Variations of water quality parameters
The variations of water quality variables in different seasons of the study sites are shown in Table 1. No significant spatial differences (p≥0.05) in water variables were found, while the temporal differences exhibited significant heterogeneity (ANOVA post hoc test, p≤0.05). Out of 11 water parameters, seven parameters (Temp., salinity, DO, turbidity, PO4−P, NO3−N and NO2−N) differed significantly (p≤0.05) among seasons. The creeks water remained alkaline throughout the study period. Lowest pH was observed at S1 (6.51), and highest at S6 (8.2) throughout the sampling period. Seasonal mean surface water temperature ranged from 20.06±3.48 ºC to 31.0±2.41 ºC. The highest (32.70 ºC) and lowest (16.9 ºC) value was recorded at S1 in June and S6 in January, respectively. A wide variation of salinity was observed during the study period. The maximum salinity was recorded at S2 in June (27.2 ppt), and lowest at S5 in September (0.60 ppt). The salinity values varied at the different sampling stations according to their distance from the sea. Differences in dissolved oxygen concentration of the stations were found to be no significant. The higher magnitude of dissolved oxygen was obtained during post-monsoon (7.20±0.27 mgl-1), and it decreased from monsoon (6.35±0.15) to pre-monsoon (5.80±0.20). Turbidity was recorded in the expected level with its peak during monsoon and lowest in pre-monsoon. Mean seasonal turbidity varied from 23.35±14.48 to 88.63±16.25 NTU with its maximum (104.0 NTU) at S6 in October, and minimum (12.0 NTU) at S3 in January.
The seasonal trend of nutrients was not uniform in the studied stations. NO3−N contents were comparatively higher during monsoon, and its concentrations were recorded maximum at S6 and minimum at S1. A similar type of seasonal trend was noticed for dissolved PO4−P and NO2−N in the sampling sites. The mean concentrations of PO4−P and NO2−N were found to be decreasing from monsoon to post-monsoon, and again it started showing reverse trend during pre-monsoon. Dissolved SiO4−Si exhibited wide variation across seasons; however, post hoc analysis portrayed no significant difference (p≥0.05) among stations. Though ANOVA (post hoc test) portrayed no significant variations of water variables across the stations; PERMANOVA analysis reflected the significant spatial (F=1.459, p = 0.046) and temporal (F=7.923, p=0.001) differences in water variables in the studies creeks. Intra-relationship among various water quality parameters is described in Table 2. Water temperature exhibited significant positive correlation with turbidity (r = 0.682; p≤0.01), PO4−P (r = 0.535; p≤0.05) and NO3−N (r = 0.575; p≤0.05). Salinity in turn, was found to have a correlation with nutrient parameters viz., PO4−P (r = 0.443), NO3−N (r = − 0.50; p≤0.05) and SiO4−Si (r = 0.75; p≤0.01). DO also had significant negative correlation with turbidity (r = − 0.601; p≤0.01), PO4−P (r = − 0.723; p≤0.01) and SiO4−Si (r = − 0.591; p≤0.01) during the study period.
Zooplankton abundance, composition and distribution
The mean abundance of total zooplankton and copepod population in the studied stations are given in Table 3. A total of sixty three zooplankton taxa were recorded throughout the study period. Among zooplankton major groups, Copepoda was most prominent in terms of species richness and numerical abundance, varying in ranges from 59.55 to 73.13% of the total zooplankton. Among zooplankton Calanoida alone contributed 46% followed by Cyclopoida (15%) and Harpacticoida (6%). The other groups contributing significantly were larvae (13%), crustacean nauplii (9%) and Mysids (4%) (Fig. 2). The mean seasonal abundance was maximum during the pre-monsoon (50,861.11±23,702.63 ind.m-3) and lowest in the monsoon (29,805.56±17,571.72 ind.m-3). On the whole, the quantitative spectrum of total zooplankton ranged from 10,000 to 1,10,000 ind.m-3 with its maximum abundance at S1 and minimum at S5 during February and October, respectively. The station-wise average numerical abundance of zooplankton in different seasons is illustrated in Fig. 3. As a whole, holoplankters dominated throughout the study period in all the stations contributing 71.4−87.9% to the total zooplankton community with the highest in monsoon and lowest in pre−monsoon. The meroplankters contributed 11.2– 34.62% to the total zooplankton density with the highest in pre−monsoon and lowest in monsoon. The abundance of holoplankters and meroplankters in the different stations are shown in Fig. 4 (a & b).
A total of 41 species of Copepoda were recorded throughout the study comprising 30 species of Calanoida (14 genera), 5 species of Cyclopoida (3 genera) and 6 species of Harpecticoida (5 genera). The family Paracalanidae and Oithonidae invariably constituted the bulk of the total copepod population, and accounted for 41% and 17% respectively. Maximum species diversity that was recorded belonged to the families of Paracalanidae (7) followed by Acartiidae (5) and Pseudodiaptomidae (4). Out of 41 species of Copepoda recorded, 14 species (Paracalanous parvus, Parvocalanous dubia, Bestiolina similis, Acrocalanous gibber, A. gracilis, Acartia erythrea, A. spinicauda, Psudodiaptomus serricaudatus, P. annandalei, P. aurivilli, Oithona brevicornis, O. similis, Longipedia weberi and Microsetella norvegica) were perennially present in the creeks. That in turn, clearly highlights the euryhaline nature, and broad range of thermal tolerance of these species. The average numerical abundance of Copepoda was maximum during pre-monsoon (31,791±7,012 ind.m-3) followed by post−monsoon (29,152±5,597 ind.m-3) and monsoon (19,236±5,912 ind. m-3). The mean percentage contribution of Copepoda was highest at S5 (73%) and lowest at S2 (65%), whereas month-wise, it accounted maximum during May (74%). The correlation coefficient values of the family Paracalanidae, Acartiidae, Eucalanidae, Temoridae and Calanidae showed positive correlation between them, explaining the possible formation of groups by themselves (Table 4). Out of 17 copepod families, six families (Paracalanidae, Acartiidae, Temoridae, Calanidae, Oithonidae and Oncaeidae) differed significantly (ANOVA; p≤0.05) between seasons, while rest of the families did not portray significant temporal variations.
Albeit, Copepoda can be considered as the major group, which comprised of a large proportion of total zooplankton population in all the months, the occasional co−dominance of larvae (comprising of the polychaete, cirripede, isopod, amphipod, gastropod and fish larvae) and mysids were also documented. The mean abundance of larvae was maximum during pre−monsoon (6,222 ind.m-3) and minimum in monsoon (2,500 ind.m-3). Throughout the study period, the numerical abundance of larvae was maximum in October and minimum in July. Rotifera and Cladocera contributed 1% each to the total zooplankton throughout the study period. The temporal abundance of Rotifera exhibited two peaks (bi-modal pattern) in August and November. There was significant increase in average numerical abundance of Rotifera from pre−monsoon (55 ind.m-3) to post−monsoon (430 ind.m-3). Cladocera did not vary significantly across the seasons. Chaetognatha was represented by a single species, Zonosagitta bedoti with an average numerical abundance of 386 ind.m-3, and exhibited its maximum (1,505 ind.m-3) in January. Its abundance exhibited unimodal pattern across the seasons. Among the others, Mysids contributed significantly to the total zooplankton density, whereas Hydromedusae were spotted sporadically during the study period. Results of ANOVA (post hoc test) showed, no significant spatial differences in zooplankton (groups), while temporal differences exhibited significant heterogeneity (p≤0.05) among the zooplankton groups such as Copepoda (F= 4.061; p = 0.039), Cladocera (F= 7.046; p = 0.007), Chaetognatha (F=3.902; p= 0.05), crustacean nauplii (F=13.442; p= 0.001) and larvae (F=7.902; p= 0.005). Other three major groups viz. Rotifera (F=1.214; p=0.324), Eggs (F=1.220; p=0.323) and Mysids (F=1.73; p=0.930) did not show statistically significant variations across seasons.
The Cluster analysis on the similarity of zooplankton species abundance in the studied stations are shown in Fig. 5. The species abundance resulted in three clusters. The maximum similarity (83.9%) was observed between the stations S1 and S2 during post−monsoon, and lowest similarity (73%) between S3 and S6 during pre−monsoon. On the whole, the species composition in the samples exhibited >65% similarity between the stations. The grouping of species composition was further substantiated by the NMDS, and the plot also gave similar compositional pattern among samples. The assemblage of pre−monsoon differed substantially from that of the other two seasons, with regard to the zooplankton composition (Fig. 6). Results of ANOSIM reflected the significant (Global R = 0.101; significance level 0.1%) spatial differences of zooplankton species composition. The zooplankton community also depicted the significant temporal variation (R= 0.565; P 0.1%). The k-dominance curve extracted for the period 2016−17 reflected that the species abundance and diversity was noticeably higher during the pre−monsoon as compared to the other seasons (Fig. 7a), whereas in the following year (2017−18) the curves overlapped indicating no differences in the species diversity, illustrating the similar dominance patterns (Fig.7b). RELATE routine analysis further substantiated the significant (ρ = 0.399; P 0.1%) inter annual differences of the zooplankton community compositions between 1st and 2nd year study period. Though, one-way ANOVA (post hoc) results portrayed no significant variation of the zooplankton groups spatially (p>0.05), PERMANOVA design depicted the significant variations of zooplankton population both spatially (F=2.313; p = 0.001) and seasonally (F = 6.107; p = 0.001). On the whole, the spatio−temporal variations of zooplankton were found to be significant (F = 2.0; p = 0.001) in the study area.
SIMPER routine was analysed based on the abundance of zooplankton described the characterizing species (within the group similarity), and discriminating species (between the group dissimilarity). The average similarity percentage varied between 44.69 and 51.1% with the highest at S3 and lowest at S6. Table 5 summarizes the information on characterizing species and their percentage contribution to the total zooplankton abundance in each group. At S3, P. parvus (12.4%), B. simils (10.9%), O. brevicornis (9.76%) and P. dubia (9.70%) were most prominent. Similarly, B. similis (20.34%), O. brevicornis (9.84%), P. parvus (8.55%) and A. spinicauda (7.07%) were the main contributors at the station S6. The dissimilarity percentage calculated ranged from 50.34 to 56.81% with the highest percentage found to be between S2 and S6 (56.81%). Paracalanous parvus (6.20%), crustacean nauplii (5.65%), polychaete larvae (5.60%), O. brevicornis (5.34%), B. similis (3.89%), P. dubia (3.28%), Longipedia weberi (3.23%) and Acrocalanous gibber (3%) were the major contributors to this group. With that of percentage contribution (β−dissimilarity) in the different assemblages, these species exhibited as key zooplankton species that differentiated between the assemblage, and their variability makes them discriminating species in the different zooplankton assemblages.
The occurrence of Copepoda species varied considerably in seasons. Maximum species richness (36 species) was recorded at the station S3 during pre−monsoon followed by S2 (34 species) and S5 (32 species) during post−monsoon and pre−monsoon, respectively. The average species richness of the Copepoda was 28±4 species across the stations. On the whole, the seasonal species richness was highest during pre−monsoon (41 species) followed by monsoon (37 species) and post−monsoon (31species) throughout the study period. In respect of total zooplankton, maximum richness was accounted for 53 taxa at S2 during post−monsoon, whereas it was minimum at S4 (34 taxa) during monsoon. Significant spatial variations in the zooplankton taxa during monsoon (F = 5.667; p = 0.000) and post−monsoon (F = 9.634; p = 0.000) was observed, whereas the variation was not significant during pre−monsoon (F = 1.474; p = 0.219).
The Shannon-Wiener diversity (H'), Margalef’s species richness (dʹ) and Pielou’s evenness index (Jʹ) in the studied stations for both the year is shown in Fig. 8. As with total zooplankton population, the dʹ was highest at S1 during post−monsoon (5.47) and lowest at S5 during monsoon (2.55). The Hʹ values also depicted similar trend with the maximum at S1 during post−monsoon (2.71) and minimum at S2 during monsoon (1.66). The stations S1 and S6 showed similar assemblage pattern with regard to the richness, evenness and diversity of zooplankton. Throughout the study period, the zooplankton assemblage was not uniformly distributed in the studied stations and fluctuated greatly. The zooplankton species evenness (Jʹ) depicted that, it was relatively higher at S4 during pre−monsoon (0.86) and lower at S5 during monsoon (0.55). Results of ANOVA (post hoc test) showed significant variations (p≤0.05) of diversity indices (dʹ, Hʹ and 1−λʹ) across seasons except Jʹ. The correlation matrix exhibited a significant positive correlation between dʹ and Hʹ (r=0.923; p≤0.01), and also dʹ with 1−λʹ (r = 0.83; p≤0.05). Similarly, Jʹ had positive correlation with Hʹ (r = 0.42; p≤0.05) and 1−λʹ (r = 0.61; p≤0.01). All the correlation coefficient values of the indices showed positive correlation between them, implying similar pattern of species interaction in the zooplankton assemblage throughout the study.
β − diversity
The relative dispersion (inter stations β−dissimilarity) value calculated using the multivariate dispersion index. The stations S4 (=1.11), S2 (=1.13) and S6 (=1.16) showed higher values (>1) indicating high β−diversity compared to S3 (=0.72) and S5 (=0.77). The zooplankton species turnover was registered in the impacted stations (S4, S2 and S6) as the values reached β =1. However, the pair-wise comparison (stations compared) of IMD depicted lower values (<1.0) indicating community variability (stability) during the study period. The β−dissimilarity index revealed the inter-stations differences in the species distribution patterns in the study area. The highest β−dissimilarity percentage was found to be between S2 and S6 with 56.81% and lowest between S3 and S5 (50.34%).
Influence of physico-chemical parameters on zooplankton distribution
The correlation matrix showed that, Copepoda was positively correlated with salinity (r = 0.557; p≤0.05), DO (r = 0.46), and negatively correlated with NO3−N (r = −0.479; p≤0.05). Cladocera had positive correlation with salinity (r = 0.367), and significant negative correlation with turbidity (r = −0.60; p≤0.01). Crustacean nauplii and larvae also portrayed significant positive correlation with the salinity, PO4−P and SiO4−Si, and negative correlation with DO, turbidity and NO3−N (Table 2). By and large, total zooplankton displayed the significant positive correlation with salinity (r = 0.617; p≤0.05), and negative correlation with NO3−N (r = −0.575; p≤0.05).
For all the sampling stations, the correlation between zooplankton abundances and environmental variables were established by multivariate BIOENV analysis. The analysis confirms a set of environmental parameters that is related to spatial variations of zooplankton. BIO-ENV analysis (employing Spearman rank correlation method) reflected that the water quality parameters viz., pH, salinity, DO, turbidity, PO4−P, NO3−N were correlated with the zooplankton community (Table 6). The best set of correlation between zooplankton and the individual environmental variable obtained in the combination of salinity and PO4−P with the maximum coefficient of 0.210, whereas the minimum correlation coefficient (0.170) was recorded only with salinity. The correlation coefficient (ρ = 0.21) with the significance level 1.6% indicated that, two variables i. e. salinity and PO4−P were the controlling factors for the distribution of zooplankton in the studied stations. To quantify the additional explained variables, Marginal tests (distance based linear model, DistLM) was performed to obtain a significant correlation between the zooplankton and each of the environmental variables. The results showed that, the significance correlation (p≤0.05) has been observed, with the variables such as Temp., salinity, DO, turbidity, PO4−P and NO3−N. The Sequential tests also revealed similar results with the former having significant correlation (p≤0.05) with three variables viz., Temp., salinity and turbidity (Table 7). The fitted model (DistLM) explained that the two variables; Temp. and salinity were the deterministic parameters to explain the zooplankton community compositions in the studied stations (Table 8).