Revisiting silicon budgets at a tropical continental shelf: Silica standing stocks in sponges surpass those in diatoms

Revisiting silicon budgets at a tropical continental shelf: Silica standing stocks in sponges surpass those in diatoms
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Revisiting silicon budgets at a tropical continental shelf: Silica standing stocks insponges surpass those in diatoms Manuel Maldonado, a,* Ana Riesgo, a Arianna Bucci, a and Klaus Ru¨tzler b a Centro de Estudios Avanzados de Blanes (CSIC), Girona, Spain b National Museum of Natural History, Smithsonian Institution, Washington, D.C. Abstract Most of the silicon (Si) in marine coastal systems is thought to recirculate under the biological control of planktonic diatoms. We challenge this view after comparing the biogenic silica (bSi) standing stocks contributedby communities of planktonic diatoms and benthic sponges in five habitats of an extensive continental shelf areaof the Mesoamerican Caribbean. In most habitats (outer reefs, patch reefs, sea grass beds, and mangroves), thesponge bSi stocks surpassed those of diatoms. Collectively, bSi in sponge communities was about 88.6 % of thetotal Si pool. Diatoms represented 4.2 % and ambient silicate about 7.2 % . Consequently, when constructing futureregional Si budgets in coastal areas, the Si standing stocks in sponge populations should be empirically examinedbefore deciding that their contribution to the total is negligible. In order to understand Si fluxes in coastal areaswhere sponges are relevant, we need additional empirical approaches to set the timescale of sponge bSi turnover,which appears to be substantially slower than that of diatom bSi. Silicate, a dissolved form of silicon (Si), is a major oceannutrient. It fuels primary production by enhancing growthof diatoms, which require silicate to construct theirskeletons of biogenic silica (bSi). Therefore, there isenormous interest in predicting the interplay betweensilicate and bSi budgets. Ever since the earliest modelsattempting to establish a general balance for Si in the ocean(Harriss 1966; Burton and Liss 1968; Calvert 1968),biological Si cycling has always been thought to revolvearound diatoms. Because diatoms are estimated to con-sume yearly nearly all the Si available in the upper ocean(about 240  3  10 12 Si mol) to build their bSi skeletons,silicon pools are thought to be under their control, withother Si-consuming organisms such as sponges, radiolari-ans, choanoflagellates, and silicoflagellates playing only anegligible role (Tre´guer et al. 1995; Ragueneau et al. 2000;Sarmiento and Gruber 2006). Since diatoms are short livedand, upon death, up to 50 %  of their frustules dissolvereadily as reusable silicate before sinking down to theaphotic ocean, it is thought that the largest Si stockrecirculates relatively rapidly in a diatom-driven uptake-dissolution loop, completing 39 uptake-dissolution eventsin about 400 yr (Nelson et al. 1995; Tre´guer et al. 1995).Nevertheless, while such a diatom-driven loop mayappropriately represent the Si cycling in open-ocean bluewater, it may not realistically reflect the situation on at leastsome continental shelves, which often are characterized byextended shallow bottoms densely populated by long-livedsiliceous (Si-consuming) sponges.In the few sponge species investigated at a populationscale so far, the Si standing stocks were surprisingly largewhen compared with those available as silicate in theambient water of the respective habitats (Ru¨tzler andMcintyre 1978; Maldonado et al. 2005). More importantly,upon death, the silica skeletal pieces of sponges (spicules)appear to be far more refractory to dissolution than diatomfrustules in both basic solutions and seawater (Hurd 1983;Hurd and Birdwhistell 1983; Maldonado et al. 2005). Suchdifferences in dissolution rates between spicules andfrustules cannot be explained only in terms of an assumedlylarger surface area of diatom frustules, given that spongespicules with very different surface areas dissolved equallyslowly in experimental conditions (Maldonado et al. 2005).Whatever the reason for a much slower dissolution of sponge bSi, the result is that populations of siliceoussponges on continental margins appear to function asbenthic Si traps that retard conversion of bSi into silicate(Maldonado et al. 2005).At first sight, these peculiarities of the Si route throughsponges may be perceived as merely anecdotal, because thecontribution of these organisms to the global Si budget incoastal areas is regarded to be negligible when comparedwith that of diatoms (Tre´guer et al. 1995; Ragueneau et al.2000; Sarmiento and Gruber 2006). However, given thatthe currently accepted model for the marine Si cycle hasemerged in the complete absence of quantitative data forsponges, we have decided to revisit empirically thewidespread assumption that the bSi standing stock insponges is negligible when compared with that of diatoms.The importance of sponges to Si cycling in continentalwaters has been established by the only study to datecomparing sponge and diatom contributions, which provedthat, contrary to expectations, sponges largely dominatedthe Si budgets in some Florida lakes (Conley and Schelske1993). Therefore, in an attempt to provide empirical data tosupport the widespread—but never tested—assumptionthat siliceous sponges are negligible when calculatingregional Si budgets in marine coastal areas, we havemeasured for the first time the relative contributions of sponges, planktonic organisms, and ambient silicate inseawater to the local Si stock of several major habitats of atropical continental shelf. * Corresponding author: Limnol. Oceanogr.,  55(5), 2010, 2001–2010 E 2010, by the American Society of Limnology and Oceanography, Inc.doi:10.4319/lo.2010.55.5.2001 2001  Methods Habitat charting—  The Smithsonian Institution’s CarrieBow MarineField Station,locatedattheBelize’s continentalshelf edge, was used as a logistic base for field work. TheBelizean section of the Mesoamerican Barrier Reef is 250 kmlong and runs north–south parallel to the mainland alongthe outer edge of a large, shallow carbonate shelf (Fig. 1).Because the shelf’s edge is distant from the mainland ( . 15 km, on average), its seaward region is by and largeunaffected by detrimental anthropogenic disturbances.Continental runoff brings in very limited (episodic) amountsof both dissolved nutrients and suspended materials, such asaluminosilicate-rich terrestrial clays, and has no substantialeffect on the predominant calcareous rock and sediments of theouter reef platform.Thisnearly pristine,nonoligotrophicenvironment consisted of four major habitats: (1) outer reefs(OR), developed along the external shelf edge, with a forereef facing the open ocean; (2) patch  +  back reefs (PR),developed at the leeside of the reef ridge, in the cuts betweenadjacent barrier-reef cays, or scattered across the lagoon; (3)island mangroves (MG), forming cays on the inner reef platform; and (4) sea grass meadows (SG), spread over thesandy bottom of the lagoon. After the four major habitatswere charted (OR, PR, SG, MG), a large bottom area(Fig. 1B) of sandy, poorly vegetated lagoon bottom was left,designated as soft bare bottom (BB). To estimate the bottomarea and average water depth at the different habitats, wecombined information from local marine charts, satelliteorthophotographs available in the public domain, and global positioning system (GPS) anddepth-sounder field data obtained during boat trips anddives (Fig. 1). For calculations of area and water volume inmangroves, we measured the length of shoreline populatedby red-mangrove trees ( Rhizophora mangle ); we estimatedthat sponge-bearing stilt roots occupied a 2-m-wide band,and that the characteristic mangrove plankton systemextended, on average, over the adjacent 5-m-wide band of water column. Measurement of seawater silicate and plankton bSi—  Wemeasured concentrations of silicate and plankton bSi in theseawater overlaying the benthic habitats. Because bothnutrient concentration and plankton abundance varyseasonally, we attempted to capture the two periods thatdiffered most by sampling in both winter (November– December 2005) and summer (July 2006). Water samplesfor nutrient analysis were collected by diver-operatedNiskin bottles from both middle depth and near bottom(i.e., within 0.5 m from bottom or mangrove roots) andpooled for the analyses ( n 5 6, per season and habitat;  n 5 4 per open-water reference values). We also sampledoceanic water at 12 m depth, 5 km seaward of the barrierreef (Fig. 1) over some 400 m bottom depth, which servedas a reference for comparison with continental shelf habitats. Water subsamples were transferred to HCl-cleaned, 20-mL polyethylene vials and refrigerated in thedark for 3 d before measuring concentration of molybdate-reactive silicate using a Bran-Luebbe, Transference Auto-mated Analysis Colorimetric System (TrAACS-2000).Samples for estimating the plankton bSi stock over thebenthic habitats were collected at various depths (2–15 m)using diver-operated Niskin bottles both in the morningand in the evening and pooled in the analyses tocompensate for the daily variability of plankton occurrencein each habitat ( n OR 5 20;  n PR 5 25;  n SG 5 22;  n MG 5 23).Known volumes of water (1–2 liters, depending on habitatand season) were filtered immediately after collectionthrough polycarbonate membrane filters (4.7 cm indiameter, 0.6- m m pore size) using a vacuum pump. Filters Fig. 1. Surveyed continental shelf area off Belize, CentralAmerica. (A) Orthophotograph of the studied shelf area and theadjacent open ocean, which is separated from the lagoon by thereef barrier. (B) Map with inset showing the position of theenlarged study area. Keys point out the location and relative sizeof the habitats. 2002  Maldonado et al.  were dried at 60 u C, folded, and stored in the refrigeratorfor 2 months, then subjected to a double, wet-alkalinedigestion to discriminate Si derived from biogenic andlithogenic sources (Ragueneau et al. 2005). After eachalkaline digestion of filters, silicate concentrations weredetermined by autoanalyzer. Aluminum concentrationswere determined using a colorimetric method (Grasshoff et al. 1983). Because lithogenic silica (lSi) was undetectablein nearly all our winter samples, the summer samples andthe final regional calculations were run without lSicorrection.Values of silicate concentration and planktonic bSi forthe BB habitat were estimated as the average of valuesmeasured for all other shelf habitats, except the peculiarmangrove environment. When required, silicate concentra-tions and bSi biomass were converted into Si biomass,according to the ratios of their respective molecular andatomic weights. Differences in rank-transformed data forboth silicate concentrations and plankton bSi concentra-tions were examined as a function of season (winter vs.summer) and habitat (OR, PR, SG, MG) by two-wayANOVAs, followed by subsequent, pairwise Student– Newman–Keuls (SNK) tests to identify groups responsiblefor significant differences in the main factors. Local Sistocks in silicate and plankton for each of the habitats werecalculated from data on bottom area, average water depth,and average Si content per unit volume for each of thesources.For the taxonomic study of microplankton (radiolarians,choanoflagellates, diatoms, silicoflagellates, dinoflagellates,and coccolithophorids), 100-mL subsamples ( n  5  3) of seawater per habitat and season were collected, fixed inLugol’s iodine solution, and stored in the refrigerator untiltaxonomic identification using an inverted compoundmicroscope. A detailed account of the composition andabundance of the major phytoplankton groups in thestudied habitats will be reported elsewhere (A. Bucciunpubl. data). Measurement of sponge b Si  —  To estimate the Si contentin the sponge communities, we identified taxonomically allsponges found in random sampling quadrats ( n  5  409),measured in vivo the volume of each siliceous individual ( n 5  1882), and calculated Si content from average Si values(mg Si cm 2 3 of living tissue) determined for each speciesthrough previous hydrofluoric acid (HF) desilicification.Aspiculate and calcareous sponge species were not consid-ered in those quantifications.The sponge communities at the OR, PR, SG, and MGhabitats were investigated by scuba diving and sampledusing random 1-m 2 polyvinylchloride quadrats, withvarious replication efforts depending on habitat size andheterogeneity of the sponge assemblages ( n OR 5 99;  n PR 5 64;  n SG 5 135;  n MG 5 111). In a conservative approach, thesandy, poorly vegetated lagoon bottom (BB) was assumedto lack sponges, although some siliceous species are knownto be adapted to this habitat (Wiedenmayer 1977; Ru¨tzler1997). In mangroves, we accounted for the sponge faunagrowing on the stilt roots inside 1-m 2 quadrats establishedat the water surface because the muddy mangrove bottoms,which often lacked sponges, were not reached by all roots.Safety rules limited the scuba survey on the outer fore reef to a maximum depth of 25 m. The volume (cm 3 ) of eachindividual was estimated using a combination of 20 and50 cm plastic rulers. The body shape was approximated toone or, more often, several geometric figures (spheres,ovals, solid cylinders, hollow cylinders, rectangular plates),and the linear parameters (length, width, diameters) weremeasured in situ to calculate volumes. To preventoverestimation, we applied a one-fourth reduction tovolume values calculated for each individual. To convertmeasured sponge volumes to bSi content, we filled a plasticmeasuring cylinder with sponge tissue, applying minimumcompression and proportional amounts of both ectosomaland endosomal regions. After drying pieces ( n  5  3 to 6,depending on availability) to constant weight at 60 u C, wedesilicified by immersion in 5 % HF for 5 h, then rinsed indistilled water three times for 2 min to remove all HF, anddried to constant weight again. We calculated bSi contentper unit sponge volume as the dry weight difference beforeand after desilicification. To prevent overestimation of skeletal bSi, we assumed that only 75 %  of the weightdifference contributes to sponge skeleton. Mean weight of bSi content per cubic centimeter sponge volume served asan estimate of total bSi content per species and bottom areain the various habitats. Differences in sponge bSi as afunction of habitat (OR, PR, SG, MG) were examined by aKruskal–Wallis analysis, followed by pairwise Dunn’s teststo identify groups responsible for the significant differencesin the main factor.To estimate sponge bSi, we preferred dissolving siliceousspicules by immersion in HF over the alternative approachof removing the organic tissue in boiling nitric acid toweigh the nondissolved residue. This was done for severalreasons: (1) Unlike HF, nitric acid treatment requiressample boiling, which in turn requires the use of heavy glasscontainers, thereby reducing the accuracy of subsequentweighing. (2) Skeletal content tends to be overestimatedwhen nitric acid is used because it does not eliminateforeign inorganic elements that sponges typically incorpo-rate in their tissue or skeletal organic fibers (sand grains,skeletal remains, etc.). (3) Boiling in nitric acid alsoproduces salts that mask skeletal content unless samplesare rinsed in distilled water, a tedious process that risksremoving sponge spicules. (4) Heating of glass containersusually causes changes in both container volume and mass,introducing an uncontrolled factor in the weighing process.(5) With HF at room temperature and low concentration,the main concern is that all spicules have been dissolvedduring the treatment. If spicules had not all been dissolvedin some tissue samples, the results would have led us tounderestimate the sponge skeletal mass, which in turnwould have merely made our estimates of sponge Si contentmore conservative. Results We surveyed a 21.7-km 2 continental shelf area locatedabout 15.5 km from the nearest mainland and includingthree small islands (Fig. 1). Our estimates of bottom Silicon budgets at a tropical continental shelf   2003  extension (m 2 ) and water-column volume (m 3 ) for each of the habitats (Fig. 1; Table 1) revealed that sandy unvege-tated bottoms (BB) and sea grass meadows (SG) were thebest represented habitats in terms of both area (51.5 % and43.5 % , respectively) and water volume (49.6 % and 42.3 % ,respectively), with only minor contributions ( ,  8 % ,collectively) by the remaining habitats (i.e., outer reef,patch reef, and mangroves).Silicate at the shelf averaged 3.6 6 0.6  m g L 2 1 (mean 6 SD,  n  5  56), with detectable differences between seasonsand habitats (Fig. 2A). Planktonic bSi content averaged113.9  6  99.7  m g L 2 1 ( n  5  91), again with differencesbetween seasons and habitats (Fig. 2B). The taxonomicstudy of plankton samples ( n 5 30; 100 mL each) revealed atotal of 210 species, with members of only two siliceousgroups (silicoflagellates and diatoms) and no choanoflag-ellate or radiolarian. Silicoflagellates of a single speciesappeared in three samples, for a total of only five cells.Diatoms consisted of 103 species totaling 14,153 cells,which accounted for 71.5 %  6  11.6 %  of all microphyto-plankton cells in winter and 87.5 %  6  20.5 %  in summer.The species  Thalassionema nitzschoides  and  Thalassionema frauenfeldii   and several  Chaetoceros  spp. dominated in bothwinter and summer.The quadrat survey ( n 5 409) of the sponge communitiesat the habitats revealed 67 siliceous species (Table 2)averaging 4.6 individuals m 2 2 , a biomass of 2.6 6 14.3 litersof living tissue m 2 2 , and a mean bSi content of 0.3  6 2.7 kg m 2 2 . There were large between-habitat differences inboth sponge biomass and bSi (Fig. 3). The skeletal contentof siliceous sponges of our tropical assemblage accountedfor 15–25 % of dry weight in some species, up to 50–65 % inothers, averaging 53.4 6 6.1 bSi mg per liter of living tissue(Table 2). In mangroves, the overabundant, midsize ‘‘firesponge,’’  Tedania ignis  (Fig. 4A) made the largest bSicontribution (0.1 6 0.8 kg m 2 2 ), while  Iotrochota arenosa , anew species discovered during our survey, was the majorbSi contributor (1  3  10 2 3 kg m 2 2 ) in sea grass beds(Fig. 4B). In reef habitats, the large and well-silicified‘‘giant barrel sponge’’  Xestospongia muta  (Fig 4F) and thecongeneric  Xestospongia rosariensis  respectively averaged0.5  6  3.2 and 0.1  6  0.9 kg bSi m 2 2 and were major bSicontributors, with large individuals of the former speciescontaining up to 28 kg bSi each.By integrating the average Si content of the threeinvestigated sources (i.e., seawater, plankton, and sponges;Table 1) over the bottom area and water depth of thevarious habitats, we calculated a regional budget of 290.7 3  10 3 kg Si, averaging 13.4  3  10 3 kg km 2 2 . Contrary towhat might be expected, diatoms did not dominate this Sistock (Table 1). Mangrove habitat, which occupied only0.2 % of the shelf area, contributed 1.5  3  10 3 kg (0.5 % ) tothe regional pool. This bSi derived largely (98.4 % ) from theskeletons of the 12 sponge species identified from red-mangrove stilt roots (Table 2; Fig. 4A); the diatomcommunity (67 species) accounted for only 0.4 % ; andsilicate in seawater for 1.1 %  (Table 1). Sea grass bedscovered nearly half (43.6 % ) of the shelf area butcontributed only 12.2 %  (35.6  3  10 3 kg) to the regional Sipool. The sponge fauna of this habitat consisted of 23       T    a      b      l    e      1 .     S   u   m   m   a   r   y   o    f    d    i   m   e   n   s    i   o   n   s   a   n    d    S    i   c   o   n    t   r    i    b   u    t    i   o   n   o    f    t    h   e   v   a   r    i   o   u   s   e   c   o    l   o   g    i   c   a    l   c   o   m   p   a   r    t   m   e   n    t   s   s    t   u    d    i   e    d   o   n    t    h   e    B   e    l    i   z   e   a   n   s    h   e    l    f .    D   a    t   a   a   r   e    b   o    t    t   o   m   a   r   e   a ,   a   v   e   r   a   g   e    d   e   p    t    h ,   a   n    d   e   s    t    i   m   a    t   e    d   w   a    t   e   r   v   o    l   u   m   e   o   v   e   r    t    h   e    h   a    b    i    t   a    t   s ,   a   s   w   e    l    l   a   s   m   e   a   n    S    i   c   o   n    t   e   n    t    (   m   g    )    i   n    d    i   s   s   o    l   v   e    d   s    i    l    i   c   a    t   e   a   n    d   p    l   a   n    k    t   o   n   o   r   g   a   n    i   s   m   s   p   e   r    l    i    t   e   r   s   e   a   w   a    t   e   r ,   m   e   a   n    S    i   c   o   n    t   e   n    t    (   m   g    )    i   n   s   p   o   n   g   e   s    k   e    l   e    t   o   n   p   e   r   s   q   u   a   r   e   m   e    t   e   r   o    f    h   a    b    i    t   a    t    b   o    t    t   o   m ,   a   n    d    t   o    t   a    l    S    i   c   o   n    t   e   n    t    (    k   g    )    i   n   e   a   c    h   s   o   u   r   c   e   s    t   u    d    i   e    d    (    i .   e . ,    d    i   s   s   o    l   v   e    d    i   n   s   e   a   w   a    t   e   r ,   p    l   a   n    k    t   o   n   o   r   g   a   n    i   s   m   s ,   a   n    d   s   p   o   n   g   e   p   o   p   u    l   a    t    i   o   n   s    ) .    O    R ,   o   u    t   e   r   r   e   e    f   s   ;    P    R ,   p   a    t   c    h   r   e   e    f   s       +     b   a   c    k   r   e   e    f   s   ;    S    G ,   s   e   a   g   r   a   s   s    b   e    d   s   ;    M    G ,   m   a   n   g   r   o   v   e   s   ;    B    B ,   s   a   n    d   y ,   p   o   o   r    l   y   v   e   g   e    t   a    t   e    d    l   a   g   o   o   n    b   o    t    t   o   m   w    i    t    h   o   u    t   n   o    t    i   c   e   a    b    l   e   s   p   o   n   g   e   s .    H   a    b    i    t   a    t    B   o    t    t   o   m   a   r   e   a    (    1    0        5    m        2     )    M   e   a   n    d   e   p    t    h    (   m    )    W   a    t   e   r   v   o    l   u   m   e    (    1    0         6    m        3     )    M   e   a   n    d    i   s   s   o    l   v   e    d    S    i    (   m   g    L    2         1     )    M   e   a   n   p    l   a   n    k    t   o   n    S    i    (   m   g    L    2         1     )    M   e   a   n   s   p   o   n   g   e    S    i    (   m   g   m    2        2     )    T   o    t   a    l    d    i   s   s   o    l   v   e    d    S    i    (    1    0        3     k   g    )    T   o    t   a    l   p    l   a   n    k    t   o   n    S    i    (    1    0        3     k   g    )    T   o    t   a    l   s   p   o   n   g   e    S    i    (    1    0        3     k   g    )    T   o    t   a    l    S    i    (    1    0        3     k   g    )    O    R    6 .    4    5    2    0 .    0    0    1    2 .    9    0    0 .    1    0    1    0 .    0    6    0    3    0    3 ,    6    7    9 .    7    5    7    1 .    3    0    2    0 .    7    7    0    1    9    5 .    9    3    8    1    9    8 .    0    1    0    P    R    4 .    8    9    1    0 .    0    0    4 .    8    9    0 .    0    9    4    0 .    0    4    4    7    8 ,    1    4    6 .    2    5    4    0 .    4    5    9    0 .    2    1    4    3    8 .    2    4    6    3    8 .    9    1    9    S    G    9    4 .    5    3    1    0 .    0    0    9    4 .    5    3    0 .    0    9    1    0 .    0    5    6    2 ,    3    0    6 .    8    1    6    8 .    5    7    1    5 .    2    5    9    2    1 .    8    0    6    3    5 .    6    3    5    M    G    0 .    4    3    3 .    0    0    0 .    1    3    0 .    1    2    9    0 .    0    5    6    8    6 ,    2    6    0 .    0    8    7    0 .    0    1    7    0 .    0    0    7    1 .    4    8    5    1 .    5    0    9    B    B    1    1    0 .    7    2    1    0 .    0    0    1    1    0 .    7    2    0 .    0    9    7    0 .    0    5    3    0 .    0    0    0    1    0 .    7    5    6    5 .    8    7    0    0 .    0    0    0    1    6 .    6    2    7    T   o    t   a    l    2    1    7 .    0    2    1    0 .    2    8    2    2    3 .    1    7    2    1 .    1    0    5    1    2 .    1    2    0    2    5    7 .    4    7    5    2    9    0 .    7    0    0 2004  Maldonado et al.  species, typically small individuals scattered at low density(Table 2; Fig. 4B,C). Yet it provided 61.2 % of the local Sistock, while the diatom community (59 species) andseawater silicate accounted for 14.8 % and 24 % , respectively(Table 1). Patch  +  back reefs occupied only 2.2 %  of thestudied shelf but contributed 38.9 3 10 3 kg (13.3 % ) to theregional Si pool (Table 1). Again, most of the Si in thesereef habitats (98.2 % ) was provided by the 22 identifiedsponge species, which occurred at relatively high density(Table 2; Fig. 4D). The 56-species diatom community(0.6 % ) and seawater silicate (1.2 % ) made only minutecontributions. The importance of sponges was even higheron outer reefs (Table 2; Fig. 4E,F). This habitat represent-ed only 2.9 % of the studied shelf area but contributed 198 3  10 3 kg Si to the regional Si pool (Table 1), by far thelargest level (68.1 % ) for all the habitats. About 98.9 %  of that Si came from the 40 identified sponge species, whereasdiatoms (42 species) and seawater silicate collectivelycontributed only 1.1 % . After charting the four majorhabitats (OR, PR, SG, MG), we examined the large (51 % )soft-bottom, barren area (BB). Despite its size, the BBhabitat contributed only a modest 5.6 % (i.e., 16.6 3 10 3 kgSi) to the regional Si pool (Table 1). In the absence of sponges, the Si stock on BB came largely from silicate inambient water (64.6 % ) and diatoms (35.4 % ).Altogether, about 88.6 % of the Si budget on the coastalshelf derived from the skeleton of living sponges, about7.2 % from silicate in the local water column, and only 4.2 % from the frustules of planktonic diatoms (Table 1). Discussion Our research has shown, contrary to expectations, thatsponges comprise the largest standing stock of bSi in thisregional pool, being clearly dominant in four out of the fivesublittoral systems investigated. Such a sponge dominancecannot be attributed to impoverished environments interms of either nutrients or phytoplankton. Both silicateand plankton bSi concentrations fell within the range of values measured for other nonoligotrophic, subtropical,and temperate coastal zones lacking large river plumes(Ragueneau et al. 2005) and were notably higher than thosetypically recorded from offshore subtropical systems(Brzezinski and Kosman 1996; Sarmiento and Gruber2006). The portion of the Mesoamerican shelf we studiedwas relatively shallow, favoring a low contribution of diatoms to the regional budget. Nevertheless, according tothe measured bSi concentrations (Table 1), the shelf depthwould have to average 220 m before the plankton bSi couldapproach the level of sponge bSi in the regional pool.Therefore, sponges would still play a relevant role in thiscoastal system, even if it was a very deep continental shelf with homogeneous diatom production through the entirewater column. Fig. 2. Summary of concentrations (mean 6 SD) of (A) ambient seawater silicate and (B) planktonic bSi and their respective two-way ANOVAs showing differences in rank-transformed data as a function of season (winter vs. summer) and habitat (OR, outer reefs;PR, patch  +  back reefs; SG, sea grass beds; MG, mangroves). (A) There was a significant ‘‘season effect’’ (  p  , 0.001) on silicate, withslightly higher concentrations in summer (4.00  6  0.56  m mol L 2 1 ) than in winter (3.37  6  0.74). This seasonal trend was also seen inreference samples of surface oceanic water (OW). Significant between-habitat differences (  p  ,  0.001), as well as a significant season– habitat interaction (  p , 0.001), indicate that between-habitat differences depend on season. Habitat keys (near top of graph and orderedin decreasing magnitude) summarize the results of pairwise SNK tests, with keys sharing underline being not significantly different fromeach other. Mangroves showed significantly higher concentration in both seasons (  p  ,  0.05), and differences among the remaininghabitats (OR, PR, SG) were evident only during winter (  p  ,  0.05). (B) Mean planktonic bSi was about eight times higher in summer(182.382 6 79.307  m g L 2 1 ) than in winter (22.795 6 13.358  m g L 2 1 ); the ANOVA revealed that significant between-habitat differences (  p 5 0.023) depended on season. (  p 5 0.005). ‘‘A posteriori’’ SNK tests indicated that small between-habitat differences were detectable onlyin winter. Silicon budgets at a tropical continental shelf   2005
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!