BENTHIC FORAMINIFERAL MORPHOGROUPS ON THE ARGENTINE
CONTINENTAL SHELF

MARTA INÉS ALPERIN
1,3, GABRIELA CATALINA CUSMINSKY

2
AND EMILIANA BERNASCONI

2

ABSTRACT

The aim of this study is to evaluate the relative influence of
abiotic factors on the association and spatial distribution of
recent benthic foraminiferal morphogroups in the Argentine
continental shelf at ,40uS environments as a first step towards
establishing their paleoecological significance. Foraminifera
are classified into five morphogroups: tapered, elongate-
flattened, milioline, planoconvex, and rounded-planispiral.
Compositional data analysis techniques were used to define
morphogroup assemblages, and Classification and Regression
Tree Analysis was used to identify environmental variables.
The distribution of the morphogroup assemblages were
recognized was driven by complex interactions between
environmental factors. The most important factor is temper-
ature, although salinity, substrate grain size, and hydrody-
namic energy also correlate with the distribution of mor-
phogroups. The morphogroups analysis shows potential for
determining present and past environments where the auto-
ecology of the species is unknown or where there is doubt
regarding their taxonomic classification.

INTRODUCTION

Foraminifera are one of the most widely used groups of
organisms for studying marine environments and recon-
structing coastal paleoenvironments, including those of
Argentina (e.g., Ferrero, 2006; Bernasconi and Cusminsky,
2007; Laprida and others, 2007). The distribution of the
different living foraminiferal species is related to, and
mainly controlled by, variations in the physicochemical
properties of the marine environment such as temperature,
salinity, substrate, dissolved oxygen, and nutrients (Bol-
tovskoy, 1966; Murray, 1991, 2006). This has led to many
studies of foraminifera involving issues related to commu-
nity ecology (e.g., Hayward and others, 1996, 2002;
Bernasconi and Cusminsky, 2005; Bernasconi, 2006;
Alperin and others, 2008; Bernasconi and others, 2009).

A morphogroup is an aggregation of forms with similar
test morphology, independent of systematic relationships
(Murray, 1973, 2006). Reolid and others (2008) note that
using morphological categories in paleoenvironmental
analyses may be preferred over the use of formal species
identifications because: 1) the morphological approach
enables reliable comparisons to be made among assemblag-
es of different ages, reducing the effect of taxonomical

divergence caused by biological evolution, 2) identifications
of species are not required, and 3) using a small number of
morphogroups instead of a large number of species reduces
the amount of data to be analyzed. In an earlier study,
Severin (1983) noted that morphogroups have great
potential for determining paleoenvironments, microhabitat
patterns, and feeding strategies because the categories are
independent of taxonomy and thereby avoid much of the
subjectivity involved in identifying species; thus, specimens
can be classified quickly and easily.

The relationship between benthic foraminiferal test
morphology and characteristics of the physical environment
has been the subject of numerous studies. Years after Bandy
(1960) pioneered the morphogroup concept, Severin (1983)
used it to infer four biofacies related to bathymetry, which
Bernhard (1986) then used in relating morphogroups to
oxygen levels. Corliss (1985) and Corliss and Chen (1988)
used shape and mode of coiling to determine microhabitat
patterns (epifaunal vs. infaunal) and linked them to organic
content. Corliss (1985) found test shape and pore distribu-
tion, are related to the microhabitat preferences of the
different species. Kaiho (1991, 1994, 1999) noted that
changes in oxygen concentration at the water-sediment
interface play an important role in controlling benthic
foraminiferal populations. Kaiho (1991) estimated oxygen-
ation conditions of the water-sediment interface by means
of the Benthic Foraminifera Oxygen Index (BFOI), which is
based on morphological and taxonomical differences
among benthic foraminifera, and extrapolated estimates
of oxygenation to the past. Bernasconi (2006), Bernasconi
and Cusminsky (2009), and Bernasconi and others (2009)
also used BFOI to infer levels of paleo-oxygenation in
Quaternary sediments of Nuevo Gulf, Argentina.

The spatial distribution of benthic foraminiferal mor-
phogroups is variable. In addition to small-scale changes
that may be caused by oxygenation conditions at the water-
sediment interface, there are other local factors that
intervene, such as salinity, temperature, nutrient availabil-
ity, substrate, bioturbation, and currents. Further studies
are needed to determine what factors cause joint occurence
of benthic foraminiferal morphogroups. Therefore we chose
to study the distribution of foraminiferal morphogroups on
the Argentine Continental Shelf, as it includes different
environments that are closely linked and suitable for the
development of benthic foraminifera. San Matı́as Gulf has
oceanographic dynamics characterized by large tide ampli-
tudes that have a major effect on the vertical mixing of
waters and circulation, while the shelf has local and
seasonal variations in water temperature, salinity, and
productivity. The aim of this study is to evaluate the relative
influence of abiotic factors on the association and spatial
distribution of benthic foraminiferal morphogroups in these
shelf environments as a first step towards establishing their
paleoecological significance.

1 Cátedra de Estadı́stica, Facultad de Ciencias Naturales y Museo,
Universidad Nacional de La Plata, Buenos Aires, Argentina; Edificio
de Institutos y Laboratorios, Calle 64 y 120, 1900 La Plata, Buenos
Aires, Argentina

3 Correspondence author: E-mail: alperin@fcnym.unlp.edu.ar;
marta.alperin@gmail.com

2 Centro Regional Universitario Bariloche, Universidad Nacional
del Comahue – INIBIOMA, Consejo Nacional de Investigaciones
Cientı́ficas y Técnicas, Calle Quintral 1250, 8400 San Carlos de
Bariloche, Rı́o Negro, Argentina

Journal of Foraminiferal Research, v. 41, no. 2, p. 155–166, April 2011

155



STUDY AREA

The study area is located between 39–43u S and 58–65u W
and includes San Matı́as Gulf and part of the Argentine
continental shelf (Fig. 1.1). The shelf gently slopes with a
fairly uniform gradient. Its sedimentary cover is mainly
sand, with mean grain size decreasing towards the
continental slope. The sediment also becomes siltier
towards Bahı́a Blanca and San Matı́as Gulf. Sandy
sediment spreads from the mouth of Negro River to the
coast (Aliotta, 1983). At the mouth of San Matı́as Gulf, the
sediment contains gravel while the coarse fraction decreases
toward the northeast (Gelós and others, 1988).

The hydrology in the region is complex. Water masses
over the continental shelf are of subantarctic origin and
mean circulation is northward, although the San Matias
Gulf and semi-protected coastal area of El Rincón are
characterized by local conditions. Martos and Piccolo
(1988) differentiated two domains on the shelf: 1) a coastal
or inner domain, and 2) an outer domain. The coastal
domain includes the littoral zone down to the 40–50 m
isobaths with vertically homogenous waters. In this area,
Guerrero and Piola (1997) and Lucas and others, (2005)
recognized Coastal Water 1) of low salinity (,33.4) in El
Rincón estuary due to discharge from the Negro and
Colorado rivers; and 2) of high salinity (33.8–34.0) at the
mouth of San Matı́as Gulf and near the coast area east of
the estuary. The coastal water is separated from the
continental shelf water by the Coastal Front (Guerrero y
Piola, 1997; Guerrero, 1998; Lucas and others, 2005).
Coastal circulation is SSW to NNE. The outer domain
extends from depths of 40–50 m to 100 m (approximate
shelf edge). This sector was described by Guerrero and
Piola (1997) as Middle Shelf Water (S 33.4–33.7) compris-
ing the northward-flowing Patagonian Current. This water
is homogenous in winter and stratified in summer. During
spring–summer, surface-water temperatures and salinity are
typical of the Rı́o de la Plata (T .15uC; S ,33) and flow

southwestward. In autumn–winter, surface waters are
cooler (mean T 12uC) and the water column is vertically
mixed by convection and influenced by the Malvinas
Current. The salinity values are typical of a continental
shelf (Martos and Pı́ccolo, 1988). Between the Middle Shelf
Water and the Malvinas Current, the Outer Shelf or
Continental Slope Water (T,15uC and S 33.7–34.0)
(Guerrero and Piola, 1997) is present. The thermohaline
shelf-break front (which separates the shelf waters from the
colder, more saline waters of the Malvinas Current) lies
between the 80–100 m isobaths (Martos and Pı́ccolo, 1988).
The Malvinas Current flows along the shelf break and carry
subantarctic water to the north and it had the same
direction over the total water column with velocities of the
order of 30 cm/s. The current has a dominating effect on
adjacent shelf waters (Martos and Piccolo, 1988).

San Matı́as Gulf, in the southwest of the study area
(40u479 S–42u139 S), is a 18,000 km2, semi-closed basin,
deeper than the adjacent continental shelf with which it
communicates over a sill at a depth of 70 m (Parker and
others, 1996). About 45% of the Gulf is shallower than
100 m, and there are large depressions that are .160 m
deep in the middle, with maximum depths .200 m (Piola
and Scasso, 1988). Substrates in the gulf are mainly mud,
gradually becoming fine sand towards the littoral zone
(Gelós and others, 1988). Sandy sediment spreads from the
mouth of Negro River to the coast (Aliotta, 1983). At the
mouth of the gulf, the substrate is sandy to gravelly (Gelós
and others, 1988; Achilli and Aliotta, 1992). There are two
different water masses in the gulf: the water to the north
and east is characterized by high temperature and salinity
(33.9–44.1) with a marked seasonal thermocline, while the
water to the south and southwest are unstratified and of
lower temperature (9–13uC) and salinity (Carreto and
others, 1995). The two water masses are separated by a
tidal front located in the northern part of the mouth of the
Gulf (Piola and Scasso, 1988; Gagliardini and Rivas, 2004).
Shelf waters flow in through the southern end of the mouth

FIGURE 1. 1 Location map. 2 Position of stations. Bathymetric contours are in meters.

156 ALPERIN AND OTHERS



and out through the northern end, producing a semi-
permanent gyre (Piola and Scasso, 1988; Scasso and Piola,
1988). This flow is more intense (,3.5 m/s) near the
entrance to the Gulf (Tonini and others, 2006).

MATERIALS AND METHODS

In 1984, 31 samples of surface sediment were obtained
using a Shipek sampler by the Argentine Institute of
Oceanography aboard the oceanographic ship Puerto
Deseado (Fig. 1.2). Table 1 presents sedimentological data
prepared by Gelós and others (1988) from these samples, as
well as physicochemical parameters of the bottom water
surveyed during the oceanographic campaign by Pucci and
others (1985). Samples were screen-washed and the .63 mm
residue was air dried. Five grams of dry sediment from each
sample were examined under a stereomicroscope. All
foraminiferal specimens (i.e., the total assemblage) were
extracted (Alperin and others, 2008).

Specimens previously classified according to taxonomic
criteria (Alperin and others, 2008) were reclassified using
only the outer morphology of their tests, a criterion
independent of traditional taxonomy. Agglutinated species
are excluded from this classification because none were
present. Specimens belonging to the species in Table 2 and
Figure 2 were sorted into the following morphogroups:
Tapered (TA): rounded or angular shapes in apertural view (Severin,

1983).

Elongate-flattened (EF): oval to compressed tests in apertural view
with parallel to subparallel sides (Severin, 1983).

Milioline (MI): ovate tests with apical aperture (Corliss and Chen,
1988), comprising porcelaneous foraminifera with a flattened
discoid coil and an elongate test (Reolid and others, 2008).

Planoconvex (PC): tests with one side flat and the other convex
(Severin, 1983).

Rounded-planispiral (RP): planispiral and trochospiral tests in which
neither the spiral or the umbilical side is visible in apertural view
(Severin, 1983).

The data matrix obtained from counting the specimens in
each sample was presented in proportions of PC, PR, MI,
AG, and EA (Appendix 1). Data expressed as proportion
are known as an Aitchison’s composition (Aitchison, 1986).
By definition, the proportions of a composition are relative
values adding up to one; thus, there is a statistical
dependence between the proportions. Aitchison (1986,
1997, 1999) warns that the study of its correlations provides
spurious results and analyzing the parts in isolation could
lead to erroneous interpretations. Instead, the analysis
should focus on the relations between the parts (Aitchison,
1986, 1999; Aitchison and others, 2000, Pawlowsky-Glahn
and Egozcue, 2001); Aitchison and Greenacre (2002),
Pawlowsky-Glahn and Egozcue, 2002; von Eynatten and
others, 2002). Utilization of log-ratios has proven to be a
good strategy (Aitchison, 1986, 1999; Kucera and Mal-
gram, 1998, Aitchison and others, 2000, Pawlowsky-Glahn
and Egozcue, 2001, Aitchison and Greenacre, 2002,
Pawlowsky-Glahn and Egozcue and others, 2002; von
Eynatten and others, 2002, Murray, 2006).

TABLE 1. Geographic location and physicochemical characteristics of water (Pucci and others, 1985) and substrate at each studied station (Gelós
and others, 1988).

Sample Latitude S Longitude W
Depth

(m)
Minimun distance

to coast (km) Silt % Sand % Gravel % Temp uC
Salinity

(psu)
D.O.
(ml/l) Cluster

1 39u19.39 59u13.29 57 65.2 5 95 0 18.23 34.08 5.72 5
2 39u30.79 58u49.49 72 93.8 3 97 0 9.71 33.65 6.05 2
3 39u42.09 58u28.09 78 123.3 4 96 0 7.34 33.74 4.82 1

13 40u32.59 58u46.59 82 206.1 8 92 0 7.13 33.77 3.70 1
14 40u20.89 59u11.59 78 175.8 5 95 0 7.55 33.71 4.48 1
15 40u10.49 59u32.29 67 150.5 2 98 0 11.20 33.65 5.60 2
16 39u58.29 59u57.29 51 122.1 1 99 0 17.22 33.82 5.33 4
17 39u47.39 60u19.39 46 96.7 6 94 0 17.87 33.95 3.87 4
18 40u30.09 60u59.59 42 97.3 1 94 5 17.16 33.75 5.59 4
19 40u41.29 60u37.79 53 130.6 2 98 0 16.91 33.66 5.12 4
20 40u51.19 60u12.69 66 167.5 3 97 0 16.99 33.71 3.96 5
21 41u01.29 59u49.39 74 202.9 5 95 0 9.08 33.69 3.53 2
22 41u10.09 59u25.59 75 239.0 8 92 0 8.56 33.79 5.40 1
27 41u18.59 61u32.99 48 80.6 5 85 10 16.85 33.68 5.40 4
28 41u26.19 61u07.29 54 118.2 4 90 6 16.19 33.61 4.80 4
31 41u52.69 59u53.29 82 231.9 5 95 0 7.18 33.67 3.25 2
39 42u24.69 60u50.99 80 209.7 6 94 0 8.05 33.51 4.19 5
41 42u09.29 61u41.39 69 149.6 9 91 0 12.85 33.57 5.52 5
42 42u01.29 62u06.19 69 120.8 6 81 13 17.39 34.04 5.21 4
44 41u45.09 62u54.59 54 67.7 1 72 27 17.12 33.98 5.40 3
45 41u36.59 63u22.59 48 49.3 1 59 40 17.14 34.20 5.80 3
46 41u28.59 63u46.29 56 34.6 1 99 0 16.05 34.12 4.92 2
47 41u18.29 64u16.39 125 32.8 85 15 0 11.62 34.05 4.53 5
48 41u07.19 64u52.49 96 24.4 49 51 0 15.71 34.08 3.89 4
49 41u37.19 64u50.29 88 15.0 18 82 0 15.57 34.03 5.07 5
50 41u32.99 64u30.69 160 42.6 98 2 0 10.68 34.09 4.68 5
51 42u04.19 64u42.69 139 16.1 98 2 0 12.85 34.02 4.52 5
52 41u59.99 64u10.19 106 20.8 13 87 0 13.38 34.03 4.65 4
53 41u43.29 64u04.99 84 43.6 20 80 0 14.55 34.04 5.79 3
54 42u27.49 63u16.49 60 27.4 2 36 62 15.54 33.72 5.16 4
55 42u35.89 62u49.39 69 62.9 4 78 18 15.01 33.66 5.14 4

MORPHOGROUPS ON THE ARGENTINE SHELF 157



Statistical Compositional Analysis was not possible
because 36% of the data matrix show zero values. As per
Martı́n-Fernández and others (2003), a replacement multi-
plicative strategy was taken in which zero values were input
as 0.005. To break statistical dependence and achieve
independent parts, we followed Aitchison (1986) by
performing the centered log ratio transformation (clr) (for
a sample log ratio between each part and sample geometric
mean):

clrxi~ ln
p1

g xið Þ
, ln

p2

g xið Þ
,:::, ln

pd

g xið Þ

� �
~z1,z2,:::,zd ;

g xið Þ~ P
d

i~1
pi

� �1=d

where: xi 5 sample i, pi 5 morphogroup proportion, and
g(xi) 5 geometric mean for sample i.

The composition of morphogroups was described ac-
cording to Aitchison (1986, 1997), using Center (closed
geometric mean of the compositional data matrix) and a
measure of total variability TotalVar (trace of the
symmetric centered logratio covariance matrix). The
variability of each morphogroup was described using the
variance clr (Var-clr).

A principal component analysis (PCA) was performed on
the clr-transformed data and the covariance matrix.
Subsequently, two-dimensional biplots were constructed
using principal components (PC). Biplots are graphical
displays of the four-dimensional morphogroups and
samples space. Biplots can be used to identify the existence
and nature of relationships between the morphogroups and
clusters of samples (Aitchison and Greenacre, 2002).

Classification and regression trees analysis (C&RT;
Breiman and others, 1984) was performed in order to
capture and model potential non-additive interactions and
nonlinear relations among the clusters of samples identified
on the biplot and six explanatory variables. The explana-
tory variables selected were location (depth and minimum
distance from shore), substrate ratio of sand:silt (ln(sand/
silt)), and three variables describing the physicochemical
conditions of the water (temperature, salinity, and dissolved
oxygen) (Table 1). The classification tree is constructed
using an algorithm that recursively searches values in all
explanatory variables to split the samples into two groups,
called nodes, such that the variance or impurity within the
node is minimized. The Gini diversity index (Gini, 1912)
was used to calculate node impurity. This splitting
procedure was repeated until node size reached a minimum
two samples or until the variance was reduced to 0.001. The
classification tree was generated when one of these criteria
was either reached or surpassed on all branches. In this
way, the final grouping was guaranteed to contain a
reasonable sample size and each variable added some
explanatory power to the model. Criteria for tree growth
and pruning are used to find out how much a tree must
grow and how many final nodes are reasonable. For a small
number of samples, Breiman and others (1984) propose a
V-fold cross-validation method with V 5 10. This
procedure starts by dividing the 31 stations in 10 equal
size groups. It continues by generating a tree with 9 of the
10 groups and utilizing the 10th group to test the tree’s
precision. This process is repeated 10 times, utilizing a
different group each time to test the tree’s precision. For
each test, the error in the tree’s prediction is given by the
increment on the number of terminal nodes. The tree with
the smallest classification risk calculated with this proce-
dure was displayed.

RESULTS

The 4690 classified specimens are distributed among the
morphogroups as 43% rounded-planispiral (RP), 41%
planoconvex (PC), 11% elongate-flattened (EF), 4% ta-
pered (TA), and 2% milioline (MI). The PC morphogroup
is present in all samples; RP is in 28 samples, and absent
only from samples from the mouth of San Matı́as Gulf (44,
45, 53); EF occurs in 20 samples from the shelf, but is
absent from samples from San Matı́as Gulf (45, 46, 53) and
the deepest part of the outer shelf (2, 3, 13, 14, 15, 21, 22,
31); MI is in 12 samples from San Matı́as Gulf (45, 48 and
52), the coastal zone, and the shallowest part of the outer
shelf (16–19, 27, 28, 42, 54, 55); TA is in eight samples from
the northernmost area that coincides with the deepest part
of the outer shelf (1, 3, 13, 14, 16, 19, 22), and in one sample
from San Matı́as Gulf (52) (Fig. 3).

TABLE 2. Morphogroup classification of benthic foraminifera: TA 5
tapered, EF 5 elongate-flattened, MI 5 milioline, PC 5 planoconvex,
RP 5 rounded-planispiral.

Genus and species Morphogroup

Ammonia beccarii PC
Angulogerina angulosa TA
Bolivina compacta EF
Bolivina lomitensis EF
Bolivina marginata EF
Bolivina ordinaria EF
Bolivina pseudoplicata EF
Buccella peruviana f. campsi PC
Bulimina gibba EF
Bulimina patagonica EF
Buliminella elegantissima EF
Cassidulina laevigata RP
Cassidulinoides parkerianus RP
Cibicides aknerianus PC
Cibicides dispars PC
Cibicides fletcheri PC
Cibicides kullenbergii PC
Cibicides mckannai PC
Cibicides sp. PC
Discorbis peruvianus PC
Elphidium discoidale RP
Elphidium sp. RP
Epistominella exigua RP
Florilus atlantica RP
Globobulimina affinis EF
Globocassidulina subglobosa RP
Lenticulina clerichii PC
Lenticulina limbosa PC
Miliolinella subrotunda MI
Nonionella auris RP
Pyrgo nasuta MI
Quinqueloculina frigida MI
Quinqueloculina patagonica MI
Quinqueloculina seminulum MI
Triloculina sp. MI
Uvigerina peregrina TA

158 ALPERIN AND OTHERS



FIGURE 2. Examples of each morphogroup. 1, 2 Planoconvex (PC): 1, Ammonia beccarii (Linné); 2; Buccella peruviana f. campsi (Boltovskoy). 3, 4
Rounded-planispiral (RP): 3, Nonionella auris (d’Orbigny); 4, N. auris (d’Orbigny). apertural view. 5, 6 Tapered (TA): 5, Angulogerina angulosa f.
angulosa (Williamson); 6, Uvigerina peregrina Cushman. 7, 8 Elongate-flattened (EF): 7, Bolivina striatula Cushman; 8, Bulimina marginata
d’Orbigny. 9, 10 Milioline (MI): 9, Quinqueloculina patagonica d’Orbigny; 10, Miliolinella subrotunda (Montagu).

FIGURE 3. Map showing morphogroup composition in stations.

MORPHOGROUPS ON THE ARGENTINE SHELF 159



Morphogroup mean composition for the samples
studied (Center) is dominated by PC (68%) and RP
(27%), followed by much lower percentages of EF (3.7%),
MI (0.6%), and TA (0.3%) (Appendix 1). The PCA results
show that the first principal component (PC1) accounts for
44%, and the second principal component (PC2) for 35%,
of the variance. The PC1 log-contrasts was 0.49 TA + 0.37
RP + 0.16 PC 5 0.67 EF + 0.36 MI, and PC2 log-contrasts
0.54 MI + 0.39 PC 5 0.58 RP + 0.45 EF (Table 4). The
most variable morphogroups are EF and TA, while the
least variable is PC (Fig. 4, Appendix 1). The most
variable ratios are EF/TA and MI/RP, followed by EF/
PC and EF/RP; the least variable are PC/TA and PC/MI
(Fig. 4). It is of particular interest to characterize the PC/
RP ratio because both morphogroups are present in most
samples and it is highly variable (variance ln(PC/RP) 5

8.55), and PC tends to be more abundant than RP (0.99),
as verified in a considerable number of samples using the
following equation:

mean ln PC=RPð Þw
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
var{ ln PC=RPð Þ

p
The samples with the highest TA/EF ratio are placed on
the positive PC1 together with those that have neither EF
nor TA, while the samples with the greatest EF/TA ratio
are placed on the negative PC1 with those that have EF
and no TA as well as those that have both but contain
higher proportions of EF. For PC2, samples with positive
PC/RP are placed on the positive PC2 and samples with
negative PC/RP are located on the negative PC2. The
biplot analysis allows for differentiating five clusters of
samples, each of which has a different morphogroup
composition. In Figure 4, there are five well-differentiated
clusters, four of which are easily discernible in the plane
shown, while the fifth (cluster 3) appears scattered from
the chosen plane.

RELATIONSHIPS BETWEEN SAMPLE CLUSTERS AND

ENVIRONMENTAL VARIABLES

The classification tree analysis showed that the most
significant variables for the morphogroup composition of
each cluster of samples were the three physicochemical
attributes of water: temperature, salinity, quantity of
dissolved oxygen with 100%, 82%, and 76% importance
respectively, and substrate, measured as ln(sand/silt) with
74% in importance. Depth, classified as having 77%
importance, is masked in this tree because the substrate’s
characteristic is negatively correlated with its depth
(r(depth; ln(sand/silt) 5 20.9194, p . 0.01). The lowest
explanatory power was attributed to distance from shore
(72%). The cross-validation procedure provided a tree with
4 partitions and 5 terminal nodes as the best fit to data (CV
cost 5 0.52; s.d. CV cost 5 0.09). The first split, based on
water temperatures, separates samples of clusters 1, 2, and 5
(T #13.1uC) from samples of clusters 3 and 4 (T .13.1uC).
The second partition, in the sample groups from colder
water, is based on substrate and separates samples of cluster
5 with finer sediment (ln(sand/silt) #2.4) from samples of
clusters 1 and 2. The third partition separates samples of
cluster 1 (salinity $33.7) from samples of cluster 2, which
has lower salinity. The samples from warmer waters
(clusters 3 and 4) were split as a function of dissolved
oxygen, with those of cluster 3 representing waters with a
higher level of oxygen (O $5.75 ml/l) than cluster 4 (Fig. 5).

It should be noted that only the terminal node labeled
cluster 1 is pure, and it comprises all of the samples in
cluster 1. In contrast, the pure terminal node labeled cluster
5 only contains those cluster 5 samples from cold and
muddy waters (47, 50, 51), leaving out four samples: one
from cold, less-saline waters (41), and three from waters
with T .13.1uC and a sandy substrate (1, 20, 49). The
terminal node labeled cluster 3 is also pure, but includes
only well-oxygenated samples (43, 44), leaving out a sample
with lower oxygen content (53). On the other hand, the
terminal node labeled cluster 2 is impure, but includes four
of the five samples in the cluster (2, 15, 21, 31) and one
cluster 5 sample (41). Finally, the node labeled cluster 4 is
the most impure, and in addition to all the samples in the

FIGURE 4. Biplot of 31 sample composition of morphogroups. TA 5
tapered, EF 5 elongate-flattened, MI 5 milioline, PC 5 planoconvex,
RP 5 rounded-planispiral.

FIGURE 5. Classification Tree model predicting sample location of
a cluster of morphogroups. Number of samples per node is shown at
left. In the node, the length of the bar indicates the number of samples
of the cluster. The variables included in this model are temperature (T),
salinity (S), dissolved O2 in water, ln(sand/silt).

160 ALPERIN AND OTHERS



cluster, it contains one cluster 2 sample (46), one cluster 3
sample (54) and three cluster 5 samples (1, 20, 49).

ASSEMBLAGES

According to Murray (2006), an assemblage is the group
of species found together in the same sample. Our study
uses morphogroup assemblages to refer to groups of
samples that have a similar composition of calcareous
benthic foraminiferal morphotypes. The absence of agglu-
tinated specimens agrees with the observations of Boltovs-
koy (1976) on the Argentine Shelf, off Tierra del Fuego
(Boltovskoy and others, 1983; Murray, 2006), and in
estuaries along the Argentine coast (Calvo Marcilese and
Protologo, 2009).

Five benthic foraminiferal morphogroup assemblages are
distinguished below according to the composition of the
clusters of samples, classification tree results, and geo-
graphical distribution of the samples. For each assemblage,
we have also determined the mean compositional mor-
phogroup composition (Fig. 6).
Assemblage I (PC-RP-TA; cluster 1 samples 3, 13, 14, 22). This

assemblage is composed only of PC (59%), RP (35%), and TA
(6.3%) (Fig. 6). The proportion of TA is the most variable, followed
by the proportion of PC. In nearly all samples there is a greater
proportion of PC than RP (mean ln(PC/RP) 5 0.51) (Fig. 4). This
assemblage corresponds to water temperature ,13.1uC, salinity
.33.7, and predominantly sandy substrate (Fig. 5, terminal node
cluster 1).

Assemblage II (RP-PC; cluster 2 samples 2, 15, 21, 31, 46). It contains
only RP (59%) and PC (41%) forms (Fig. 6). Although the PC/RP
ratio is variable among the samples in the group, it is always slightly
below the mean for the set of samples, indicating that there is a
greater proportion of RP than PC (mean ln(PC/RP) 5 20.35)
(Fig. 4). This assemblage matches to water temperature ,13.1uC,
salinity ,33.7, and predominantly sandy substrate (Fig. 5, terminal
node cluster 2).

Assemblage III (PC; cluster 3 samples 44, 45, 53). 99% are PC forms,
with very low percentages of MI and EF (0.5% and 0.4%
respectively) (Fig. 6). This assemblage corresponds to water
temperatures .13.1uC and oxygen content $5.7 (Fig. 5, terminal
node cluster 3).

Assemblage IV (PC-RP-EF-MI; cluster 4 samples 16–19, 27, 28, 42, 48,
52, 54, 55). This is the most diverse of the assemblages because it
includes the five morphogroups, PC (67.3%), RP (13.5%), EF
(13.2%), MI (5.7%), and TA (0.2%) (Fig. 6) The PC/RP ratio is
variable, although in most samples there is a higher proportion of
PC than RP (mean ln(PC/RP) 5 1.60). When ln(PC/RP).1, EF
forms appear (Fig. 4). The assemblage is present in the coastal
domain (extending the coastal domain to the 55 m isobath) and

from the shallowest part of the outer shelf (Fig. 7). This assemblage
occurs where water temperature is .13.1uC and oxygen content is
#5.7 ml/l (Fig. 5, terminal node cluster 4) Substrates are gravelly for
sample 54 and sandy-gravelly for 55, 42, 27, 28, and 18.

Assemblage V (RP-EF-PC; cluster 5 samples 1, 20, 39, 41, 47, 49–51).
Includes RP (58%), EF (27%), and PC (13%) (Fig. 6). Most samples
contain a higher proportion of RP than PC (mean ln(PC/RP) 5

21.46). RP predominates in four of the samples, PC and EF in two.
The ratios between morphogroups vary a lot (Fig. 4). The
assemblage is present in the outer shelf and in samples from San
Matı́as Gulf (Fig. 7). No similarities were found among the
environmental characteristics of the samples that would allow the
assemblage to be related to a single set of abiotic features. Samples
39, 47, 50, and 51 are found at water temperatures ,13.11uC and
salinity .33.70 (Fig. 5, terminal node cluster 5). Sample 41 also has
water temperature under 13.11uC, but salinity ,33.70 (Fig. 5,
terminal node cluster 2). Samples 1, 20, and 49 have water
temperature .13.11uC and oxygen content #5.7 ml/l (Fig. 5,
terminal node cluster 4). The substrate also varies; it is silty for
samples from San Matı́as Gulf, and sandy for samples from the
outer shelf.

DISCUSSION AND INTERPRETATION

Planoconvex (PC) and rounded-planispiral (RP) calcar-
eous foraminiferal morphogroups predominate throughout
the study area. This finding agrees with others investiga-
tions such us Nigam and others (2000) and Fernández
(2006). However, the abundance distribution of these forms
is contrary to the spatial pattern found by Severin (1983),
where the relative abundance of PC on the Texas coast
decreases with increasing depth. As mentioned previously,
the PC/RP ratio is highly variable without any determined
relation to any explanatory factor. A plausible explanation
for this behavior may be found in the morphological
symmetry of both morphogroups (similarities in apertural
view, which is biconvex to planoconvex in both groups).
Severin (1983) pointed out that the morphologies of PC and
RP may be those of generalists, as symmetric forms are
more common than asymmetic forms in undisturbed
sediment, while the reverse occurs in disturbed sediment.
Elongate-flattened (EF) forms have been found over an
extensive shelf area corresponding to rather shallow depths.
Tapered forms (TA) are found in a low proportion and are
limited geographically to deeper parts of the gulf zone and
the outermost, deepest sector of the shelf. EF and TA are
only found together in four samples, with a higher
proportion of EF. It is worth noting that in most samples,
when EF is present, TA is lacking, and vice versa.

Morphogroups PC and RP, which have no predominant
axis in their morphology, may be considered symmetrical,
while those like TA and EF, which do have a predominant
axis, are considered asymmetrical. Undisturbed sediments
(Reineck-Singh, 1980) favor the development of asymmet-
rical forms because they can maintain a preferential
position, while disturbed sediments are more likely to have
symmetrical forms, which can more easily reposition
themselves after disturbance (Severin, 1983; Nigam and
others, 2000; Fernández, 2006). The increase in asymmet-
rical forms with depth has been related to bottom-water
turbulence and consequential disturbance of the sediment
(Severin, 1983). In this context, the presence of TA forms,
limited to deeper shelf zones, suggests an environment with
little turbulence and low disturbance rate, but the presence

FIGURE 6. Mean morphogroup composition of each assemblage.

MORPHOGROUPS ON THE ARGENTINE SHELF 161



of EF forms in shallower parts of the shelf cannot be
interpreted in the same way. The MI forms are limited to
samples from the mouth of San Matı́as Gulf and the
shallowest part of the inner shelf nearest the littoral zone.
Principal Component Analysis is interpreted by analyzing
asymmetry. Asymmetric forms (EF and TA) are the
morphogroups with the greatest weight in PC1, while
symmetric forms (RP and PC) and miliolines (MI) have the
greatest weight in PC2 (Fig. 4, Table 3). Clusters 1, 3, and 4

are found on the positive PC2s with PC and MI forms. On
the positive PC1s are clusters 1 and 2, whose samples have a
clearly sandy substrate, suggesting environments with
similar energy. This differs from those of cluster 5, which
are placed on the negative PC2s and whose samples have
sandy or sandy-gravelly sandy-silty to silty substrates.

Many factors have been proposed to explain the presence
of benthic foraminifera, such as temperature, salinity,
microhabitat, food availability, and substrate type. Water
temperature, salinity, and dissolved oxygen, as well as the
type of substrate play an important part in the structure of
the morphogroup assemblages we have identified. Others
authors suggest the food availability is one of the most
important factor of the benthic foraminiferal distribution
(Jorissen and others, 1995; Van der Zwaan and others,
1999). We do not have information about food supply to
include in our analysis to consider as proxy data. It follows
that:

Assemblage I (PC-RP-TA) is dominated by the symmet-
rical morphogroups PC and PR and is the only assemblage
that contains asymmetrical TA individuals. It appears on
the outer shelf, on a sandy substrate, where Middle Shelf
Water is influenced by the Malvinas Current, which is
associated with the accumulation production of detritus
between the 80-100 m, just before the shelf break. The

FIGURE 7. Map showing distribution of sample clusters based on morphogroups.

TABLE 3. Eigen values. Eigen vectors and proportion of the
accounted variation of each principal component (PC) using the
covariance matrix derived from the centered log-ratio data. TA 5

tapered, EF 5 elongate-flattened, MI 5 milioline, PC 5 planoconvex,
RP 5 rounded-planispiral.

Morphogroup

Eigen vector

PC1 PC2 PC3 PC4

TA 0.496 0.099 0.579 0.456
EF 20.673 20.450 0.371 20.079
MI 20.363 0.544 20.449 0.410
PC 0.165 0.3892 0.064 20.785
RP 0.375 20.582 20.565 20.001

Eigen value 8.356 6.628 2.553 1.487
Absolute total variance (%) 43.92 34.84 13.42 7.81
Cumulative total variance (%) 43.92 78.76 92.18 100.00

162 ALPERIN AND OTHERS



presence of TA forms coincides with the greatest depths of
the outer shelf for the area. The preponderance these
symmetrical forms on sandy substrates, indicates that they
tolerate an environment disturbed by the Malvinas current,
which has velocities in the order of 30 cm/s.

Assemblage II (RP-PC) is composed exclusively of the
symmetrical forms PC and RP. This assemblage occurs on
the outer shelf where Middle Shelf Water is above a sandy
substrate. The exclusive presence of symmetrical mor-
phogroups suggests that these forms are able to inhabit a
high-energy environment because they can adapt to
disturbance.

Assemblage III (PC) comprises mainly symmetrical
morphogroup PC and a few milioline and asymmetrical
TA forms. The preponderance of PC forms, the presence of
MI forms, which tolerate high energy environments
(Boltovskoy, 1966; Gómez and others, 2005; Cusminsky
and others, 2009), plus a coarse-grained substrate suggest
the fauna tolerates intensive deposit-erosion processes
(Scasso and Piola, 1988) resulting from the strong currents
(,3.5 m/s) generated by the tidal front (Tonini and others,
2006).

Assemblage IV (PC-RP-EF-MI) is composed of has a
large proportion of symmetrical morphogroups PC and
RP, a smaller proportion of asymmetrical EF, and few TA
and miliolines. The assemblage present in the San Matı́as
Gulf is linked to the saline gulf waters. The assemblages
located in the shallower part of the outer shelf could be
related to the Coastal Water with high salinity, and
associated with the production of detritus generated at
the Coastal Front zone between the 40–50 m isobaths. The
predominance of symmetrical forms and the presence of
miliolines on sand to gravelly sand suggest that this
association prospers in a high-energy environment.

Assemblage V (RP-TA-PC) comprises mainly of sym-
metrical RP, subordinate PC forms, and asymmetrical EF
forms. This assemblage occurs in the deep parts of San
Matı́as Gulf and on the outer shelf. It is found in a wide
range of water temperature and salinity, on silty and sandy
substrates, but does not appear to be linked in any
definitive way to any of these variables. In the deep parts
of the Gulf, the assemblage inhabits silty substrates and its
generalist character would allow the specimens to behave
like as though they were asymmetric, suggesting either a
eutrophic environment most likely caused by a deficit in
current circulation or waters with restricted circulation
typical of deep, engulfed zones (Bernasconi and Cusminsky,

2005; Bernasconi and others, 2009). On the shelf, the
assemblage develops on sandy sediments where there is an
EF morphogroup, and its patchy spatial pattern may be
associated with local factors.

Alperin and others (2008) studied the same material with
statistical analysis of compositional foraminiferal species
census data, and defined six assemblages of benthic
foraminiferal species whose characteristics are believed to
be responses to oceanographic conditions at the site. This
would show that the inferences based on the autoecology of
the species are similar to the ones in this work that are
based on shell morphology (Table 4). On the other hand,
several workers, including Corliss and Chen (1988), Jorissen
and others (1995), and Van der Zwaan and others (1999),
have found food supply to be a major driver of benthic
foraminifera assemblage composition and distribution.
Food availability is a complex factor with different
components (e.g., type, quality, and quantity) that enable
different morphogroups to coexist. The coastal and shelf-
break fronts, which are areas of high primary productivity
favorable to the development of benthic foraminifera,
would be the main sources of food supply. Food could
still be another driver in this case but there was no proxy
information on food supply needed to include it in these
analyses.

CONCLUSIONS

The distribution pattern of benthic foraminiferal mor-
phogroups on the Argentine continental shelf between 39–
43uS is the result of complex interactions with different
environmental parameters. Temperature appears to be the
most important factor in structuring the assemblages,
although salinity, substrate grain size, and environmental
energy also play important roles. These factors, especially
temperature and salinity, could be related to different water
masses.

The geographical distribution pattern of the benthic
foraminiferal assemblages defined by autoecological char-
acteristics is maintained when the analysis uses morpho-
logical criteria. Both species assemblages and foraminiferal
morphogroups allow inferences to be made about water
mass such as temperature and salinity, substrate character-
istics, and environmental energy.

Our study is the first attempt to use taxon-free criteria of
benthic foraminifera on the Argentine continental shelf.

TABLE 4. Comparison between morphogroups (this paper) and species assemblages according to Alperin and others (2008) (sample number
in brackets).

This paper Alperin and others. 2008 Oceanographic conditions

Assemblage I (3, 13, 14, 22) Group 1 (3, 13, 14, 21, 22, 39) Outer shelf; cooler waters strong influence of Malvinas
Current

Assemblage II (2, 15, 21, 31, 46) Group 2 (2, 15, 17, 20, 31, 41) Transition between outer and inner shelf; little influence of
Malvinas Current

Assemblage III (44, 45, 53) Group 5 (44, 45, 53) Coastal water with higher salinity and temperature; high
sedimentary dynamics

Assemblage IV (16, 17, 18, 19, 27, 28, 42,
48, 52, 54, 55)

Groups: 4 (18, 19, 28, 42, 52, 54, 55);
2 (17); 3 (16); 6 (48)

Influence of middle-shelf water and coastal water of high
salinity

Assemblage V (1, 20, 39, 41, 47, 49, 50, 51) Groups: 1 (39); 2(20, 41); 3 (1, 49);
6 (47, 50, 51)

Wide range of temperature salinity and environment energy

MORPHOGROUPS ON THE ARGENTINE SHELF 163



Foraminiferal morphogroup analysis shows potential for
determining recent environments and paleoenvironments
when species autoecology is unknown or when there is
uncertain taxonomy. Further study of the spatial patterns
and factors that structure modern benthic foraminiferal
morphogroup assemblages in different environments and
zones of the Argentine continental shelf will allow them to
be integrated with taxonomically based studies in order to
achieve more precise ecologic and paleoecological interpre-
tations.

ACKNOWLEDGMENTS

This study was partially funded by the National Agency
for the Promotion of Science and Technology (ANPCyT),
PICT 07-14653, 26057 and by the National Research
Council (CONICET), PIPS 6416 and 00819. This work is
also a contribution to the Project UNComahue (B940,
B001). We are grateful to C. Cotaro for preparing the
samples of scanning electron microscope (SEM) from
Centro Atómico Bariloche. This paper benefited substan-
tially from comments and suggestions given by Drs.
Kenneth Severin, Martin Buzas, Bruce Hayward, Kenneth
L. Finger, and an anonymous referee.

REFERENCES

ACHILLI, S. M., and ALIOTTA, S., 1992, Caracterı́sticas morfológicas de
ondas de arena de la plataforma continental al sur de la provincia
de Buenos Aires, Argentina. Proceedings of the 3ras Jornadas
Geológicas Bonaerenses (La Plata, Buenos Aires, Argentina),
p. 207–212.

AITCHISON, J., 1986, The statistical analysis of compositional data:
Monographs on Statistics and applied Probability: Chapman and
Hall Ltd., London, 416 p.

———, 1997, The one-hour course in compositional data analysis or
compositional data analysis is simple, in Pawlowsky-Glahn, V.
(ed.), Proceedings of IAMG’97: The 3u Annual Conference of
International Associations for Mathematical Geology Interna-
tional Center for Numerical Methods in Engineering (CIMNE),
Barcelona, v. 1, p. 3–35.

———, 1999, Logratios and natural laws in compositional data
analysis: Mathematical Geology, v. 131, p. 563–580.

———, and GREENACRE, M., 2002, Biplots of compositional data:
Applied Statistics, v. 51, p. 375–392.

———, BARCELÓ-VIDAL, C., MARTı́N-FERNÁNDEZ, J. A., and PAW-

LOWSKY-GLAHN, V., 2000, Logratio analysis and compositional
distance: Mathematical Geology, v. 32, p. 563–580.

ALIOTTA, S., 1983, Estudio sedimentológico y de deriva litoral entre la
desembocadura del Rı́o Negro y Playa Bonita (Pcia. de Rı́o
Negro): Universidad Nacional del Sur Instituto Argentino de
Oceanografı́a, Graduate Degree Thesis, 58 p.

ALPERIN, M., BERNASCONI, E., and CUSMINSKY, G., 2008, Asocia-
ciones de foraminı́feros bentónicos recientes de la plataforma
continental argentina (39u–43uS y 58u–65uO) analizados con
métodos estadı́sticos para datos composicionales: Ameghiniana,
v. 45, p. 443–461.

BANDY, O. L., 1960, General correlation of foraminiferal structure
with environment. International Geological Congress, 21, Copen-
hagen, v. 22, p. 7–19.

BERNASCONI, E., 2006, Los foraminı́feros del Holoceno de Testigos de
la Plataforma Continental Argentina (40u 309–42u 489 LS y 59u
259–64u 409): Ph.D. Dissertation, Universidad Nacional del
Comahue, Bariloche, 210 p.

———, and CUSMINSKY, G., 2005, Distribución de Nonionella auris
(d’Orbigny) Orden Foraminiferida en el Golfo San Matı́as,
Provincia de Rı́o Negro, Argentina: Ameghiniana, v. 42,
p. 167–174.

———, and ——— , 2007, Estudio de los foraminı́feros bentónicos de
un testigo holoceno proveniente de la cuenca del Colorado:
Ameghiniana, v. 44, p. 271–278.

———, and ———, 2009, Estudio paleoecológico de Foraminı́feros de
testigos del Holoceno de Golfo Nuevo (Patagonia, Argentina):
Geobios, v. 42, p. 435–450.

———, ——— , and GÓMEZ, E., 2009, Foraminı́feros bentónicos del
holoceno del golfo Nuevo, Argentina: inferencias paleoclimáticas:
Revista Española de Micropaleontologı́a, v. 41, p. 21–34.

BERNHARD, J. M., 1986, Characteristic assemblages and morphologies
of benthic foraminifera from anoxic, organic-rich deposits:
Jurassic through Holocene: Journal of Foraminiferal Research,
v. 16, p. 207–215.

BOLTOVSKOY, E., 1966, Los Foraminı́feros Recientes, EUDEBA (ed.),
Buenos Aires, Argentina, 510 p.

———, 1976, Distribution of Recent Foraminifera of the South
America Region: Foraminifera, v. 2, p. 171–236.

———, GIUSSANI DE KAHN, G., and WATANABE, S., 1983, Variaciones
estacionales y standing crop de los foraminı́feros bentónicos de
Ushuaia, Tierra del Fuego: Physis, Sección A, v. 41, p. 113–127.

BREIMAN, J. H., FRIEDMAN, J. H., OLSHEN, R. A., and STONE, C. J.,
1984, Classification and Regression Trees: Wadsworth & Brooks/
Cole Advanced Books & Software, Monterey, California.

CALVO MARCILESE, L., and PRATOLONGO, P., 2009, Foraminı́feros de
marismas y llanuras de marea del estuario de Bahı́a Blanca,
Argentina: distribución e implicaciones ambientales: Revista
Española de Micropaleontologı́a, v. 41, p. 315–332.

CARRETO, J. I., LUTZ, V. A., CARIGNAN, M. O., CUCCHI COLLEONI, A.
D., and DE MARCO, S. G., 1995, Hydrography and chlorophyll a
in a transect from the coast to the shelf-break in the Argentinian
Sea: Continental Shelf Research, v. 15, p. 315–336.

CORLISS, B. H., 1985, Microhabitats of benthic foraminifera within
deep-sea sediments: Nature, v. 314, p. 435–438.

———, and CHEN, C., 1988, Morphotype patterns of Norwegian Sea
deep-sea benthic foraminifera and ecological implications: Geol-
ogy, v. 16, p. 716–719.

CUSMINSKY, G., BERNASCONI, E., and CALVO-MARCILESE, L., 2009,
Holocene benthic foraminifera from Bahı́a Blanca estuary: a
review and update of systematic and palaeoenvironmental aspects:
The Holocene, v. 19, p. 1221–1231.

FERNÁNDES, J., 2006, Evolución paleoambiental de la formación
Cubagua (Penı́nsula de Araya) a partir de patrones de distribución
de morfogrupos de foraminı́feros bénticos: Revista de la Facultad
de Ingenierı́a de la U.C.V., v. 21, p. 23–34.

FERRERO, L., 2006, Micropaleontologı́a y Paleoecologı́a del Cuater-
nario del sudeste de la provincia de Buenos Aires: Ph.D. Thesis,
Universidad Nacional de Mar del Plata, 371 p.

GAGLIARDINI, D. A., and RIVAS, A. L., 2004, Environmental
characteristics of San Matı́as Gulf obtained from LANDSAT-
TM and ETM+ DATA: Gayana (Concepción), v. 68(2) supl, T I,
Proc Concepción 2004, p. 186–193. ISSN 0717-6538.

GELOS, E., SPAGNUOLO, J., and LIZASOAIN, G., 1988, Mineralogı́a y
caracterización granulométrica de sedimentos actuales de la
plataforma argentina entre los paralelos 39u y 43u de latitud Sur
y del golfo San Matı́as: Revista de la Asociación Geológica
Argentina, v. 63, p. 63–79.

GINI, C. W., 1912, Variabilita e mutabilita, Studi Economico-Giuridici
della R. Universita di Cagliari,, v. 3, p. 3–159.

GÓMEZ, E., MARTı́NEZ, D., BOREL, M., GUERSTEIN, G. R., and
CUSMINSKY, G. C., 2005, Submarine evidences of Holocene sea-
level fluctuations in the Bahı́a Blanca estuary, Argentina: Journal
of South America Earth Sciences, v. 20, p. 139–155.

GUERRERO, R. A., 1998, Oceanografı́a fı́sica del estuario del Rı́o de la
Plata y el sistema costero de El Rincón, Noviembre, 1994:
INIDEP Informe Técnico 21, p. 29–54.

———, and PIOLA, A. R., 1997, Masas de agua en la Plataforma
Continental: In: El Mar Argentino y sus recursos pesqueros,
Contribución INIDEP 998, v. 1, p. 107–118.

HAYWARD, B. W., GRENFELL, H., CAIRNS, G., and SMITH, C., 1996,
Environmental controls on benthic foraminiferal and thecamoe-
bian associations in a New Zealand tidal inlet: Journal of
Foraminiferal Research, v. 26, p. 150–171.

———, NEIL, H., CARTER, R., GRENFELL, H., and HAYWARD, J., 2002,
Factors influencing the distribution patterns of Recent deep-sea

164 ALPERIN AND OTHERS



benthic foraminifera, east of New Zealand, Southwest Pacific
Ocean: Marine Micropaleontology, v. 46, p. 139–176.

JORISSEN, F. J., STIGTER, H. C., and WIDMARK, J. G. W., 1995, A
conceptual model explaining benthic foraminiferal microhabitats:
Marine Micropaleontology, v. 26, p. 3–15.

KAIHO, K., 1991, Global changes of Paleogene aerobic/anaerobic
benthic foraminifera and deep-sea circulation: Palaeogeography,
Palaeoclimatology, Palaeoecololgy, v. 83, p. 65–85.

———, 1994, Benthic foraminiferal dissolved-oxygen index and
dissolved-oxygen levels in the modern ocean: Geology, v. 22,
p. 719–722.

———, 1999, Effect of organic carbon flux and dissolved oxygen on
the benthic foraminiferal oxygen index (BFOI): Marine Micro-
palaeontology, v. 37, p. 67–76.

KUCERA, M., and MALMGREN, B. A., 1998, Logratio transformation of
compositional data: a resolution of the constant sum constraint:
Marine Micropaleontology, v. 34, p. 117–120.

LAPRIDA, C., GARCÍA-CHAPORI, N., VIOLANTE, R., and COMPAGNUCCI,
R., 2007, Mid-Holocene evolution and paleoenvironments of the
shoreface-offshore transition, north-eastern Argentina: new evi-
dence based on benthic microfauna: Marine Geology, v. 240,
p. 43–56.

LUCAS, A. J., GUERRERO, R., MIANZAN, H., ACHA, E. M., and LASTA,
C., 2005, Coastal oceanographic regimes of the Northern
Argentine Continental Shelf (34–43uS): Estuarine, Coastal and
Shelf Science, v. 65, p. 405–420.

MARTÍN-FERNÁNDEZ, J. A., BARCELÓ-VIDAL, C., and PAWLOWSKY-
GLAHN, V., 2003, Dealing with zeros and missing values in
compositional data sets using nonparametric imputation: Math-
ematical Geology, v. 35, p. 253–277.

MARTOS, P., and Pı́CCOLO, M. C., 1988, Hydrography of the Argentine
continental shelf between 38u and 42u S: Continental Shelf
Research, v. 8, p. 1043–1056.

MURRAY, W., 1973, Distribution and Ecology of Living Benthic
Foraminiferids: Heinemann Educational Books, London, 271 p.

———, 1991, Ecology and Paleoecology of Benthic Foraminifera:
Longman, Wiley, Harlow/Essex, New York, 397 p.

———, 2006, Ecology and Applications of Benthic Foraminifera:
Cambridge University Press, Cambridge, UK, 426 p.

NIGAM, R., KHARE, N., and MAYENKAR, D. N., 2000, Can bathymetry
be a discriminatory factor for the distribution of benthic
foraminiferal morpho-groups in modern marine sediments?:
ONGC Bulletin 37, v. 1, p. 47–51.

PARKER, G., VIOLANTE, R. A., and PATERLINI, M. C., 1996, Fisiografı́a
de la plataforma continental: in Ramos, V., and Turic, M. (eds.),

Geologı́a y Recursos Naturales de la Plataforma Continental
Argentina: Asociación Geológica Argentina, Buenos Aires,
Argentina, p. 1–16.

PAWLOWSKY-GLAHN, V., and EGOZCUE, J. J., 2001, Geometric
approach to statistical analysis on the simplex: Stochastic
Environmental Research and Risk Assessment, v. 15, p. 384–398.

———, and ———, 2002, BLU estimators and compositional data:
Mathematical Geology, v. 34, p. 259–274.

PIOLA, A. R., and SCASSO, L. M. L., 1988, Circulación del Golfo San
Matı́as: Geoacta v. 15, v. 1, p. 33–51.

PUCCI, E., ASTEASUAIN, R., VILLA, N., LARA, R., RUSANSKY, C.,
GARCÍA, O., MARCOVECCHIO, J., ASTEASUAIN, A., and CEJAS, J.,
1985, Nutrientes, factores quı́micos y fı́sicos relacionados y
perfiles en aguas de la plataforma continental argentina: Informe
técnico IADO 11, Bahı́a Blanca, 179 p.

REINECK, H. E., and SINGH, I. B., 1980, Depositional Sedimentary
Enviornments. Second, Revised and Updated Edition: Springer-
Verlag, Berlin; Heidelberg, New York, 542 p.

REOLID, M., RODRÍGUEZ-TOVAR, F., NAGY, J., and OLÓRIZ, F., 2008,
Benthic foraminiferal morphogroups of mid to outer shelf
environments of the Late Jurassic (Prebtic Zone, southern Spain):
characterization of biofacies and environmental significance:
Palaeogeography, Palaeoclimatology, Palaeoecology, v. 261,
p. 280–299.

SCASSO, L. M. L., and PIOLA, A. R., 1988, Intercambio neto de agua
entre el mar y la atmósfera en el Golfo San Matı́as: Geoacta, v. 15,
p. 13–31.

SEVERIN, K., 1983, Test morphology of benthic foraminifera as a
discriminator of biofacies: Marine Micropaleontology, v. 8,
p. 65–76.

TONINI, M., PALMA, E., and RIVAS, A., 2006, Modelo de alta
resolución de los Golfos Patagónicos: Mecánica Computacional,
v. 25, p. 1441–1460.

VAN DER ZWAAN, G. J., DUIJNSTEE, L. A. P., DEN DULK, M., ERNST, S.
R., JANNINK, N. T., and KOUWENHOVEN, T. J., 1999, Benthic
foraminifers: proxies or problems? A review of paleoecological
concepts: Earth-Science Reviews, v. 46, p. 213–236.

VON EYNATTEN, H., PAWLOWSKY-GLAHN, V., and EGOZCUE, J. J.,
2002, Understanding perturbation on the simplex: a simple
method to better visualizes and interpret compositional data in
ternary diagrams: Mathematical Geology, v. 34, p. 249–257.

Received 20 November 2009
Accepted 5 January 2011

MORPHOGROUPS ON THE ARGENTINE SHELF 165



APPENDIX 1. CompositionAL data of morphogroups. TA 5 tapered, EF 5 elongate-flattened, MI 5 milioline, PC 5 planoconvex, RP 5
rounded-planispiral. Var-clr 5 variance of centered log-ratio data.

Sample
Number of
specimens Cluster

Morphogroup

TA EF MI PC RP

1 58 5 0.017 0.397 0.000 0.310 0.276
2 22 2 0.000 0.000 0.000 0.136 0.864
3 78 1 0.090 0.000 0.000 0.603 0.308

13 462 1 0.320 0.000 0.000 0.357 0.323
14 179 1 0.011 0.000 0.000 0.642 0.346
15 12 2 0.000 0.000 0.000 0.333 0.667
16 45 4 0.022 0.267 0.044 0.556 0.111
17 264 4 0.000 0.227 0.030 0.125 0.617
18 224 4 0.000 0.045 0.134 0.786 0.036
19 65 4 0.015 0.185 0.031 0.538 0.231
20 48 5 0.000 0.125 0.000 0.146 0.729
21 86 2 0.000 0.000 0.000 0.442 0.558
22 105 1 0.038 0.000 0.000 0.638 0.324
27 27 4 0.000 0.037 0.111 0.630 0.222
28 46 4 0.000 0.130 0.130 0.696 0.043
31 31 2 0.000 0.000 0.000 0.806 0.193
39 27 5 0.000 0.037 0.000 0.741 0.222
41 66 5 0.000 0.394 0.000 0.136 0.470
42 372 4 0.000 0.054 0.013 0.892 0.040
44 19 3 0.000 0.158 0.000 0.842 0.000
45 11 3 0.000 0.000 0.273 0.727 0.000
46 10 2 0.000 0.000 0.000 0.400 0.600
47 133 5 0.000 0.549 0.000 0.045 0.406
48 244 4 0.000 0.270 0.020 0.275 0.434
49 15 5 0.000 0.400 0.000 0.467 0.133
50 900 5 0.000 0.082 0.000 0.031 0.887
51 419 5 0.000 0.143 0.000 0.002 0.854
52 250 4 0.016 0.112 0.084 0.696 0.092
53 300 3 0.000 0.000 0.000 1.000 0.000
54 134 4 0.000 0.119 0.052 0.784 0.045
55 38 4 0.000 0.053 0.026 0.842 0.079

Samples 31 8 20 12 31 28
Specimens 4690 168 505 93 1916 2008
Geometric mean 0.003 0.038 0.006 0.682 0.271
Var-clr 3.29 5.49 3.84 2.16 4.24

166 ALPERIN AND OTHERS