Temporal variation on environmental variables and ...

Science of the Total Environment 573 (2016) 841–853

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Temporal variation on environmental variables and pollution indicators in marine sediments under sea Salmon farming cages in protected and exposed zones in the Chilean inland Southern Sea Mauricio A. Urbina Departamento de Zoología, Facultad de Ciencias Naturales y Oceanográficas, Universidad de Concepción, Concepción, Chile

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Salmon farming impacts vary depending on the hydrodynamic characteristics. • The impacts on the sediments varied in magnitude and temporally between exposed and protected zones. • Redox potential, sulphurs and phosphorus are the best to reflect the impacts in both protected and exposed zones. • Organic carbon in the sediments is not an accurate predictor of the salmon farming impacts. • Oxygen availability in the sediments seems to be a major driver of the impacts.

a r t i c l e

i n f o

Article history: Received 2 September 2014 Received in revised form 22 August 2016 Accepted 22 August 2016 Available online xxxx Editor: D. Barcelo Keywords: Salmon farming Fish farming Hydrodynamic conditions Marine sediments Environmental impacts Aquaculture impacts

E-mail address: [email protected].

http://dx.doi.org/10.1016/j.scitotenv.2016.08.166 0048-9697/© 2016 Published by Elsevier B.V.

a b s t r a c t The impacts of any activity on marine ecosystems will depend on the characteristics of the receptor medium and its resilience to external pressures. Salmon farming industry develops along a constant gradient of hydrodynamic conditions in the south of Chile. However, the influence of the hydrodynamic characteristics (weak or strong) on the impacts of intensive salmon farming is still poorly understood. This one year study evaluates the impacts of salmon farming on the marine sediments of both protected and exposed marine zones differing in their hydrodynamic characteristics. Six physico-chemical, five biological variables and seven indexes of marine sediments status were evaluated under the salmon farming cages and control sites. Our results identified a few key variables and indexes necessary to accurately evaluate the salmon farming impacts on both protected and exposed zones. Interestingly, the ranking of importance of the variables and the temporality of the observed changes, varied depending on the hydrodynamic characteristics. Biological variables (nematodes abundance) and environmental indexes (Simpson's dominance, Shannon's diversity and Pielou evenness) are the first to reflect detrimental impacts under the salmon farming cages. Then the physico-chemical variables such as redox, sulphurs and phosphorus in both zones also show detrimental impacts. Based on the present results we propose that the hydrodynamic regime is an important driver of the magnitude and temporality of the effects of salmon farming

842

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

on marine sediments. The variables and indexes that best reflect the effects of salmon farming, in both protected and exposed zones, are also described. © 2016 Published by Elsevier B.V.

1. Introduction Aquaculture production has constantly increased during the last decade. World aquaculture production reached its maximum of around 60 MT in 2010 with a value of USD 119,000 million (FAO, 2012). Within the American continent Chile is the main aquaculture producer with around 701,062 tons in 2010 representing approximately 27% of the total production of the American continent (FAO, 2012). In 2007, 73% of this production corresponded to salmon farming, which since its beginnings in the 1970′s has experienced an abrupt growth (Buschmann et al., 2009). In 2011 Norway and Chile produced 77% of the farmed Atlantic salmon worldwide with a shared production of 1,324,307 tons of a world total of 1,721,254 tons (FAO, 2013). This is in part due to the exceptionally good water conditions of the deep and protected inland fjords of Chile and Norway (Wilding et al., 2012). In Chile the good water quality, environmental and physical conditions of the southern inland channels and fjords have greatly contributed to the development of salmon farming. In recent years, however, salmon farming zones have become scarce since most of the protected zones are currently in use. This low availability of suitable zones (protected) for salmon farming in the south of Chile has been disadvantageous for the salmon farming industry, forcing companies to move to more exposed zones. Intensive aquaculture generates diverse effects on the environment, which then relies on ecosystem services such as the recycling of nutrients and maintenance of water quality for mitigation (Folke and Kautsky, 1989; Beveridge, 1996; Soto and Norambuena, 2004; Mulsow et al., 2006). In salmon farming for example, organic matter and nutrients coming from unconsumed food and feces have to be oxidized and recycled in both water column and marine sediments. The production of one ton of salmon generates around 33 kg of nitrogen (N) and 7 kg of phosphorus (P) pollutants in marine sediment (Buschmann and Pizarro, 2001). For example, an average salmon farm producing about 3000 tons over a period of 17 months (2118 tons per year) generates about 186 tons of organic matter, ~ 70 tons of N and ~ 15 tons of P per year. According to Folke et al. (1994), this is equivalent to the nitrogenous waste produced by 9000 people and the phosphorous waste produced annually by 27,000 people in developing countries. Therefore, one of the main negative environmental effects caused by salmon farming is a change in the structure of marine sediments, impacting benthic communities (Pereira et al., 2004; Tomassetti and Porrello, 2005; Edgar et al., 2010). The magnitude and the consequences of the environmental impacts caused by intensive salmon farming depends on several productive characteristics such as the level of production and time under production, but it also depends on environmental conditions. These include salinity, pluviosity, wind, waves, tidal influence, hydrodynamic characteristics, depth, nutrient availability and the capacity of the sediments to recycle them (Brooks and Mahnken, 2003). Therefore, among these factors salmon farming impacts may differ between zones with different oceanographic conditions. These conditions will ultimately determine the potential accumulation of unconsumed food and feces and their degradation (Cromey et al., 2002; Hevia et al., 1996). In fact, it has been shown that the sediments below salmon farms located in zones with high speed currents are better able to recycle the nutrients than sediments below salmon farms located in low speed currents areas (Findlay and Watling, 1997). However, the opposite pattern has also been documented (Mayor

et al., 2010), and even a recent study have reported no effect of the hydrodynamic regime (Sweetman et al., 2014). Therefore, while the importance of evaluating how the hydrodynamic regime modulates the effects of salmon farming on the sediments has been recognized (Sweetman et al., 2014), only a handful of studies have explored this. Marine sediments are temporally more stable than the water column, in part due to the biotic component such as microbes, algae, and fungal biofilm living on it (Amos et al., 2004; Friend et al., 2003). Therefore, marine sediments may show the impacts of salmon farming for longer. Under aquaculture production, however, the regular input of organic matter and nutrients disturbs the bio stabilized natural sediments leading to erosion (instability) and the formation of colonies of new organisms (Droppo et al., 2007; Karthikeyan et al., 1999). Marine sediments have been described to be negatively impacted by the degradation of salmon farming waste products, leading to a reduction in the concentration of oxygen (Dissolved oxygen, DO) as a result of the aerobic oxidation of organic matter. When oxygen is no longer available (anoxia), the organic matter is oxidized by anaerobic pathways producing ammonium and sulphide by-products (Hargrave et al., 1993; Wildish et al., 2001). Impacts on marine sediment involve changes in physico-chemical variables in turn leading to biological changes. Evaluating the communities that inhabit marine sediments is insightful since their dynamic may integrate the effects of pollutants and disturbances over relatively long time periods (months to years). Furthermore, since different taxa have differing sensitivity to pollutants, their abundance, species richness and dominance allow for a precise evaluation of the effects of salmon farming on impacted sites (Warwick, 1986). Environmental impacts can also be evaluated through the use of indexes (Madec, 2003), which provide useful information about the environmental status before, during and after an event or intervention. Since bias often exists in relation to the use of a particular index, it has been recommended to simultaneously use several indexes to describe the impacts on the marine environments (UICN, 2009). This challenge could be addressed using a multivariate approach and therefore combining several environmental variables into simple indexes (Hartwell, 1998). Given the current importance and extent of fish farming worldwide and its expected growth in the next years, the present study aims to evaluate the effectiveness of using environmental indexes and variables as indicators of salmon farming impacts between zones differing in hydrodynamic characteristics. This is crucial not only to develop a sustainable industry, but also to enhance food security globally. We selected two marine farming zones with differing hydrodynamic characteristics. The exposed zone presented strong hydrodynamic characteristics facilitating the degradation and dilution of contaminants, while the protected zone presented weak hydrodynamic conditions and therefore not facilitating nutrient degradation and their dilution. Physicochemical, biological variables (environmental variables) and diversity indexes were quantified and calculated for sediments of control and farming cages sites identifying the ones best describing salmon farming impacts and their temporal variability over a one year period. For both zones, ecological status was assessed and potential relationships between variables and indexes were evaluated. We hypothesized that sediments below salmon farms located in exposed zones (high speed currents) will show less impact than sediments below salmon farms located in protected zones (low speed currents). We also hypothesized that the effects of salmon farming in the sediments of exposed zones will take longer to develop, and they will recover quicker than in the protected zones.

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

843

Table 1 Hydrographic characteristics of the areas of study. Values presented are the average (±SD) conditions during the year of study.

Exposed Protected

Depth (m)

Speed current (m s−1)

Wind speed (m s−1)

Fetch (mi)

DO (mg L−1)

Wave height (m)

Granulometry % of medium and fine sand (0.250–0.125 mm)

3 23 3 23

1.29 ± 0.35 0.25 ± 0.072 0.37 ± 0 0.016 0.012 ± 0.008

1.67 ± 2.04

8.6

7.2 ± 1.5

1.1 ± 1.4

62.8 ± 6.2

8.26 ± 4.86

4.7

4.5 ± 1.0

0.98 ± 0.7

68.3 ± 5.7

2. Materials and methods 2.1. Sampling and experimental design Both exposed (speed current: 1.29 ± 0.35 m s− 1; DO: 7.2 ± 1.5 mg O2 L−1) and protected (speed current: 0.37 ± 0.02 m s−1; DO: 4.5 ± 1.0 mg O2 L− 1) zones (full details in Table 1), were located in the Los lagos region of Chile. The exposed zone was located in the Reloncaví estuary (42° 06′ 04.8″ S) and the protected zone was located in the Chiloé island (45° 27′ 51.7″ S, Fig. 1). In both zones, two farming installations of similar age and production level were chosen to avoid potential differences due to these factors. As sediment characteristics could also influence the degree of impact of a given area, areas and sites with similar granulometry (particle size) were chosen (Table 1). Furthermore, in order to minimize potential influences of anthropogenic organic matter inputs due to human settlements, sites far away from main cities were chosen. A control station for each salmon farming installation was designated (in both sites and both zones). Physico-chemical, biological variables and environmental indexes were evaluated in the marine sediments of impact and control sites every two months, over a one year period (four seasons). Sediment samples were collected in triplicates using an Ekman sampler of 0.1 m2 sampling area. A weight on top of the sampler (5 kg), further aided the penetration of the Ekman sampler into the sea floor. The sampler was “slammed” into the sea floor and then closed. Ocular inspection of the samples once onboard, guaranteed a minimum of 5 cm depth of undisturbed sediment was collected. The “farming cages” sites were chosen from the oldest cage in production based on the bathymetry, considering the speed and direction of the

currents, based on proximity to where most pollutants accumulate following the methodology proposed by Soto and Norambuena (2004). Control sites were always located approximately 1 km away from the salmon farming cages, but presenting similar bathymetric and oceanographic conditions such as granulometry, depth and currents. Sediment samples (0.1 m2 area x ~ 5 cm depth) were individually placed on black plastic bags, labeled and stored at 4 °C during field work and then frozen at − 20 °C until analysis of benthic fauna (≤5 days). Sub-samples (30 g) from each sample were also preserved onboard by the addition of a 2 mL solution of 19 mM H2SO4 and stored at 4 °C for analysis of the physic-chemical variables (Section 2.2). All other variables were assessed “in situ” immediately after collection (see following sections), and they did not involve the removal of sediment. 2.2. Physico-chemical variables Using the preserved sub-samples described above, organic nitrogen and phosphorus compounds were digested in a solution of mercuric oxide (0.37 M) and sulfuric acid (3 M). Nitrogen was determined by the Kjeldahl method (AOAC, 1980), and phosphorus was quantified spectrophotometrically at 855 nm after addition of molibdate and ascorbic acid (Pearson et al., 1984; Jackson, 1964). Organic matter was determined by gravimetry based on the difference in weight between samples dried for 48 h at 50 °C and after combustion at 450 °C for 4 h (Byers et al., 1978; Urbina et al., 2010). The percentage of total organic carbon was determined following the method of Walkley and Black (1934). Organic matter was oxidized with potassium dichromate and subsequently titrated using a ferrous solution generating diphenylamine

Fig. 1. Study area. The box represents where the majority of Chilean salmon farming occurs (A), exposed (1) and protected zones (2) (B), and location of each salmon farming installation sampled (C).

844

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

(Jackson, 1964). The redox potential was measured using a combined Redox/ORP electrode (model 9678) connected to a Orion TM 290ª meter, immediately after the samples were onboard. The electrode was placed in the first 3 cm of sediment obtained by the Ekman sampler (in an undisturbed part of the sediment) and values were taken once readings were stable. The electrode was rinsed with distilled water between samples to prevent contamination between samples (Brooks, 2001; Orion Research Inc., 2000). Free sulphurs were quantified in a sub-sample of 10 mL of sediment, taken from the top 3 cm of the sediment. This sub-sample was quickly mixed with a buffer solution (SAOB) at a 1:1 ratio then free sulphurs were measured using a combined electrode Silver/Sulphurs (model 9616 BNC) connected to a Orion TM 290ª meter.

and indexes by a Shapiro-Wilk and a Cochran and Bartlett test, respectively (Conover et al., 1981). Data failing parametric assumptions were transformed using 4 √, Log2 (x + 1) or 1/x (Sokal and Rohlf, 1995). If after transformation these variables or indexes still did not meet parametric assumptions, a non-parametric Kruskal-Wallis ANOVA for each individual factor was used (Kruskal and Wallis, 1952). Potential associations between environmental variables and indexes were further evaluated by Pearson correlations (Pearson, 1896). All statistical analyses were performed in Statistic 8.0 software and differences were considered statistically significant with a p value b 0.05.

2.3. Biological variables

3.1. Physico-chemical, biological variables and diversity indexes of marine sediments

Benthic fauna was collected after sieving from mesh sizes ranging from 4 to 0.5 mm and stored following the methodology proposed by Sellanes et al. (2003). Sieving was performed in order to facilitate collecting all benthic fauna. The abundance of the most representative taxa in both impact and control sites throughout the year was determined by SIMPER (similarity of percentages, see Section 2.5 for details). 2.4. Diversity indexes Marine macroinvertebrate data collected from all samplings during the year was included in an abundance matrix. Community structure and therefore environmental status was evaluated by using the following indexes: Species richness (R: total number of species in the community), Shannon-Wiener index (H´: describes the community based on species richness and abundance; Shannon and Weaver, 1949), Diversity of Simpson 1 − λ′ (Species diversity in a community; Thomas, 2009), Dominance of Simpson (describes species dominance; Simpson, 1949) and Pielou evenness index (compares observed diversity against the maximum expected diversity; Pielou, 1969). The ecological status was further evaluated using the AZTI Marine Biotic Index (AMBI) (Glémarec and Hily, 1981; Hily, 1984; Hily et al., 1986; Majeed, 1987). Briefly, species were classified in one of the proposed groups following the guidelines described and software AMBI available from AZTI-Tecnalia (www.azti.es), which includes 4466 taxa worldwide. The AMBI index was then independently calculated for each zone, sampling time and treatment (farming cages and control). The index was calculated using the Biotic Coefficient (Borja et al., 2000). The multivariate index M-AMBI (Muxika et al., 2007) was also determined, evaluating the ecological status using a factorial analysis previously proposed by Bald et al. (2005) by integrating three indexes 1) species richness (S), 2) Shannon-Wiener diversity (H′) and 3) AMBI index. 2.5. Statistical analysis Data is presented as means ± standard deviation (SD). The macroinvertebrate species where abundance during the year was used as biological variable were chosen considering their representativeness, which was evaluated using SIMPER (Similarity Percentages) (PRIMER, Clarke and Warwick, 1994). This analysis evaluates the similarities and dissimilarities between farming cages and control sites. Based on those results the two taxa explaining the biggest proportion of the similarities and dissimilarities between farming cages and control sites were used. Potential differences in the environmental variables and indexes evaluated during the year between farming cages and control sites were analyzed using a two-way ANOVA based on farming conditions (farming cages and control) and season (September–October, Spring I; November–December, Spring II; January–February, Summer I; March– April, Summer II; May–June, Autumn; and July–August, Winter) as factors. Data was then subjected to Tukey post hoc analysis. Normal distribution and homogeneity of variances was previously tested in all data

3. Results

In the exposed zone differences between controls and farming sites were evident in organic nitrogen, phosphorus, sulphurs, redox, abundance of nematodes and N. gayi, richness, Shannon's diversity, Diversity of Simpson, Pielou, dominance of Simpson, Ambi, and M-Ambi (Tables 2 and 4). From these variables and indexes, data for seasons only showed a significant effect on sulphurs, species richness, Shannon's diversity and Dominance of Simpson (Table 2). The interaction of these two factors (season and farming condition) was only significant for sulphurs, Redox and the abundance of nematodes. In the protected zone however, differences between controls and farming sites were evident in organic nitrogen, sulphurs, redox, abundance of nematodes, richness, Shannon's diversity, Pielou, Ambi, and M-Ambi (Tables 2 and 4). From these variables and indexes, season only showed a significant effect on abundance of nematodes, richness, and Shannon's diversity (Table 1), while the interaction between farming status and season was significant on sulphurs, redox, abundance of nematodes and richness (Table 2). In exposed zones differences between controls and farming sites were more evident and lasted during all four seasons in the dominance of Simpson, diversity of Simpson and Pielou evenness index (Tables 3 and 4). In protected zones, however, differences were more evident and lasted during all four seasons in Shannon diversity, Ambi, M-Ambi and Pielou evenness index (Tables 3 and 4). Simper analysis showed that the taxa that best described the impacts of salmon farming are nematode spp. and Nassarius gayi, representing ~53 and 22% of the accumulated similarity in the exposed zone, and ~ 80 and 9% of the accumulated similarity in the protected zone, respectively. The taxa that better described the control sites were nematode spp. (~ 12%) and Ostrácoda (~ 9%) in the exposed zone, and Chlamys patagonicus (~ 11%) and nematodes (~ 10%) in the protected zone. The abundance of the taxa that better described the impact and control sites in both zones are also the ones presenting the highest dissimilarity among them. In the exposed zone these taxa were represented by nematode spp., Nassarius gayi and Ostrácoda, and nematode spp., Nassarius gayi and Chlamys patagonicus in the protected zone. The polychaetes family Capetillidae sp. was also included in the analysis as it has been widely recognized as opportunistic and indicative of impacted marine sediments. Interestingly, the environmental variables and indexes that best described the salmon farming impacts differed between exposed and protected zones. In the exposed zone the most impacted of the physico-chemical variables, and also where the impacts lasted for longer, were phosphorus (Fig. 3C), sulphurs (Fig. 2A) and redox potential (Fig. 2C). On the other hand the most impacted taxa, and also where the impacts lasted for longer, were opportunistic nematode spp. (Fig. 4A) and Nassarius gayi (Fig. 4C). The indexes that consistently maintained statistical differences between farming cages and control sites throughout the whole year were the Simpson's dominance (Fig. 6E) followed by the evenness Pielou index (Fig. 6G).

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

845

Table 2 Results from a parametric two-way ANOVA for farming impact (I) and seasonality (E) as factors, and their interaction (I × E). F value and their probability (indicated by asterisks) for each variable and environmental index in exposed and protected zones. Exposed zone

I

E

I×E

Protected zone

I

E

I×E

Indicator

F

F

F

Indicator

F

F

F

N (mg/kg)a P (mg/kg)b C.O.T. (%)a Sulphur (mV)a Redox (mV) Nematodac Richnessc Shannon diversityc Dominance of Simpsonc

14.74⁎⁎⁎ 55.61⁎⁎⁎ 0.01 56.22⁎⁎⁎ 66.84⁎⁎⁎ 141.03⁎⁎⁎ 48.25⁎⁎⁎ 136.38⁎⁎⁎ 207.72⁎⁎⁎

1.34 2.20 7.53⁎⁎⁎ 8.31⁎⁎⁎

0.94 1.24 2.73⁎ 2.49⁎ 5.64⁎⁎⁎ 6.53⁎⁎⁎

N (mg/kg)b M.O. (%)b C.O.T. (%)a Sulphur (mV)a Redox (mV) Nematodaa Richnessa Shannon diversity

11.50⁎⁎ 2.96 2.71 225.89⁎⁎⁎ 285.50⁎⁎⁎ 99.79⁎⁎⁎ 86.88⁎⁎⁎ 267.43⁎⁎⁎

0.08 0.43 0.35 0.38 1.59 5.27⁎⁎ 4.72⁎⁎ 11.43⁎⁎⁎

0.32 0.38 0.74 6.11⁎⁎⁎ 12.69⁎⁎⁎ 5.8⁎⁎⁎ 3.99⁎⁎

1.60 0.69 4.19⁎ 4.97⁎⁎ 5.73⁎⁎⁎

1.58 1.58 1.75

1.02

Data are transformed where appropriate. N: organic nitrogen; P: phosphorus; C.O.T (%): percentage of total organic carbon; M.O. (%): percentage of organic matter. ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001. a Fourth root. b 1/x. c Log 2 (x + 1).

In the protected zone the most impacted of the physico-chemical variables were the same as in the exposed zone although their order of importance varied. In order of importance, the most affected physico-chemical variable and where the impacts lasted for longer were the redox potential (Fig. 2D), sulphurs (Fig. 2B) and phosphorus (Fig. 3D). Differences in biological variables were also evident, for example, the taxa nematode consistently reflected the impacts of salmon farming in the protected zone (Fig. 4B). In order of importance, the indexes that best reflected the impacts of salmon farming on the protected zone were the Shannon's diversity (Fig. 5D), Ambi (Fig. 6B), M-Ambi (Fig. 6D) and Pielou evenness index (Fig. 6H). Important temporal changes were also evident in both zones, exposed and protected. For example in the farming cages, sulphur levels

increased during Autumn and Winter (March–August) in the exposed zone, while they increased in Spring and Summer (September–April) in the protected zone (Fig. 2A,B). In the control stations, the redox potential increased during Summer in the exposed zone, while it remained with no change in the protected zone. Phosphorus and nitrogen were quite variable all year around in the farming cages in both exposed and protected zones, phosphorus only showing a clear decrease in summer in the exposed zone (Fig. 3 A, B, C, D). All other physico-chemical variables were quite variable across seasons, not showing a clear pattern. The abundance of nematods spp., N. gayi and capetilidae spp., showed no major changes across seasons in the control sites, but it was very variable in the farming cages (Fig. 4). The only consistent seasonal change observed in the farming cages was an increased

Table 3 Tukey post hoc analyses to determine significant differences between farming and control sites during the year in the environmental variables and indexes measured in exposed and protected zones. Table 2 shows F value and probability (indicated by asterisks). Exposed zones

Farming impact

Tukey post hoc analysis

Variable

F

SO

ND

JF

MA

MJ

JA

N (mg/kg)a P (mg/kg)b C.O.T. (%)a Sulphur (mV)a Redox (mV) Nematodac Richnessc Shannon diversityc Dominance of Simpsonc

14.74⁎⁎⁎ 55.61⁎⁎⁎ 0.01 56.22⁎⁎⁎ 66.84⁎⁎⁎ 141.03⁎⁎⁎ 48.25⁎⁎⁎ 136.38⁎⁎⁎ 207.72⁎⁎⁎

ns. ns. ns. ns. ns. ns. 0.0140⁎ 0.0232⁎ 0.0126⁎

ns. 0.0157⁎ ns. ns. ns. 0.0002⁎⁎⁎ ns. ns. 0.0023⁎⁎

ns. ns. ns. ns. 0.0012⁎⁎ ns. ns. 0.0006⁎⁎⁎ 0.0001⁎⁎⁎

ns. ns. ns. 0.0037⁎⁎ 0.0001⁎⁎⁎ 0.0001⁎⁎⁎ ns. 0.0011⁎⁎ 0.0001⁎⁎⁎

ns. 0.0347⁎ ns. 0.0190⁎ ns. 0.0002⁎⁎⁎ ns. 0.0024⁎⁎ 0.0001⁎⁎⁎

ns. 0.0302⁎ ns. 0.0028⁎⁎ ns. 0.0001⁎⁎⁎ 0.0029⁎⁎ 0.0001⁎⁎⁎ 0.0002⁎⁎⁎

Protected Zones

Farming impact

Tukey post hoc analysis

Variable

F

SO

ND

JF

MA

MJ

JA

N (mg/kg)b M.O. (%)b C.O.T. (%)a Sulphur (mV)a Redox (mV) Nematodaa Richnessa Shannon diversity

11.50⁎⁎ 2.96 2.71 225.89⁎⁎⁎ 285.89⁎⁎⁎ 99.79⁎⁎⁎ 86.88⁎⁎⁎ 267.43⁎⁎⁎

ns. ns. ns. 0.0001⁎⁎⁎ 0.0169⁎ ns. 0.0002⁎⁎⁎ 0.0001⁎⁎⁎

ns. ns. ns. 0.0001⁎⁎⁎ 0.0001⁎⁎⁎ 0.0002⁎⁎⁎ ns. 0.0001⁎⁎⁎

ns. ns. ns. 0.0001⁎⁎⁎ 0.0001⁎⁎⁎ 0.0001⁎⁎⁎ ns. 0.0001⁎⁎⁎

ns. ns. ns. 0.0001⁎⁎⁎ 0.0001⁎⁎⁎ ns. 0.0022⁎⁎ 0.0006⁎⁎⁎

ns. ns. ns. ns. ns. 0.0011⁎⁎ 0.0417⁎ 0.0001⁎⁎⁎

ns. ns. ns. ns. 0.0016⁎⁎ 0.0203⁎ 0.0004⁎⁎⁎ 0.0001⁎⁎⁎

n.s.: not significant. Data are transformed where appropriate. N: organic nitrogen; P: phosphorus; C.O.T (%): percentage of total organic carbon; M.O. (%): percentage of organic matter. Seasonality: SO: Sep–Oct; ND: Nov–Dec; JF: Jan–Feb; MA: Mar–Apr; MJ: May–Jun; JA: Jul–Aug. ⁎ p b 0,05. ⁎⁎ p b 0,01. ⁎⁎⁎ p b 0.001. a Fourth root. b 1/x. c Log 2 (x + 1).

846

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

Table 4 Kruskal-Wallis ANOVA for farming condition (salmon farming and control) as factor. H Value and their probability (indicated by asterisks) for the variables and environmental indexes in exposed and protected zones. Exposed Zone: Farming impact

Kruskal-Wallis (H)

Variable

SO

ND

JF

MA

MJ

JA

M.O. (%) Nassarius gayi Capetilidae Ostracodo Chlamys patagonicus Ambi M-Ambi Diversity of Simpson Pileou evenness index

3.00 5.53⁎ 0.00 2.33 2.33 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

0.08 5.40⁎ 0.00 3.50 1.00 3.00 3.00 5.33⁎ 5.33⁎

0.08 0.02 0.00 1.82 0.04 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

0.02 5.67⁎ 2.33 2.33 0.00 5.53⁎ 5.33⁎ 5.33⁎ 5.33⁎

0.19 1.10 2.29 6.22⁎ 0.10 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

0.00 6.14⁎ 0.39 6.40⁎ 1.00 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

Protected Zone: Farming impact

Kruskal-Wallis (H)

Variable

SO

ND

JF

MA

MJ

JA

3.00 0.63 0.00 0.00 4.08⁎ 5.33⁎

5.33⁎

5.33⁎

5.33⁎

5.33⁎

0.00 0.00 3.94⁎ 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

0.00 0.00 4.98⁎ 5.40⁎ 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

0.00 0.09 4.00⁎ 5.60⁎ 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

2.90 1.38 0.35 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

3.00 4.05⁎ 6.40⁎

P (mg/kg) Nassarius gayi Capetilidae Chlamys patagonicus Ambi M-Ambi Diversity of Simpson Dominance of Simpson Pielou evenness index

1.33 1.33 5.33⁎

2.33 5.40⁎ 5.33⁎ 5.33⁎ 5.33⁎ 5.33⁎

P: phosphorus; M.O. (%): percentage of organic matter. Seasonality: SO: Sep–Oct; ND: Nov–Dec; JF: Jan–Feb; MA: Mar–Apr; MJ: May–Jun; JA: Jul–Aug. ⁎ p b 0.05.

abundance of capetilidae spp. during Autumn and Winter in both exposed and protected zones (Fig. 4 E, F). Species richness decreased in the control sites during summer months in both zones, exposed and protected (Fig. 5. A, B), while its remained low all year around in the farming cages sites in both zones. A similar trend, but less marked was evident in Shannon´s diversity index in the control sites of both zones. In the farming cages sites, Sipmson's diversity index decreased in summer, a drop that was more dramatic in the protected zone than in the exposed zone (Fig. 5 E, F).

3.2. Association between physico-chemical, biological variables and environmental indexes Several strong correlations were found between physico-chemical, biological variables and environmental indexes in both protected and exposed zones (Table 5). The most significant relationships in the exposed zone among the physico-chemical variables were found between nitrogen and phosphorus (r = − 0.62), sulphurs and redox potential (r = − 0.60), and phosphorus and sulphurs (r = − 0.59). Between chemical and biological variables, the most significant associations found were between phosphorus and Nassarius gayi (r = 0.63), phosphorus and nematode, (r = 0.61), and redox potential and nematode (r = −0.63). Between taxa, the only significant association was found between nematode and Nassarius gayi (r = 0.68). Between taxa and indexes the only association found was between nematode and Simpson's diversity (r = 0.79). Among indexes, associations were found between Shannon and Simpson’ dominance (r = − 0.97), and between Simpson's dominance and Pielou evenness index (r = −0.95). In the protected zone more correlations were found between variables, the most significant were between nitrogen and organic matter (r = 0.91), nitrogen and phosphorus (r = − 0.65), phosphorus and redox potential (r = − 0.68), sulphurs and redox potential (r = − 0.86), phosphorus and sulphurs (r = 0.73), phosphorus and nematode, (r = 0.63), redox potential and nematode (r = − 0.61), nematode and Simpson's diversity (r = 0.91), Shannon's diversity and Simpson's dominance (r = − 0.88), and Simpson's dominance and Pielou evenness index (r = 0.90). The strong correlations found among and between the aforementioned variables and indexes in both protected and exposed zones not only support the results obtained from the SIMPER and ANOVA analysis, but it also indicates that the variables used in the present study to assess the impacts of salmon farming on marine sediments were properly chosen. 4. Discussion Overall, our results do support the idea that zones with different hydrographic characteristics (exposed vs protected) differ on the level and timescale where the impacts salmon farming on the sediments are evident, supporting the findings of Findlay and Watling (1997).

Fig. 2. Annual variability of physico-chemical variables on the sediments of salmon farming (solid circles) and control sites (open circles) in exposed and protected zones. Asterisks denote statistical differences between salmon farming and control sites.

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

847

Fig. 3. Annual variability of physico-chemical variables on the sediments of salmon farming (solid circles) and control sites (open circles) in exposed and protected zones. Asterisks denote statistical differences between salmon farming and control sites.

The present results confirm previous findings indicating that the physico-chemical variables that best reflect the impacts of salmon farming in the marine sediments during the year are the redox potential, sulphurs and phosphorus in both protected and exposed zones. Similar results have been reported in Tasmania (Edgar et al., 2005, 2010) and on the coast of Canada (Hargrave et al., 1993), where the variable that best reflects the effects of salmon farming on the marine sediments is the redox potential. Yet, to our knowledge, this is the first study to report that the order of importance of those variables changed between protected and exposed zones. In the exposed zone the variable that best evidenced the impacts of salmon farming is phosphorus followed by sulphur and redox potential, while in the protected zone it is exactly the opposite. The redox potential variable showed the greatest impact on salmon farming cages in the protected zone followed by sulphur and then phosphorus. This finding might well explain differences among studies (Findlay and Tenore, 1982; Li et al., 1997; Piepenburg

et al., 1997; Karakassis et al., 2000; Soto and Norambuena, 2004; Kalantzi and Karakassis, 2006) and also highlights the importance of considering the hydrodynamic characteristics of the site (see below). Our results also suggest that the percentage of organic carbon present in the sediment is not an accurate predictor of the temporal changes of the marine macro-benthic community. This supports previous findings from marine sediments close to salmon farms in Scotland (Pereira et al., 2004). In our study, we propose this could be due an elevated influx of organic matter associated with natural climatic conditions (prevalent elevated rainfall conditions). Fluvial inputs in the south of Chile are often rich in organic matter mainly from natural and terrestrial origins, and this condition prevails along the entire year. This organic matter is carried through the rivers to the sea where it sinks in proximity to the coast therefore potentially masking any potential input of organic matter from salmon farming activities.

848

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

Fig. 4. Annual variability in the abundance of different taxa on the sediments of salmon farming (solid circles) and control sites (open circles) in exposed and protected zones. Asterisks denote statistical differences between salmon farming and control sites.

Although organic matter content might not readily be a good predictor of the community composition, it might change other physicochemical characteristics that might strongly influence community structure. Both zones evaluated here (exposed and protected) had a similar level of salmon production and length of time under production and therefore it would be expected that both zones experienced similar inputs of organic matter from the salmon farming activity. Sites were also located further away from human settlements aiming to minimize potential anthropogenic inputs of organic matter. However, since the exposed zone had higher speed currents (1.29 ± 0.35 m s−1) and oxygen availability (7.2 ± 1.5 mg O2 L−1) than the sediments of the protected zone (0.37 ± 0.02 m s−1, 4.5 ± 1.0 mg O2 L−1), the exposed zone might be better able to recycle the nutrients inputs from the salmon farming activities. Oxygen undoubtedly is crucial to aquatic life (Hochachka, 1980; Urbina and Glover, 2013) and its availability is likely to cause differences not only in the degradation rates of organic matter, but also in the macro faunal composition. In fact changes at community structure and escape behavioral responses to hypoxia have been commonly documented in fish (Urbina et al., 2011), but also in marine benthic invertebrates (Riedel et al., 2008). In fact, a recent study found no effect of the hydrodynamic regime on the ecosystem functioning in the sediments beneath salmon farming installations (Sweetman et al., 2014), but did show benthic community changed in abundance, composition and biomass.

In both protected and exposed zones it was found that the abundance of the opportunistic species belonging to the phylum nematode are the best to describe the impacts of salmon farming on marine sediments. Interestingly, the opportunistic snail Nassarius gayi also coexisted with the opportunistic nematodes in the exposed zone, but its abundance was a lot lower in the protected zone. This observation might be related to the oxygen concentrations found in both zones. The genus Nassarius sp. has been described to be very efficient inhabiting sites rich in fresh organic matter (labile), yet it is less tolerant to sulphurs and hypoxia than nematodes spp. (Crawford et al., 2002). Nematodes however, are well recognized as hypoxia-tolerant species (Warwick and Price, 1979; Gambi et al., 2009; Armenteros et al., 2010) allowing them to inhabit and exploit the rich organic matter sediments under hypoxic conditions. Although both species are mobile and therefore able to seek out and utilize better living conditions (Armenteros et al., 2010; Black et al., 2012), Nasarius sp. is considerably more active and less tolerant to hypoxia. Therefore, one hypothetical explanation for our observations would be that in protected zones the lower oxygen availability would impede Nasarius sp. to co-exist with nematodes which are more hypoxia tolerant, forcing Nasarius sp. to move to more oxygenated conditions competing with nematodes only in exposed zones. Evidence to support this also comes from an increase in the abundance of capetilidae spp. in the farming cages sites starting in Autumn and maintained through the Winter in both exposed and

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

849

Fig. 5. Annual variability of the diversity indexes on the sediments of salmon farming (solid circles) and control sites (open circles) in exposed and protected zones. Asterisks denote statistical differences between salmon farming and control sites.

protected zones. Winter conditions such as lower temperatures and higher winds also favor better mixing and oxygenation of the water column and therefore also oxic conditions near the sediments. Benthic meiofauna comprises several species of burrowing animals that inhabit marine sediments such as bivalves, echinoderms, polychaetes, crustaceans, ostracods, nematodes and protozoans (Giere, 2009). Of these invertebrates the most dominant taxa are nematode comprising individuals that generally range from 45 μm to 2.5 mm of length. Nematodes have been widely used as indicators of environmental disruption and pollution, mainly due its wide distribution, high abundance and taxonomic diversity (Higgins and Thiel, 1988; Bongers and Ferris, 1999). A previous study carried out in Concepción bay, Chile, showed that nematodes are particularly prolific in sediments rich in organic matter and poor in dissolved oxygen. Under such conditions nematodes reached abundances as high as 95% of the total community playing a crucial role in the degradation and recycling of organic matter (Sellanes et al., 2003). However, not all nematodes are tolerant to hypoxia. A previous study evaluating the effects of marine fish farming cages in the Gulf of Creta by changes in species density, community structure and size of nematodes, revealed deep impacts at community level (Mirto et al., 2002). The authors identified that species belonging to genus Setosabatieria, Latronema and Elzalia were sensitive to fish farming pollution while species belonging to genus Sabatieria, Dorylaimopsis and Oxystomina were opportunistic and succeeded in impacted sites. This study not only found that species richness decreased in farming cages sites compared to control sites, but it also that the

genus of nematodes dominating in farming cages sites were bigger in size. Our results support these findings as we found nematodes bigger than 0.5 mm in the farming cages sites, but not in the controls sites. Other studies have also reported the presence nematodes belonging to a genus of bigger size in sediments rich in organic matter compared to controls sites (Moore and Bett, 1989; Tsujino, 1998; Porter et al., 1996). It has been previously found that the dominant genus in sediments impacted for salmon farming activities in Australia is Capitelidae sp., a polychaete, an observation that the authors attributed to its tolerance to low redox potential (Crawford et al., 2002). In the present study, however, we found that the dominant taxon in the farming cages sites was nematodes. Interestingly, the authors also found that in farming sites located in exposed zones the dominant species was Nassarius nigellus (Crawford et al., 2002). A similar result was found in the present study, and in fact, we found that the second more abundant species in the sediments of exposed zones was Nassarius gayi, a snail belonging to the same genus to that reported in Australia (Crawford et al., 2002). The effects of salmon farming on the marine sediments of both protected and exposed zones were best reflected in an increase in the Simpson's dominance index, and a decrease in the Shannon diversity and Pielou evenness index in agreement with previous studies (Kalantzi and Karakassis, 2006). Other indexes, such as Ambi, M-Ambi and Simpson's diversity, also provided evidence of the impacts of salmon farming in the present study, yet with less significance. In the present study most impacted sites scored values around 2 for the Shannon's diversity index compared to values close to 4 found in control sites.

850

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

Fig. 6. Annual variability of the diversity indexes on the sediments of salmon farming (solid circles) and control sites (open circles) in exposed and protected zones. Asterisks denote statistical differences between salmon farming and control sites.

Similar results were also found in Scotland where Shannon's diversity index was always lower than 3 in the salmon farming sites during the period of study (Pereira et al., 2004). Simpson's diversity, however, evidenced a very marked seasonal decrease in summer in both zones, but this decrease was considerably larger in the protected than in the exposed zone. The present results also show a strong seasonal pattern, with negative impacts from salmon farming increasing between the end of Spring and Summer. This pattern was further altered between protected and exposed zones. In the protected zone the detrimental effects of salmon farming increased between late Spring and early Summer, while in the exposed zone the salmon farming impacts reached its maximum only during Summer months (January and February). This temporal lag in the maximum impact on marine sediments suggest that exposed zones are more resilient to the effects of salmon farming than protected

zones where worst conditions were reached earlier. Protected zones have low or moderated hydrodynamic characteristics which seem to be crucial in defining the maximum carrying capacities of a farming site based on the present results. Furthermore, the observed temporal differences found in the present study between protected and exposed zones also agreed with our previously proposed hypothesis of oxygen availability. The solubility of oxygen in water decrease as temperature increase (Benson and Krause, 1984). Therefore, as protected zones are characterized by lower speed currents and therefore less oxygen exchange, the quality of the sediments is more sensitive to changes in the oxygen dissolution in the water column as a result of the first increases in temperature during spring time. Conversely, the reductions in the oxygen availability as consequence of decreases in the oxygen dissolution during spring and summer can be compensated by a higher exchange rates at the sediments in the exposed zones as a result of

Table 5 Correlation matrix (Pearson's product-moment correlation coefficient r) for all variables and indexes used in both exposed and protected zones.

Correlation matrix a

1 −0.62⁎⁎⁎ 0.56⁎⁎ 0.37⁎⁎ 0.4⁎⁎ −0.2 0.35⁎ 0.44⁎⁎ 0.02 −0.55⁎⁎⁎ 0.11 0.31⁎ −0.39⁎⁎ −0.39⁎⁎ −0.44⁎⁎ −0.39⁎⁎ 0.43⁎⁎ −0.35⁎

P 1 −0.13 −0.18 −0.59⁎⁎⁎ 0.48⁎⁎⁎ −0.61⁎⁎⁎ −0.63⁎⁎⁎ −0.2 0.49⁎⁎⁎ −0.02 −0.5⁎⁎⁎ 0.53⁎⁎⁎ 0.44⁎⁎ 0.54⁎⁎⁎ 0.52⁎⁎⁎ −0.58⁎⁎⁎ 0.51⁎⁎⁎

M.O.

C.O.T.

Sulphur

Redox

Nematoda

N. gayi

1 0.25 0.24 0.09 −0.16 0.04 −0.05 −0.17 0.17 −0.15 0 −0.13 −0.05 −0.01 0.01 0.05

1 0.07 0.08 0.07 0.12⁎ −0.32 −0.01 0.11 −0.1 0.1 0.06 0.08 0.08 −0.1 0.1

1 −0.6⁎⁎⁎ 0.47⁎⁎⁎ 0.51⁎⁎⁎ 0.4⁎⁎ −0.4⁎⁎ −0.02 0.51⁎⁎⁎ −0.46⁎⁎⁎ −0.32⁎ −0.41⁎⁎ −0.37⁎⁎ 0.45⁎⁎ −0.41⁎⁎

1 −0.63⁎⁎⁎ −0.45⁎⁎ −0.41⁎⁎ 0.36⁎ 0.01 −0.71⁎⁎⁎ 0.62⁎⁎⁎ 0.42⁎⁎ 0.59⁎⁎⁎ 0.62⁎⁎⁎ −0.66⁎⁎⁎ 0.61⁎⁎⁎

1 0.68⁎⁎⁎ 0.3⁎ −0.55⁎⁎⁎ −0.2 0.76⁎⁎⁎ −0.74⁎⁎⁎ −0.53⁎⁎⁎ −0.73⁎⁎⁎ −0.74⁎⁎⁎ 0.79⁎⁎⁎ −0.77⁎⁎⁎

1 0.37⁎⁎ −0.49⁎⁎ −0.02 0.64⁎⁎⁎ −0.61⁎⁎ −0.51⁎⁎ −0.56⁎⁎ −0.44⁎⁎ 0.52⁎⁎⁎ −0.45⁎⁎

Capetilidae

1 −0.32⁎ 0.21 0.45⁎⁎ −0.33⁎ −0.22 −0.22 −0.16 0.23 −0.18

Ostracodo

C. patagonicus

Ambi

M-Ambi

Richness

D Shannon

D. Simpson

Dom. Simpson

Pielou

1 0.02 −0.52⁎⁎⁎ 0.57⁎⁎⁎ 0.48⁎⁎⁎ 0.59⁎⁎⁎ 0.57⁎⁎⁎ −0.57⁎⁎⁎ 0.53⁎⁎⁎

1 −0.14 0.11 0.15 0.01 −0.11 0.05 −0.14

1 −0.9⁎⁎⁎ −0.69⁎⁎⁎ −0.8⁎⁎⁎ −0.74⁎⁎⁎ 0.79⁎⁎⁎ −0.7⁎⁎⁎

1 0.91⁎⁎⁎ 0.96⁎⁎⁎ 0.85⁎⁎⁎ −0.92⁎⁎⁎ 0.75⁎⁎⁎

1 0.88⁎⁎⁎ 0.69⁎⁎⁎ −0.77⁎⁎⁎ 0.49⁎⁎⁎

1 0.94⁎⁎⁎ −0.97⁎⁎⁎ 0.84⁎⁎⁎

1 −0.98⁎⁎⁎ 0.95⁎⁎⁎

1 −0.92⁎⁎⁎

1

Data are transformed where appropriate. a: fourth root; b: 1/x; c: Log 2 (x + 1). Exposed Correlation matrix

N

P

N (mg/kg)a P (mg/kg) MO (%)a C.O.T. (%)b Sulphurb Redox Nematodab Nassarius gayi Capetilidae Chlamys patagonicus Ambi M-Ambi Richnessb D. Shannon D. Simpson Dom. Simpson Pielou Protected

1 −0.65⁎⁎⁎ 0.91⁎⁎⁎ −0.83⁎⁎⁎ −0.67⁎⁎⁎ 0.48⁎⁎⁎ −0.48⁎⁎⁎

1 −0.56⁎⁎⁎ 0.55⁎⁎⁎ 0.73⁎⁎⁎ −0.68⁎⁎⁎ 0.63⁎⁎⁎

−0.16 −0.29⁎ 0.24 −0.56⁎⁎⁎ 0.41⁎⁎ 0.22 0.33⁎ 0.4⁎⁎ −0.4⁎⁎ 0.33⁎

0.09 0.24 −0.26 0.67⁎⁎⁎ −0.63⁎⁎⁎ −0.47⁎⁎⁎ −0.6⁎⁎⁎ −0.62⁎⁎⁎ 0.62⁎⁎⁎ −0.57⁎⁎⁎

M.O.

1 −0.9⁎⁎⁎ −0.49⁎⁎⁎ 0.26 −0.32⁎ −0.04 −0.29⁎ 0.15 ⁎1 0.19 0.05 0.11 0.2 −0.2 0.14

C.O.T.

1 0.45⁎⁎ −0.32⁎ 0.28 0.03 0.17 −0.22 0.33⁎ −0.15 0 −0.08 −0.19 0.19 −0.13

Sulphur

Redox

Nematoda

1 −0.86⁎⁎⁎ 0.67⁎⁎⁎ 0.2 0.24 −0.39⁎⁎ 0.82⁎⁎⁎ −0.77⁎⁎⁎ −0.55⁎⁎⁎ −0.72⁎⁎⁎ −0.71⁎⁎⁎ 0.71⁎⁎⁎ −0.72⁎⁎⁎

1 −0.61⁎⁎⁎ −0.16 −0.1 0.44⁎⁎ −0.81⁎⁎⁎ 0.77⁎⁎⁎ 0.58⁎⁎⁎ 0.72⁎⁎⁎ 0.72⁎⁎⁎ −0.72⁎⁎⁎ 0.7⁎⁎⁎

1 0.17 0.11 −0.33⁎ 0.78⁎⁎⁎ −0.73⁎⁎⁎ −0.44⁎⁎ −0.74⁎⁎⁎ −0.91⁎⁎⁎ 0.91⁎⁎⁎ −0.84⁎⁎⁎

N. gayi

1 0.37⁎⁎ −0.1 0.39⁎⁎ −0.3⁎ −0.21 −0.2 −0.18 0.18 −0.11

Capetilidae

1 0.2 0.36⁎ −0.19 −0.1 −0.04 0.01 −0.01 0.07

C. patagonicus

1 −0.51⁎⁎⁎ 0.37⁎⁎ 0.19 0.32⁎ 0.35⁎ −0.35⁎ 0.39⁎⁎

Ambi

M-Ambi

Richness

Shannon

D. Simpson

Dom. Simpson

Pielou

1 −0.87⁎⁎⁎ −0.58⁎⁎⁎ −0.76⁎⁎⁎ −0.81⁎⁎⁎ 0.81⁎⁎⁎ −0.74⁎⁎⁎

1 0.88⁎⁎⁎ 0.97⁎⁎⁎ 0.86⁎⁎⁎ −0.86⁎⁎⁎ 0.85⁎⁎⁎

1 0.87⁎⁎⁎ 0.63⁎⁎⁎ −0.63⁎⁎⁎ 0.64⁎⁎⁎

1 0.88⁎⁎⁎ −0.88⁎⁎⁎ 0.92⁎⁎⁎

1 −1⁎⁎⁎ 0.9⁎⁎⁎

1 −0.9⁎⁎⁎

1

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

N (mg/kg) P (mg/kg)b MO (%) C.O.T. (%)a Sulphura Redox Nematodac Nasarius gayi Capetilidae Ostra codo Chlamys patagonicus Ambi M-Ambi Richnessc D. Shannonc D. Simpson Dom. Simpsonc Pielou

N

Data are transformed where appropriate. a: 1/x; b: fourth root. Protected ⁎ p b 0.05. ⁎⁎ p b 0.01. ⁎⁎⁎ p b 0.001.

851

852

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853

higher speed currents. Therefore, the impacts of salmon farming on marine sediments is detected earlier in the protected zones than in an exposed zones. Most indexes also varied temporally as a result of a cascade of causeand-effect changes. A decrease in the Shannon's diversity and an increase in the dominance of opportunistic species are the first signs of the environmental impacts of salmon farming and two months later the redox potential reached its lowest point. Likewise, sulphurs follow the same pattern as the redox potential increasing between two and four months following a decrease in diversity. In both protected and exposed zones the lowest impact took place at the beginning of spring characterized by the highest diversity, the lowest value of the Simpson's dominance and a redox potential closest to the control site. Similar results have been found in Scotland where the best redox values in sediments under salmon farming cages were found in winter followed by a decrease in summer (Brown et al., 1987). Several correlations between the variables measured were also found in both protected and exposed zones. For example, the abundance of the phylum nematode was tightly correlated with the redox potential and phosphorus suggesting that these opportunistic species proliferate well in hypoxic sediments rich in organic matter. This finding also agrees with previous studies (Findlay and Tenore, 1982; Li et al., 1997; Piepenburg et al., 1997). In the protected zone we found a strong correlation between the nitrogen and the percentage of organic matter. This might be related with low hydrodynamic characteristics and likely low dissolved oxygen in agreement with previous studies (Snelgrove and Butman, 1994). In agreement with the present results, the effects of salmon farming at biological level have also been correlated with physico-chemical variables in British Columbia (Canada) (Brooks, 2001). The authors, by using simple Pearson correlations, showed a strong correlation between organic matter, abundance and species richness and sulphurs and redox potential. An extensive review on 41 articles integrating environmental data from fish farming cages from several parts of the world evaluated the correlations and regressions between 123 variables and biological indexes (Kalantzi and Karakassis, 2006). The authors found that the most important variables were the redox potential, organic nitrogen, total organic carbon, and dissolved oxygen in the sediments. These variables were also associated with several other independent variables such as depth and latitude. In agreement with our findings, the authors also warned caution in considering total organic carbon since it causal relationship could depend on several other variables (Kalantzi and Karakassis, 2006). The results from the present study not only agree with the variables that these authors identified as more accurately explaining causal relationships, but also support the idea that total organic carbon is not a good explanatory variable when evaluating the impacts of salmon farming. In the present study total organic carbon did not show any sign of negative impact of salmon farming, a results that could potentially be due to the constantly high natural inputs of organic carbon in the south of Chile. 5. Conclusion The present study shows that the impacts of salmon farming on marine sediments are dynamic, varying temporally and also strongly modulated by hydrodynamic characteristics. In both protected and exposed zones the impacts of salmon farming are first evidenced by a decrease in diversity and an increase in the dominance of opportunistic species (nematodes in this case). At the same time the salmon farming impacts are evidenced in physico-physico-chemical variables of the sediments such as redox potential and sulphurs. The effects of salmon farming on marine sediments varied temporally and this also varied depending on the hydrodynamic regime of the emplacement. The maximum impacts on marine sediments were found between late Spring and early Summer in the protected zone while the maximum impacts were evident during the second half of the Summer in the exposed zone. This

lag on the impacts showed in the exposed zone compared to the protected zone is likely the result of a bigger resilience in the exposed zone for oxidizing nutrients inputs. In both zones the lowest impacts took place in winter. We propose that the temporal variation and differences between protected and exposed zones are mainly related to the availability of oxygen in the marine sediments, which ultimately fuels the aerobic degradation of organic matter and determines community structure. These changes at community structure level also highlight the importance of specie-specific physiological capabilities, in this study for example, particularly related to aerobic/anaerobic metabolism and probably acid-base regulation. The results from the present study suggest that exposed zones are more resilient to the impacts of salmon farming and so proper regulations and monitoring programs should address those differences in order to guarantee sustainability and environmentally friendly fish farming practices. Acknowledgements I personally thank MSc. Diego Ortiz for gathering the original data of this study and to the reviewers for their feedback and constructive criticism. There is no conflict of interest. References Amos, C., Bergamasco, A., Umgiesser, G., Cappucci, S., Cloutier, D., Denat, L., Flindt, M., Bonardi, M., Cristante, S., 2004. The stability of tidal flats in Venice Lagoon – the results of in-situ measurements using two benthic, annular flumes. J. Mar. Syst. 51, 211–241. AOAC, 1980. Official Methods of Analysis. 13th ed. Association of official analytical chemists, Washington, D.C. USA. Armenteros, M., Perez, J., Ruiz, A., Diaz, L., Helguera, Y., Vincx, M., Decraemer, W., 2010. Effects of organic enrichment on nematode assemblages in a microcosm experiment. Mar. Environ. Res. 70, 374–382. Bald, J., Borja, A., Muxika, I., Franco, J., Valencia, V., 2005. Assessing reference conditions and physico-chemical status according to the European Water Framework Directive: a case-study from the Basque Country (Northern Spain). Mar. Pollut. Bull. 50, 1508–1522. Benson, B., Krause, D., 1984. The concentration and isotopic fractionation of oxygen dissolved in freshwater and seawater in equilibrium with the atmosphere. Limnol. Oceanogr. 29, 620–632. Beveridge, M., 1996. Cage Aquaculture. Second edition. Fishing News Book, Oxford 346 pp. Black, K., Calder, L., Nickell, T., Sayer, M., Orr, H., Brand, T., Cook, E., Magill, S., Katz, T., Eden, N., Kennet, J., Tsapakis, M., Angel, D., 2012. Chlorophyll, lipid profiles and bioturbation in sediments around a fish cage farm in the Gulf of Eilat, Israel. Aquaculture 356–357. Bongers, T., Ferris, H., 1999. Nematode community structure as a bioindicator in environmental monitoring. Trends Ecol. Evol. 14, 224–228. Borja, A., Franco, J., Perez, V., 2000. A marine biotic index to establish the ecological quality of soft-bottom benthos within European estuarine and coastal environments. Mar. Pollut. Bull. 40, 1100–1114. Brooks, K., 2001. An Evaluation of the Relationship between Salmon Farm Biomass, Organic Inputs to Sediments, Physicophysico-Chemical Changes Associated with those Inputs and the Infaunal Response—With Emphasis on Total Sediment Sulfides, Total Volatile Solids, and Oxidation–Reduction Potential as Surrogate Endpoints for Biological Monitoring. Available from Technical Advisory Group. Ministry of Environment, Nanaimo, BC, Canada. Brooks, K., Mahnken, C., 2003. Interactions of Atlantic salmon in the Pacific northwest environment II. Organic wastes. Fish. Res. 62, 255–293. Brown, J., Gowen, R., Mclusky, D., 1987. The effect of salmon farming on the benthos of a Scottish sea loch. J. Exp. Mar. Biol. Ecol. 109, 39–51. Buschmann, A., Pizarro, R., 2001. El costo ambiental de la salmonicultura en Chile. Análisis de políticas públicas número Vol. 5. Publicaciones Terram, pp. 5–8. Buschmann, A., Cabello, F., Young, K., Carvajal, J., Varela, D., Henríquez, L., 2009. Salmon aquaculture and coastal ecosystem health in Chile: analysis of regulations, environmental impacts and bioremediation systems. Ocean Coast. Manag. 52, 243–249. Byers, S., Mills, E., Stewart, P., 1978. A comparison of methods of determining organic carbon in marine sediments, with suggestions for a standard method. Hydrobiologia 58, 43–47. Clarke, K., Warwick, R., 1994. Change in Marine Communities: An Approach to Statistical Analysis and Interpretation. Natural Environment Research Council. Plymouth, United Kingdom, p. 144. Conover, W., Johnson, M.E., Johnson, M., 1981. A comparative study of tests for homogeneity of variances, with applications to the outer continental shelf bidding data. Technometrics 23, 351–361. Crawford, C., Macleod, C., Mitchell, I., 2002. Evaluation of Techniques for Environmental Monitoring of Salmon Farms in Tasmania. TAFI Technical Report Vol. 8. Tasmanian Aquaculture and Fisheries Institute, University of Tasmania, Hobart, Australia. Cromey, C., Nickell, T., Black, K., 2002. DEPOMOD – modelling the deposition and biological effects of waste solids from marine cage farms. Aquaculture 214, 211–239.

M.A. Urbina / Science of the Total Environment 573 (2016) 841–853 Droppo, I., Jaskot, C., Nelson, T., Milne, J., Charlton, M., 2007. Aquaculture waste sediment stability: implications for waste migration. Water Air Soil Pollut. 183, 59–68. Edgar, G., Macleod, C., Mawbey, R., Shields, D., 2005. Broad-scale effects of marine salmonid aquaculture on macrobenthos and the sediment environment in southeastern Tasmania. J. Exp. Mar. Biol. Ecol. 327, 70–90. Edgar, G., Davey, A., Shepherd, C., 2010. Application of biotic and abiotic indicators for detecting benthic impacts of marine salmonid farming among coastal regions of Tasmania. Aquaculture 307, 212–218. FAO: Food and Agriculture Organization of the United Nations, 2012n. El estado mundial de la pesca y acuicultura 2012. FAO: Food and Agriculture Organization of the United Nations, 2013n. FAO Yearbook. Vol. 2011. Fishery and Aquaculture Statistics. Findlay, S., Tenore, K., 1982. Effect of a free-living marine nematode (Diplolaimella chitwoodi) on detrital carbon mineralization. Mar. Ecol. Prog. Ser. 8, 161–166. Findlay, R.H., Watling, L., 1997. Prediction of benthic impact for salmon net pens based on the balance of benthic oxygen supply and demand. Mar. Ecol. Prog. Ser. 155, 147–157. Folke, C., Kautsky, N., 1989. The role of ecosystems for a sustainable development of aquaculture. Ambio 18, 234–243. Folke, C., Kautsky, N., Troell, M., 1994. The cost of eutrophication from salmon farming: implications for policy. J. Environ. Manag. 40, 173–182. Friend, P., Collins, M., Holligan, P., 2003. Day-night variation of intertidal flat sediment properties in relation to sediment stability. Estuar. Coast. Shelf Sci. 58, 663–675. Gambi, C., Bianchelli, S., Pérez, M., Invers, O., Ruiz, J., Danovaro, R., 2009. Biodiversity response to experimental induced hypoxic-anoxic conditions in seagrass sediments. Biodivers. Conserv. 18, 33–54. Giere, O., 2009. Meiobenthology. The Microscopic Motile Fauna of Aquatic Sediments, 2nd edition Springer. Glémarec, M., Hily, C., 1981. Perturbations apportées á la macrofaune benthique de la baie de Concarneau par les effluents urbains et portuaires. Acta Oecol. Oec. Appl. 2, 139–150. Hargrave, B., Duplisea, E., Pfiffer, E., Wildish, D., 1993. Seasonal changes in benthic fluxes of dissolved oxygen and ammonium associated with cultured Atlantic salmon. Mar. Ecol. Prog. Ser. 96, 249–257. Hartwell, S., 1998. Biological Habitat Quality Indicators for Essential Fish Habitat. Workshop Proceedings, 14–15 July 1997, Charleston, SC. NOAA Technical Memorandum NMFS/SPO-32. NOAA/NMFS, Silver Spring, MD. Hevia, M., Rosenthal, H., Gowen, R., 1996. Modeling benthic deposition under fish cages. J. Appl. Ichthyol. 12, 71–74. Higgins, R., Thiel, H., 1988. Introduction to the Study of Meiofauna. Smithsonian Institution Press, Washington, DC. Hily, C., 1984. Variabilite de la macrofaune benthique dans les milieus hypertrophiques de la Rade de Brest These de Doctorat d'Etat Univ. Bretagne Occidentale vol. 1 359 pp.; Vol. 2, 337 pp. Hily, C., Le Bris, H., Glemarec, M., 1986. Impacs biologiques desemissaires urbains sur les écosistemes benthiques. Oceanis 12, 419–426. Hochachka, P. 1980. Livingwithout Oxygen. Closed and Open Systems in Hypoxia Tolerance. 1st ed. Harvard University Press, Cambridge (MA). 181 pp. Jackson, M., 1964. Análisis químicos de los suelos. 2nd ed. Editorial Omega, Barcelona, España 666 pp. Kalantzi, I., Karakassis, I., 2006. Benthic impacts of fish farming: meta-analysis of community and geochemical data. Mar. Pollut. Bull. 52, 484–493. Karakassis, I., Tsapakis, M., Hatziyanni, E., Papadopoulou, K.-N., Plaiti, W., 2000. Impact of cage farming of fish on the seabed in three Mediterranean coastal areas. ICES J. Mar. Sci. 57, 1462–1471. Karthikeyan, S., Wolfaardt, G., Korber, D., Caldwell, D., 1999. Functional and structural responses of a degradative microbial community to substrates with varying degrees of complexity in chemical structure. Microb. Ecol. 38, 215–224. Kruskal, W., Wallis, W., 1952. Use of ranks in one-criterion variance analysis. J. Am. Stat. Assoc. 47, 583–621. Li, J., Vincx, M., Herman, P., 1997. Carbon flows through meiobenthic nematodes in the Westerschelde estuary. Fundam. Appl. Nematol. 20, 487–494. Madec, P., 2003. Les indicateurs de développement durable. INRA-Université de Montpellier II 118 pp. Majeed, S., 1987. Organic matter and biotic indices on the beaches of North Brittany. Mar. Pollut. Bull. 18, 490–495. Mayor, D.J., Zuur, A.F., Solan, M., Paton, G.I., Killham, K., 2010. Factors affecting benthic impacts at Scottish fish farms. Environ. Sci. Technol. 44, 2079–2084. Mirto, S., La Rosa, T., Gambi, C., Danovaro, R., Mazzola, A., 2002. Nematode community response to fish-farm impact in the western Mediterranean. Environ. Pollut. 116, 203–214.

853

Moore, C., Bett, B., 1989. The use of meiofauna in marine pollution impact assessment. Zool. J. Linnean Soc. 96, 263–280. Mulsow, S., Krieger, Y., Kennedy, R., 2006. Sediment profile imaging (SPI) and microelectrode technologies in impact assessment studies: examples from two fjords in Southern Chile used for fish farming. J. Mar. Syst. Muxika, I., Borja, A., Bald, J., 2007. Using historical data, expert judgment and multivariate analysis in assessing reference conditions and benthic ecological status, according to the European Water Framework Directive. Mar. Pollut. Bull. 55, 16–29. Pearson, K., 1896. Mathematical contributions to the theory of evolution – on the form of spurious correlation which may arise when indices are used in the measurements of organs. Proc. R. Soc. Lond. 60, 489–498. Pearson, T., Maita, Y., Lalli, C., 1984. A Manual of Chemical and Biological Methods for Seawater Analysis. Pergamon Press, Britain 169 pp. Pereira, P., Black, K., Mclusky, D., Nickell, T., 2004. Recovery of sediments after cessation of marine fish farm production. Aquaculture 235, 315–330. Pielou, E., 1969. An Introduction to Mathematical Ecology. Wiley-Interscience, Wiley and Sons 285 pp. Piepenburg, D., Ambrose, W., Brandt, A., Renaud, P., Ahrends, M., Jensen, P., 1997. Benthic community patterns reflect water column processes in the Northern Water polynya (Greenland). J. Mar. Syst. 10, 467–482. Porter, C., Krost, P., Gordin, H., Angel, D., 1996. Preliminary assessment of greymullet (Mugil cephalus) as a forager of organically enriched sediments below marine fish farms. Isr. J. Aquacult. Bamidgeh 48, 47–55. Riedel, B., Zuschin, M., Haselmair, A., Stachowitsch, M., 2008. Oxygen depletion under glass: behavioural responses of benthic macrofauna to induced anoxia in the Northern Adriatic. J. Exp. Mar. Biol. Ecol. 367, 17–27. Sellanes, J., Neira, C., Quiroga, E., 2003. Composición, estructura y flujo energético del meiobentos frente a Chile central. Rev. Chil. Hist. Nat. 76, 401–415. Shannon, C., Weaver, W., 1949. The Mathematical Theory of Communication. University Illinois Press, Urbana, IL. Simpson, E., 1949. Measurement of diversity. Nature 163, 688. Snelgrove, P., Butman, C., 1994. Animal-sediment relationships revisited: cause versus effects. Oceanogr. Mar. Biol. Annu. Rev. 32, 111–177. Sokal, R., Rohlf, F., 1995. Biometry. The Principles and Practice of Statistics in Biological Research, 3rd ed WH freman, New York. Soto, D., Norambuena, S., 2004. Evaluation of salmon farming effects on marine systems in the inner seas of southern Chile: a large-scale mensurative experiment. J. Appl. Ichthyol. 20, 493–501. Sweetman, A.K., Norling, K., Gunderstad, C., Haugland, B.T., Dale, T., 2014. Benthic ecosystem functioning beneath fish farms in different hydrodynamic environments. Limnol. Oceanogr. 59, 1139–1151. Tomassetti, P., Porrello, S., 2005. Polychaetes as indicators of marine fish farm organic enrichment. Aquac. Int. 13, 109–128. Tsujino, M., 1998. Distribution of free-living nematodes in relation to bottom environments in Hiroshima Bay. Bull. Nansei Natl. Fish. Res. 31, 25–32. UICN, 2009. Guía para el Desarrollo Sostenible de la Acuicultura Mediterránea 2. Acuicultura: Selección y Gestión de Emplazamientos. UICN. VIII, Gland, Suiza y Málaga, España (221-233), 332 pp.). Urbina, M.A., Glover, C., 2013. Relationship between fish size and metabolic rate in the oxyconforming inanga Galaxias maculatus reveals size-dependent strategies to withstand hypoxia. Physiol. Biochem. Zool. 86, 740–749. Urbina, M.A., Paschke, K., Gebauer, P., Chaparro, O.R., 2010. Physiological energetics of the estuarine crab Hemigrapsus crenulatus (Crustacea: Decapoda: Varunidae): responses to different salinity levels. J. Mar. Biol. Assoc. UK 90, 267–273. Urbina, M.A., Foster, M., Glover, C., 2011. Leap of faith: voluntary emersion behaviour and physiological adaptations to aerial exposure in a non-aestivating freshwater fish in response to aquatic hypoxia. Physiol. Behav. 103, 240–247. Walkley, A., Black, I., 1934. An examination of Degtjareff method for determining soil organic matter and a proposed modification of the chromic acid titration method. Soil Sci. 37, 29–37. Warwick, R., 1986. A new method for detecting pollution effects on marine macrobenthic communities. Mar. Biol. 92, 557–562. Warwick, R.M., Price, R., 1979. Ecological and metabolic studies on free-living nematodes from an estuarine mud-flat. Estuar. Coast. Mar. Sci. 9 (3), 257–271. Wilding, T., Cromey, C., Nickell, T., Hughes, D., 2012. Salmon farm impacts on muddysediment megabenthic assemblages on the west coast of Scotland. Aquacult. Environ. Interact. 2, 145–156. Wildish, D., Hargrave, B., Pohle, G., 2001. Cost effective monitoring of organic enrichment resulting from salmon mariculture. ICES J. Mar. Sci. 58, 469–476.