*2.2. Sampling*

The sampling stations were located as follows: "Oglio" in the Oglio River (left tributary of the Po River), "Po" (within the Po River, just before the beginning of the Po-di-Goro River, which is part of the Po delta and flows into the Sacca di Goro Lagoon), "Gorino" (at the end of the Po-di-Goro river), "Giralda" (inside the Sacca di Goro Lagoon), "Delta" (located at the mouth of the Sacca di Goro Lagoon), and "Cesenatico" (located south of the lagoon, in the marine coastal environment) (Figure 1). The sampling was carried out in two different seasons: summer (September 2011) and winter (February 2012), identified as representative of periods of low (late summer) vs. high (winter) river outflow, respectively.

The bottom temperature and salinity were measured in situ by means of a multiparametric probe. Sediment samples were collected (three independent replicates per station and sampling period, for both OM and meiofauna) using plexiglass corers operated manually, kept at in situ temperature after being brought to the laboratory (within 4 h), and then immediately frozen once in the laboratory and kept at −20 ◦C until analysis (within 2 weeks). Only sediment aliquots for the measurement of OM degradation rates were immediately treated, as described below.

#### *2.3. Sedimentary Organic Matter Quantity, Nutritional Quality, and Degradation Rates*

Once in the laboratory, the top centimeter from each sediment core was used for analyses of the OM biochemical composition, in terms of the total phytopigment, protein, carbohydrate, and lipid contents. Chlorophyll-a and phaeopigments were analysed fluorometrically [48]. Total phytopigment concentrations were defined as the sum of chlorophyll-a and phaeopigment concentrations and utilized as an estimate of the organic material of algal origin [12]. Sediment phytopigment concentrations were converted into C equivalents using 40 μg C μg phytopigment−<sup>1</sup> as a conversion factor [3]. Protein, carbohydrate, and lipid analyses were carried out spectrophotometrically [49]. For the analysis of each biochemical class of organic compound, blanks were made with the same sediment samples previously treated in a muffle furnace (450 ◦C, 2 h). Protein, carbohydrate, and lipid concentrations were converted into C equivalents using the conversion factors 0.49, 0.40, and 0.75 mg C mg<sup>−</sup>1, respectively, and their sum is referred to as the biopolymeric C (BPC) [50].

The fraction of biopolymeric C represented by relatively fresh algal material was assessed as the percentage contribution of phytopigment C to biopolymeric C contents and referred to as the algal fraction of biopolymeric C [3]. The algal and protein fractions of biopolymeric C and the values of the protein to carbohydrate ratio were also used as descriptors of OM nutritional quality (algal and protein fractions) and ageing (protein to carbohydrate ratio) [50,51].

OM degradation rates were estimated from aminopeptidase and beta-glucosidase activities determined by the cleavage of fluorogenic substrates (L-leucine-4-methylcoumarinyl- 7-amide, Leu-; 4-methylumbelliferone-β-D-glucopyranoside, respectively) at saturating concentrations. Briefly, 2.5 mL of sediment subsamples was incubated at in situ temperature in the dark for 2 h with 2.5 mL of filtered, sterile water containing 200 μM L-leucine-4-methylcumarinyl-7-amide and 50 μM 4-methylumbelliferyl β-D-glucopyranoside, respectively, separately for aminopeptidase and β-glucosidase determinations. After incubation, the sediment slurries were centrifuged and the supernatants were analysed fluorometrically [49]. Protease and glucosidase activities (μmol of substrate g−<sup>1</sup> h<sup>−</sup>1) were converted into C equivalents using 72 as a conversion factor [13], and their sum is reported as the C degradation rate (μgC g−<sup>1</sup> h<sup>−</sup>1).

#### *2.4. Prokaryotic Abundance, Biomass, and Diversity*

The prokaryotic abundance and biomass were determined as described by Danovaro (2009) [49]. Briefly, prokaryotic cells were extracted from the sediments according to standard procedures, stained with SYBR Green I, and counted by epifluorescence microscopy. For determination of the prokaryotic biomass, the cell biovolume was converted into the carbon content assuming 310 fg C μm−<sup>3</sup> as a conversion factor [49]. The prokaryotic abundance and biomass were normalized to the sediment dry weight after desiccation (60 ◦C, 24 h).

The prokaryotic diversity was assessed according to Danovaro (2009) [49]. DNA was extracted from sediment with the UltraClean soil DNA isolation kit (MoBio Laboratories Inc., Carlsbad, CA USA). In all samples, extracted DNA was determined spectrofluorimetrically using SYBR Green I (Molecular Probes) and quantified vs. standard solutions of genomic DNA from *Escherichia coli*. The extracted DNA was amplified using universal bacterial primers 16S-1392F and 23S-125R, and the latter was fluorescently labeled with the fluorochrome HEX (MWGspa Biotech). PCRs were performed in 50-μL volumes in a thermal cycler (Biometra, Germany) using 30 PCR cycles. PCR products were checked on agarose–Tris-borate-EDTA (TBE) gels (1%). Four different reactions were run for each sample and then combined to form two duplicate PCRs, which were subsequently utilized for Automated Ribosomal Intergenic Spacer Analysis (ARISA). The quality of amplified fragments was checked, and the PCR products were purified and quantified spectrofluorimetrically. For each ARISA, about 5 ng of amplicons was mixed with 14 μL of internal size standard (GS2500-ROX; Applied Biosystems, Foster City, CA, USA) and the automated detection of ARISA fragments was carried out using the ABI Prism 3100 Genetic Analyzer (Applied Biosystems). ARISA fragments in the range of 390 to 1400 bp were determined using GeneScan analytical software version 2.02 (Applied Biosystems). Despite the fact that the DNA fingerprinting approach utilized in the present manuscript does not provide specific taxonomic information for the identified Operational Taxonomic Units (OTUs), it is still a largely utilized approach when assessing the patterns of prokaryotic diversity in environmental samples [52–55].

#### *2.5. Meiofauna*

Once in the laboratory, sediment samples for meiofaunal analyses were sliced into five sediment layers (i.e., 0–1, 1–3, 3–5, 5–10, and 10–15 cm), fixed with 4% buffered formalin, and stained with Rose Bengal (0.5 g L<sup>−</sup>1) until analysis. Sediments were sieved through a 500-μm mesh, and a 20-μm mesh was used to retain the smallest organisms. The fraction remaining on the latter sieve was re-suspended and centrifuged three times with Ludox HS40 [21,49] (diluted with water to a final density of 1.18 g cm<sup>−</sup>3). All animals remaining in the surnatant were again passed through a 20-μm mesh net, washed with tap water and, after staining with Rose Bengal, sorted under a stereomicroscope [49] (×40 magnification).

The meiofaunal biomass was assessed by bio-volumetric measurements for all specimens encountered. The nematode biomass was calculated from the biovolume, using the formula reported in Andrassy (1956) [56]: V = L <sup>×</sup> W2 <sup>×</sup> 0.063 <sup>×</sup> 10−<sup>5</sup> (in which body length, L, and width, W, are expressed in mm). The body volumes of all other taxa were derived from measurements of body length (L, in mm) and width (W, in mm), using the formula V = L <sup>×</sup> W2 <sup>×</sup> C, where C is the approximate conversion factor for each meiofaunal taxon [57]. Each body volume was multiplied by an average density (1.13 g cm−3) to obtain the biomass (mg DW: mg WW = 0.25) and the carbon content was considered to be 40% of the dry weight [58]. The biomass was expressed as μgC 10 cm<sup>−</sup>2.

#### *2.6. Statistical Analyses*

For all of the investigated variables, differences among sampling stations and sampling periods were assessed using distance-based permutational nonparametric analyses of variance (PERMANOVA) in a univariate context [59]. When significant differences were observed, pairwise tests were also carried out to ascertain patterns of differences among stations and/or sampling times. The sampling design

included two fixed orthogonal factors: Station (n = 6: Oglio, Po, Gorino, Giralda, Delta, and Cesenatico) and season (n = 2: summer and winter). Although time is typically a continuous source of variation, in this study, we considered the two sampling periods as levels of a fixed factor, assuming that they represented contrasting periods of the Po River discharge regime.

The same experimental design was used to test variations in i) the biochemical composition (in terms of protein, carbohydrate, lipid, and phytopigment contents) and nutritional quality (in terms of the protein to carbohydrate ratio, protein, and chlorophyll-a contributions to biopolymeric C) of sedimentary organic matter, and ii) the composition of OTUs and meiofaunal assemblages (based on abundance data only), again using PERMANOVA, in a multivariate context. The PERMANOVA analyses were based on matrixes of the Euclidean distance after normalization of the data (OM) and Bray Curtis similarity matrixes after square root transformations (prokaryotes and meiofauna) [60].

To visualize differences among stations and seasons in the biochemical composition, the nutritional quality of sedimentary organic matter, and the composition of OTUs and meiofaunal assemblages, bi-plots after a canonical analysis of the principal coordinates (CAP) were also produced [61]. Additionally, a Similarity percentage (SIMPER) analysis was carried out to assess the percentage dissimilarity in the meiofaunal taxonomic composition among systems and seasons. All statistical analyses were performed with the software PRIMER 6+ [62].

To assess whether the sedimentary organic matter content or nutritional quality explained significant differences in the prokaryotic and meiofaunal community composition, non-parametric multivariate multiple regression analyses, based on Euclidean distances, were also carried out using the DISTLM forward routine [60]. The forward selection of predictor variables was carried out with tests by permutation. P values were obtained using 9999 permutations of raw data for the marginal tests (tests of individual variables), whereas, for all of the conditional tests, the routine used 9999 permutations of residuals under a reduced model. Linear regressions were carried out using the Excel software.

#### **3. Results**

#### *3.1. Environmental Parameters*

The water temperature and salinity measured during the study period are reported in Table 1A. In summer, the temperature ranged from 17 to 25 ◦C, with the lowest values at sea (Cesenatico) and highest at the river (Oglio) sampling station, respectively. In winter, the temperature ranged from 8 to 13 ◦C, with the lowest values at the river and lagoon stations (Oglio, Gorino, and Giralda) and highest at the lagoon mouth station (Delta), respectively. In both summer and winter, the salinity ranged from 0 to 35, at the sampling stations located within the rivers (i.e., Oglio and Po) and at sea (Cesenatico), respectively.


**Table1.**Temperature,salinity,organicmattercontents(**A**),descriptorsoforganicmatternutritionalquality,anddegradationrates(**B**)inthesediments

#### *Diversity* **2020** , *12*, 189

#### *3.2. Content, Biochemical Composition, Nutritional Quality, and Degradation Rates of Organic Matter*

The chlorophyll-a, phaeopigment, total phytopigment, protein, carbohydrate, lipid, and biopolymeric C contents; algal and protein contributions to biopolymeric C; values of the protein to carbohydrate ratio; and OM degradation rates are reported in Table 1A,B.

The results of one-way PERMANOVA tests reveal a significant effect of the factor Station × Season for organic matter contents (Table S1A). In both seasons, the contents of almost all investigated variables, with only a few exceptions, were significantly the highest at the lagoon stations (Giralda and Gorino). At each station, almost all variables displayed contents in winter that were significantly higher than those in summer, with the exception for those at sea (Cesenatico), where the highest values occurred in summer.

The results of the multivariate PERMANOVA test also show a significant effect of the factor Station × Season on the OM biochemical composition, with significant differences among sampling stations in both seasons and between seasons at each station (Figure 2A).

**Figure 2.** Output of the canonical analysis of principal coordinates (CAP) on the sedimentary organic matter biochemical composition (**A**) and nutritional quality (**B**).

The results of univariate PERMANOVA tests on the nutritional quality and ageing of OM are reported in Table S1B. The factor Station × Season had a significant effect on all descriptors of OM nutritional quality, with values varying significantly among stations in both seasons and between seasons at almost all stations. Specifically, in summer, the algal fraction of biopolymeric C was the highest at the Po River station and lowest at the Oglio River, lagoon (Goro), and sea (Cesenatico) stations; in winter, values at the lagoon stations Delta and Giralda were higher than those in all other stations, whereas the lowest values occurred at the Po River station. The algal fraction of biopolymeric C was only higher in summer than in winter at the Po River station.

In both seasons, the protein fraction of biopolymeric C and the values of the protein to carbohydrate ratio were highest at sea (Cesenatico) and higher in summer than in winter at all sampling stations, except for the marine station (Cesenatico). Overall, the nutritional quality of OM varied significantly among all sampling stations in each season and between seasons at each sampling station (Figure 2B).

The results of one-way PERMANOVA and the consequent pair wise tests carried out on enzymatic activities are reported in Table S1C. In both seasons, the aminopeptidase activity was highest at the lagoon stations Giralda and Gorino, and significantly higher in summer than in winter at the Po River and Cesenatico (sea) stations. The β-glucosidase activity was highest at the lagoon station Giralda in summer and at the lagoon Delta and Po and Oglio River stations in winter. Values of the β-glucosidase activity were higher in winter than in summer at Oglio and Po River stations and at the lagoon Delta station, whereas at sea (Cesenatico), the highest values occurred in summer. In both seasons, the alkaline-phosphatase activity was highest at the lagoon Giralda station. Moreover, it was higher in summer than in winter at Po, Giralda, and Cesenatico stations, with the opposite pattern at Oglio and Delta stations.
