Meiobenthic assemblage of the grey mangrove (Avicennia marina) along the Saudi Arabian coast of the Red Sea with emphasis on free-living nematodes

Twenty different sites, representing various mangrove ecosystems with the grey mangrove (Avicennia marina) along the 1500 km coast of the Saudi Arabian Red Sea, were selected for the present study (Table 1 and Fig. 1). The sites are located between 16°30’ & 25°30’N and 36°10’ & 42°20’E, which covers a stretch of coast from Jazan in the south to Duba in the north (Fig. 1). All sampling sites were located in the leeward part of the subtidal zone. Of the five sites in the Jazan region, two of them (sites II and III) contain a dense population of dead mangrove trees, and another one (site IV) is located near the outlet of a sewage treatment plant. Four sites were selected in the Al-Birk region, with site VI located near a fish farm outlet. The Jeddah region was represented by one site (XI) located near a sewage treatment plant, and the Rabigh region was also represented by one site located near a fish farm effluent (XVI). Both the sewage treatment plant and the fish farm usually discharge the effluent into the coastal areas, after proper treatment and settlement. Al-Wajh and Duba regions comprise the most pristine mangroves in the Red Sea and they were represented by three different sites.

Figure 1

All field sampling work was carried out in February–March 2017 at low tide using a plastic corer with a length of 10 cm and a diameter of 7 cm. At each site, three replicates of sediment samples were collected for the meiofaunal studies and another three for analyzing the grain size as well as other abiotic parameters. The collected meiofaunal samples were immediately fixed with 4% neutral formalin in plastic bottles. In the laboratory, sediment samples were washed with tap water to remove the excess formalin and the meiofaunal organisms were sorted out using the documentation and illustration technique (Platt & Warwick 1983). The floating content was then sieved through a 40 μm mesh and the organisms retained on the sieve were collected in Petri dishes for sorting and later preserved in 70% ethyl alcohol along with 5% glycerol (Tolhurst et al. 2010). Small drops of Rose Bengal were also added to this solution to facilitate the counting process. All meiofaunal organisms were sorted into main groups and counted using a stereomicroscope. Nematodes were mounted on glass slides following the methodology of Somerfield & Warwick (1996). Identification of nematodes up to the genus level was carried out following the standard protocols of Platt & Warwick (1983, 1988) and Warwick et al. (1998). Clustering of nematodes based on their feeding status was carried out following the procedures of Wieser (1953). They were mainly classified into four feeding types: selective deposit feeders (1A), non-selective deposit feeders (1B), epigrowth (epistrate) feeders (2A) and omnivores/predators (2B).

In the laboratory, sediment samples were air dried in an oven at 80°C until constant weight and then mechanically sieved through a set of sieves with 2, 1, 0.5, 0.250, 0.125 and 0.063 mm mesh size and granulometric characteristics were determined following Morgans (1956). Total organic carbon (TOC) of dried subsamples was determined according to Parker (1983) by calculating the difference between the dry weight of the sediment (60°C, 24 h) and the residue remained after combustion at 550°C for 4 h. Seawater temperature and salinity were measured using a multi-parameter water quality analyzer (Horiba U50). Electrical conductivity EC (mS cm⁻¹) was measured in 1:2.5 soil:water extract and soil pH (saturated soil paste) as explained by Richards (1954).

Different univariate and multivariate tests were performed to find out the possible relations and interactions between the meiofaunal assemblages and other environmental parameters. One-way ANOVA was performed to determine significant differences within the study sites for different aspects such as meiofaunal abundance, nematode feeding habits, as well as for biodiversity indices using SPSS V23. Values of the Pearson correlation coefficient (r) were also determined to analyze significant correlations between meiofaunal assemblages and environmental parameters. Principal component analysis (PCA) was applied to provide detailed information on the interaction between the environmental variables and various meiofaunal taxa using the PAST software (V3). Analysis of the similarity (ANOSIM) test was carried out to determine the similarity of nematodes among different mangrove habitats. Furthermore, the nematode diversity indices were calculated based on the Shannon-Wiener index (H’), Pielou’s evenness index (J’) and Margalef’s species richness index (d). The Bray-Curtis similarity index and group average linkage were used to perform the multidimensional scaling (MDS) ordination of the abundance of different meiofaunal groups as well as the nematode genera. All these multivariate analyses (except PCA) were performed using the PRIMER 6.0 software (Clarke & Gorley 2006) and the data were transformed to square root prior to each analysis.

All abiotic factors at the study sites are presented in Table 2. The minimum seawater temperature and salinity were recorded at site IV (27.9°C and 38.2, respectively), while the maximum temperature (36.2°C) was measured at site XVIII and the maximum salinity (43.6) at site IX. The lowest pH value was determined at site VII (6.75), while the highest (8.08) at site XX. The EC values varied considerably at the surveyed sites and ranged from 2.31 and 22 mS cm⁻¹ at sites XIV and III, respectively, with an average of 8.89 mS cm⁻¹ (Table 2). Regarding the sediment characteristics, the sand fraction dominated at all sites, ranging from 88.1 (site II) to 99.2% (XIV). The contribution of the silt-clay fraction was considerably low and ranged between 0.76 and 11.96% at sites XIV and II, respectively. Values of the mean grain size ranged from 96 μm at site IV to 747 μm at site XIV, with an overall average of 288 μm (Table 2). Furthermore, the sediment sorting index showed that a total of 16 sites were characterized by poorly sorted texture, while four sites – by moderately sorted texture (Table 2). The total organic carbon content was generally low (average 2.4%) and ranged between 1% and 3.7% at sites IV and V, respectively (Table 2).

The meiofaunal abundance fluctuated between 119 and 1380 ind. 10 cm⁻² with an average value of 643 ± 462 ind. 10 cm⁻² (Fig. 2). Higher abundance of meiofauna was determined at six sites (I, VI, XII, XIII, XVI and XVII – 1370, 1008, 1350, 1110, 1350 and 1380 ind. 10 cm⁻², respectively; Fig. 3a). On the other hand, site XV showed the lowest abundance of 119 ind. 10 cm⁻² (Fig. 3a). A total of seven main taxa were identified in the sediment samples of mangrove stands (Fig. 2). The largest number of taxa (6) was observed at sites XIII and XVI, while the smallest number (1) was reported from site V (average: 4 ± 1 SD). Free-living nematodes, harpacticoid copepods and polychaetes were the most abundant taxa, accounting for almost 90.4% of the total meiofaunal abundance. Nematodes were the most abundant group (average abundance 413 ind. 10 cm⁻²) and accounted for about 64.3% of the total meiofauna. Their abundance varied between 72 and 1100 ind. 10 cm⁻² at sites II and XVII, respectively (Fig. 3b). Their contribution ranged between 29 and 100% of the total meiofaunal abundance recorded at each site. In terms of abundance, harpacticoid copepods were the next most dominant group and accounted for 13.2% of the total meiofaunal abundance at 14 sites, where it was present with an average of 85 ind. 10 cm⁻². Their abundance varied greatly at different sites, with particularly high abundance recorded at sites: XVII, XIII, VI, VII and XVI (Fig. 3c). Polychaetes were yet another relevant group, which accounted for about 12.9% of the total meiofaunal abundance. They were observed at 15 sites with an average abundance of 83 ind. 10 cm⁻². Their abundance was relatively higher at some sites (I, XII, XIII and XVI; Fig. 3d) with a maximum of 485 ind. 10 cm⁻² at site XIII, which in turn represented almost 43.7% of the total meiofauna obtained from that particular site. The contribution of oligochaetes and turbellarians to the total meiofauna was only 5.1 and 4.1% with mean counts of 33 and 27 ind. 10 cm⁻², respectively. Their maximum abundance was observed at site XVI with densities of 153 and 104 ind. 10 cm⁻², respectively (Fig. 3e–f). Other groups, such as ostracods and amphipods, were recorded at only one site with very low abundance.

Figure 2

Figure 3

A total of 25 genera belonging to 15 families and three orders were recorded from the study area (Table 3). Chromadorida were the dominant order and were represented by 12 genera, which in turn accounted for 40.2% of the total nematode assemblages. They were followed, in terms of abundance and distribution, by Enoplida represented by five families and seven genera (accounting for 34.3% of the total nematode counts) and Monhysterida with two families and six genera (25.5% of the total nematode counts). Microlaimidae, Oxystominidae, Ironidae and Desmodoridae were the most abundant families and accounted for 19.8%, 16.2%, 15% and 14.7%, respectively (Table 3), while Xyalidae was the richest family (four genera), followed by Desmodoridae (three genera; Table 3). Interestingly, about 42% of all these nematode genera were observed only once during the entire study period. Microlaimus had the highest relative abundance, accounting for 17.4% of the total identified nematodes, followed by Halalaimus with the relative abundance of 16.2%. Both Syringolaimus (14%) and Terschellingia (10.3%) were also present with considerable relative abundance (Table 3), making these four genera the most dominant ones, which together contributed about 58% to the total abundance of nematodes recorded in this study. Concerning the trophic structure of nematode assemblages, epistrate feeders were the dominant group (52.1%), followed by deposit feeders (44.4%), while predators/omnivores were represented by only 3.5% of the total nematode community. Epistrate-feeding (epigrowth) nematodes were significantly abundant at sites I, II, IV, V, VIII, X, XI, XVIII and XIX (Fig. 4). While selective deposit feeders dominated, in terms of abundance, at sites III, VI, IX, XIV, XV, XVI and XX, non-selective feeders showed greater abundance at sites VII, XIII and XX (Fig. 4).

Figure 4

One-way ANOVA showed significant differences between the sites in the total abundance of meiofauna as well as other individual groups (p < 0.001). Based on nematode feeding habits, one-way ANOVA also revealed significant differences between the

sites (p < 0.001). Pearson’s correlation (r) analysis revealed some significant correlations between the meiofaunal groups and various nematode genera with changing environmental parameters. The sand fraction showed significant positive correlation with the total meiofauna (r = 0.675, p < 0.01) and with different meiofaunal groups such as nematodes (r = 0.59, p < 0.01) and harpacticoid copepods (r = 0.61, p < 0.01) (Table 4). Turbellarians showed a negative correlation (r = −0.502, p < 0.05) with the percentage of the very fine sand fraction (Table 4). Some nematode genera like Paralongicyatholaimus (r = 0.73, p < 0.01), Dolicholaimus (r = 0.57, p < 0.01) and Metachromadora (r = 0.57, p < 0.01) are significantly positively correlated with temperature, while Halalaimus (r = 0.43, p < 0.05) and Microlaimus

(r=0.45, p < 0.05) – with the percentage of sand. PCA ordination (based on correlation) of environmental variables and meiofaunal groups showed that the first two components accounted for 52.14% of the total variance (PC1: 34.89% and PC2: 17.25%). The main environmental variables that define PC1 were sand, silt/clay, pH and EC along with all meiofaunal taxa, while PC2 was mainly represented by TOC, salinity and temperature (Fig. 5). Most of the meiofaunal groups were positively correlated with sandy substrate and pH, but negatively correlated with such factors as silt-clay fraction, EC and TOC (Fig. 5). ANOSIM values obtained for the distribution of different nematode genera in various habitats clearly indicate that there are no significant differences between habitats (p > 0.05). The largest number of nematode genera (S) was determined at three sites (XII, XVII and XVIII; Table 5). Margalef’s species richness index (d) varied between 0.27 and 2.15 at sites I and XVII, respectively (average 0.65 ± 0.63), while values of Pielou’s evenness index (J’) showed a considerable variation between the sites (F = 4.41, p < 0.05; Table 5). The Shannon diversity index (H’) fluctuated between 0.66 and 2.37 without any significant variation among the sites (F=3.95, p > 0.05; Table 5). The multidimensional scaling ordination (MDS) based on the Bray-Curtis similarity of the meiofaunal group abundance showed a 40% similarity between different sampling sites and two main groups (60% of similarity; Fig. 6a). The generic structure of nematodes also produced four main clusters with 40% similarity (Fig. 6b).

Figure 5

Figure 6