- Email: [email protected]
Sustainable impact of mussel farming in the Adriatic Sea (Mediterranean Sea): evidence from biochemical, microbial and meiofaunal indicators
Marine Pollution Bulletin 49 (2004) 325–333 www.elsevier.com/locate/marpolbul
Sustainable impact of mussel farming in the Adriatic Sea (Mediterranean Sea): evidence from biochemical, microbial and meiofaunal indicators R. Danovaro *, C. Gambi, G.M. Luna, S. Mirto Department of Marine Sciences, Faculty of Science, Polytechnic University of Marche, Via Brecce Bianche, Monte D’Ago, Ancona 60131, Italy
Abstract We have investigated the impact of a large mussel farm on the benthic environment using a battery of benthic indicators of environmental quality (including biochemical, microbial and meiofaunal parameters). These were analysed through a multi-control sampling strategy over one year. The differences across the seasons are typically higher than those between the impacted and the control stations. No effects are seen in terms of the sediment oxygen penetration and the downward fluxes (as the total mass, organic and phytopigment fluxes). The indicators based on the biochemical compositions of the sediment organic matter and the microbial parameters also show no evidence of the eutrophication process, except as a slight increase in the bacterial density in the sediments beneath the long-lines of the farm during the period of highest mussel stocks. Finally, no effects are observed in terms of the benthic faunal indicators, as the meiofaunal abundance, the community structure and the taxa richness are all indistinguishable between the farm sediments and the controls. These results show that mussel farming in the investigated system is eco-sustainable and does not significantly alter the coastal marine ecosystem, both in terms of the functioning and the trophic state. The battery of indicators selected in this study represents a useful tool for the monitoring of the potential ecological impact of mussel farms, towards guaranteeing the sustainable development of aquacultures in shallow coastal environments. 2004 Elsevier Ltd. All rights reserved. Keywords: Bio-indicators; Aquaculture; Mussel farms; Sediment organic matter; Microbial indicators; Meiofaunal indicators
1. Introduction Shellfish production in aquaculture is increasing significantly throughout the world (FAO, 1997). Considering only the Mediterranean Sea, mussel production is over 700,000 ton y1 (FAO, 2000). Shellfish farms are typically located in estuarine areas and/or inshore coastal waters, which harbour high population densities and a broad range of economic interests. As a consequence, the opposition to the expansion of shellfish farming is increasing, due to the growing concern about ecosystem degradation, and their potential conflicts with the tourism industry and coastal urbanization (Kaiser et al., 1998).
* Corresponding author. Tel.: +39-71-220-4654; fax: +39-71-2204650. E-mail address: [email protected] (R. Danovaro).
0025-326X/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2004.02.038
The impact of bivalve culturing, which is mostly due to mussel and oyster farms, is related to the intensive biodeposition of the faeces and pseudo-faeces that modify the physical and chemical characteristics of the benthic environment as they accumulate in the bottom sediments (Kaspar et al., 1985; Gilbert et al., 1997; Mirto et al., 1999). This phenomenon is well known and documented for intensive fin-fish aquaculture. However, the impact of mussel farming is expected to be less relevant than fish farming, because in the latter the impact due to the accumulation of biodeposits is further increased by the accumulation of organic matter due to uneaten food (Mazzola et al., 2000; Crawford et al., 2003). Nonetheless, a wide variety of negative effects have generally been reported for mussel farming since biodeposition can induce severe organic matter accumulation and reducing conditions in the sediments beneath the cages (Kaspar et al., 1985; Tenore et al., 1985;
326
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
Kaiser et al., 1998; Mirto et al., 2000). This, in turn, affects benthic biodiversity and community structure, altering trophic interactions and pathways of energy transfer from bacteria and picoeukaryotes (Mirto et al., 2000; Danovaro et al., 2003) to meiofauna (Dinet et al., 1990; Mirto et al., 2000); and macrofauna (Castel et al., 1989; Stenton-Dozey et al., 1999). Assessing the potential impact of mussel farming is extremely important for developing an eco-sustainable management of aquaculture (FAO, 2000). For this, we need to apply the widest possible range of scientific tools and environmental indicators that are able to detect all possible kinds of ecological alterations that can be caused by the aquaculture activities. Besides the historical use of bioindicators of impact based on macrofaunal assemblages, more recently new indicators have been proposed for identifying the short-term impacts of aquaculture. Indicators based on the biochemical composition of the sediment organic matter provide important insights into the eutrophication processes occurring in the area (Dell’Anno et al., 2002). Microbial indicators have been suggested for highlighting the alterations in the biogeochemical processes occurring in farm sediments (La Rosa et al., 2001). Finally, due to their fast generation times and their ability to respond quickly to any changes in environmental conditions, biological indicators based on meiofauna can reveal alterations at the community level. The simultaneous use of these indicators provides elements for an integrated approach through which we can cover various levels of biological organization of the ecosystem. In this study, we have investigated the impact of a large mussel farm located in the Adriatic sea (Western Mediterranean) using a sampling strategy based on a preliminary survey. This was then followed by a multicontrol approach to evaluate: (a) the effects of mussel biodeposition on the eutrophication processes; (b) the alterations in terms of the microbial compartment; (c) the effects on the faunal community structure.
2. Materials and methods 2.1. Study site and sampling The present study was conducted from June 2001 to February 2002 in a mussel farm located 1.5–2.0 miles from the coast, facing Cattolica-Rimini in the Middle Adriatic Sea (Central Mediterranean). The area covered by the farm was ca. 2 km2 and the average bottom depth was 11 m. The mussels were cultivated using a long-line technique, in which the mussel larvae settle on ropes hanging down from long horizontal anchored lines that are suspended by buoys. The locations of the sampling stations was defined after a preliminary survey, in which 25 stations were
Fig. 1. The sampling stations for the mussel farm located in the Adriatic Sea. The shadow box illustrates the extent of the mussel farm.
sampled in a grid covering the entire mussel farm area and surroundings up to 600 m around. This allowed the creation of a map of biodeposition accumulation, and according to a multi-control sampling strategy, we defined the position of 3 control stations outside the mussel farm (Control 1, 2 and 3; located ca. 600 m out of the farming area) and three impacted stations within the area of the long-lines of the mussel farm (Impact 1, 2 and 3; Fig. 1). The sediment samples were collected using a boxcorer and/or a grab; this was done in June (after the phytoplankton bloom), September (the period of long-line seeding), December 2001 (the period of highest mussel standing stocks) and February 2002 (the period of the highest mixing of the water column, and the lowest temperatures). Hereafter, these sampling periods will be termed spring, summer, autumn and winter, respectively. Five deployments were carried out at each station, and independent replicate cores were collected by different deployments. All of the cores were collected from the inner part of the box corer (using a plexiglass liner with an inner diameter of 4.7 cm). For the biochemical analyses, the surface layer of each core (the top 2 cm of the sediment) was homogenised in a Petri dish and stored at )20 C until analysis. To estimate the downward fluxes at each sampling time, two sediment trap moorings were deployed in correspondence to the Impact 1 and Control 3 stations. Each mooring had three sediment traps (Technicapmodel; PVC with an inner diameter of 13 cm and a height of 70 cm) placed at approximately 70 cm from the sea bottom. The trap deployments lasted 24 h at each sampling period, and the material collected was filtered immediately after recovery using GF/F glass-fibre filters (0.45 lm pore size), and stored at )20 C until analysis.
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
For the analysis of the microbiological variables, replicate sub-samples (n ¼ 3, ca. 1 ml) of surface sediment were withdrawn from each core and fixed with 4 ml of prefiltered (0.2 lm) and buffered 2% formalin. The samples were then stored at 4 C until the analysis, which was performed within 4 weeks of sampling. The meiofaunal samples were collected in replicate Plexiglas cores (n ¼ 3, 3.7 cm diameter, 10.7 cm2 surface area) down to a depth of 8 cm, and then fixed with 4% buffered formalin in 0.4 lm prefiltered seawater solution and stained with Rose Bengal (0.5 g l1 ). This depth appeared to be optimal for the quantitative analysis as the evaluation of the top 15 cm of the sediment core revealed that <1% of the total meiofaunal density was present in the lower 8–15 cm sediment horizon. 2.2. Environmental variables and biochemical indicators The redox potential discontinuity (RPD) layer was visually estimated as the depth at which the sediment colour turned from brown to grey-black. The grain size was determined according to standard protocols using a dry sieve technique. The total organic matter (TOM) was determined as the difference between the dry weight of the sediment (80 C, 24 h) and the residue left after combustion at 450 C for 4 h (Parker, 1983). Before the TOM analysis, the sediment samples were treated with an excess of 10% HCl to remove carbonates that may have interfered with the TOM determination. The biochemical analyses (protein, carbohydrate and lipid) and chloroplastic pigment determinations (chlorophyll-a and phaeopigments) of both the trap material and the surface sediments were carried out as described by Dell’Anno et al. (2002). For each parameter, the blanks used were from the same previously calcinated sediments (450 C, 2 h). All analyses were carried out in three replicates. The sediment organic matter concentrations were normalized to dry weight after desiccation at 60 C. The protein, carbohydrate and lipid concentrations were converted to their carbon equivalents, assuming the conversion factors of 0.49, 0.40 and 0.75 lgC lg1 , respectively (Fichez, 1991). The sum of the protein, carbohydrate and lipid carbon is referred as the biopolymeric carbon (BPC; Fabiano and Danovaro, 1994). 2.3. Microbial parameters The total bacterial counts were performed on subsamples that had been sonicated three times (Branson Sonifier 2200; 60 W for 1 min) and diluted 250–500 times using sterile (0.2 lm prefiltered) formalin (2% final concentration). An aliquot of 1 ml was stained with Acridine Orange (5 mg l1 final concentration, for 5 min in the dark) and filtered under vacuum (<100 mmHg) using 0.2 lm pore size black Nucleopore Polycarbonate
327
filters. The filters were then washed twice with 3 ml of 0.2 lm prefiltered and sterilized Milli-Q water, mounted on microscope slides, and analysed using epifluorescence microscopy (Zeiss Axioskop 2; magnification ·1000). For each slide, at least 10 microscope fields were observed, for a total of at least 400 cells counted per filter. The frequency of dividing bacterial cells (FDC, defined as cells with a clearly visible invagination) was also determined (Danovaro et al., 1999). All data were normalized to sediment dry weight after desiccation (60 C, 24 h). 2.4. Meiofauna For meiofaunal analyses, the sediments were sieved through 1000 and 37 lm mesh nets. The fraction remaining on the 37 lm mesh sieve was centrifuged three times with Ludox HS (density 1.18 g cm3 ; Heip et al., 1985). All meiobenthic animals were sorted, counted and classified into their taxa under a stereo microscope. 2.5. Statistical analyses The differences among the stations (impacted and control) and the sampling periods were analysed by means of two-way analysis of variance (ANOVA). When a significant difference for the main effect was observed ðp < 0:05Þ, a Tukey’s pairwise comparison test was also performed.
3. Results 3.1. Sediment characteristics The characteristics of the benthic environment (RPD, grain size) are reported in Table 1. The RPD depth varied widely between the seasons, ranging from 0.7 cm in summer to 6.5 cm in winter in the impacted stations, and from 1.5 cm in summer to 8.0 cm in winter in the control stations. No significant differences were observed between the RPD values comparing impacted and control stations (ANOVA; ns). The surface sediments were typically characterized by muddy sands, as they were composed of similar fractions of mud (on average 44.1%) and sand (on average 55.9%). 3.2. Fluxes The seasonal changes in the vertical fluxes in the areas subjected to mussel biodeposition (as total and organic matter flux and chloropigment fluxes) and in the control sites are illustrated in Fig. 2. The total mass flux ranged from 4.5 ± 1.5 to 17.0 ± 5.5 g m2 d1 in the impacted
328
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
Table 1 Sediment parameters in the investigated sites Redox potential discontinuity
Station
Spring (cm)
Summer (cm)
Autumn (cm)
Impact 1 Impact 2 Impact 3
2.2 5.5 4.5
1.8 2.3 0.7
4 2 3.5
Average
4.1
1.6
Control 1 Control 2 Control 3
5.5 6.7 3.8
Average
5.3
Grain size Winter (cm)
Spring
Summer
Autumn
Winter
Mud (%)
Sand (%)
Mud (%)
Sand (%)
Mud (%)
Sand (%)
Mud (%)
Sand (%)
6.5 4.3 4
15.8 15.2 32.0
84.2 84.8 68.0
68.2 34.9 69.3
31.8 65.1 30.7
48.0 15.8 70.1
52.0 84.2 29.9
77.1 34.0 70.0
22.9 66.0 30.0
3.2
4.9
21.0
79.0
57.5
42.5
44.6
55.4
60.4
39.6
4.7 1.5 2.7
4.3 4.5 3.3
15.7 8 6.7
8.3 34.2 50.8
91.7 65.8 49.2
18.9 86.4 65.8
81.1 13.6 34.2
5.7 24.9 50.1
94.3 75.1 49.9
18.4 69.7 75.4
81.6 30.3 24.6
3.0
4.0
6.8
31.1
68.9
57.1
42.9
26.9
73.1
54.5
45.5
The redox potential discontinuity depth and grain size composition (as the sand and mud fraction contributions, expressed as percentages) across the four sampling periods.
Total mass flux
25
Control
Impact
g m-2 d-1
20 15 10 5 na 0
(a)
spring
autumn
winter
Organic matter flux
1600
mg m-2 d-1
summer
1200 800 400
na
(b)
0 spring
summer
autumn
winter
Chloropigment flux
7
mg m-2 d-1
6 5 4 3 2 1
(c)
ences were observed in fluxes between the impacted and the control stations (ANOVA; ns). The organic matter flux in the mussel farm area was quite constant across the sampling periods (range: 947.6 ± 379.3 to 1146.5 ± 270.2 mg m2 d1 ), although in the control stations it was more variable (range: 499.0 ± 117.6 to 968.2 ± 216.2 mg m2 d1 ), and followed a pattern similar to that observed for the total mass flux. No significant differences were observed between the organic fluxes measured in the mussel farm and those measured in the control sites (ANOVA; ns). The total chloropigment fluxes ranged from 0.91 ± 0.06 to 5.97 ± 0.28 mg m2 d1 in the impacted stations (in spring and winter, respectively), and from 0.32 ± 0.05 to 4.06 ± 0.52 mg m2 d1 in the control stations (in spring and winter, respectively). In winter and spring, total chloropigment fluxes were significantly higher in the mussel station (ANOVA; p < 0:01).
0
na spring
summer
autumn
winter
Fig. 2. Downward fluxes of (a) total mass, (b) organic matter, and (c) chloropigments. Standard deviations are shown. na ¼ not available.
stations (in winter and spring, respectively), and from 6.9 ± 1.8 to 9.4 ± 2.9 g m2 d1 in the control stations (in spring and autumn, respectively). No significant differ-
3.3. Phytopigment and organic matter concentrations in the sediment The phytopigment concentrations (chlorophyll-a and phaeopigments) are shown in Fig. 3. The chlorophyll-a sediment concentrations ranged from 1.0 ± 0.2 to 11.2 ± 2.5 lg g1 in the impacted stations (in autumn and summer, respectively), and from 1.4 ± 0.1 to 8.3 ± 0.5 lg g1 in the control stations (in autumn and summer, respectively). The phaeopigment concentrations ranged from 2.5 ± 0.6 to 35.1 ± 3.2 lg g1 in the impacted stations (in autumn and summer, respectively), and from 2.6 ± 0.0 to 26.2 ± 17.5 lg g1 in the control stations (in winter, at the Control 1 and 2 stations, respectively). Similar temporal patterns were displayed by the total chloropigment concentrations. The concentrations of chlorophyll-a and phaeopigments in both
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
the impacted and the control stations were significantly higher in summer than in all of the other sampling periods (ANOVA; p < 0:01), but they did not show differences between the impacted and the control sites themselves. The total organic matter (TOM), protein, carbohydrate, lipid and biopolymeric carbon (BPC) concentrations are shown in Fig. 4. The TOM ranged from 1.2 ± 0.1 to 6.1 ± 3.3% of sediment dry weight (DW) in the impacted stations (in summer and winter, respectively), and from 1.7 ± 0.1 to 7.8 ± 0.8% in the control stations (in summer and spring, respectively). Significant differences were observed only in spring, when the organic matter content was lower in the impacted than in the control sites (ANOVA; p < 0:01). Protein and carbohydrate were the two main biochemical classes of organic compounds. The protein ranged from 717.3 ± 113.9 to 2581.9 ± 457.0 lg g1 in the impacted stations (in autumn and winter, respectively), and from 646.0 ± 134.6 to 2999.1 ± 890.2 lg g1 in the control stations (in spring and winter, respectively). The total carbohydrate ranged from 372.1 ± 55.0 to 1983.4 ± 34.2 lg g1 in the impacted stations (in spring
µ
µ
µ
µ
Fig. 3. Phytopigment content in the sediments investigated. (a) Chlorophyll-a concentrations, and (b) phaeopigment content. Standard deviations are shown.
329
Fig. 4. The quantity and quality of the organic matter in the sediments investigated. (a) Total organic matter, (b) protein, (c) carbohydrate, (d) lipid, and (e) biopolymeric carbon concentrations. Standard deviations are shown.
330
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
and autumn, respectively), and from 461.3 ± 54.4 to 2313.1 ± 266.9 lg g1 in the control stations (in spring and winter, respectively). The lipid ranged from 198.5 ± 14.7 to 1073.7 ± 129.8 lg g1 in the impacted stations (in summer and winter, respectively), and from 113.7 ± 33.0 to 1033.5 ± 323.4 lg g1 in the control stations (both in winter, at Control 1 and 2 stations, respectively). None of the biochemical variables investigated displayed any significant differences between the impacted and control stations at each sampling period, nor between the sampling periods (ANOVA; ns). The BPC concentrations (expressed as the sum of the protein, carbohydrate and lipid carbon) ranged from 675.6 ± 112.3 to 2394.0 ± 317.1 lgC g1 in the impacted stations (in spring and summer, respectively), and from 639.2 ± 137 to 3049.9 ± 749.9 lgC g1 in the control stations (both in winter, at Control 1 and 2 stations, respectively). A significant temporal variability was observed for BPC, with the highest concentrations in winter in both the impacted and the control stations (ANOVA; p < 0:01). 3.4. Microbial parameters The total bacterial counts (TBC) and the frequency of dividing cells (FDC) are shown in Fig. 5. The TBC ranged from 3.53 ± 0.49 to 15.35 ± 2.19 · 108 cells g1 sediment DW in the impacted stations (in spring and autumn, respectively), and from 5.12 ± 0.08 to 10.85 ± 1.24 · 108 cells g1 in the control stations (in summer and spring, respectively). Significant differences were observed between sampling periods (ANOVA;
Fig. 5. Microbial parameters in the impacted and the control sediments. (a) Total bacterial counts, and (b) frequency of dividing cells. Standard deviations are shown.
p < 0:01), but no significant differences were observed between the impacted and the control sites, with the exception of the autumn sampling, in which the bacterial abundance was significantly higher in the impacted sediments than in the control stations (on average: 11.85 ± 1.85 vs 7.39 ± 0.96 · 108 cells g1 , respectively; ANOVA; p < 0:01). The frequency of dividing cells (FDC) ranged from 1.8 ± 0.1% to 7.0 ± 0.9% in the impacted sediments (in spring and autumn, respectively), and from 1.9 ± 0.1% to 6.6 ± 0.6% in the control sediments (in spring and summer, respectively). Significant differences were observed among the sampling periods (ANOVA; p < 0:01), but no differences were observed between the impacted and the control sediments (ANOVA; ns). 3.5. Meiofaunal abundance and community structure The meiofaunal densities in the top 8 cm of the sediments ranged from 381 ± 32 to 2185 ± 344 ind.10 cm2 in the impacted stations (in winter and summer, respectively, Fig. 6), while in the control stations they ranged from 218 ± 32 to 2326 ± 396 ind.10 cm2 (in winter and summer, respectively). Significant differences were observed among the sampling periods (ANOVA; p ¼ 0:01), but no significant differences were observed between the impacted and the control stations at each sampling time (ANOVA; ns). Temporal changes in the meiofaunal community structure in the impacted and the control stations are illustrated in Fig. 7. The meiofaunal community struc-
Fig. 6. Meiofaunal parameters in the sediments investigated. (a) Meiofaunal abundance, and (b) number of taxa.
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
spring
(a)
summer
100%
100%
95%
95%
90%
90%
85%
85%
80%
331
(b)
80% impact 1 impact 2 impact 3 control 1 control 2 control 3
autumn
impact 1 impact 2 impact 3 control 1 control 2 control 3
(c)
winter
100%
100%
95%
95%
90%
90%
85%
85%
80%
(d)
80% impact 1 impact 2 impact 3 control 1 control 2 control 3
Nematoda Gastrotricha
Copepoda Ostracoda
Polychaeta Others
Turbellaria
impact 1 impact 2 impact 3 control 1 control 2 control 3 Nematoda Gastrotricha
Copepoda Ostracoda
Polychaeta Others
Turbellaria
Fig. 7. Meiofaunal community structures in the four sampling periods: (a) spring, (b) summer, (c) autumn, and (d) winter.
ture was similar in the impacted and the control stations (i.e., no significant differences were observed; ANOVA; ns). In the impacted stations, nematodes accounted for 83.7 to 97.6% of the total meiofaunal abundance (at Impact 1 station, in spring and autumn, respectively). They were followed by copepods (on average: 1.7%; range: 0.2–4.1%), gastrotrichs (on average: 1.5%; range: 0.0–9.8%) and polychaetes (on average: 1.1%; range: 0.3–.1.8%). The contribution of all of the other remaining taxa was lower than 1%. Similarly, in the control stations, the meiofaunal community structure was dominated by nematodes (accounting for 87.6– 98.5% of total meiofaunal abundance at Control 2 station, in summer and winter, respectively), followed by copepods (on average: 1.8%; range: 0.7–6.3%), polychaetes (on average: 1.4%; range 0.5–4.9%) and gastrotrichs (on average: 0.8%; range: 0.0–4.6%).
4. Discussion Aquaculture activities are generally viewed as having major negative impacts on the coastal environment as they can modify the ecosystem equilibria, and hence increase the degradation rates of coastal marine systems (Wu, 1995; Danovaro, 2003). Detrimental effects of intensive aquaculture (including both fish and shellfish rearing) have been documented using a wide range of indicators under several conditions and at various latitudes (Castel et al., 1989; Grenz et al., 1990; Gilbert et al., 1997; Mirto et al., 2000; Danovaro et al., 2003). The impact of the intensive fish farming on the benthic environment would be expected to be higher than that of mussel farming (Mazzola et al., 1999) as mussels
feed on natural resources (suspended particles) and are not sustained by any additional intensive feeding (Inglis et al., 2000). However, previous investigations in mussel farms located in the Mediterranean (Gaeta Gulf, Tyrrhenian Sea) have reported that mussel biodeposition can also have serious impacts on the farm sediments. Indeed, such biodeposition has been seen to be responsible for a significant accumulation of biopolymeric carbon, which induced significant changes in microbial and meiofaunal assemblages (Mirto et al., 2000). At the same time, other studies based on biogeochemical descriptors have reported very low impacts of mussel farming on the benthic environment (Baudinet et al., 1990; Crawford et al., 2003). These last results are confirmed by our study, since we do not observed any significant differences in the oxygen penetration of the sediment between the impacted and the control stations. In this regard, the application of a multi-control sampling strategy has revealed the presence of rather heterogeneous sediment features, which in most cases were not taken into account in previous studies. The analysis of total vertical fluxes and organic inputs to the bottom also do not reveal any signs of more intense biodeposition in the farm sediments. Only in December, the period of highest mussel stocks in the farm, is an increase in biodeposition fluxes observed. These results can probably be explained by the shallow depth of the sampling sites (on average ca. 11 m), which when coupled with the hydrodynamic regime, would result in a continuous re-suspension and/or export of mussel biodeposits. The vertical fluxes measured in the present study are significantly lower than such fluxes observed for other mussel farms (Hatcher et al., 1994; Stenton-Dozey et al., 1999), but they are comparable
332
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
with those reported for other shellfish (oyster and mussel) farms in which no impacts were detected (Crawford et al., 2003). As a consequence, also the biochemical indicators do not show any clear evidence of the eutrophication processes, as no significant differences are found when comparing mussel-farm and control sediments in terms of the organic matter accumulation and the biochemical composition or the trophic conditions (sensu Dell’Anno et al., 2002). This does not mean that the effects of mussel farming are not detectable at all. Indeed, an accumulation of phytopigments and organic matter was observed in the farm sediments in December, just before mussel harvesting, although these are not statistically relevant when compared to the background conditions. In this regard, it is also reasonable to expect that the impact on the sediment chemistry of long-line mussel culture is less relevant than of mussel dredging, as the mechanical disturbance of the sediment can cause changes in the diagenetic processes occurring in the surface and deeper sediment layers (Spencer et al., 1998; Holmer et al., 2003). Similarly, no microbial response to biodeposition is evident, as we only saw a slight increase in the bacterial density in the sediments beneath the mussel farm. In this case, it is difficult to discern whether the bacteria responded to slight differences in the organic matter accumulation in the sediment, or were released in the sediment from faeces through biodeposition. However, the lack of differences in terms of potential growth rates (as inferred from the frequencies of the dividing cells) would suggest that the bacteria in the farm sediment do not display a different metabolism than those in control sediments. Meiofaunal assemblages are known to respond quickly to biodeposition. Previous investigations have revealed that the impact of organic matter accumulation and the formation of reducing conditions in surface sediments can cause a significant reduction in the meiofaunal abundance (up to 50%) and changes in the community structure (Mazzola et al., 1999). In the present study, in comparing the sediments beneath the long-lines with the control sites, no differences are seen in terms of the meiofaunal abundance, nor are there any changes evident in terms of the meiofaunal community structure or the number of taxa. These results are likely to reflect the absence of any evident organic matter accumulation in the mussel farm sediments. All the variables investigated generally display wider differences across the seasons within the same station than between the impacted and the control stations during the same sampling periods. As such, it is evident that the changes related to the temporal dynamics of the biological components are more relevant than the potential impact associated with the mussel production process itself.
None of the environmental descriptors investigated in the impacted sites are significantly different from those values observed in the control areas, thus suggesting the absence of any effects on the benthic domain due to the mussel farming. These results suggest that mussel farming in the investigated system is eco-sustainable, as it does not significantly alter the coastal marine ecosystem, both in terms of the functioning and the trophic states, and it does not affect the biota at the different levels of the biological organization. It remains to be clarified why mussel farming is detrimental for ecosystem quality in certain cases (as in the Tyrrhenian Sea; Mirto et al., 2000), while in others (as in the present study) it is apparently inoffensive. The impact on the benthic environment appears to be strictly dependent upon several factors, including: (i) the culturing method; (ii) the density of the cultivated mussels; (iii) the water depth; and (iv) the hydrographical conditions in the area. In addition, two other factors could have affected these results: (a) the mussel farm location, which was in an area characterized by relatively high trophic conditions and a high background concentration of sediment organic matter; and (b) a multicontrol sampling strategy, which has allowed the problem of understanding the variability of the background conditions to be addressed. The results presented in the present study can now represent a reference point for better approaches towards the identification of the conditions for the selecting of sites where mussel farms should be established, and towards the sustainable development of aquaculture (i.e., mussel farms) in shallow coastal environments.
Acknowledgements This work was financially supported by the MIPAF (Ministero per le Politiche Agricole e Forestali, Italy). Dr. G. Fabi ISMARE (Ancona) kindly supported all sampling activities and coordinated the field work.
References Baudinet, D., Alliot, E., Berland, B., Grenz, C., Plante-Cuny, M., Plante, R., Salen-Picard, C., 1990. Incidence of mussel culture on biogeochemical fluxes at the sediment-water interface. Hydrobiologia 207, 187–196. Castel, J., Labourg, P.J., Escaravage, V., Auby, I., Garcia, M.E., 1989. Influence of seagrass beds and oyster parks on the abundance and biomass patterns of meio-and macrobenthos in tidal flats. Estuarine, Coastal and Shelf Science 28, 71–85. Crawford, C.M., Macleod, C.K.A., Mitchell, I.M., 2003. Effects of shellfish farming on the benthic environment. Aquaculture 224, 117–140. Danovaro, R., 2003. Pollution threats in the mediterranean sea: an overview. Chemistry and Ecology 19, 15–32.
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333 Danovaro, R., Marrale, D., Della Croce, N., Parodi, P., Fabiano, M., 1999. Biochemical composition of sedimentary organic matter and bacterial distribution in the Aegean Sea: trophic state and pelagic– benthic coupling. Journal of Sea Research 42, 117–129. Danovaro, R., Corinaldesi, C., La Rosa, T., Luna, G.M., Mazzola, A., Mirto, S., Vezzulli, L., Fabiano, M., 2003. Aquaculture impact on benthic microbes and organic matter cycling in coastal Mediterranean sediments: a synthesis. Chemistry and Ecology 19, 59–65. Dell’Anno, A., Mei, M.L., Pusceddu, A., Danovaro, R., 2002. Assessing the trophic state and eutrophication of coastal marine systems: a new approach based on the biochemical composition of sediment organic matter. Marine Pollution Bulletin 44, 611–622. Dinet, A., Sornin, J.M., Sabliere, A., Delmas, D., Feuillet-Girard, M., 1990. Influence de la biodeposition de bivalves filtreurs sur les peuplements meiobenthiques d’un marais maritime. Cahiers de Biologie Marine 31, 307–322. Fabiano, M., Danovaro, R., 1994. Composition of organic matter in sediment facing a river estuary (Tyrrhenian Sea): relationships with bacteria and microphytobenthic biomass. Hydrobiologia 277, 71– 84. FAO, 1997. Review of the State of World Aquaculture. FAO Fisheries Circular 886, pp. 1–163. FAO, 2000. The state of world fisheries and aquaculture. p. 142. Fichez, R., 1991. Composition and fate of organic matter in submarine cave sediments: implications for the biogeochemical cycle of organic carbon. Oceanologica Acta 14, 369–377. Gilbert, F., Souchu, P., Bianchi, M., Bonin, P., 1997. Influence of shellfish farming activities on nitrification, nitrate reduction to ammonium and denitrification at the water–sediment interface of the Thau lagoon, France. Marine Ecology Progress Series 151, 143–153. Grenz, C., Hermin, M.N., Baudinet, D., Daumas, R., 1990. In situ biochemical and bacterial variation of sediments enriched with mussel biodeposits. Hydrobiologia 207, 153–160. Hatcher, A., Grant, J., Schofield, B., 1994. The effects of suspended mussel culture (Mytilus spp.) on sedimentation, benthic respiration, and sediment nutrient dynamics in a coastal bay. Marine Ecology Progress Series 115, 219–235. Heip, C., Vincx, M., Vranken, G., 1985. The ecology of marine nematodes. Oceanography and Marine Biology Annual Review 23, 399–489. Holmer, M., Ahrensberg, N., Jørgensen, N.P., 2003. Impacts of mussel dredging on sediment phosphorus dynamics in a eutrophic Danish fjord. Chemistry and Ecology 19, 343–361.
333
Inglis, G.J., Hayden, B.J., Ross, A.H., 2000. An overview of factors affecting the carrying capacity of coastal embayments for mussel culture. Report NIWA (National Institute of Water and Atmospheric Research, New Zealand), p. 31. Kaiser, M.J., Laing, I., Utting, S.D., Burnell, G.M., 1998. Environmental impacts of bivalve mariculture. Journal of Shellfish Research 17, 59–66. Kaspar, H.F., Gillespie, P.A., Boyer, I.C., MacKenzie, A.L., 1985. Effects of mussel aquaculture on the nitrogen cycle and benthic communities in Kenepuru Sound, Malborough Sounds, New Zealand. Marine Biology 85, 127–136. La Rosa, T., Mirto, S., Mazzola, A., Danovaro, R., 2001. Differential responses of benthic microbes and meiofauna to fish-farm disturbance in coastal sediments. Environmental Pollution 112, 427–434. Mazzola, A., Mirto, S., Danovaro, R., 1999. Initial fish-farm impact on meiofaunal assemblages in coastal sediments of the western Mediterranean. Marine Pollution Bulletin 38, 1126–1133. Mazzola, A., Mirto, S., Danovaro, R., Fabiano, M., 2000. Fish farming effects on benthic community structure in coastal sediments: analysis of the meiofaunal recovery. ICES Journal of Marine Science 57, 1454–1461. Mirto, S., Fabiano, M., Danovaro, R., Manganaro, A., Mazzola, A., 1999. Use of meiofauna for detecting fish farming disturbance in coastal sediments: preliminary results. Biologia Marina Mediterranea 6, 331–334. Mirto, S., La Rosa, T., Danovaro, R., Mazzola, A., 2000. Microbial and meiofaunal response to intensive mussel-farm biodeposition in coastal sediments of the western Mediterranean. Marine Pollution Bulletin 40, 244–252. Parker, J.G., 1983. A comparison of method used for the measurement of organic matter in marine sediments. Chemistry and Ecology 1, 201–210. Spencer, B.E., Kaiser, M.J., Edwards, D.B., 1998. Intertidal clam harvesting: benthic community change and recovery. Aquaculture Research 29, 429–437. Stenton-Dozey, J.M.E., Jackson, L.F., Busby, A.J., 1999. Impact of mussel culture on macrobenthic community structure in Saldanha Bay, South Africa. Marine Pollution Bulletin 39, 1–12. Tenore, K.R., Corral, J., Gonzales, N., 1985. Effects of intense mussel culture on food chain patterns and production in coastal Galicia, North Spain. ICES CM 62, 9. Wu, R.S.S., 1995. The environmental impact of marine fish culture: towards a sustainable future. Marine Pollution Bulletin 31, 159– 166.
Sustainable impact of mussel farming in the Adriatic Sea (Mediterranean Sea): evidence from biochemical, microbial and meiofaunal indicators R. Danovaro *, C. Gambi, G.M. Luna, S. Mirto Department of Marine Sciences, Faculty of Science, Polytechnic University of Marche, Via Brecce Bianche, Monte D’Ago, Ancona 60131, Italy
Abstract We have investigated the impact of a large mussel farm on the benthic environment using a battery of benthic indicators of environmental quality (including biochemical, microbial and meiofaunal parameters). These were analysed through a multi-control sampling strategy over one year. The differences across the seasons are typically higher than those between the impacted and the control stations. No effects are seen in terms of the sediment oxygen penetration and the downward fluxes (as the total mass, organic and phytopigment fluxes). The indicators based on the biochemical compositions of the sediment organic matter and the microbial parameters also show no evidence of the eutrophication process, except as a slight increase in the bacterial density in the sediments beneath the long-lines of the farm during the period of highest mussel stocks. Finally, no effects are observed in terms of the benthic faunal indicators, as the meiofaunal abundance, the community structure and the taxa richness are all indistinguishable between the farm sediments and the controls. These results show that mussel farming in the investigated system is eco-sustainable and does not significantly alter the coastal marine ecosystem, both in terms of the functioning and the trophic state. The battery of indicators selected in this study represents a useful tool for the monitoring of the potential ecological impact of mussel farms, towards guaranteeing the sustainable development of aquacultures in shallow coastal environments. 2004 Elsevier Ltd. All rights reserved. Keywords: Bio-indicators; Aquaculture; Mussel farms; Sediment organic matter; Microbial indicators; Meiofaunal indicators
1. Introduction Shellfish production in aquaculture is increasing significantly throughout the world (FAO, 1997). Considering only the Mediterranean Sea, mussel production is over 700,000 ton y1 (FAO, 2000). Shellfish farms are typically located in estuarine areas and/or inshore coastal waters, which harbour high population densities and a broad range of economic interests. As a consequence, the opposition to the expansion of shellfish farming is increasing, due to the growing concern about ecosystem degradation, and their potential conflicts with the tourism industry and coastal urbanization (Kaiser et al., 1998).
* Corresponding author. Tel.: +39-71-220-4654; fax: +39-71-2204650. E-mail address: [email protected] (R. Danovaro).
0025-326X/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2004.02.038
The impact of bivalve culturing, which is mostly due to mussel and oyster farms, is related to the intensive biodeposition of the faeces and pseudo-faeces that modify the physical and chemical characteristics of the benthic environment as they accumulate in the bottom sediments (Kaspar et al., 1985; Gilbert et al., 1997; Mirto et al., 1999). This phenomenon is well known and documented for intensive fin-fish aquaculture. However, the impact of mussel farming is expected to be less relevant than fish farming, because in the latter the impact due to the accumulation of biodeposits is further increased by the accumulation of organic matter due to uneaten food (Mazzola et al., 2000; Crawford et al., 2003). Nonetheless, a wide variety of negative effects have generally been reported for mussel farming since biodeposition can induce severe organic matter accumulation and reducing conditions in the sediments beneath the cages (Kaspar et al., 1985; Tenore et al., 1985;
326
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
Kaiser et al., 1998; Mirto et al., 2000). This, in turn, affects benthic biodiversity and community structure, altering trophic interactions and pathways of energy transfer from bacteria and picoeukaryotes (Mirto et al., 2000; Danovaro et al., 2003) to meiofauna (Dinet et al., 1990; Mirto et al., 2000); and macrofauna (Castel et al., 1989; Stenton-Dozey et al., 1999). Assessing the potential impact of mussel farming is extremely important for developing an eco-sustainable management of aquaculture (FAO, 2000). For this, we need to apply the widest possible range of scientific tools and environmental indicators that are able to detect all possible kinds of ecological alterations that can be caused by the aquaculture activities. Besides the historical use of bioindicators of impact based on macrofaunal assemblages, more recently new indicators have been proposed for identifying the short-term impacts of aquaculture. Indicators based on the biochemical composition of the sediment organic matter provide important insights into the eutrophication processes occurring in the area (Dell’Anno et al., 2002). Microbial indicators have been suggested for highlighting the alterations in the biogeochemical processes occurring in farm sediments (La Rosa et al., 2001). Finally, due to their fast generation times and their ability to respond quickly to any changes in environmental conditions, biological indicators based on meiofauna can reveal alterations at the community level. The simultaneous use of these indicators provides elements for an integrated approach through which we can cover various levels of biological organization of the ecosystem. In this study, we have investigated the impact of a large mussel farm located in the Adriatic sea (Western Mediterranean) using a sampling strategy based on a preliminary survey. This was then followed by a multicontrol approach to evaluate: (a) the effects of mussel biodeposition on the eutrophication processes; (b) the alterations in terms of the microbial compartment; (c) the effects on the faunal community structure.
2. Materials and methods 2.1. Study site and sampling The present study was conducted from June 2001 to February 2002 in a mussel farm located 1.5–2.0 miles from the coast, facing Cattolica-Rimini in the Middle Adriatic Sea (Central Mediterranean). The area covered by the farm was ca. 2 km2 and the average bottom depth was 11 m. The mussels were cultivated using a long-line technique, in which the mussel larvae settle on ropes hanging down from long horizontal anchored lines that are suspended by buoys. The locations of the sampling stations was defined after a preliminary survey, in which 25 stations were
Fig. 1. The sampling stations for the mussel farm located in the Adriatic Sea. The shadow box illustrates the extent of the mussel farm.
sampled in a grid covering the entire mussel farm area and surroundings up to 600 m around. This allowed the creation of a map of biodeposition accumulation, and according to a multi-control sampling strategy, we defined the position of 3 control stations outside the mussel farm (Control 1, 2 and 3; located ca. 600 m out of the farming area) and three impacted stations within the area of the long-lines of the mussel farm (Impact 1, 2 and 3; Fig. 1). The sediment samples were collected using a boxcorer and/or a grab; this was done in June (after the phytoplankton bloom), September (the period of long-line seeding), December 2001 (the period of highest mussel standing stocks) and February 2002 (the period of the highest mixing of the water column, and the lowest temperatures). Hereafter, these sampling periods will be termed spring, summer, autumn and winter, respectively. Five deployments were carried out at each station, and independent replicate cores were collected by different deployments. All of the cores were collected from the inner part of the box corer (using a plexiglass liner with an inner diameter of 4.7 cm). For the biochemical analyses, the surface layer of each core (the top 2 cm of the sediment) was homogenised in a Petri dish and stored at )20 C until analysis. To estimate the downward fluxes at each sampling time, two sediment trap moorings were deployed in correspondence to the Impact 1 and Control 3 stations. Each mooring had three sediment traps (Technicapmodel; PVC with an inner diameter of 13 cm and a height of 70 cm) placed at approximately 70 cm from the sea bottom. The trap deployments lasted 24 h at each sampling period, and the material collected was filtered immediately after recovery using GF/F glass-fibre filters (0.45 lm pore size), and stored at )20 C until analysis.
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
For the analysis of the microbiological variables, replicate sub-samples (n ¼ 3, ca. 1 ml) of surface sediment were withdrawn from each core and fixed with 4 ml of prefiltered (0.2 lm) and buffered 2% formalin. The samples were then stored at 4 C until the analysis, which was performed within 4 weeks of sampling. The meiofaunal samples were collected in replicate Plexiglas cores (n ¼ 3, 3.7 cm diameter, 10.7 cm2 surface area) down to a depth of 8 cm, and then fixed with 4% buffered formalin in 0.4 lm prefiltered seawater solution and stained with Rose Bengal (0.5 g l1 ). This depth appeared to be optimal for the quantitative analysis as the evaluation of the top 15 cm of the sediment core revealed that <1% of the total meiofaunal density was present in the lower 8–15 cm sediment horizon. 2.2. Environmental variables and biochemical indicators The redox potential discontinuity (RPD) layer was visually estimated as the depth at which the sediment colour turned from brown to grey-black. The grain size was determined according to standard protocols using a dry sieve technique. The total organic matter (TOM) was determined as the difference between the dry weight of the sediment (80 C, 24 h) and the residue left after combustion at 450 C for 4 h (Parker, 1983). Before the TOM analysis, the sediment samples were treated with an excess of 10% HCl to remove carbonates that may have interfered with the TOM determination. The biochemical analyses (protein, carbohydrate and lipid) and chloroplastic pigment determinations (chlorophyll-a and phaeopigments) of both the trap material and the surface sediments were carried out as described by Dell’Anno et al. (2002). For each parameter, the blanks used were from the same previously calcinated sediments (450 C, 2 h). All analyses were carried out in three replicates. The sediment organic matter concentrations were normalized to dry weight after desiccation at 60 C. The protein, carbohydrate and lipid concentrations were converted to their carbon equivalents, assuming the conversion factors of 0.49, 0.40 and 0.75 lgC lg1 , respectively (Fichez, 1991). The sum of the protein, carbohydrate and lipid carbon is referred as the biopolymeric carbon (BPC; Fabiano and Danovaro, 1994). 2.3. Microbial parameters The total bacterial counts were performed on subsamples that had been sonicated three times (Branson Sonifier 2200; 60 W for 1 min) and diluted 250–500 times using sterile (0.2 lm prefiltered) formalin (2% final concentration). An aliquot of 1 ml was stained with Acridine Orange (5 mg l1 final concentration, for 5 min in the dark) and filtered under vacuum (<100 mmHg) using 0.2 lm pore size black Nucleopore Polycarbonate
327
filters. The filters were then washed twice with 3 ml of 0.2 lm prefiltered and sterilized Milli-Q water, mounted on microscope slides, and analysed using epifluorescence microscopy (Zeiss Axioskop 2; magnification ·1000). For each slide, at least 10 microscope fields were observed, for a total of at least 400 cells counted per filter. The frequency of dividing bacterial cells (FDC, defined as cells with a clearly visible invagination) was also determined (Danovaro et al., 1999). All data were normalized to sediment dry weight after desiccation (60 C, 24 h). 2.4. Meiofauna For meiofaunal analyses, the sediments were sieved through 1000 and 37 lm mesh nets. The fraction remaining on the 37 lm mesh sieve was centrifuged three times with Ludox HS (density 1.18 g cm3 ; Heip et al., 1985). All meiobenthic animals were sorted, counted and classified into their taxa under a stereo microscope. 2.5. Statistical analyses The differences among the stations (impacted and control) and the sampling periods were analysed by means of two-way analysis of variance (ANOVA). When a significant difference for the main effect was observed ðp < 0:05Þ, a Tukey’s pairwise comparison test was also performed.
3. Results 3.1. Sediment characteristics The characteristics of the benthic environment (RPD, grain size) are reported in Table 1. The RPD depth varied widely between the seasons, ranging from 0.7 cm in summer to 6.5 cm in winter in the impacted stations, and from 1.5 cm in summer to 8.0 cm in winter in the control stations. No significant differences were observed between the RPD values comparing impacted and control stations (ANOVA; ns). The surface sediments were typically characterized by muddy sands, as they were composed of similar fractions of mud (on average 44.1%) and sand (on average 55.9%). 3.2. Fluxes The seasonal changes in the vertical fluxes in the areas subjected to mussel biodeposition (as total and organic matter flux and chloropigment fluxes) and in the control sites are illustrated in Fig. 2. The total mass flux ranged from 4.5 ± 1.5 to 17.0 ± 5.5 g m2 d1 in the impacted
328
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
Table 1 Sediment parameters in the investigated sites Redox potential discontinuity
Station
Spring (cm)
Summer (cm)
Autumn (cm)
Impact 1 Impact 2 Impact 3
2.2 5.5 4.5
1.8 2.3 0.7
4 2 3.5
Average
4.1
1.6
Control 1 Control 2 Control 3
5.5 6.7 3.8
Average
5.3
Grain size Winter (cm)
Spring
Summer
Autumn
Winter
Mud (%)
Sand (%)
Mud (%)
Sand (%)
Mud (%)
Sand (%)
Mud (%)
Sand (%)
6.5 4.3 4
15.8 15.2 32.0
84.2 84.8 68.0
68.2 34.9 69.3
31.8 65.1 30.7
48.0 15.8 70.1
52.0 84.2 29.9
77.1 34.0 70.0
22.9 66.0 30.0
3.2
4.9
21.0
79.0
57.5
42.5
44.6
55.4
60.4
39.6
4.7 1.5 2.7
4.3 4.5 3.3
15.7 8 6.7
8.3 34.2 50.8
91.7 65.8 49.2
18.9 86.4 65.8
81.1 13.6 34.2
5.7 24.9 50.1
94.3 75.1 49.9
18.4 69.7 75.4
81.6 30.3 24.6
3.0
4.0
6.8
31.1
68.9
57.1
42.9
26.9
73.1
54.5
45.5
The redox potential discontinuity depth and grain size composition (as the sand and mud fraction contributions, expressed as percentages) across the four sampling periods.
Total mass flux
25
Control
Impact
g m-2 d-1
20 15 10 5 na 0
(a)
spring
autumn
winter
Organic matter flux
1600
mg m-2 d-1
summer
1200 800 400
na
(b)
0 spring
summer
autumn
winter
Chloropigment flux
7
mg m-2 d-1
6 5 4 3 2 1
(c)
ences were observed in fluxes between the impacted and the control stations (ANOVA; ns). The organic matter flux in the mussel farm area was quite constant across the sampling periods (range: 947.6 ± 379.3 to 1146.5 ± 270.2 mg m2 d1 ), although in the control stations it was more variable (range: 499.0 ± 117.6 to 968.2 ± 216.2 mg m2 d1 ), and followed a pattern similar to that observed for the total mass flux. No significant differences were observed between the organic fluxes measured in the mussel farm and those measured in the control sites (ANOVA; ns). The total chloropigment fluxes ranged from 0.91 ± 0.06 to 5.97 ± 0.28 mg m2 d1 in the impacted stations (in spring and winter, respectively), and from 0.32 ± 0.05 to 4.06 ± 0.52 mg m2 d1 in the control stations (in spring and winter, respectively). In winter and spring, total chloropigment fluxes were significantly higher in the mussel station (ANOVA; p < 0:01).
0
na spring
summer
autumn
winter
Fig. 2. Downward fluxes of (a) total mass, (b) organic matter, and (c) chloropigments. Standard deviations are shown. na ¼ not available.
stations (in winter and spring, respectively), and from 6.9 ± 1.8 to 9.4 ± 2.9 g m2 d1 in the control stations (in spring and autumn, respectively). No significant differ-
3.3. Phytopigment and organic matter concentrations in the sediment The phytopigment concentrations (chlorophyll-a and phaeopigments) are shown in Fig. 3. The chlorophyll-a sediment concentrations ranged from 1.0 ± 0.2 to 11.2 ± 2.5 lg g1 in the impacted stations (in autumn and summer, respectively), and from 1.4 ± 0.1 to 8.3 ± 0.5 lg g1 in the control stations (in autumn and summer, respectively). The phaeopigment concentrations ranged from 2.5 ± 0.6 to 35.1 ± 3.2 lg g1 in the impacted stations (in autumn and summer, respectively), and from 2.6 ± 0.0 to 26.2 ± 17.5 lg g1 in the control stations (in winter, at the Control 1 and 2 stations, respectively). Similar temporal patterns were displayed by the total chloropigment concentrations. The concentrations of chlorophyll-a and phaeopigments in both
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
the impacted and the control stations were significantly higher in summer than in all of the other sampling periods (ANOVA; p < 0:01), but they did not show differences between the impacted and the control sites themselves. The total organic matter (TOM), protein, carbohydrate, lipid and biopolymeric carbon (BPC) concentrations are shown in Fig. 4. The TOM ranged from 1.2 ± 0.1 to 6.1 ± 3.3% of sediment dry weight (DW) in the impacted stations (in summer and winter, respectively), and from 1.7 ± 0.1 to 7.8 ± 0.8% in the control stations (in summer and spring, respectively). Significant differences were observed only in spring, when the organic matter content was lower in the impacted than in the control sites (ANOVA; p < 0:01). Protein and carbohydrate were the two main biochemical classes of organic compounds. The protein ranged from 717.3 ± 113.9 to 2581.9 ± 457.0 lg g1 in the impacted stations (in autumn and winter, respectively), and from 646.0 ± 134.6 to 2999.1 ± 890.2 lg g1 in the control stations (in spring and winter, respectively). The total carbohydrate ranged from 372.1 ± 55.0 to 1983.4 ± 34.2 lg g1 in the impacted stations (in spring
µ
µ
µ
µ
Fig. 3. Phytopigment content in the sediments investigated. (a) Chlorophyll-a concentrations, and (b) phaeopigment content. Standard deviations are shown.
329
Fig. 4. The quantity and quality of the organic matter in the sediments investigated. (a) Total organic matter, (b) protein, (c) carbohydrate, (d) lipid, and (e) biopolymeric carbon concentrations. Standard deviations are shown.
330
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
and autumn, respectively), and from 461.3 ± 54.4 to 2313.1 ± 266.9 lg g1 in the control stations (in spring and winter, respectively). The lipid ranged from 198.5 ± 14.7 to 1073.7 ± 129.8 lg g1 in the impacted stations (in summer and winter, respectively), and from 113.7 ± 33.0 to 1033.5 ± 323.4 lg g1 in the control stations (both in winter, at Control 1 and 2 stations, respectively). None of the biochemical variables investigated displayed any significant differences between the impacted and control stations at each sampling period, nor between the sampling periods (ANOVA; ns). The BPC concentrations (expressed as the sum of the protein, carbohydrate and lipid carbon) ranged from 675.6 ± 112.3 to 2394.0 ± 317.1 lgC g1 in the impacted stations (in spring and summer, respectively), and from 639.2 ± 137 to 3049.9 ± 749.9 lgC g1 in the control stations (both in winter, at Control 1 and 2 stations, respectively). A significant temporal variability was observed for BPC, with the highest concentrations in winter in both the impacted and the control stations (ANOVA; p < 0:01). 3.4. Microbial parameters The total bacterial counts (TBC) and the frequency of dividing cells (FDC) are shown in Fig. 5. The TBC ranged from 3.53 ± 0.49 to 15.35 ± 2.19 · 108 cells g1 sediment DW in the impacted stations (in spring and autumn, respectively), and from 5.12 ± 0.08 to 10.85 ± 1.24 · 108 cells g1 in the control stations (in summer and spring, respectively). Significant differences were observed between sampling periods (ANOVA;
Fig. 5. Microbial parameters in the impacted and the control sediments. (a) Total bacterial counts, and (b) frequency of dividing cells. Standard deviations are shown.
p < 0:01), but no significant differences were observed between the impacted and the control sites, with the exception of the autumn sampling, in which the bacterial abundance was significantly higher in the impacted sediments than in the control stations (on average: 11.85 ± 1.85 vs 7.39 ± 0.96 · 108 cells g1 , respectively; ANOVA; p < 0:01). The frequency of dividing cells (FDC) ranged from 1.8 ± 0.1% to 7.0 ± 0.9% in the impacted sediments (in spring and autumn, respectively), and from 1.9 ± 0.1% to 6.6 ± 0.6% in the control sediments (in spring and summer, respectively). Significant differences were observed among the sampling periods (ANOVA; p < 0:01), but no differences were observed between the impacted and the control sediments (ANOVA; ns). 3.5. Meiofaunal abundance and community structure The meiofaunal densities in the top 8 cm of the sediments ranged from 381 ± 32 to 2185 ± 344 ind.10 cm2 in the impacted stations (in winter and summer, respectively, Fig. 6), while in the control stations they ranged from 218 ± 32 to 2326 ± 396 ind.10 cm2 (in winter and summer, respectively). Significant differences were observed among the sampling periods (ANOVA; p ¼ 0:01), but no significant differences were observed between the impacted and the control stations at each sampling time (ANOVA; ns). Temporal changes in the meiofaunal community structure in the impacted and the control stations are illustrated in Fig. 7. The meiofaunal community struc-
Fig. 6. Meiofaunal parameters in the sediments investigated. (a) Meiofaunal abundance, and (b) number of taxa.
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
spring
(a)
summer
100%
100%
95%
95%
90%
90%
85%
85%
80%
331
(b)
80% impact 1 impact 2 impact 3 control 1 control 2 control 3
autumn
impact 1 impact 2 impact 3 control 1 control 2 control 3
(c)
winter
100%
100%
95%
95%
90%
90%
85%
85%
80%
(d)
80% impact 1 impact 2 impact 3 control 1 control 2 control 3
Nematoda Gastrotricha
Copepoda Ostracoda
Polychaeta Others
Turbellaria
impact 1 impact 2 impact 3 control 1 control 2 control 3 Nematoda Gastrotricha
Copepoda Ostracoda
Polychaeta Others
Turbellaria
Fig. 7. Meiofaunal community structures in the four sampling periods: (a) spring, (b) summer, (c) autumn, and (d) winter.
ture was similar in the impacted and the control stations (i.e., no significant differences were observed; ANOVA; ns). In the impacted stations, nematodes accounted for 83.7 to 97.6% of the total meiofaunal abundance (at Impact 1 station, in spring and autumn, respectively). They were followed by copepods (on average: 1.7%; range: 0.2–4.1%), gastrotrichs (on average: 1.5%; range: 0.0–9.8%) and polychaetes (on average: 1.1%; range: 0.3–.1.8%). The contribution of all of the other remaining taxa was lower than 1%. Similarly, in the control stations, the meiofaunal community structure was dominated by nematodes (accounting for 87.6– 98.5% of total meiofaunal abundance at Control 2 station, in summer and winter, respectively), followed by copepods (on average: 1.8%; range: 0.7–6.3%), polychaetes (on average: 1.4%; range 0.5–4.9%) and gastrotrichs (on average: 0.8%; range: 0.0–4.6%).
4. Discussion Aquaculture activities are generally viewed as having major negative impacts on the coastal environment as they can modify the ecosystem equilibria, and hence increase the degradation rates of coastal marine systems (Wu, 1995; Danovaro, 2003). Detrimental effects of intensive aquaculture (including both fish and shellfish rearing) have been documented using a wide range of indicators under several conditions and at various latitudes (Castel et al., 1989; Grenz et al., 1990; Gilbert et al., 1997; Mirto et al., 2000; Danovaro et al., 2003). The impact of the intensive fish farming on the benthic environment would be expected to be higher than that of mussel farming (Mazzola et al., 1999) as mussels
feed on natural resources (suspended particles) and are not sustained by any additional intensive feeding (Inglis et al., 2000). However, previous investigations in mussel farms located in the Mediterranean (Gaeta Gulf, Tyrrhenian Sea) have reported that mussel biodeposition can also have serious impacts on the farm sediments. Indeed, such biodeposition has been seen to be responsible for a significant accumulation of biopolymeric carbon, which induced significant changes in microbial and meiofaunal assemblages (Mirto et al., 2000). At the same time, other studies based on biogeochemical descriptors have reported very low impacts of mussel farming on the benthic environment (Baudinet et al., 1990; Crawford et al., 2003). These last results are confirmed by our study, since we do not observed any significant differences in the oxygen penetration of the sediment between the impacted and the control stations. In this regard, the application of a multi-control sampling strategy has revealed the presence of rather heterogeneous sediment features, which in most cases were not taken into account in previous studies. The analysis of total vertical fluxes and organic inputs to the bottom also do not reveal any signs of more intense biodeposition in the farm sediments. Only in December, the period of highest mussel stocks in the farm, is an increase in biodeposition fluxes observed. These results can probably be explained by the shallow depth of the sampling sites (on average ca. 11 m), which when coupled with the hydrodynamic regime, would result in a continuous re-suspension and/or export of mussel biodeposits. The vertical fluxes measured in the present study are significantly lower than such fluxes observed for other mussel farms (Hatcher et al., 1994; Stenton-Dozey et al., 1999), but they are comparable
332
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333
with those reported for other shellfish (oyster and mussel) farms in which no impacts were detected (Crawford et al., 2003). As a consequence, also the biochemical indicators do not show any clear evidence of the eutrophication processes, as no significant differences are found when comparing mussel-farm and control sediments in terms of the organic matter accumulation and the biochemical composition or the trophic conditions (sensu Dell’Anno et al., 2002). This does not mean that the effects of mussel farming are not detectable at all. Indeed, an accumulation of phytopigments and organic matter was observed in the farm sediments in December, just before mussel harvesting, although these are not statistically relevant when compared to the background conditions. In this regard, it is also reasonable to expect that the impact on the sediment chemistry of long-line mussel culture is less relevant than of mussel dredging, as the mechanical disturbance of the sediment can cause changes in the diagenetic processes occurring in the surface and deeper sediment layers (Spencer et al., 1998; Holmer et al., 2003). Similarly, no microbial response to biodeposition is evident, as we only saw a slight increase in the bacterial density in the sediments beneath the mussel farm. In this case, it is difficult to discern whether the bacteria responded to slight differences in the organic matter accumulation in the sediment, or were released in the sediment from faeces through biodeposition. However, the lack of differences in terms of potential growth rates (as inferred from the frequencies of the dividing cells) would suggest that the bacteria in the farm sediment do not display a different metabolism than those in control sediments. Meiofaunal assemblages are known to respond quickly to biodeposition. Previous investigations have revealed that the impact of organic matter accumulation and the formation of reducing conditions in surface sediments can cause a significant reduction in the meiofaunal abundance (up to 50%) and changes in the community structure (Mazzola et al., 1999). In the present study, in comparing the sediments beneath the long-lines with the control sites, no differences are seen in terms of the meiofaunal abundance, nor are there any changes evident in terms of the meiofaunal community structure or the number of taxa. These results are likely to reflect the absence of any evident organic matter accumulation in the mussel farm sediments. All the variables investigated generally display wider differences across the seasons within the same station than between the impacted and the control stations during the same sampling periods. As such, it is evident that the changes related to the temporal dynamics of the biological components are more relevant than the potential impact associated with the mussel production process itself.
None of the environmental descriptors investigated in the impacted sites are significantly different from those values observed in the control areas, thus suggesting the absence of any effects on the benthic domain due to the mussel farming. These results suggest that mussel farming in the investigated system is eco-sustainable, as it does not significantly alter the coastal marine ecosystem, both in terms of the functioning and the trophic states, and it does not affect the biota at the different levels of the biological organization. It remains to be clarified why mussel farming is detrimental for ecosystem quality in certain cases (as in the Tyrrhenian Sea; Mirto et al., 2000), while in others (as in the present study) it is apparently inoffensive. The impact on the benthic environment appears to be strictly dependent upon several factors, including: (i) the culturing method; (ii) the density of the cultivated mussels; (iii) the water depth; and (iv) the hydrographical conditions in the area. In addition, two other factors could have affected these results: (a) the mussel farm location, which was in an area characterized by relatively high trophic conditions and a high background concentration of sediment organic matter; and (b) a multicontrol sampling strategy, which has allowed the problem of understanding the variability of the background conditions to be addressed. The results presented in the present study can now represent a reference point for better approaches towards the identification of the conditions for the selecting of sites where mussel farms should be established, and towards the sustainable development of aquaculture (i.e., mussel farms) in shallow coastal environments.
Acknowledgements This work was financially supported by the MIPAF (Ministero per le Politiche Agricole e Forestali, Italy). Dr. G. Fabi ISMARE (Ancona) kindly supported all sampling activities and coordinated the field work.
References Baudinet, D., Alliot, E., Berland, B., Grenz, C., Plante-Cuny, M., Plante, R., Salen-Picard, C., 1990. Incidence of mussel culture on biogeochemical fluxes at the sediment-water interface. Hydrobiologia 207, 187–196. Castel, J., Labourg, P.J., Escaravage, V., Auby, I., Garcia, M.E., 1989. Influence of seagrass beds and oyster parks on the abundance and biomass patterns of meio-and macrobenthos in tidal flats. Estuarine, Coastal and Shelf Science 28, 71–85. Crawford, C.M., Macleod, C.K.A., Mitchell, I.M., 2003. Effects of shellfish farming on the benthic environment. Aquaculture 224, 117–140. Danovaro, R., 2003. Pollution threats in the mediterranean sea: an overview. Chemistry and Ecology 19, 15–32.
R. Danovaro et al. / Marine Pollution Bulletin 49 (2004) 325–333 Danovaro, R., Marrale, D., Della Croce, N., Parodi, P., Fabiano, M., 1999. Biochemical composition of sedimentary organic matter and bacterial distribution in the Aegean Sea: trophic state and pelagic– benthic coupling. Journal of Sea Research 42, 117–129. Danovaro, R., Corinaldesi, C., La Rosa, T., Luna, G.M., Mazzola, A., Mirto, S., Vezzulli, L., Fabiano, M., 2003. Aquaculture impact on benthic microbes and organic matter cycling in coastal Mediterranean sediments: a synthesis. Chemistry and Ecology 19, 59–65. Dell’Anno, A., Mei, M.L., Pusceddu, A., Danovaro, R., 2002. Assessing the trophic state and eutrophication of coastal marine systems: a new approach based on the biochemical composition of sediment organic matter. Marine Pollution Bulletin 44, 611–622. Dinet, A., Sornin, J.M., Sabliere, A., Delmas, D., Feuillet-Girard, M., 1990. Influence de la biodeposition de bivalves filtreurs sur les peuplements meiobenthiques d’un marais maritime. Cahiers de Biologie Marine 31, 307–322. Fabiano, M., Danovaro, R., 1994. Composition of organic matter in sediment facing a river estuary (Tyrrhenian Sea): relationships with bacteria and microphytobenthic biomass. Hydrobiologia 277, 71– 84. FAO, 1997. Review of the State of World Aquaculture. FAO Fisheries Circular 886, pp. 1–163. FAO, 2000. The state of world fisheries and aquaculture. p. 142. Fichez, R., 1991. Composition and fate of organic matter in submarine cave sediments: implications for the biogeochemical cycle of organic carbon. Oceanologica Acta 14, 369–377. Gilbert, F., Souchu, P., Bianchi, M., Bonin, P., 1997. Influence of shellfish farming activities on nitrification, nitrate reduction to ammonium and denitrification at the water–sediment interface of the Thau lagoon, France. Marine Ecology Progress Series 151, 143–153. Grenz, C., Hermin, M.N., Baudinet, D., Daumas, R., 1990. In situ biochemical and bacterial variation of sediments enriched with mussel biodeposits. Hydrobiologia 207, 153–160. Hatcher, A., Grant, J., Schofield, B., 1994. The effects of suspended mussel culture (Mytilus spp.) on sedimentation, benthic respiration, and sediment nutrient dynamics in a coastal bay. Marine Ecology Progress Series 115, 219–235. Heip, C., Vincx, M., Vranken, G., 1985. The ecology of marine nematodes. Oceanography and Marine Biology Annual Review 23, 399–489. Holmer, M., Ahrensberg, N., Jørgensen, N.P., 2003. Impacts of mussel dredging on sediment phosphorus dynamics in a eutrophic Danish fjord. Chemistry and Ecology 19, 343–361.
333
Inglis, G.J., Hayden, B.J., Ross, A.H., 2000. An overview of factors affecting the carrying capacity of coastal embayments for mussel culture. Report NIWA (National Institute of Water and Atmospheric Research, New Zealand), p. 31. Kaiser, M.J., Laing, I., Utting, S.D., Burnell, G.M., 1998. Environmental impacts of bivalve mariculture. Journal of Shellfish Research 17, 59–66. Kaspar, H.F., Gillespie, P.A., Boyer, I.C., MacKenzie, A.L., 1985. Effects of mussel aquaculture on the nitrogen cycle and benthic communities in Kenepuru Sound, Malborough Sounds, New Zealand. Marine Biology 85, 127–136. La Rosa, T., Mirto, S., Mazzola, A., Danovaro, R., 2001. Differential responses of benthic microbes and meiofauna to fish-farm disturbance in coastal sediments. Environmental Pollution 112, 427–434. Mazzola, A., Mirto, S., Danovaro, R., 1999. Initial fish-farm impact on meiofaunal assemblages in coastal sediments of the western Mediterranean. Marine Pollution Bulletin 38, 1126–1133. Mazzola, A., Mirto, S., Danovaro, R., Fabiano, M., 2000. Fish farming effects on benthic community structure in coastal sediments: analysis of the meiofaunal recovery. ICES Journal of Marine Science 57, 1454–1461. Mirto, S., Fabiano, M., Danovaro, R., Manganaro, A., Mazzola, A., 1999. Use of meiofauna for detecting fish farming disturbance in coastal sediments: preliminary results. Biologia Marina Mediterranea 6, 331–334. Mirto, S., La Rosa, T., Danovaro, R., Mazzola, A., 2000. Microbial and meiofaunal response to intensive mussel-farm biodeposition in coastal sediments of the western Mediterranean. Marine Pollution Bulletin 40, 244–252. Parker, J.G., 1983. A comparison of method used for the measurement of organic matter in marine sediments. Chemistry and Ecology 1, 201–210. Spencer, B.E., Kaiser, M.J., Edwards, D.B., 1998. Intertidal clam harvesting: benthic community change and recovery. Aquaculture Research 29, 429–437. Stenton-Dozey, J.M.E., Jackson, L.F., Busby, A.J., 1999. Impact of mussel culture on macrobenthic community structure in Saldanha Bay, South Africa. Marine Pollution Bulletin 39, 1–12. Tenore, K.R., Corral, J., Gonzales, N., 1985. Effects of intense mussel culture on food chain patterns and production in coastal Galicia, North Spain. ICES CM 62, 9. Wu, R.S.S., 1995. The environmental impact of marine fish culture: towards a sustainable future. Marine Pollution Bulletin 31, 159– 166.