386
Romeroetal.
Factor 6, which explains -6% of the variance,
is dominated by Fragilariopsis doliolus and
Thalassionema nitzschioides var. nitzschioides,
and is related to factor 2 (see overlapping contours
in Fig. 9). In addition, the common occurrence of
F. kerguelensis in the southernmost cores may
have played a role in the grouping of samples (compare with factor 5 of Pokras and Molfino 1986)
(Fig. 9, Table 4). This antarctic/subantarctic species is associated with displacement and northward
transport by bottom currents (Burckle and Stanton
1975; Pokras and Molfino 1986). Results from
Treppke et al. (1996b) show that transportation of
F. kerguelensis may also happen at intermediate
depths within the Antarctic Intermediate Water.
Highest loadings of this factor are confined to a
very few cores south of the Walvis Ridge (Table
5).
Results from the traps and the surface
sediments suggest that the mixture of species with
different ecological affinities recorded between the
Walvis Ridge (factor 2) and the Mid Cape Basin
(factor 6) (Fig. 9) mirrors the complicated hydrology of this transitional region of interplay between
oceanic and coastal conditions.
Factor 3 accounts for -14% of the variance,
and is dominated by Chaetoceros resting spores,
accompanied by Thalassionema nitzschioides
var. nitzschioides as subdominant (Table 4). It extends between the Congo and the Cuanza river
mouths, and southwards from the Cunene to the
Orange river mouths (Fig. 9). It reflects the high
productivity coastal band associated with lowered
sea surface salinity and nutrient input (at the Congo;
Pokras and Molfino 1986; van Iperen et al. 1987),
and the characteristic upwelling cells and filaments
of the Benguela System (Shannon and Nelson
1996). This factor coincides with observations by
Schiitte and Schrader (1981) concerning resting
spores of Chaetoceros as the main contributors to
the diatom assemblage between 19° and 24 OS offshore SW Africa, as well with Pokras and Molfmo's
(1986) factor 3 in the high abundance of T.
nitzschioides var. nitzschioides.
Factor 7, which accounts for-4% of the variance, represents the continental signal given by
freshwater diatoms (mainly Aulacoseira
granulata and Cyclotella spp.) (Table 4). It ineludes only two sites, in the N-Guinea Basin and
offshore the Congo river mouth. The presence of
freshwater diatoms in the N-Guinea Basin was
associated with eolian input by the NE trades rather
than riverine influence from the Niger (Lange et
al. 1994; Treppke et al. 1996a). The high abundance of freshwater taxa at the mouth of the Congo
is obviously related to river input (van Iperen et al.
1987). The anomaly that the freshwater diatoms
having highest loadings not directly at the river
mouth but in water depths below 2000 m may be
explained by dilution with resting spores elose to
the coast (van Iperen et al. 1987), and/or may be a
consequence of the particular hydrology and nutrient distribution of the Congo-fan area (Eisma and
van Bennekom 1978; van Bennekom et al. 1978).
Moderate loadings of T. nitzschioides var.
nitzschioides within this factor (Table 4) may reflect the so-called river induced upwelling (van
Iperen et al. 1987).
Finally, factor 4 (-6% oftotal variance) is represented by oceanic warm water forms, T.
nitzschioides var. parva, Nitzschia bicapitata
group and Roperia tesselata (Table 4). However,
we believe that this factor is a statistical artifact
due to the scarcity of samples. Also, with the exception of the site 1701 (Fig. 8), all other sites within
factor 4 have components from as many as four
other factors. We therefore chose to represent this
factor as a dottted area without a defined contour
in Fig. 9.
Summary ofthe Principal Findings ofthe
Tropical and Equatorial Atlantic
As part of the Bremen University trapping
program and in an attempt to contrast different
upwelling areas (coastal vs. oceanic) and
oligotrophic gyres, we have investigated the temporal variations in the fluxes of biogenic opal and
diatoms offNW Africa, Itt the Walvis Ridge, and
in the eastern and western equatorial Atlantic. We
observed that:
(1) Regional differences in diatom fluxes and
assemblages reflect (a) atmospheric and oceanographic circulation patterns related to the Trade
Wind system and seasonal migration of the ITCZ,
Romeroetal.
Factor 6, which explains -6% of the variance,
is dominated by Fragilariopsis doliolus and
Thalassionema nitzschioides var. nitzschioides,
and is related to factor 2 (see overlapping contours
in Fig. 9). In addition, the common occurrence of
F. kerguelensis in the southernmost cores may
have played a role in the grouping of samples (compare with factor 5 of Pokras and Molfino 1986)
(Fig. 9, Table 4). This antarctic/subantarctic species is associated with displacement and northward
transport by bottom currents (Burckle and Stanton
1975; Pokras and Molfino 1986). Results from
Treppke et al. (1996b) show that transportation of
F. kerguelensis may also happen at intermediate
depths within the Antarctic Intermediate Water.
Highest loadings of this factor are confined to a
very few cores south of the Walvis Ridge (Table
5).
Results from the traps and the surface
sediments suggest that the mixture of species with
different ecological affinities recorded between the
Walvis Ridge (factor 2) and the Mid Cape Basin
(factor 6) (Fig. 9) mirrors the complicated hydrology of this transitional region of interplay between
oceanic and coastal conditions.
Factor 3 accounts for -14% of the variance,
and is dominated by Chaetoceros resting spores,
accompanied by Thalassionema nitzschioides
var. nitzschioides as subdominant (Table 4). It extends between the Congo and the Cuanza river
mouths, and southwards from the Cunene to the
Orange river mouths (Fig. 9). It reflects the high
productivity coastal band associated with lowered
sea surface salinity and nutrient input (at the Congo;
Pokras and Molfino 1986; van Iperen et al. 1987),
and the characteristic upwelling cells and filaments
of the Benguela System (Shannon and Nelson
1996). This factor coincides with observations by
Schiitte and Schrader (1981) concerning resting
spores of Chaetoceros as the main contributors to
the diatom assemblage between 19° and 24 OS offshore SW Africa, as well with Pokras and Molfmo's
(1986) factor 3 in the high abundance of T.
nitzschioides var. nitzschioides.
Factor 7, which accounts for-4% of the variance, represents the continental signal given by
freshwater diatoms (mainly Aulacoseira
granulata and Cyclotella spp.) (Table 4). It ineludes only two sites, in the N-Guinea Basin and
offshore the Congo river mouth. The presence of
freshwater diatoms in the N-Guinea Basin was
associated with eolian input by the NE trades rather
than riverine influence from the Niger (Lange et
al. 1994; Treppke et al. 1996a). The high abundance of freshwater taxa at the mouth of the Congo
is obviously related to river input (van Iperen et al.
1987). The anomaly that the freshwater diatoms
having highest loadings not directly at the river
mouth but in water depths below 2000 m may be
explained by dilution with resting spores elose to
the coast (van Iperen et al. 1987), and/or may be a
consequence of the particular hydrology and nutrient distribution of the Congo-fan area (Eisma and
van Bennekom 1978; van Bennekom et al. 1978).
Moderate loadings of T. nitzschioides var.
nitzschioides within this factor (Table 4) may reflect the so-called river induced upwelling (van
Iperen et al. 1987).
Finally, factor 4 (-6% oftotal variance) is represented by oceanic warm water forms, T.
nitzschioides var. parva, Nitzschia bicapitata
group and Roperia tesselata (Table 4). However,
we believe that this factor is a statistical artifact
due to the scarcity of samples. Also, with the exception of the site 1701 (Fig. 8), all other sites within
factor 4 have components from as many as four
other factors. We therefore chose to represent this
factor as a dottted area without a defined contour
in Fig. 9.
Summary ofthe Principal Findings ofthe
Tropical and Equatorial Atlantic
As part of the Bremen University trapping
program and in an attempt to contrast different
upwelling areas (coastal vs. oceanic) and
oligotrophic gyres, we have investigated the temporal variations in the fluxes of biogenic opal and
diatoms offNW Africa, Itt the Walvis Ridge, and
in the eastern and western equatorial Atlantic. We
observed that:
(1) Regional differences in diatom fluxes and
assemblages reflect (a) atmospheric and oceanographic circulation patterns related to the Trade
Wind system and seasonal migration of the ITCZ,