Spatial Distribution of Benthic Foraminifera in the Neretva Channel (Croatia Coast): Faunal Response to Environmental Parameters

Abstract

Benthic foraminiferal assemblages have been studied at 11 sediment surface samples located in the Neretva Channel covering the delta habitat and the adjacent open sea areas. The major objective of the investigation was to explore the main environmental parameters affecting the benthic foraminifera compositional changes. To this end, a statistical approach was applied that integrates micropaleontological data with physical, geochemical and sedimentological parameters (total organic carbon and grain size). Statistical analyses identified four distinct groups (cluster A1, A2, B1, B2) corresponding to different environmental settings. Cluster A1 groups samples under Neretva river influence and is characterized by Aubignyna perlucida, Nonionella turgida, Eggerelloides scaber and Rectuvigerina sp.; species able to live in organic-matter-rich sediments and in a wide range of oxygen content. Cluster A2 includes samples distant from the fluvial outlet and samples along the NW coast partially influenced by the Neretva river plume. In these environmental conditions, Ammonia beccarii, Bulimina marginata, Nonionella turgida and Textularia sp. resulted as the most characteristic taxa. Cluster B1 distinguishes the deepest stations which are in connection with the open Adriatic Sea. Here Asterigerinata mamilla, Buccella granulata, Cibicides group, Reussella spinulosa and Textularia sp. reach their maximum abundance associated with coarse-grained sediments. Cluster B2 groups samples collected in the inner bay of the southernmost sector of the studied area characterized by silt and clay and a negligible influence by river inputs. The benthic microfauna is principally composed of Miliolids, Porosononion granosum and Textularia sp.

1. Introduction

Benthic foraminifera (unicellular protists) have been widely used to reconstruct paleoenvironmental changes since they are very sensitive to ambient conditions including organic matter fluxes, oxygen content, salinity, type of substrate and others [1,2,3,4,5,6]. Moreover, foraminiferal assemblages are influenced by their immediate environment and have often served as modern and past analogues to characterize paleoenvironments [7,8,9]. Therefore, documenting and understanding the regional distribution and ecological response of foraminifera is of pivotal importance. The majority of surveys on the current distribution of benthic foraminifera have been carried out in the western part of the Adriatic Sea [10,11,12,13,14,15,16,17,18,19], whereas less is known about their presence on the eastern coast [20,21,22,23,24,25,26]. Here we report for the first time the distribution patterns of benthic foraminifera in the Neretva Channel, a semi-enclosed basin located along the southernmost part of the Croatian coast. The Neretva river is the fifth largest Mediterranean river, ranked by annual water discharge [27]. The river mouth and its adjacent semi-enclosed marine environment provide a good example of land–sea interactions, particularly between sediment type and organic matter content in a microtidal, low-wave energy, and river-dominated coastal sedimentary system [28]. While the sediment dynamics and geochemistry were examined [28,29,30], no investigation of the microfauna was performed. The aims of this study are: (1) to examine the quantitative distribution of the benthic foraminifera taxa in the Neretva Channel and (2) to identify the main factors affecting the spatial compositional changes of the foraminiferal assemblages. This kind of study adds information on benthic foraminiferal ecological preferences useful for a better reconstruction of past environmental conditions.

2. Study Area

The Neretva Channel is bounded to the NE by the Croatian coastline, to the NW by the open Adriatic Sea through the Korcula Channel, and to the SW by the Peljesac Peninsula (Figure 1). The hydrological setting is influenced by the intrusion of the highly saline oligotrophic waters from the open Adriatic Sea [31] and by the freshwater discharge from the Neretva River [28,32,33]. The Neretva is the largest river on the Croatian part of the eastern Adriatic coast, and the only one forming a deltaic system [34]. The Neretva Channel is a microtidal low-wave energy environment characterized by river-dominated sedimentation processes [30,35].

The surface sediment is composed mainly of clayey silt, in particular in the area in front of the river delta and in the northeast part of the channel. Sedimentation is uniform in most of the area; this is because the river creates a hypopycnal plume at the mouth that distributes fine-grained particles over the entire channel [23,29]. However, two coarser-grains depositional zones with different natures are observed [28]. In detail, in the north-western part of the channel, the sediment is composed mainly of sand, probably related to the influence of higher energy currents from the Adriatic Sea that remove the finer part of the sediments [28]. The southernmost part of the southeastern sector of the channel is instead characterized by relict sandy-silt sediment as this zone only receives river sediment sporadically during events of strong hydro-dynamism in the channel [28,30,36].

The mineral composition of sediments is quartz, calcite and clay, with a secondary presence of plagioclase and dolomite [30]. Marine sediment compositions are mainly controlled by the local geology, but can also reflect anthropogenic discharges from industrial and urban activities, thus increasing concentrations of trace metals and organic matter [37]. Most of the mainland coast is built of highly permeable limestone without a natural barrier between the groundwater of the karstic fields in the hinterland and the freshwater submarine springs [38]. Higher discharge of the submarine springs in winter causes a spreading of the less saline surface waters into the Neretva Channel and a compensatory inflow in the bottom layer. The strong NE “Bora” wind can, however, reverse that circulation. During the summer, the circulation pattern depends entirely on the wind direction [39].

3. Materials and Methods

Sediment samples were collected in 2006 during the NERES06 cruise on board R/V Bios DVA along three parallel NW–SE transects [28] (Figure 1;

Table 1. List of the 11 sites considered in this study with geographical coordinates and water depth.

Site	Latitude	Longitude	Water Depth (m)
1 BC	43°00.702′	17°25.080′	24
2 BC	43°00.017′	17°23.998′	29
6 BC	43°05.946′	17°14.193′	47
7 G	43°04.629′	17°11.948′	51
8 G	43°07.120′	17°16.524′	46
9 BC	43°04.878′	17°19.591′	39
11 G	43°02.939′	17°21.937′	32
12 G	43°01.496′	17°22.343′	34
14 BC	43°08.256′	17°22.343′	29
15 G	42°58.147′	17°27.579′	24
16 G	42°56.645′	17°29.626′	26

) in the frame of the Project ADRICOSM_NERES: ADRIatic sea integrated COastal areaS and river basin Management system—Environmental Regeneration and Sustainable Development of the delta of the NEretva river.

Here we report micropaleontological results of benthic foraminifera from the upper 0–2 cm surface sediment layer collected in 11 sites using Van Veen grab (G) and box-corer (BC) samplers (Table 1). In detail, small cores (about 30 cm long and 8 cm in diameter) were collected by subsampling the central part of the grab samples, which was assumed as relatively undisturbed.

3.1. Environmental Parameters

The environmental parameters used in this study are from the dataset provided and described in [28]. Here we consider sedimentological (percentage of clay/silt/sand) and geochemical (TOC, C/N) data performed on the same samples analysed for micropaleontological investigation (surface levels 0–2 cm).

3.2. Benthic Foraminiferal Analysis

In this study, we considered the total benthic foraminiferal assemblages (living and dead specimens). The total assemblage represents a more reliable picture of the population as it integrates many multiple generations (seasonal variation) of foraminifera [40,41]. Consequently, the considered foraminiferal content (living and dead specimens) corresponds to average environmental conditions throughout the deposition of the collected sediment layer (0–2 cm).

For foraminiferal investigation, samples were washed over a 63 μm sieve then oven-dried at 50 °C and weighed. Identification and counting analysis were performed under a Leica MZ 8 stereomicroscope on the sediment fraction > 125μm. We considered all specimens present in the whole sample or in representative aliquots using an Otto microsplitter, containing at least 300 specimens. Taxonomy identification of foraminifera was based mainly on systematic studies from Adriatic and extra-Mediterranean areas [42,43,44]. Online World Modern Foraminifera Database [45,46] was also consulted.

Less abundant species with similar ecological characteristics were merged together into major species groups. Adelosina mediterranensis and Adelosina sp. were merged into the Adeloisina group. Both species are indicative of shallow marine environments (e.g., [47]). Other porcelaneous taxa were grouped in Miliolids (Miliolinella, Pyrgo, Quinqueloculina, Spiroloculina and Triloculina genera).

Although phylogenetically distinct species, Lobatula lobatula (i.e., Cibicies lobatulus, [48]), and Cibicides refulgens are often grouped together in environmental studies due to their notable morphological similarity and intermediate forms between the two species [49]. They are both typical taxa of the infralittoral environment with strong near-bottom currents [50,51,52,53]. Here they are referred to as the Cibicides group.

Identified Elphidium specimens were divided into two groups based on different life strategies (having a keel or not, being infaunal or epifaunal, according to [24,26,54,55]). Keeled elphidiids species are here represented by Elphidium crispum and Elphidium macellum. They are known as typical epiphyte shallow marine taxa [54,56,57]. Unkeeled elphidiids species in our data set include only Porosononion granosum, an infaunal organism, abundant in marginal-marine environments that tolerates wide salinity changes (euryhaline) [54,58].

The data are reported in the percentage of specimens on the total assemblage (relative abundance %) and the number of specimens per gram of dry bulk sediment for each of the examined samples (BFN) (Supplementary Table S1). Foraminiferal biodiversity was estimated using different diversity indices: Shannon (H) (that varies from 0 for communities with only a single taxon to high values for communities with many taxa, each with comparable abundance); Dominance (D) evidences the occurrence of the opportunistic taxa (it ranges from 0, all taxa are equally present, to 1 one taxon dominates the community completely), and Simpson’s Index of Diversity (1-D) (the value of this index also ranges between 0 and 1, the greater the value, the greater the sample diversity. The index represents the probability that two individuals randomly selected from a sample will belong to different species).

All indices were calculated using the PAlaeontological STatistics (PAST) package [59,60].

3.3. Statistical Analyses

Statistical analyses include all taxa with relative abundance > 3% in at least one sample (total 23 taxa/groups) in order to reduce background noise due to the infrequent taxa.

As physical and chemical variables TOC, grain size and water depth were considered.

The foraminiferal assemblage was analysed by means of a cluster analysis based on Ward’s criterion [61]. In order to investigate patterns of covariance between the foraminiferal data and the chemical and physical variables, we performed a two-block partial least squares analysis (2B-PLS). This statistical technique is suitable for studying matrices with a relatively low sample size and highly correlated variables [62]. Recently it was successfully applied to examine relationships between benthic foraminiferal communities and seafloor morphology in the southern Mediterranean area [63]. The two-block PLS method differs from other established multivariate statistics (e.g., Canonical Correlation Analysis) in that it aims to find latent variables that can explain the covariance between two multivariate matrices, returning variables that explain the covariance between two sets of variables as much as possible [64]. All analyses were performed in R 4.2.0 (R Core Team 2022- Foundation for Statistical Computing, Vienna, Austria)

4. Results

4.1. Environmental Data

Sediment grain size analyses indicated that the Neretva Channel deposits are generally composed of clayey silts with a silty dominant fraction and a low fraction of sand (Figure 2). In particular, the percentage of clay is overall high in the entire channel, with a minimum of 52.30% in site 7 G and a maximum of 69.90% in site 8 G. In the upper 2 cm surface sediment, the coarser materials were found in the southernmost zone (sand 31% in site 15 G) and in the north-western sector of the channel (sand 51% and 39% at sites 7 G and 6 BC, respectively; Figure 3; [28]). As a consequence, the silt fraction was the lowest in the 7 G and 6 BC sites (3.02% and 4.24%, respectively).

Surficial Total Organic Carbon (TOC) (Figure 2) varies from 0.6% (6 BC) to 1.06% (1 BC). The maximum values of TOC are detected in samples in the sites close to the river mouth (1 BC and 2 BC) (Figure 2). The C/N ratios range from 6.81% (16 G) to 9.6 (1 BC) (Figure 2; [28]). Samples closer to the river mouth display significantly higher C/N ratios (Figure 2) indicative of a terrestrial signature due to the fluvial input [28].

4.2. Benthic Foraminifera

A total of 50 taxa were identified in the surface samples, including 4 groups (Adelosina, Cibicides, keeled and unkeeled elphidiis) and generic Miliolids (Table S1). The recorded individuals were predominantly calcareous forms (mainly Ammonia beccarii, Bulimina marginata, Bulimina elongata, Porosononion granosum, Nonionella turgida and Valvulineria bradyana) and agglutinated forms (e.g., Eggerelloides scaber, Lagenammina difflugiformis and Textularia sp.).

Regarding microhabitat, the most predominant were infaunal species, [4,14,65,66,67] such as Ammonia beccarii, Bulimina basispinosa, Bulimina elongata, Bulimina marginata, Nonionella turgida, Reussella spinulosa, Eggerelloides scaber and Textularia sp. Less abundant species were epifaunal taxa belonging to the keeled elphidiids and Quinqueloculina genera [47,54]. Miliolids ranged largely in the studied samples, from 4.55% (7 G) to 15.85% (16 G). The relative abundance of the common species (>3% in at least one sample) were listed in

Table 2. Relative percentages of the most common species (>3% in at least one sample) identified in the surface sediment layer (0–2 cm).

Sample	1 BC	2 BC	6 BC	7 G	8 G	9 BC	11 G	12 G	14 BC	15 G	16 G
Adelosina group	0.91	1.15	1.42	3.24	0.95	1.32	0.57	0.81	0.89	0.55	3.08
Ammonia beccarii	4.97	5.95	2.84	2.23	6.64	5.38	3.41	9.80	8.93	3.02	2.20
Aubignyna perlucida	3.66	5.18			0.19	0.45	2.70	1.84	3.87	1.78	1.76
Asterigerinata mamilla		0.19	1.83	3.04	1.33	0.45	0.43	1.43		2.47	0.44
Buccella granulata	0.39	0.77	4.67	4.86		0.34				2.61
Bulimina basispinosa	2.09	1.34	2.84	2.23	4.17	3.81	6.83	3.06	1.79	1.65	1.76
Bulimina elongata	7.85	2.88	1.42	1.72	5.12	3.36	5.69	4.08	3.27	3.57	6.17
Bulimina marginata	2.36	1.20	2.84	2.53	10.25	6.05	3.70	6.33	4.17	0.55	4.41
Cibicides group	0.65	1.54	4.67	7.09	1.90	2.58	1.28	2.04	3.57	1.92	2.20
Hanzawaia boueana	0.13	0.19	1.42	3.75	0.19	0.34				2.61
Keeled elphidiids species	1.04	1.15	1.42	2.23	2.09	3.25	2.84	1.84	2.38	0.55	3.96
Nonionella turgida	7.85	4.03	1.62	0.71	3.80	9.19	6.83	6.12	7.44	4.66	1.76
Miliolids	4.70	6.13	5.08	4.55	7.03	10.21	9.66	7.55	8.64	12.07	15.85
Rectuvigerina sp.	8.77	6.91	1.62	0.71	2.85	3.59	3.41	2.65	3.57	4.12	2.64
Reussella spinulosa	0.65	3.65	14.40	11.65	4.93	3.36	3.56	1.63	3.27	5.76	6.17
Rosalina sp.	0.92	3.84	1.42	4.36	3.61	1.46	1.42	1.84	2.08	5.08	4.41
Unkeeled elphidiids species	9.55	8.06	17.04	15.91	12.52	11.66	11.1	12.04	10.42	20.58	13.22
Valvulineria bradyana	7.33	8.06			2.09	3.59	7.40	7.76	5.65	1.92	1.76
Adercotryma glomeratum	0.39	3.45	0.20		0.19	0.56			1.19
Eggerelloides scaber	8.64	3.07	0.20	0.10	2.66	3.25	2.99	2.24	2.98	0.41	0.44
Lagenammina difflugiformis	6.41	6.53	1.22	0.10	0.76	1.91	1.28	0.61	2.98	1.78	0.44
Sigmoilopsis schlumbergeri	0.39	2.30	1.01	1.42	2.28	2.69	2.70	3.67	1.19	0.27	3.52
Textularia sp.	11.78	11.13	23.73	20.36	15.75	14.01	16.64	17.96	15.48	13.31	17.18

and their spatial distribution is shown in Figure 3. These species were selected for statistical analyses.

Diversity indexes for the different assemblages are reported in

Table 3. Diversity indexes for the foraminifera assemblages in the different samples: number of Taxa, number of specimens/gr sediment (BFN), Shannon–Wiener entropy index (H), Dominance (D) and Simpson’s Index of Diversity (1-D).

Sample	Taxa	BFN	Shannon_H	Dominance_D	Simpson_1-D
1 BC	50	38	3.06	0.07	0.93
2 BC	50	38	3.29	0.05	0.95
6 BC	41	663	2.77	0.11	0.89
7 G	42	394	2.87	0.10	0.90
8 G	41	67	3.06	0.07	0.93
9 BC	50	76	3.19	0.06	0.94
11 G	44	45	3.06	0.07	0.93
12 G	35	33	2.94	0.08	0.92
14 BC	38	29	3.07	0.07	0.93
15 G	44	199	3.03	0.08	0.92
16 G	37	54	3.06	0.07	0.93

and Figure 4. On average, 43 taxa were identified per sample, ranging from a minimum of 35 (12 G) to a maximum of 50 taxa off the Neretva river mouth (1 BC, 2 BC). The number of specimens per gram of dry sediment mass (BFN) was largely variable, ranging from a minimum of 33 (12 G) to a maximum of 663 specimens (6 BC).

Benthic foraminiferal assemblages are characterized by the Shannon (H) index ranging between 2.77 in 6 BC and 3.29 in 2 BC, with a median of 3.03; Dominance (D) varied between 0.05 in 2 BC and 0.11 in 6 BC, with a median of 0.07; finally, the Simpson index (S) is not significantly different in the investigated area with comparable values throughout the samples, the extremes found in samples 6 BC and 2 BC (0.89–0.95, respectively) have a median of 0.92.

The cluster analysis on foraminiferal fauna enabled the distinction of two major groups of samples (A and B). Both of them can be subdivided into two more (A1, A2; and B1, B2) (Figure 5).

Cluster A1 comprises two box-corer samples (1 BC and 2 BC) sited in front of the Neretva river mouth. In these assemblages, the most common species are P. granosum (8.06–9.55%), V. bradyana (7.33–8.06%), Rectuvigerina sp. (6.91–8.77%), N. turgida (4.03–7.85%), A. beccarii (4.97–5.95%), A. perlucida (3.66–5.18%) and the agglutinants taxa Textularia sp. (11.13–11.78%) and E. scaber (3.07–8.64%). Abundance ranged from 29 to 38 individuals, H ranged from 3.06 to 3.29, D ranged from 0.05 to 0.07, and S from 0.93 to 0.95.

Cluster A2 includes samples 8 G, 9 BC, 11 G, 12 G and 14 G located along the northern part of the Croatian coastline and in the central sector of the Neretva channel. The benthic assemblages are mainly defined by the presence of Textularia sp. (14.01–17.96%), P. granosum (11.1–12.5%), Miliolids (7.03–9.66%), A. beccarii (5.38–9.80%), N. turgida (3.80–9.19%), B. marginata (3.70–10.25%), B. elongata (3.27–5.12%) and V. bradyana (2.09–7.76%). Abundance ranged from 33 to 76 individuals, H ranged from 2.93 to 3.19, D ranged from 0.06 to 0.08, and S from 0.93 to 0.94.

Cluster B1 groups the two deeper samples (6 BC and 7 G), located in the north-western sector, principally represented by Textularia sp. (20.36–23.73%), P. granosum, (15.9–17%), R. spinulosa (11.65–14.40%), Cibicides group (4.67–7.09%) and B. granulata (4.67–4.86%). Abundance ranged from 394 to 663 BFN, H ranged from 2.77 to 2.86, D ranged from 0.09 to 0.11, and S from 0.89 to 0.90.

Cluster B2 gathers two grab samples (15 G and 16 G) both located in the southeastern sector of the study area. It was mostly characterized by Textularia sp. (13.31–17.18%), P. granosum (13.2–20.6%), and R. spinulosa (5.76–6.17%). On the whole, in this cluster Miliolids are generally abundant (12.7–15.85%) and are represented by the Miliolinella subrotunda (3.52–6.31%). Abundance ranged from 54 to 199 individuals, H ranged from 3.03 to 3.06, D ranged from 0.07 to 0.08, and S from 0.92 to 0.93.

4.3. Two-Block Partial Least Squares

The results from the first axis of the 2B-PLS analysis are reported in Figure 6. The R2 for Block 1 (i.e., x-axis of the plot, Foraminifera Community) and Block 2 (i.e., y-axis of the plot, Chemical and Physical Variables) were 0.43 and 0.63, respectively. The analysis showed a gradient from the lower left quadrant to the upper right quadrant of the graph, with no notable outliers. Most samples are found in the lower left quadrant, where the most extreme points are represented by samples 1 BC and 2 BC located in front of the Neretva river. The variables most correlated with these observations (considering the covariance between the matrices) are certain taxa of the foraminifera community (led by V. bradyana, E. scaber, Rectuvigerina sp. and N. turgida) and chemical–physical variables such as silt content and total organic carbon (TOC%). On the contrary, in the upper right quadrant, only two samples (6 BC and 7 G) are found, which not only occur at greater depths and with a higher sand content, but also showed a different foraminiferal community, led by a different group of taxa (mostly R. spinulosa, B. granulata, and Asterigerinata mamilla). Some taxa (such as Bulimina sp., Miliolids, and Sigmoilopsis schlumbergeri) contribute little to the covariance, as do physico-chemical variables such as sand content.

5. Discussion

In the investigated area, the benthic foraminifera distribution results primarily correlated with three main factors: sediment grain size, total organic carbon and water depth. (Figure 6). This is in agreement with studies performed in tanatocenosis from the North Adriatic river-influenced shelf [14,19]. Both cluster analysis and 2B-PLS discriminate two distinct faunal groups with two additional sub-groups reflecting different depositional settings (clusters A1, A2, B1, B2).

Foraminiferal assemblages in samples collected in proximity to Neretva river (cluster A1) result primarily controlled by organic carbon content and by fine-grained substrate (silt). The benthic foraminiferal microfauna is characterized by the presence of shallow infaunal taxa (V. bradyana, N. turgida and B. elongata), commonly regarded as eutrophic species feeding on the low-quality organic matter [4,14,65,66,67]. The detected abundance of these taxa was favoured by the nutrient availability linked to the Neretva river discharge. In particular, the high relative presence of V. bradyana (Figure 3) was already documented as associated with high organic carbon content in the Rone prodelta [68], in front of the Ombrone River mouth (Italy) [69] and the Po delta [70]. N. turgida (Figure 3), an opportunistic and stress-tolerant species, is able to live in hypoxic sediments enriched in low-quality organic matter [16,70]. Similarly, B. elongata (Figure 3) is reported to thrive in shallow marine, silty to sandy substrates under the riverine influence [15,69,70,71].

In addition, A. beccarii and A. perlucida show relevant importance in this cluster (Figure 3). A. beccarii is well known to colonize environments under river influence, widely distributed in intertidal and subtidal zones [72]; it survives under a wide range of values of dissolved oxygen [73], salinity, and temperature [74]. In the Adriatic Sea, A. beccarii inhabits shallow and nutrient-rich environments near the coast under the direct influence of the Po River [17]. A. perlucida is an infaunal species tolerant of stressed environments [13], related to low salinity and shallow water settings [75]. Van der Zwaan and Jorissen [65] report that this species is typical of northern Adriatic clayey sediments where nutrient inputs by the rivers lead to periodic oxygen depletion events.

In the same way, the distribution pattern of the euryhaline taxa as E. scaber can be correlated to the Neretva freshwater input [76,77], decreasing with distance from the shore (Figure 3). The detected high abundance of E. scaber was documented in shallow, highly energetic, and organic-matter-rich environments [54,78]. Additionally, the abundance values of the agglutinated L. difflugiformis (mostly in samples 1 BC and 2 BC; Figure 3) are in agreement with fine-grained sediment enriched in organic matter [79]. The decrease in percentages of Rectuvigerina sp. from sample 1 BC to 2 BC (Figure 3) well correlates with the decrease in organic carbon supply moving offshore (Figure 2). This trend is in agreement with available studies showing that Rectuvigerina is a typical shallow continental shelf genus [80,81,82], generally living in sandy to muddy shelf environments with relatively high organic matter content [83,84]. It was also recorded in different oxic environments [77,85]. We do not have available data on oxygen content, however, the described assemblages suggest periodical oxygen deficiency. Within these areas, oxygen deficiency can be developed in response to the high organic matter concentrations [65].

The foraminiferal assemblage in samples collected in the central part and in the north-western direction along the Croatian coastline (cluster A2) is mainly composed of species characteristic to the infra-circalittoral zones of the Mediterranean and northern Adriatic seas [12,14,20,21,25,43,55,86,87]. Here, the significant presence of eutrophic taxa such as N. turgida, V. bradyana and Bulimina, associated with the percentages of A. beccarii (Figure 3) still reflect the riverine influx in the form of organic matter and sediment inputs, but to a much lesser degree than observed in sub-cluster A1. It results that foraminiferal distribution is affected by the position of the Neretva river plume that spreads along the coast in a NW direction during most of the year, decreasing its influx northward [28].

The data indicate that the principal factor influencing microfauna distribution, (group A) is the organic matter content from the fluvial input. In detail, the differences observed on clusters A1 and A2 are well correlated with the amount and the type of organic matter. Sediments of samples BC 1 and BC 2 (cluster A1) close to the river mouth are characterized by large amounts of terrestrial organic carbon with a clear terrestrial signature [28] whereas sediments at an intermediate distance from the river (cluster A2) contain lower contents of organic carbon showing the transition between terrestrial and marine origins [28].

Water depth and coarse-grained substratum (sand) are identified as the main drivers influencing the foraminiferal content in the northern sector (cluster B1), with open marine water entering the gulf. In this area, the lithology is mainly composed of sand due to the influence of higher energy currents from the Adriatic Sea that remove the finest part of the sediment [28]. Here R. spinulosa, Textularia sp. and B. granulata reach their maximum abundance (Figure 3). These species characterize sandy/muddy substrates along shelf edges subject to bottom currents [88] with slight tolerance of variable salinities [50,74]. Additionally, in the area we detect A. mamilla commonly related to high hydrodynamic energy levels with coarse-grained sediments [46,86]. The relatively deeper water and hydrodynamic conditions favoured also the presence of the Cibicides group (Figure 3) which is known to be associated with well-oxygenated environments with stable physico-chemical sea bottom conditions [50,51,52,53]. Cluster B1 hosts the highest number of species (Table 3), confirming favourable environmental conditions at the sea floor, and allowing the development of diversified benthic foraminiferal communities.

Cluster B2 is located in the southern sector south of the river mouth. This sector is an area where wave-generated longshore currents laterally redistribute the sediments, leading to the presence of relict deposits [28,30]. The relatively low amount of A. beccarii (Figure 3) and C/N values (Figure 2) correlate with the sediment grain size in this sector where the river influence is minimal. In this cluster, the higher percentages of Miliolids (Figure 3) may be explained by the shallow depth with low organic matter content at the sea bottom as similarly observed in other Mediterranean shallow marine settings [4,47]. The abundance of M. subrotunda supports available records documenting this species in an inner-shallow gulf environment with possible confined conditions [47]. The detected relatively high percentages of P. granosum, unkeeled elphidiis species, suggest the preference of this taxon for the coarse sediment and wide salinity change in agreement with observation performed in shallow water depth (8–22 m) around the Po delta area [19,67]. In the literature, this species is found in brackish-hypersaline bays, marshes, coastal lagoons, estuaries, and inner shelves [7,56,57,58,74]. In our dataset, the relatively low percentages of P. granosum observed in samples 1 BC and 2 BC suggest that this species does not correlate with high organic matter content. This agrees with [89] that considers P. granosum a third-order opportunist, increasing its abundance as a response to higher organic matter supply, but disappearing with maximum organic matter enrichment [89].

6. Conclusions

Modern benthic foraminiferal assemblages were investigated in the Neretva Channel to evidence their relationships with the environmental factors characterizing this area. Statistical analysis and an integrative partial least square approach (2B-PLS) highlighted that the diversity and distribution of benthic foraminifera studied in 11 sample sites are principally correlated with substrate grain size, organic matter content, and water depth.

The data showed that samples collected in front of the Neretva delta contain opportunistic species, well adapted to high input of organic matter and silt content such as A. perlucida, E. scaber, N. turgida and Rectuvigerina sp. (cluster A1). The areal foraminiferal species distribution indicates that the influence of the river discharge is a function of distance from the shore. At the samples furthest away from the river mouth, the assemblages were characterized by A. beccarii, B. marginata, N. turgida and Textularia sp. (cluster A2). These species living there seem to profit from the combination of a more labile organic matter with a transition between terrestrial and marine signatures.

In the north-west sector, thus in connection with the open Adriatic Sea, the coarse (sand) sediment samples include a highly diversified foraminiferal assemblage with the presence of A. mamilla, B. granulata, Cibicides group, R. spinulosa, and Textularia sp. (cluster B1) typical of well-oxygenated and hydrodynamic environments.

On the other hand, in the inner bay, in the south of the studied area, the samples are mainly composed of P. granosum and Miliolids reflecting the inner-shallow gulf with more confined environmental conditions (cluster B2).