INTRODUCTION

Estuaries are coastal ecosystems that provide important services for global socioeconomic development. Services generally range from tourism and recreation to coastal protection, erosion control, nutrient recycling, water purification, carbon sequestration, and food and raw material production (^{Dauvin, 2007}). However, coastal zones are one of the most highly impacted marine ecosystems due to human activities (^{Halpern et al., 2008}).

The effects of human activities on estuarine ecosystems have been evaluated according to the diversity of macrobenthos because these communities are sensitive to anthropogenic pressures or natural stressors (^{Dauvin, 2007}). In addition, the benthic macroinvertebrates are not only important for evaluating the quality of estuaries and for management and conservation planning, but worldwide, fishing of 3.4 million tons of shrimp and prawns represents 10 billion dollars and approximately 16% of global fish exports (FAO, 2009).

The pressures on macrobenthic communities from shrimp trawling can be summarized as follows: changes in the physical structure, sediment suspension, chemical composition, and changes in the benthic community due to bycatch (^{Dauvin, 2007}; FAO, 2009). In addition, the proximity to the coast makes the estuarine bottom one of the main recipients of anthropogenic contamination (^{Bayen, 2012}). However, different adaptations that macrobenthos undergo in estuaries make it difficult to determine when a change in the community is due to an anthropogenic stressor or when it is due to the environmental variability of the ecosystem (FAO, 2009; ^{Lambert et al., 2011}) because the distribution and abundance of benthic organisms is influenced by environmental variables such as salinity, pH, depth, temperature, and the texture of the sediment and organic matter (^{Carvalho and Santos, 2013}; ^{Furlan et al., 2013}; ^{Martins et al., 2014}).

In the estuary of Buenaventura Bay, artisanal fishing plays an important socioeconomic role (^{Carvajal et al., 2011}). This estuarine ecosystem is fed by the Dagua and Anchicayá rivers, which contribute a large amount of sediment that accumulates over time and is mostly rich in organic matter (^{Lucero-Rincón et al., 2008}). Soft bottoms predominate in the Bay (^{Neira and Cantera, 2005}).

This study emphasizes the macroinvertebrates that inhabit the soft bottoms; however, in the rocky intertidal habitats, soft bottoms, sands, and mangroves of the Pacific of Valle del Cauca, 226 species of crustaceans belonging to 56 families and five orders have been reported (Sessilia, Stomatopoda, Amphipoda, Isopoda, and Decapoda) (^{Lazarus and Cantera, 2007}). For echinoderms, nine of the 66 species of the Pacific of the Cauca Valley have been reported in Buenaventura Bay, distributed in seven families and four orders (Paxillosida, Ophiurida, Cnidaria, and Scutellina). The poor representation of this group in the bay is mainly due to the preference of echinoderms for hard bottoms (^{Neira and Cantera, 2005}).

This study seeks to contribute to the research conducted in Buenaventura Bay on how environmental variables influence the dynamics of macroinvertebrates (^{Lucero et al., 2006}; ^{Hernández, 2015}), considering that these organisms are of socioeconomic importance. In addition, understanding the dynamics of communities as a function of the environmental variation will allow the impact from anthropogenic and natural pressures on the macrobenthos to be predicted.

MATERIALS AND METHODS

Field phase

Four samplings were carried out in each climatic season in 2015 (April-November) in four stations distributed in Buenaventura Bay according to a salinity gradient. The River Estuary (RE) station, located at 77°6'33.1"W and 3°50'5L5"N, was characterized as internal and influenced by the discharge of the Dagua River; the Inner Estuary (IE) station, located at 77°7'24.9''W and 3°52'4.4''N, was characterized as internal and consisting of a rocky shore; the External Estuary (EE) station, located at 77°9'35.9''W and 3°50'58.7''N, was characterized as external and consisting of a rocky shore; and finally, the Marine Estuary (ME) station, located at 77°8'58.4''W and 3°48'56.5''N, was characterized as external and influenced by the discharge of the Anchicaya River. The stations were separated by an average of 4 km (Figure 1).

Source: Own elaboration adapted from SIGOT.

Figure 1: Study area. Buenaventura Bay. (RE: River Estuary, IE: Inner Estuary, EE: External Estuary, ME: Marine Estuary.) 

For each station and season, water samples were taken to measure the physicochemical variables; sediments, for granulometry; and macroinvertebrates, for structure and abundance. Each sample had three replicates for a total of 48 samples. To determine the sediment composition, samples were collected from the substrate at a depth of 5 cm by means of a PVC corer (4.5 cm in diameter and 20 cm long). A Thermo Scientific brand portable multimeter was used to measure the physicochemical variables from the samples of surface water: temperature (°C), salinity, pH, and dissolved oxygen (DO mg Additionally, the transparency was measured with a Secchi disk, depth was measured with a Depth Sounder floating depth gauge, and the geographical coordinates were recorded with a GPS unit (GARMIN).

The samples of epibenthic macroinvertebrates were collected from a boat using a trawling net with 1-inch mesh and 8-m working width, with each trawl being independent and lasting 10 minutes; each experimental unit had three replicates, for a total of 48 trawls. The organisms and sediments were stored in plastic bags, refrigerated, and then taken to the laboratory to be stored at a temperature of -20 °C.

RESULTS

In Buenaventura Bay, environmental variables such as salinity, pH, dissolved oxygen, transparency, and sediment composition varied temporally (MANOVA, F=25.4, p<0.0001) across the stations (MANOVA, F=7.1, p<0.0001), and a seasonal influence also acted on the environmental variables at the stations (MANOVA, F=16.4, p<0.0001). The salinity was highest in June in RE, but this station also had the lowest salinity in November (ANOVA, p<0.0001). The pH was highest in April at ME, but in November, it had the lowest value in the same season (ANOVA, p<0.0001). Dissolved oxygen was highest in September at the OE station and lowest in November at the RE station (ANOVA, p<0.0001). Transparency was highest in June in the IE and lowest in November in the ME (ANOVA, p<0.0001) (Table 1).

Table 1: Averages of the environmental variables by season and station, estimated by least squares (± SD). The letters read vertically indicate significant differences (Tukey), in descending order, for each environmental variable with a significant two-way interaction (p≤0.05). Each average was calculated from three samples, for a total of 48 samples. RE = River Estuary. IE = Inner Estuary. OE = Outer Estuary. ME = Marine Estuary. Salt = Salinity, DO = Dissolved oxygen, TR = Transparency, %MO =Percentage of organic matter, %G = Percentage of gravel and very coarse sand, %SC = Percentage of coarse sand, %SM = Percentage of medium sand, % SF = Percentage of fine sand, and %A = Percentage of clay. 

Regarding sediments, the percentage of organic matter was highest in September at the OE station and lowest at the ME station during September and June (ANOVA, p<0.0001). The fraction of the coarse sediment was highest in September at the RE and the lowest in June at the IE (ANOVA, p<0.0001). The percentage of sand was highest in June at the ME station and lowest in September at the IE (ANOVA, p<0.0001).

Finally, the percentage of clay was highest in November at the ME and lowest at the same station but in September (ANOVA, p<0.0001) (Table 1).

In Buenaventura Bay, 48 trawls were carried out from April to November at four stations, and a total of 532 individuals, distributed in 17 species and nine families, were recorded. Abundance varied from 0.7+1.2 to 29.7±7.4 individuals per trawl, richness varied from 0.3+1.2 to 4.7+1.2 species per trawl, and the maximum Shannon diversity index was 1.6+0.3. By class, crustaceans predominated (91%), followed by mollusks (8%) and finally echinoderms (1%).

In the group of crustaceans, the most abundant family was Portunidae (n=383 ; 78%), followed by the families Penaeidae (n=53; 11%) and Squillidae (n=43; 9%). The most abundant species were Callinectes arcuatus, Squilla aculeata aculeata, and Litopenaeus occidentalis (Figure 2).

Figure 2 Macroinvertebrate structure and composition. 

In the group of mollusks, the most abundant family was Loliginidae (n=42; 96%), whereas the families Arcidae and Corbiculidae were poorly represented (n=1; 2%). Piangua (Anadara reinharti) from the family Arcidae was recorded, and from the family Corbiculidae, the species Polymesoda inflata was recorded. In the group of echinoderms, only the species Luidia columbia (n=3) of the family Luidiidae was captured.

The total abundance of macroinvertebrates was highest at the OE station (38.39%), followed by the RE station (28.12%), the ME station (23.72%), and the IE station (9.78%). By season, the peak of abundance occurred in June (39.36%), followed by April (26.89%), November (20.78%), and September (12.96%) (Figure 3). Four of the 17 species of macroinvertebrates were found at all four stations (Table 2), and the most abundant species were found in the outer zone of the bay (ME and OE). Of the remaining species, six were exclusive to the outer zone of the bay (ME and OE), and four were in the inner zone (RE and IE).

Figure 3 Box-and-whisker plot of the distribution of the relative abundance of macroinvertebrates over the sampling periods. 

Table 2: Scale of macroinvertebrate abundance, relative frequency (%F) at the sampling stations, size range (mm, total length for shrimp, starfish and hermit crabs and carapace length for crabs), and average relative abundance (ind/trawl, average of 48 samples) of macroinvertebrates collected from the RE, IE, OE, and ME stations. 

The structure and composition of macroinvertebrates did not vary by season (MANOVA, F=2.1, p=0.06) or by station (MANOVA, F=1.7, p=0.14). However, the period had a significant influence on the variation of abundance and richness of the macroinvertebrates across stations (MANOVA, F=3.9, p<0.0001). The highest abundance and richness were both observed in June at the RE station, whereas the lowest abundance and richness were observed in April at IE (Table 3) (ANOVA, p<0.0001, for both). No interaction was observed between the season and the station in terms of diversity.

Table 3: Average relative abundance of macroinvertebrate assemblages, species richness and diversity by Season and Station, estimated by least squares (±SD). The letters read vertically indicate significant differences (Tukey) for each descriptor of the community with a significant 2-way interaction (p≤0.05). Each average was calculated from 3 samples for a total of 48 samples. RE = River Estuary. IE = Inner Estuary. OE = Outer Estuary. ME = Marine Estuary. 

In the univariate multiple regression analysis, the abundance was greater with higher salinity of the water, whereas the richness was greater when salinity was high and the percentage of clay was low. Thirteen environmental variables were used to predict the relative abundance, richness, and diversity (Table 4). Two of the three models were significant (p<0.05). The best model explained 16% of the variation in the biological descriptors. Variables that were not collinear were included in the model, and the most notable variables had significant correlations. The Pearson correlation of salinity was significant for the abundance model (p=0.03). According to the Pearson correlations, salinity explained 31% of the variance of macroinvertebrate abundance and 23% of the variance in macroinvertebrate richness.

Table 4: Multiple regression analysis of the descriptors of the macroinvertebrate assembly in relation to environmental variables. The variables were reported in the order they entered the model; that is, variable 1 had the highest F value (p≤0.05). The type of relations between the biological and environmental variables were represented by signs, and the Pearson correlations were represented in parentheses. The level of significance to retain the variables in each model was p≤0.05, except for the variables in italics 0.15≤p≤0.05). Significant Pearson correlations are indicated in bold. 

DISCUSSION

The environmental conditions of the study were considered to be marine when the salinity, pH, dissolved oxygen, and transparency variables showed high values and when medium and fine sand predominated in the sediment. The environmental conditions of lower salinity were the opposite of the above-described conditions. This classification is based on several studies that describe the physicochemical characteristics of the water and sediments of the bay (^{Cantera and Blanco, 2001}; ^{Lucero et al., 2006}). According to these studies, the percentage of gravel and very coarse sand was high in the sediment when the salinity, pH, and transparency were low, which indicated the influence of fresh water discharges from the Dagua and Anchicayá rivers. Gravel and very coarse sand were the materials that predominated in the sediments of stations upstream of the mouth of the Dagua River (Lucero, et al., 2006). The low transparency was possibly due to the hydrodynamic conditions that allow the settlement of silt and clays. In general, the periods of greatest precipitation in the estuary and at the stations closest to the river drainage areas affected the salinity, pH, dissolved oxygen, and sediment fraction generating changes and variability in the physicochemical contributions of the estuarine microhabitats.

In this study, the taxonomic group that was most representative in the macroinvertebrate assembly were the crustaceans, represented by the crab C. arcuatus, which was the dominant species in the Buenaventura Bay estuary. This result agrees with macroinvertebrate inventories ^{Norse and Estevez, 1977}; ^{Neira and Cantera, 2005}; ^{Lazarus and Cantera, 2007}) and studies on populations of economic interest in the area (^{Pineda and Madrid, 1993}; ^{Baltazar, 1997}). The crabs are euryhaline organisms; that is, they have the physiological adaptation of tolerating a wide range of salinity (Norse and Estevez, 1977; ^{Hernández and Arreola-Lizárraga, 2007}); hence, their distribution throughout the estuary is unrestricted by salinity.

The second most abundant crustacean was Squilla aculeata aculeata. The abundance of S. aculeata did not have a clear trend because this species is a resident of the mudflats and sandy mudflat areas of the bay (^{Murillo, 1988}), and the females reportedly bury under the sediment (^{Wortham, 2009}), which could impede their capture in the sampling. In general, bravo shrimp were abundant in all seasons, whereas white shrimp (P. occidentalis), and Carabali shrimp (R. byrdi) were also collected during all seasons but were not as abundant. The remaining species were absent in one or more seasons.

In the group of mollusks, the bivalves Polymesoda and Anadara and cephalopods of the genus Lolliguncula were collected. Regarding the thimble squid Lolliguncula panamensis of Panama, individuals were only found in April at the OE station; in the Gulf of California, the species has been reported to move according to the food supply (^{Arizmendi-Rodríguez et al., 2012}), and studies have been conducted on their eating habits along the Colombian Pacific coast (^{Barragán, 1977}), which report that fish and crustaceans are the main source of food. The results of the present work, i.e., the greatest abundance of L. panamensis being observed in April in the OE, suggest that, during this period, the food supply at the OE favored the abundance of this species.

As for the group of echinoderms, only three individuals of the starfish Luidia columbia of the family Luidiidae were found, representing one of the seven families of echinoderms that inhabit soft bottoms in the Pacific (^{Neira and Cantera, 2005}). The poor representation agrees with that reported by other studies in the area that found a scarcity of these families in soft bottoms due to their preference for the hard bottoms such as found in rocky and coral reefs and rocky shores (Neira and Cantera, 2005).

Polychaetes, which are predominant among the soft-bottom macrofauna (^{Valencia et al., 2014}), were not collected in this investigation because they were not large enough to be captured by the trawl nets of 2.54 cm (one inch) mesh.

The relative abundance of macroinvertebrates was highest in June, lowest in September, and positively associated with salinity. In June, the abundance peak of the crab C. arcuatus was observed. In periods of highest salinity, the recruitment peak of crabs of the genus Callinectes occurs in the Caribbean (^{Valencia and Campos, 1996}). Regarding the species Penaeus occidentalis, the highest salinity was observed in June, coinciding with the presence of a few juveniles (n=5). The main pulse of juvenile migration from nursery areas to adult shrimp areas are known to occur between May and June (^{Pineda, 1992}). In summary, during June, the estuary presented high salinities that favored the presence and abundance of macroinvertebrates.

The species richness was highest in June, lowest in September, and weakly influenced by salinity and the percentage of clay. The regression model showed that species richness was greater with higher water salinity and lower percentage of clay in the sediment. In the highest rainfall months in the bay (September- November), the introduction of clays reportedly increases due to erosion and runoff (^{Cantera and Blanco, 2001}). In general, the results of the study suggest that, in June, conditions of higher salinity predominated in the estuary; consequently, a lag existed in the influx of fresh water, and marine species were favored, which increased the number of species in the estuary.

In this study, spatial trends were observed in the macroinvertebrate richness, which was higher in the outer zone of the bay, and these trends were characterized by marine conditions (OE and ME). This gradient was also reported in other studies in the Pacific because species such as hermit crabs (genus Clibanarius) and starfish (genus Luidia) were found (^{Norse and Estevez, 1977}) farther from the mouths of the rivers. Six species were exclusive to the outer zone compared to four that were exclusive to the inner bay. At the mouth of the bay, corresponding to the marine stations, macroinvertebrates with wide distributions have been reported (Portunidae, Penaeidae, Palaemonidae, Calappidae, and Lolliguncula) (^{Cantera and Blanco, 2001}); in addition, a mudflat exists where the Anchicaya River empties at the ME station. This interaction and supply of microhabitats favors macroinvertebrate richness because it is an ecotone between the open sea and the estuary, where there is a presence of marine species, estuarine species, and species from the mouth of the Anchicaya River.

CONCLUSIONS

In Buenaventura Bay, the relative abundance of epibenthic macroinvertebrates was higher at higher salinities, species richness was greater in higher salinity waters and in sediments with low clay content, and diversity was greater in sediments with high gravel content and high water pH. The richness and abundance of the macroinvertebrates were greatest in June and in the outer stations because, in this season and at these stations, the higher salinity conditions of the estuary prevailed, characterized by sediments with little organic matter and medium-to-fine sand-grain size.