Changes in fish community structure associated with cage aquaculture in Lake Malawi, Africa

    Changes in fish community structure associated with cage aquaculture in Lake Malawi, Africa A.M. Macuiane, R.E. Hecky, S.J. Guildford PII: DOI: Reference:

S0044-8486(15)30007-7 doi: 10.1016/j.aquaculture.2015.05.015 AQUA 631670

To appear in:

Aquaculture

Received date: Revised date: Accepted date:

4 November 2014 8 May 2015 11 May 2015

Please cite this article as: Macuiane, A.M., Hecky, R.E., Guildford, S.J., Changes in fish community structure associated with cage aquaculture in Lake Malawi, Africa, Aquaculture (2015), doi: 10.1016/j.aquaculture.2015.05.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Changes in fish community structure associated with cage aquaculture in Lake Malawi, Africa

1

PT

Macuiane1,2, A. M., Hecky1, R. E, and Guildford1, S. J. Large Lakes Observatory, University Minnesota Duluth, 2205 East 5th St, Duluth, MN, 55812,

Centro de Estudos Costeiros, Escola Superior de Ciências Marinhas e Costeiras da Universidade

SC

2

RI

USA. Tell: (218) 726-8522, Fax: (218) 726-6979.

Eduardo Mondlane, Av. 1 de Julho, Chuabo Dembe, C.P. 128 Quelimane, Mozambique.

Corresponding author: [email protected]

MA

Abstract

NU

Tel: (+258)823056104, Cell: (+258)842264947, Fax: (+258)21020129.

Local fishermen claim that introduction of a commercial cage aquaculture farm in Lake

D

Malawi resulted in low fish catches as fish take refuge within the farm. Fish specimens were

TE

caught in three fishing sites, one at the farm and two 5 km southeast and northwest of the farm in February, April, June, and August 2012 using four experimental multi-mesh gillnets with similar

AC CE P

dimensions to catch different fish species and sizes. Data was used to determine changes of fish community composition, abundance, biomass, and fish diversity. Site 2 located at the farm and 3 to the southeast were dominated numerically by many small bodied fishes which often is an indication of disturbance by fishing; therefore, they were classed as disturbed while site 1 in northwest was classed as undisturbed site. A near field impact of the farm on fish community structure was detected as revealed by significantly different fish community structure from site 2 compared with that found at sites 1 and 3, but with similar number of fish species and diversity as the remote sites. Overall community structure and the abundance of the fish community in the vicinity of the cage site were improved primarily through providing protection from fishing while incidence of large number of small bodied fish at site 3 resulted from removal of large fish by fishing pressure. This study suggests that protected areas such as the cage site can be a practical strategy to reduce fishing pressure in Lake Malawi and allow recovery of native fish stocks.

Key words: cage aquaculture, wild fish community, fish diversity, fisheries management

ACCEPTED MANUSCRIPT 1. Introduction Lake Malawi/Niassa/Nyasa, hereafter Lake Malawi, lies in south/central Africa. The lake is famous for harboring the highest number of freshwater fish species in the world estimated

PT

between 500 and 1000 species, more than 99% of which are endemic haplochromine cichlids (Fryer and Iles, 1972; Ribbink et al., 1983; Eccles and Trewavas, 1989; Konings, 1995; Snoeks,

RI

2000). According to Snoeks (2000), more fish species are yet to be identified in the lake. The

SC

haplochromine cichlid species are more abundant and diverse in the nearshore areas of the lake, particularly around rock crops. These species are locally referred to as mbuna (Konings, 1990;

NU

Bootsma and Hecky, 1993). The fish dominating the offshore pelagic waters of Lake Malawi belong to the Cyprinidae, Clariidae and Mochokidae families as well as a few pelagic

MA

haplochromines (Thompson et al., 1995).

Fishing is one of the most important livelihood activities in Lake Malawi particularly among poor riparian communities using low cost fishing gears (Jamu et al., 2011). It is estimated

D

that fish provides 70% of animal protein supply and 40% of total protein intake in Malawi

TE

(Banda et al., 2005) while the fisheries sector in the country employs about 60,000 people directly and over 450,000 people indirectly (GoM, 2009). This contribution is especially

AC CE P

valuable in Malawi which is constrained by increased food and nutritional insecurity due to low agricultural productivity, low rural income and rampant poverty (Jamu and Chimatiro, 2005). However, the combined effects of overfishing, habitat destruction, use of illegal gear, violation of closed seasons and protected areas and catchment degradation have significantly reduced fish catches in the lake, particularly the more lucrative Chambo fishery (Oreochromis lidole, O. squamipinnis, and O. karongae) and associated role of the fish as a source of protein in the country (Banda et al., 2005). The market price of fish has significantly increased over the past decades such that fish is currently unaffordable to low income people in Malawi. The Chambo Restoration Strategic Plan implemented by the Malawi government has proposed several management interventions to restore the Lake Malawi fisheries in general but with special attention to the Chambo fishery. The restoration strategies include restriction of fishing gears (mesh size and head line length), introduction of cage aquaculture farming in Lake Malawi and other water bodies, establishment of fish sanctuaries, artificial reefs, community property rights to fishing areas, and restocking of the fishery (Banda et al., 2005) to reduce pressure for sustainable fisheries exploitation. The implementation of the plan has met resistance

ACCEPTED MANUSCRIPT and has had limited success because some of the strategies have been persistently ignored by local artisanal fishermen who are constantly challenged by lack of alternative livelihood strategies.

PT

The current study focused on one of the strategies of the restoration plan, the cage culture of the native Chambo introduced in the south-east arm (SEA) of Lake Malawi in 2004. Over the

RI

past years fish harvest from aquaculture production increased from 666 tonnes in 2004 to 3,232

SC

tonnes in 2012 (FAO, 2005-2014) and it is predicted to increase further as the number of farms in the lake increases and farming strategies improve. However, an increase in the number of

NU

cages may as well result in decline in fish production and fish kills as it did in Saguling and Citarata reservoirs in west Java, Indonesia (Abery et al., 2005). There have been environmental

MA

concerns about the impacts of the fish farm in Lake Malawi on water quality and local fish community structure that support subsistence capture fisheries. The impact of the fish farm on water quality in Lake Malawi was investigated by Gondwe et al. (2011) who reported minimal

D

impacts but warned that the impact may increase as the industry expands in the lake.

TE

While the impacts of cage aquaculture may vary from one ecosystem to another, local fishermen operating in areas surrounding Maldeco farm in Lake Malawi claim that the

AC CE P

introduction of the farm has reduced their fish catches because fish take refuge in the farm. While fish community structure in the SEA prior to the introduction of the cage aquaculture was studied by Banda et al. (1996), Duponchelle et al. (2003) and Weyl et al. (2005), very little information on the same area is available following the introduction of cage aquaculture. The current study was conducted in 2012; eight years after the introduction of cage aquaculture in the lake to investigate the impact of cage aquaculture farming on fish community structure in the SEA. This knowledge is important to guide lake wide aquaculture expansion and ensure sustainable management of the wild capture fishery and diverse fish community in the lake.

2. Materials and methods 2.1 Study area Lake Malawi (Figure 1) is the third largest African Great Lake by area after Lakes Victoria and Tanganyika. The catchment area of Lake Malawi is 97, 740 km2, of which 64, 373 km2 lies in Malawi, 26, 600 km2 lies in Tanzania, and only 6, 768 km2 in Mozambique

ACCEPTED MANUSCRIPT (Department of Water/UNDP, 1986). The lake is 550 km long with a mean width of 48-60 km (Patterson and Kachinjika, 1995; Kanyika, 2000). The maximum depth of 700 m is found in the northern region from which the basin continuously shoals to the south and into the South East

PT

Arm from which the Shire River exits. The lake surface area is 29, 743 km2, a mean depth of 474 m and an estimated volume of 7,723 km3 (Department of Water/UNDP, 1986). Lake Malawi is

RI

connected to downstream Lake Malombe through Shire River which flows out of the SEA. The

SC

Shire River runs through the southern region of Malawi and flows into the Zambezi River in Mozambique.

NU

Maldeco Aquaculture Limited is the only cage aquaculture operation allowed by the Government of Malawi to operate at a commercial scale in Lake Malawi. The farm is dedicated

MA

to the monoculture or polyculture of Chambo species in circular cages (16 m diameter and 6 m deep) moored in the lake where water depths range between 12and 22 m. As of July-August 2010, each active cage was stocked with an average of 130,000 fish corresponding to 108 fish

D

per m3. The study was conducted when only between 11 and 30 cages out of 51 cages were

AC CE P

2.2 Sample collection

TE

stocked with fish during the sampling.

Fishing was done at three fishing sites (Figure 1, right panel) during day time in February, April, June, and August 2012 using four experimental gillnets with similar dimensions to catch different fish species and sizes in order to determine changes of fish community composition, abundance, biomass, and fish diversity. Fishing site 1 was located 5 km northwest of the fish farm; site 2 was located at the farm while site 3 was located 5 km to the southeast of the farm. Fishing sites 1 and 3 were at similar distance from the shoreline but had the same water depth range as site 2 at the fish farm. The gillnets were 76.2m long, 1.83 m deep, constructed of mono-filament nylon, with nylon floating lines. Each net consists of five randomly placed panels each with 15.2 m and 19.1 mm, 25.4 mm, 31.8 mm, 38.1 mm, and 50.8 mm webbing. Two gillnets were deployed simultaneously as replicates at each site as shown in Table 1. After harvesting, fish were immediately identified and sorted to species level, weighed in bulk and individually to the nearest 0.01g and counted.

ACCEPTED MANUSCRIPT 2.3 Data analysis Multivariate analysis using PRIMER (Plymouth Routines in Marine Ecological Research) program (version.6.1.8) (Clarke and Warwick, 2001; Clarke and Gorley, 2006) was

PT

employed to investigate the spatial and temporal changes of fish community structure. Raw abundance data (number of fish per site per net) were square root transformed to reduce the

RI

variance. The Bray-Curtis similarity index, which is widely used in similarity analysis of species

SC

composition in community ecology (Bloom, 1981) was employed to test the null hypothesis of no spatial and temporal difference in fish community composition at fishing sites (Bray and

NU

Curtis, 1957). Cluster analysis and multidimensional scaling (MDS) were used for graphical visualization of fish community structure onto two and three dimensional plots, respectively.

MA

Analysis of Similarity (ANOSIM) (Clarke and Warwick, 2001; Clarke and Gorley, 2006) was used to examine changes in fish community.

Similarity of Percentage Analysis (SIMPER) (Clarke and Warwick, 2001; Clarke and

D

Gorley, 2006) was used to determine which species, if any, contributed to spatial and temporal

TE

dissimilarities by reporting their individual rank contribution to overall average dissimilarity, contribution to dissimilarity (δ), and cumulative percentage contribution (∑δi %). The cut-off

AC CE P

dissimilarity was set at 50% cumulative contribution. Disturbed and undisturbed sites were detected by the Abundance Biomass Comparison (ABC) method (Warwick, 1986). The ABC routine generated cumulative abundance/biomass curves with either positive or negative W (Warwick) statistic values. Positive values classed the fishing sites as undisturbed when biomass curves were above the abundance curves and negative W values classed fishing sites as disturbed/polluted when abundance curves were above the biomass curves (Clarke and Warwick, 2001). Shannon (H’=∑Pi log (Pi)) (Shannon and Weaver, 1963) and Margalef diversity indices (d=(S-1)/ log N) (Margalef, 1958) measured fish diversity from three fishing sites.

ACCEPTED MANUSCRIPT 3. Results A total of 628 individual fishes and 51 species were identified from all catches. Table 2 shows the fish community composition and number of individuals sampled in February 2012.

PT

There were 10, 11, and 7 species in sites 1, 2, and 3 respectively. However, despite this small difference, the species caught at each site were not similar, e.g. about 50 percent of fish species

RI

found at site 1 were not found at sites 2 and 3. About 73 percent of fish species found at site 2

SC

were not found at sites 1 and 3. All species found at site 3 were either found at site 1 or 3 in February (Table 2), except Nyasachromis argyrosoma and Placidochromis suboccularis (Table

NU

2). Nyasachromis argyrosoma was only found once at site in February and never in another site. April had the highest number of fish species caught once; the species only caught at site 2

MA

include: Trematocranus microstoma, Caprichromis liemi, L. cf. parvidens, and Tramitichromis lituris while Labeo mesops, Chilotilapia rhoadesii, Otopharynx speciosus, Lethrinops oliveri, and Rhamphochromis esox were caught once at site 1. The incidence of new species resulted in

D

the number of species reaching maxima in April (Table 3). Fishing sites 1 and 2 had 23 different

TE

fish species each. L. parvidens, O. karongae, O. squamipinnis, Taeniolethrinops praeorbitalis were found at both sites 1 and 2. Site 2 consisted of 11 fish species absent in sites 1 and 3.

AC CE P

Likewise 9 fish species found in site 1 were not recorded in sites 2 and 3. Species recorded in all fishing sites include: L. lethrinus, Otopharynx argyrosoma, Pseodotropheus livingstonii, P. suboccularis, and T. placodon (Table 3). Site 3 had fewer species than sites 1 and 2. All fish species found in site 3 were either found in sites 1 or 2. There was a decline in fish species composition in June (Table 4). Fishing sites 1, 2, and 3 recorded only 11, 13, and 7 fish species respectively. Hemitilapia oxyrhynchus and P. suboccularis were the only two species found in all three fishing sites. Over 50 percent of fish species found in fishing site 1 were not found in sites 2 and 3. Similarly, over 50 percent of species found in fishing site 2 were found in sites 1 and 3 (Table 4). L. parvidens and P. livingstonii were the only two fish species found in sites 1 and 2 and not in site 3 in June, while S. woodi was only found in fishing sites 1 and 3. Rhamphochromis (long fin yellow) was only caught at site 3 in June during the study period. Buccochromis lepturus and Protomelas triaenodon were caught once at site 1. The last two species, Hemitaeniochromis urotaenia and Rhamphochromis ferox were caught at site 2 in August (Figure 5). The number of fish species declined in August (Table 5). Fishing sites 1, 2, and 3 had 7, 12, and 8 fish species respectively.

ACCEPTED MANUSCRIPT C. crysonotus, C. virginalis, Ctenopharynx intermedius, L. lethrinus, O. auromarginatus, and T. placodon were found in all fishing sites whilst Bathyclarias species was only found in site 1. Overall, fish community composition (Tables 2, 3, 4and5) fluctuated over the sampling

PT

period. Some fish species were represented by few individuals whilst others had ≥ 5 individuals per site or month. The highest number of individual fish was recorded in fishing site 2

RI

throughout the study and the lowest number of individuals was observed at site 1 (see Tables 2,

SC

3, 4and5). On the other hand fishing site 3 had relatively higher fish numbers than fishing site 1, except in April when both sites had 42 fish each (Table 4). C. chrysonotus, C. intermedius, L.

NU

lethrinus, O. argyrosoma, P. elegans, and T. placodon were represented by ≥5 individuals in site 2 in February. C. chrysonotus was the only species amounting to ≥5 individuals in site 1 and 3

MA

(Table 2, February 2012). In April (Table 3), the number individuals contributing with ≥5 fish was relatively higher in site 2 than site 1 with similar number of species but different in composition. Only P. suboccularis and T. placodon contributed with ≥5 individuals in site 1

D

while there were 9 species contributing with ≥5 species in site 2 where C. chrysonotus, O.

3).

TE

auromarginatus, and O. argyrosoma contributed with over 10 individuals in fishing site 2 (Table

AC CE P

June and August recorded low individual numbers (Tables 4 and 5). C. chrysonotus, C. pleurostigma, L. lethrinus, O. auromarginatus, P. livingstonii were represented with ≥5 individuals in site 2 (Table 4) while site 3 was only represented by C. chrysonotus and O. argyrosoma with ≥5 individuals. None of the 11 fish species found in site 1 accounted to ≥5 individuals (Table 4). Similarly, none of the 7 fish species found in site 1 in August had ≥5 individuals (Table 5). The presence of species with ≥5 individuals was reduced in sites 2 and 3, as C. chrysonotus and O. argyrosoma are the only species with ≥5 individuals in site 2 while only C. chrysonotus had ≥5 individuals in site 3 (Table 5). It is important to mention that the number of species doesn’t necessarily mean more individuals per species, e.g. in April, site 3 had almost half of fish species found in site 1, but each had 42 individuals (Table 3). Table 4 (June) shows 11 and 7 fish species in sites 1 and 3 respectively indicating relatively more species in site 1, but the individuals found in site 3 outnumbers those in site 1. However, regardless of variability in the number of fish species caught per site per net, the average number of fish species and biomass caught in experimental nets were not significantly different at the three

ACCEPTED MANUSCRIPT sampling sites (p>0.05), but the average abundance (number of fish per site per net) was

PT

significantly higher at site 2 than at sites 1 and 3 (Table 11) (p<0.05).

3.1 Fish community structure

RI

Visual inspection of the dendrogram plot (Figure 2a) for the whole fish community composition shows scattering of fish species from sites 1 and 3 to margins with a tendency for

SC

fish communities from site 2 to occur near the center. Similar to Figure 2a, the MDS plot (Figure 2b) shows a clear visualization of grouping near the center of fish species from site 2 and

NU

scattering of species from sites 1 and 3 in the upper and lower left and right sides in the three dimensional presentation. Scattering and grouping indicate a degree of similarity and

MA

dissimilarity in fish community at the fishing sites.

There was a significant spatial difference in fish community composition (Global R = 0.472, p=0.005). Pairwise test detected significant differences between fishing sites 1 and 2

D

(R=0.563, p=0.03) and between fishing sites 2 and 3 (R=0.563, p=0.04). No significant

TE

difference was observed between sites 1 and 3 (R=0.250, p=0.22). Significant temporal differences in fish community structure were also detected (Global test R=0.472, p=0.004).

AC CE P

Pairwise tests reveal significant temporal difference of fish community structure between (February and April, R=0.917, p=0.037), (February and June, R=0.667, p=0.037), and between (June and August, R=0.583, p=0.037). There was no temporal significant difference of fish community structure between (February and August, R=-0.167, p=0.667), and (April and June, R=167, p=0.37), and between (April and August, R=0.417, p=0.148). The list of fish species responsible for both spatial and temporal dissimilarity is presented in Tables 6 to 10. Table 6 shows eight fish species (C. chrysonotus, O. auromarginatus, C. intermedius, L. lethrinus, O. argyrosoma, T. placodon, P. suboccularis, and C. virginalis) representing 50% of cumulative contribution of species dissimilarity between sites 1 and 2. The dissimilarity between fishing sites 2 and 3 was made of seven fish species, the same species that contributed to dissimilarity between sites 1 and 2 with exception of the absence of C. virginalis. The cumulative contribution of species dissimilarity between February and April was made by nine species (see Table 8). While S. njassae added to the number of already known species contributing to dissimilarity in April (Table 8), H. oxyrhynchus and L. parvidens also added the number of species contributing to community dissimilarity between February and June

ACCEPTED MANUSCRIPT (Table 9). The species that contributed to dissimilarity between June and August (Table 10) are similar to those contributing to species dissimilarity in other months; however, S. woodi is contributing to community dissimilarity for the first time.

PT

Overall, many species contributed to a fairly high certain degree of spatial and temporal dissimilarity. C. chrysonotus, C. intermedius, L. lethrinus, O. argyrosoma, T. placodon, and P.

RI

suboccularis are the six fish species that contributed to both spatial and temporal dissimilarity. C.

SC

chrysonotus accounted for the highest number of individuals whilst the numbers of other species fluctuated in space and time. Regardless of species dissimilarities, no significant difference was

NU

noted in Shannon and Margalef diversity indices among the fishing sites (Table 11).

MA

3.2 Health status of fishing sites

The fishing sites had significantly different W-statistic values (Table 11). Fishing site 1 had significantly higher W-statistic value (0.30±0.01) than fishing sites 2 (0.007±0.020) and 3

D

(0.006±0.05) (p<0.05) and was classed as undisturbed as revealed by dominance of biomass

TE

curves over abundance curves over the sampling period (Figure 3, F1, AP1, J1, and AU1). The biomass and abundance curves in sites 2 and 3 were closely coincident and crossed each other, at

AC CE P

least once when the sites were moderately disturbed; and, when the sites were grossly disturbed the abundance dominated over biomass curve and sometimes overlapped (Figure 3). Fishing site 2 exhibited moderate disturbance in February (Figure 3, F2), April (Figure 3, AP2), and June (Figure 3, J2) and gross disturbance in August (Figure 3, AU2). Fishing site 3 exhibited moderate disturbance in February (Figure 3, F3) and April (Figure 3, AP3) and gross disturbance in June (Figure 3, J3) and August (Figure 3, AU3).

4. Discussion The total number of fish species (51) reported in the current study is an aggregate number of all species caught in experimental nets at sites 1, 2, and 3 in February, April, June, and August 2012. The numbers of fish species were variable at the site level. Fish species were either found at all sites in a specific month or were just found at one site and not at the other sites. In some cases, replicate experimental nets caught different fish species at the same site. It is known that some fish species may exhibit uniform or random distribution which rarely occur in nature, while

ACCEPTED MANUSCRIPT others may have aggregated and gradient distribution (King, 1995), or their size structure, frequency and spatial distribution may reflect human exploitation (Haddon, 2001). The cage aquaculture industry in SEA of Lake Malawi adds food in this oligotrophic lake (Gondwe et al.,

PT

2012; Macuiane, 2014), suggesting that the species distribution found in the current study may reflect differential resource availability at each site.

RI

Site 2 located at the fish farm had consistently higher fish abundance compared to sites 1

SC

and 3 (p<0.05), thus supporting the complaints from local fishermen that fish abundance is skewed towards the farm where the fish probably take refuge and does not depend on local

NU

variability in lake conditions at the site. High fish abundance, using underwater camera in the same farm have been reported by Gondwe et al. (2011, 2012), and elsewhere, aggregation of

MA

wild fish populations have been reported to increase in cage aquaculture farms with increase in the number of cages (Dempster et al., 2002; Tuya et al., 2006) and also during the feeding time of captive fish (Tuya et al., 2006; Bacher, 2015). There were more wild fish around the farm in

D

2010 when the farm had more cages stocked than in 2012 when the farm had few stocked fish.

TE

Gondwe et al. (2011) estimated large losses of artificial feed supplied to captive fish in the cages into the surrounding environment and applied stable isotope analysis of carbon and nitrogen to

AC CE P

prove that the wild fish around Maldeco Aquaculture farm consume uneaten feeds and other organic wastes emanating from the cages, suggesting that feed losses may be responsible for high abundance of wild fish at site 2. Consumption of cage wastes by wild fish and other fauna in the vicinity of fish farms has been demonstrated all around the world (Beveridge, 1984; Dempster et al., 2002; Sudirman et al., 2009). For instance, in the Mediterranean Sea, cage wastes accounted for about 80% of the diet of clams cultivated around a fish farm (Mazzola and Sara, 2001). In the Mediterranean Sea, wild fish consumed an estimated 80% of organic wastes that sank from fish cages (Vita et al., 2004) while Phillips et al. (1985) estimated that feed pellets from fish cages constituted about 98% of the gut contents of wild fish around a salmon fish farm. Elsewhere, wild fish populations have been reported to consume 27% of the lost food pellets (Sudirman et al., 2009). The percentage of lost pellets might even be higher at Maldeco Aquaculture Limited farm because the farm uses sinking pellets which are easily lost to the environment; but the impact of feed pellets depends on currents which may sweep them away from the feeding site, thus increasing the area of the impact but diminishing its impact as a local food resource as found by Macuiane

ACCEPTED MANUSCRIPT (2014). Dependence on artificial feeds emanating from the cages instead of consumption of natural diets may change the feeding behavior of wild fish populations and may lead to a change in the body composition and condition of fish (Skog et al., 2003; Fernandez-Jover et al., 2007).

PT

While our field study cannot provide conclusive results about the behavior of wild fish populations in regard to changes in feeding behavior, it suggests that wild fish populations stay

RI

long enough at the farm to result in higher average numbers of fish compared to sites remote

SC

from the farm. Wild fish follow boats during the feeding period, a behavior expressed by captive fish when they hear the noise of engine boats when the feeding crew approaches the farm.

NU

Similarities in fish community structures at remote sites 1 and 3 located along a 10 km transect through the farm (R=0.250, p=0.222) suggests that the farm has little or no effect on fish communities at this scale and suggests the farm impacts on fish community may be spatially

MA

quite limited. This study didn’t find any statistical evidence of biomass differences among the three sites which could be due to high variability of fish sizes caught in the experimental nets.

D

The catches at Site 1 were characterized by few numbers of large bodied species such as B.

TE

meridionalis and Bathyclarias species while catches at sites 2 and 3 were dominated by small bodied species such as C. chrysonotus, L. lethrinus, and O. auromarginatus (Tables 2 to 5).

AC CE P

Thus, based on ABC criteria (Figure 3) for size distribution, site 1 was classified as undisturbed while sites 2 and 3 are considered disturbed. Site 2 was attested as disturbed mainly due to abundance of large numbers of small bodied fish species. These small fishes may be attracted and maintained by the cages/farm resources (uneaten feeds, feces, shelter). The farm provides protection of fish from overexploitation by local fishermen, thus it serves as a site where fish may have both trophic benefits and protection. The grossly disturbed status of fishing site 3 may be due to overfishing in this site. Illegal (non-selective) seine nets, locally known as “kandwindwi” prohibited by the Government of Malawi are actively used in the area of this site by local fishermen who do not comply fisheries laws. While restoration strategies that include restriction of fishing gears (mesh size and head line length), establishment of fish sanctuaries, artificial reefs, community property rights to fishing areas, and restocking of the fishery to reduce fishing pressure in Malawian waters have been implemented with no success (Banda et al., 2005); the Maldeco Aquaculture site seems to act as a sanctuary that protects fish communities and maintain high fish abundance,

ACCEPTED MANUSCRIPT therefore, fishing regulations should also recognize the protection of wild fish populations at the cage farm sites where fishing access and exploitation are controlled by cage farm operators. Non-significant difference in the number of fish species at sites 1, 2, and 3 (p=0.069)

PT

does not necessarily suggest similar fish communities and/or similar abundance at these sites, but habitat preferences of fish probably to prevent niche overlap thereby maintaining coexistence of

RI

some fish populations. The Shannon and Margalef diversity indices which are widely applied

SC

measures of species diversity showed no significant difference in species diversity among the fishing sites (Table 11), indicating that the cage aquaculture farm at site 2 and fishing pressure at

NU

sites 1 and 3 have not yet affected fish diversity although intensification of farming activity may eventually increase the risk.

MA

Previous studies in SEA (Tweddle and Turner, 1977; Banda et al., 1996; Weyl et al., 2010) observed that overexploitation and increasing fishing pressure have depleted the larger commercial fish species, changed species composition of catches, and caused local extirpations,

D

reduced the contribution of fish to less than 30% of Malawi’s total animal protein supply, and

al., 2005).

TE

per capita fish consumption dropped from 14 kg in mid-1970 to below 3 kg in 2003 (Banda et

AC CE P

To reduce pressure on the wild fish artisanal and commercial fish populations, Weyl et al. (2010), recommended no addition of pair-trawl effort in SEA, effort elimination in Area “A” in the SEA, and the artisanal and pair-trawl fisheries should be managed similarly. However, despite reports that the fish diversity of SEA is already affected, comparisons made on the three fishing sites suggest no effect on wild fish diversity at these inshore locations yet; however, the major concern emerging for our study area is the incidence of small bodied fish populations which are unevenly distributed among these fishing sites which may indicate continuing disturbance. It possible that the increased abundance of small fishes at the two sites (2 and 3) may have different causes. If site 2 is acting as a sanctuary and food source it may be allowing young fish to have increased survival and abundance while at site 3 high fishing pressure have increased mortality on larger fishes leaving smaller fish to dominate. In both cases small fish become more prominent but causes are different.

ACCEPTED MANUSCRIPT 5. Conclusion 1. The current study has established that Maldeco Aquaculture farm has a near field impact on fish community structure in the SEA of Lake Malawi. Site 2 had significantly

PT

different fish community structure with that found at sites 1 and 3.

2. Margalef and Shannon diversity indices gave no evidence of differences in fish species

RI

diversity among the locations studied in SEA of Lake Malawi.

SC

3. The ABC criteria classed site 1 as undisturbed due to abundance of large bodied species and sites 2 and 3 as moderately and grossly disturbed, respectively due the incidence of

NU

small sized fish. Disturbance in the current study may suggest a different interpretation of classical use of ABC criteria. Protection and food attract young fish in site 2 while high

MA

mortality of large fish leave small fish dominant.

4. The Maldeco Aquaculture site acts as a sanctuary that protects fish communities and maintains high fish abundance. Recognition of cage sites as sites for protection from high

D

fishing exploitation as they are controlled by cage farm operators seems to be a valuable

TE

strategy. Thus, the Fisheries Department of Malawi should consider the role of cage sites as potential sanctuaries that protect fish stocks attracted to cage aquaculture farms in the

AC CE P

SEA of Lake Malawi where the catch per effort data used by fisheries managers may not be reliable to quantify the abundance/biomass of available fish stock in the lake and to maintain them for future generations.

Acknowledgement

Funding was provided by the Office of International Programs at the University of Minnesota. We wish to thank the Malawi Department of Fisheries for allowing us conduct this study in Malawi. Special thanks go to the managing board of Maldeco Aquaculture Limited and staff for granting us permission and support to conduct the study at the farm premise site. We thank the Malawi College of Fisheries for provide laboratory facility and accommodation during the sampling period. We are grateful to Mr. Manivesta and his crew members for joining us during sampling. We thank Dr. Mangaliso Gondwe for proving valuable inputs to this manuscript.

ACCEPTED MANUSCRIPT References Abery, N.W., Sukadi, F., Budhiman, A.A., Kartamihardja, E.S., Koeshendrajana, S., De Silva,

PT

S.S., 2005. Fisheries and cage culture of three reservoirs in west Java, Indonesia; a case study of ambitious development and resulting interactions. Fish. Manag. Ecol. 12, 315-

RI

330.

Bacher, K., Gordoa, A, Sagué, O., 2015. Feeding activity strongly affects the variability of wild

SC

fish aggregations within fish farms: a sea bream farm as a case study. Aquacult Res. 46(3), 552-564.

NU

Banda, M.C., Kanyerere, G.Z., Rusuwa, B.B., 2005. The status of the chambo in Malawi: fisheries and biology, in: Banda, M., Jamu, D., Njaya, F., Makuwila, M., Maluwa, A.

MA

(Eds.), The Chambo Restoration Strategic Plan. Proceedings of the National Workshop: 13–16 May 2003, Mangochi: WorldFish Center Conference Proceedings, 71, pp. 1–7.

D

Banda, M.C., Tomasson, T., Tweddle, D., 1996. Assessment of the deep water trawl fisheries of

TE

the southeast arm of Lake Malawi using exploratory surveys and commercial catch data, in: Cowx, I.G. (Eds.), Stock Assessment in Inland Fisheries. Blackwell Science, Oxford,

AC CE P

pp. 53–75.

Beveridge, M.C.M., 1984. Cage and pen fish farming. Carrying capacity models and environmental impact. FAO Fish. Tech. Pap., (255): pp. 131. Bloom, S.A., 1981. Similarity indices in community studies: potential pitfalls, Mar. Ecol. Prog. Ser. 5, 125-128.

Bootsma, H.A., Hecky, R. E., 1993. Conservation of the African Great Lakes: a limnological perspective. Conserv. Biol. 7(3), 644–656. Bray, J.R., Curtis, J.T., 1957. An ordination of the upland forest communities of Southern Wisconsin. Ecol. Monogr. 4(27), 325-349. cage culture on inland fisheries: present status and future trends. J. Fish Biol. 27, 123– 137. Clarke, K.R., Warwick, R.M., 2001. Change in marine communities: an approach to statistical analysis and interpretation, second ed. PRIMER-E, Plymouth. Dempster, T., Sanchez-Jerez, P., Bayle-Sempere, J.T., Gime T., Gime.,Gim, F., Valle, C., 2002. Attraction of wild fish to sea-cage fish farms in the south-western Mediterranean Sea: spatial and short-term temporal variability. Mar. Ecol. Prog. Ser. 242, 237–252.

ACCEPTED MANUSCRIPT Department of Water/UNDP., 1986. National water resources master plan – Reports and Appendices, UNDP Projects MLW-79-015/MLW-84 – 003, Malawi. Duponchelle, F., Ribbink, A.J., Msukwa, A., Mafuka, J., Mandere, D., 2003. Seasonal and

PT

spatial patterns of experimental trawl catches in the south east arm of Lake Malawi. J. Great. Lakes Res. 2, 216-231.

RI

Eccles, D.H., Trewavas, E., 1989. Malawian cichlid fishes; the classification of some

SC

haplochromine genera. Lake Fish Movies, Herten, Germany.

FAO., 2005-2014. National aquaculture sector overview. Malawi. National Aquaculture sector

NU

overview fact sheets, in: Chimatiro, S.K., Chirwa, B.B. (Eds.), FAO Fisheries and Aquaculture Department. Rome. Updated 1 January 2005.

MA

http://www.fao.org/fishery/countrysector/naso_malawi/en Fernandez-Jover, D., Lopez-Jimenez, J. A., Sanchez-Jerez, P., Bayle-Sempere, J., GimenezCasalduero, F., Martinez-Lopez, F.J., Dempster, T., 2007. Changes in body condition and

D

fatty acid composition of wild Mediterranean horse mackerel (Trachurus mediterraneus,

TE

Steindachner, 1868) associated with sea cage fish farms. Mar. Environ. Res. 63, 1-18. Fryer, G., Iles, T.D., 1972.The cichlid fishes of the Great Lakes of Africa; their biology and

AC CE P

evolution. Edinburgh: Oliver and Boyd. Gondwe, M.J., 2009. Environmental impacts of cage aquaculture in the southeast arm of Lake Malawi: water and sediment quality and

food web changes. PhD Thesis. University

of Waterloo., Ontario, Canada. Gondwe, M.J., Guildford, S.J., Hecky, R.E., 2011. Physical–chemical measurements in the water column along a transect through a

tilapia cage fish farm in Lake Malawi, Africa. J.

Great. Lakes Res 37 (1), 102–113. Gondwe, M.J., Guidford, S.J., Hecky, R.E., 2012. Tracing the flux of aquaculture-derived organic wastes in the southeast arm of Lake Malawi using carbon and nitrogen stable isotopes. Aquaculture 350-353, 8-18. Haddon, M., 2001. Modeling and quantitative methods in fisheries. Chapman and Hall/CRC. Government of Malawi (GoM)., 2009. Fish Frame Survey Report. Department of Fisheries, Lilongwe. Jamu, D., Banda, M., Nyaja, F., Hecky, R., 2011. Challenges to sustainable management of the lakes of Malawi. J. Great. Lakes Res. Supplement 1, 3-14.

ACCEPTED MANUSCRIPT Jamu, D., Chimatiro, S., 2005.Sustainable Agro-pisciculture in Malawi. Agriculture and Rural Development: Contributing to International Cooperation, 12, (2), 45-46. Kanyika, G.Y., 2000. Water balance model for Lake Malawi, M.Sc. thesis, University of Dar es

PT

Salaam, Tanzania.

King, M.,1995. Fisheries biology, assessment and management. Fishing News Books, p 341.

RI

Konings, A., 1990. Cichlids and all the Fishes of Lake Malawi. Tropical Fish Hobbyist

SC

Publication, Neptune City, New Jersey, USA.

Konings, A., 1995. Malawi cichlids in their natural habitat. 2nd ed. Germany: Cichlid Press.

NU

Macuiane, M.A., 2014. Cage aquaculture and environment in Lake Malawi: an assessment of water quality, food web shifts, and development of a decision support tool for sustainable

MA

aquaculture. A PhD dissertation submitted to the Faculty of the Graduate School of the University of Minnesota, pp. 229.

Margalef, R., 1958. Information theory in ecology. General Systems Yearbook 3, 36-71.

D

Mazzola, A., Sarà, G., 2001. The effect of fish farming organic waste on food availability for

TE

bivalve molluscs (Gaeta Gulf, Central Tyrrhenian, MED): stable carbon isotopic analysis. Aquaculture 192, 361–379.

AC CE P

Patterson, G., Kachinjika, O., 1995. Limnology and phytoplankton ecology, in: Menz, A (Eds.), The fishery potential and production of the pelagic zone of Lake Malawi/Niassa. Chatham, UK: Natural Resources Institute, pp. 1-67. Phillips, M.J., Beveridge, M.C.M., Ross, L.G., 1985. The environmental impact of salmonid cage culture on inland fisheries: present status and future trends. J. Fish Biol. Vol. 27. Suppl. A, 123-137.

Ribbink, A. J., Marsh, B. A., Marsh, A. C., Ribbink, A. C., Sharp, B. J. 1983. A preliminary survey of the cichlid fishes of rocky habitats in Lake Malawi. S. Afr. J. Zool. 18, 149310. Skog, T.E., Hylland, K., Torstensen, B.E., Berntssen, M.H.G., 2003. Salmon farming affects the fatty acid composition and taste wild sai the Pollachiusvirens L. Aquacult Res. 34, 9991007. Snoeks, J., 2000. How well known is the ichthyodiversity of the large East African lakes? Adv. Ecol. Res. 31, 17-38.

ACCEPTED MANUSCRIPT Sudirman., Halide, H., Jompa, J., Zulfikar., Iswahyudin., Mcknnin, A. D., 2009. Wild fish associated with tropical sea cage aquaculture in South Sulawesi, Indonesia. Aquaculture 286, 233-239.

PT

Thompson, A.B., Allison, E.H., Ngatunga, B.P., 1995. Spatial and temporal distribution of fish in the pelagic waters, in: Menz, A. (Ed.). The fishery potential and productivity of pelagic

RI

zone of Lake Malawi/Niassa. Chathan, Uk: Natural Resources Institute, pp., 201- 232.

SC

Tuya, F., Sanchez-Jerez, P., Dempster, T., Boyra, A., Haroun, R. J., 2006. Changes in demersal wild fish aggregations beneath a sea-cage farm after the cessation of farming. J. Fish

NU

Biol. 69, 682–697.

Tweddle, D., Turner, J.L., 1977. Age, growth and natural mortality rates of some cichlidfishes of

MA

Lake Malawi. J. Fish Biol. 10, 385–398.

Vita, R., Martin, A., Madrid, J.A., Jimenez-Brinquis, B., Cesar, A., Marin-Guirao, L., 2004. Effects of wild fishes on waste exportation from a Mediterranean fish farm. Mar. Ecol.

D

Prog. Ser. 277, 253–261.

TE

Warwick, R.M., 1986. A new method for detecting pollution effects on marine macrobenthic communities. Mar. Biol. 92, 557-562.

AC CE P

Weyl, O.F., Ribbink, A.J., Tweddle, D., 2010. Lake Malawi: fishes, fishery, biodiversity, health and habitat. Aquat. Ecosys. Health Manage. 13(3), 241-254.

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

Figure 1 Lake Malawi (left panel) and fishing sites (ST-1,1 is fishing site 1 replicate 1, ST-1,2 is fishing site 1 replicate 2, ST-2,1 is fishing site 2 replicate 1, ST-2,2 is fishing site 2 replicate 2, ST-3,1 is fishing site 3 replicate 1, and ST-3,2 is fishing site 3 replicate 2) including Maldeco Aquaculture farm (right panel).

ACCEPTED MANUSCRIPT 0

Site

a)

1 2 3

PT

40

AU12

F12

F32

AP31

AP22

J32

AP21

AU22

AU21

F22

J21

F21

AU32

F31

F11

AP32

J12

AP11

AP12

J22

J11

AU11

100

NU

80

J31

SC

RI

60

AU31

Bray-Curtis similarity

20

MA

Fishing site

3D Stress: 0.14

TE

AC CE P

J22 AP21

Site 1 2 3

D

b)

J12

J11

AP11 AP22 J21 AU21 AP32 J32 F21 F22 AU31 J31 AU22 AU32 AU11 F32AP31 AU12 F31 F12

AP12

F11

Figure 2 a) Cluster analysis dendrogram based on Bray-Curtis similarity between samples for whole fish assemblage and b) Non-metric multi-dimensional scaling (NMDS) plot of spatial and temporal and distribution of fish community structure. Symbols represent sites, open circles represent site 1, black circles represent site 2, and the x represent site 3. Symbol IDS are described in Table 1.

ACCEPTED MANUSCRIPT

F1

100

AP1

100 80

80

W = 0.366

40

20

RI

20 0

1

10

J1

10

100

80

80

60

NU

60

W = 0.442

W = 0.316

40

40

20 1

10

1

100

F2

100

MA

20

0

100

AU1

SC

100

1

10

AP2

100

D

80 80

TE

60

W = -0.013

60

W = 0.011

40 20

40 1

AC CE P

C U M M U L A TI V E % D O M I N A N C E

W = 0.544

40

PT

60

60

10

J2

100 80

0 1

100

10

100

AU2

100 90

60

W = 0.071

40

80

W = -0.056

70

20

60 1

10

100

1

10

SPECIES RANK

100

ACCEPTED MANUSCRIPT F3

100

AP3

100 80

80 60

60

W = -0.012

PT

W = 0.043

40

1

J3

10

100 90

80

80

W = -0.14

40

W = -0.139

NU

60

70 60 50

1

10

MA

20

100

AU3

SC

100

1

10

RI

20

40

1

10

SPECIES RANK

D

Figure 3 Abundance Biomass Curves (ABC) curves for fish community in Lake Malawi.

AC CE P

TE

Abundance ( ) and biomass ( ). Symbol ID’s are described in Table 1.

ACCEPTED MANUSCRIPT Table 1 Sample IDs for the various fish samples collected from the three sites along a transect through the fish farm in Lake Malawi. Site 3 5 km southeast Replicate 1 Replicate 2 F31 F32 A31 A32 J31 J32 AU31 AU32

PT

Site 2 at the fish farm Replicate 1 Replicate 2 F21 F22 A21 A22 J21 J22 AU21 AU22

NU

SC

February April June August

Site1 5 km northwest Replicate 1 Replicate 2 F11 F12 A11 A12 J11 J12 AU11 AU12

RI

Month

D

Site 2 Copadichromis chrysonotus(52)+ Caprichromis orthognathus(2) Corematodus taeniatus(1) Ctenopharynx intermedius(16)# Lethrinops lethrinus(10)+ Otopharynx auromarginatus(6) Otopharynx sp(1) Pseudotropheus elegans(7) Protomelas similis(2) Trematocranus microstoma(1) Trematocranus placodon(6)x No. of species 11(104)

AC CE P

TE

Site 1 Bagrus meridionalis(1) Copadichromis chrysonotus(5)+ Copadichromis quadrimaculatus(2) Copadichromis virginalis(2)* Lethrinops lethrinus(1)+ Mylochromis anaphyrmus(2) Oreochromis karongae (1) Placidochromis platyrhynchos(2) Stigmatochromis guttatus(1)* Trematocranus placodon(1)x

MA

Table 2 Fish species and numbers (in parentheses) caught in February 2012.

No. of species 10(18) +

Site 3 Copadichromis chrysonotus(16)+ Copadichromis virginalis(1)* Ctenopharynx intermedius(7)# Lethrinops lethrinus(1)+ Nyasachromis argyrosoma(2) Placidochromis suboccularis(1) Stigmatochromis guttatus(1)*

No. of species 7(29)

Fish species found in fishing sites 1, 2, and 3; *Fish species found only in fishing sites 1 and 3 Fish species found only in fishing sites 1 and 2; #Fish species found only in sites 2 and 3.

x

ACCEPTED MANUSCRIPT Table 3 Fish species and numbers (in parentheses) caught in April 2012.

PT

RI

SC

NU

MA

Sites 3 Buccochromis rhoadesii(1)# Clarias gariepinus(1)* Copadichromis virginalis(6)* Ctenopharynx intermedius(5)* Lethrinops lethrinus(10)+ Mylochromis melanotaenia(2)* Otopharynx argyrosoma(10)+ Pseodotropheus livingstonii(1)+ Protomelas similis(1)# Placidochromis suboccularis(1)+ Stigmatochromis woodi(1)# Synodontis njassae(1)* Trematocranus placodon(2)+

+

AC CE P

TE

Copadichromis pleurostigma(2) Copadichromis virginalis(1)* Ctenopharynx intermedius(4)* Lethrinops parvidens(2)x Labeo mesops(1) Lethrinops lethrinus(4)+ Lethrinops oliveri(1) Mylochromis anaphyrmus(1) Mylochromis melanotaenia(1)* Oreochromis karongae(1)x Oreochromis squamipinnis(1)x Otopharynx argyrosoma(2)+ Otopharynx speciosus(1) Pseodotropheus livingstonii(1)+ Placidochromis suboccularis(6)+ Protomelas kirkii(1) Rhamphochromis esox(1) Synodontis njassae(1)* Taeniolethrinops praeorbitalis(2)x Trematocranus placodon(5)+ No. of species 23(42)

Site 2 Buccochromis rhoadesii(1)# Caprichromis liemi(1) Copadichromis chrysonotus(27) Copadichromis quadrimaculatus(4) Dimidiochromis kiwinge(5) Fossorochromis rostratus(6) Hemitaeniochromis spilopterus(4) Hemitilapia oxyrhynchus(6) L. cf. parvidens(6) Lethrinops parvidens(1)x Lethrinops lethrinus(3)+ Oreochromis karongae(1)x Oreochromis squamipinnis(1)x Otopharynx auromarginatus(14) Otopharynx argyrosoma(14)+ Pseodotropheus livingstonii(3)+ Protomelas similis(7)# Placidochromis suboccularis(4)+ Stigmatochromis woodi(4)# Tramitichromis lituris(1) Taeniolethrinops praeorbitalis(1)x Trematocranus microstoma(4) Trematocranus placodon(5)+ No. of species 23 (123)

D

Site 1 Bathyclarias species(1) Chilotilapia rhoadesii(1) Clarias gariepinus(1)*

No. of species 13(42)

Fish species found in fishing sites 1, 2, and 3; *Fish species found only in fishing sites 1 and 3 Fish species found only in fishing sites 1 and 2; #Fish species found only in sites 2 and 3.

x

ACCEPTED MANUSCRIPT Table 4 Fish species and numbers (in parentheses) caught in June 2012.

No. of species 11(21)

Site 3 Copadichromis chrysonotus(15)# Hemitilapia oxyrhynchus(1)+ Lethrinops lethrinus(3)# Otopharynx argyrosoma(14) Placidochromis suboccularis(3)+ Rhamphochromis(long fin yellow) (1) Stigmatochromis woodi(2)*

NU

SC

RI

PT

Site 2 Buccochromis nototaenia(4) Copadichromis chrysonotus(17)# Copadichromis pleurostigma(5) Dimidiochromis kiwinge(2) Fossorochromis rostratus(3) Hemitaeniochromis spilopterus(2) Hemitilapia oxyrhynchus(4)+ Lethrinops parvidens(2)x Lethrinops lethrinus(8)# Otopharynx auromarginatus(5) Pseudotropheus elegans(1) Pseodotropheus livingstonii(5)x Placidochromis suboccularis(2)+ No. of species 13(60)

MA

Site 1 Buccochromis lepturus(2) Copadichromis virginalis(1) Hemitilapia oxyrhynchus(2)+ Lethrinops parvidens(3)x Mylochromis melanotaenia(2) Pseodotropheus livingstonii(2)x Protomelas triaenodon(1) Placidochromis suboccularis(4)+ Protomelas kirkii(1) Stigmatochromis woodi(1)* Trematocranus placodon(2)

+

No. of species 7(39)

Fish species found in fishing sites 1, 2, and 3; *Fish species found only in fishing sites 1 and 3 Fish species found only in fishing sites 1 and 2; #Fish species found only in sites 2 and 3

TE

D

x

Table 5 Fish species and numbers caught (in parentheses) in August 2012.

AC CE P

Site 1 Bathyclarias species(1) Copadichromis chrysonotus(2)+ Copadichromis virginalis(2)+ Ctenopharynx intermedius(1)+ Lethrinops lethrinus(3)+ Otopharynx auromarginatus(2)+ Trematocranus placodon(2)+

No. of species 7(13) +

Site 2 Copadichromis chrysonotus(65)+ Copadichromis virginalis(1)+ Ctenopharynx intermedius(1)+ Hemitaeniochromis Urotaenia(2) Lethrinops parvidens(1) Lethrinops lethrinus(4)+ Otopharynx auromarginatus(3)+ Otopharynx argyrosoma(8) Placidochromis suboccularis(2) Rhamphochromis ferox(1) Stigmatochromis woodi(1) Trematocranus placodon(2)+ No. of species 12(91)

Site 3 Buccochromis nototaenia(1) Copadichromis chrysonotus(35)+ Copadichromis virginalis(1)+ Ctenopharynx intermedius(2)+ Lethrinops lethrinus(2)+ Otopharynx auromarginatus(2)+ Placidochromis suboccularis(2) Trematocranus placodon(1)+

No. of species 8(46)

Fish species found in fishing sites 1, 2, and 3; *Fish species found only in fishing sites 1 and 3 Fish species found only in fishing sites 1 and 2; #Fish species found only in sites 2 and 3

x

ACCEPTED MANUSCRIPT Table 6 SIMPER analysis for fish community structure between fishing sites 1 and 2, with dissimilarity cut-off set at 50% cumulative contribution. Mean abundance (square rooted data) expressed as individuals per site.

a

Cumulative % contribution (∑δi %) 19.16 26.41 31.90 37.22 41.88 46.39 49.64 52.85

RI

PT

Contributiona (δi ) 1.82 1.52 0.92 0.98 0.75 1.11 0.96 0.94

SC

Copadichromis chrysonotus Otopharynx auromarginatus Ctenopharynx intermedius Lethrinops lethrinus Otopharynx argyrosoma Trematocranus placodon Placidochromis suboccularis Copadichromis virginalis

Mean abundance Site 1 Site 2 0.57 4.29 0.18 1.57 0.48 0.83 0.68 1.53 0.18 1.14 0.76 0.88 0.75 0.73 0.60 0.13

NU

Species

TE

D

MA

Ranked contribution of each species to overall average dissimilarity (δ= 78.40) between sites 1 and 2.

AC CE P

Table 7 SIMPER analysis for fish community structure between fishing sites 2 and 3, with dissimilarity cut-off set at 50% cumulative contribution. Abundance (square rooted data) expressed as individuals per site. Mean abundance Contributiona Cumulative % Site 2 Site 3 (δi ) contribution (∑δi %) Copadichromis chrysonotus 4.29 2.38 1.63 11.33 Otopharynx auromarginatus 1.57 0.18 1.53 20.21 Otopharynx argyrosoma 1.14 1.14 0.98 28.30 Lethrinops lethrinus 1.53 1.15 1.61 35.64 Trematocranus placodon 0.88 0.30 1.02 40.61 Placidochromis suboccularis 0.73 0.73 0.99 45.43 Ctenopharynx intermedius 0.83 0.79 0.75 50.12 a Ranked contribution of each species to overall average dissimilarity (δ= 63.53) between sites 2 Species

and 3.

ACCEPTED MANUSCRIPT Table 8 SIMPER analysis for fish community structure between February and April, with dissimilarity cut-off set at 50% cumulative contribution. Abundance (square rooted data) expressed as individuals per month.

a

Cumulative % contribution (∑δi %) 11.13 19.62 27.94 34.33 40.30 44.66 49.00 51.57

RI

PT

Contributiona (δi ) 1.34 1.40 1.64 1.24 1.13 1.12 0.86 0.66

SC

Copadichromis chrysonotus Ctenopharynx intermedius Otopharynx argyrosoma Lethrinops lethrinus Placidochromis suboccularis Trematocranus placodon Copadichromis virginalis Synodontis njassae

Mean abundance February April 3.12 1.21 1.38 0.84 0.00 1.82 1.04 1.47 0.17 1.21 0.74 0.98 0.00 0.74 0.00 0.33

NU

Species

TE

D

MA

Ranked contribution of each species to overall average dissimilarity (δ= 78.52) between February and April 2012.

AC CE P

Table 9 SIMPER analysis for fish community structure between February and June, with dissimilarity cut-off set at 50% cumulative contribution. Abundance (square rooted data) expressed as individuals per month. Mean abundance Contributiona Cumulative % February June (δi ) contribution (∑δi %) Copadichromis chrysonotus 3.12 1.82 2.31 9.74 Ctenopharynx intermedius 1.38 0.00 0.87 17.35 Placidochromis suboccularis 0.17 1.09 1.26 24.33 Otopharynx argyrosoma 0.00 0.81 0.61 30.99 Lethrinops lethrinus 1.04 0.87 1.20 36.52 Hemitilapia oxyrhynchus 0.00 0.74 0.91 41.26 Lethrinops parvidens 0.00 0.64 0.87 45.74 Trematocranus placodon 0.74 0.24 0.97 50.18 a Ranked contribution of each species to overall average dissimilarity (δ= 73.95) between February and June 2012. Species

ACCEPTED MANUSCRIPT Table 10 SIMPER analysis for fish community structure between June and August, with dissimilarity cut-off set at 50% cumulative contribution. Abundance (square rooted data) expressed as individuals per month. Mean abundance Contributiona Cumulative % June August (δi ) contribution (∑δi %) Copadichromis chrysonotus 1.82 3.50 1.52 10.42 Otopharynx argyrosoma 0.81 0.64 0.95 19.90 Lethrinops lethrinus 0.87 1.09 1.70 28.58 Placidochromis suboccularis 1.09 0.47 1.20 36.67 Otopharynx auromarginatus 0.52 0.76 0.91 43.10 Lethrinops parvidens 0.64 0.17 0.88 48.42 Stigmatochromis woodi 0.50 0.17 1.20 53.67 a Ranked contribution of each species to overall average dissimilarity (δ= 72.55) between June

NU

SC

RI

PT

Species

MA

and August 2012.

Parameter

TE

D

Table 11 One-Way ANOVA and mean fish abundance, biomass, number of fish species, diversity indices, and W-statistic values at fishing sites 1, 2, and 3. Significant levels were estimated with. (mean±standard error). Sites 1

Site 2

Site 3

F value

p-value

AC CE P

Abundance 11.8±2.2a 48.0±6.6b 19.5±3.1c 18.701 0.001 Biomass (Kg) 18.1±7.5 33.6±8.8 17.8±7.2 2.106 0.147 Number of fish species 7.4±1.3 10.1±1.6 6.6±2 3.037 0.069 Margalef diversity index 2.50±0.32 2.37±0.35 2.10±0.27 2.802 0.083 Shannon diversity index 1.80±0.15 1.75±0.20 1.51±0.17 3.424 0.052 W-statistics 0.30±0.01a 0.007±0.02b 0.006±0.05b 8.427 0.002 Different superscripts within a row indicate significant differences amongst means (p <0.05).

ACCEPTED MANUSCRIPT Highlights for reviewers Corresponding Author

PT

Messias Alfredo Macuiane

SC

RI

Current address Centro de Estudos Costeiros, Escola Superior de Ciências Marinhas e Costeiras da Universidade Eduardo Mondlane , Av. 1 de Julho, Chuabo Dembe, C.P. 128 Quelimane, Mozambique. Tel: (+258)823056104, Cell:(+258)842264947, Fax: (+258)21020129. [email protected]

MA

NU

Changes in fish community structure associated with cage aquaculture in Lake Malawi/Niassa, Africa

1. Cage aquaculture is becoming popular in Africa, particularly in Lake Malawi where it is the flagship for aquaculture development owing to the dwindling stocks of the capture

D

fisheries which no longer meet the demand for fish.

TE

2. Knowing impact of cage aquaculture on the wild fish community structure, especially in species-rich communities with high endemismsuch as Lake Malawi, is a pre-requisite for

AC CE P

sustainable development of this industry. 3. In Lake Malawi local fishermen traditionally derive income and livelihood from the capture fisheries, andthe expansion of cage aquaculture privatizes fishing grounds and consequently raises conflicts among user groups mainly due to the lack of legal framework for cage aquaculture development. 4. The current study of species composition, abundance and biomass of the fish community near to and away from a cage farm confirmed concerns raised by the local fishermen, and the results will contribute towards a holistic management approach for aquaculture in Lake Malawi.

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Statement of relevance The research promotes sustainable cage culture in Malawi