Macroinvertebrate Communities Associated with Hydrilla verticillata (Royle, 1839) and Relationship with Environmental Factors in Ono Lagoon, Southeast of Côte d’Ivoire

The macroinvertebrates associated with Hydrilla verticillata was studied in Ono lagoon, South-eastern of Côte d'Ivoire. Monthly samples of macrophytes with their associated macroinvertebrates were collected in upstream, centre and downstreamusing a Van veen grab of 0.314 m2 internal area. The environmental variables (temperature, transparency, depth, conductivity, TDS, pH, dissolved oxygen, , , and ) were also recorded. A total of 71 taxa belonging to 28 families, 11 orders, 05 classes and 03 phyla of which 40 taxa were recorded in upstream, 45 taxa in centre and 44 taxa in downstream. Insects numerically dominated the capture, comprising 91.55% of the collectedtaxa with Odonata and Coleoptera being the most diverse and abundant groups. The density was higher in upstream (1407ind. per 100 g d.w.) and lower in downstream (1062 ind. per 100 g d.w.), whist theLibellulidae and Corduliidae exhibited the highest density communities. The rarefied richness did not show spatial variation but vary significantly between seasons. The Evenness did not show spatial and seasonal variations. However, Shannon diversity index varied significantly between sites and seasons. From the results of RDA analysis, conductivity and pH showed a strong environmental gradient and had a structuring effect on macroinvertebrate communities.


Data collection and laboratory procedure
Sampling in Ono Lagoon spanned a period of one year (September 2015 to August 2016). Samples of water and macrophytes with their associated macroinvertebrates were monthly collected in upstream, centre and downstream (Fig.  1). Hydrilla verticillatawas sampled using a Van veen grab of 0.314 m 2 internal area. The great transparency of water allowed us to position and close the grab with a relative precision at the level of the portion to be studied. Parts of macrophytes reported in excess and overflowing the jaws of the grab were removed by cutting, to keep only those located inside. All plant material and associated macroinvertebrates of the grab were washed into a bowl, then filtered through a 0.2 mm mesh sieve. Subsequently, the samples were preserved in a 10% formaldehyde solution in a plastic container for further analysis. At laboratory, preserved samples were washed to remove formaldehyde solution and then screened through a 500 µm mesh size to collect all macroinvertebrates on white plates. They were then fixed in a 70% alcohol solution for identification , counting and weighing. Separated and washed plants were drained of excess water, weighed to estimate plant wet biomass and dried up to 105°C for 2 daysto express the dry weight. Large macroinvertebrates were sorted by the naked eye while smaller fauna was sorted under a binocular loupe.

2.3.Measurement of environmental variables
Simultaneously to the biological examinations, a number of physical and chemical analyses were conducted. On each sampling site, the temperature, transparency, water depth, pH, Total Dissolved Solids (TDS), conductivity and dissolved oxygen were monthly measured in situ between 08.00 am and 10.00 am. Water samples were taken, stored in polyethylene bottles (500 mL) and kept at a temperature below 4°C and conducted to laboratory where analyses of dissolved inorganic nutrients (ammonium + 4 NH , nitrite 2 NO  , nitrate 3 NO  and phosphorus 3 4 PO  ) were carried using a spectrophotometer Model HACH DR 6000.

Data analysis
The macroinvertebrate density was characterised based on the total number of individuals (N) per 100 g dry weight (d.w.) of macrophytes. Invertebrate diversity was assessed as: taxon richness, Shannon-Wiener diversity and evenness indices. The total taxa were rarefied for each site for a given number of individuals drawn randomly from a sample (Magurran, 2004). The rarefaction was used to avoid any bias related to differences inabundances between samples using the lowest abundance (56 individuals for this s tudy) found in all sites asthe target number of individuals following Oksanen et al., 2013. Before performing comparison analyses, data normality was tested using Shapiro test. As the biotic and abiotic data distribution was not normal (P<0.05), the non-parametric test of Kruskal-Wallis was performed to compare data between sampling sites. When Kruskal-Wallis test is significant, Mann-Whitney U test was used for pairwise comparison. Redundancy Analysis (RDA) was used to assess relationships between macroinvertebrate distribution and environmental variables. A Monte Carlo permutation test was performed to assess the statistical significance of the environment variables and the full model to arrive at the significance of the first two axes. All analyses were conducted using the R package.

Environmental variables
The mean values of Ono environmental variables for study period are summarize in TABLE 1. Significant variation in water parameters was not observed among the sampling zones (ANOVA results of Kruskal-Wallis test, P > 0.05), except for temperature. This parameter varied significantly among the sampling sites (H2,36 = 3.37, p = 0.048) with thehighest value occurring in downstream (27,91 ± 1,51°C).For the other parameters, namely water depth, transparency, dissolved oxygen, nitrate, nitrite and phosphate, the values were slightly high in downstream.
The mean values of conductivity and total dissolved solids were rather high in upstream than in downstream.  17.30% in centre), Diptera (from 6.49% in upstream to 10.38% in centre) and Lepidoptera (from 3.29% in centre to 11.79% in downstream). The lowest densities were observed in Ep hemeroptera, Decapoda, Pharyngobdelliformes, Architaenioglossa and Plecoptera.Libellulidae (from 19% in both centre and downstream to 24% in upstream) and Corduliidae (from 10% upstream to 16.64% in centre) exhibited the highest family densities in sampling stations.     (Table.2). This richness did not show spatial variation (Fig. 2) but vary significantly between seasons (upstream and centre). In downstream, there were no significant differences amongst seasons (Fig.3 A). The Shannon diversity index showed a significant difference between sites ,withhighest values recorded in centre (3.18) and lowest values in downstream (2.21). The Shannon-Weaver index is greater than 2 in all sampling sites (Fig.2). Significant seasonal variations were found and the values were significantly highest in rainy season (centre and downstream) and lowest in dry season (upstream, centre and downstream) (Fig.3 B). The Evenness index ranged from 0.7 in upstream to0.98in downstream and did not show spatial and seasonal variations ( Fig.2&Fig.3 C).

Macroinvertebrate communities and environmental factors influence
The results of the Redundancy Analysis (RDA) showed that the correlation between environmental variables and macroinvertebrate taxa was mainly explained by the first two axes with 49.96% of total variance (Fig.4). From the RDA ordination diagram, two factors (conductivity and pH) had a significant impact on the macroinvertebrate communities. The temperaturewas the only parameter positively correlated with the axis 1 and no parameters were negatively correlated with this axis. The second axis, was positively correlated with conductivity and phosphate but negatively with pH, dissolved oxygen, nitrate and nitrite.  IV. DISCUSSION Analysis of the physical and chemical parameters of Ono lagoon reveals that the parameters (water depth, pH, dissolved oxygen, transparency, TDS, conductivity, nitrate, nitrite and phosphate)show no significant variation between sites, except for the temperature. Temperature plays an important role in the physical and chemical characteristics of lagoon environment, affecting the rate of CO2 fixation by phytoplankton (primary productivity) and solubility of gases such as O2, CO2 and  (Kraet al., 2018). The contamination of surface waters by total phosphorus can be induced by leaching of cropland containing phosphate fertilizers and some pesticides. Indeed, the Ono lagoon watershed closes several industrial plantations (rubber, palm oil, pineapple) that require the use of fertilizers and pesticides over large areas. A total of 71 macroinvertebrates taxa belonging to 28 families, 11 orders, 05 classes and 03 phyla were recorded demonstrating a relatively similar taxa to those recorded in Manasbal Lake (Sami et al., 2012) and Atchafalaya River Basin (Colon-Gaud, 2003). This indicates that macrophytes provide excellent microhabitats that enhance the establishment and colonisation of many invertebrates. Rakhi et al. (2014) reported that H. verticillate represents a suitable well-illuminated substrate in the water column. The dominant groups of our study were Odonata and Coleoptera whereas Diptera, Gasteropoda, Ephemeroptera, Decapoda, Amphipoda and Hemiptera dominated macroinvertebrate community of Atchafalaya River Basin. However, the number of taxafound in this study was higher than that of Scott and Osborne (1981)in Central Florida Lake (54 taxa), Heather et al. (2008) in Earthen experimental ponds (24 taxa) and Sami et al. (2012) in Manasbal lake (15 taxa). Insects numerically dominated the capture, comprising 91.55% of the collected taxa with Odonata and Coleoptera being the most diverse and abundant groups. The high insect taxonomic representativeness and abundance of this group was a pattern also observed in other studies (Tomazet al., 2008;Lucca et al., 2010). According to Tachetet al. (2003), insects represent one of the most important groups of freshwater invertebrates especially due to their diversity. Some macroinvertebrate taxa (Ephemeroptera and Plecoptera) were not recorded in upstream where lowest values of dissolved oxygen were measured. Also, differences in taxon composition and abundance within the same taxonomic group (class or order) were observed. Similarly, Merritt and Cummins (1996)found that macroinvertebrate abundance and species composition was strongly influenced by water quality. The total density of macroinvertebrates found in the microhabitat created by H.verticillatawas higher (1407 ind. per 100 g d. w.) in upstream and lower (1062 ind. per 100 g d. w.) in downstream. The groups which had the highest densities were Insects (from94,44% to 97,45%) of all the organisms collected, with Odonata (from 44,81% to 48,54%) being the most abundant groups. Our results are similar to findings of Poi De Neiff and Carignan (1997) who observed the same situation in floodplain of the Paraná River. Odonata larvae are known to use the aquatic plants as their egg laying site (Singh, 1989) and as ambush points to capture their prey (Merritt and Cummins, 1996). Density of coleoptera was found to be the second dominant as reported in SantragachiJheel Lake by Patra et al. (2012), attesting that Odonata and Coleoptera are well associated with macrophyte. The rarefied richness did not vary significantly among sites but showed a significant seasonal variation. The rarefied richness shows that in absence of any bias in samples, flood and rainyseasons were rich in the number of species . The Shannon diversity index showed significant spatial and seasonal variationsfrom a minimum of 2.2 (downstream) to a maximum of 3.18 (centre),suggesting that the centre was able to sustain a richer associated community. In addition, the lowest and highest values were respectively recorded in dry and rainy seasons, indicating that rainy and flood seasons were able to sustain a richer associated macroinvertebrates community than dry season. The maximum of Shannon diversity index takes place when species richness increases as found by (Brown and Lomolino, 1998). Concerning the Evenness values, no significant spatial and seasonal variations was observed. However, Evenness values varied from a minimum of 0, 74 in upstream to a maximum of 0, 96 in downstream. These values are high when compared with those recorded in subtropical lakes of south Brazil(Albertoniet al., 2007), in Central Florida Lake (Scott et al., 1981) and in Lake Nasser Khors of Egypt(Gawad and Abdel-Aal, 2018), showing the equitability in the distribution of individuals among the species. According to Schäfer (1980), high levels of evenness indicate an environment with heterogeneous conditions regulated by a community which is rich in the number of species and the multiplicity of their mutual relationships. From the results of RDA analysis, macroinvertebrate communities of Ono lagoon were most influenced by water quality variables such as conductivity, pH, temperature, phosphate, ammonium-nitrogen and nitrate. However, conductivity and pH showed a strong environmental gradient and had a structuring effect on macroinvertebrate communities. These communities were similar in rainy season and different in flood season.

V.
CONCLUSION In this investigation, we collected an important number of species (71) associated with Hydrilla verticillata. Insects with 91.55% of total species was the most diverse class. This class recorded the highest density community (94.44% to 97.45%) and macroinvertebrate assemblages were qualitatively (22 taxa) and quantitatively (44.81% to 48.54% of total density) dominated by Odonata. The centre recorded the greatest taxonomic richness (45 taxa). However, the higher density was reported in upstream (1407 ind.per 100 g d.w.). Distribution of aquatic macroinvertebrates by H. verticillata was best explained by conductivity, pH, temperature, ammonium-nitrogen, phosphate and nitrate. The spatio-temporal values of