RESULTS

Stained Benthic Foraminifera >250 µm

Of the 35 samples obtained from the inner Namibian shelf (Table 1), four were found completely barren of foraminiferal tests, and 15 were barren of stained foraminifers in the size class >250 µm. Of the remaining 16 samples, Virgulinella fragilis (Grindell and Collen) was detected in 12 samples (Table 2, Figure 2.1-3.), and thus can be considered a common species. Nonionella stella Cushman and Moyer (Figure 2.7-8.) and Discammina compressa (Goes) occur in five samples. Crithionina pisum Goes and Hyperammina sp. each were found once, i.e., at station 680. All named species occur within low dysoxic to anoxic conditions (0 to 0.3 ml l^-1). On the outermost edge of the DMB, at 152 m water depth, the oxygen values rise to 0.5 ml l^-1. Ammonia beccarii (Linné) exclusively occurs in this area, where it co-occurs with D. compressa. The standing stock of this community (station 212, Table 2) is elevated by about two orders of magnitude, in comparison to the more oxygen depleted DMB stations from 74 to 132 m water depth. For these shallower sites, the standing stock of the populations consistently ranges below 1 ind. per 10 cm³, with V. fragilis as a constant member. This depth range corresponds to the depth range of the bacterial blooms and anoxic bottom water masses described by Lavik et al. (2009). Towards nearshore, at 70-50 m water depth, N. stella and D. depressa display an increased standing stock, while V. fragilis is absent. The shallowest station yielding stained foraminifera occurs at 35 m water depth and contains both V. fragilis and N. stella (station 830, Table 2). Enhanced methane production by submarine seepage of Namibian groundwater aquifers has been postulated for the shallow shelf, but reliable near shore seepage quantifications are available for only one aquifer at present (see Altenbach and Struck 2006).

Unstained Tests >250 µm

The most common species at 35 m water depth, as well as for the depth range from 74 to 132 m is Virgulinella fragilis, and sporadically co-occuring Nonionella stella and Discammina compressa. At 50-70 m water depth, D. compressa and N. stella are most common in the unstained assemblage (Figure 3). This corresponds with the distributional pattern observed for the stained species (Table 2). Ammonia beccarii is most common at the outermost edge of the DMB (station 212), where it co-occurs with D. compressa. In addition, Cibicides lobatulus Walker and Jacob (referred to Lobatula by some authors), Uvigerina mediterranea Hofker, Cancris auriculus (Fichtel and Moll), Hyalinea balthica (Schroeter), two elphidid, and two quinqueloculinid species are indicative of more open marine conditions. However, a large number of heavily corroded and fragmented tests of textulariid, miliolid, or calcitic origin (gen. et. sp. indet.) suggest that reworking and onshore transportation from the outer shelf may have occurred. These intruders are found sporadically in the DMB, either introduced by temporal migration during increased oxygen availability, or by lateral advection. They usually occur close to the edges of the mud, at water depths of either above 74 m (greenish marker colors in Figure 3) or deeper than 132 m (yellow to reddish marker colors in Figure 3). The only exception is represented by Cibicidoides pseudoungerianus (Cushman) and C. lobatulus (stations 730 and 810 in Figure 3). These taxa, which have a typical epibenthic lifestyle are most commonly advected over broad distances within shelf environments and down slope (Hald and Vorren 1987, Mudie et al. 1984). At station 810, elevated oxygen concentrations (0.7 ml l^-1) indicate the local intrusion of more oxygen saturated water masses contemporary to our sampling. In water depths shallower than 74 m, Nonion asterizans (Fichtel and Moll) and Quinqueloculina venusta Karrer are the most important taxa (Figure 3). Light to dark blue marker colors indicate the species that also were recovered stained from the DMB, and thus can be considered autochthonous (Figure 3). Four stations were omitted in Figure 3: stations 187 and 199 only contained test fragments (gen. et sp. indet.), while station 181 yielded only a single test of Cassidulina cf. teretis (or the sibling shallow water morphotype described by Seidenkrantz 1995), and from station 184 one slightly corroded test of Cribrostomoides cf. jeffreysi (referred to Labrospira by some authors) was recovered.

Stained Benthic Foraminifera 150-250 µm

Stained benthic foraminifers were recovered from seven of 19 stations (Table 3). Four of these produced Virgulinella fragilis, Nonionella stella, or Ammonia beccarii at identical depth ranges as compared to the larger fraction. Due to shape and size, Bolivina pacifica Cushman and McCulloch (Figure 2.6.) was only recovered from this size fraction at five stations with O₂ concentrations from 0-0.4 ml l^-1. Fursenkoina fusiformis (Williamson) (Figure 2.4-5., referred to Stainforthia by some authors) was observed from four stations, with high standing stock (>20 ind. per 10 cm³) at stations 178 and 188 in spite of the microxic conditions. This species was also recovered from the larger size fraction, but did only occur in the dead count, together with a number of unstained tests of other species.

Stained V. fragilis, N. stella, F. fusiformis and B. pacifica were also recovered deep infaunal from core 199 (Figure 4). F. fusiformis occurs in all samples down to 15 cm, V. fragilis and B. pacifica occur at 9-11 cm, and N. stella settles in the deepest layer at 11-19 cm sediment depth. The standing stock of these populations increases at 3-5 cm sediment depth by nearly one order of magnitude, in comparison to the sediment surface; all shifts of standing stock down core are closely connected to the amount of unstained tests preserved (Figure 4).

Unstained Tests 150-250 µm

A pattern of co-occuring Fursenkoina fusiformis, Virgulinella fragilis, Nonionella stella, and Bolivina pacifica stretches from 70 to 143 m water depth. Intruders are found seaward from 152-143 m water depth, and from 44-74 m on the landward flank of the DMB (Figure 5), similar to the larger size fraction (Figure 3). The dominance of unstained Discammina compressa at 50-70 m water depth (Figure 3, >250 µm) is not recognizable in the smaller fraction (Figure 5). The abundance of V. fragilis in the larger size fraction (Figure 3) is less significant in the smaller size fraction (Figure 5).

The taphocoenoses represented by core 199 are composed of the four species recovered stained downcore, together with sporadic occurrences of Uvigerina mediterranea at 3 to 4 cm, Cibicides lobatulus at 4 to 5 cm, Ehrenbergina undulata Parker at 9 to 11 cm, and Nonionella grateloupi (d'Orbigny) at 15 to 19 cm sediment depth. The number of unstained tests increases by about two orders of magnitude at 3 to 5 cm sediment depth (Figure 4).

Imprint of Size Classes

Of a total of 16 surface samples from cruise M57/3, unstained tests were detected in 12 samples representing the size fraction >250 µm (Figure 3). This recovery increases by 25% if the size fraction considered is extended down to >150 µm (15 stations, Figure 5). Six additional taxa are recorded, which is inevitably the case for all early ontogenetic stages, and all taxa with a small or slender growth, such as many bolivinids and buliminids. However, this advantage can only be retained when the samples are scrutinized completely. If smaller grid sizes are used for sieving foraminifera, even a small split fraction of the sample may contain an abundance of small-sized species. This may suggest that the analysis of a small sample volume derived from a split fraction is sufficient in order to adequately assess the diversity of foraminifera. However, by restricting the screening process, the likelihood of discovering a rare species is reduced because these species may only be represented by one individual in the entire sample. The abundance of stained foraminifera within the DMB is usually near 1 to 0.1 individual per 10 cm³ sediment. Dale and McMillan (1998) studied the small fraction (i.e., >125 micron) from the DMB based on sample sizes of 10 g of dry sediment and discovered a diverse fauna with many small growing species. However, the two taxa most commonly stained in our samples, Virgulinella fragilis and Nonionella stella, were not discovered by these authors. Based on the preceding considerations we submit that the analysis of foraminifera from extreme environments should be based on the screening of particularly large samples.

Smaller specimens of Virgulinella fragilis, Nonionella stella, and Fursenkoina fusiformis were recorded stained for six stations in the size fraction 150-250 µm (Table 3), but none of these stations contained stained tests in the size fraction >250 µm (Table 2). Vice versa, adult stages of V. fragilis were recovered from three stations in the size fraction >250 µm, but these stations lacked specimens in the smaller size fraction. Such isolated cohorts result from rare and irregular reproduction, or from repeated and complete loss of recruitment.

Teratological tests of Virgulinella fragilis were commonly observed from the size fraction >250 µm in the year 2000 (cruise M48/2), ranging at 7-63% for the counts of stained and unstained tests. Most of the teratological tests display fusion of two embryonic macrospheres, or attachment of a juvenile to an adult test (figured in Altenbach and Struck 2006). With decreasing standing stock of the population, proportion of teratological tests increases significantly (Figure 6, black circles; linear regression, r = -0.85, p >99%; exponential regression exp(-0.0634 * x), r = 0.95, p >99.9%). No correlation is indicated for the number of unstained tests (Figure 6, open circles). Three years later (cruise M57/3), not a single teratological structure was recovered from the size fraction of >250 µm, neither from the stained, nor from unstained tests. In the size fraction 150-250 µm, however, unstained juvenile specimens with teratological tests were observed from 4 surface samples (Figure 2.2.; Figure 6, open squares). All these tests were slightly to heavily corroded or broken.