ARTICLE IN PRESS
Continental Shelf Research 25 (2005) 1249–1264 www.elsevier.com/locate/csr
Opal content in the Rı´ a de Vigo and Galician continental shelf: biogenic silica in the muddy fraction as an accurate paleoproductivity proxy Patricia Berna´rdeza,b,, Ricardo Pregob, Guillermo France´sa, Raquel Gonza´lez-A´lvareza a
Departmento de Geociencias Marinas y Ordenacio´n del Territorio, Facultad de Ciencias del Mar, Campus Lagoas Marcosende s/n, Universidad de Vigo, 36310 Vigo, Spain b Instituto de Investigaciones Marinas (CSIC). C/ Eduardo Cabello 6, 36208 Vigo, Spain
Received 21 November 2003; received in revised form 8 December 2004; accepted 21 December 2004 Available online 26 February 2005
Abstract Biogenic silica (BSi) content was determined in both superficial marine sediments from the Rı´ a de Vigo and gravity core samples (core CGPL00-1) from the adjacent continental shelf. Samples were processed following the alkaline leaching procedure. The standard deviation for opal-rich samples is very low (70.2), whereas for opal-poor samples (o1.3 wt.%) the relative standard deviation can reach up to 16%. Opal percentages in superficial dry bulk sediments and gravity core samples range between 0.2–5.1 wt.% and 1.1–2.0 wt.%, respectively. Maximum opal percentages are found in the inner part of the ria around San Simo´n Inlet. Values of 2–3 wt.% typify the inner-central part of the ria. Throughout the ria longitudinal axis opal content is about 2 wt.%. Smaller values are found in the margins at the mouth of the ria. Opal distribution throughout the core is irregular, but there is a general tendency for higher values in upper muddy level and lower values in sandy sequence. Opal analyses were performed for the total and o63 mm fractions of both ria and core sediment samples. For the core samples, there is no correlation between opal content in the fine and bulk fractions, but opal percentage in the muddy fraction is an useful parameter to standardize results and to apply as a paleoproductivity proxy. For the ria surface sediments there is a good correlation between BSi content in both fractions (R2 ¼ 0:90). This fact suggests that the information provided by total and fine fraction analysis is similar; as a result, the opal content analysis in the fine fraction does not supply any new information concerning diatom productivity. r 2005 Elsevier Ltd. All rights reserved. Keywords: Opal; Paleoproductivity; Grain size; Ria; Galician continental shelf; Spain
Corresponding author. Facultad de Ciencias, Departamento de Geociencias Marinas y Ordenacio´n del Territorio, Universidad de
Vigo. Campus Lagoas-Marcosende s/n, 36200 Vigo, Spain. Tel.: +34986812633; fax: +34986812556. E-mail address:
[email protected] (P. Berna´rdez). 0278-4343/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.csr.2004.12.009
ARTICLE IN PRESS 1250
P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
1. Introduction Silicid acid is one of the major nutrients in the marine environment since marine planktonic microorganisms (diatoms, radiolaria and silicoflagellates) build amorphous silica shells from this nutrient, which they must extract from surface seawater undersaturated with respect to solid amorphous silica. A large fraction of biogenic silica (BSi) production in superficial waters is recycled via dissolution within the upper 100 m of the water column (Nelson et al., 1995; Tre´guer et al., 1995; Ragueneau et al., 2000), though dissolution can continue on the sea floor (Willey and Spivack, 1992; Tre´guer et al., 1995; Rickert et al., 2002). As a result, only a small fraction of the original opal reaches the sediment surface (Nelson et al., 1995; Tre´guer et al., 1995; Ragueneau et al., 2000). In spite of this fact, BSi accumulation in the sediments still reflects the general pattern of primary productivity in the overlying waters (Lisitzin, 1972; Banahan and Goering, 1986; Leinen et al., 1986) and can be used as a proxy for paleoproductivity studies (Charles et al., 1991; Mortlock et al., 1991; Ragueneau et al., 1996; De La Rocha et al., 1998; Masque´ et al., 2003). Spatial distribution of opal content has been widely studied, both in coastal-shelf areas, (DeMaster, 1981; Kamatani and Oku, 2000; Emelyanov, 2001; Gehlen and van Raaphorst, 2002; Liu et al., 2002) and in the deep ocean (Schlu¨ter and Rickert, 1998; Dixit et al., 2001; Ragueneau et al., 2001; Rathburn et al., 2001; van der Weijden and van der Weijden, 2002). However, only a few studies have been done in sediment cores, with the purpose to determine changes in the biosiliceous paleoproductivity of the superficial water column (Charles et al., 1991; Mortlock et al., 1991; Abrantes, 1996; Mortyn and Thunell, 1997; Anderson et al., 1998; De La Rocha et al., 1998; Weber and Pisias, 1999; Gorbarenko et al., 2002; Masque´ et al., 2003). Numerous techniques and modifications have been used in the determination of BSi in sediments. In general, they can be divided into four broad areas. Some procedures are based on structural analysis of the solid phase, e.g., direct IR spectroscopy of amorphous opal (Chester and Elderfield, 1968; Fro¨hlich, 1989) and X-ray dif-
fraction, direct diffraction of amorphous opal (Eisma and van der Gaast, 1971) and the conversion of opal to cristobalite at high temperature (Bareille et al., 1990). Other methods are based on point counts of siliceous microfossils (Leinen, 1985; Pokras, 1986), and on normative calculations estimating BSi by subtracting mineral silicates calculated from the Al and Mg concentration from the total Si content, to determine the excess of BSi (Leinen, 1977). However, the potentially most sensitive technique for determining BSi is wet-chemical leaching (Mu¨ller and Schneider, 1993). This technique is based on the assumption that the BSi and aluminosilicates have different dissolution rates, even in a weak alkaline solution, with fastdissolving amorphous BSi and slow-dissolving non-BSi phases, such as clays, feldspars and quartz. The advantages and limitations of these various methods have been discussed by DeMaster (1991), who basically proposed the use of the wetalkaline extraction technique as the most versatile procedure for opal measurements of samples with different origins and compositions. These procedures are also simple and economical for processing (Kamatani and Oku, 2000). Accordingly, wetalkaline methods are the most used techniques (Hurd, 1973; Eggimann et al., 1980; Kamatani, 1980; DeMaster, 1981; Shemesh et al., 1988; Mortlock and Froelich, 1989; Gehlen and van Raaphorst, 1993; Mu¨ller and Schneider, 1993; Kamatani and Oku, 2000; Fabres et al., 2002; Koning et al., 2002; Liu et al., 2002). The alkaline leaching technique has diverse problems, and several factors may affect the precision and accuracy on BSi content measurement. The recovery of BSi depends on extraction conditions, leaching procedure, sediment composition and dissolution by non-BSi compounds. Therefore, variability in opal measurements should be recognized and discussed. The main objectives of the present study are (1) to establish the precision of the Mortlock and Froelich (1989) method for the opal percentage range found in Galician sediment samples, (2) determine the distribution of the BSi content in surficial sediment of the Rı´ a de Vigo and in core samples of the Galician continental shelf, (3)
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
determine the influence of grain size distribution in opal determination, and (4) discuss the importance of the use of BSi percentage in fine fraction as a paleoproductivity proxy. 2. Regional framework The study area is located in the northwestern Iberian Peninsula and embraces two zones: the Rı´ a de Vigo, one of the chief embayments of the Galician coast, and the Galician continental shelf (Fig. 1). The ria can be divided into several zones according to the degree of continental or oceanic influence. The innermost zone includes the San Simo´n Inlet, that is under a strong tidal influence (average tidal range 3 m) and freshwater supply from the Verdugo-Oitave´n River (annual average flow of 27.5 m3 s1, Rı´ o and Rodrı´ guez, 1992). The middle zone is influenced by both continental and oceanic contributions, whereas in the outermost zone freshwater input is negligible (Nogueira et al., 1997). The ria behaves as a partially mixed estuary (Beer, 1983) with a two-layered positive residual circulation pattern, maintained in winter by the freshwater flow and in summer by upwelling (Prego and Fraga, 1992). Off the ria, the Galician continental shelf is relatively narrow: the 200 m isobath lying at 15–30 km from the coastline (Fig. 1). Sediment supply comes mainly from the Min˜o River (the rias act as sediment traps (Prego, 1993)), and defines a sedimentary body at 110–120 m depth known as the Galicia Mud Patch (Dias et al., 2002). The hydrography of the ria-shelf region is strongly influenced by a seasonal wind-driven upwelling (Fraga, 1981; Blanton et al., 1984; A´lvarez-Salgado et al., 1993), SW storm surges (Vitorino et al., 2002) and the presence of a poleward current during winter months (Frouin et al., 1990; Haynes and Barton, 1990).
3. Material and methods 3.1. Sample recovering and processing Fifty-one sediment samples from the uppermost oxic layer (0–1 cm) were collected in the Rı´ a de
1251
Vigo by a Van Veen grab on board the R/V Mytilus during a research cruise in October 1999 (Fig. 1). After collection, surface sediment samples were dried in an oven below 40 1C and kept in plastic storage for subsequent analyses. Afterwards, bulk sediment was size-fractionated by dry sieving into mud, sand and gravel through sieves of 63 and 2000 mm mesh size. Opal determinations were done on bulk sediment for all samples, and for selected sites, analyses were carried out also in the o63 mm fraction (Fig. 1). Opal analyses were also carried out on samples from a 96 cm gravity core (CGPL00-1) retrieved from the Galician continental shelf (42150 15.11500 N, 9130 46.38000 W, 130.8 m water depth) on board the R/V Mytilus during a cruise in May 2000 (Fig. 1). The core was sealed just after collection and kept in storage at 4 1C until analyses were performed in the laboratory. The core was split longitudinally in two sections: one was used for determination of the BSi percentage and correlative subsamples of 2 cm thickness were taken for grain size analyses. After removing the organic matter with H2O2, the coarse fraction was separated from the fine fraction by wet sieving through a sieve of 63 mm mesh size. The 463 mm residue was dry sieved through 2, 1, 0.5, 0.250, 0.125 and 0.063 mm sieves and weighed. Mud content was determined following the method outlined by McManus (1991). X-ray nephelometry (Micromeritics, Sedigraph 5100) was used to separate silt and clay. Finally, determinations of opal percentage were carried out on the dry bulk sediment and in the o63 mm fraction. Measurements of the BSi content in the bulk sediment were done every 2 cm, whereas in the muddy fraction opal analyses were done every 5 cm. 3.2. Opal analysis Opal concentration was measured using the alkaline leaching technique outlined by Mortlock and Froelich (1989). Sediment samples were kept in storage at 4 1C and softly dried to constant weight at 40 1C. Preceding leaching, samples were crushed with a mortar and pestle and homogenized. At least two replicates of each sediment sample were treated in each leaching experiment. In
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
200
100
150
500
1000
4690000
S IXA BA
LATITUDE N (UTM, m)
S RIA
4730000
1252
VIGO
10 km
1500
4650000
CGPL00-1
455000
475000
495000
515000
535000
LONGITUDE W (UTM, m)
7
5
6
SAN SIMÓN INLET
4680000
37
22
Home Cape
23
4675000
Cíes Islands 21
30
2
45 42 44 43
Rande Strait
REDONDELA
38 32
27
20
40
31
25 24 26
41
36
CANGAS
3
9
10 11 51 50 1 49 1412 13 46 48 47
MOAÑA
35
39
A Guía Point
33 34
29 28
15
Samil Beach
VIGO
19
4670000
LATITUDE N (UTM, m)
4685000
8 4
′ RIA DE VIGO
18 16 17
5 km 510000
515000
520000
525000
530000
LONGITUDE W (UTM, m) Fig. 1. Chart of the study area. Above: core CGPL00-1 location in the Galician continental shelf. Contour lines in this map show depth in meters. Below: map of the Rı´ a de Vigo showing the 51 sampling sites (circles). Samples of sediment were taken from the uppermost oxic layer (0–1 cm). Opal analyses were performed in bulk and muddy fractions for selected samples (black circles).
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
addition, a blank leaching solution was included in all runs. Dried sediment was weighted (approximately 200 mg) into a polypropylene centrifuge tube. Carbonates and organic matter were removed by hydrochlorhydric acid 1M and peroxide (pharmaceutical grade). 40 ml of 2M Na2CO3 solution was added to the samples. The tube was closed with caps having pin-holes for ventilation, sonified, homogenized, and placed in a covered constant-temperature water bath preheated to 85 1C for 5 h. The tubes were agitated vigorously to suspend the solids at 2 and 4 h and incubated again in the water bath. All steps after removal of the tube from the hot bath were done quickly to minimize irreversible loss of dissolved silica to solid surfaces. After a total of 5 h, samples were agitated and then removed from the water bath and centrifuged at 5000 rpm for 5 min at 22 1C. Immediately after centrifugation, approximately 10 ml of the leaching solution were transferred to plastic tubes and stored for subsequent analysis. The resulting extract was measured for the dissolved silicate concentration by the molybdate blue spectrophotometric method according to Hansen and Grashoff (1983) using an AutoAnalyser Technicon II. The sample was previously acidified with HCl to neutralize the Na2CO3 leaching solution and segmented by air bubbles to enhance mixing with reactive agents (ascorbic acid, oxalic acid and molybdate). The analyser was calibrated and checked for linearity by running the blank leaching solution and the working standard solution. Once linearity was checked, an appropriate working standard solution was run (106.58 mM Si(OH)4) before and after the sample series. Precision of the analytical system is based on replicate measurements of silicon standard solutions. Dissolved silicate content for each sample was calculated taking into account the dissolved silicate concentration of the working standard solution. Finally, the percentage of opal in dry sediment is expressed by % opal ¼
ðC m C 0 Þ 100 K. P
(1)
C m is the dissolved silicate concentration of the sample in mM, C 0 is the dissolved concentration of the blank in mM, P is the dry sediment weight in mg
1253
and K (2.7 gmol1 l) is a constant value comprising the digestion volume (l), the molecular mass of Si (gmol1) and a correction factor that depends on the opal water content (10% in this case because of their recent origin). 3.3. Accuracy and precision of the opal determination Due to the absence of certified BSi standards, establishing absolute accuracy with existing methodologies is not possible. As there is not an amorphous silica standard or purified BSi available to add to the sediment matrix, some authors have established accuracy by comparison with other techniques (Mortlock and Froelich, 1989; Conley, 1998; Liu et al., 2002) or analysis of artificial sediment standards that contain a known fraction of pure biogenic opal (Mortlock and Froelich, 1989; Mu¨ller and Schneider, 1993; Koning et al., 2002). However, it is difficult to prepare an artificial matrix similar to natural sediments, to mix various sedimentary components in the laboratory or to manufacture an internal laboratory standard. Thus, the assumptions pointed out by Mortlock and Froelich (1989) about accuracy have been accepted and the opal measurements done are considered as exact. The precision of the overall method was estimated from analyses of selected superficial samples from the Rı´ a Vigo (samples 10, 49, 26, 39 and 31) with opal contents covering the entire opal range found in the ria and shelf sediments (Table 1). The standard deviation of the analyzed samples ranges between 0.14 and 0.22 and increases slightly as the opal content diminishes, but the general trend is nearly constant. Precision of the method is about 70.2 for the opal range studied. It decreases when opal percentages are relatively low, but according to the stability of the standard deviation value (70.2) precision is satisfactory for all sediment samples analyzed, regardless of their opal content. When the relative standard deviation (RSD) about the mean is examined, this value rises as the BSi content diminishes down to 1.3 wt.%. Mortlock and Froelich (1989) analyzed sediments from Atlantic, Antarctic, Pacific and Indian Oceans and they
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
1254
Table 1 Study of precision of the method Sample code
Longitude (UTM)
Latitude (UTM)
Number of analysis
Opal mean (wt.%)
Standard deviation
Relative standard deviation (%)
10 49 26 39 31
530,005.31 528,921.71 516,124.18 523,588.86 518,330.63
4,683,159.26 4,682,062.81 4,676,928.81 4,677,136.80 4,678,200.05
6 6 6 8 6
5.01 4.12 2.97 2.14 1.37
0.18 0.19 0.14 0.21 0.22
3.49 4.73 4.88 9.95 16.30
Table shows samples used in this work, location (latitude and longitude in UTM units), number of analysis, mean, standard deviation and relative standard deviation. Opal content determinations for each sample were done in different runs.
obtained similar RSD results, i.e. around 8% in samples with opal concentrations lower than 15 wt.%.
4. Results and discusion 4.1. Opal distribution in surface sediments of the Rı´a de Vigo BSi analysis of bulk sediment samples from the Rı´ a de Vigo shows that, as a general tendency, maximum opal percentage is found in the inner part of the ria (Fig. 2A), clearly decreasing towards the mouth of the ria. Higher percentages (about 4–5 wt.%) are found in the San Simo´n Inlet. In the inner part of the ria and near the Rande Strait adjacent to the northern coast opal percentage is around 2–3 wt.%. Similar values are registered in the middle zone of the ria (about 2 wt.%), especially throughout the longitudinal axis. In contrast, lower percentages (o0.5 wt.%) are found in the ria margins, particularly, at the ria mouth. In this sector, sediment structure (Fig. 2B) comprises coarser grain sizes such as quartz sand and biogenic carbonates (Nombela et al., 1987; Vilas et al., 1995). It is interesting to compare our results with previous works by different authors in the surrounding area. Prego et al. (1995) studied the biogeochemical silicon cycle in the Rı´ a de Vigo and also described a decrease in opal content from the inner part to mouth in the superficial sediments. According to these authors, opal distribution shows high percentages along the
central axis (42.5 wt.%), especially in the central zone, whereas low opal contents typify the margins to the ria. Compared to this study, we have increased the number of sampling stations, and the innermost part of the ria (San Simo´n Inlet) has also been surveyed. This higher sampling resolution permits us to establish a more accurate account of opal distribution of the ria and, in fact, remarkable differences with the Prego et al. (1995) results are found. For example, the highest values (even 5 wt.%) are located in the inner part of the ria (San Simo´n Inlet) and in the innercentral zone. Subsequently, Prego and Bao (1997) evaluated BSi content in superficial sediment samples recovered from the whole Galician continental shelf, and Barciela et al. (2000) extended this study to two other rias: Pontevedra and Arousa. Recently, Dale and Prego (2002) have determined the opal content in the uppermost oxic layer (1 cm of sediment) in sampling stations of the Rı´ a de Pontevedra. This work shows that higher opal percentages are recorded in the inner ria and towards the northern coast, whereas, at the ria mouth, BSi content is relatively low, mainly due to the coarser grain size distribution in this zone. Comparisons with Rı´ a de Pontevedra and Arousa illustrate that opal content is similar between them, ranging from 0.5 to 5 wt.%. BSi distribution is analogous between the Rı´ a de Vigo and Pontevedra, higher percentages occurring in the inner part and towards the north shore in the inner-central zone. This particular trend is due to a dilution effect caused by residual organic matter in the sediments of the southern margin of the Rı´ a de
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
1255
Fig. 2. (A) Opal distribution in the bulk sediment throughout the Rı´ a de Vigo. (B) Detailed map of the superficial sediment distribution of the Rı´ a de Vigo (modified from Vilas et al., 1995).
Vigo (Vilas et al., 1995) and by discharges of particulate organic matter from a nearby paper mill (Arbones et al., 1992) and urban wastewater in the Rı´ a de Pontevedra (Barciela et al., 2000; Dale and Prego, 2002). On the other hand, BSi distribution of the Rı´ a de Arousa is quite singular because higher opal percentages in the north coast are not detected and opal in this ria shows a general decreasing trend from head to mouth. This pattern is explained by the higher fertilization by
silicate provoked by the freshwater input of the Ulla River (Vergara and Prego, 1997). Taking into account these previous studies and the results of the present work, it is possible to discriminate three areas in the Rı´ a de Vigo in relation to opal content of seabed sediment: the inner ria (San Simo´n Inlet), a central area where opal content is still relatively high, but less than in the internal zone and, finally, an outer area characterized by lower BSi percentages.
ARTICLE IN PRESS 1256
P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
Data concerning primary productivity in the Rı´ a de Vigo are relatively scarce and have a wide range of variability depending on the ria sector and the sampling period considered. Mean annual value of net primary production throughout the ria, at an annual scale, is about 350 mgCm2 d1 (Prego, 1993), although 40% of this quantity is remineralized. During the upwelling season in 1997, Gago et al. (2003) found values of net ecosystem production around 790 mgCm2 d1, but 68% of production is trapped in the sediments. Elevated values from spring to autumn range between 700 and 1200 mgCm2 d1 in the inner part of the ria (Vives and Fraga, 1961), even 2400 mgCm2 d1 in summer (Tilstone et al., 1999). Fraga (1976) found values of about 2800 mgCm3 d1 in the inner zone. A mean seasonal water column gross primary production of 2100–2700 mgCm2 d1 (Moncoiffe´ et al., 2000) is also found in the inner ria, but only 33% is transferred to sediment on the shelf or to superior trophic levels. Thus, elevated values of primary production are found when upwelling occurs form April to October. As a general rule, productivity in the Rı´ a de Vigo shows a decreasing trend from head to mouth (Fraga, 1976) with the highest values in the innercentral part (adjacent to Rande Strait). Primary production is 3–4 times higher in the inner zone than in the outer zone (Vives and Fraga, 1961). Prego (1993) and Gago (2000) also registered this spatial and temporal differences in their calculations about carbon fluxes to the sediment. Considering data about carbon primary production in the ria, the opal content in the sediment is correlated with the productivity in the water column, principally diatoms, typical biosiliceous organisms in upwelling areas (Bao et al., 1989; Lisitzin, 1996; Bao et al., 1997). However, in San Simo´n Inlet the opal content found in the sediment is higher than expected, and it cannot be explained only by water column productivity. This semi-enclosed area is characterized by shallower depths, which permits seabed illumination, tidal mixing, and introduction of dissolved silicate input from river discharge (Prego and Fraga, 1991; Vergara and Prego, 1997). Thus, proliferation of benthic diatoms in this sector is
enhanced. Moreover, Varela (1984) and Varela and Penas (1985) pointed out the increasing in the phytomicrobenthic production in intertidal environments, and the higher frustule resistance of benthic diatoms to dissolution. Pazos et al. (2000) revealed the presence of green algae in particulate matter of the Verdugo-Oitave´n River. Margalef (1956) and Bao et al. (1989) found freshwater diatoms in the San Simo´n Inlet and in the vicinity of Rande Strait. Therefore, the high opal percentage found on the sediment of the San Simo´n Inlet could be due to the growth of benthic diatoms and the input of freshwater diatoms. Looking at the sediment cartography of the Rı´ a de Vigo (Vilas et al., 1995) (Fig. 2B), the distribution of BSi in the bed-sediment appears closely correlated with the sediment structure throughout the ria. Seafloor sediment is composed of mixed siliciclastic and skeletal gravels in both the outer area and the margins of the ria. Grain size distribution in the entrance channels, north and south of the Cı´ es Islands, is characterized by a high sand proportion, especially in the northern mouth, whereas along the longitudinal axis, sediment is finer. Towards the shoreline, the sediments become coarser, grading through various intermediate sediment types into clear carbonate skeletal sands or mixed carbonate siliciclastic sands. In some places, especially in the outer and northern parts of the ria, patches of the calcareous algae Lithothamnion corralloides and Phymatolithon calcareum are found (Nombela et al., 1987, 1995; Vilas et al., 1995). The central and inner parts of the ria are dominated by clay and silt fractions. Particularly in the inner parts, finegrained sediments persist up to the shoreline. Opal content appears to be related to the fine fractions distribution, showing higher percentages throughout the longitudinal axis and in the inner and central parts of the ria (Figs. 2A and B). Carbonate content also shows a good agreement with the grain size distribution. Maximum percentages are recorded in the south entrance channel, whereas minimum carbonate content is recorded in the central axis of the ria. Organic matter content appears to be related to the fine fraction distributions, showing its maximum percentages in the inner and middle part of the ria (Vilas et al., 1995).
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
1257
Opal distribution all along the ria appears to be influenced by the primary production in the water column, as well as the biogenic carbonate and lithogenic fractions of the sediment, as a result of the ria hydrodynamic (Prego and Fraga, 1992; Taboada et al., 1998) and sedimentation processes (Prego, 1993; Dale and Prego, 2002).
To quantify the difference between the opal content in both fractions, the variation percentage is calculated using the expression
4.2. Opal content in the muddy fraction of surface sediments of the Rı´a de Vigo
Variation percentage (Table 2) shows that the opal percentage in the bulk sediment is relatively lower than in the o63 mm fraction, except for sample 20 (variation percentage 2.5%). Samples 44 and 37 contain similar opal percentages in the bulk and muddy sediment. Mean variation percentage for all samples is approximately 17%; therefore differences between opal percentage in the bulk and muddy fractions are remarkable. Variation percentage can even reach 40%, demonstrating that BSi is more concentrated in the muddy fraction. Fig. 3A shows the linear correlation between the opal content in the bulk sediment and in the finer fraction. The correlation coefficient is high (R2 ¼ 0:90). This excellent relationship shows that variations in the opal content in both sediment fractions are analogous. Thus, we conclude that analysis of opal in the muddy fraction is unnecessary, and variations in biosiliceous productivity in the recent sedimentary record of the ria can be detected by the analysis of opal in the bulk sediment. This methodological information can be also applied to the other Galician Rı´ as or similar environments as estuaries or embayments.
Grain size distribution can affect opal content in sediment because coarser grains provide a dilution effect that must be eliminated, as it was demonstrated by Luoma et al. (1990) in the heavy-metal case. In order to remove this we evaluated the opal percentage in the o63 mm fraction because we assumed that most opal, principally diatoms, is retained here. Therefore, important information can be obtained when opal is normalized and quantified in this fraction. This hypothesis is supported by two conditions: (1) Fertilisation by nutrients during upwelling process leads to an increase in the total primary production, particularly in the biosiliceous production. Tilstone et al. (2000) in a study in this area show that biosiliceous production is due to a larger contribution by diatoms. (2) Diatom size depends on a few parameters, such as the characteristic size of the species, water column temperature (Margalef, 1956, 1959), and the essential nutrient supply (Burckle, 1998). Nevertheless, it is unlikely that diatom frustules can reach sizes 463 mm. However, other BSi contributions in the 463 mm fraction would be possible, e.g., sponge spicules remains, but this type of BSi is quite refractory and the dissolution in the alkaline leaching solution is difficult (Mu¨ller and Schneider, 1993). In order to test the above hypothesis, a comparison between opal percentage in total and muddy fractions was carried out for 19 sediment samples, with an opal content in the bulk sediment of 1–5 wt.% and a mean value about 2.9 wt.% (Table 2). In these samples, opal percentage in the fine fraction ranged between 1.9 and 5.9 wt.%, with a mean value of 3.5 wt.% (Table 2).
Variation percentage % opalo63 mm % opal bulk ¼ 100. % opalo63 mm
ð2Þ
4.3. Opal content in the Galician continental shelf: an accurate paleoproductivity proxy To establish the temporal variability of the sedimentary record of BSi, a gravity core CGPL00-1 was retrieved from the Galician continental shelf. Lithology of the core is characterized by the presence of two well-differentiated sections: the lower half of the core consists of laminated glauconitic sand overlying a bioclastic gravel interval 2 cm thick, whereas the upper one is
1258
Table 2 Opal percentage in the bulk sediment and in the o63 mm fraction from selected surface samples from the Rı´ a de Vigo Longitude (UTM)
Latitude (UTM)
o63 mm fraction (%)
63–2000 mm fraction (%)
19 31 20
Homogeneous dark grey mud Muddy sand Non homogeneous dark brown mud Homogeneous dark grey mud Brilliant, viscous and light brown mud Dark brown mud Greyish black mud Compact, viscous and light brown mud Dark brown mud with mussel shells Black sandy mud with shells and strong smell Brilliant black mud Dark brown mud with dispersed shells Dark brown mud with dispersed shells and seaweed — Mud with dispersed shells and seaweed Greyish brown mud Brilliant black mud — —
512,133.87 518,330.63 511,964.66
4,671,233.86 4,678,200.05 4,674,435.12
38.81 12.02 62.97
30.67 79.73 36.41
509,373.22 515,530.51
4,674,344.21 4,674,891.79
39.93 46.96
517,687.66 523,258.38 520,657.24
4,675,833.30 4,678,940.08 4,677,138.68
526,308.65
21 27 30 40 32 44 23 41 37 47 2 51 26 46 10 8 Mean
Opal o63 mm Opal bulk (wt.%) (wt.%)
Variation percentage
0.52 8.25 0.61
1.88 2.09 2.16
1.59 1.21 2.21
15.42 41.75 2.54
56.83 53.04
3.24 0.00
2.25 2.45
1.80 2.08
19.82 15.28
45.75 27.65 37.42
53.97 68.05 62.49
0.29 4.30 0.09
2.51 2.55 2.73
1.98 2.18 2.05
21.32 14.39 24.72
4,679,729.96
28.19
71.12
0.69
2.89
2.85
1.58
512,968.07
4,677,083.29
6.02
82.21
11.77
3.34
2.01
39.99
524,239.04 522,423.69
4,680,390.60 4,680,738.04
33.10 40.10
66.19 55.84
0.71 4.06
3.65 3.67
3.24 3.48
11.11 5.19
527,467.23
4,681,181.61
45.53
52.91
1.56
4.10
3.50
14.63
528,944.91 529,252.87
4,683,765.52 4,682,721.17
16.19 50.35
80.36 49.65
3.45 0.00
4.22 4.29
3.68 3.89
12.88 9.34
516,124.18 526,619.32 530,005.31 530,801.27
4,676,928.81 4,681,537.37 4,683,159.26 4,686,493.92
20.47 36.59 28.82 13.54
70.38 63.41 71.18 75.53
9.15 0.00 0.00 10.93
4.30 5.02 5.52 5.87
2.98 4.22 5.07 4.82
30.58 15.87 8.11 18.02
33.18
62.10
3.14
3.45
2.89
16.71
42000 mm fraction (%)
Table also shows a description of the sediment samples, the percentage of each fraction and the variation percentage between opal in bulk and in muddy fraction. Variation percentage is calculated following the Eq. (2).
ARTICLE IN PRESS
Sample description
P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
Sample code
ARTICLE IN PRESS P. Berna´rdez et al. / Continental Shelf Research 25 (2005) 1249–1264
fine fractions were done throughout both sandy and muddy sequences of the core. As a general pattern, the trend of opal abundance in the bulk sediment throughout the whole core shows two different modes (Table 3, Fig. 4). In sandy sequence, opal distribution shows a very irregular profile, whereas in the muddy sequence BSi values stabilize around 1.5 wt.%. Profile of the BSi percentage in the o63 mm fraction displays a relatively regular tendency. From the bottom of the core to 60 cm, opal content increases progressively. Values became stable from 60 to 25 cm, averaging 1.6 wt.%, and from this level to the core top, opal percentage decreases considerably (Table 3, Fig. 4). This trend closely resembles the bulk sediment fraction profile in the muddy sequence, whereas in the sandy sequence both profiles differ.
OPAL
