Foraminiferal assemblages as bioindicators to assess potential pollution in mangroves used as a natural biofilter for shrimp farm effluents (New Caledonia)

Marine Pollution Bulletin 93 (2015) 103–120

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Foraminiferal assemblages as bioindicators to assess potential pollution in mangroves used as a natural biofilter for shrimp farm effluents (New Caledonia) J.-P. Debenay a, C. Marchand b,⇑, N. Molnar b,c, A. Aschenbroich b,c, T. Meziane c a b c

UMR 7159, IPSL/LOCEAN, Centre IRD France Nord, 32 Avenue Henri Varagnat, 93143 Bondy Cedex, France IRD, UR 206 – UMR 7590 IMPMC, F-98848 New Caledonia, France UMR BOREA MNHN-CNRS 7208-IRD-UPMC, Muséum National Histoire Naturelle, CP 53, 61 rue Buffon, 75231 Paris cedex 05, France

a r t i c l e

i n f o

Article history: Available online 7 March 2015 Keywords: Foraminifera Shrimp farming Effluents Mangrove Fatty acids SW Pacific

a b s t r a c t In New Caledonia, semi-intensive shrimp farms release untreated effluents into the mangrove. Foraminiferal assemblages were analyzed for assessing the impact of effluent release on the benthic compartment. Comparison was made between samples collected (1) in an effluent receiving mangrove before and after the rearing cycle, and (2) for one-year monitoring an effluent receiving and a control mangrove. The distribution of foraminiferal assemblages was primarily driven by the gradient between Rhizophora stands and salt-flats, related to salinity and tidal elevation, and by seasonal cycles. The potential impact of effluent release was due to the combined effects of normal-saline effluents on surface salinity, and of nutrient input and microbial stimulation on food availability. Foraminiferal assemblages did not indicate a substantial impact of farm effluents and suggest that semi-intensive shrimp farming using mangrove for effluent discharge may appear as a sustainable solution in New Caledonia, when considering only the impact on the mangrove itself. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction The increasing demand for shrimp in the developed countries led to an exponential expansion of shrimp farming, mainly in the subtropical and tropical lowlands. In addition to direct loss of saltmarsh and/or mangrove ecosystems for pond construction, shrimp aquaculture is increasingly criticized for a number of negative impacts (Martinez-Porchas and Martinez-Cordova, 2012). These impacts include the adverse effects of effluents on receiving ecosystems (review in Bui et al., 2012) and the introduction of alien species, either intentionally imported or accidentally introduced. A large portion of nutrients added to a shrimp pond as feed is not converted to shrimp biomass but can be exported from the pond system as particulate and dissolved nutrients, where it can be responsible for excess primary productivity, and even harmful algal blooms (Jackson et al., 2003; Costanzo et al., 2004; CasillasHernández et al., 2007; Miranda-Baeza et al., 2007; Hasani et al., 2012; Bui et al., 2013). Shrimp pond effluents, high in organic matters, also have a high biological oxygen demand and can cause oxygen depletion in receiving waters (EJF, 2003), where eutrophication ⇑ Corresponding author. E-mail address: [email protected] (C. Marchand). http://dx.doi.org/10.1016/j.marpolbul.2015.02.009 0025-326X/Ó 2015 Elsevier Ltd. All rights reserved.

is often cited as a major concern (Páez-Osuna et al., 1998; Mckinnon et al., 2002; Bui et al., 2012; Herbeck et al., 2013). Thus, spatial and temporal assessment of coastal aquatic environments in shrimp farming areas is essential for protecting estuarine and marine ecosystems and promoting a sustainable economic development (Boyd and Green, 2002). The use of natural mangroves as biofilters for shrimp pond effluents is considered as an efficient tool for reducing the impact of shrimp farming (Páez-Osuna, 2001; Gautier, 2002). They remove significant percentages of total suspended solids (Gautier et al., 2001), and several authors have reported their effectiveness in removing nutrients and pollutants from effluents (e.g. Tam and Wong, 1999; Wang et al., 2010; Zaldívar-Jiménez et al., 2012). Constructed on salt-flats, shrimp ponds are not responsible for mangrove deforestation in New Caledonia, but induce topographical and hydrological transformations that may cause changes in faunal composition at all levels (meiofauna, macrofauna, megafauna) (Virly et al., 2005). Semi-intensive shrimp farming of the blue shrimp Litopenaeus stylirostris typically uses a flowthrough system with water exchange rates as a tool to maintain optimum hydrological and biological parameters for the crop (Della Patrona and Brun, 2008; Thomas et al., 2010). Water supplies consist of open-sea water that keeps salinity of the ponds

104

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

between 32‰ and 39‰. Water renewal varies from 0% to 30% per day, depending on shrimp biomass. Untreated effluents, mostly composed of normal saline water, are released into the mangrove that has been demonstrated to be only a partial filter for the shrimp farm effluent leaving part of the nutrient loads exported to the adjacent bay (Molnar et al., 2013). In addition to nutrient enrichment, the release of effluent within the Avicennia stand, by modifying the length of water-logging, modifies the redox conditions as well as the salinity of pore-waters, and thus modify the ecological conditions of the mangrove benthic compartment (Marchand et al., 2011a; Molnar et al., 2014). Because of their short life and reproductive cycles, high biodiversity and specific ecological requirements, benthic foraminifera are particularly sensitive to changing environment. Density, diversity and composition of the communities may change rapidly in response to changes in environmental conditions, making them valuable environmental bioindicators of environmental stress, both natural and anthropogenic (e.g., review in Frontalini and Coccioni, 2008; Carnahan et al., 2009). Despite the difficulty in deconvoluting the impact of pollution from natural stress (Armynot du Châtelet and Debenay, 2010), a number of studies conducted since the 1960s has demonstrated their usefulness as proxies for coastal monitoring (reviews in Nigam et al., 2006 and Pati and Patra, 2012). Foraminiferal assemblages from low-latitude intertidal mangrove swamps have been studied for long in various regions (reviews in Debenay et al., 2004 and Woodroffe et al., 2005). They can show an intertidal zonation (e.g. Hayward et al., 1999; Horton et al., 2003; Woodroffe et al., 2005), but major factors independent from elevation, such as their sensitivity to spatial (horizontal) or temporal changes in salinity may have a greater influence on their distribution trends (e.g., Debenay and Guiral, 2006; Culver et al., 2012). Only a few studies have been carried out on benthic foraminifera specifically related to aquaculture (e.g., Schafer et al., 1995; Scott et al., 1995; Angel et al., 2000; Bouchet et al., 2007; Vidovic´ et al., 2009), and even less related to the impact of shrimp farms (e.g., Luan and Debenay, 2005; Souza et al., 2010). Recently, studies carried out in shrimp ponds from New Caledonia have shown that they are rapidly colonized by a few foraminiferal species, but that the individuals growing in the ponds may be strongly deformed, indicating adverse environmental conditions mostly due to the accumulation of easily oxidized material (Debenay et al., 2009a,b). The aim of this study was to investigate the impact of shrimp farm effluent discharge on the benthic compartment of a mangrove using foraminiferal assemblages as bioindicators. Firstly, samples were collected throughout the different mangrove stands after a period of several months without effluent discharge and after a period of several months of rearing; assemblages of the two periods were compared. The results were compared to geochemical analyses (Chl-a content, fatty acids distribution, carbon and nitrogen stable isotopes) carried out on the same samples in previous studies (Molnar et al., 2014; Aschenbroich et al., 2015). Secondly, changes in the assemblages from an Avicennia stand was monitored over a one-year period, during and after the rearing period, in the impacted mangrove and in a neighboring control mangrove. The Avicennia stand was chosen because it is the closest stand from the salt-flat where the ponds are constructed.

2. Material and methods

season is from January to March, while a cooler, drier time occurs from May to October, with the winter season in July to August. However, the dates and duration of each season, as well as rainfall intensity may considerably vary from one year to another. Yearly rainfall on the leeward southwest coast, where the study was carried out, varies from 800 to 1200 mm. During our study, rainfalls were particularly intense at the end of the rainy season 2009 (February–March), and in July, but were particularly low during the rainy season of 2010 (Fig. 1). The tidal regime is semi diurnal, unequal, and low mesotidal (maximum tide value is 1.6 m) (Douillet, 1998). During the study period, the levels of both high tides and low tides were the lowest during the dry season (May– October), leading to a downward shift of about 15 cm of the intertidal zone (Fig. 1). Owing to the very flat slope of the mangrove, this vertical shift may produce significant horizontal offsets resulting in differences in the duration of tidal immersion. Extensive mangroves are fringing 88% of the low western coastline of the main island, sheltered from the easterlies winds. They cover 35,100 ha, including 9200 ha of salt marshes and salt-flats (Virly, 2008), and exhibit the typical zonation of mangroves in semi-arid conditions with Rhizophora spp. in the low intertidal zone, Avicennia spp. in the intertidal zone, Sarcocornia quinqueflora on the salt marshes, and bare salt-flats at the higher tidal elevation. The Rhizophora and Avicennia stands are subjected to tidal cycles whereas salt-flats are more irregularly immersed, which leads to higher salinity levels. The study was conducted in two bays located on the west coast of New Caledonia, both occupied by a mangrove forest and with insignificant freshwater input (Fig. 2). The first one (A) of about 20 ha in surface area and virtually free of anthropogenic influences was used as a control area (Fig. 2A). The second one (B) of about 29 ha in surface area receives the effluent discharges from the FAO shrimp farm (Ferme Aquacole de la Ouenghi). The FAO shrimp farm opened in 1989. It operates two 1 m deep rearing ponds of 10.5 ha (L) and 7.5 ha (K) respectively. The rearing cycle lasts about eight months (December–July), and ends during the cool season in order to prevent diseases and bacterial growth. At the end of the cycle, the ponds are partially drained during each of the 2–12 partial harvests. After the last harvest, they are completely drained out and left to dry for several weeks. Untreated effluents are discharged at multiple points, either released directly into the mangrove or collected in the effluent channel, often spilling out over the channel banks at high tide (Fig. 2B and C). A levee partially separates the inner bay, which receives shrimp effluents, from the outer bay, which opens to Saint Vincent bay. Five reference stations were selected in the Avicennia zone of each mangrove A (control) and B (impacted) for a one-year monitoring. They were sampled eight times during the 2009 rearing cycle, when both ponds L and K were in production (Figs. 1 and 2). In addition, surface sediment samples were collected at low tide, at 51 locations in the whole mangrove B, for a dual-season mapping of foraminiferal assemblages. Samples were collected before the beginning of a rearing cycle in December 2009 (Non Active Period – NAP), and at the end of a rearing cycle in July 2010 (Active Period – AP) (Fig. 2C). Due to a shortage of shrimp larvae, only pond K was in production during AP. The sites were selected following a random sampling procedure, but because of the high density of trees and aerial roots, several sites were not accessible and the use of a systematic sampling approach (Caeiro et al., 2003) was therefore not always possible. Sampling points were recorded using a handheld GPS (Colorado 300, Garmin).

2.1. Study site and sampling 2.2. Methods New Caledonia (south-west Pacific) typically experiences easterly to southeasterly winds, with a tropical humid climate on the east coast, and a semi-arid climate on the west coast. The hot rainy

Each sample for foraminiferal analyses was composed of surface sediments collected randomly over a 1 m2 surface area. The upper

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

105

Fig. 1. Time scale of sampling collections and rearing operations; tides, temperature and rainfall at Nouméa during the period of study.

five millimeters of sediment were scraped with a stainless steel spoon, and placed into a plastic bag. This approach that was referred to as a pseudoreplication procedure (Debenay et al., 2002; Morvan et al., 2006) may more properly be likened to sample pooling since each spoonful can be considered as a subsample, and all subsamples are pooled into one homogeneous sample (see chapter 2.3 below). The samples were preserved in 70% ethanol and kept at least 3 days in the Rose Bengal stain at a concentration of 2 g l 1 (Walton, 1952; Murray and Bowser, 2000). A subsample of 50 cm3 was extracted from each sample after homogenization and washed through 350 and 63 lm sieves to remove coarse material, mud, alcohol, and excess stain. Foraminifera from the 63– 350 lm fraction were separated by flotation on ethylene trichlorure. When available, about 300 individuals were identified and counted, distinction been made between stained (usually considered as living) and unstained specimens. In case of abundant material, the sample was split, and the results were extrapolated for the whole sample. Standardized abundances (density) were expressed as total number of specimens per 50 cm3 of sediment. The relative abundance of each species (proportion) was calculated, as well as the proportion of abnormal tests. The data set for statistical analyses of the dual-season mapping was restricted to the eighteen species that occurred in at least 10

samples and reached a relative abundance of 5% in at least one sample at each period. No transformation or standardization of the data was done. A hierarchical clustering analysis using Ward’s linkage method and Euclidean distances (Ward, 1963) was carried out on species relative abundance in all samples. The data were also explored in relation with geochemical parameters from previous studies, using a canonical correspondence analysis (CCA) (ter Braak, 1986; ter Braak and Verdonschot, 1995). Owing to the availability of geochemical data, the CCA was carried out on 31 samples for the first sampling period, and 28 for the second period. As this study focused on the impact of the shrimp farm, we investigated the relationships between species relative abundance and parameters related to farm activity, at the expense of the usually considered natural variables. The selected parameters were C:N that indicates the origin (continental-marine) and the relative degradation state (fresh-decaying) of organic matter, d13C and d15N that indicate the type of vegetation from which originated the organic matter, Chl-a and phaeopigments markers of primary production, and fatty acids (FAs) used as biomarkers (Table 1). FAs, considered relevant for the present study, on the basis of Aschenbroich et al. (2015), were a bacterial marker (18:1Á7), a fungial marker (18:2x6), a diatoms marker (20:5x3), the sum of the FAs 18:3x6 + 20:3x6 that revealed the presence of another,

106

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Fig. 2. Location map; (A) map of mangrove A (virtually free of anthropogenic influences); (B) map of mangrove B (receiving effluents of FAO shrimp farm, and location of sampling areas used for the one-year monitoring); (C) localization of the discharges areas in mangrove B, and location of the sampling stations used for the dual-season mapping.

unidentified, microalgal community, and the FA 18:1x9. This FA has been considered as a good effluent pathway tracer during farm activity (Aschenbroich et al., 2015) but most probably results from the presence of fungi able to synthesize it in small quantity during the non active period (Chen et al., 2001). The multivariate analysis package Ginkgo was used for statistical analyses (http://biodiver. bio.ub.es/vegana/index.html; De Cáceres et al., 2003). 3. Results Including mapping and monitoring sampling stations, 173 samples were collected, which yielded a total of 53 foraminiferal taxa (Tables A1–A3). All species are illustrated in Debenay (2012). Forty-five taxa were recorded as stained (living). The dominant species over the whole area were Ammonia tepida and Quinqueloculina seminula. The other abundant species were the hyaline calcareous forms Elphidium excavatum, Gyroidina lamarckiana, Helenina anderseni, and the agglutinated forms Monotalea salsa, Trilocularena patensis and Trochammina inflata. The proportion of abnormal tests never exceeded 2% of the assemblage. 3.1. Dual-season mapping 3.1.1. General characteristics of the assemblages In December 2009 (Non Active Period – NAP), densities of total and stained assemblages were higher around the inner bay, mostly

over the Avicennia zone, and in the outer bay (Fig. 3). The proportion of living specimens followed the same trend. Species richness was higher in the center of the inner bay, in the Rizophora zone, and in the outer bay. In all cases, values were lower at the transition between the inner and outer bays, in front of the levee. In July 2010 (Active Period – AP), the same general trends were seen, the most significant differences being a general decrease in density with higher values in the salt-flat and Sarcocornia zones, along the shrimp ponds, instead of being in the Avicennia. The species richness tended to increase, mostly by the addition of agglutinated species (Tables A1 and A2), and the area of higher species richness extended over the whole Rhizophora zone, except at the transition between the inner and outer bays. 3.1.2. Distribution of the main species The dominant species A. tepida and Q. seminula made up as much as 84% and 86% of the total assemblage, respectively (Tables A1 and A2). In NAP samples, A. tepida exceeded 25% of the assemblage over all the area except 8 samples, with higher abundances in the Rizophora zone (Fig. 4). Conversely, Q. seminula was less abundant in the Rhizophora zone, becoming dominant at the periphery of the bay in the Sarcocornia zone and on the saltflats. The same general trends were found in AP samples, but the abundance of A. tepida increased in the areas of effluent discharge and the abundance of Q. seminula showed a general slight decrease. Among the most abundant species, three showed a noticeable

Table 1 Geochemical data from surface sediment used for statistical analyses (‘‘–’’ indicates undetectable concentration of FAs). Stations

2

5

7

9

10

12

13

14

16

18

20

21

22

23

27

Chl a (lg g 1) Phaeo (lg g 1) d13C‰ d15N‰ C/N 20:5x3 (lg g 1) 18:2x6 (lg g 1) 18:1x9 (lg g 1) 18:1x7 (lg g 1) 18:3x6 + 20:3x6 (lg g 1)

8.44 2.75 23.01 4.54 12.00 – – – 0.80 2.76

9.81 6.19 21.11 3.47 11.29 – – – 1.20 10.70

5.50 2.97 14.23 4.76 10.98 – – – 0.42 –

11.41 9.99 25.50 3.31 14.51 – – 0.41 1.37 6.01

14.42 24.15 24.78 4.22 14.33 0.55 – 2.26 3.14 8.16

22.67 20.49 22.52 4.70 11.96 – – – 0.91 17.30

27.92 12.97 26.16 2.11 22.57 – – 0.65 1.70 10.50

14.84 9.99 25.87 1.73 17.95 – – – 0.48 4.58

22.19 23.12 25.03 4.48 12.75 – – 1.51 2.36 6.21

21.95 25.30 24.32 4.63 14.03 – – – 0.57 3.82

4.20 1.61 14.39 5.97 9.97 – – – 0.47 0.10

12.06 13.21 26.49 2.95 16.60 – – 0.85 1.47 4.38

27.12 23.66 25.19 4.78 15.31 – – – 1.49 11.90

9.44 17.43 26.10 4.37 15.44 0.25 – 0.57 1.07 3.06

31.06 42.21 26.20 4.01 14.35 18.90 13.59 22.61 35.24 7.83

14.49 18.99 28.47 3.26 30.25 3.73 4.67 14.00 40.01 3.68

June 2010 (Active Period – AP)

Chl a (lg g 1) Phaeo (lg g 1) d13C ‰ d15N ‰ C/N 20:5x3 (lg g 1) 18:2x6 (lg g 1) 18:1x9 (lg g 1) 18:1x7 (lg g 1) 18:3x6 + 20:3x6 (lg g 1)

9.47 2.82 22.87 3.66 8.78 14.10 3.67 20.74 28.20 1.61

2.92 1.07 22.05 3.62 9.06 5.61 1.72 12.45 11.96 0.58

7.37 1.54 22.27 2.77 8.95 8.15 1.87 14.35 16.25 0.62

7.80 5.39 25.72 3.49 14.52 26.20 18.68 42.62 79.22 13.40

23.95 6.60 25.74 2.85 12.98 24.80 15.63 37.82 73.46 4.19

10.27 7.00 25.56 3.61 12.43 21.50 17.70 55.04 70.13 5.13

16.29 4.05 26.64 2.07 19.40 20.40 31.39 37.87 57.51 3.17

6.21 2.18 25.38 2.38 17.42 17.10 6.37 18.44 28.47 2.66

5.72 3.05 24.42 2.96 11.99 11.10 3.47 17.78 17.94 1.45

41.56 16.80 25.91 3.12 13.81 33.30 24.68 50.85 109.48 6.35

11.62 10.13 25.55 3.64 12.23 25.90 13.32 32.75 49.35 5.02

14.12 6.99 26.92 2.67 16.22 20.50 21.20 34.35 64.69 3.86

13.35 5.71 25.73 4.02 14.52 15.90 15.65 33.91 51.93 2.37

7.85 12.88 26.58 3.79 14.23 6.25 12.27 26.61 51.41 1.59

34.36 13.30 27.64 2.93 17.00 27.00 29.15 47.99 132.91 7.27

4.87 9.04 28.48 2.83 21.67 6.47 16.12 35.41 118.11 3.50

28

30

31

32

33

34

36

37

39

41

42

43

45

48

49

50

December 2009 (Non Active Period – NAP)

Chl a (lg g 1) Phaeo (lg g 1) d13C ‰ d15N ‰ C/N 20:5x3 (lg g 1) 18:2x6 (lg g 1) 18:1x9 (lg g 1) 18:1x7 (lg g 1) 18:3x6 + 20:3x6 (lg g 1)

7.74 20.02 26.95 3.10 17.97 1.84 2.05 9.24 23.34 3.68

8.18 16.48 27.51 3.95 22.18 3.09 6.94 15.12 36.22 2.89

15.34 17.06 23.21 2.41 12.37 5.43 5.11 15.03 43.15 3.77

22.14 17.06 26.87 3.18 27.32 8.95 3.29 11.87 34.61 4.82

23.08 21.90 26.74 3.32 21.45 8.36 4.72 12.97 27.44 4.85

24.16 33.04 27.62 3.30 17.69 10.20 8.81 19.83 40.97 4.45

34.10 17.66 27.08 2.64 18.57 12.70 6.50 17.10 33.52 7.86

19.46 14.86 26.81 3.04 20.62 7.74 3.80 11.69 27.29 5.12

7.95 15.07 27.49 3.41 23.30 1.51 4.65 11.47 27.74 4.06

14.11 16.82 27.86 2.79 20.85 8.44 3.81 12.66 32.30 5.69

22.29 13.72 26.35 3.46 20.68 2.28 1.31 5.08 8.30 11.00

16.41 9.10 25.16 3.03 18.79 – – 1.49 2.15 13.90

8.83 8.29 25.25 3.39 17.07 – – 0.30 0.66 6.07

19.70 11.63 26.50 2.48 19.45 0.49 0.65 2.70 4.89 4.93

22.18 12.85 25.76 2.56 16.83 – – 1.20 2.00 2.61

June 2010 (active period – AP)

Chl a (lg g 1) Phaeo (lg g 1) d13C ‰ d15N ‰ C/N 20:5x3 (lg g 1) 18:2x6 (lg g 1) 18:1x9 (lg g 1) 18:1x7 (lg g 1) 18:3x6 + 20:3x6 (lg g 1)

7.46 7.33 28.13 2.92 21.24 7.24 11.30 27.08 75.34 2.97

4.70 5.41 27.93 3.19 21.41 6.28 19.84 37.57 99.64 2.52

21.51 8.17 27.29 3.06 22.69 29.70 12.10 39.87 113.53 2.85

14.76 7.82 27.21 3.29 21.13 16.00 29.41 40.33 82.45 3.00

15.11 13.02 28.17 2.76 20.78 10.20 54.28 76.90 160.51 2.72

25.07 7.57 27.62 2.23 20.94 20.30 16.49 28.48 68.95 3.79

11.28 3.20 27.25 2.92 21.20 17.40 10.62 20.44 39.37 2.36

13.31 3.67 25.51 3.08 18.95 13.90 5.24 17.32 29.82 1.87

15.31 3.04 25.72 2.94 14.46 12.70 6.83 13.66 14.90 1.63

7.97 3.40 26.75 2.67 22.63 8.32 16.10 27.21 49.32 1.96

8.70 4.42 26.52 2.59 18.81 9.51 14.27 18.38 35.01 2.60

4.07 2.90 27.19 2.26 21.79 6.55 12.65 21.98 58.21 1.49

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

1

December 2009 (Non Active Period – NAP)

107

108

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Fig. 3. General characteristics of the assemblages in December 2009 (Non Active Period – NAP) and in July 2010 (Active Period – AP).

change between NAP and AP samples. The abundance of T. patensis, which lived in the inner bay, increased; the abundance of H. anderseni, mostly living in the Rhizophora zone, decreased; Caronia exilis, which live in the Rhizophora zone, significantly extended its area within the inner bay (Fig. 5). The other species showed little changes in their distribution between NAP and AP samples. Glomospira gordialis was mostly found in the Rhizophora zone of

the outer bay; G. lamarckiana was found in the Rhizophora and Avicennia zones, except at the transition between the inner and the outer bay; M. salsa was widely distributed in the Rhizophora and Avicennia zones (Fig. 5). Bolivina striatula was found in the outer part of the bay, under marine influence, with a very small extension of its area in AP samples; T. inflata was widely distributed over all the vegetation zones, also with a very small extension of

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

109

Fig. 4. Distribution maps of Ammonia tepida and Quinqueloculina seminula in December 2009 (Non Active Period – NAP), and in July 2010 (Active Period – AP).

its area in AP samples; E. excavatum was widely distributed, except at the transition between the inner and the outer bay, and its abundance slightly decreased in AP samples (Fig. 5). 3.1.3. Cluster analyses 3.1.3.1. Non active period samples. The Q-mode cluster analysis identifies four clusters (Fig. 6). The first Cluster (I) groups the samples from the periphery of the inner bay, including the salt-flats, and the Avicennia and Sarcocornia zones (Fig. 7; Table 2). The dominant species is Q. seminula, associated with A. tepida and E. excavatum. Cluster II is composed of samples from Rhizophora and mixed Rhizophora/Avicennia zones. The dominant species is A. tepida associated with Q. seminula, E. excavatum, H. anderseni and G. lamarckiana. Cluster III groups samples from Rhizophora zones where A. tepida is strongly dominant, associated with G. gordialis. The last Cluster (IV) also groups samples from Rhizophora zones. The assemblage is dominated by A. tepida, associated with H. anderseni and G. lamarckiana. The main characteristics of the assemblages are displayed in Table 3. Average densities are the highest in Cluster I while the species richness is the lowest in this cluster. The lowest densities are in Cluster III, and the highest species richness is in Cluster IV. 3.1.3.2. Active period samples. The Q-mode cluster analysis identifies four clusters (Fig. 8). Cluster (I) comprises only three samples from Avicennia and Sarcocornia zones located close to the output channel (Fig. 7; Table 2). Q. seminula is strongly dominant, associated with A. tepida. The four samples of Cluster II are from Rhizophora and mixed Rhizophora/Avicennia zones. Located around the inner bay, they suggest that assemblages from the center of the bay may be associated with this cluster. The dominant species is T. patensis associated with A. tepida, Q. seminula, and several agglutinated species. Cluster III groups samples from Rhizophora zones where A. tepida is dominant, associated with Q. seminula, G.

lamarckiana, E. excavatum and several agglutinated species. The last Cluster (IV) groups samples from almost all possible environments. The assemblage is dominated by A. tepida, associated with Q. seminula. The main characteristics of the assemblages are displayed in Table 3. Average densities are the highest in Cluster I while the species richness is the lowest in this cluster. The lowest densities are in Cluster II, and the highest species richness is in Cluster III.

3.1.4. Canonical correspondence analysis 3.1.4.1. Non active period samples. The two first ordination axes of the CCA explain 29% of the total variance in the foraminiferal data (axis 1 = 20%, axis 2 = 9%) (Fig. 9). As indicated by the arrows, the bacteria and fungi markers (FAs 18:1x7, 18:2x6 and 18:1x9), and C:N are positively correlated with axis 1. Conversely, d13C and d15N are negatively correlated with this axis. Photosynthetic pigments (Phaeopigments and Chl-a) are negatively correlated with axis 2, as well as the microalgal markers 20:5x3, 18:3x6 and 20:3x6. In addition, the diatom marker 20:5x3 is positively correlated with axis 1, whereas other microalgae (18:3x6 and 20:3x6) are negatively correlated with this axis (Fig. 9B). Except T. patensis all the agglutinated species are correlated with fungi markers (18:2x6 and 18:1x9) and C:N (axis 1positive). Q. seminula is strongly correlated with d13C and d15N (axis 1 negative). T. patensis is negatively correlated with phaeopigments, Chl-a and microalgae (axis 2 positive), and Elphidium williamsoni is positively correlated with these variables (axis 2 negative). Several species are positively correlated with both axes (G. gordialis, C. exilis and T. inflata), showing a positive correlation with C:N and bacteria, and a negative correlation with photosynthetic pigments and microalgae. Others are positively correlated with axis 1 and negatively with axis 2 (H. anderseni, Paratrochammina sp. and M. salsa), showing a correlation with diatoms. Discorinopsis agayoi is positively correlated with d13C and d15N (axis 1 negative)

110

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Fig. 5. Distribution maps of Trilocularena patensis, Helenina anderseni, Caronia exilis Glomospira gordialis, Gyroidina lamarckiana, Monotalea salsa, Bolivina striatula, Trochammina inflata and Elphidium excavatum in December 2009 (Non Active Period – NAP), and in July 2010 (Active Period – AP).

and negatively with photosynthetic pigments and microalgae (axis 2 positive). The other species are less strongly correlated with the two axes, most of them negatively with axis 2. On the plot of the sampling stations (Fig. 9C), the clusters identified by hierarchical analysis are clearly separated from each other. Cluster I plots negatively with axis one, showing its correlation with d13C and d15N. Clusters III and IV plot positively on axis 1,

Cluster III also being negatively correlated with photosynthetic pigments and microalgae and Cluster IV being positively correlated with these variables. Plots of samples of Cluster II are widely skewed along axis one. 3.1.4.2. Active period samples. The two first axes of the CCA explain 27% of the total variance in the foraminiferal data (axis 1 = 15%,

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

111

axis 2 = 12%). Most of the arrows that represent explanatory variables are widely angled with the ordination axes, indicating a moderate to weak influence of these variables on foraminiferal assemblages, and a weak correlation of the variables with each other (Fig. 10). The first axis is dominated by phaeopigments, bacteria (18:1x7), fungi (18:2x6), and the FA 18:1x9 that mostly represents effluent discharge during AP (Aschenbroich et al., 2015). This FA, however, has lesser influence on foraminiferal assemblages during AP than during NAP, as shown by the shorter arrow. The second axis shows an opposition between C:N towards negative values, and d13C, d15N, Chl-a and microalgae (FAs 20:5x3, 18:3x6 and 20:3x6) towards positive values. Diatoms (FA 20:5x3) that plotted opposite to d13C and d15N during NAP were correlated with these variables during AP. Except T. patensis, agglutinated species, are positively correlated with C:N (axis 2 negative) and with the bacteria and fungi markers (axis 1 positive). The calcareous species B. striatula, G. lamarckiana and H. anderseni also correlate with C:N. Q. seminula and E. excavatum correlate positively with d13C and diatoms (20:5x3). G. lamarckiana and E. williamsoni strongly correlate with d15N and Chl-a, E. williamsoni positively and G. lamarckiana negatively. Cornuspira planorbis and D. aguayoi negatively correlate with phaeopigments, bacteria and fungi (axis 1 negative). The opportunistic A. tepida is near the center of the plane, weakly influenced by the environmental variables considered in this analysis. Sampling stations of Clusters I plot negatively with axis 1 and positively with axis 2, being positively correlated with d13C and diatoms. Cluster II plot positively with axis 1, and then with phaeopigments, bacteria and fungi. Clusters III and IV overlap and are located near the center of the ordination plane, suggesting intermediate conditions with respect to the environmental variables. 3.2. One-year monitoring within the Avicennia stand

Fig. 6. Q-mode cluster dendogram (above) and R-mode cluster dendogram (left) based on relative abundances of foraminiferal assemblages from December 2009 (Non Active Period – NAP). The Q-mode dendogram identifies four Clusters (I, II, III, IV). Relative abundance of each species in each sample is summarized in the chart.

A total of twenty-two species were collected over the ten stations (five in the impacted Avicennia stand and five in the control Avicennia stand) during the period of monitoring, including 13 agglutinated, 7 hyaline and 2 porcelaneous species (Table A3). The most frequent and abundant species were A. tepida, Q. seminula, T. patensis and T. inflata. The species richness of each sample ranged from 3 to 16, and the density (number of tests in 50 cm3) from 10 to 80,000, minimum values being from the control mangrove and maximum values from the impacted mangrove. Foraminiferal assemblages show major changes between June and August, at the end of the rearing period, in both the control and impacted mangroves. During the period of maximum effluent discharge (April–June 2009), species richness was high and density was low (Fig. 11). After June 2009, a general decrease of species richness was recorded in both mangroves while the density increased at all stations, with a great disparity between stations. This trend, however, was somewhat weaker in the control mangrove, where the density was much lower at the beginning of the monitoring. The proportion of agglutinated species, which was greater in the control mangrove, declined in the two mangroves after June 2009. The abundance of the most frequent and abundant species evidences the great heterogeneity between stations in space and time (Fig. 12), but also reveals temporal trends. In the impacted mangrove, the absolute abundance of A. tepida slightly increased at some stations after the last harvest and the end of effluent discharge, and the heterogeneity between stations increased. Two stations had very high abundance in February 2010. In the control mangrove, the absolute abundance of A. tepida clearly increased at all stations after the last harvest. The absolute abundance of Q.

112

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Fig. 7. Distribution of the four clusters identified by cluster analyses during December 2009 (NAP), and July 2010 (AP).

Table 2 Vegetation and foraminiferal species characteristic of each cluster. Dominant species

Associated species

Mapping of December 2009 (Non Active Period) Cluster I Saltpans, Sarcocornia, Avicennia Cluster II Rhizophora or mixed Rhizophora/Avicennia Cluster III Rhizophora Cluster IV Rhizophora

Dominant vegetation

Q. seminula A. tepida A. tepida (strong dominance) A. tepida, H. anderseni

A. tepida, E. excavatum Q. seminula, E. excavatum, H. anderseni, G. lamarckiana G. gordialis G. lamarckiana

Mapping of July 2010 (Active Period) Cluster I Sarcocornia, Avicennia Cluster II Rhizophora or mixed Rhizophora/Avicennia Cluster III Rhizophora Cluster IV Almost all possible environments

Q. seminula (strong dominance) T. patensis A. tepida A. tepida

A. tepida A. tepida, Q. seminula and agglutinated species Q. seminula, G. lamarckiana, E. excavatum and agglutinated species Q. seminula

seminula increased sharply after the last harvest, in both the impacted and control mangroves. Differences between the two mangroves are: a greater heterogeneity in the impacted mangrove; the quasi absence of Q. seminula during the rearing period in the control mangrove, which corresponded to the period of seasonal temperature decrease in New Caledonia, while it was present in the impacted mangrove, making up more than 20% of the assemblage at most of the stations in May and June, end of the rearing period. Despite the great heterogeneity, Fig. 12 shows that T. patensis and T. inflata were more abundant during the rearing period in both mangroves. This feature is even more obvious when considering relative abundance (Table A3). T. patensis was better represented in the control mangrove and the difference in its abundance between the two periods was stronger than for T. inflata, which was better represented in the impacted mangrove. Other species remarkable for their distribution were E. williamsoni, well represented at the impacted stations, with higher relative abundance from June to October, but almost absent from the control stations; A. salsum, better represented in the control mangrove with a relative abundance decreasing with time; E. excavatum, roughly equally represented in both mangroves, with a relative abundance that increased with time. 4. Discussion 4.1. Sampling foraminiferal assemblages in mangrove swamps Intertidal sediments of mangrove swamps are extremely heterogeneous due to: litter accumulation; the existence of hard substrata (pneumatophores, roots and tree trunks) interspersed in the soft substratum (Beck, 2000); presence of roots and bioturbation by burrower crabs (up to 40 burrows m 2 in the Rhizophora stand) that induces/maintains biogeochemical heterogeneity in sediments (Kristensen, 2008); and the existence of depressions in

the sediment, with considerably different physico-chemical conditions (Olafsson et al., 2000). This heterogeneity promotes small-scale (cm) patchiness in the distribution of the meiobenthos, and particularly of foraminiferal assemblages, which complicates the observations (e.g., Boltovskoy and Lena, 1969; Buzas, 1970), and raised the key problem concerning spatial scale: which sample size is most appropriate (Lörz et al., 1998). Results of large-scale surveys can be biased if sampling design is not adapted to such small-scale patchiness (Lang, 1989). One mean to overcome the bias introduced by patchiness should be to use replicates. However, such heterogeneous environments require a special sampling effort to overcome the ineluctable heterogeneity of the collected data (Parravicini et al., 2010), and the use of three or even two replicate samples, often recommended to obtain reliable information on foraminiferal assemblages (e.g., Hayek and Buzas, 1997; Murray and Alve, 2000; Schönfeld et al., 2012), would be insufficient. It may be stated that a dispersed sampling effort with many small replicate samples will characterize a local assemblage better than a few large samples (Kahlert, 2001; Bennington, 2003), but the level of replication required may be prohibitively high (Findlay, 1981; Ware and Kenny, 2011). Some authors consider that the distribution of meiofauna might be easier to recognize when collecting sample over a larger area that covers sample patchiness (Winkelmann and Ziemer, 1998), while, even repeated, smaller samples are not as representative (Lörz et al., 1998). The pooling method used for this study was based on the studies mentioned above and on more than 30 years of field experience of highly patchy mangroves environments. If one wants to collect two replicates, the question will be: where? One at the bottom of a depression, the other between aerial roots? For one replicate more, would it be collected near a crab burrow? Collecting replicates in such an environment is likely to introduce substantial bias. A pooling sampling strategy encompasses this variability (Baker et al., 2009). It is clear from previous studies that it provides a satisfactory image of the assemblages. Even if studies based on other

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

113

biological groups have demonstrated that especially rare taxa are not well represented in pooled samples (Ohman and Lavaniegos, 2002), it may be sufficient, when targeting the most abundant taxa, to analyze one large and homogenized sample instead of numerous individual biological replicates (Engel et al., 2012). 4.2. Individual behavior of foraminiferal species

Fig. 8. Q-mode cluster dendogram (above) and R-mode cluster dendogram (left) based on relative abundances of foraminiferal assemblages from July 2010 (Active Period – AP). The Q-mode dendogram identifies four Clusters (I, II, III, IV). Relative abundance of each species in each sample is summarized in the chart.

The dominant species A. tepida and Q. seminula collected in the Avicennia zone of the two monitoring areas are cosmopolitan species that may be found in a variety of restricted environments (review in Debenay et al., 2000), sometimes in very high proportion (e.g., Abu-Zied et al., 2007). A. tepida tolerates hyposaline to hypersaline, organic-rich sediments (e.g. Almogi-Labin et al., 1992; review in Murray, 2006; Wennrich et al., 2007), and is very tolerant to pollution (review in Frontalini and Coccioni, 2008). It has been reported in mangrove swamps all over the world (review in Debenay et al., 2004). Widely distributed in lagoons and estuaries, Q. seminula is generally better represented in slightly hypersaline environments, and may be abundant on hypersaline saltflats (Debenay and Guillou, 2002; Murray, 2006). Species of the Elphidium (Cribroelphidium) genus have also been widely reported from mangrove swamps. E. excavatum can live in sediments of varying grain size and organic content (e.g., Alve and Murray, 1999; Takata et al., 2006), and can survive under low salt conditions (Abu-Zied et al., 2007). In the present study, it tolerates slightly hypersaline waters. E. williamsoni may be found on high marshes subject to highly changing conditions (review in Murray, 2006). In the present study, its distribution is consistent with the general trend of Chl-a, and is probably related to the same causes since it husbands functioning algal chloroplasts with a significant amount of chlorophyll a (Knight and Mantoura, 1985). The species found during this study are generally reported in mangrove swamps, but some species typical of mangrove habitats or salt marshes were absent. The most typical of them are Miliammina fusca and Trochamminita irregularis, usually found in very low salinity environments (Hayward et al., 1999; Debenay et al., 2002), and Arenoparrella mexicana and Haplophragmoides wilberti, both growing in organic-rich sediments. The two dominant agglutinated species T. inflata and T. patensis are very irregularly represented in space and time, which probably reflects their highly patchy distribution within the litter. T. inflata is generally reported as characteristic of vegetated areas at high tidal elevation (e.g., Scott and Medioli, 1978; Jennings and Nelson, 1992). Its presence in both mangroves all along the monitoring period reflects its high tolerance to changing environmental conditions. The higher abundance of T. patensis in the control mangrove before the last harvest can be attributed to its preference for lowsalinity resulting from abundant rainfall, which is greatly attenuated by the discharge of normal-saline effluents in the impacted mangrove. As far as we know, G. lamarckiana and T. patensis had never been reported before in mangrove swamps from New Caledonia. They are suspected to be accidentally introduced alien species. About half of the species collected in the impacted and control Avicennia mangroves were not found in shrimp ponds (Debenay et al., 2009a, 2009b). Conversely, species usually associated with organic matter enrichment, such as H. wilberti and A. mexicana, were found in shrimp ponds, but not in the mangrove. It is consistent with the exceptional proportion of abnormal tests (often >50%, sometimes >80%) in the shrimp ponds, due to high concentration of organic, mostly easily oxidized, matter (Debenay et al., 2009a), while the very low proportion of abnormal tests in the impacted mangrove attested the limited organic enrichment of the mangrove sediment by the effluents (Molnar et al., 2014). It also suggests a weak impact of pollutants, since organic matter and/or

114

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Table 3 Main characteristics of foraminiferal assemblages in each cluster. Number of samples

Density total assemblage

Density stained assemblage

Species richness

% Stained individuals

Average

sd

Average

sd

Average

sd

Average

sd

Mapping of December 2009 (Non Active Period) Cluster I 14 Cluster II 22 Cluster III 5 Cluster IV 9

12,000 9300 640 6000

12,000 8200 220 3800

3000 2400 31 1500

4100 2700 38 1500

6 10 7 11

1.8 2.9 0.8 3.7

19 24 4.4 24

17 16 4.5 15

Mapping of July 2010 (Active Period) Cluster I 3 Cluster II 4 Cluster III 18 Cluster IV 18

17,000 2300 5600 7200

14,000 1900 6400 4400

3300 500 1700 1800

2800 220 2200 1200

6 11 14 9

1.5 2.1 4.4 4

14 26 30 23

14 8.5 14 15

Fig. 9. (A) Representation of the first plane (axes 1 and 2) of the canonical correspondence analysis triplot for December 2009 (Non Active Period – NAP); (B) enlargement of the arrows representing explanatory variables; (C) enlargement of the plots representing the stations, grouped according to the clusters identified by the cluster analysis. See text for explanation.

pollutants generally induce test deformations (e.g., review in Frontalini and Coccioni, 2008; Romano et al., 2008), and of the relatively high salinity in the two mangroves (Fig. 13) whereas salinity stress may cause abnormalities (e.g., Almogi-Labin et al., 1992). 4.3. Foraminiferal assemblages in relation to mangrove stand and farm activity 4.3.1. Foraminiferal assemblages in the whole mangrove during the non-active period In many estuaries and lagoons, foraminiferal distributions can be explained, in large part, by spatial or temporal variations in salinity due to the relative influence of riverine and coastal waters

(review in Culver et al., 2012). In the study area, however, freshwater inputs are limited to rain and runoff. As a result, salinity of surface sediments increases landwards from 37.7 ± 1.0 in Rhizophora stands to 53.8 ± 10.5 on the bare or low vegetated salt-flats (Molnar, 2012), in relation with soil elevation (Marchand et al., 2011b, 2012). In such a case, vertical elevation is the main parameter in controlling foraminiferal distributions (e.g., Horton et al., 2003; Culver et al., 2013). Generally, agglutinated forms dominate higher intertidal elevations (usually vegetated and thus organic enriched), and calcareous taxa are increasingly abundant downwards (e.g., Woodroffe et al., 2005). The study area shows a reverse trend, which might be due to the presence of bare or low vegetated salt-flats, unfavorable to agglutinated species (review in Berkeley et al., 2007). Lower density of foraminiferal

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

115

Fig. 10. (A) Representation of the first plane (axes 1 and 2) of the canonical correspondence analysis triplot for July 2010 (Active Period – AP); (B) enlargement of the plots representing the stations, grouped according to the clusters identified by the cluster analysis; (C) enlargement of the arrows representing explanatory variables. See text for explanation.

assemblages in the Rhizophora stands is consistent with the lowest values of both the meiobenthic and the microphytobenthic biomasses. It may result from the low primary production detected within these stands, and/or to the anoxic conditions prevailing during the whole year in the sediments whereas sediments are well oxygenated in the Avicennia stands and on salt-flats (Molnar, 2012). Higher species richness in the Rhizophora stands probably results from the very strong spatial heterogeneity due to the roots, litter accumulation and the high density of crab burrows (40 m 2), which provides more potential ecological niches available to foraminiferal species. The correlations with axis 1 of the CCA (Fig. 9), positive for C:N and fungal markers, characteristic of Rhizophora stands, and negative for d13C and d15N, characteristic of salt-flats (Molnar, 2012), indicate that axis 1 represents a gradient between these two contrasting environments, which acts as the major factor controlling

the distribution of foraminiferal assemblages. Within this gradient foraminifera are also sensitive to the presence of diatoms and other microalgae, as shown by the negative correlation of microalgal markers and of the photosynthetic pigments with axis 2 (Fig. 9). The correlation of the agglutinated species with the Rhizophora stands (axis 1 positive) probably reflects their detritivorous feeding behavior, which allow them to benefit from minute phytodetritus originating from litter decay (high C:N ratio). Moreover, scanning electron microscope observations (Lepoint et al., 2006), as well as FAs analyses, indicate that leaf litter is highly colonized by diverse diatoms, bacteria and fungi, which also constitute potential food sources and may favor higher species diversity. Indeed, Paratrochammina sp., M. salsa and H. anderseni that are strongly correlated with 20:5x3 presumably feed on diatoms, while G. gordialis, C. exilis and T. inflata that plot positively with axis 2 presumably feed mainly on phytodetritus and bacteria.

116

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Fig. 11. Species richness and density at the height periods of sampling from February 2009 to February 2010.

Plotting negatively on axis 1, Q. seminula is strongly correlated with salt-flats, which is in agreement with its living habits (discussion above). Due to the husbanding of functional chloroplasts, E. williamsoni may have similar requirements as microphytobenthos, which may explain its positive correlation with photosynthetic pigments. E. excavatum and A. tepida are close to the center of the ordination plane, which show their tolerance to a variety of environments and food sources. E. excavatum, however, is correlated with microalgae other than diatoms, while the position of A. tepida suggests a preference for phytodetritus and bacteria. Sampling stations of Cluster I are localized on salt-flats and in the Avicennia stands (Fig. 7), which is consistent with their negative correlation with axis 1. Cluster II covers a wide range of the Rhizophora stands, and therefore a wide range of habitats, which explains its central position on the ordination plane. Clusters III and IV comprise stations from the Rhizophora stands and are logically positively correlated with axis 1. Sampling stations of Cluster III were collected in front of the levee in an area of low concentrations in Chl-a and phaeopigments, which explain their positive correlation with axis 2. They have low species richness, with mostly agglutinated forms. Conversely, the stations of Cluster IV, collected in the axis of the bay are negatively correlated with axis 2, which suggests that foraminifera feed primarily on diatoms and other microalgae. This cluster shows the highest average species richness, hosting a mixture of calcareous and agglutinated species. 4.3.2. Changes in foraminiferal assemblages in the whole mangrove from the non active period to the active period The decreases of species density between NAP and AP is probably related, directly or not, to the general decrease of the

microphytobenthic biomass, which was more than twice greater in December 2009 than in June 2010 (Molnar et al., 2014). Indeed, reproduction periods have often been considered as a response to high food supply resulting from phytoplankton blooms (e.g., Alve and Murray, 1994). The increased abundance of the highly tolerant A. tepida in the main impacted area, located near the levee (Fig. 2), is consistent with the widely reported preference of this species for organic matter inputs, and its tolerance to pollution. It indicates a localized impact of farm effluents. At both periods, foraminiferal assemblages of thearea located in front of the levee had peculiar characteristics with lower density than elsewhere. The presence of higher relative abundance of T. inflata, which live attached on hard substrates, suggests that higher speed of tidal currents, resulting from the narrowing of the bay in front of the levee, may be responsible of the peculiarity of the area. The gradient between the Rhizophora stands and the salt-flats evidenced by the CCA, which was the major factor controlling foraminiferal assemblages during NAP, was weakened during AP and appeared only along axis 2 (Fig. 10). The weakening of this gradient, associated with the drop of d15N and d13C values, and lower C:N ratios, may result from higher litter decay in the Rhizophora stands (Aschenbroich et al., 2015). It traduces the homogenization of the organic matter in the sediments of the whole mangrove towards a Rhizophora-like area (Aschenbroich et al., 2015), which was favorable to the detritivorous species T. inflata, and considerably reduced the strong correlation found during NAP between this species and the Rhizophora stands. The higher litter decay was attributed to enhanced fungal and bacterial activities favored by nutrient input. The resulting fungal and bacterial biomasses were potential food sources for foraminifera, which probably explain their

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

117

Fig. 12. Absolute abundance of A. tepida, Q. seminula, T. patensis and T. inflata at the height periods of sampling from February 2009 to February 2010.

prominent role on the distribution of foraminiferal assemblages (axis 1). As for NAP, agglutinated species were correlated with bacterial and fungal biomass. The distinction made during NAP between agglutinated species presumably feeding on diatoms and species presumably feeding on bacteria was not found during AP. It suggests that these species are mainly detritivores and that bacteria and diatoms were not a primary food source for them. Q. seminula and E. excavatum were still on the salt-flats (negatively correlated with C:N), but were correlated with diatoms, while they were correlated with other microalgae during NAP (Figs. 9 and 10). They seem to have change of feeding source from one period to the other, which was probably made necessary by the increase of diatom growth on the salt-flat at the expense of other microalgae (Aschenbroich et al., 2015). Such an adaptability to food availability has been reported for Q. seminula in other environments (Suhr et al., 2003). E. williamsoni was still correlated with Chl-a, which confirms that it may have similar requirements as microphytobenthos. The FA 18:1x9 has been recognized as the

most reliable tracer of effluent impact during AP (Aschenbroich et al., 2015). Its widespread redistribution under tidal influence during AP probably explains its weaker correlation with foraminiferal assemblages than during NAP, when it mostly resulted from fungal activity within the Rhizophora stands. Except two stations from the salt-flats (Cluster I) and two from the Rhizophora stands of the inner bay, correlated with the phaeopigments and characterized by T. patensis (Cluster II), all the other stations are grouped within two clusters that strongly overlap in the field (Fig. 7) as well as on the ordination plane (Fig. 10), and are distributed around the center of the plane. It indicates a weakening of the environmental zonation from NAP to AP that led to less constraint for foraminiferal assemblages and somewhat obscured their natural zonation found during NAP and usually observed in mangrove environments. To summarize, the key parameter that governed the distribution of foraminiferal assemblages in the bay was the gradient of salinity from Rhizophora stands to the salt-flats. Subordinate parameters were the

118

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Fig. 13. Salinity at the height periods of sampling from February 2009 to February 2010 in the impacted and control mangroves (from Molnar, 2012).

availability of ecological niches and food sources, the effect of increased tidal current velocities in front of the levee, the impact of effluent discharge close to outputs, the presence of low vegetated areas at higher intertidal elevations. The potential impact of the farm is found in food sources and effluent outputs. The most striking changes between NAP and AP were the weakening of the gradient from Rhizophora stands to salt-flats, and the inferred adaptation of some species to changes in food availability by changing their feeding source. As shown by the CCA, foraminiferal assemblages do not indicate a substantial impact of shrimp effluents on the mangrove meiobenthos, all the more than the seasonal cycle is superimposed on the rearing cycle.

4.4. Foraminiferal assemblages in the Avicennia stand: respective influence of season and farm activity The general decrease of species richness, associated with increasing density, after June 2009, is consistent with the Pearson–Rosenberg model prediction in case of increasing organic input (Pearson and Rosenberg, 1978). However, minimum species richness and maximum density in the Avicennia stand occurred when the shrimp farm had stopped discharging nutrient-rich effluents to the mangrove, and the trends are similar in both mangroves. Therefore, they cannot be related to nutrient input from the shrimp farm, but are more likely to be related to the natural cycles of primary productivity in the lagoon of New Caledonia. Maximum biomass and production of micro- and macro-algae, which participate to foraminiferal food sources, occur during the hot season (October to February) (Clavier and Garrigue, 1999). However, slightly higher density in the impacted mangrove than in the control one during the rearing cycle (Fig. 11) may be favored by enriched effluents discharge. Molnar et al. (2014) showed that the development of microphytobenthos was driven both by the release of effluent and the climatic seasonal evolution. They suggested that the final drain of the shrimp ponds, which occurred just before the seasonal temperature increase, induced a boosted algal bloom. Other parameters may have influenced the benthic communities. They are: the downward shift of about 15 cm of the intertidal area from February to July, which modify exposure during tidal cycles; the contrast between rainfalls higher than average values from February to July but particularly low from August to December 2009. The combination of the two parameters led to higher salinity of pore-waters after August (Molnar et al., 2014; Fig. 13). Moreover, the heavy rains that occurred particularly at

the beginning of the monitoring period drastically lowered the salinity at sediment surface, favoring agglutinated species able to live in low-saline environments, but disfavoring species usually living in normal saline or hypersaline waters, leading to low assemblage density (Fig. 11). In a nutshell, global changes in the foraminiferal assemblages with time seem much more directly dependent on the natural seasonal cycle than on the shrimp farm rearing cycle. We infer that the main differences between the two mangroves during the rearing period (February–June), with lower proportions of T. patensis and T. inflata, higher proportion of Q. seminula, and higher density in the impacted mangrove than in the control one partly result from the interplay between effluent discharge, rainfall and period of tidal immersion. As discussed above, the natural trend leads to lowered salinity during the rearing period. It is greatly attenuated in the impacted mangrove, due to the discharge of normal-saline effluents that offset the rainfall impact. The increasing proportion of Q. seminula in May and June in the impacted mangrove is attributed to the fact that the saline water of the effluent may evaporate under the joint effect of decreasing rainfall and longer tidal exposure time due to lower tides (Fig. 1). It results in higher salinity favorable to Q. seminula. Unfortunately, salinity measurements, carried out at each sampling period on the pore water of surface sediments, cannot evidence these surficial changes that occur at the time scale of tidal cycles, or under the temporary effect of showers. Another difference between the two mangroves is that higher concentrations of Chl-a were found in the impacted mangrove where microphytobenthic biomass was three times greater, and meiobenthic biomass and production two times greater than in the control mangrove (Molnar et al., 2014). It suggests greater food availability for foraminifera that may explain the higher densities of foraminiferal assemblages measured in the impacted mangrove during the rearing cycle. To summarize, the differences between the two mangroves deduced from foraminiferal assemblages is inferred to result from the combined effects of normal-saline farm effluents on surface salinity, and of nutrient input on food availability, both key parameters for foraminiferal distribution.

5. Conclusions The key parameters that govern the distribution of foraminiferal assemblages are the three interconnected parameters: gradient from sea to salt-flats, salinity and tidal elevation, while changes in foraminiferal assemblages with time mostly depend on natural seasonal cycles. All these parameters are independent from the shrimp farm activity. However, differences in foraminiferal assemblages between the two mangroves evidence a moderate impact of shrimp farming that results, on the one hand, from the effect of normal-saline farm effluents on surface salinity, and, on the other hand, from the effect of nutrient input on food availability. The absence of foraminiferal species usually associated to high organic enrichment and the very low proportion of deformed tests indicates that the receiving mangrove is not adversely impacted by organic and/or chemical pollution, which suggest a limited cumulative effect of 25 years of rearing activities on the impacted mangrove. However, two species G. lamarckiana and T. patensis are suspected to be alien species accidentally introduced in the bay. Correlations could be identified between changes in foraminiferal assemblages from the non-active period to the rearing period and changes in potential food sources evidenced by geochemical parameters, and particularly by FAs concentrations. Changes in food sources were driven both by the release of effluent and the climatic seasonal evolution. The relative influence of these two parameters, and thus the direct impact of effluents on foraminiferal

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

assemblages was not clearly established. Only the pollution indicator A. tepida evidenced the direct effect of effluents close to an outflow area. This study suggests that semi-intensive shrimp farming using mangrove for effluent discharge may appear as a sustainable solution in New Caledonia, when considering only the impact on the mangrove itself. But the efficiency of the mangrove as a biofilter is debated and the effect of shrimp farming on the neighboring lagoon is still poorly known. Acknowledgements This work was supported by the ZONECO Program, the Northern Province and Southern Province of New Caledonia. We thank the staff of the ‘‘Ferme Aquacole de la Ouenghi’’ farm for allowing us the access to the ponds and for the background information concerning the farm activity. We thank the staff of the IFREMER, Saint Vincent station for their valuable help, in particular L. Della Patrona for field assistance. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.marpolbul.2015. 02.009. References Abu-Zied, R.H., Keatings, K.W., Flower, R.J., 2007. Environmental controls on foraminifera in lake Qarun, Egypt. J. Foramin. Res. 37, 136–149. Almogi-Labin, A., Perelis-Grossovicz, L., Raab, M., 1992. Living Ammonia from a hypersaline inland pool, Dead Sea area, Israel. J. Foramin. Res. 22, 257–266. Alve, E., Murray, J.W., 1994. Ecology and taphonomy of benthic foraminifera in a temperate mesotidal inlet. J. Foramin. Res. 24, 18–27. Alve, E., Murray, J.W., 1999. Marginal marine environments of the Skagerrak and Kattegat: a baseline study of living (stained) benthic foraminiferal ecology. Palaeogeogr. Palaeoclimatol. Palaeoecol. 146, 171–193. Angel, D.L., Verghese, S., Lee, J.J., Saleh, A.M., Zuber, D., Lindell, D., Symons, A., 2000. Impact of a net cage fish farm on the distribution of benthic foraminifera in the northern Gulf of Eilat (Aqaba, Red Sea). J. Foramin. Res. 30, 54–65. Armynot du Châtelet, E., Debenay, J.-P., 2010. Anthropogenic impact on the western French coast as revealed by foraminifera: a review. Rev. Micropaléontol. 53, 129–137. Aschenbroich, A., Marchand, C., Molnar, N., Deborde, J., Hubas, C., Rybarczyk, H., Meziane, T., 2015. Spatio-temporal variations in the composition of organic matter in surface sediments of a mangrove receiving shrimp farm effluents (New Caledonia). Sci. Total Environ. 512–513, 296–307. Baker, K.L., Langenheder, S., Nicol, G.W., Ricketts, D., Killham, K.S., Campbell, C.B., Prosser, J.I., 2009. Environmental and spatial characterisation of bacterial community composition in soil to inform sampling strategies. Soil Biol. Biochem. 41 (11), 2292–2298. Beck, M.W., 2000. Separating the elements of habitat structure: independent effects of habitat complexity and structural components on rocky intertidal gastropods. J. Exp. Mar. Biol. Ecol. 249, 29–49. Bennington, J.B., 2003. Transcending patchiness in the comparative analysis of paleocommunities: a test case from the Upper Cretaceous of New Jersey. Palaios 18, 22–33. Berkeley, A., Perry, C.T., Smithers, S.G., Horton, B.P., Taylor, K.G., 2007. A review of the ecological and taphonomic controls on foraminiferal assemblage development in intertidal environments. Earth Sci. Rev. 83, 205–230. Boltovskoy, E., Lena, H., 1969. Seasonal occurrences, standing crop and production in benthic foraminifera of Puerto Deseado: Contributions from the Cushman Foundation for Foraminiferal Research 20, pp. 87–95. Bouchet, V., Debenay, J.-P., Sauriau, P.-G., Radford-Knoery, J., Soletchnik, P., 2007. Effects of short-term environmental disturbances on living benthic foraminifera during the Pacific oyster summer mortality in the Marennes-Oléron Bay (France). Mar. Environ. Res. 64, 358–383. Boyd, C.E., Green, B.W., 2002. Coastal water quality monitoring in shrimp farming areas, an example from Honduras. Report prepared under the World Bank, NACA, WWF and FAO Consortium Program on Shrimp Farming and the Environment. Work in Progress for Public Discussion, published by the Consortium. Bangkok: Network of Aquaculture Centers in Asia Pacific (NACA). . Bui, T.D., Luong-Van, J., Austin, C.M., 2012. Impact of shrimp farm effluent on water quality in coastal areas of the world heritage-listed ha long bay. Am J. Environ. Sci. 8, 104–116.

119

Bui, T.D., Luong-Van, J., Maier, S.W., Austin, C.M., 2013. Assessment and monitoring of nutrient loading in the sediments of tidal creeks receiving shrimp farm effluent in Quang Ninh, Vietnam. Environ Monit. Asses. 185, 8715–8731. Buzas, M.A., 1970. Spatial homogeneity: statistical analyses of unispecies and multispecies populations of Foraminifera. Ecology 51, 874–879. Caeiro, S., Painho, M., Goovaerts, P., Costa, H., Sousa, S., 2003. Spatial sampling design for sediment quality assessment in estuaries. Environ. Modell. Softw. 18, 853–859. Carnahan, E.A., Hoare, A.M., Hallock, P., Lidz, B.H., Reich, C.D., 2009. Foraminiferal assemblages in Biscayne Bay, Florida, USA: responses to urban and agricultural influence in a subtropical estuary. Mar. Pollut. Bull. 59, 221–233. Casillas-Hernández, R., Nolasco-Soria, H., García-Galano, T., Carrillo-Farnes, O., Páez-Osuna, F., 2007. Water quality, chemical fluxes and production in semiintensive Pacific white shrimp (Litopenaeus vannamei) culture ponds utilizing two different feeding strategies. Aquacult. Eng. 36 (2), 105–114. Chen, F.H., Scow, K.M., Graham, K.J., 2001. Composition and dynamics of selected fungal-feeding nematodes and Fungi. Comp. Biochem. Physiol. 130, 135–144. Clavier, J., Garrigue, C., 1999. Annual sediment primary production and respiration in a large coral reef lagoon (SW New Caledonia). Mar Ecol.-Prog. Ser. 191, 79– 89. Costanzo, S.D., O’Donohue, M.J., Dennison, W.C., 2004. Assessing the influence and distribution of shrimp pond effluent in a tidal mangrove creek in north-east Australia. Mar. Pollut. Bull. 48, 514–525. Culver, S.J., Mallinson, D.J., Corbett, D.R., Leorri, E., Rouf, A.A., Shazili, N.A.M., Yaacob, R., Whittaker, J.E., Buzas, M.A., Parham, P.R., 2012. Distribution of foraminifera in the Setiu estuary and lagoon, Malaysia. J. Foramin. Res. 42, 109–133. Culver, S.J., Leorri, E., Corbett, D.R., Mallinson, D.J., Shazili, N.A.M., Rouf, A.A., Nasir, M., Parham, P.R., Rosnan, Y., 2013. Infaunal mangrove swamp foraminifera in the Setiu wetland, Terengganu, Malaysia. J. Foramin. Res. 43, 262–279. Debenay, J.-P., Guiral, D., 2006. Mangrove swamp foraminifera, indicators of sea level or paleoclimate. Rev. Paléobiol. 25, 567–574. Debenay, J.-P., 2012. A guide to 1000 foraminifera from southwestern Pacific: New Calédonia. Editions IRD, Marseille/Publications Scientifiques du Museum, Paris, 383pp. Debenay, J.-P., Guillou, J.-J., 2002. Ecological transitions indicated by foraminiferal assemblages in paralic environments. Estuaries 25, 1107–1120. Debenay, J.-P., Guillou, J.-J., Redois, F., Geslin, E., 2000. Distribution trends of foraminiferal assemblages in paralic environments: a base for using foraminifera as bioindicators. In: Martin, R.E. (Ed.), Environmental Micropaleontology: The Application of Microfossils to Environmental Geology. Kluwer Academic/Plenum Publishers, New York, pp. 39–67. Debenay, J.-P., Guiral, D., Parra, M., 2002. Ecological factors acting on the microfauna in mangrove swamps. The case of foraminiferal assemblages in French Guiana. Estuar. Coast. Shelf Sci. 55, 509–533. Debenay, J.-P., Guiral, D., Parra, M., 2004. Behaviour and taphonomic loss in foraminiferal assemblages of mangrove swamps of French Guiana. Mar. Geol. 208, 295–314. Debenay, J.-P., Della Patrona, L., Herbland, A., Goguenheim, H., 2009a. The impact of Easily Oxidized Material (EOM) on the meiobenthos: Foraminifera abnormalities in Shrimp ponds of New Caledonia, implications for environment and paleoenvironment survey. Mar. Pollut. Bull. 59, 323–335. Debenay, J.-P., Della Patrona, L., Herbland, A., Goguenheim, H., 2009b. Colonization of coastal environments by Foraminifera: insight from shrimp ponds in New Caledonia. J. Foramin. Res. 39, 249–266. De Cáceres, M., Font, X., García, R., Oliva, F., 2003. VEGANA, un paquete de programas para la gestión y análisis de datos ecológicos. VII Congreso Nacional de la Asociación Española de Ecología Terrestre. Barcelona. Julio, 2003, pp. 1484–1497. Della Patrona, L., Brun, P., 2008. Elevage de la Crevette Bleue en Nouvelle-Calédonie Litopenaeus stylirostris: bases biologiques et zootechnie. Institut français de recherche pour l’exploitation de la mer (Ifremer), 312pp. Douillet, P., 1998. Tidal dynamics of the south-west lagoon of New Caledonia: observations and 2D numerical modelling. Oceanol. Acta 21 (1), 69–79. EJF, 2003. Risky Business: Vietnamese Shrimp Aquaculture–Impacts and Improvements. Environmental Justice Foundation, London, UK, 44pp. Engel, M., Behnke, A., Bauerfeld, S., Bauer, C., Buschbaum, C., Volkenborn, N., Stoeck, T., 2012. Sample pooling obscures diversity patterns in intertidal ciliate community composition and structure. FEMS Microbiol. Ecol. 79, 741–750. Findlay, S.E.G., 1981. Small scale spatial distribution of meiofauna on a mud and sandflat. Estuar. Coast. Shelf Sci. 12, 471–484. Frontalini, F., Coccioni, R., 2008. Benthic foraminifera for heavy metal pollution monitoring: a case study from the central Adriatic Sea coast of Italy. Estuar. Coast. Shelf Sci. 76, 404–417. Gautier, D., 2002. The Integration of Mangrove and Shrimp Farming: A Case Study on the Caribbean Coast of Colombia. Report prepared under the World Bank, NACA, WWF and FAO Consortium Program on Shrimp Farming and the Environment. Work in Progress for Public Discussion. Published by the Consortium. 26 pages. Gautier, D., Amador, J., Newmark, F., 2001. The use of mangrove wetland as a biofilter to treat shrimp pond effluents: preliminary results of an experiment on the Caribbean coast of Colombia. Aquac. Res. 32, 787–799. Hasani, Q., Adiwilaga, E.M., Pratiwi, N.T.M., 2012. The relationship between the Harmful Algal Blooms (HABs) phenomenon with nutrients at shrimp farms and fish cage culture sites in Pesawaran District Lampung Bay. Makara J. Sci. 16 (3), 183–191.

120

J.-P. Debenay et al. / Marine Pollution Bulletin 93 (2015) 103–120

Hayek, L.A.C., Buzas, M.A., 1997. Surveying Natural Populations. Columbia University Press, New York, 563pp. Hayward, B.W., Grenfell, H.R., Scott, D.B., 1999. Tidal range of marsh foraminifera for determining former sea-level heights in New Zealand. NZ J. Geol. Geophys. 42, 395–413. Herbeck, L.S., Unger, D., Wu, Y., Jennerjahn, T.C., 2013. Effluent, nutrient and organic matter export from shrimp and fish ponds causing eutrophication in coastal and back-reef waters of NE Hainan, tropical China. Cont. Shelf Res. 57, 92–104. Horton, B.P., Larcombe, P., Woodroffe, S.A., Whittaker, J.E., Wright, M.R., Wynn, C., 2003. Contemporary foraminiferal distributions of a mangrove environment, great barrier reef coastline, Australia: implications for sea-level reconstructions. Mar. Geol. 198, 225–243. Jackson, C., Preston, N., Thompson, P., Burford, M., 2003. Nitrogen budget and effluent nitrogen components at an intensive shrimp farm. Aquaculture 218, 397–411. Jennings, A.E., Nelson, A.R., 1992. Foraminiferal assemblage zones in Oregon tidal marshes – relation to marsh floral zones and sea level. J. Foramin. Res. 22, 13– 29. Kahlert, M., 2001. Biomass and Nutrient Status of Benthic Algae in Lakes. Acta Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 649, 35. Knight, R., Mantoura, R.F.C., 1985. Chlorophyll and carotenoid pigments in Foraminifera and their symbiotic algae: analysis by high performance liquid chromatography. Mar. Ecol. Prog. Ser. 23 (3), 241–249. Kristensen, E., 2008. Mangrove crabs as ecosystem engineers; with emphasis on sediment processes. J. Sea Res. 59, 30–43. Lang, C., 1989. Effects of small-scale sedimentary patchiness on the distribution of tubificid and lumbriculid worms in Lake Geneva. Freshw. Biol. 313, 477–481. Lepoint, G., Cox, A.S., Dauby, P., Poulicek, M., Gobert, S., 2006. Food sources of two detritivore amphipods associated with the seagrass Posidonia oceanica leaf litter. Mar. Biol. Res. 2 (5), 355–365. Lörz, A.N., di Renzo, A., Nickel, J., 1998. Comparative analysis of three sampling gear types for marine macrobenthos. In: Brandt, A., Thomsen, H.A., Heide-Jorgensen, H., Kristensen, R.M., Ruhberg, H. (Eds), The 1998 Danish-German Excursion to Disko Island, West Greenland, pp. 134–151. Luan, B.T., Debenay, J.P., 2005. Foraminifera, environmental bioindicators in the highly impacted environments of the Mekong delta. Hydrobiologia 548, 75–83. Marchand, C., Lallier-Vergès, E., Allenbach, M., 2011a. Redox conditions and heavy metals distribution in mangrove forests receiving effluents from shrimp farms (Teremba Bay, New Caledonia). J. Soils Sediments 11, 529–541. Marchand, C., Allenbach, M., Lallier-Vergès, E., 2011b. Relationships between heavy metals distribution and organic matter cycling in mangrove sediments of Conception Bay, New Caledonia. Geoderma 160, 444–456. Marchand, C., Fernandez, J.M., Moreton, B., Landi, L., Lallier-Vergès, E., Baltzer, F., 2012. The partitioning of transitional metals (Fe, Mn, Ni, Cr) in mangrove sediments downstream of a ferralitized ultramafic watershed (New Caledonia). Chem. Geol. 300, 70–80. Martinez-Porchas, M., Martinez-Cordova, L.R., 2012. World aquaculture: environmental impacts and troubleshooting alternatives. The Scientific World Journal 2012. Article ID 389623, 9 pages. McKinnon, A.D., Trott, L.A., Cappo, M., Miller, D.K., Duggan, S., Speare, P., Davidson, A., 2002. The trophic fate of shrimp farm effluent in mangrove creeks of North Queensland, Australia. Estuar. Coast. Shelf Sci. 55, 655–671. Miranda-Baeza, A., Voltolina, D., Brambila-Gámez, M.A., Frías-Espiricueta, M.G., Simental, J., 2007. Effluent characteristics and nutrient loading of a semiintensive shrimp farm in NW México. Life Environ. 57, 21–27. Molnar, N., 2012. Impact des effluents de la crevetticulture sur la dynamique de la matière organique benthique et leurs implications sur les processus biogéochimiques dans une mangrove (Nouvelle-Calédonie). PhD Thesis Museum National d’Histoire Naturelle Paris, 268pp. Molnar, N., Welsh, D.T., Marchand, C., Deborde, J., Meziane, T., 2013. Impacts of shrimp farm effluent on water quality, benthic metabolism and N-dynamics in a mangrove forest (New Caledonia). Estuar. Coast. Shelf Sci. 117, 12–21. Molnar, N., Marchand, C., Deborde, J., Patrona, L., Meziane, T., 2014. Seasonal pattern of the biogeochemical properties of mangrove sediments receiving shrimp farm effluents (New Caledonia). J. Aquat. Res. Dev. 5 (5). http://dx.doi.org/10.4172/ 2155-9546.1000262. Morvan, J., Debenay, J.-P., Jorissen, F., Redois, F., Bénéteau, E., Delplancke, M., Amato, A.S., 2006. Patchiness and life cycle of intertidal foraminifera: implication for environmental and paleoenvironmental interpretation. Mar. Micropaleontol. 61, 131–154. Murray, J.W., Alve, E., 2000. Major aspects of foraminiferal variability (standing crop and biomass) on a monthly scale in an intertidal zone. J. Foramin. Res. 30, 177– 191. Murray, J.W., Bowser, S.S., 2000. Mortality, protoplasm decay rate, and reliability of staining techniques to recognize ‘‘living’’ foraminifera: a review. J. Foramin. Res. 30, 66–70. Murray, J.W., 2006. Ecology and Applications of Benthic Foraminifera. Cambridge University Press, Cambridge, UK, 426pp. Nigam, R., Saraswat, R., Panchang, R., 2006. Application of foraminifers in ecotoxicology: Retrospect, perspect and prospect. Environ. Int. 32, 273–283. Ohman, M.D., Lavaniegos, B.E., 2002. Comparative zooplankton sampling efficiency of a ring net and bongo et with comments on pooling of subsamples. California Cooperative Oceanic Fisheries Invest. Rep. 43, 162–173.

Olafsson, E., Carlström, S., Ndaro, S.G.M., 2000. Meiobenthos of hypersaline tropical mangrove sediment in relation to spring tide innundation. Hydrobiologia 426, 57–64. Páez-Osuna, F., 2001. The environmental impact of shrimp aquaculture: causes, effects, and mitigating alternatives. Environ. Manage. 28 (1), 131–140. Páez-Osuna, F., Guerrero-Galvan, S.R., Ruiz-Fernandez, A.C., 1998. The environmental impact of shrimp aquaculture and the coastal pollution in Mexico. Mar. Pollut. Bull. 36, 65–75. Parravicini, V., Micheli, F., Montefalcone, M., Villa, E., Morri, C., Bianchi, C.N., 2010. Rapid assessment of epibenthic communities: a comparison between two visual sampling techniques. J. Exp. Mar. Biol. Ecol. 395, 21–29. Pati, P.K., Patra, P.K., 2012. Benthic foraminiferal response to coastal pollution: a review. Int. J. Geol., Earth Environ. Sci. 2, 42–56. Pearson, T.H., Rosenberg, R., 1978. Macrobenthic succession in relation to organic enrichment and pollution of the marine environment. Oceanogr. Mar. Biol. Annu. Rev. 16, 229–311. Romano, E., Bergamin, L., Finoia, M.G., Carboni, M.G., Ausili, A., Gabellini, M., 2008. Industrial pollution at Bagnoli (Naples, Italy): benthic foraminifera as a tool in integrated programs of environmental characterisation. Mar. Pollut. Bull. 56, 439–457. Schafer, C.T., Winters, G.V., Scott, D.B., Pocklington, P., Cole, F.E., Honig, C., 1995. Survey of living foraminifera and polychaete populations at some Canadian aquaculture sites: potential for impact mapping and monitoring. J. Foramin. Res. 25, 236–259. Schönfeld, J., Alve, E., Geslin, E., Jorissen, F., Korsun, S., Spezzaferri, S., 2012. The FOBIMO (FOraminiferal BIo-MOnitoring) initiative—Towards a standardised protocol for soft-bottom benthic foraminiferal monitoring studies. Mar. Micropaleontol. 94–95, 1–13. Scott, D.B., Medioli, F.S., 1978. Vertical zonations of marsh foraminifera as accurate indicators of former sea-levels. Nature 272, 528–531. Scott, D.B., Schafer, C.T., Honig, C., Younger, D.C., 1995. Temporal variations of benthic foraminiferal assemblages under or near aquaculture operations: documentation of impact history. J. Foramin. Res. 25, 224–235. Souza, V.M.de, Laut, L.L.M, Silva, F.S.da, Figueiredo Jr., A.G.de, Vital, H., Frazão, E., 2010. Benthic Foraminifera and Bacterial Activity as a Proxy for Environmental Characterization in Potengi Estuary, Rio Grande do Norte, Brazil. Anuário do Instituto de Geociências, UFRJ 33, 1, pp. 20–34. Suhr, S.B., Pond, D.W., Gooday, A.J., Smith, C.R., 2003. Selective feeding of Antarctic benthic foraminifera. Mar. Ecol. Prog. Ser. 262, 153–162. Takata, H., Takayasu, K., Hasegawa, S., 2006. Foraminifera in an organic-rich, brackish-water lagoon, lake Saroma, Hokkaido, Japan. J. Foramin. Res. 36, 44– 60. Tam, N.F.Y., Wong, Y.S., 1999. Mangrove soils in removing pollutants from municipal wastewater of different salinities. J. Environ. Qual. 28 (2), 556–564. Ter Braak, C.J.F., 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67, 1167–1179. Ter Braak, C.J.F., Verdonschot, P., 1995. Canonical correspondence analysis and related multivariate methods in aquatic ecology. Aquat. Sci. 57, 1015–1621. Thomas, Y., Courties, C., El Helwe, Y., Herbland, A., Lemonnier, H., 2010. Spatial and temporal extension of eutrophication originated by shrimp farm wastewater discharges in New Caledonia lagoon. Mar. Pollut. Bull. 61, 387–398. Vidovic´, J., Cosovic, V., Juracic, M., Petricioli, D., 2009. Impact of fish farming on foraminiferal community, Drvenik Veliki Island, Adriatic Sea, Croatia. Mar. Pollut. Bull. 58, 1297–1309. Virly, S., 2008. Atlas des mangroves de Nouvelle-Caledonie. ZoNéCo, 207pp. Virly, S., Buisson, D., Clough, B., Lemonnier, H., Richer de Forges, B., 2005. Evaluation de l’impact de l’aquaculture de crevettes sur les mangroves de NouvelleCalédonie. Rapport final du Programme ZoNéCo, 99pp. Walton, W.R., 1952. Techniques for recognition of living foraminifera. Contrib. Cushman Found. Foramin. Res. 3, 56–60. Wang, M., Zhang, J., Tu, Z., Gao, X., Wang, W., 2010. Maintenance of estuarine water quality by mangroves occurs during flood periods: a case study of a subtropical mangrove wetland. Mar. Pollut. Bull. 60 (11), 2154–2160. Ward, J.H., 1963. Hierarchical grouping to optimize an objective function. J. Am. Stat. Assoc. 58, 236–244. Ware, S.J., Kenny, A.J., 2011. Guidelines for the Conduct of Benthic Studies at Marine Aggregate Extraction Sites, 2nd ed. Marine Aggregate Levy Sustainability Fund, 80pp. Wennrich, V., Meng, S., Schmiedl, G., 2007. Foraminifers from Holocene sediments of two inland brackish lakes in central Germany. J. Foramin. Res. 37, 318–326. Winkelmann, J., Ziemer, O., 1998. Distribution and Composition of Meiofauna with special attention to the Kinorhyncha in the Mellemfjord, Disko Island (West Greenland). In: Brandt, A., Thomsen, H.A., Heide-Jorgensen, H., Kristensen, R.M., Ruhberg, H. (Eds.), The 1998 Danish-German Excursion to Disko Island, West Greenland, pp. 170–182. Woodroffe, S.A., Horton, B.P., Larcombe, P., Whittaker, J.E., 2005. Intertidal mangrove foraminifera from the central great barrier reef shelf, Australia: implications for sea-level reconstruction. J. Foramin. Res. 35 (3), 259–270. Zaldívar-Jiménez, A., Herrera-Silveira, J., Pérez-Ceballos, R., Teutli-Hernández, C., 2012. Evaluation of use mangrove wetland as a biofilter of shrimp pond effluent in Yucatan, Mexico. Rev. Biol. Marin. Oceanogr. 47 (3), 395–405.