Living ( stained ) calcareous benthic foraminifera from recent sediments off Concepción , central-southern Chile ( ~ 36 o S ) Foraminíferos bentónicos calcáreos vivos ( teñidos ) en sedimentos recientes de Concepción , Chile centro-sur ( ~ 36 o S )

This study examines onshore-offshore and vertical distribution of living (Rose Bengal stained) benthic foraminifera (> 180 μm fraction) from three sediment stations along a bathymetric transect off Concepción, Chile (station 18 = 88 m water depth, station 26 = 120 m, station 40 = 1,030 m), within and below the oxygen minimum zone. All cores were collected in austral winter. Calcareous foraminifera dominated the three stations. The species composition, living foraminifera density, and vertical distribution patterns within the sediment changed in accordance with bottom water dissolved oxygen concentration and food availability. Onshore-offshore pattern revealed overall highest living foraminiferal densities at shelf stations 18 and 26 where bottom water dissolved oxygen was lowest (~ 0.2 mL-1) and content in labile organic matter highest. Within the sediment, maximum relative abundances (50-60 %) of living organisms were found in the 0-1 cm interval at the organic-rich and oxygen-poor shelf stations 18 and 26. In the well-oxygenated (2.7 mL-1) slope station 40, 70 % of living foraminifera were observed deeper than the first centimeter. The number of species and the contribution of the > 250 μm fraction to the total fauna larger than 180 μm increased offshore. Nonionella auris (d’Orbigny) dominated at stations 18 and 26 while a more diverse foraminifera fauna was found at station 40. This study provides the first quantitative data on living benthic foraminifera in the area; seasonal and interannual changes are not addressed.


INTRODUCTION
The vertical distribution of benthic foraminifers in the sediment column seems to be determined mainly by two factors: food availability and oxygen content at the seafloor (Gooday & Rathburn 1999, Gooday 2003).This is supported by a number of field studies and experiments that demonstrate the effects of organic supply and changes in bottom water oxygen concentration on the microhabitat selection (e.g., Rathburn & Corliss 1994, Ernst & van der Zwaan 2004).
All these observations were summarized in the "TROX-model" of Jorissen et al. (1995), which relates the microhabitat "selection" of a given taxa to a balance between food and oxygen availability.According to this model, oligotrophic systems are food-limited and organisms are concentrated close to the sediment surface, where the majority of the food is located.Although eutrophic systems are oxygenlimited, organisms are also concentrated near the surface in order to avoid anoxic conditions deeper in the sediment.This general scheme has been refined by Jorissen (1999), van der Zwaan et al. (1999), andFontanier et al. (2002) who suggested that: (1) the organic matter flux is the main parameter controlling foraminiferal microhabitats, (2) oxygen is not the limiting factor for deep infaunal species, and that (3) biological interactions, particularly competition for labile food material, may play a role in microhabitat selection.
Oxygen minimum zones (OMZs) are significant midwater features in the eastern Pacific Ocean (e.g., Helly & Levin 2004).It has been shown that, where the OMZ intercepts the continental margin, strong gradients are formed in both bottom water dissolved oxygen concentrations and organic matter input (Levin et al. 2000).These gradients influence the biogeochemical properties of the underlying sediments and the distribution of benthic organisms within the sedimentary column.Usually organic-rich, oxygen-poor environments are dominated by meiofauna, mainly calcareous foraminifera and nematodes (e.g., Levin et al. 2002).In the last decades, significant advances have been made in the understanding of ecological aspects of benthic foraminifera from oxygen deficient settings in the world (see Bernhard & Sen Gupta 1999 for a review), including off Peru in the eastern south Pacific (Phleger & Soutar 1973, Levin et al. 2002).
Substantial progress in the knowledge of modern benthic foraminifera from the Chilean continental margin has been made since the early works of Bandy & Rodolfo (1964), Boltovskoy &Theyer (1970), andIngle et al. (1980).More recent studies (e.g., Zapata & Moyano 1997, Figueroa et al. 2005, 2006, Hromic et al. 2006) have focused on taxonomic and/or biogeographical aspects, emphasizing the identification and characterization of zoogeographic provinces along the Chilean coast.But none of these studies investigated the relationship between living benthic foraminiferal fauna and environmental characteristics.
To provide new insight into benthic foraminiferal ecology off Concepción, Chile (3 6º S), we present results on total and living benthic foraminifera from a bathymetric transect (80, 120, 1,030 m water depth; Table 1, Fig. 1A) that includes stations within (stations 18 and 26) and below the OMZ (statation 40).Our study focused on (1) onshore-offshore, and (2) vertical distribution of calcareous foraminiferal communities within the sedimentary column (upper 10 cm) with regard to bottom water oxygen content and organic matter availability.

Regional setting
The area off central-southern Chile has been studied in detail by various authors (Ahumada & Chuecas 1979, Sobarzo et al. 2001) including plankton variability, biogeochemical processes, and physical and chemical characteristics (Escribano & Schneider 2007).
Due to diverse factors (coastal winds, upwelling events, shelf width, and topographic settings), this area is one of the most productive sectors of the central-southern Chilean coast (35-45º S, Shaffer et al. 1999) with primary productivity values among the highest in the world ocean (Daneri et al. 2000, Schubert et al. 2000).Upwelling is strongly seasonal.It intensifies during austral spring-summer, fertilizing the photic zone with the salty, nutrient-rich, oxygen-poor waters of the Equatorial Subsurface Water mass (ESSW) (Ahumada & Chuecas 1979).Upwelled ESSW can be found over the continental shelf and sometimes even at the inner part of Concepción Bay.Downwelling usually occurs in winter due to the prevalence of strong northerly winds (Sobarzo 1993).
Total organic carbon content (TOC) in the sediment is usually high and varies between 2 and 5 % (Table 2), with maximum TOC values in austral summer (Muñoz et al. 2007).Organic matter is mainly of phytoplanktonic origin (Schubert et al. 2000).Chlorophyll-a (Chl-a) concentrations in the surface sediments clearly decrease offshore (Table 2).On a yearly basis, C/N molar values fluctuate between 8.4 and 11.3 (Muñoz et al. 2007).Modern sedimentation rates are high (0.09-0.25 cm yr -1 , Muñoz et al. 2004).Sediments are usually composed of silt and clay minerals with a mean grain size of 8.5-13.0μm (Molina et al. 2004).
The study area is located on the widest portion of the continental margin off centralsouthern Chile (50 km) (Fig. 1A).The shelf is limited by two submarine canyons and their corresponding rivers: the Itata canyon (Itata River) to the north and the Biobío canyon (Bio-Bío River) to the south.In general, three different water masses can be observed here: Subantarctic Water (SAAW) between the surface and about 80 m depth, ESSW at about 80 to 400 m with minimum dissolved oxygen concentrations, and Antarctic Intermediate Water (AAIW) below 400 m (Molina et al. 2004) impinging on the continental slope (Fig. 1B).The OMZ is associated with oxygendepleted ESSW and intensifies by the local biogeochemical O 2 consumption (Paulmier et al. 2006).Off Concepción, the OMZ is less intense and its upper boundaries are deeper (about 100 m) than off northern Chile (Pizarro et al. 2002); its strength and the depth of the upper boundary vary seasonally.The OMZ's upper boundary is shallow in spring-summer (at 20 m from October to March) during the high upwelling activity period with O 2 concentrations lower than 0.2 mL -1 , but it deepens in autumn (about 50 m in April-May) (Paulmier et al. 2006).

MATERIAL AND METHODS
The sediment samples used in this study were collected during winter cruises as part of the COPAS Center Time Series Study off Concepción (Table 1).A SMBA-type minimulticorer (MUC) with six sampling tubes (inner diameter = 9.5 cm) was used for sampling at Sta. 18 (middle shelf, 88 m water depth) and 26 (outer shelf, 120 m water depth) (Fig. 1A).A Rhumor corer (RC) -a type of gravity corer that recovers the sediment-water interface -was used at Sta. 40 (middle slope, 1,030 m water depth).The RC uses an acrylic of 7 cm internal diameter and ~ 100 cm in length.One core tube was analyzed for each station.Due to the limited data set, interannual and seasonal changes cannot be addressed.
The cores were sliced at 0.5 cm intervals for the upper 2-3 cm, and every 1 cm downcore.Immediately after slicing, the samples were preserved and stained according to the methodology of Rathburn & Corliss (1994); samples were allowed to stain for at least one week before processing for foraminifera analysis.Samples were not split but examined entirely for their content of live foraminifera.The use of Rose Bengal stained foraminifera to interpret living associations was critically discussed by Bernhard (1988).One problem of this technique is the fact that Rose Bengal may stain the protoplasm of dead foraminifera, which may be relatively well preserved under anoxic conditions leading to an overestimation of the living fauna.However, it is still the most frequently used method in foraminifera research and the most practical one (Bernhard 2000).Our definition of "stained" (living) foraminifera was very strict and only specimens with all chambers, except for the last one (the youngest) stained bright pink were counted as stained.
After staining, samples were wet-sieved using 63, 180, and 250 μm mesh-opening.The data presented here are from the > 250 and 180-250 μm fractions and refer to the upper 10 cm of each core.The mesh sizes were selected in keeping with the methodology previously used for the COPAS Time Series Study off Concepción (Tapia 2003).The fraction 63-180 μm was preserved for future studies.
We are aware of the fact that the selection of the > 180 μm size fraction may have introduced some bias regarding the abundance of total foraminifers as well as the abundance of smaller organisms.Although the smaller size fraction (e.g., 63-150 μm) may contain a large number of foraminiferal species and juvenile forms that are rare or absent in the larger size fraction (e.g., Schröder et al. 1987), their removal is extremely time-consuming.Additionally, smaller specimens are often more problematic to identify to the species level.
To our knowledge, there is only one quantitative (unpublished) study of living benthic foraminifera available for the shelf off Concepción (Mayor et al. unpublished results).Our results may have underestimated the total number of living foraminifera in the study area (shelf stations) by ca.20 % as compared with the abundances obtained by Mayor et al. (unpublished results) for the > 150 μm fraction.However, the total numbers of benthic foraminifera (stained + dead) and species that we found for the slope area off Concepción falls within the range obtained by Figueroa et al. (2005Figueroa et al. ( , 2006) ) who analyzed the > 150 μm at nearby slope and deep stations.An important aspect is that a scan of the > 63 μm fraction revealed that, with the exception of tiny Globobulimina and Stanforthia, the overall species composition for the Concepción area is comparable in both small and coarser size fractions.
The total density of living calcareous foraminifera was determined by integrating the numbers of live individuals picked at all levels from 0 to 10 cm sediment depth.Percentages of individual taxa were calculated from the raw density data.Concerning vertical profiles, foraminiferal densities were normalized to 50 cm 3 sediment volume for each 0.5 and 1 cm layer.Finally, with the purpose of comparison with other areas, living calcareous foraminifera were normalized to 50 cm 2 for the first centimeter.
We assigned the benthic foraminifera to two different microhabitat groups (shallow and deep infauna).None of the species off Concepción could be considered to be strictly epifauna as defined by Buzas et al. (1993) who suggested that only species living on elevated substrates can be assigned to this category.
In order to provide an overview of the environmental characteristics in the study area, we present available water column and surface sediment data in Table 2; in those cases when environmental data were not available for the dates we collected the cores, we resorted to previously published data.Water column data derive mostly from the COPAS Time Series Study off Concepción (http://copas.udec.cl/eng/research/timeseries.php), and include CTDO data (2002)(2003)(2004)

RESULTS
The sediments off Concepción were typically greenish-brown in color (7.5 Y 3/2-7.5 Y 4/2 Munsell table color).Fecal pellets were numerous at the shallower stations (Sta. 18 and 26).A grey layer of very fine material was observed at 7-9 cm depth in the core from station 40.Available Chl-α and TOC data (although not always coincident with our August samplings) reveal that the sediments of the shelf stations are generally characterized by higher levels of Chl-a concentrations and TOC (> 20 μg g -1 and > 3.3 %, respectively) than the slope station 40 (Table 2).Bottom water dissolved oxygen (BWDO) conditions differed along the transect.Minimum BWDO values (0.17 mL -1 ) were encountered at station 18 (88 m water depth) and maximum values at station 40 (2.7 mL -1 ) during our August samplings (Table 2), in agreement with known OMZ boundaries in the area.Redox potential values at the sediment surface (Eh) agree with increasing bottom water dissolved oxygen concentrations from onshore to offshore (Table 2).
Using the BWDO values given in Table 2, the stations could be grouped into two categories according to Sen Gupta & Machain-Castillo (1993) classification of environments: (1) shelf stations 18 and 26, with dysoxicsuboxic conditions (< 2 mL -1 ), located at and immediately below the upper boundary of the OMZ, respectively; and (2) slope station 40, with oxygenated conditions (> 2 mL -l ), located several hundred meters below the OMZ.Bottom water temperature, salinity, and BWDO (Table 2) showed the influence of the ESSW at Sta. 18 and 26, whereas AAIW affected Sta.40.

Onshore-offshore distribution of benthic foraminifera
The analysis of benthic foraminifera in the cores off Concepción showed that the contribution of the > 250 μm fraction to the total fauna (larger than 180 μm) increased offshore (Table 3) from 1.8 % (88 m) to 16 % (120 m) and 29 % (1030 m water depth).
The relative abundance of stained calcareous foraminifera (> 180 μm, Table 3) as well as their densities within the first centimeter (Table 6) decreased drastically from shelf to slope (i.e., from 161-198 at shelf stations 18 and 26 to 96 specimens per 50 cm 2 at Sta. 40).

Downcore pattern and microhabitat of benthic foraminifer species
Stations 18 and 26 showed the highest concentrations of stained organisms near the sediment surface (Fig. 2); indeed, 50 and 60 % (cumulative percent) of total living calcareous foraminifera were observed within the first centimeter at each station.In contrast, station 40 contained only 30 % of the total stained calcareous foraminifera at the surface, and a maximum was observed at 1-1.5 cm (~ 22 % of total living calcareous foraminifera) (Fig. 2).In addition, minor secondary abundance peaks were observed at station 18 and 40 between 5-6 cm (~ 5 %) and 3-4 cm (~ 9 %), respectively (Fig. 2).
Identification of species microhabitat was based on the presence or absence of an infaunal maximum, and on the shape of the distribution profile of each species (position with respect to other taxa; Fig. 3).Based on these assumptions, we defined the species N. auris, B. spissa, U. peregrine, B. seminuda and B. costata as shallow infauna, and G. affinis and C. oolina as deep infauna.The high abundance of shallow infauna and the absence of deep infauna characterized the shelf stations located under the OMZ's influence (station 18 and 26).In contrast, shallow and deep infaunal species were observed at the slope station 40 (1,030 m water depth).Living benthic foraminiferal census data of the more representative taxa in the > 180 μm fraction at Sta. 18, 26, and 40 (numbers are not standardized for a given sediment volume and refer to total organisms in each interval) within the upper 10 cm of the sediment column.
According to Sen Gupta & Machain-Castillo (1993), about ten calcareous foraminifer species typically live in oxygenpoor areas, with two or three species usually comprising as much as 80 % of the fauna.In our study, the number of living calcareous species for shelf stations 18 and 26 varied between 3 and 10 (Table 3 and 6); ~ 90 % of the total calcareous living benthic foraminiferal fauna were composed by one species (Table 5).The number of species is similar to that found in the Santa Barbara basin, California, by Phleger &Soutar (1973) andBernhard et al. (1997), who reported 7-13 stained species of calcareous foraminifera in the > 63 μm fraction (Table 6), and to Peru, where one species (Bolivina cf.pacifica) composes > 95 % of the living fauna (Phleger & Soutar 1973).In contrast, our data are much lower than those reported for the Oman Margin, where the total number of stained species observed by Gooday et al. (2000) varied between 49 (> 125 μm) and 64 (> 63 μm).Distribución vertical de las especies de foraminíferos calcáreos vivos más importantes (teñidos con Rosa de Bengala, fracción > 180 μm, estandarizados a 50 cm 3 de sedimento) dentro de los primeros 10 cm de cada testigo de sedimento.

Onshore-offshore distribution of living foraminifera
It is well known that benthic foraminiferal abundances in marine sediments are closely related to food availability, and foraminifera generally prosper where food is abundant (Gooday et al. 2000).Along our transect, we detected a decreasing trend in living foraminifera (> 180 mm) within the first centimeter with increasing water depth (from 161-198 at stations 18 and 26 to 96 specimens per 50 cm 2 at station 40, Table 6).This is most probably induced by lower and less labile organic matter flux reaching the sea floor in deeper settings such as those of Sta. 40 (Schubert et al. 2000, Molina et al. 2004, Muñoz et al. 2004), and/or a reduction in predation pressure at oxygen-depleted settings such as stations 18 and 26 (Gooday 2003).Although available Chl-α and TOC measurements do not coincide with the sampling times for foraminifera, on hand data show that station 40 is generally characterized by lower levels of Chl-α concentrations and TOC (~ 13 mg g -1 and < 3 %, respectively) than the shelf stations (Table 2).
In order to gain access to food, benthic foraminifers must tolerate the oxygen depletion that frequently accompanies abundant organic matter.Reduced oxygen concentrations and/or recurring anoxia events will influence the size distribution and taxonomic composition of foraminiferal assemblages, eliminating the less tolerant species, whereas abundance and dominance increase for the remaining species (Sen Gupta & Machain-Castillo 1993, Gooday et al. 2000).Our observations are consistent with these concepts, showing minimal presence of large foraminifera and the dominance of a low diversity assemblage in the oxygen-poor conditions of the shallowest stations (18 and 26), which are associated with the upper boundary of the OMZ.Here, the shelf restricted N. auris (Resig 1981) (in fraction > 180 μm, Table 5) composed over 90 % of the total living calcareous foraminifera fauna.A quick scan of the > 63 μm fraction revealed that N. auris was dominant (data not shown) at these shelf stations and was accompanied by juvenile bolivinids (about 30 %) at station 26.
The dominance of N. auris at the two stations associated with the OMZ differs from observations by Ingle et al. (1980), who found typical OMZ foraminiferal fauna (> 250 μm fraction) in central Chile to be primarily composed by bolivinids such as Bolivina seminuda and Bolivina costata.Bolivinids had a minimal presence or were absent in the 180 μm fraction at Sta. 18 and 26 and, as stated above, juveniles were only observed in the > 63 μm fraction at station 26.The discrepancy between our data and those of Ingle et al. (1980) may be explained by the fact that bolivinids are commonly associated with stable OMZ conditions (Heinze & Wefer 1992), which is not the case for our study area with a seasonally variable OMZ (Paulmier et al. 2006).In contrast, the living assemblage observed at station 40 (dominated by B. spissa, G. affinis, U. peregrina, C. oolina) agrees with the one reported previously by Ingle et al. (1980) at similar water depths.

Vertical distribution of living foraminiferal fauna within the upper sediment column (10 cm)
The relationship of the benthic foraminiferal microhabitat with food availability and oxygenation of the benthic realm has been schematized in a conceptual model by Jorissen et al. (1995).The TROX-model explains that, in oligotrophic environments, species adapted to low organic flux conditions will thrive close to the sediment-water interface in a welloxygenated setting.The low flux of organic matter to the sediment prevents colonization of the deeper sediment layers by infaunal taxa.In eutrophic conditions, on the other hand, where the principal redox front is positioned close to the sediment surface, infaunal taxa are limited to the first millimeters or centimeters of the sediment (Murray 2001); in this case, they have only a limited tolerance for low oxygen conditions.
Faunal penetration is maximal in "mesotrophic" settings because the oxygen penetration is relatively deep and more or less labile food particles are introduced at depth in the sediment by bioturbating macrofauna creating suitable microhabitats for benthic foraminifera (Jorissen 1999, Fontanier et al. 2002).In such environments, the subsurface accumulation of organic matter (Rathburn & Corliss 1994) and/or grazing on populations of anaerobic bacteria associated with redox boundaries (Fontanier et al. 2002) could explain the microhabitat selection of a given foraminifer.
In general terms, the vertical patterns observed at our stations agree with the TROXmodel.At stations 18 and 26, the living foraminifera are located preferentially near the sediment surface (50 and 60 % of the total living foraminifera within the first centimeter, Fig. 2).Particularly, species tolerant to low oxygen conditions with shallow infaunal microhabitat preferences (Bernhard & Sen Gupta 1999) at stations 18 and 26 indicate eutrophic conditions.On the other hand, the deepest penetration of foraminiferal fauna at station 40 is more typical of mesotrophic conditions (70 % of total living calcareous foraminifera occur below the first centimeter).Here, an important proportion of the living population is composed of deep infaunal species such as G. affinis (~ 22 %) and C. oolina (~ 12 %) (Table 5).This general scenario would suggest that BWDO is the main controlling agent at stations 18 and 26, whereas food supply controls the distribution of benthic foraminifera at station 40.
The extensive vertical distribution of N. auris at station 18, even with a small peak of stained organisms in the 5-6 cm level (Fig. 2), is interesting to note.This feature may have two explanations.On the one hand, it could refer to an artifact of the Rose Bengal method which may have stained the protoplasm of dead foraminifera (Material and Methods section) at this shallow station.On the other hand, it may be explained by the strong seasonal variation in bioturbation affecting the upper 2-5 cm of the sediment column associated with highly seasonally variable BWDO concentrations and organic matter supply (Muñoz et al. 2007, Table 2).In fact, 210 Pb activity profiles show an average mixed layer depth of 5 cm at station 18 (Muñoz et al. 2007).An additional factor to the pattern observed at station 18 could include the denitrification capabilities of N. cf stella (most probably N auris, according to published photographs, Risgaard-Petersen et al. 2006) which, although pushed deeper down the sedimentary column by bioturbation, would allow them to remain alive.
In summary, our results show that a clear difference exists between shelf stations 18 and 26 which are dominated mostly by one species tolerant to oxygen depletion, and slope station 40 where oxygen availability and food introduction within the sediment allow the occurrence of shallow and deep infaunal species.

Fig. 1 :
Fig.1: (A) Study area showing the position of the sampling stations off Concepción used in this study: station 18 (middle shelf), 26 (outer shelf), and 40 (on the middle slope), and bathymetry; (B) T-S diagram (April 2003) for station 40 showing the different water masses present: Subantarctic Water (SAAW), Equatorial Subsurface Water (ESSW), and Antarctic Intermediate Water (AAIW).See text for details.

(
Fig.3: Vertical distribution of the most important taxa of living benthic calcareous foraminifers (Rose Bengal stained, fraction > 180 μm, standardized to 50 cm 3 sediment volume) within the upper 10 cm of each core.

TABLE 4
Living (stained) benthic foraminifers identified in this study