proximate body composition of juvenile Penaeus Õannamei and Penaeus monodon grown at different dissolved oxygen levels J World Maricult Soc 16, 222–346 Sokal, R.R., Rohlf, F.J., 1995 Biometry Freeman, New York, 887 pp Teichert-Coddington, D.R., Green, B.W., Phelps, R.P., 1992 Influence of site and season on water quality and tilapia production in Panama and Honduras Aquaculture 105, 297–314 Velji, M.I., Albright, L.J., 1993 Improved sample preparation for enumeration of aggregated aquatic substrate bacteria In: Kemp, P.F., Sherr, B.F., Sherr, E.B., Cole, J.J ŽEds , Handbook of Methods in Aquatic Microbial Ecology CRC Press, Boca Raton, pp 139–142  ... crucial role in maintaining poor shrimp yields in these ponds An earlier and more extensive survey of 138 D.M Alongi et al.r Aquaculture 175 (1999) 121–141 shrimp ponds in the Mekong delta ŽBinh et... cycle in each pond and in the river adjacent to Pond 23 Deployment was similar in November 1997, but including Pond 22 In May 1997, dataloggers were deployed for days in Ponds 12 and 23 only during... are declining As in other countries experiencing declining yields, poor shrimp yields in the Mekong delta have been attributed to several factors, such as pollution, poor management and infrastructure,